Isolation of Bluetongue virus from canine abortions

Abstract

Three aborted canine fetuses were submitted to the Animal Health Diagnostic Center at Cornell University in November 2011 and September 2012 for diagnostic workups to determine the causes of the reproductive difficulties. Histological assessments of the sampled tissues were inconclusive due to the autolysis. Tests to detect bacterial causes of the abortions were also negative. Virus isolation testing on pooled tissues from the fetuses identified a cytopathogenic agent in cell cultures. Fluorescent antibody tests on the infected cells gave a positive reaction for Bluetongue virus, and subsequent serotype specific reverse transcription polymerase chain reaction assays identified the isolates as Bluetongue virus serotype 11. The current report describes the isolation of Bluetongue virus from dogs unrelated to contaminated vaccines and in the absence of a raw meat diet.

Bluetongue virus (BTV; family Reoviridae, subfamily Sedoreovirinae, genus Orbivirus) is predominantly a pathogen of ruminants that become infected by way of the feeding activity of various species of the genus Culicoides. Various serotypes of the virus exist in nearly all parts of the world. In the United States, the predominant serotypes are 10, 11, 13, and 17 with sporadic reports of other serotypes mainly in the Southeast. The susceptibility of carnivores to infection by BTV has been documented by serological studies in enzootic areas and by the identification of clinical disease in dogs vaccinated with a BTV-11–contaminated vaccine.^1-4,6,8 Serological data are unable to define the manner in which animals become infected. Early speculation was that BTV-contaminated raw meat might provide the means for initiating the infection in dogs. ² A 2011 report from Morocco identified a significant percentage of dogs serologically positive for BTV by competitive enzyme-linked immunosorbent assay (cELISA) and virus neutralization. ⁶ The study population was fed only canned commercial feed and had no access to raw meat products, suggesting infection by a local Culicoides species. The current report documents the isolation of BTV from 2 separate cases of canine abortion. The abortion events were separated by a 14-month interval representing 2 transmission seasons.

Case 1.A 32-month-old pregnant female Rottweiler was presented to a veterinary clinic in Texas in October 2011, in the early stages of aborting a litter of puppies. She had been artificially inseminated 6 weeks earlier. An ultrasound 5 days prior to the crisis indicated multiple viable fetuses. Prior to arrival at the veterinary clinic, 4 nonviable fetuses were expelled, and ultrasound at the clinic indicated that the remaining fetuses were nonviable. An additional 6 fetuses were expelled, but 1 remained, and a caesarian section was performed to remove the remaining fetus. The serosal surface of the uterus was normal, but the uterine wall was edematous, and the lumen of the uterus contained large amounts of a malodorous black and green discharge. Aerobic and anaerobic cultures of uterine swabs taken at the time of the caesarian section were negative for bacterial growth. The female recovered normally from the surgery but was found dead in her cage the following morning. No postmortem diagnostic testing was performed. She had been current on the core canine vaccines, which were administered solely by the veterinary practice. All of the owner’s dogs had access to an outdoor area that was adjacent to a pasture with cattle.

A fetus from this abortion episode was submitted to the Animal Health Diagnostic Center (AHDC) at Cornell University (Ithaca, New York) for a full necropsy with ancillary testing. Histological examination of the tissues from the fetus and the placenta did not detect any significant lesions, but all samples exhibited moderate to severe autolysis. Bacterial cultures on unfixed lung and placenta did not detect the presence of any organisms related to canine abortions, and fluorescent antibody testing (FAT) for leptospira ^a was negative.

A 10% tissue pool of lung, liver, kidney, and placenta was prepared in Eagle minimal essential medium (MEM-E) ^b plus 0.5% bovine serum albumin ^c and 10 μg/ml ciprofloxacin. ^d After tissue disruption and low speed centrifugation, 1 ml of the unfiltered supernatant was inoculated onto monolayers of A-72 ^e and immortalized canine kidney cells ^f in 25-cm² flasks. Cells were cultured in MEM-E plus 10% gamma-irradiated fetal bovine serum. ^g The initial assessments of the cell cultures revealed some nonspecific degeneration of the cells perhaps due to toxicity of the inoculum. Subsequent passage of freeze–thaw culture medium (extracts) continued to show cell degeneration not seen in the control cultures. Inoculation of the extracts onto African green monkey kidney epithelial (Vero) cells produced a more robust cytopathology caused by an agent that was chloroform resistant. Negative-stain electron microscopy identified an agent in the cell cultures as a “rotavirus,” but FAT of the affected Vero cells demonstrated a positive reaction with a BTV conjugate ^h and a negative reaction for reovirus. ⁱ The identity of the agent was confirmed by serotype-specific reverse transcription polymerase chain reaction (RT-PCR) assays to be BTV-11. ⁵

Case 2. A breeding kennel of Bulldogs in Kansas experienced 2 abortion episodes approximately 5 days apart at 8 weeks gestation in August–September of 2012. Both dogs were bred by artificial insemination using semen from a male that was part of a 20-dog kennel. Two other dogs that were bred using the same semen source had fetal and neonatal issues. One week after the second abortion, a caesarian section delivery resulted in 2 dead puppies and 4 live puppies with 1 subsequent death. Two weeks later, another caesarian delivery produced 1 dead puppy and 2 weak ones. No testing was done related to the caesarian section dogs. The aborting females both survived without significant difficulties. No vaccines had been administered for 2 years prior to the abortions, and no raw meat was fed to any of the dogs.

Two fetuses from the second abortion were submitted to the AHDC for full necropsy. As with case no. 1, there were no significant findings on any tissue samples primarily due to severe autolysis with both fetuses. Aerobic bacterial cultures on 1 of the fetuses grew a number of environmental contaminants, and a culture for Brucella canis was negative. Fluorescent antibody staining for leptospira ^a was also negative. Pooled tissue samples from 1 fetus were processed for virus isolation as described above. Cultures of A72 cells ^e and Vero dog SLAMtag cells ⁷ in 25-cm² flasks were inoculated with 1 ml of a 10% tissue extract. By 6 days on culture, the Vero cells were showing evidence of virus growth with areas of small rounded cells. Cytospin preparations of affected cells tested negative by FAT for Canid herpesvirus 1 (CaHV-1), ⁱ Canine distemper virus, ⁱ Canine adenovirus A, ⁱ and reovirus, ⁱ but were positive for BTV. Serotype-specific RT-PCR assays on the isolate again confirmed the presence of BTV-11.

Sera were collected from the 20 dogs in the affected kennel for CaHV-1 serum neutralization testing. Neither of the aborting females had evidence of a recent CaHV-1 infection (titers < 12). Of the 20 sera, 8 tested positive for antibodies to BTV by cELISA. ^j As there are no validated tests for BTV using canine sera, the sera were also tested by agar gel immunodiffusion ^k (AGID). Of the 8 cELISA positives, 3 were scored as strong positive on AGID while of the remaining 5, 4 were scored as weak positive and 1 negative. The 2 aborting females were the most highly reactive in both tests.

To further examine the utility of the BTV tests using canine sera, 100 sera from a non-BTV enzootic area (New York) and 119 from BTV enzootic areas (Florida, Texas, and Kansas) were tested using the cELISA. Samples from the enzootic areas represented 63 dogs, 16 of which were sampled over multiple years. None of the sera from the nonenzootic area were positive for BTV antibodies, while 5 sera from the enzootic area were positive. An examination of the sources of the positive sera revealed that 3 of the 5 positive samples were from the same dog spanning a period of 2 years. Two additional samples from this same animal were identified with the earliest time point being November 2007. Both additional samples were also positive for BTV antibodies, thus indicating that antibody titers to BTV can persist in a dog for a period of at least 5 years.

The current study describes the isolation of BTV from dogs not vaccinated with a BTV-contaminated canine vaccine. The speculation that infection of carnivores by BTV-contaminated raw meat can also be eliminated as the source of the infection as in neither situation was raw meat fed to the affected dogs. In case no. 2, there was some concern as to whether semen from a stud dog could have transmitted the virus during the artificial insemination procedure. Virus isolation on a semen sample, collected after the abortions, was negative for BTV, ruling out a chronic BTV infection of the stud dog. The 2 cases reported herein were approximately 500 miles apart in distance and were over a year apart in the timing of the abortions. The most logical conclusion as to the source of the virus in these instances is the Culicoides vector of BTV. What is somewhat curious is the identification of BTV-11 in both cases, which is the same serotype as in the contaminated vaccine. Does this particular serotype affect dogs differently than the other enzootic serotypes? Given the apparent rarity of BTV infections of dogs, experimental challenge studies will be needed to resolve this issue.

Reverse transcription PCR amplification and sequencing of the 2 canine isolates resulted in partial sequences of the outer capsid (VP2) genome. The amplicons of the 2 isolates were 99% homologous to each other (2 base mismatches) over the approximately 750-bp sequenced region. The 2 dog isolates also shared a 99% similarity within the same VP2 region to a 2012 BTV-11 isolate from Iowa and three 2011 BTV-11 isolates from Texas (DJ Johnson, unpublished data, 2012). It should be noted that this region of the genome appears to be highly conserved, as BTV-11 isolated from Texas in 1962 (1962 TX BT-11) was 97% homologous to the canine viruses (GenBank accession no. M17437).

While not the intent of the present investigation, it does appear that the cELISA is a valuable tool for screening canine samples for antibodies to BTV. In a somewhat blinded situation, the test detected a dog that was seropositive for at least a 5-year period. The strong positive cELISA sera were also positive by AGID, lending support to the validity of the cELISA results. The seroprevalence rate in samples of convenience from enzootic areas in the current report was low as also noted in a previous study, ⁴ but more extensive studies are needed to define the risk of BTV infections in canines. The identification of 8 infected dogs in a kennel of 20 dogs is noteworthy, but with evidence of long-term persistence of BTV antibodies in dogs, the timing of the infections is unknown.