CLINICAL, CULTURE, SEROLOGY, AND HISTOPATHOLOGY OUTCOMES OF BIGHORN SHEEP EXPERIMENTALLY INFECTED WITH BRUCELLA OVIS

Rocky mountain bighorn sheep (BHS; Ovis canadensis canadensis) are an icon of the American West. Effective management of BHS is difficult due to periodic disease-related die-offs and habitat loss or fragmentation (Buechner, 1960; Geist, 1971; Foreyt and Jessup, 1982; Jessup, 1985; Singer et al., 2000). As part of herd management strategies, BHS are captured and translocated to form new herds or supplement existing herds. When translocations cross state lines, state animal health authorities usually require testing for selected domestic sheep (Ovis aries) pathogens including Brucella ovis.

Brucella ovis, a gram negative coccobacillus infecting domestic sheep, occurs in most sheep-raising areas of the world. In domestic sheep, B. ovis produces male reproductive tract lesions including epididymitis, testicular atrophy, adhesions of the tunica vaginalis, and decrease in semen quality (Hartley et al., 1955; Blasco, 1990; Bulgin, 1990). Infection in ewes is less severe, causing occasional abortions and increased perinatal mortality; however, infection rarely extends from one pregnancy to the next (Hartley et al., 1955; Ris, 1970). Natural transmission in domestic sheep primarily results from homosexual behavior between rams and passive venereal transmission from ram to ram via infected semen in the ewe's vagina (Buddle, 1955; Hartley et al., 1955). Mating by infected rams causes a low percentage of ewes to become infected, and although infected females shed B. ovis vaginally, they have not been shown to transmit to naïve males (Buddle, 1955; Grillo et al., 1999). Lambs born to infected dams often have maternal antibody but are rarely infected (Buddle, 1955; Keogh et al., 1958; Clapp, 1962).

Among wildlife, natural infection with B. ovis resulting in lowered semen quality has been detected in red deer (Cervus elaphus elaphus) in New Zealand (Scott, 1998). Experimental infection of white-tailed deer (Odocoileus virginianus) resulted in epididymal lesions and lowered semen quality (Barron et al., 1985). An attempt to experimentally infect mouflon (Ovis musimon) with B. ovis was unsuccessful (Cerri et al., 2002).

Serum samples from wild BHS destined for translocation are rarely positive for B. ovis antibodies on serologic testing. When this occurs, it places regulatory officials in a quandary as to the appropriate action because B. ovis has never, to our knowledge, been isolated from BHS despite serologic evidence of exposure. In Montana there were three antibody-positive and three suspect cases out of 1,660 samples tested from 1991 to 2010 (N. Anderson, pers. comm.). In Idaho, there were seven antibody-positive and six suspect animals out of 1,458 samples tested from 1988 to 2010 (M. Drew, pers. comm.). The purpose of this study was to determine the susceptibility of BHS to experimental infection with B. ovis and to document the reproductive outcomes and lesions produced from an experimental infection.

Wild-caught and captive-raised BHS and domestic sheep were used in this study. Fourteen BHS (nine ewes and two rams >2-yr-old; two yearling ewes, and one yearling ram) were captured by helicopter netgun in southwest Montana (45°85′N, 113°94′W) in the spring of 2004 and immediately transported to Colorado State University, Animal Population Health Institute, Wildlife Research Facility (WRF), Fort Collins, Colorado, USA (40°35′N, 105°14′W). Twenty captive-raised adult (>2-yr-old) BHS (10 ewes and 10 rams) housed for 26 mo or longer at the Wildlife Health Laboratory, Caldwell, Idaho, USA (43°38′N, 116°40′W) were transported to the WRF.

Wild-caught and captive-raised BHS contained in three 30×30 m pens with free choice access to hay and water were allowed to acclimate at WRF for 6 and 7 wk, respectively. A squeeze chute designed for restraint and handling of deer (Delclayna Mfg, Swanville, Minnesota, USA) was used for handling of the sheep for sample collection.

Seven domestic ewes and 11 rams obtained from two commercial flocks in Colorado were kept on a 50×70 m pasture near the WRF. Domestic sheep were allowed to acclimate for 2 wk; they were handled using manual restraint. All animals were negative for antibodies to B. ovis and were housed, handled, and cared for under protocols approved by the Institutional Animal Care and Use Committee at Colorado State University (IACUC project 04-034A-01).

Semen was collected via electro-ejaculation (Lane Ram Ejaculator, Lane Manufacturing, Denver, Colorado, USA) from a B. ovis antibody-positive domestic ram originating from a commercial sheep flock from Colorado, USA. Brucella ovis was isolated from the semen sample by the National Veterinary Services Laboratories (NVSL), Ames, Iowa, USA, where the sample was prepared for inoculation after one passage.

Bighorn sheep were assigned to three treatment groups. Wild-caught and captive-raised BHS were housed separately to facilitate a study examining stress that was conducted concurrently (Coburn et al., 2010). Twenty BHS (nine males, 11 females, of which nine were pregnant) were inoculated intraconjunctivally (IC) in each eye with 2.7×10⁸ colony forming units (cfu) B. ovis suspended in 0.5 mL phosphate-buffered saline (PBS) at pH 7.2. Six BHS (three males, three pregnant females) received 1 mL PBS IC and were housed with inoculated animals to serve as in-contact controls. The remaining eight BHS (one male, seven females, of which five were pregnant) received 0.5 mL PBS IC in each eye and were housed in a separate pen to serve as noncontact controls. On the following day, 14 domestic sheep (nine males, five pregnant females) were inoculated in the same manner to serve as positive controls to verify the pathogenicity of the organism at this concentration and to compare B. ovis infection in the two species. Five domestic sheep (three males, two pregnant females) received 1 mL PBS IC and were housed with the inoculated domestic sheep to serve as contact controls. At inoculation, bighorn and domestic sheep were within 4–8 wk and 1–4 wk, respectively, of lambing.

Blood samples were collected by venipuncture from wild-caught BHS at the time of capture, from captive-raised BHS prior to transport, and from domestic sheep on arrival at WRF. Subsequent samples were collected on the day of inoculation, every 2 wk for 6 wk, and monthly thereafter until final collection at necropsy. Blood was collected in heparinized tubes for B. ovis culture and held at −70 C until shipment on dry ice to NVSL. Serum was separated and stored at −70 C until shipment on dry ice to Rocky Mountain Regional Animal Health Laboratory, Denver, Colorado, USA, for a panel of tests consisting of two enzyme-linked immunosorbent assays (ELISAs) and a compliment fixation test. We report here only the results of the standard ELISA used for regulatory purposes in the USA at the time. In 2008 an updated ELISA was instituted in the USA. Each plate was interpreted based on the optical density (OD) of the negative control and the OD of the weak positive control on that individual plate. Any test sample that had an OD greater than the weak positive control was identified as positive. Suspects were defined as samples producing an OD greater than the negative control but less than the weak positive control. A comparison of multiple serologic tests will be reported elsewhere.

Within 2 days of parturition, stillbirth, abortion, or 19 wk postinoculation (PI), B. ovis–inoculated ewes of both species, and their lambs were euthanized and necropsied. Blood was collected for culture from control ewes and their lambs within 2 days of parturition. Inoculated rams and control ewes and their lambs were euthanized and necropsied between 14 and 19 wk PI for BHS and 24 wk PI for domestic sheep. Tissues collected for culture and histopathology included liver, lung, spleen, kidney, heart, blood, and the following lymph nodes: retropharyngeal, parotid, superficial inguinal, mandibular, internal iliac, mesenteric, mediastinal, and hepatic. In addition, the uteri and mammary tissues were collected from the ewes. Testes, epidydimides (head, body, and tail), ductus deferens, prostate, ampulla, and seminal vesicles were collected from the rams. Tissues collected from fetal and neonatal lambs included spleen, liver, lung, kidney, placenta, and abomasal fluid. Tissues for culture were stored at −70 C and subsequently sent on dry ice to NVSL, where tissues were processed and cultured using standard techniques (Ewalt, 1989). Inoculated media were incubated at 37 C in 10% CO₂ for 7 days (Mayfield et al., 1990) and observed for growth at 4 and 7 days. Colonies with typical B. ovis morphology were transferred to tryptic soy agar and tryptose agar with 5% sheep blood and incubated for 2–5 days until there was enough growth to run the biochemical, agglutination, and phage typing tests (Mayfield et al., 1990). Specimens for histopathologic examination were fixed in 10% neutral buffered formalin and sent to NVSL where they were processed in paraffin, sectioned at 5 µm, and routinely stained with hematoxylin and eosin. Immunohistochemical (IHC) staining of selected sections to demonstrate B. ovis antigen was conducted at the National Wildlife Research Center in Fort Collins, Colorado, USA, using a commercially available system (EnVision®+System, DakoCytomation, Carpinteria, California, USA). Sections were incubated at 37 C with a primary rabbit antiserum to B. ovis at a 1∶500 dilution for 30 min. Nonimmune rabbit serum was substituted for the primary antibody as a control for nonspecific reactions, and tissues from noninfected BHS were used as negative controls.

Fisher's exact test was used to compare BHS rams to domestic rams based on tissue culture, blood culture, and histologic lesions (Proc FREQ; SAS 9.1, SAS Institute, Cary, North Carolina, USA). Data obtained from BHS ewes could not be statistically compared to those of domestic ewes due to differences in gestational stages at the time of inoculation. Statistical priority was placed on minimizing the chance of type II error in order to detect differences that might be biologically valid. Due to the pilot status of this experiment, statistical significance was considered when P≤0.1.

No differences were observed between wild-caught and captive raised BHS; therefore those data were pooled for analysis. Six of 10 BHS lambs born to nine inoculated dams were stillborn or died within 48 hr. This included one set of stillborn twins. Seven of eight lambs born to eight control BHS dams were apparently healthy. One control BHS ewe had a uterine tear, and both the ewe and lamb died. Eight of the 10 BHS lambs born to inoculated ewes were culture positive. All lambs born to uninoculated BHS dams were culture negative. Two of eight lambs born to inoculated domestic ewes were stillborn or nonviable (one twin and one triplet); neither was culture positive. None of the six healthy lambs born to inoculated domestic ewes were culture positive. The three lambs born to control domestic ewes were healthy and culture negative.

Eighteen of 20 B. ovis–inoculated BHS (8/9 rams and 10/11 ewes) became blood culture positive at least once during the course of the study. Most animals were blood culture positive 4 wk PI and remained positive for 4 wk before testing negative. Tissues from 19 of 20 inoculated BHS were culture positive at necropsy. Tissues most often found to be positive were epididymides, ductus deferens, ampullae, and testes in males, and internal iliac lymph node (ln), superficial inguinal ln, mandibular ln, retropharyngeal ln, and spleen in females. Among inoculated BHS rams, none were blood culture positive at the time of inoculation or 2 wk PI. Eight of nine were culture positive 4 wk PI, six of nine at 8 wk, two of nine at 12 wk, and one of seven at 16 wk. Two contact control BHS, one ewe and one unrelated lamb, were positive on one serology test and subsequently did not test positive. No control BHS (contact and noncontact) were positive on culture at any time during the study. Sera from 19 of 20 inoculated BHS were positive on ELISA for B. ovis within 4 wk PI and remained positive until the end of the study (Table 1). A subset of serum samples from the time of necropsy were inadvertently sent to the wrong lab and subsequently lost to the study.

Of the 14 inoculated domestic sheep, two rams and a ewe became blood culture positive. The ewe was positive 2 wk PI and negative at 4 wk PI while the rams were positive 4 wk PI and negative thereafter. That same ewe was culture positive on numerous tissues, and five of nine inoculated domestic rams were culture positive on at least five of the tissues collected. Sera from 10 inoculated domestic sheep were positive on ELISA for B. ovis from 2 wk up to 12 wk PI (Table 2). One of five inoculated domestic ewes was positive on ELISA, and six of nine inoculated rams were positive. No control domestic rams, ewes, or associated lambs were positive on serology or culture.

The number of culture positive BHS rams (9/9) was significantly higher than the number of culture positive domestic rams (5/9) (P = 0.0824). The number of blood culture positive BHS rams (8/9) was significantly higher than the number of blood culture positive domestic rams (2/9) (P<0.0152).

At necropsy, the BHS rams were in good to fair body condition. Epididymal lesions, including swelling and abscesses, were grossly observed in eight of nine inoculated rams, while none of four control rams showed lesions. Three of nine inoculated domestic rams had gross epididymal lesions, while none of three control rams showed lesions.

Microscopic examination of reproductive tract tissues from nine B. ovis–inoculated bighorn rams revealed epididymitis in all, ampullitis in six, seminal vesiculitis in five, and orchitis in two. Epididymitis, ampulitis, and seminal vesiculitis were focal or segmental and ranged from mild to severe. Lesions were characterized by mild to moderate, focal accumulations of lymphocytes and plasma cells scattered or perivascularly located in interstitial connective tissue; mild to moderate, focal accumulations of lymphocytes, plasma cells, and neutrophils with necrotic debris located in ductal and vesicular lumina; focal epithelial necrosis; and intraepithelial cysts with microabscesses in some epididymides. Sections of epididymides from two rams also contained partially mineralized casseous abscesses containing extravasated sperm, necrotic eosinophilic material, and cellular debris surrounded by infiltrates of macrophages, with moderate numbers of multinucleated giant cells, lymphocytes, plasma cells, and scattered neutrophils. Severe lesions involving much of the testicular parenchyma were present in two bighorn rams. Lesions were morphologically similar and were characterized by numerous round to elongate or tortuous, frequently coalescing areas of casseous necrosis (Fig 1a). Necrotic foci often contained central mineralization and small accumulations of neutrophils. They appeared to have originated in seminiferous tubules and followed the tubular architecture. Each area of necrosis was surrounded by a cellular mantle consisting primarily of macrophages with focal accumulations of neutrophils and few multinucleated giant cells. There was diffuse fibrosis and a moderate to severe cellular infiltrate consisting predominantly of lymphocytes with fewer plasma cells in the interstitium surrounding the necrotic foci.

Of the nine B. ovis–inoculated domestic rams, there was microscopic evidence of seminal vesiculitis and epididymitis in five and ampulitis in four. Orchitis was not noted in any of the rams. Seminal vesiculitis, ampullitis, and epididymitis were morphologically identical to that in the BHS. Epididymis of one ram had a large irregular casseous abscess containing numerous focal accumulations of neutrophils surrounded by macrophages, lymphocytes, plasma cells, and neutrophils. Microscopic lesions were not present in reproductive tract tissues from the control bighorn or domestic rams.

Microscopic lesions were present in reproductive tract tissues from nine of nine BHS rams and five of nine domestic rams; the number of lesioned BHS being statistically greater than the number of domestics (P = 0.0824). Immunohistochemical stained sections contained particulate B. ovis antigen in epididymal lesions of domestic and bighorn sheep, in sections of testes from two bighorn rams with orchitis, and in uterine and placental exudate and epithelial cells of BHS (Fig. 1a, b).

Of the inoculated BHS (male and female) that were monitored for 16 wk PI, 10 of 10 seroconverted and remained positive, while four of nine (male) domestics monitored for the same period seroconverted and remained positive (Tables 1 and 2).

Among inoculated BHS ewes, uterine lesions were noted in six of 10 and mastitis in five of eight animals from which uterus and mammary gland were examined. Uterine lesions consisted of mild to severe, multifocal accumulations of lymphocytes and plasma cells in the endometrial lamina propria, often located around glands. Glandular lumina contained neutrophils and degenerate cellular debris (Fig. 1b). Necrosis of endometrial epithelium was multifocal or diffuse. In some cases, caruncles were denuded and overlaid with a coagulum of fibrin, neutrophils, and necrotic cellular debris. Mastitis was characterized by mild to severe accumulations of lymphocytes and plasma cells in the intersitium and in some cases by neutrophils and lesser numbers of macrophages in glandular acini.

Three of five and one of four B. ovis–inoculated domestic ewes examined had uterine lesions and mastitis, respectively. Lesions were similar to those in the BHS. Uterine lesions were not noted in the control BHS or domestic ewes; mastitis, consisting of mild lymphoplasmacytic infiltrates in the glandular interstitium, was noted in one control bighorn and two control domestic ewes.

Tissue sections from two of the three stillborn BHS lambs and the three BHS lambs that died as neonates were examined. Mild to moderate pneumonia was present in two stillborn BHS lambs. The pneumonia was characterized by mild to moderate multifocal accumulations of lymphocytes and plasma cells, mixed with neutrophils in one lamb, located in the interstitium primarily around blood vessels. Small numbers of neutrophils and macrophages were present in some airways. Lung sections from one of the neonatal lambs had a lymphocytic pneumonia similar to those in the stillborns. Brucella ovis was isolated from the lungs of the three lambs with histologic evidence of pneumonia, from the stillborn from which tissues were not examined, and from one stillborn lamb without histologic evidence of pneumonia. Additionally, B. ovis was isolated from the spleen of three BHS lambs that were euthanized as neonates and one that was stillborn. Lung sections from two stillborn lambs from B. ovis–inoculated domestic ewes were unremarkable, and neither of those lambs was culture positive.

Placentas were examined from three B. ovis–inoculated BHS (one culture positive stillborn, one culture positive neonatal death, and one culture negative stillborn) and two BHS controls (Table 3). Necropurulent placentitis, present in the two culture-positive BHS, consisted of multifocal necrosis of placental epithelium, well-demarcated necrosis of cotyledonary villi accompanied by multifocal accumulations of neutrophils and degenerate cells in the chorionic stroma at the margin of necrosis, and diffuse, focally marked, infiltrates of mixed inflammatory cells in the chorionic stroma. Additionally, neutrophils mixed with necrotic cellular debris were adherent to the cotyledonary surface. Placentas from the culture negative inoculated lamb and from control BHS as well as domestic lambs were unremarkable.

Our results show that BHS are susceptible to infection and disease caused by experimental inoculation with B. ovis. Prior to this study, B. ovis had not been isolated from BHS nor have lesions in BHS ever been associated with B. ovis. Kreeger et al. (2004) documented infection of captive BHS with B. abortus through contact with an aborted elk fetus.

Controlled experiments involving rare, large, highly stressed species in confined conditions are typically limited by space, cost, and availability of study animals. Power and significance levels should be carefully considered prior to carrying out such a study, including the analysis of the study results. Studies with low sample sizes are susceptible to outcomes that fail to statistically demonstrate differences between or among treatment groups when a true treatment effect exists. We therefore increased the alpha level and, hence, the P value in order to increase the power of the test. The purpose of this pilot study was to investigate the susceptibility of BHS to B. ovis. Despite the small sample size, results indicate that B. ovis is equally or perhaps more pathogenic in BHS rams as compared to domestic rams. Bacteremia was detected in the majority of BHS rams compared to none of the domestic rams. Additionally, fewer domestic rams were tissue-culture positive than were BHS rams, and more bighorn rams had microscopic lesions in reproductive organs than domestic rams. Lesions in the epididymides, ampullae, and seminal vesicles were consistent with those previously described (Blasco, 1990; Ladds, 1993; Preziuso et al., 2009). The microscopic characteristics of the orchitis were similar to those observed with B. abortus infection in cattle (Ladds, 1993) and bison (Rhyan et al., 1997).

In this study, BHS ewes were exposed to the organism for a longer time before lambing and euthanasia than domestic ewes. This may be the reason that we were unable to culture the organism and document reproductive tract lesions from more than one domestic ewe. We were unable to acquire domestic ewes that were at a similar stage of gestation to the BHS ewes; therefore, direct comparison of BHS ewes to domestic ewes was not feasible since BHS ewes were inoculated in mid-gestation while domestics were inoculated in late gestation. However, prior studies investigating infections in domestic sheep at various stages of gestation are consistent with data from our study. In a domestic sheep study, the rate of abortion or poor doing lambs in IV-inoculated mid-gestation ewes was four of nine (Meinershagen et al., 1974). In another study of 26 lambs born to ewes inoculated 2–4 mo gestation, 13 were stillborn or poor-doing, and B. ovis was cultured from 20 (McGowan et al., 1961). Results from McGowan's late-gestation-inoculated ewes match well with our results for late-gestation-inoculated domestic ewes. Results from our BHS ewes are similar to those from the mid-gestation-inoculated domestic ewes in McGowan's study. Ewes inoculated 1–2 mo prior to parturition in McGowan's study had poor-doing lambs, though none of them aborted.

Uterine and placental lesions in the BHS were consistent with those reported for domestic sheep (Molello and Flint, 1963; Marco, 1994). Mild pneumonic changes in infected fetuses are associated with B. ovis infection in domestic sheep (Kennedy and Miller, 1993). The cause of the mild to moderate mastitic lesions in treatment and control bighorn and domestic sheep in this study is unknown. Among the contact control ewes, there was one transient seroconversion with a low titer that may be attributable to contact with the inoculated animals or aborted fetuses. Further research is needed to document intraspecies transmission.

Brucella ovis is primarily a pathogen of domestic rams (OIE, 2009), and based on our findings, this also may hold true for BHS rams. Furthermore, the dispersal and sexual behaviors of BHS rams would likely further increase risk of pathogen transmission. Studies have shown BHS rams to range over large areas, therefore increasing chances of comingling with infected animals. In a study of BHS rams in Alberta, Festa-Bianchet (1986) found them to range up to 48 km. Bighorn sheep have been known to breed with domestic sheep (Young and Manville, 1960), and it has been observed that domestic ewes in estrus attract BHS rams (USAHA Joint Working Group, 2009). In addition, similar to domestic sheep, BHS rams display polygamous and homosexual behavior, with the dominant rams treating subordinate rams identical to females in estrus (Geist, 1971). These behaviors of BHS rams may also contribute to a greater likelihood of infection through breeding with recently exposed ewes, as well as infected BHS rams.

In summary, following experimental infection with B. ovis, BHS had similar or greater susceptibility to infection as concurrently inoculated domestic sheep. Reproductive tract lesions were similar to or more severe than in domestics, and B. ovis was isolated from the blood, semen, tissues, and aborted fetal tissues of BHS. These findings indicate that naturally occurring B. ovis infection is possible in BHS and suggest the need for further investigation when antibody is detected in wild populations. Studies on the persistence of infection and duration of antibody are also needed.

We thank several individuals who contributed time, expertise, and hard work to the completion of this project. The Animal Population Health Institute, Colorado State University, and the Foundation for North American Wild Sheep provided funding. K. Held at the US Department of Agriculture, Animal and Plant Health Inspection Service (USDA APHIS) Veterinary Services, R. Thompson, and D. Miller at Colorado State University assisted with all aspects of animal handling, sampling, and sample processing. J. George and M. Vierra, and many other Colorado Division of Wildlife personnel assisted with equipment and sampling. S. Templin-Hladky (USDA APHIS Wildlife Services) conducted IHC tissue staining. Many others at the National Wildlife Research Center and the Center for Epidemiology and Animal Health assisted with sample collection. Mark Drew with Idaho Game and Fish and Neil Anderson with Montana Fish, Wildlife and Parks allowed us access to unpublished data or personal communications. T. Spraker with Colorado State University helped with sample collection and with editing and revising. S. Olsen with the National Animal Disease Center, Ames, Iowa, provided a primary rabbit antiserum to B. ovis for use in IHC staining.