INTRODUCTION
Sarcocystis species have an obligatory 2-host life cycle with carnivores as definitive hosts and mostly omnivores or herbivores as intermediate hosts (Dubey et al. Reference Dubey, Speer and Fayer1989). A single herbivore may serve as the intermediate host for several Sarcocystis species, often concurrently, and this is very common in cervids. Thus, 5 Sarcocystis species have been described from moose (Alces alces) in Norway; 3 species (Sarcocystis alces, Sarcocystis ovalis, and Sarcocystis scandinavica), on the basis of both cyst morphology and molecular data, and 2 species (Sarcocystis sp. Type E and Type D) only on the basis of unique DNA sequences (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008). However, a species with the same ssu rRNA gene sequence as the unnamed Sarcocystis sp. Type E in moose (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008) was later described by both cyst morphology and molecular data from red deer and named Sarcocystis hjorti (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2010). It then also became obvious that sarcocysts of S. hjorti, as seen in red deer, were morphologically indistinguishable from those of Sarcocystis alceslatrans, which had been obtained from 2 Canadian moose and examined by light microscopy by Dahlgren and Gjerde (Reference Dahlgren and Gjerde2008). Molecular analysis did, however, prove that S. alceslatrans was different from Sarcocystis sp. Type E. S. alceslatrans had originally been reported from moose in USA (Dubey, 1980) and Canada (Colwell and Mahrt, Reference Colwell and Mahrt1983), and had been found to use coyotes and dogs as definitive hosts (Dubey, Reference Dubey1980; Colwell and Mahrt, Reference Colwell and Mahrt1983). By contrast, the life cycle of all Sarcocystis species identified in Norwegian moose was unknown.
S. alces was found to be the most prevalent Sarcocystis species in Norwegian moose (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008). In addition, several moose in that study were heavily infected, which is likely to have had a negative impact on the health of the animals. The high prevalence and infection intensity of S. alces suggested that the definitive hosts were common in Norwegian forests. Moreover, the phylogenetic position of S. alces indicated that canids were its most likely definitive hosts (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008). Since red foxes (Vulpes vulpes) commonly feed on carcasses of moose that have died from accidents or disease, we were particularly interested in using this species in the transmission experiments, but we also included the arctic fox (Vulpes lagopus) in the study. Thus, the aim of this study was to determine whether the red fox and the arctic fox would act as definitive hosts for S. alces, through the identification of any sporocysts/oocysts in the intestine or faeces of the inoculated foxes by molecular methods, using Sarcocystis sequence data obtained in previous studies for this purpose (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2007, Reference Dahlgren and Gjerde2008, Reference Dahlgren and Gjerde2009, Reference Dahlgren and Gjerde2010).
MATERIALS AND METHODS
Experimental animals
A total of 12 foxes were used in 2 separate experiments; 8 foxes in Experiment 1 (October 2007), and 4 foxes in Exp. 2 (October 2008). An equal number of silver foxes (colour mutants of the red fox (Vulpes vulpes)) and blue foxes (colour mutants of the arctic fox (Vulpes lagopus)), were used in each experiment. In Exp. 1, all foxes were females, 1–5 years old, whereas in Exp. 2, 2 female silver foxes, 1–2 years old, and 2 male blue foxes, about 6 months old, were used. All foxes had been born and reared at the Unit for Fur-bearing Animals at the Animal Production Experimental Centre at the Norwegian University of Life Sciences, Ås, Norway. The foxes were caged individually and fed their ordinary food throughout the experimental period, except on day 0, when they were only fed fresh muscle tissue. The ordinary food consisted mainly of heat-treated and ground offal, mostly from broilers. The foxes had never eaten fresh muscle tissues before the experiments. The foxes were killed with electrocution according to standard euthanasia methods. Both experiments had been approved by the Norwegian Animal Research Authority (NARA).
Collection and preparation of infective material
Experiment 1
Fresh portions from the oesophagus, diaphragm, and abdominal muscles were obtained from 5 moose that had been killed in Nordmarka, near Oslo, during the regular hunt in October 2007. The carcasses were processed at an abattoir-like plant and had been kept there at around 4°C, for a maximum of 7 days before sampling. Muscle tissues from moose were screened for sarcocysts under a stereomicroscope. A presumptive species identification of several cysts of each type, as seen in situ, was made by excising the cysts and examining their surface morphology in wet mounts under a light microscope. A few sarcocysts of each cyst type were also excised and their species identity subsequently determined by molecular methods as described previously (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2007). Samples from all animals contained cysts (from a few to numerous) consistent with those of S. alces; samples from 3 animals contained a few cysts of S. ovalis and, in 1 animal, a single cyst with hair-like protrusions was found (Fig. 1). The muscle tissues from all moose were cut into small pieces, thoroughly mixed and divided into 8 portions of about 200 g each, which were fed to each of 8 foxes the next day in a single feeding.
Sarcocystis hjorti in moose. (A) Nearly complete cyst, which was subsequently determined by sequence analysis to belong to S. hjorti. (B) Detail from surface of cyst in (A), showing delicate hair-like protrusions. Phase contrast. (C) Another small cyst with hair-like protrusions consistent with S. hjorti. Scale bars=150 μm in (A); 10 μm in (B) and (C).
Experiment 2
Fresh portions from the oesophagus, diaphragm, and abdominal muscles were obtained from 7 moose that had been killed during the regular hunt in Nordmarka, near Oslo, in October 2008. Samples were collected at the same processing plant as in 2007, from 1 moose that had been killed just a few hours earlier, from 2 carcasses that had been kept at about 4°C for a maximum of 4 days, and from offal from 3–4 carcasses, which had been kept in a container outside the plant for 3–4 days at about 10°C. The muscle samples were examined as in Exp. 1. In the most recently killed moose, numerous cysts of S. alces and a few cysts of S. ovalis were seen. In the other samples, only a few cysts of these two species were observed. The muscle tissues were cut into small pieces, then thoroughly mixed and split into 4 portions of about 700 g each, which were fed to each of 4 foxes the next day in a single feeding.
Animal monitoring and blood sampling
The animals were monitored every day for clinical symptoms, faecal consistency and appetite. Blood samples (serum and full blood) were taken from Vena cephalica from all experimental animals, starting at day 0 (control samples), and collected with 7 days interval until euthanasia. The blood samples were analysed at the Central Laboratory at the Norwegian School of Veterinary Science. Serum samples were examined for total protein, albumin, globulin, creatinine, bile acids, total bilirubin, cholesterol, glucose, urea, enzymes (aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, creatinine kinase, amylase, lipase), minerals (inorganic phosphate, calcium, sodium, potassium, chloride), and hormones (cortisol). Full-blood samples were examined for concentration of erythrocytes and different types of leucocytes (lymphocytes, neutrophils, monocytes, eosinophils, basophils). Blood parameters of an uninfected fox sampled concurrently with the experimental foxes, control samples of all experimental animals on day 0, and reference values from cats and dogs were used as estimates of normal range values for the foxes.
Faecal examination
In 2007, faecal sampling started on day 6 after feeding and faeces were collected every day until euthanasia. In 2008, due to a misunderstanding by the staff at the research facility, daily samples were only collected on days 6 and 7, whereas faeces from days 8–15 were combined into 1 pooled sample per animal. The faecal samples were stored at 4°C for 1–3 months and then examined under a light microscope after flotation of the samples in saturated sucrose. Small portions (about 0·5 ml) of each faecal sample were transferred to 1·5 ml Eppendorf tubes and stored at −20°C for subsequent molecular analysis.
Intestinal examination
Two randomly selected animals, 1 silver fox and 1 blue fox, were killed on days 7, 14, 21, and 28 post-inoculation (p.i.) in Exp. 1, and on days 8 and 15 p.i. in Exp. 2. The entire intestine was removed, cut open lengthwise, and the contents and mucosal lining examined for parasites along the entire length of the intestine (about 180–200 cm). Samples from the intestinal mucosa were taken at approximately 10 cm intervals, by scraping with a scalpel blade. The scrapings were smeared onto a microscope slide and examined for Sarcocystis oocysts/sporocysts under a light microscope. Pictures were taken with a Leica MPS60 photomicrography system (Exp. 1) or a Leica DC480 digital camera (Exp. 2).
In Exp. 2, a few oocyst-positive scrapings were directly transferred from the microscope slide to 1·5 ml Eppendorf tubes containing about 0·5 ml of distilled water and stored at −20°C until DNA isolation and molecular examination. The remaining intestinal mucosa from the foxes in Exp. 2, and all mucosal samples in Exp. 1, were treated as follows: areas where oocysts had been observed microscopically were scraped with a scalpel blade and the mucosal lining was transferred to 50 ml plastic tubes containing either 70% alcohol or distilled water. The samples were stored at −20°C until molecular examination.
DNA isolation
Two samples from each fox were examined by molecular methods. DNA was isolated from faecal and/or mucosal samples containing oocysts/sporocysts.
Intestinal mucosa containing Sarcocystis oocysts, which had been stored frozen in distilled water, was thawed and portions of about 0·7 ml of this mixture were transferred from the 50 ml tubes to 1·5 or 2·0 ml Eppendorf tubes. These tubes were centrifuged at 12 000 g for 1 min, and the supernatant was discarded. A few samples that had been stored in 70% alcohol were transferred to 1·5 ml Eppendorf tubes, and left open at room temperature until most of the fluid had evaporated. To evaluate whether temperature treatment would improve DNA recovery, a few samples were incubated in TE-buffer at 100°C for 60 min. Genomic DNA from the samples was isolated either by using QIAamp® DNA Mini Kit (Qiagen GmbH, Germany) according to the manufacturer's tissue protocol, or by using QIAamp® Stool DNA Mini Kit (Qiagen GmbH, Germany) according to the protocol described for the extraction of DNA from Tritrichomonas foetus (Gookin et al. Reference Gookin, Birkenheuer, Breitschwerdt and Levy2002). The ‘Stool kit’ and associated protocol were also used for DNA isolation from faecal samples that had been stored frozen in Eppendorf tubes. The samples containing intestinal scrapings that had been transferred directly to, and stored at −20°C in 1·5 ml Eppendorf tubes, were thawed, centrifuged for 60 s at 12 000 g, and the supernatant decanted. DNA was then isolated using QIAamp® DNA Mini Kit, according to the manufacturer's tissue protocol.
DNA amplification and sequencing
The ssu rRNA gene was selected as a target molecule to test all DNA samples for the presence of coccidian DNA by PCR using forward primer ERIB1 (Barta et al. Reference Barta, Martin, Liberator, Dashkevicz, Anderson, Feighner, Elbrecht, Perkins-Barrow, Jenkins, Danforth, Ruff and Profous-Juchelka1997) and reverse primer S4r (Fischer and Odening, Reference Fischer and Odening1998). PCR products were separated on 1% agarose gels and visualized under UV light after staining with ethidium bromide to check for appropriately sized products. To verify the results, positive PCR products of 1 DNA sample from each fox were sent for sequencing to Eurofins MWG Operon, Germany. Additional primers used in the sequencing reactions were S3f, S5f (Fischer and Odening, Reference Fischer and Odening1998) and Primer B (Fenger et al. Reference Fenger, Granstrom, Langemeier, Stamper, Donahue, Patterson, Gajadhar, Marteniuk, Xiaomin and Dubey1995). PCR protocols and DNA purification have been described previously (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2007). Vector NTI Advance 11 software (Invitrogen, Scotland) was used to assemble sequence information from forward and reverse primers into contiguous sequences.
All DNA samples were also tested for the presence of Hammondia DNA, using the Hammondia specific primers AT264/AT9, which target the alpha tubulin gene (Abel et al. Reference Abel, Schares, Orzeszko, Gasser and Ellis2006), followed by examination for PCR products on agarose gels. The identity of the coccidia present in a few samples could not be clearly resolved by ssu rRNA gene sequencing alone, and these samples were therefore checked for the presence of Neospora caninum-DNA, using the N. caninum specific primers Np21/Np6 (Yamage et al. Reference Yamage, Flechtner and Gottstein1996), which amplify part of the pNc5-gene. DNA from N. caninum, S. alces, S. hjorti, and the Hammondia sp. detected in this study were included in the PCR as positive and negative controls.
Design of species-specific ssu rRNA gene primers for S. alces and S. ovalis
Species-specific primers were designed for those Sarcocystis species which were readily seen in the moose musculature, i.e., S. alces and S. ovalis. Primer-BLAST was used to search for suitable primers along the entire ssu rRNA gene with the following parameters for primer design: optimal melting temperature of 60±3°C, optimal primer length of 20 bases, PCR product of minimum 200 bases and maximum 1000 bases. Primer specificity was checked using the ‘nr’ Nucleotide Sequence Database at NCBI. Three sense and 2 antisense oligonucleotides were designed and obtained from Invitrogen™ (Table 1). The suitability of the primers was tested by using for each reaction: 3 μl DNA template, 12·5 μl HotStarTaq Master Mix (Qiagen GmbH, Germany), 10 pmol of each primer, 4 μg bovine serum albumin, and RNase-free water to make a final volume of 25 μl. The annealing temperature was optimized by increasing the temperature stepwise from 52°C to 60°C. Sensitivity of the PCR assay was determined by diluting genomic DNA preparations, starting with a 1:1 dilution and increasing the proportion of water stepwise to a final 1:100 dilution. PCR products (10 μl) were separated and visualized on 1% agarose gels. In addition to the specificity check by Primer-BLAST, the primer pairs were also tested with DNA from all Sarcocystis species that we previously had characterized or identified by molecular methods from cervids (S. alces, S. alceslatrans, S. hjorti, Sarcocystis gracilis, Sarcocystis grueneri, Sarcocystis hardangeri, S. ovalis, Sarcocystis oviformis, Sarcocystis rangi, Sarcocystis rangiferi, S. scandinavica, Sarcocystis tarandi, Sarcocystis tarandivulpes, and Sarcocystis sp. Type D of moose) (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2007, Reference Dahlgren and Gjerde2008, Reference Dahlgren and Gjerde2009, Reference Dahlgren and Gjerde2010). All primers were also tested using DNA from the Hammondia sp. detected in this study and with fox-DNA obtained from a liver sample.
Specific primers developed to selectively amplify regions of the ssu rRNA gene of Sarcocystis alces and Sarcocystis ovalis/S. hardangeri
* f=forward; r=reverse.
The faecal and mucosal samples from the foxes were finally tested using the following optimized PCR protocol. Each reaction mixture contained: 4 μl DNA template, 25 μl HotStarTaq Master Mix (Qiagen GmbH, Germany), 20 pmol of each primer, 8 μg bovine serum albumin, and RNase free water to make a final volume of 50 μl. PCR-cycling conditions were: initial hot start at 95°C for 15 min, followed by 35 cycles of 94°C for 40 s, 58°C for 45 s, and 72°C for 60 s; and a final extension at 72°C for 10 min.
Examination of moose for sarcocysts of S. hjorti
Due to the finding of S. hjorti-infection in several foxes (see later), the oesophagus and/or portion of the diaphragm from an additional 11 moose were examined in October 2009, in an attempt to determine whether S. hjorti was a common species in moose and whether the cysts of this species could be differentiated from those of S. alces in situ at low magnification. The samples were obtained at the abattoir-like plant described above from moose killed in Nordmarka during the regular hunt in 2009. The tissue samples were examined under a stereomicroscope and a large number of sarcocysts were excised and further examined under a light microscope for hair-like surface protrusions consistent with those of S. hjorti in red deer (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2010).
RESULTS
Sarcocystis species identified by light microscopy and molecular methods in muscle tissues from moose used in the experiments
Experiments 1 and 2
Numerous 2–5 mm long and spindle-shaped cysts were found in the oesophagus, diaphragm, and abdominal muscles. The size, shape, and surface morphology of these cysts were consistent with those of S. alces (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008), and molecular examination of one such cyst, confirmed its identity as S. alces (100% identity with another S. alces sequence in GenBank, Accession number EU282018). A small cyst, measuring 3·0×0·15 mm, that was found in the muscles of one of the moose used in Exp. 1, had similar delicate hair-like protrusions, about 10 μm long (Fig. 1A, B) as cysts of S. alceslatrans of moose and S. hjorti of red deer (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008, Reference Dahlgren and Gjerde2010). Molecular examination revealed that this cyst belonged to the previously reported Sarcocystis sp. Type E in moose (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008), which more recently has been described as S. hjorti in red deer (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2010) (100% identity with another S. hjorti sequence and Sarcocystis Type E sequence in GenBank, Accession numbers GQ250990 and EU282017). A few 1–2 mm long and oval cysts, consistent with those of S. ovalis in moose (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008) were also observed in the moose tissues used in both experiments.
Clinical observations
All foxes were clinically normal at the beginning of the experiments. Six of the 8 foxes in Exp. 1 had reduced feed intake, soft faeces and/or vomited between day 6 and 10 p.i. One fox in Exp. 2 had reduced feed intake and soft faeces on day 8 p.i., and 1 fox had diarrhoea and vomited on day 13 p.i. The blood parameters in each fox varied slightly between sampling days, but fluctuated within the same range as the control samples from foxes and also within the normal range for cats and dogs. No increase or decrease in any blood value seemed to be associated with the experimental infections.
Microscopic findings in faecal samples from foxes
Experiment 1
A few Sarcocystis sporocysts, measuring approximately 12·5×8 μm, were detected in the faeces of 1 silver fox on day 13 p.i. and in the faeces of another silver fox on day 14 p.i.; all other samples were negative for sporocysts (Table 2). Unsporulated Hammondia oocysts, about 10–12 μm in diameter (Fig. 2), were detected in the faeces of all foxes from days 6–8 p.i. onwards, except in 1 of the foxes killed on day 7 p.i., which, however, had numerous oocysts in the ileum contents at that time. The Hammondia oocysts were shed in large numbers during the first 1–5 days of patency, later on only in low numbers or intermittently, until the animals were euthanized (Table 2).
Unsporulated oocysts of Hammondia sp. from small intestinal contents of a blue (arctic) fox 7 days post-feeding with muscle tissue from moose. Scale bar=10 μm.
Experimental infections of foxes with Sarcocystis-infected muscle tissue from moose (Microscopic and molecular findings in faecal samples and mucosal scrapings from the small intestine.)
1 Day refers to day after feeding moose tissues.
2 Blue fox=arctic fox, silver fox=red fox.
3 ++ Days with high numbers of Hammondia oocysts.
4 Probably derived from the food, which included offal.
Experiment 2
No Sarcocystis oocysts/sporocysts were detected in any of the daily or pooled faecal samples, or in faeces retrieved from the colon of the 2 foxes killed on day 15 p.i. However, a single sporulated Sarcocystis oocyst was found by flotation of ileum contents from 1 of the foxes killed on day 15 p.i., indicating that oocyst shedding was about to begin (Table 2). A few unsporulated Hammondia-like oocysts were seen in the faeces of 3 of the foxes on day 8 p.i. In 1 of these foxes, a few large, oval oocysts, measuring about 35×27 μm, and resembling those of Isospora canivelocis, were also found on day 8 p.i.
In both experiments, a few ascarid-type nematode eggs and oocysts without a recognizable sporont inside were seen in some samples. These were considered to be spurious parasites that the foxes had ingested with their ordinary feed, which contained viscera of slaughtered animals.
Microscopic findings in intestinal scrapings from foxes at necropsy
Experiment 1
A few scattered unsporulated, ellipsoidal Sarcocystis oocysts, measuring about 17·5×12·5 μm (Fig. 3A), were seen in the anterior part of the intestine in both foxes killed on day 7 p.i. (Table 2). A moderate number of sporulated, thin-walled Sarcocystis oocysts were seen in the mucosa throughout the length of the small intestine of all animals killed on day 14, 21, and 28 p.i. (Fig. 3C). The 2 sporocysts inside the oocysts each measured around 14–15×10 μm. Each sporocyst contained 4 sporozoites and a compact granular residual body.
Oocysts of Sarcocystis alces and/or S. hjorti in scrapings from the intestinal mucosa of foxes fed muscle tissues of moose. (A) Unsporulated oocysts in a blue fox 7 days post-infection. (B) Partly sporulated oocysts in a blue fox 8 days post-infection. (C) Fully sporulated oocysts in a silver (red) fox 14 days post-infection. The oocysts have a thin wall (arrows) and each oocyst contains 2 sporocysts with 4 sporozoites (sz) and a granular residual body (rb). (D) Sporulated oocysts in a silver fox 15 days post-infection. Scale bars=10 μm in (A−D).
Numerous unsporulated, subspherical Hammondia oocysts were observed in the mucosa along the entire posterior half of the small intestine in both foxes killed on day 7 p.i. A few Hammondia oocysts were also found in the ileum of 1 fox killed on day 14 p.i. (Table 2).
Experiment 2
A moderate number of unsporulated and partly sporulated, ellipsoidal Sarcocystis oocysts (Fig. 3B), measuring 15–20×10–15 μm, were seen along almost the entire length of the small intestine in both foxes killed on day 8 p.i. (Table 2). A moderate number of sporulated Sarcocystis oocysts, measuring 15–16×20 μm, were seen in the small intestinal mucosa of both animals killed on day 15 p.i. (Fig. 3D). The 2 sporocysts inside the oocysts each measured approximately 15–16×10 μm and had the same general morphology as in Exp. 1.
Identification of parasite stages in faeces or intestinal scrapings by molecular methods
DNA recovery from faeces or intestinal scrapings
DNA extraction from Sarcocystis oocysts/sporocysts in both the intestinal mucosa and faeces was most efficient with the QIAamp® DNA Mini Kit, using the standard ‘tissue protocol’ as described by the manufacturer. DNA isolation from faeces or from the intestinal mucosa using QIAamp® DNA Stool Mini Kit resulted in a lower DNA concentration. Heat treatment of oocysts/sporocysts before DNA isolation did not seem to improve the DNA yield, but only increased the duration of DNA isolation. There was no difference in DNA yield from sporocysts/oocysts between those that had been stored in water or ethanol before DNA isolation.
Specificity of the specific primer pairs and sensitivity of the PCR assay
Primer combination SD1f/SD1r produced a single band of approximately 620 bp on the agarose gels for S. alces. This primer combination did not amplify any other DNA tested in the PCR assay. Primer combination SD2f/SD1r produced a band of approximately 400 bp on the agarose gels and was specific for S. alces. Primer set SD3f/SD3r produced a band of approximately 680 bp on agarose gels for both S. ovalis and S. hardangeri. No PCR product was produced for any other DNA tested in the assay.
PCR products from DNA templates diluted 1:100 (or 0·01 ng/μl of S. alces and 0·13 ng/μl of S. ovalis) were still clearly visible on agarose gels after amplification using any of the 3 above-mentioned primer combinations. The sensitivity of the primers was not tested at further dilutions.
Experiment 1
Samples from 7 of the 8 foxes were positive on agarose gels after PCR using general coccidian primers. Samples from 3 foxes were positive on agarose gels after PCR using S. alces-specific primers. Samples from 7 animals were positive on agarose gels after PCR using Hammondia-specific primers. All PCRs using S. ovalis-specific primers and Neospora-specific primers were negative. The Neospora-specific primers were initially evaluated using DNA from S. alces, S. hjorti, and Hammondia sp., and all of these PCRs were negative as expected.
One sample from each of the 7 foxes that had been positive for coccidian DNA using the universal primers was sequenced. For most PCR products, fine single peak sequencing-chromatograms indicated the presence of DNA from either of S. alces, S. hjorti or Hammondia sp. However, sequencing reactions of a few other samples resulted in multiple peaks on chromatograms from 1 of the 2 sequencing primers (forward or reverse), or the sequencing primers produced 2 different DNA sequences. In total, 3 of the foxes were proven by sequence analysis to be infected by S. alces and 2 foxes by S. hjorti (Table 2).
Experiment 2
All 4 foxes were positive on agarose gels after PCR using general coccidian primers, whereas 1 fox was positive on agarose gel after PCR using the S. alces-specific primers. All PCRs using S. ovalis-specific primers, Hammondia-specific primers, or Neospora-specific primers were negative.
One sample from each of the 4 foxes which had been positive for coccidian DNA using the universal primers was sequenced. Most sequencing reactions resulted in fine single peak chromatograms which indicated the presence of DNA from either S. alces or S. hjorti, but a few samples resulted in multiple peaks on the chromatogram for one of the sequencing primers (forward or reverse), or the sequencing primers produced 2 different DNA sequences. Sequence analysis indicated that all 4 foxes were infected with S. hjorti and at least 1 fox also with S. alces (Table 2).
Examination of muscle samples of moose in 2009 for sarcocysts of S. hjorti
In 4 of 11 moose examined, altogether 7 cysts were consistent with those of S. hjorti, i.e., they were small, slender, and thread-like, measuring about 2×0·2 mm, with about 10 μm long, hair-like protrusions (Fig. 1C). The surrounding host cell material was difficult to separate from these cysts and some of the hair-like protrusions were torn off during this process, or only parts of the cyst surface were exposed. S. hjorti-like cysts were only found in the diaphragm. All 11 moose examined, had numerous cysts consistent with those of S. alces, i.e., they were spindle-shaped, approximately 1–4×0·2–0·3 mm, with no visible surface protrusions, or with barely visible knob-like protrusions. Ten moose also had large oval sarcocysts consistent with those of S. ovalis (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008).
DISCUSSION
The present study revealed that red foxes and arctic foxes may serve as definitive hosts for 2 Sarcocystis species, i.e., S. alces and S. hjorti, as well as 1 Hammondia sp., using moose as intermediate host. The foxes were not suitable definitive hosts for the third Sarcocystis species, i.e. S. ovalis, present in the moose musculature used in the experiments. Such mixed natural infections with several Sarcocystis species, using different definitive hosts, are very common both in domestic animals (Dubey et al. Reference Dubey, Speer and Fayer1989) and in free-ranging cervids (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2007, Reference Dahlgren and Gjerde2008, Reference Dahlgren and Gjerde2009, Reference Dahlgren and Gjerde2010). By microscopy, it was only possible to differentiate between Hammondia-type oocysts and Sarcocystis oocysts, whereas molecular methods were necessary to further identify the Sarcocystis oocysts to species.
This seems to be the first study in which molecular methods have been used to verify the results of transmission experiments aimed at determining the definitive host(s) of Sarcocystis species of cervids. The life cycle of some species in cervids has been determined by feeding potential definitive hosts Sarcocystis-infected muscle tissue followed by screening of the faeces and/or the intestinal mucosa of the carnivores for oocysts/sporocysts, and then simply assuming that the sporocysts belonged to the (single) species identified as sarcocysts in the intermediate host (e.g., S. alceslatrans in moose: Dubey, Reference Dubey1980; Colwell and Mahrt, Reference Colwell and Mahrt1983). In a few experiments, micro-isolated sarcocysts of a single species have been used to ascertain that the sporocysts belonged to a particular species (e.g., S. rangi of reindeer: Gjerde, Reference Gjerde1985 a), or the identity of the sporocysts has been verified by their ability to induce formation of sarcocysts consistent with particular species after inoculation into intermediate hosts (e.g., S. gracilis and S. capreolicanis in roe deer: Erber et al. Reference Erber, Boch and Barth1978). The oocysts/sporocysts of only a few Sarcocystis species, particularly those of S. neurona and S. falcatula in opossums, have previously been identified by molecular methods (Fenger et al. Reference Fenger, Granstrom, Langemeier, Stamper, Donahue, Patterson, Gajadhar, Marteniuk, Xiaomin and Dubey1995; Tanhauser et al. Reference Tanhauser, Yowell, Cutler, Greiner, MacKay and Dame1999; Cheadle et al. Reference Cheadle, Dame and Greiner2001; Elsheikha et al. Reference Elsheikha, Murphy and Mansfield2004; Xiang et al. Reference Xiang, Chen, Yang, He, Jiang, Rosenthal, Luan, Attwood, Zuo, Zhang and Yang2009).
In the present study, molecular methods proved to be crucial in detecting a mixed infection with S. alces and S. hjorti in the foxes, and hence in the moose used as a source of the infective feed. Surprisingly, more foxes appeared to be infected with S. hjorti than with S. alces, even though the moose musculature used in both experiments seemed to contain mainly cysts of S. alces. An examination of muscle samples from an additional 11 moose subsequent to the infection experiments, with this new knowledge in mind, confirmed that cysts of S. alces vastly outnumbered those of S. hjorti in moose. The more frequent finding of S. hjorti in the foxes could be valid and, if so, was possibly due to a higher infectivity of S. hjorti than of S. alces, either as an inherent property of the species, or related to a better survival of the cysts after the death of the host, or it could be an artefact of the molecular methods used.
In the present study, sarcocysts of S. hjorti were described for the first time from moose. The small size and scarcity of sarcocysts of S. hjorti in moose compared with those of S. alces, may explain why no cysts of S. hjorti were detected by microscopy in a previous study of Sarcocystis species in Norwegian moose, although 1 cyst of S. hjorti (referred to as Sarcocystis sp. Type E) was detected by molecular methods (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008). The seemingly higher prevalence and infection intensity of S. hjorti in red deer as compared with moose, suggest that this species is better adapted to red deer than to moose. This study also confirmed that sarcocysts of S. hjorti in Norwegian moose are morphologically very similar to those of S. alceslatrans previously described from Canadian moose (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008), which had been expected from the appearance of S. hjorti in red deer and the close phylogenetic relationship between S. hjorti and S. alceslatrans (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2010).
Based on their phylogenetic position (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008, Reference Dahlgren and Gjerde2010), we had predicted that S. alces and S. hjorti would use canids as definitive hosts, and these predictions were confirmed in the present study. Previous studies by others have also indicated that different Sarcocystis species co-evolved with their definitive hosts (Holmdahl et al. Reference Holmdahl, Morrison, Ellis and Huong1999; Doležel, Reference Doležel, Koudela, Jirků, Hypsa, Oborník, Votýpka, Modrý, Slapeta and Lukes1999; Slapeta et al. 2003; Elsheikha et al. Reference Elsheikha, Lacher and Mansfield2005), and phylogenetic analysis may therefore be a useful complementary tool in the search for the definitive host(s) of those Sarcocystis species for which they are unknown (Dahlgren et al. Reference Dahlgren, Gouveia-Oliveira and Gjerde2008). Presumably, dogs and other canids may also act as definitive hosts for S. alces and S. hjorti in addition to foxes. Since S. alceslatrans has previously been found to use dogs and coyotes as definitive hosts (Dubey, Reference Dubey1980; Colwell and Mahrt, Reference Colwell and Mahrt1983), there are currently 3 known Sarcocystis species in moose that are being transmitted by canids. However, there might be still another species in moose using canids as definitive hosts. As noted previously (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2008), Sedlaczek and Zipper (Reference Sedlaczek and Zipper1986) described, by use of transmission electron microscopy, sarcocysts of a species in a moose in Germany, which they believed were S. alceslatrans, but which are more similar to S. grueneri of reindeer (Gjerde, Reference Gjerde1985 b), S. cervicanis of European red deer (Hernández et al. Reference Hernández, Martínez-Gómez, Navarrete and Acosta-García1981) and S. wapiti of North American wapiti (Speer and Dubey, Reference Speer and Dubey1982), all of which use canids as definitive hosts.
In this study, sensitive primer pairs were developed, which specifically amplify parts of the ssu rRNA gene of either S. alces or S. hardangeri and S. ovalis. DNA from S. hardangeri/S. ovalis was not detected in the faeces or intestinal mucosa of any fox, and hence canids are apparently not suitable definitive hosts for these two species, which is not very likely based on their phylogenetic positions (Dahlgren and Gjerde, Reference Dahlgren and Gjerde2010). The use of species-specific primers provides a fast and easy method for screening multiple samples for a particular Sarcocystis species. However, as seen from this study, it is necessary to use more general primers and sequence a few samples in order to detect a mixed infection with unexpected species, not targeted by species-specific primers.
In Experiment 1, all foxes shed or harboured Hammondia heydorni-like oocysts. PCR-assays using Hammondia-specific and Neospora caninum-specific primers, confirmed that these oocysts belonged to a Hammondia species, and not to the morphologically indistinguishable species N. caninum, which, however, does not seem to use foxes as definitive hosts (Dubey et al. Reference Dubey, Barr, Barta, Bjerkås, Bjørkman, Blagburn, Bowman, Buxton, Ellis, Gottstein, Hemphill, Hill, Howe, Jenkins, Kobayashi, Koudela, Marsh, Mattsson, McAllister, Modrý, Omata, Sibley, Speer, Trees, Uggla, Upton, Williams and Lindsay2002; Schares et al. Reference Schares, Heydorn, Cüppers, Mehlhorn, Geue, Peters and Conraths2002). Recent molecular comparisons of Hammondia isolates from foxes and dogs, respectively, have indicated that they are different at the D2/D3 domain of the large subunit rRNA gene and in the intron 1 region of the alpha tubulin gene (Schares et al. Reference Schares, Heydorn, Cüppers, Mehlhorn, Geue, Peters and Conraths2002; Mohammed et al. Reference Mohammed, Davies, Hussein, Daszak and Ellis2003; Abel et al. Reference Abel, Schares, Orzeszko, Gasser and Ellis2006), and that Hammondia sp. of foxes might therefore be different from Hammondia heydorni of dogs. Hence, it will be of considerable interest to examine the present Hammondia isolate from foxes more extensively at various genetic markers to determine its relationship to other Hammondia isolates from foxes and dogs. Moose has previously been reported to be an intermediate host of H. heydorni infecting dogs (Dubey and Williams, Reference Dubey and Williams1980), but this seems to be the first report of moose as intermediate hosts of a Hammondia sp. using foxes as definitive hosts. Previously, sheep (Ashford, Reference Ashford1977), roe deer (Entzeroth et al. Reference Entzeroth, Scholtyseck and Greuel1978), reindeer (Gjerde, Reference Gjerde1983), sheep and goats (Schares et al. Reference Schares, Heydorn, Cüppers, Mehlhorn, Geue, Peters and Conraths2002) and the Arabian mountain gazelle (Mohammed et al. Reference Mohammed, Davies, Hussein, Daszak and Ellis2003) have been reported to be natural intermediate hosts of a Hammondia sp. infecting foxes. The oocyst morphology, pre-patent period and pattern of oocyst shedding in the foxes in the present study were similar to those reported from Hammondia sp. in foxes in the above-mentioned papers, but also similar to those of H. heydorni in dogs (summarized in Table 2 in Dubey et al. Reference Dubey, Barr, Barta, Bjerkås, Bjørkman, Blagburn, Bowman, Buxton, Ellis, Gottstein, Hemphill, Hill, Howe, Jenkins, Kobayashi, Koudela, Marsh, Mattsson, McAllister, Modrý, Omata, Sibley, Speer, Trees, Uggla, Upton, Williams and Lindsay2002).
In Experiment 1, several blue foxes and silver foxes developed reduced feed intake or gastrointestinal symptoms of 1 to a few days duration, about 1 week after infection. These symptoms were most likely associated with the Hammondia infection. Thus, H. heydorni infections in dogs have been associated with diarrhoea (Abel et al. Reference Abel, Schares, Orzeszko, Gasser and Ellis2006). The intestinal development of the 2 Sarcocystis species and the Hammondia sp. did not seem to affect any of the blood parameters analysed in this study. It is possible that the infection intensity was too low to produce any recognizable changes in the blood parameters, or that infection with Sarcoystis or Hammondia only caused local damage to the intestinal mucosa, without affecting any component of the blood. Sarcocystis species are generally considered to be non-pathogenic for their definitive hosts (Dubey et al. Reference Dubey, Speer and Fayer1989).
In conclusion, the definitive host of S. alces and S. hjorti was determined by experimental infection and molecular diagnosis to be the red fox and the arctic fox. The first experiment also showed that moose may act as intermediate host for a Hammondia species using foxes as definitive hosts, but further molecular studies are necessary in order to determine whether this Hammondia isolate should be regarded as a species different from H. heydorni of dogs.
ACKNOWLEDGEMENTS
We would like to thank Løvenskiold Skog for allowing us to collect samples from moose at their processing plant for moose-carcasses in Nordmarka. We also thank Målfrid Stadaas and other staff members at the Animal Production Experimental Centre, for surveillance of the foxes during the experimental periods and for collecting faecal samples and assisting us during blood sampling.
