Corresponding author: J. Walter, Klinik für Reproduktionsmedizin, Vetsuisse-Fakultät Universität Zürich, Winterthurerstrasse 260, 8057 Zürich, Switzerland; email: firstname.lastname@example.org.
Equid herpesvirus 1 (EHV-1) is a highly prevalent pathogen in horse populations worldwide. Oronasal infection represents the classic route of disease transmission. Venereal shedding of EHV-1 is not regarded relevant in terms of virus spreading, which is in contrast to the close relatives of EHV-1, bovine and suid alphaherpesvirus, for which artificial insemination is a well-documented and accepted means of virus spread.
Documentation of venereal EHV-1 shedding in 3 naturally infected stallions.
Three stallions were infected during an acute outbreak by an EHV-1 strain with the G2254/D752 Pol genotype.
In this observational study, 12 semen samples from these 3 stallions were tested for EHV-1 to determine venereal shedding. EHV-1 was diagnosed by conventional PCR and paired serum neutralization tests in 42 horses. Semen samples were separated into sperm and seminal plasma fractions and tested for EHV-1 by conventional and quantitative PCR as well as virus isolation by cell culture.
Acute EHV-1 infection was diagnosed on the premise. Five semen samples collected from 2 of the 3 stallions tested positive for EHV-1 by (q)PCR. On days 18 and 20 after onset of fever, the last positive samples were retrieved. All samples were positive in seminal plasma, only three in sperm fraction. Virus isolation attempts were unsuccessful.
Conclusions and Clinical Importance
The data presented here document shedding of EHV-1 in semen of naturally infected stallions for close to 3 weeks, which seems not to be directly associated with spermatozoa.
Equid herpesvirus type 1 represents, with a prevalence of up to 88%, a common pathogen in horse populations worldwide and causes abortion, perinatal mortality, respiratory disease, and neurological disorders. Horizontal transmission of EHV-1 by aerosol and fomites is a documented mode of virus transmission. Aborted fetuses and fetal membranes also constitute an important source of infectious virus. The entry port for EHV-1 is the respiratory tract, usually early in life. Horses establish latent infections upon initial infection, which can be reactivated during periods of stress to result in clinical disease and viral shedding. Semen is a known risk for spreading of bovine herpesvirus 1 (BoHV-1) in cattle, but the role of venereal transmission in horses is not elucidated until now. In a recent study on 390 healthy stallions, 13% tested positive for EHV-1 in semen by PCR targeting gB, with significant more shedders in the unvaccinated group (25%) compared with the vaccinated group (15%). In contrast, a similar study in the United States detected no EHV-1 in semen of 50 healthy stallions. Experimental intranasal infection of pony stallions (n = 3) resulted in 1 shedder until day 25 after virus exposure. The EHV-1-positive semen samples contained neutrophils, plasma cells, and lymphocytes in sperm-rich fractions.
The aim of this investigation was to evaluate whether viral shedding of EHV-1 occurs as consequence of acute EHV-1 infection in stallions under natural outbreak conditions.
Material and Methods
Animals and Premises
The stud housed a total of 149 horses, of which 79 were in virus-exposed areas and easily accessible for thorough examination and documentation. The premise is divided into 11 separated areas for sport and visiting horses, broodmares, breeding stallions, and adolescent horses (1–3 years of age). With the exception of the breeding stallions, the same staff attended all horses and the same equipment was used, such as a tractor for hay transport and oat trolleys. The clinical outbreak occurred in 7 of the 11 different areas of the farm. Active breeding stallions were not affected and it was logistically impossible to monitor adolescent horses continually. In the remaining barns, a total of 79 horses were housed. Ten mares (7 in foal, 3 barren) and 1 foal were boarded in the broodmare section, 7 stallions, 26 geldings, 34 mares, and 1 foal in the remaining barns. A neighboring stud in 50 m distance to the broodmare barn remained uninvolved in the outbreak. The vaccination status varied between the different barns. Only 21 (32%) of the horses in the sport and visiting horse section were vaccinated in 6-monthly intervals with an EHV-1/-4 inactivated vaccine1 as recommended by the supplier. All 10 broodmares were current on their immunization schedule with vaccinations in the 5th, 7th, and 9th month of gestation. Ten active breeding stallions, housed in a separate building, were also properly vaccinated.1 The 60 adolescent horses (20 per age-group 1-, 2- and 3-year-olds) in the surrounding areas of the stud in a distance of 500–1000 m were not vaccinated. The clinical course of the outbreak was thoroughly documented with strict definitions for fever (body temperature of ≥38.1°C [≥100.6°F]) and EHM (equid herpesvirus-associated myeloencephalopathy, graded according to Reed and Andrews).
EHV-1 shedding in semen was tested in 3 stallions (4 years of age) that presented with fever for periods of 5 (n = 2) and 3 days (n = 1) during the outbreak. The stallions were housed separate from the active breeding stallions and were the only affected stallions in which semen was collected. All stallions were current on their EHV-1/-4 vaccination schedule with an inactivated vaccine,1 and the last booster vaccination was performed less than 6 months before the outbreak.
Collection of Samples and Laboratory Tests
Detection of EHV-1 from aborted fetuses
All aborted fetuses were examined by gross pathology and histology. Tissue samples of fetal lung and liver as well as fetal membranes were examined by cell culture isolation and ORF30 (polymerase, Pol)-specific PCR. To determine the Pol genotype, the amplification product was digested with restriction enzyme SalI and sequenced.
Detection of EHV-1 antibody titers and EHV-1 in nasal secretions and PBMC
Serum neutralization tests (SNT) in 42 horses were performed on day 12 after the initial abortion (considered the index case) and testing was repeated on day 28. Samples were taken on the same day for all horses, regardless of the time point when clinical symptoms first appeared. On day 12, nasal swabs, sera, and peripheral blood mononuclear cells (PBMC) from all 42 horses were tested by PCR (see below). After confirmation of EHV-1 as the causative agent, all horses showing elevated body temperature, EHM symptoms, or both during the acute phase of the outbreak were considered EHV-1-infected.
Detection of EHV-1 from ejaculates
During and up to 1 month after the acute febrile phase, a total of 10 semen samples of the 3 affected young nonbreeding stallions were examined for EHV-1 by conventional PCR targeting gC. Semen from stallion 1 was tested on days 7, 20, and 24; from stallion 2 on days 4, 10, 18, and 31; and from stallion 3 on days 1, 10, and 23 after onset of fever. In stallions 1 and 2, semen samples were also examined 1 year after the outbreak. Semen was collected using a phantom in a separate area not connected to the affected barns. For each collection, the artificial vagina was covered with a new inner sheath to exclude contamination. For nucleic acid extraction, 200–500 μL of a semen sample was centrifuged for 10 minute at 1,485 × g to separate the seminal fluid from the cellular fraction. DNA isolation and purification were performed using the QIAamp DNA Blood Mini Kit,2 according to the manufacturer's instructions. To increase the analytical specificity of conventional PCR, restriction fragment length polymorphism (RFLP) analysis was performed on positive PCR amplicons: PCR products were digested with DdeI3 and electrophoresed in 2% agarose gels. In each PCR, a known amount of plasmid DNA near the detection limit of the PCR (10 copies), derived from positive EHV-4 clinical samples, was included as a positive control. A no template control (NTC) as well as water treated like patient samples were included as negative and extraction controls.
Two different qPCR assays targeting the gB and gC genes were used to determine viral DNA copy numbers in sperm. To relate qPCR signals in sperm to viral DNA copy numbers and infectious virus, respectively, aliquots of the sperm samples were spiked with 10-fold dilutions of EHV-1 ranging from 100,000 to 1 plaque-forming units. Viral DNA copies in untreated sperm were then calculated based on the resulting calibration curve. Virus isolation was attempted using rabbit kidney RK13 and equine fibroblast cells (NBL-6) exactly as previously described. Virus isolation was considered negative if no cytopathic effect was observed after 3 blind passages that were performed in 3-day intervals.
After initial confirmation of EHV-1 infection on the premise by PCR and virus isolation, 61 (77%) of the 79 stabled horses were clinically affected as defined by presentation of at least 1 of 3 symptoms: fever, abortion, or EHM. The index case was an abortion (Fig 1), and the 61 clinical cases occurred over a time period of 44 days after the index case. The EHV-1 typical clinical course of the outbreak with fever (n = 55), abortions (n = 6), and neurological deficits (n = 9) was confirmed by laboratory results.
Detection of EHV-1 from Aborted Fetuses
Gross pathological and histological alterations of the 6 aborted fetuses revealed typical lesions in spleen and liver with necrosis and intranuclear inclusion bodies. Cell culture isolation of the virus was successful from 4 aborted fetuses, while EHV-1-specific PCR amplification was positive in all fetuses.
Detection of EHV-1 Antibody Titers and Virus Shedding
A ≥4-fold increase of SNT (n = 42) was detected in 12 (29%) of clinically affected horses, whereas 19 (45%) horses showed no change in titers, and 11 (26%) a slight decrease. The overall change in serum neutralizing titers is presented in Figure 2. Nasal swabs, sera, and PCR from PBMC yielded EHV-1-positive results in 11, 6, and 9 cases, respectively. In 3 of the 42 horses sampled, all 3 materials tested positive by PCR.
Detection of EHV-1 from Ejaculates
Semen samples of 2 stallions tested positive for EHV-1 by PCR after the onset of fever; EHV-4 was not detectable in any of the samples as determined by DdeI digestion of the amplification products that resulted in fragments of 197 and 49 bp in the case of EHV-1, while the EHV-4 PCR product remained undigested (Fig 3). The 1st stallion shed virus in semen on days 7 and 20 after 1st detection of fever, the 2nd stallion in collections from days 4, 10, and 18. On days 24 (stallion 1) and 31 (stallion 2), virus shedding in semen was no longer detectable. In the remaining stallion, all semen samples (days 1, 10, and 23) tested negative for EHV-1 (Fig 4). One year after the outbreak, semen samples of the 2 shedding stallions 1 and 2 were still negative by PCR. In all 5 samples containing EHV-1 virus, detection was successful in seminal plasma, while the sperm fraction yielded positive results only in 3 samples (1st collection in stallion 1; 2nd and 3rd collection in stallion 2). While EHV-1 isolation attempts from semen failed despite continuous blind passaging, quantitative PCR analyses were successful. Two independent qPCR protocols confirmed presence of EHV-1 at low levels totaling approximately 50–500 viral genome equivalents or 5–50 infectious units per mL of sperm as calculated based on sperm samples that had been spiked with infectious EHV-1 and were used as controls.
This report documents venereal shedding of EHV-1 in 3 stallions, which were infected during a large EHV-1 outbreak caused by a G2254/D752 strain. The classical mode of EHV-1 transmission is horizontal by aerosol and fomites, venereal shedding is not considered in the epidemiology of this disease.
Venereal shedding of BoHV-1 is a known risk for disease transmission in cattle, but has not been investigated with respect to EHV-1. Only limited data are available on occasional virus shedding in the semen of EHV-1-infected stallions.[3-5] To our knowledge, this is the 1st report on EHV-1 shedding in semen during a clinical outbreak on a stud. The 3 stallions examined in our study were current on their EHV-1 vaccination schedule with an inactivated vaccine,1 but nonetheless were infected during the outbreak. A significant reduction in the risk for EHV-1 shedding by vaccination was observed in latently infected stallions. However, concluding from the observations in this natural outbreak, vaccination was not capable of preventing shedding during acute infection in 2 of the 3 stallions tested. As the stallions were not used for breeding at the time of infection, venereal transmission to mares could be excluded.
The mechanism of EHV-1 shedding in sperm will require further investigation. EHV-1 causes a cell-associated viremia and has a well-documented endothelial cell tropism. Viral antigen was detectable in endothelial cells of testes and epididymis by virus isolation, immunoperoxidase staining, and PCR on days 8 and 9 postinfection. By histopathology, necrotizing vasculitis and thrombosis at endothelial sites in testes and epididymis were observed. With regard to pathogenesis, the authors suggested a leakage of free or cell-associated virus through the blood-semen barrier, but our results seem to support the notion of the presence of free, non–cell-associated virus, which might stem from the accessory glands. We base our interpretation on the fact that virus was detected in seminal plasma, but not in the cellular fraction in 2 cases, which may indicate that virus is “added” in a cell-free form to ejaculates after replication in glandular epithelia. The sperm fraction and seminal plasma were only separated by centrifugation and not by density-gradient centrifugation. As a consequence, the positive results in the sperm fraction in 3 samples can also be a result of contamination by seminal plasma. Alternatively, we cannot formally exclude that spermatozoa were ruptured during processing and virus could have been released and contaminated seminal plasma. The fact that we identified fewer positive sperm-rich fractions than seminal plasma fractions seems to argue against this alternative interpretation, however. A survey where EHV-1 was not detectable in washed sperm cells of 50 stallions supports the hypothesis of cell-free EHV-1 shedding. In contrast, 15% (n = 390) of whole fresh and frozen semen samples of healthy stallions were EHV-1 positive. Certainly, other factors including different horse populations and viral strains could also explain the discrepancy between the conflicting results of these 2 investigations.
There are no data available on the infectious dose that may be needed to cause infection in the mare by natural breeding or artificial insemination. Some data are available for transmission of BoHV-1 by artificial insemination. BoHV-1 is a close relative of EHV-1 and doses higher than 105.3 TCID50 invariably resulted in infection, while doses below 200 TCID50 infected only 6 of 25 cows. We report here that the amount of virus in stallion semen was only 50–500 viral genome equivalents. One must consider, however, that these qPCR assays were performed 2 years after sperm isolation on frozen samples. In cattle, levels of BoHV-1 as low as 32 TCID50 in semen can result in venereal transmission, a virus load comparable to that detected in equine semen. BoHV-1 remains infective during storage in liquid nitrogen, which could be the same for EHV-1 and is certainly of interest to the horse industry. Unfortunately, we were unable to isolate virus from semen samples, but again on the basis of studies with BoHV-1, one cannot assume that the negative tissue culture isolation indicates that there is no infectious virus in the ejaculate.[2, 12] For detection of BoHV-1 in semen, qPCR protocols showed a superior sensitivity over cell culture isolation techniques. We are aware that positive (q)PCR results do not necessarily correspond with infectious virus. The OIE (World Organization for Animal Health) gives detailed recommendations in the Terrestrial Animal Health Code for testing donor bulls for BoHV-1 prior to semen collection. Bulls need to be serologically negative or––in case of seropositivity––an aliquot of semen from each collection is to be subjected to virus isolation. EHV-1 is not considered an infectious agent with regard to semen by the OIE, but due to the high percentage of latently infected horses, the necessity of testing in stallions that serve as semen donors should be considered.
Consequences and Conclusions
The presence of EHV-1 was confirmed in semen of naturally infected nonbreeding stallions for up to 20 days, which may suggest the possibility of venereal transmission of virus with unknown consequences. Our findings warrant further research to clarify if venereal EHV-1 transmission is a serious risk and if it could be necessary to implement a screening of AI stallions for EHV-1 in semen.
The clinical work was performed at Birkhof Stud, Donzdorf, Germany. Diagnostic procedures were performed at Chemisches und Veterinäruntersuchungsamt (CVUA) Fellbach, Germany; CVUA Freiburg, Germany; Institut für Virologie, Freie Universität Berlin, Germany; Vet Med Labor GmbH, Ludwigsburg, Germany.
Conflict of interest: Authors disclose no conflict of interest.
Duvaxyn, Fort Dodge, Animal Health, Würselen, Germany