Yersinia Enterocolitica: an Unlikely Cause of Positive Brucellosis Tests in Greater Yellowstone Ecosystem Bison (Bison Bison)

Journal of Wildlife Diseases, 48(3), 2012, pp. 537–541 # Wildlife Disease Association 2012

YERSINIA ENTEROCOLITICA: AN UNLIKELY CAUSE OF POSITIVE BRUCELLOSIS TESTS IN GREATER YELLOWSTONE ECOSYSTEM BISON (BISON BISON)

Wade See,1 William H. Edwards,2 Stacey Dauwalter,2 Claudia Almendra,6 Martin D. Kardos,1,7 Jennifer L. Lowell,1 Rick Wallen,3 Steven L. Cain,4 William E. Holben,1,7,8 and Gordon Luikart1,5,6,7 1 Division of Biological Sciences, University of Montana, Missoula, Montana 59812, USA 2 Wyoming Game and Fish Wildlife Disease Laboratory, 1174 Snowy Range Rd., Laramie, Wyoming 82070, USA 3 National Park Service, P.O. Box 168, Yellowstone National Park, Wyoming, 82190 USA 4 Grand Teton National Park, USDI, P.O. Drawer 170, Moose, Wyoming 83012, USA 5 CIBIO, Centro de Investigaçaõ em Biodiversidade e Recursos Gene´ticos, Campus Agra´rio de Vairaõ, 4485-661 Vairaõ, Portugal 6 Flathead Lake Biological Station, University of Montana, Missoula, Montana 59812, USA 7 Montana—Ecology of Infectious Diseases Program, University of Montana, Missoula, Montana 59812, USA 8 Corresponding author (email: [email protected])

ABSTRACT: Yersinia enterocolitica serotype O:9 has identical O-antigens to those of Brucella abortus and has apparently caused false-positive reactions in numerous brucellosis serologic tests in elk (Cervus canadensis) from southwest Montana. We investigated whether a similar phenomenon was occurring in brucellosis antibody–positive bison (Bison bison) using Y. enterocolitica culturing techniques and multiplex PCR of four diagnostic loci. Feces from 53 Yellowstone bison culled from the population and 113 free-roaming bison from throughout the Greater Yellowstone Ecosystem (GYE) were tested. Yersinia enterocolitica O:9 was not detected in any of 53 the bison samples collected at slaughter facilities or in any of the 113 fecal samples from free-ranging bison. One other Y. enterocolitica serotype was isolated; however, it is not known to cause cross-reaction on B. abortus serologic assays because it lacks the perosamine synthetase gene and thus the O-antigens. These findings suggest that Y. enterocolitica O:9 cross-reactivity with B. abortus antigens is unlikely to have been a cause of false-positive serology tests in GYE bison and that Y. enterocolitica prevalence was low in bison in the GYE during this study. Key words: Bison, brucellosis, elk, Greater Yellowstone Ecosystem, perosamine synthetase, Yellowstone, Yersinia enterocolitica O:9.

INTRODUCTION elk (Weynants et al., 1996; Munoz et al., 2005; Atkinson et al., 2007). Yersinia enterocolitica is an enteric, In 2004–2005, staff from Montana Fish, gram-negative, pathogenic bacterium Wildlife, and Parks detected high numbers found in domestic animals, wildlife, and of B. abortus antibody–positive elk (11.4%) humans (Bottone, 1997). Yersinia enter- in the Pioneer Mountains on the northwest ocolitica serotype O:9 and Brucella abor- perimeter of the Greater Yellowstone tus, the etiologic agent of brucellosis in elk Ecosystem (GYE), ,100 km northwest of (Cervus canadensis), bison (Bison bison), Yellowstone National Park where B. abor- cattle (Bos taurus), and certain other tus occurs in bison and some elk (e.g., Beja- ungulates, possess an identical O-antigen Pereira et al., 2009). In the Madison Valley structure encoded by the gene perosamine Elk Management Unit, approximately synthetase (Godfroid et al., 2002). This 20 km east of the Pioneer Mountains, genetic similarity is known to cause cross- antibody prevalence in sampled elk rose to reactivity between the two bacteria during 6.9% in 2004–2005 and to 17.5% in 2005– serologic testing. As a result, infection with 2006 samples (Atkinson, et al., 2007). Y. enterocolitica serotype O:9 can lead to Before 2004–2005, annual estimates of false-positive Brucella serology tests, mak- antibody prevalence were consistently less ing identification of Brucella-positive ani- than 2% in both areas. The rise in antibody mals difficult in livestock and wildlife prevalence prompted further investigation populations including cattle, bison, and to determine whether Y. enterocolitica O:9

537 538 JOURNAL OF WILDLIFE DISEASES, VOL. 48, NO. 3, JULY 2012 cross-reactivity on B. abortus serology culture-positive for B. abortus. Fecal samples testing was causing artificially high Brucella alone were collected from 113 free-ranging bison from the GYE, including 60 Yellowstone antibody prevalence estimates in elk. bison sampled in July and August 2008 and 53 Using Y. enterocolitica–specific, anti- Jackson-area bison sampled from near Grand body-based, Western immunoblot tests, Teton National Park (Moose, Wyoming, USA) Atkinson et al. (2007) found antibodies to in March through August 2008. Frozen tissue Y. enterocolitica in all of the Brucella and fecal samples from slaughtered bison were sent to the National Veterinary Science antibody-positive samples from the two Lab (NVSL; Ames, Iowa, USA) for Brucella areas. With the use of the Y. enterocolitica culture. After thawing, each sample was Western immunoblot, Brucella antibody individually homogenized, and aliquots of prevalence estimates were revised to each homogenate were smeared on Brucella- 1.93% in the Madison Elk Management selective media. Plates were incubated at 37 C % for up to 10 days. Species and biovar Unit and 0.0 in the Pioneer Mountains. identification was performed according to The authors concluded that most B. standard procedures (Alton et al., 1988) by abortus antibody–positive elk from the NVSL staff. initial Brucella serology (based on the Fecal samples collected from free-ranging O:9 antigen) could represent false-posi- bison were gathered within 20 min of defeca- tives caused by cross-reactivity between Y. tion after the individuals had moved away. Samples were immediately placed on wet ice enterocolitica serotype O:9 and B. abortus and transported to a 10 C refrigerator or O:9. However, we note that Western 220 C freezer within a few hours of collection. blotting is not a well-validated test, and Most samples were stored frozen at 220 C its sensitivity and specificity have not been (for ,2 mo) before processing for cultivation, evaluated against the gold standard: culti- with the exception of 28 samples that were held at 10 C and never frozen before vation of Y. enterocolitica. cultivation. Earlier research with meat samples False-positive brucellosis tests as report- (De Zutter et al., 1994), and our own ed by Atkinson et al. (2007) could have experiments holding feces spiked with Y. severe social and political repercussions if enterocolitica at 220 C before cultivation the same error rate were discovered for (data not shown), showed that storage at 10 C or 220 C for ,2 mo did not affect the Yellowstone bison, where antibody-positive cultivation results. One hundred and thirty- individuals are sometimes culled when eight fecal samples were sent to Wyoming captured at the park boundary and anti- Game and Fish Department Wildlife Disease body-negative animals are afforded release Laboratory (Laramie, Wyoming, USA), and 28 back onto the conservation areas. Because samples were sent to Colorado State Univer- sity Veterinary Diagnostic Laboratory (Fort the incidence of bison infected with B. Collins, Colorado, USA). abortus in the GYE has fluctuated between 40% and 60% during the past 20 yr Enrichment and culture to detect Yersinia (Treanor et al., 2007), we investigated To enrich for cultivation of Y. enterocolitica, whether seemingly antibody-positive bison 1 g of feces was placed in 9 ml of 0.1% were actually infected with Y. enterocolitica peptone water and homogenized in a Stoma- O:9, producing false-positive reactions in cherH laboratory blender (Metrohm USA, B. abortus serologic assays. Riverview, Florida, USA) for 30 sec, as described by Bhaduri et al. (2005). One milliliter of the homogenized mixture was MATERIALS AND METHODS added to 9 ml of Irgasan broth, mixed by inverting, and incubated at room temperature Sample collection for 48 hr, as previously described (De Zutter Fecal and tissue (lymph node) samples et al., 1994). After incubation in enrichment were collected from 53 GYE bison during broth, tubes were inverted several times and slaughter (February through April 2008) centrifuged for 1 min at 430 3 G to remove at private livestock facilities. Thirty-five of large particulates. The remaining suspen- these (66%) had detectable antibodies for sion was then plated on Cefsulodin-Irgasan brucellosis. Seven of the 35 (20%) were also Novobiocin/Yersinia-selective agar (CIN/YSA) WADE ET AL.—YERSINIA ENTEROCOLITICA AND FALSE-POSITIVE TESTS FOR BRUCELLOSIS 539 plates, as described by Bhaduri (2005), with RESULTS AND DISCUSSION the addition of 0.028 g of CaCl2 per liter of agar medium as described by Li et al. (1998). When staff from Montana Fish, Wild- The plates were incubated for 24 hr at 28 C, life, and Parks reported on possible false- then checked for colony growth. Cultures were restreaked on CIN/YSA agar if individual positive brucellosis tests in several of their colonies were not obtained on the original southern elk herds assumed to be caused plates. by Y. enterocolitica O:9, we questioned Colonies were identified as putative Y. whether the same phenomenon could be enterocolitica based on morphology, with occurring in GYE bison. In 2008, we colonies of 1–2 mm diameter, having a deep- red center surrounded by a clear, colorless performed cultivation assays on 166 fecal edge scored as positive as described by samples, including from 35 slaughtered Weagant and Feng (2007). Single colonies Yellowstone bison that were antibody- demonstrating these characteristics were se- positive for brucellosis. Seven of these 35 lected for DNA extraction, followed by PCR antibody-positive animals (20%) were also analysis for confirmation of identity. culture-positive for B. abortus. Molecular methods for Y. enterocolitica Yersinia enterocolitica O:9 was not confirmation isolated from any of the 166 Yellowstone DNA was extracted from the isolated and Jackson bison sampled in winter and bacterial colonies described above using the summer. This suggests that Y. enterocoli- QIAampH DNA Mini Kit (Qiagen, Valencia, tica O:9 cross-reactivity with B. abortus California, USA) according to the manufac- antigens was unlikely to cause false-positive turer’s instructions. We conducted multiplex serology tests in GYE bison during this PCR to confirm the preliminary identification of Y. enterocolitica from cultures in 20-ml study, and that Y. enterocolitica prevalence reaction mixtures consisting of 13 Promega was apparently low in bison from this area. Master Mix (Promega Corporation, Madison, Possible explanations for this observation Wisconsin, USA), 2.5 ml of DNA template, and include that some bacterial infections likely m 0.25 M each of forward and reverse primer. infect primarily cervids and not bovids or The primers were designed to target four separate diagnostic gene fragments for the ail, vice versa. It is also possible that elk and yst, and virF genes (356, 134, and 231 base bison immune systems have different pairs [bp], respectively), as previously de- efficiencies for clearing Yersinia. Finally, scribed (Harnett et al., 1996), and a 312-bp elk more often encounter livestock outside fragment of the perosamine synthetase gene the park and could be occasionally infected (per) as described by Lu¨ beck et al. (2003). Thermalcyclerconditionsforrespective or reinfected from livestock, whereas bison primers were identical to those in Harnett are not allowed to leave the confines of the et al. (1996) and Lubeck et al. (2003). park to overlap with cattle or other Amplified products were separated on a livestock. Ultimately, the explanation for 1.5% agarose gel in 13 Tris-Borate–ethylene- this observation can only be resolved diaminetetraacetic acid buffer. The amplified bands were visualized in the presence of through additional research into this phe- ethidium bromide under ultraviolet transillu- nomenon. mination and digitally captured using Quantity Two isolates of Y. enterocolitica biotype OneH software (BioRad, Hercules, California, O:2,3 were identified by the Centers for USA). The size and number of amplified Disease Control, but those biotypes are not fragments were used to determine the presence or absence of Y. enterocolitica as known to cross-react with B. abortus described by Harnett et al. (1996) and Lubeck serology tests because of a lack of the O- et al. (2003). Isolates identified as Y. enter- antigen structure encoded by the perosa- ocolitica by PCR were submitted for confir- mine synthetase gene. Both of these isolates mation via serotyping to the Division of were obtained from samples collected at Foodborne, Bacterial, and Mycotic Diseases at the National Center for Zoonotic, Vector- Antelope Flats in Grand Teton National Borne, and Enteric Diseases, Centers for Park. Yersinia enterocolitica O:9 has been Disease Control (Atlanta, Georgia, USA). showntobeshedathighlevelsfromcattle 540 JOURNAL OF WILDLIFE DISEASES, VOL. 48, NO. 3, JULY 2012 during the first month following infection, outbreaks or epizootics. Quantitative PCR- with continued shedding for months until based tests should provide more sensitivity slaughter (Garin-Bastuji, 1999). The ab- for the reliable detection of low numbers of sence of Y. enterocolitica O:9 in feces from Y. enterocolitica in intestinal tissue, gut- GYE bison samples indicated that the bison associated lymphoid tissue, or feces. Al- sampled were not infected at a detectable though one disadvantage of DNA-based level with Y. enterocolitica O:9 and that PCR assays is that they might detect DNA cross-reactivity by Y. enterocolitica O:9 on from nonviable pathogens, positive results B. abortus serology tests should not be of in these tests would still provide clues as to concern to bison managers in the GYE. whether Y. enterocolitica O:9 is present in Any decision to remove animals from GYE wildlife. the population based on brucellosis serol- Given the recent high incidence of ogy results is controversial. It is, therefore, apparently false-positive brucellosis tests important for wildlife managers to under- in elk populations on the northwest side of stand the probability of false-positive data, the GYE, careful distinction between true and the implications of the negative Y. brucellosis infection and false-positive bru- enterocolitica O:9 culture results reported cellosis tests due to Y. enterocolitica O:9 is herein are important to the management important. In summary, our study indicat- and conservation of bison. There remains ed that Y. enterocolitica O:9 was absent or a possibility that, although Y. enterocoli- at low prevalence in the GYE and that tica O:9 was not cultured in the fecal cross-reactivity with B. abortus antigens samples we collected, the organism may was unlikely to have caused false-positive still be present in the ecosystem. More serology tests in Yellowstone and Jackson detailed sampling and testing of intestinal bison populations during this study. tissues and perhaps gut-associated lymphoid tissue would be needed to com- ACKNOWLEDGMENTS pletely rule out the presence of this We thank Flo Gardipee, Doug Blanton, organism in these populations of bison. Jenny Jones, and Chris Geremia for assistance Another important implication of these in collecting, storing, and processing field samples. Also thanks to Irene Wesley, USDA, findings is in the management of herd and Ames, Iowa, USA, and Nancy Strockbine, human health. Yersinia enterocolitica has Centers for Disease Control and Prevention, increasingly been implicated in human Atlanta, Georgia, USA, for control DNA and diarrheal disease and may approach the biotype analysis, respectively. W.S. was sup- level of Salmonella spp. infection in north- ported by the Montana Integrative Learning Experience for Students (MILES) program, ern climates (Burnens et al., 1996). Al- through Grant 52005905 to the University of though most cases of yersinosis are linked Montana from the Howard Hughes Medical to pork (Ostroff et al., 1995), the substantial Institute—Undergraduate Science Education rate of seroconversion in Montana elk Program. G.L. was partially supported by grants raises questions about its prevalence in from the U.S. National Science Foundation (NSF; grant DEB 074218), the Walton Family Montana wildlife and the potential transfer Foundation, the Luso-American Foundation, to humans through contact, such as in the the Portuguese Science and Technology Foun- hunting and processing of elk. dation (Grant SFRH/BD30737/2006; PTDC/ We recommend that future studies CVT/69438/2006 and CIBIO-UP. M.D.K. was include testing for Y. enterocolitica in supported by NSF Integrative Graduate Edu- cation and Research Traineeship grant 0504628. different seasons, during the course of multiple years and to use both cultivation LITERATURE CITED and quantitative PCR to complement ALTON, G., L. JONES,R.ANGUS, AND J. VERGER. 1988. serology and Western blots. Testing bison Techniques for the brucellosis laboratory. In- in different seasons and years might detect stitut National de la Recherche Agronomique, Y. enterocolitica O:9 appearing only during Paris, France, 190 pp. WADE ET AL.—YERSINIA ENTEROCOLITICA AND FALSE-POSITIVE TESTS FOR BRUCELLOSIS 541

ATKINSON, M., N. ANDERSON,K.AUNE, AND R. using the multiplex polymerase chain reaction. CLARKE. 2007. The epidemiological challenge of Epidemiology and Infection 117: 59–67. evaluating brucellosis seroprevalence in free- LI, L., S. BHADURI, AND W. E. MAGEE. 1998. Maximiz- ranging Rocky Mountain elk in Montana. 2007. ing plasmid stability and production of released In Proceedings of the 110 annual meeting of the proteins in Yersinia enterocolitica. Applied and United States Animal Health Association, Min- Environmental Microbiology 64: 1812–1815. neapolis, Minnesota, 12–18 October 2006, LU¨ BECK, P. S., M. SKURNIK,P.AHRENS, AND J. pp. 200–202; United States Animal Health HOORFAR. 2003. A multiplex PCR-detection assay Association, St. Joseph, Missouri, 756 pp. for Yersinia enterocolitica serotype O:9 and BEJA-PEREIRA, A., B. BRICKER,S.CHEN,C.ALMENDRA, Brucella spp. based on perosamine synthetase P. J. WHITE, AND G. LUIKART. 2009. DNA genes: The genus Yersinia. Kluwer Academic/ genotyping suggests recent brucellosis outbreaks Plenum Publishers, New York, pp. 451–454. near Yellowstone National Park originate from MUNOZ, P. M., C. M. MARIN,D.MONREAL,D. elk. Journal of Wildlife Diseases 45: 1174–1177. GONZALEZ,B.GARIN-BASTUJI,R.DIAZ,R.C. AINAR AIME ORIYON AND LASCO BHADURI, S., I. WESLEY, AND E. BUSH. 2005. M -J ,I.M , J. M. B . Prevalence of pathogenic Yersinia enterocolitica 2005. Efficacy of several serological tests and strains in pigs in the United States. Applied and antigens for diagnosis of bovine brucellosis in the Environmental Microbiology 71: 7117–7121. presence of false-positive serological results due to Yersinia enterocolitica O:9. Clinical and BOTTONE, E. J. 1997. Yersinia Enterocolitica: The charisma continues. Clinical Microbiology Re- Diagnostic Laboratory Immunology 12: 141–51. OSTROFF, S. 1995. Yersinia as an emerging infection: views 10: 257–276. Epidemiologic aspects of yersiniosis. Contribu- BURNENS, A. P., A. FREY, AND J. NICOLET. 1996. tions to Microbiology and Immunology 13: 5–10. Association between clinical presentation, TREANOR, J., R. WALLEN,G.PLUMB,J.JOHNSON,S. biogroups and virulence attributes of Yersinia CILLES,P.CROWLEY, AND D. MAEHR. 2007. enterocolitica strains in human diarrhoeal dis- Brucellosis in Yellowstone bison: An individual- ease. Epidemiology and Infection 116: 27–34. based simulation model of vaccination strategies. DE ZUTTER, L., L. LE MORT,M.JANSSENS, AND G. In Proceedings of the 110 annual meeting of the WAUTERS. 1994. Shortcomings of irgasan ticar- United States Animal Health Association, Min- cillin chlorate broth for the enrichment of neapolis, Minnesota, 12–18 October 2006, pp. Yersinia enterocolitica biotype 2, serotype 9 192–195; United States Animal Health Associa- from meat. International Journal of Food tion, St. Joseph, Missouri, 756 pp. Microbiology 23: 231–237. WEAGANT, S. D., AND P. FENG. 2007. Yersinia entero- GARIN-BASTUJI, B., N. HUMMEL,G.GERBIER,C.CAU, colitica. Chapter 8 in Bacteriological analytical R. POUILLOT,M.DA COSTA, AND J. FONTAINE. manual. 8th Edition. United States Food and 1999. Non specific serological reactions in the Drug Administration, Center for Food Safety diagnosis of bovine brucellosis: Experimental and Applied Nutrition, Silver Spring, Maryland, oral infection of cattle with repeated doses of http://www.fda.gov/Food/ScienceResearch/ Yersinia enterocolitica O:9. Veterinary Microbi- LaboratoryMethods/BacteriologicalAnalytical ology 66: 222–233. ManualBAM/default.htm. Accessed December GODFROID, J., C. SAEGERMAN,V.WELEMANS,K. 2011. WALRAVENS,J.J.LETESSON,A.TIBOR,A.MC WEYNANTS, V., A. TIBOR,P.A.DENOEL,C.SAEGERMAN, MILLAN,S.SPENCER,M.SANNA,D.BAKKER,R. J. GODFROID,P.THIANGE, AND J. J. LETESSON. POUILLOT, AND B. GARIN-BASTUJI. 2002. How to 1996. Infection of cattle with Yersinia enteroco- substantiate eradication of bovine brucellosis litica O:9: A cause of false positive serological when aspecific serological reactions occur in the reaction in bovine brucellosis diagnostic tests. course of brucellosis testing. Veterinary Micro- Veterinary Microbiology 48: 101–112. biology 90: 461–477. HARNETT, N., Y. P. LIN, AND C. KRISHNAN. 1996. Submitted for publication 12 January 2011. Detection of pathogenic Yersinia enterocolitica Accepted 30 January 2012.