1228

Journal of Food Protection, Vol. 71, No. 6, 2008, Pages 1228–1231 Copyright ᮊ, International Association for Food Protection

Research Note Isolation of Campylobacter jejuni and Coliform Bacilli from Bile and Liver Obtained from Slaughter in Western Japan

NORIKO MATSUMOTO,1 TAE TANIWAKI,1 MINAE KINUTA,1 AND TOSHIYUKI MURASE2,3*

1Division of Health Science, The Public Health Institute of Kochi Prefecture, 2-4-1, Marunouchi, Kochi 780-0850, Japan; and 2Laboratory of Veterinary Microbiology and 3Avian Zoonosis Research Center, Faculty of , Tottori University, 4-101, Koyama, Tottori 680-8553, Japan Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1228/1682087/0362-028x-71_6_1228.pdf by guest on 27 September 2021

MS 07-545: Received 9 October 2007/Accepted 14 November 2007

ABSTRACT

A total of 290 bile samples from 143 Japanese Brown, 97 , and 50 Holstein cattle, and a total of 148 liver samples from 81 Japanese Brown, 49 Japanese Black, and 18 Holstein cattle were examined for the presence of Cam- pylobacter jejuni by direct plating. The bile samples were also subjected to enumeration of coliform bacilli. Sixty-eight (23%) bile samples and 2 (1.4%) liver samples were positive for C. jejuni. A significantly higher isolation rate was observed from bile samples from Holstein (50%) than from Japanese Black (22%) and Japanese Brown (15%) cattle. C. jejuni was isolated from 52 of 232 bile samples that contained Ͻ30 CFU/ml (under the detection threshold) of coliform bacilli. The presence of C. jejuni from bile was observed throughout the year. Fifty-four of the 68 bile isolates were serologically typed into eight groups. Serotypes O:4 complex (28 isolates) and O:2 (11 isolates), which were commonly isolated from human patients in Japan, accounted for 57% of the isolates. These observations suggest that bile can be a cause of contamination with C. jejuni even though it contains only a low number of coliforms.

Campylobacter jejuni is one of the major enteric hu- consumption should be negative for fecal coliforms and man pathogens. Although the bacterium has been recog- Salmonella, but Campylobacter species were not included. nized to cause foodborne disease that is epidemiologically In the present study, the authors monitored bovine liver and linked to the consumption of poultry products (3, 16, 17), bile samples obtained from slaughterhouses in western Ja- the association of campylobacteriosis in humans with the pan for Campylobacter and fecal coliform contamination to large number of campylobacters being shed in feces of underscore the importance of testing for Campylobacter or- feedlot cattle is a source of concern (9). Recent studies ganisms. suggest that cattle can be considered as a possible source of C. jejuni (1, 10, 20). There have been reports of bovine MATERIALS AND METHODS enteritis (19) and abortions (22) due to C. jejuni, although Bile and liver samples. A total of 290 bile samples from this bacterium is generally present in cattle without any 143 Japanese Brown, 97 Japanese Black, and 50 Holstein cattle symptoms of diseases (6, 7). Saito et al. (20) demonstrated and a total of 148 liver samples from 81 Japanese Brown, 49 that identical serotypes and pulsed-field gel electrophore- Japanese Black, and 18 Holstein cattle were obtained at two sis–based genotypes were found in C. jejuni isolates ob- slaughter houses in western Japan between June 2004 and Decem- tained from human patients and bovine samples in northern ber 2006. Japan. It was reported that substantial numbers of Cam- Bacteriological examination. A 10-ml sample of bile was pylobacter were present in chicken (4) and sheep (23) liver, aseptically aspirated from each gall bladder with a syringe and although further investigations need to evaluate the asso- subjected to bacteriological examination as described below. Sur- ciation of this organism in the livers of those animals with faces of posterior hepatic segments were disinfected with a cotton human campylobacteriosis. Eating raw internal organs such swab immersed in 70% ethanol, and 20 g of the inner part of the as bovine liver may cause foodborne infection due to Cam- tissue were aseptically weighed into a sterile stomacher bag. Tryp- pylobacter (2) and dishes with bovine liver and intestine ticase soy broth (180 ml) (Becton Dickinson, Sparks, Md.) was added, and the contents were emulsified for 60 s using a Stom- are commonly served in certain restaurants in Japan (20). acher 400LAB BLENDER (Seward Medical, London, UK). A Therefore, contamination of bovine meat and liver with this loopful of the bile samples and a loopful of the rinse fluid from organism is of considerable public health concern. The each of the liver samples were streaked onto Columbia blood agar Ministry of Health, Labour and Welfare, Japan (former plates supplemented with 5% horse blood, 250 IU/ml bacitracin, Ministry of Health and Welfare) issued a notification in 50 ␮g/ml cycloheximide, 10 IU/ml colistin, 15 ␮g/ml cefazolin 1998 that bovine raw liver and meat destined for human sodium salt, and 5 ␮g/ml novobiocin (Campylobacter selective supplement [code no. SR0085]; Oxoid Ltd., Basingstoke, UK), * Author for correspondence. Tel: 81-857-31-5430; Fax: 81-857-31-5430; and incubated under microaerobic atmosphere at 42ЊC for 48 h. E-mail: [email protected]. Suspect colonies were identified as Campylobacter species by J. Food Prot., Vol. 71, No. 6 ISOLATION OF C. JEJUNI FROM BOVINE BILE 1229

TABLE 1. Results of isolation of C. jejuni from bile samples with TABLE 2. Association between C. jejuni from bile samples from different cattle breeds and serotyping slaughter cattle and the slaughter age or sex of the cattle Breed: Slaughter age (mo)a:

Japanese Japanese Breed Sex 18–24 25–36 37–72 73–198 Property Black Brown Holstein Japanese Black Steer 0/3 11/56 No. of samples tested 97 143 50 Female 0/1 10/35 0/2 No. (%) of samples positive 21 22 25 Japanese Brown Steer 0/1 11/86 0/1 for C. jejuni (22) (15) (50a) Female 1/1 8/44 0/2 2/8 Serotype Holstein Steer 22/39 Female 1/1 2/2 0/2 1/6 1,44 2 2461a Values are number of C. jejuni–positive samples/number of bile 34samples tested. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1228/1682087/0362-028x-71_6_1228.pdf by guest on 27 September 2021 4,13,16,43,50 12 6 10 51 19 1 C. jejuni isolates, followed by O:2 (11 of 68; 16%), where- 23,26,53 2 2 as 13 isolates were untypeable. Of the 148 liver samples, 27 1 1 1 only 2 samples, from a Japanese Black and a Japanese Untypeable 3 3 7 Brown, were positive for the organism. These isolates be- Not tested 1 longed to O:2 serotype. Serotype O:2 isolate was simulta- a The isolation rate in Holstein cattle (50%) was significantly (P neously recovered from one of two bile sample from the Ͻ 0.01) higher than that of either Japanese Black (22%) or Jap- two individuals with positive liver samples. anese Brown (15%) cattle. Isolation of C. jejuni and age of cattle from which bile samples were obtained is shown in Table 2. Japanese Black and Japanese Brown cattle are and are usually characteristic Gram staining morphology, the positive oxidase re- slaughtered approximately 32 to 34 months of age in steers action, hippurate hydrolysis, and susceptibility to nalidixic acid and 34 to 36 months of age in females. Older animals, Ͼ36 and cephalothin performed with disk susceptibility tests (SN Sen- months of age, used for breeding (reproduction) are occa- sidisk, Nissui Pharmaceutical, Tokyo, Japan). Species identifica- sionally slaughtered as beef cattle. Differences in the iso- tion was carried out with the apiCampy system (bioMe´rieux, Mar- lation rates among age of those cattle were not obvious. cy-l’Etoile, France). The C. jejuni isolates were serotyped using a commercially available kit (Denka-Seiken, Tokyo, Japan) ac- Holstein steers are usually reared as beef cattle and pro- cording to the scheme described by Penner and Hennessy (18). cessed in slaughterhouses approximately 24 to 26 months This kit contains 21 individual O sera and four types of combi- of age in Japan. Because the number of the older animals nation sera, e.g., O:1,44; O:4,13,16,43,50 (O:4 complex); O:6,7; was small, differences in the isolation rates due to age were and O:23,36,53. not analyzed using statistical analyses. The isolation rates of C. jejuni in bile samples were Enumeration of coliform bacilli in bile samples. To deter- not correlated with the number of coliform bacilli in the mine the number of coliform bacilli, each bile sample was serially samples from all the breeds (Table 3). C. jejuni was isolated diluted 10-fold in saline, and 1-ml aliquots of each dilution were mixed with molten desoxycholate agar (Kanto Chemical, Tokyo, from bile samples from Japanese Black (19%), Japanese Japan) plates. Plates yielding 30 to 300 colonies were used for Brown (13%), and Holstein (55%) cattle, respectively, that the estimation of viable cell counts of coliform bacilli. When Ͻ30 contained Ͻ30 CFU/ml (under the detection threshold) of colonies were observed in plates that were inoculated with undi- coliform bacilli. Except for bile samples with this level of luted bile samples, the number of coliform bacilli in those samples were designated as Ͻ30 CFU/ml (under the detection threshold). TABLE 3. Association between the level of C. jejuni and coliform Statistical analysis. Fisher’s exact test was used to determine bacilli from bile samples correlation of the overall frequency of Campylobacter isolation No. of coliform bacilli No. of samples positive with the number of coliform bacilli in the samples. Differences in in the bile sample No. of samplesa for C. jejunia the isolation frequency were also determined among samples from different breeds. Differences were considered significant at P Ͻ Ͻ30 232 (79/113/40) 52 (15/15/22) 0.05. 30–102 2 (1/1/0) 0 (0/0/0) 102–103 1 (1/0/0) 1 (1/0/0) RESULTS 103–104 2 (0/1/1) 1 (0/0/1) C. jejuni was isolated from 23% (68 of 290) of the bile 104–105 6 (1/3/2) 2 (1/1/0) samples. The isolation rate in Holstein cattle (50%) was 105–106 8 (2/4/2) 3 (1/2/0) 6 7 significantly (P Ͻ 0.01) higher than either Japanese Black 10 –10 25 (7/15/3) 8 (2/4/2) 7 8 (22%) or Japanese Brown (15%) cattle (Table 1). C. jejuni 10 –10 13 (6/5/2) 1 (1/0/0) Ͼ108 1 (0/1/0) 0 (0/0/0) isolates obtained from bile samples were divided into eight serotypes (Table 1). Twenty-eight isolates of O:4,13,16, a Numbers in parentheses are the number of bile samples from 43,50 (O:4 complex) accounted for 41% of the total of 68 Japanese Black/Japanese Brown/Holstein cattle. 1230 MATSUMOTO ET AL. J. Food Prot., Vol. 71, No. 6

Isolation rates from bile samples varied according to cattle breeds. Although the exact reason(s) are unknown, it is likely to be associated with the differences in age and sex in each of the cattle breeds that came to the slaughter- houses. Such factors may change the combination of con- stituents of bile, resulting in the differences in isolation rates from bile samples. There was variation in the preva- lence of C. jejuni in bile samples with peaks in March, July, and November. Fecal shedding of Campylobacter species with peaks in the spring and autumn has been observed in dairy cattle (21). Stanley et al. (21) considered that changes in shedding patterns may reflect hormonal disturbances in

the gut flora because the spring and autumn roughly cor- Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1228/1682087/0362-028x-71_6_1228.pdf by guest on 27 September 2021 relate with traditional milk flushes and period of calving, although these peaks were absent in beef cattle. Indirect factors, such as migratory birds, rodents, and insects (11), FIGURE 1. Monthly isolation rates of C. jejuni from bile samples, which are dependent on changes in temperature and hours obtained during the period of June 2004 to December 2006, which of sunlight due to seasonal variation, may contribute to en- were contaminated with Ͼ100 CFU/ml coliforms (solid columns) vironmental contamination of beef cattle farms with Cam- Ͻ and 30 CFU/ml (open columns). The figures given under each pylobacter species. month indicate the number of bile samples tested. It is of interest that C. jejuni was isolated from bile samples irrespective of levels of contamination with coli- coliform bacilli, the number of coliform bacilli in C. jejuni– forms. To our knowledge, it is the first report that the con- positive samples ranged from 4.2 ϫ 102 (CFU/ml) to 1.1 tamination of bile from cattle with C. jejuni was simulta- ϫ 107 in Japanese Black, from 4.8 ϫ 104 to 6.4 ϫ 106 in neously monitored with coliforms. Markedly, overall 22% Japanese Brown, and from 1.0 ϫ 103 to 4.7 ϫ 106 in Hol- of the bile samples (52 of 232) with Ͻ30 CFU/ml coliforms stein cattle. were positive for C. jejuni, suggesting that testing of cattle C. jejuni were isolated from bile samples throughout products for coliforms is not always adequate to warrant the year, although the frequency was Ͻ0.2% in early and the quality of the products. Moreover, contamination of bile late summer (Fig. 1). From February to June, C. jejuni was samples with C. jejuni was found throughout the year, al- exclusively isolated from bile samples that were apparently though the monthly isolation rates varied in this study. Be- negative for coliform bacilli. cause bile in cattle is considered to be an important con- tamination source of Campylobacter species in processing DISCUSSION plants (5), further investigations, such as typing studies of C. jejuni was isolated from a total of 68 bile samples isolates from bovine bile and human cases that evaluate from 290 slaughter cattle despite a low frequency of iso- etiological significance and a risk for human infection, are lation from liver samples of the same animals. Previous needed. reports (5, 7) also demonstrated that C. jejuni had been isolated from substantial numbers of bile or gallbladder ACKNOWLEDGMENT samples obtained from apparently healthy cattle. In several We are grateful to Dr. Peter S. Holt for reviewing this manuscript. studies (1, 5, 7), the rate of samples positive for C. jejuni in liver from cattle was lower than that in gallbladder or REFERENCES bile. Hugdahl et al. (8) determined the chemotactic behavior 1. Ac¸ik, M. N., and B. C¸ etinkaya. 2005. The heterogeneity of Cam- of C. jejuni and revealed that the soluble mucin in bile is pylobacter jejuni and Campylobacter coli strains isolated from a significant factor in the attraction of this organism to bile. healthy cattle. Lett. Appl. Microbiol. 41:397–403. Moreover, tolerance of C. jejuni to the bactericidal effects 2. Anonymous. 2006. Campylobacter enteritis in Japan, 1999–2005. Infect. Agents Surveill. Rep. 27:167–168. of bile has recently been demonstrated (6, 15). 3. Cogan, T. A., S. F. Bloomfield, and T. J. Humphrey. 1999. The ef- Serotypes O:4 complex and O:2 accounted for 57% of fectiveness of hygiene procedures for prevention of cross-contami- C. jejuni isolates obtained in this study. Saito et al. (20) nation from chicken carcasses in the domestic kitchen. Lett. Appl. reported that these serotypes were predominant in the iso- Microbiol. 29:354–358. lates from not only bovine bile and feces, but also retail 4. Cornelius, A. J., C. Nicol., and J. A. Hudson. 2005. Campylobacter spp. in New Zealand raw sheep liver and human campylobacteriosis poultry and humans with sporadic diarrhea in Japan. Sim- cases. Int. J. Food Microbiol. 99:99–105. ilar distribution of serotypes in chicken, bovine, and human 5. Enokimoto, M., M. Kubo, Y. Bozono, Y. Mieno, and N. Misawa. isolates was demonstrated by the other investigations con- 2007. Enumeration and identification of Campylobacter species in ducted in Denmark (14) and Canada (13). However, Kramer the liver and bile of slaughtered cattle. Int. J. Food Microbiol. 118: et al. (12) examined the serotypes and phage types of C. 259–263. 6. Fox, E. M., M. Raftery, A. Goodchild, and G. L. Mendz. 2007. jejuni isolates in the United Kingdom and found that a sig- Campylobacter jejuni response to ox-bile stress. FEMS Immunol. nificant portion of the chicken and lamb isolates shared Med. Microbiol. 49:165–172. identical subtypes with the human isolates. 7. Garcia, M. M., H. Lior, R. B. Stewart, G. M. Ruckerbauer, J. R. R. J. Food Prot., Vol. 71, No. 6 ISOLATION OF C. JEJUNI FROM BOVINE BILE 1231

Trudel, and A. Skljarevski. 1985. Isolation, characterization, and se- study of bile responses in Campylobacterales. FEMS Immunol. Med. rotyping of Campylobacter jejuni and Campylobacter coli from Microbiol. 49:101–123. slaughter cattle. Appl. Environ. Microbiol. 49:667–672. 16. Ono, K., and K. Yamamoto. 1999. Contamination of meat with Cam- 8. Hugdahl, M. B., J. T. Beery, and M. P. Doyle. 1988. Chemotactic pylobacter jejuni in Saitama, Japan. Int. J. Food Microbiol. 47:211– behavior of Campylobacter jejuni. Infect. Immun. 56:1560–1566. 219. 9. Inglis, G. D., L. D. Kalischuk, and H. W. Busz. 2004. Chronic shed- 17. Park, S. F. 2002. The physiology of Campylobacter species and its ding of Campylobacter species in beef cattle. J. Appl. Microbiol. 97: relevance to their role as foodborne pathogens. Int. J. Food Micro- 410–420. biol. 74:177–188. 10. Ishihara, K., T. Yamamoto, S. Satake, S. Stakayama, S. Kubota, H. 18. Penner, J. L., and J. N. Hennessy. 1980. Pasive hemagglutination Negishi, A. Kojima, T. Asai, T. Sawada, T. Takahashi, and Y. Tamura. technique for serotyping Campylobacter fetus subsp. jejuni on the 2006. Comparison of Campylobacter isolated from humans and basis of soluble heat-stable antigens. J. Clin. Microbiol. 12:732–737. 19. Prescott, J. F., and D. L. Munroe. 1982. Campylobacter jejuni en- food-producing animals in Japan. J. Appl. Microbiol. 100:153–160. teritis in man and domestic animals. J. Am. Med. Assoc. 181:1524– 11. Jacobs-Reitsma, W. F., N. M. Bolder, and R. W. A. W. Mulder. 1994. 1530. Cecal carriage of Campylobacter and Salmonella in Dutch broiler 20. Saito, S., J. Yatsuyanagi, S. Harata, Y. Ito, K. Shinagawa, N. Suzuki, flocks at slaughter; a one year study. Poult. Sci. 73:1260–1266. K. Amano, and K. Enomoto. 2005. Campylobacter jejuni isolated Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1228/1682087/0362-028x-71_6_1228.pdf by guest on 27 September 2021 12. Kramer, J. M., J. A. Frost, F. J. Bolton, and D. R. A. Wareing. 2000. from retail poultry meat, bovine feces and bile, and human diarrheal Campylobacter contamination of raw meat and poultry at retail sale: samples in Japan: comparison of serotypes and genotypes. FEMS identification of multiple types and comparison with isolates from Immunol. Med. Microbiol. 4:311–319. human infection. J. Food Prot. 63:1654–1659. 21. Stanley, K. N., J. S. Wallace, J. E. Currie, P. J. Diggle, and K. Jones. 13. Munroe, D. L., J. F. Prescott, and J. L. Penner. 1983. Campylobacter 1998. The seasonal variation of thermophilic campylobacters in beef jejuni and Campylobacter coli serotypes isolated from chickens, cat- cattle, dairy cattle and calves. J. Appl. Microbiol. 85:472–480. tle, and pigs. J. Clin. Microbiol. 18:877–881. 22. Welsh, R. D. 1984. Campylobacter jejuni abortion in a heifer. J. Am. 14. Nielsen, E. M. 2002. Occurrence and strain diversity of thermophilic Med. Assoc. 185:549–551. campylobacters in cattle of different age groups in dairy herds. Lett. 23. Whyte, R., J. A. Hudson, and C. Graham. 2006. Campylobacter in Appl. Microbiol. 35:85–89. chicken livers and their destruction by pan frying. Lett. Appl. Mi- 15. Okoli, A. S., T. Wadstrom, and G. L. Mendz. 2007. Bioinformatic crobiol. 43:591–595.