MOLECULAR DETECTION OF BABESIA ODOCOILEI IN WILD, FARMED, AND ZOO CERVIDS IN ONTARIO, CANADA Open Access

Babesia odocoilei is a tick-borne protozoan parasite that is endemic in wild white-tailed deer (Odocoileus virginianus) in the southeastern US (Waldrup et al. 1990). Ixodes scapularis ticks, the definitive hosts and vectors for B. odocoilei, are widely distributed along the Atlantic seaboard and across the southeastern US (Keirans et al. 1996). Infections with B. odocoilei do not appear to cause severe disease in immunocompetent white-tailed deer (Perry et al. 1985). However, B. odocoilei infection can cause fatal hemolytic anemia in other members of the family Cervidae, namely wapiti (also known as North American elk: Cervus canadensis; previously Cervus elaphus canadensis), Eurasian tundra reindeer (Rangifer tarandus tarandus), and woodland caribou (Rangifer tarandus caribou; Petrini et al. 1995; Holman et al. 2000). Babesia odocoilei has recently emerged as a cause of morbidity and mortality in captive Canadian cervids and potentially threatens vulnerable wild populations of boreal woodland caribou. For example, between 2012 and 2015, five wapiti and three reindeer died due to cervid babesiosis at the Toronto Zoo in southern Ontario, Canada (Mathieu et al. 2018).

The geographic range of I. scapularis is expanding northward within Canada in association with global climate change and anthropogenic alterations in land use (Ogden et al. 2008). Exposure of susceptible cervids to I. scapularis infected with B. odocoilei may underlie the recent emergence of cervid babesiosis in Canada (Mathieu et al. 2018). The current geographic distribution of B. odocoilei in wild and captive Canadian cervids is unknown because surveillance has not yet been performed. Further, the role of noncervid hosts in regional B. odocoilei epidemiology is uncertain. These knowledge gaps result in a poor understanding of the risk that this pathogen poses to captive and free-ranging cervids in Canada.

Because cervid babesiosis is an emerging disease in Canada, improved knowledge of the geographic distribution, potential wildlife reservoirs, and host range of B. odocoilei is urgently needed. The objective of this research study was to survey wild, farmed, and zoo cervids in Ontario for B. odocoilei infection. In addition, surveillance of bovids housed at the Toronto Zoo was performed to assess for B. odocoilei infection in noncervid hosts. Data gathered will aid in understanding the regional epidemiology of this emerging pathogen and help to inform management strategies for at-risk cervids.

Postmortem samples were collected opportunistically from wild, farmed, and zoo cervids in Ontario, Canada. Sources included diagnostic cases submitted to the Canadian Wildlife Health Cooperative (CWHC–Ontario/Nunavut) and the Animal Health Laboratory at the University of Guelph (Ontario, Canada) as well as hunter-harvested animals, those killed on the road or by predators, farmed cervids sent for slaughter at Ontario abattoirs, and deaths at the Toronto Zoo. Additionally, postmortem samples were collected from animals of the family Bovidae that died naturally or were euthanized at the Toronto Zoo during the study.

Samples were collected from cervids and bovids from May 2016 to January 2018. The latitude and longitude of the animal's location at the time of death was recorded when known; otherwise, the nearest street intersection was used. Fresh spleen samples collected within 8 h of death were frozen to –20 C either prior to transport or immediately upon arrival to a research laboratory in the Department of Pathobiology, University of Guelph. Samples were subsequently thawed and a 1-cc piece of spleen was placed into a sterile cryovial and frozen to –80 C until testing.

We performed DNA isolation on spleens from 270 cervids and seven bovids. From each spleen, 50-mg samples were minced with a sterile scalpel blade and lysed by repetitive pipetting in 1 mL of DNAzol^® Reagent (Invitrogen Life Technologies, Carlsbad, California, USA). We extracted DNA from each spleen sample according to the manufacturer's instructions for the tissue protocol. The DNA quality and concentration were determined using a Nanodrop 1000 Spectrophotometer (Thermo Scientific, Wilmington, Delaware, USA) and stored at 4 C.

A piroplasm-specific 18S ribosomal (r)DNA primer (Piro_18S_300F; Table 1) was designed using an alignment of publicly available piroplasm 18S rDNA sequences that also included the 18S rDNA sequence (EU823286.1) of one of the target hosts, white-tailed deer (O. virginianus), so that amplification of host DNA was avoided. Primer design used Primer3 (Koressaar and Remm 2007; Untergasser et al. 2012) executed from within Geneious^® version 6.1 or later (Biomatters Limited, Auckland, New Zealand; Kearse et al. 2012).

Extracted DNA was used as a template in a standard PCR assay designed to amplify 18S rDNA of Babesia spp. (Mathieu et al. 2018). Amplification primers (Table 1) were Piro_18S_300F and Piro_18S_1688R (Mathieu et al. 2018; similar to primer BN1700 of Ramos et al. 2010). PCR-positive samples produced a primary amplicon of 1,290 base pairs (bp). The PCR was performed in a 25-µL reaction containing 3 mM MgCl₂, 0.4 mM dNTPs, 10× PCR buffer, 2U Platinum Taq polymerase (Invitrogen Life Technologies), 1.25 µM of each amplification primer, nuclease-free water, and 100–200 ng of sample DNA. Negative and positive controls were included in each run, with water as the negative control and B. odocoilei DNA extracted from the spleen of a Toronto Zoo reindeer that died from acute babesiosis as the positive control (Mathieu et al. 2018). Amplification was carried out in a T100 thermal cycler (Bio-Rad, Mississauga, Canada) after the following reaction conditions: initial melt of 94 C for 3 min followed by 35 amplification cycles (denature at 94 C for 30 s, anneal at 60 C for 30 s, extend at 72 C for 90 s) and a final extension of 72 C for 5 min. The PCR products were separated by electrophoresis using a 2% agarose gel with 0.5× TAE buffer (120 mL) and 12 µL of SYBR Safe DNA gel stain (Invitrogen Life Technologies) and visualized under ultraviolet light. The GeneRuler 1 kb Plus DNA size ladder (Thermo Scientific) was used to determine product fragment length. An additional reamplification step was performed on samples that produced a band of appropriate size by subsequent agarose gel electrophoresis.

To obtain sufficient product for subsequent Sanger sequencing, the resulting primary amplicons were used as templates in a secondary, heminested PCR reaction using primers Cocci_18S_595F and Piro_18S_1688R (Table 1). This PCR was performed in a 20-µL reaction containing 10 µL HotStar Taq Master Mix (Qiagen, Toronto, Ontario, Canada), 1.25 µM of each amplification primer, and 9 µL of reaction product in a T100 thermal cycler using the above reaction conditions at an annealing temperature of 58 C and an extension time of 60 s.

The resulting primary or nested amplicons were purified using a QIAquick PCR Purification Kit (Qiagen). Direct sequencing of Babesia spp. amplicons was performed by the Genomics Facility Advanced Analysis Centre, University of Guelph, using amplification or sequencing primers (Table 1). For one sample, multiple sequencing chromatograms were assembled into a consensus sequence using the Geneious bioinformatics program. The identities of the other five PCR-positive samples were confirmed by direct sequencing using the Lank_18S_1278R primer (Table 1). Sequences were trimmed to remove primers and then searched against public sequence databases using the BLAST algorithm (Altschul et al. 1990).

From May 2016 to January 2018, 270 wild, farmed, and zoo cervids of six species and one hybrid species (wapiti-red deer) were sampled as well as seven zoo bovids of five species (Table 2). In total, spleens from 2% (6/270) of cervids tested positive by PCR and were confirmed as B. odocoilei by sequencing of the 18S rDNA fragment. There was no clinical suspicion of babesiosis in any of the animals sampled. Cervids originated from across Ontario, but the number of submissions was highest in southern Ontario (Fig. 1).

Wild cervid samples were from 68 white-tailed deer and one moose (Alces alces). This included 39 hunter-harvested white-tailed deer culled for management purposes under advisory from the Ministry of Natural Resources and Forestry and Parks Canada at Point Pelee National Park in January 2017 and January 2018. Sixteen wild white-tailed deer and one moose were submitted to the CWHC for necropsy. Seven white-tailed deer were killed by vehicles in the Greater Toronto Area. Hunters submitted spleens from five white-tailed deer in the 2016 and 2017 hunting seasons. One wild white-tailed deer was killed by coyotes on the Toronto Zoo grounds. Three of these 68 wild white-tailed deer (4%), including two from the CWHC and one hunter-harvested, tested positive for B. odocoilei by PCR (Table 2). The wild moose tested negative for B. odocoilei.

Commercially farmed cervid samples originated from red deer (n=142), wapiti (n=28), wapiti-red deer hybrids (n=13), fallow deer (Dama dama; n=12), and white-tailed deer (n=3) sent for slaughter at Ontario abattoirs. Among these, 1% (2/142) of farmed red deer tested PCR-positive for B. odocoilei. All farmed wapiti, wapiti-red deer hybrids, fallow deer, and farmed white-tailed deer tested PCR-negative for B. odocoilei (Table 2). Zoo cervids sampled included two from the Toronto Zoo; one moose was euthanized due to age-related disease and one wapiti was euthanized following a traumatic injury. One caribou from a private zoo was submitted for necropsy to the Animal Health Laboratory. Zoo bovids sampled included one Barbary sheep (Ammotragus lervia), one chamois (Rupicapra rupicapra), one west Caucasian tur (Capra caucasica), three wood bison (Bison bison athabascae), and one yak (Bos grunniens) that died or were euthanized at the Toronto Zoo. Of the zoo cervids and bovids, only wapiti tested positive for B. odocoilei by PCR. Overall, 3% (1/29) of captive (zoo and farmed) wapiti were B. odocoilei positive (Table 2).

The partial 18S rDNA sequence (1,290 bp) from B. odocoilei infecting a white-tailed deer sampled in this study was submitted to GenBank (accession no. MH366302). This partial 18S rDNA sequence had 100% identity over its entire length to numerous B. odocoilei sequences in GenBank from various hosts and geographic locations. These included wild white-tailed deer in Texas, US (GenBank no. U16369.2), farmed wapiti in Saskatchewan (GenBank no. KC460321), and the recent Toronto Zoo clinical babesiosis cases (Mathieu et al. 2018) in wapiti (no. MF357056) and reindeer (no. MF357057) in Canada. The other five PCR-positive samples were direct-sequenced using primer Lank_18S_1278R, generating 571-bp reads that spanned the highly variable region (bp 609–645 of the Babesia odocoilei reference sequence U16369.2) that discriminates this species from other closely related Babesia species; all samples had 100% sequence identity with the reference sequence over this 571-bp region.

We documented B. odocoilei infection in wild white-tailed deer in Ontario, supporting the hypothesis that this species could serve as a natural reservoir of infection in this geographic region as it does in the southeastern US (Waldrup et al. 1989). An infection prevalence of 4% was found in wild white-tailed deer in Ontario; although no comparable PCR-based surveillance for B. odocoilei has been performed in wild cervids in the US, examination of blood smears from wild white-tailed deer in Texas and Oklahoma found a prevalence of B. odocoilei parasitemia of 2% (Waldrup et al. 1989). Asymptomatic B. odocoilei infection was also identified in two species in the genus Cervus in Ontario, with an infection prevalence of 1% in red deer and 3% in wapiti. Babesia odocoilei-infected wapiti are known to be susceptible to acute hemolytic crisis, which may be triggered by stress or immunosuppression (Gallatin et al. 2003). The pathogenicity of B. odocoilei in red deer is unknown. Red deer, a European species, are farmed for venison in Ontario, and wapiti are present in zoos, on farms, and as reintroduced wild populations in Ontario. A B. odocoilei-like parasite was detected in blood samples from three wild red deer in Ireland, although the species of this parasite was not confirmed (Zintl et al. 2011). It is possible that Cervus spp. may serve as additional reservoirs of B. odocoilei. Samples were obtained only from captive Ontario wapiti, and thus we make no comment regarding subclinical infections or disease in the wild wapiti population.

All sequenced samples in our study were 100% identical to each other and to B. odocoilei isolates from several different host species and geographic locations, including the partial 18S rDNA sequence isolated from a fatal case of babesiosis in a Toronto Zoo reindeer in 2012 (Mathieu et al. 2018). Comparison of nuclear 18S rDNA sequences is widely accepted as a method for resolving questions of relatedness and identity of Babesia spp. isolates (Holman et al. 2000). As the 18S rDNA fragment of B. odocoilei appears to be well-conserved among cervid species in Ontario, further molecular characterization of the isolates using primers specific for a mitochondrial target, such as the cytochrome c oxidase subunit I, would be desirable to further elucidate the epidemiology of this parasite.

Prior to the first cases in 2012 of hemolytic anemia due to babesiosis in Saskatchewan farmed wapiti (Pattullo et al. 2013) and in reindeer and wapiti at the Toronto Zoo (Mathieu et al. 2018), B. odocoilei was not recognized as a clinical problem in Canadian cervids, and surveillance for this pathogen was not conducted in captive or free-ranging wildlife. While it is possible that the parasite was present in Canada prior to 2012, a lack of reports of disease in captive or wild cervids makes that unlikely. Based on limited home ranges of white-tailed deer in many areas (Nixon et al. 1991), as well as the rapid emergence of clinical cases of cervid babesiosis in several widely disparate locations across North America (Mathieu et al. 2018), the natural movements of free-ranging white-tailed deer are unlikely to have been a primary driver in the emergence of B. odocoilei in Canada. The underlying cause(s) of the likely incursion of B. odocoilei into Canada remains unknown, but possibilities include translocation of an asymptomatic infected cervid or range expansion of B. odocoilei-infected ticks, perhaps through adventitious carriage on migratory birds (Ogden et al. 2008).

The locations of historic Canadian cervid babesiosis cases and of the B. odocoilei PCR-positive cervids in our study coincide with the distribution of known I. scapularis populations, with the exception of the wapiti cases in Saskatchewan where I. scapularis is not known to be established (Pattullo et al. 2013). Dermacentor spp. ticks are present in Saskatchewan, but the vector competence of this tick species for B. odocoilei has not been investigated. Ixodes scapularis serves as both the definitive host and vector for B. odocoilei (Waldrup et al. 1990). Following gametogony in the tick gut, B. odocoilei gametes fuse and form a zygote that develops into a kinete, which subsequently invades and replicates in the tick hemolymph. Sporogony follows invasion of the kinetes into the tick salivary glands. Susceptible cervids are exposed to B. odocoilei by the bite of an infected tick that introduces infective stages into the cervid host's blood stream (Homer et al. 2000). These stages invade the cervid erythrocytes and undergo merogony, primarily through binary fission. In a parasitemic cervid, B. odocoilei merogonic stages appear on a blood smear as single, paired, or tetrad pyriform and ring-shaped organisms, usually in the accolé position at the periphery of the erythrocyte. The infected erythrocytes eventually rupture, liberating merozoites that invade new erythrocytes and either become trophozoites that undergo merogony again by binary fission or alternatively develop into nondividing gamonts that are infective to a tick when it feeds on the cervid intermediate host. Maintenance of B. odocoilei within an ecosystem is therefore dependent both on I. scapularis as the definitive host and a mammalian intermediate host.

We did not detect B. odocoilei infection in any bovids at the Toronto Zoo, but the significance of this finding is limited by small sample sizes and limited range of species. Subclinical B. odocoilei infection has been identified in endemic and exotic ruminants belonging to the family Bovidae, including zoo markhor (Capra falconeri), zoo yak (Bartlett et al. 2009), and wild bighorn sheep (Ovis canadensis nelsoni) in the US (Schoelkopf et al. 2005).

In our study, B. odocoilei was identified in both captive and wild cervids. Transmission of disease between the two groups is well recognized, with the management of cervids in captivity on private farms or in zoologic institutions posing a recognized health risk to free-ranging wild cervids and vice versa (Gerhold and Hickling 2016). Captive cervid facilities provide ample opportunities for transmission of pathogens and ectoparasites between captive and wild cervids including direct contact through fence lines, escape of captive animals, and ingress of wild cervids into inadequately fenced enclosures. In the absence of reliable premovement disease testing, the translocation of captive cervids across provincial and international borders necessarily results in the simultaneous translocation of their pathogens, including ticks. Babesia odocoilei requires a tick host to complete its life cycle and, therefore, direct transmission of babesiosis between cervids does not occur, with the possible exception of iatrogenic transmission such as via blood transfusion (Pastor and Milnes 2018).

The lack of reliable commercially available diagnostic tests for screening and premovement testing of cervids for hemoparasites such as babesiosis is a major problem for the captive cervid industry, and the development of a commercial PCR assay for blood testing of carrier animals would help to alleviate this problem (Holman et al. 2000). Molecular testing of postmortem spleen samples for B. odocoilei DNA through PCR, as used in this study, is a highly sensitive and specific method for detecting latently infected animals with a low-level parasitemia that may not be evident on microscopic examination of blood smears (Holman et al. 2000). This method is also suitable for use on whole blood samples and has been used successfully to identify sub-clinically infected cervids in a zoo setting (Bos et al. 2017).

In conclusion, the results of this study suggest that B. odocoilei has an established wildlife reservoir in wild white-tailed deer in Ontario, Canada. Given the expanding range and increasing density of the arthropod host and vector, I. scapularis, continued expansion of transmission and thus infection among the while-tailed deer population is likely. The finding of B. odocoilei in apparently healthy captive cervids emphasizes the importance of proactive tick prevention and biosecurity measures (e.g., double fencing) in cervid facilities to prevent disease and vector transmission between captive and wild animals as well as losses due to clinical disease in more-susceptible species. In addition, viable options for prophylaxis and treatment of clinical disease in captive cervids should be explored. Cervid managers should be aware of the risks involved in translocating animals between herds.

We are grateful to Alexandra Reid, the hunters of Caldwell First Nation, Don Thornton, Erin Harkness, Kent Charlton, Marg Stalker, Rob Kirkpatrick, Samantha Allen, Simon Hollamby, Stephanie Sparling and the staff of Toronto Animal Services, and Tammy Dobbie and the staff of Parks Canada at Point Pelee National Park for the submission of deer samples. Hannah Bagnall and Janessa Price assisted with sample processing, and Tami Sauder, Dorothee Bienzle, and Mary Ellen Clark provided laboratory support. This study was supported by the British Veterinary Zoological Society Zebra Foundation; the Toronto Zoological Foundation; the Natural Sciences and Engineering Research Council of Canada; the Wilson Ornithological Society; and the Canadian Foundation for Innovation.