Prevalence and Antimicrobial Resistance of Campylobacter Jejuni and Campylobacter Coli Isolated from Raw Chicken Meat and Human Stools in Korea

2915

Journal of Food Protection, Vol. 69, No. 12, 2006, Pages 2915–2923 Copyright , International Association for Food Protection

Prevalence and Antimicrobial Resistance of Campylobacter jejuni and Campylobacter coli Isolated from Raw Chicken Meat and Human Stools in Korea

YUN-SOOK KANG,1,2 YONG-SUN CHO,3 SUN-KYUNG YOON,1 MYEONG-AE YU,4 CHANG-MIN KIM,5 JONG-OK LEE,2 AND YU-RYANG PYUN1*

1Department of Biotechnology, Yonsei University, Seoul, 120-749, Korea; 2Korea Food and Drug Administration Seoul, 122-704, Korea;

3Food Analysis Research Team, Korea Food Research Institute, Gyunggi-do, 463-746, Korea; 4International Life Sciences Institute of Korea, Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021 Seoul, 150-889, Korea; and 5Dongwon F&B Food R&D Center, Gyunggi-do, 462-120, Korea

MS 05-524: Received 17 October 2005/Accepted 9 May 2006

ABSTRACT

Prevalence of Campylobacter in raw chicken meat and human stools and subsequent antibiotic resistance profiles of the pathogenic isolates obtained from 2000 through 2002 were investigated. Campylobacter jejuni and Campylobacter coli were isolated from 570 of the 923 raw chicken meat samples collected from traditional markets, large retail stores, or department stores in Korea, resulting in the isolation rate of 61.8%. A total of 579 Campylobacter isolates were obtained from raw chicken (36.3% for C. jejuni and 26.4% for C. coli) with the average population of 335.6 CFU/g. From 513 human stool samples, 15 isolates of Campylobacter were detected. Seasonal variation in the quantification of C. coli was not noticeable throughout the year, while the isolation rate of C. jejuni was the highest in September through October (840 CFU/g) followed by that of July through August and May through June in decreasing order, showing a significant seasonal effect (P Ͻ 0.05). Contamination of Campylobacter was more severe in raw chicken meat sold in traditional markets than in those sold in large retail stores and department stores. Prevalence of Campylobacter in raw chicken sold in traditional markets was significantly influenced by seasonal changes (P Ͻ 0.05), whereas the samples obtained from other places was less affected by the seasonal changes. Susceptibilities of the 594 chicken isolates to ciprofloxaxin, chloramphenicol, erythromycin, kanamycin, nalidixic acid, and tetracycline were determined by an E-test. Campylobacter isolates were the most resistant to nalidixic acid (91.4%) followed by ciprofloxaxin (87.9%), tetracycline (87.2%), kanamycin (30.6%), erythromycin (19.4%), and chloramphenicol (1.3%). Hu- man isolates showed a similar resistance to the six antibiotics tested. The proportion of Campylobacter isolates with multidrug resistance to four or more antimicrobials obtained from 2000 through 2002 ranged from 28 to 43.5%, indicating that it could be a serious health-threatening factor. This study suggests that it is prudent to establish an effective National Monitoring Program in Korea for the prevention and control of Campylobacter spp.

Campylobacteriosis characterized by diarrhea, fever, Campylobacter and Salmonella, major human pathogens and abdominal cramps is the most common bacterial gas- worldwide, share common food vehicles in sporadic human troenteritidis in humans and animals worldwide. Over 2.4 infection (3). Poultry meat, especially undercooked and million human infections per year have been estimated in cross-contaminated with ready-to-eat foods, has been the the United States alone (1). Although campylobacteriosis is major cause of Campylobacter and Salmonella infections the most common infectious disease in the United States (3). In The Netherlands, 61% of raw chicken products were and other regions of the world, it has been rarely reported contaminated with C. jejuni, which was due to cross-con- in Korea, possibly because of fastidious culture conditions tamination during food preparation via cutting boards, of the pathogen, which made it overlooked in the past. plates, and food handlers (5). Approximately 71% of raw Since foodborne campylobacteriosis was first reported in retail chickens sold in the United Kingdom were contami- 2001, however, a few outbreaks have been continuously nated with Campylobacter spp. while only 8% of the chick- occurring in Korea (14), which implies that preventative en meat harbored Salmonella (20). It was reported that of and control measures for the illness are required. 415 raw poultry meat samples sold in Switzerland, the path- The causative microorganism, Campylobacter, has ogens were detected from 91 samples with a contamination been the most common infectious agent transmitted to hu- rate of 21.9% (17). mans via raw meats and poultry (22). Campylobacter spe- Antibiotic susceptibility of Campylobacter spp. has cies of clinical importance include jejuni, coli, lari, fetus, been examined for effective clinical treatment of campy- and upsaliensis, among which C. jejuni and C. coli are lobacteriosis (1, 5, 6). Among the antibiotics, either a fluo- mainly implicated to foodborne disease of humans (21, 22). roquinolone, such as ciprofloxacin, or a macrolide is the drug of choice for the therapeutic treatment of Campylo- * Author for correspondence. Tel: ϩ82-2-2123-4098; Fax: ϩ82-2-362- bacter gastroenteritis (10). Fluoroquinolone antibiotics that 7265; E-mail: [email protected]. were first approved for the use in humans in 1986 in the 2916 KANG ET AL. J. Food Prot., Vol. 69, No. 12

United States were also approved for the use in poultry in posed of Bolton broth base (9.5 ml), 1 ml of fetal bovine serum 1995. Campylobacter isolates resistant to ﬂuoroquinolones (ﬁltered, 0.22 ␮M), and 1 ml of glycerol (10%). have been detected from both humans and poultry in most Samples. Raw chicken meat was purchased from small local regions of world and have been increasing in the regions markets, large retail stores, or department stores in Seoul, Daejon, where the antibiotics were abused (13). In recent years, ex- Busan, and Gwangju in Korea from March 2000 through Decem- tensive researches have been carried out to investigate the ber 2002. Human stool samples were obtained from patients with antibiotic resistance of Campylobacter, revealing informa- diarrhea who visited hospitals in Seoul and Jeolla Nam-do from tion that may be used to reduce the resistance of the path- January 2000 through December 2001. A total of 923 raw chicken ogen to therapeutic drugs (1, 2). meat and 513 human stool samples were collected and examined In this study, we investigated the prevalence of C. je- for the presence of Campylobacter species. Samples were stored Ͻ Њ juni and C. coli in raw chicken meat from 2000 through at refrigerated temperatures ( 5 C) until examination. Seasonal temperatures at the time of sample purchases were recorded for 2002, which can be used for estimating potential risk of

calculating the average temperature for each 2-month period. The Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021 foodborne campylobacteriosis in Korea. Human stool sam- average temperatures measured and calculated throughout the year ples were also monitored for the presence of Campylobac- were then used as the temperature of seasonal bimonthly periods. ter spp. in order to determine whether there might be any relationship between the food contamination level and clin- Strain isolation and identification. From each raw chicken ical implications. Among the food supply chains, raw sample, 25 g cut from the surface was added to 100 ml of selective chickens sold through traditional markets were expected to broth (e.g., Bolton broth) containing Campylobacter selective supplement (5% lysed horse blood and four antibiotics mentioned be highly contaminated. Therefore, prevalence of Campylo- above) and homogenized using a stomacher (model 400, Norfolk, bacter in raw chickens sold in three different retail places, UK). The mixture was preincubated at 35ЊC for 4 h under mi- including traditional markets, large retail stores, and de- croaerobic conditions and subsequently at 42ЊC for 24 to 48 h partment stores, was examined throughout different seasons under the same conditions. Human stool samples were subjected when temperature changes were distinct among other en- without further enrichment to the isolation procedure. vironmental factors. The results would contribute to the de- Enriched cultures were streaked onto Abeyta-Hunt agar con- velopment and enforcement of preventative and control taining antibiotic supplement and incubated microaerobically for measures that can effectively reduce the contamination of 24 to 48 h. Translucent white, moist, and glistening colonies, typ- raw chickens during distribution, storage, and sales. Anti- ical of Campylobacter, were selected and streaked onto Abeyta- biotic susceptibility and multidrug resistance of both chick- Hunt agar with antibiotic supplement. Campylobacter colonies were selected and used for further identification using biochemical en and clinical isolates were examined using six antibiotics tests. Gram staining, catalase and oxidase reactions, and hippurate that have been applied for food animals as well as for the tests were carried out for the identification as previously described clinical treatment of human infection in Korea. Results on (19). API Campy Kit (bioMe´rieux, Inc., Marcy l’Etoile, France) the antibiotic resistance profiles of Campylobacter isolates was used for final identification. may be used for the development of a database from which From each Campylobacter-positive raw chicken meat sam- information necessary for effective clinical treatment of the ple, 25 g was weighed, added to 225 ml of sterile PBS (pH 7.0), illness can be extracted during outbreaks of foodborne dis- and homogenized. After 10-fold serial dilutions, diluted samples ease. Antibiotic resistance profiles of food and clinical were streaked onto Abeyta-Hunt agar containing antibiotic sup- Њ Campylobacter isolates may also serve as a sound scientific plements and incubated microaerobically at 42 C. Typical Cam- pylobacter colonies were picked for further identification using basis of a comprehensive system for antibiotic manage- biochemical confirmative tests as described above. Final cell pop- ment. ulations of C. jejuni and C. coli in positive samples were calculated by multiplying dilution folds by average cell counts of each MATERIALS AND METHODS sample. For those samples collected in 2002, cell populations were Bacterial strains. C. jejuni subsp. jejuni ATCC 33560, C. determined for all the chicken samples tested. jejuni subsp. doylei ATCC 49349, and C. coli ATCC 33559 were Susceptibility test. Antibiotic resistances of Campylobacter purchased from American Type Culture Collection (ATCC) and isolates were determined using an E-test system (AB Biodisk, So- used as control strains for the study. Campylobacter strains were lana, Sweden) as described (19, 23). Campylobacter isolates were cultured as previously described (12). Briefly, lyophilized cells of cultivated on Abeyta-Hunt agar without antibiotic supplement as each strain were aseptically rehydrolyzed using 1.0 ml of sterile described above. A single typical colony was picked and sus- phosphate-buffered saline (PBS), transferred into Bolton broth pended in Mueller-Hinton broth (Difco) until the McFarland num- (Oxoid AM7625, Hampshire, UK) containing 5% lysed horse ber became 0.5 that corresponds to approximately 1 ϫ 108 to 4 blood and four antibiotics (i.e., sodium cefoperazone, trimetho- ϫ 108 CFU/ml (24). The cell suspension was streaked, using a prim lactate, vancomycin, and amphotericine B), and incubated sterile swab, onto Mueller-Hinton agar (Difco) containing 5% Њ microaerobically (5% O2, 10% CO2, 85% N2)at42C for 24 to horse blood. After the surfaces of the agar plates were completely 48 h. The cultures were streaked onto Abeyta-Hunt agar contain- dried, an E-test strip was placed onto the plates and then incubated ing 40 g of heart infusion agar (Difco, Becton Dickinson, Sparks, at 42ЊC for 24 h under microaerobic conditions. The E-test strips Md.),2gofyeast extract, and 1,000 ml of distilled water with used for the study included ciprofloxaxin (0.002 to 32 ␮g/ml), an antibiotic supplement for Campylobacter and incubated under chloramphenicol (0.016 to 256 ␮g/ml), erythromycin (0.016 to the same conditions as mentioned above to check the purity. Cam- 256 ␮g/ml), kanamycin (0.016 to 256 ␮g/ml), nalidixic acid pylobacter cultures were dispensed into 2.0-ml cryogenic vials (0.016 to 256 ␮g/ml), and tetracycline (0.016 to 256 ␮g/ml). The and stored at Ϫ70ЊC until use. The freezing medium was com- MIC of each antibiotic for Campylobacter was determined by J. Food Prot., Vol. 69, No. 12 RESISTANCE OF CAMPYLOBACTER SP. ISOLATED FROM MEAT AND STOOLS 2917

TABLE 1. Prevalence of Campylobacter jejuni and Campylobacter coli in raw chicken meat and human stool samples isolated from 2000 through 2002 in Korea C. jejuni C. coli No. of samples No. of contaminated with Isolation No. of Isolation No. of Isolation Samples samples tested Campylobactera rate (%) positive samples rate (%) positive samples rate (%)

Raw chicken meat 923 570b 61.8 335 36.3 244 26.4 Human stoolsc 513 15 2.9 13 2.5 2 0.4 a Campylobacter spp. include C. jejuni and C. coli exclusive of other Campylobacter species. b From 570 positive samples, both C. jejuni and C. coli were detected in nine chicken meat samples. c Human stool samples only were collected in 2000. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021 measuring the concentration in the elliptical zone around the strip. Statistical analysis. The enumeration data were analyzed sta- If cells did not grow sufficiently after 24 h of incubation, the tistically using Sigma-Stat software (version 2.03, Systat software, plates were incubated an additional 24 h under the same condi- Inc., Point Richmond, Calif.) with one-way analysis of variance. tions prior to checking the results. The breakpoint of resistance is Significance was determined using Dunnett’s test in 95% confi- defined, according to National Antimicrobial Resistance Monitor- dence level (P Ͻ 0.05). ing System, as the MIC above which microorganisms are consid- ered to be resistant to the antibiotics tested. The breakpoints for RESULTS resistance were applied as follows: ciprofloxaxin, Ն4 ␮g/ml; chloramphenicol, Ն32 ␮g/ml; erythromycin, Ն8 ␮g/ml; kana- Prevalence of C. jejuni and C. coli. Raw chicken meat mycin, Ն64 ␮g/ml; nalidixic acid, Ն32 ␮g/ml; and tetracycline, purchased from traditional markets, large retail stores, or Ն16 ␮g/ml. Amounts of five antibiotics sold for use in the Korean department stores located in Seoul, Daejeon, Busan, and poultry industry from 2001 through 2002 were obtained from the Gwangju regions in Korea and human stool samples from National Antimicrobial Resistance Management Program Report patients with diarrhea were used to investigate the contam- 2004 prepared by the Korea Food and Drug Administration (16). ination levels of C. jejuni and C. coli. Isolation rates of C. jejuni and C. coli from raw chicken meat and human stool samples are presented in Table 1. Campylobacter was detected in 570 of 923 whole chicken meat samples, resulting in the isolation rate of 61.8%. A total of 579 Campylobac- ter isolates were collected from the 570 chicken meat samples, 9 of which were contaminated with both C. jejuni and C. coli. When Campylobacter isolates were further identi- fied for C. jejuni and C. coli, respective species were isolated from 335 and 244 samples with isolation rates of 36.3 and 26.4%, respectively. Campylobacter strains were isolated from 15 human stool samples, resulting in an isolation rate of 2.9%, which was far less than that of raw chicken meat. Isolation rates of C. jejuni and C. coli from human stool samples were 2.5 and 0.4%, respectively. Results showed that C. jejuni was slightly more prevalent than C. coli in both raw chicken samples and human stool samples. Seasonal changes in the isolation rate and cell populations of C. jejuni and C. coli in raw chicken meat are shown in Figure 1. Of the 400 raw chicken samples collected in 2002, C. jejuni and C. coli were isolated from 159 samples and 59 samples, respectively. Isolation rates of C. jejuni ranged from 28.4 to 35.3% from January through June when the average seasonal temperatures were 2.3 to 15.4ЊC, but sharply increased to 68.6 and 65.6% in July through August (average temperature of 25.2ЊC) and in September through October (average temperature of FIGURE 1. Monthly variations of Campylobacter jejuni and Cam- 17.8ЊC), respectively, reaching the highest in the year (Fig. pylobacter coli in retail-sold raw chicken meat during January through December, 2002. (A) Isolation rates (%) of () 1A). The isolation rate in November through December C. jejuni Њ and C. coli (□). (B) Populations (CFU per gram) of C. jejuni (average temperature of 4.6 C) was 37.9%, higher than that () and C. coli (). Asterisks indicate significant differences (P in January through February (average temperature of Ͻ 0.05). Cell populations were determined for all the chicken 2.3ЊC). In contrast, the isolation rates of C. coli ranged samples tested in 2002, not just for positive samples. between 10 and 25% throughout the year without increases 2918 KANG ET AL. J. Food Prot., Vol. 69, No. 12

FIGURE 2. Comparison of isolation rates (A) and cell populations (B) of Campylo- bacter jejuni and Campylobacter coli from raw chicken samples in traditional markets (), large retail stores (), department stores (᭝). Samples were collected from January through December 2002. Aster- isks indicate signiﬁcant differences from control (P Ͻ 0.05). Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021

during summer seasons. The cell population of C. jejuni in raw chicken were more manifest in the traditional mar- demonstrated that seasonal effects led to rapid increases to kets than in large retail stores and department stores. The 588 and 840 CFU/g in July through August and September average isolation rates of three retail places, such as tradi- through October, respectively (Fig. 1B). C. jejuni cells were tional markets, large retail stores, and department stores, detected at low levels with 82 and 53 CFU/g in January were 64.5, 43.0, and 43.2%, respectively, with the respec- through February and March through April, respectively, tive cell populations of 572.5, 102.6, and 287.3 CFU/g. In while those of 471 and 186 CFU/g from raw chicken sam- the case of traditional markets where raw chickens were ples collected in May through June and November through sold on ice at tables exposed to atmosphere, the isolation December were detected, respectively. These seasonal in- rates of Campylobacter species sharply increased even in creases in the cell population from May through October, March through April (65.2%) and reached up to 93.3% in typical summer and early fall seasons in Korea, were sta- July through August and 82.1% in September through Oc- tistically signiﬁcant (P Ͻ 0.05). On the other hand, cell tober (Fig. 2A). On the other hand, isolation rates of C. populations of C. coli were almost unchanged throughout jejuni and C. coli from the products sold in large retail the year ranging from 23 to 95 CFU/g, showing no signif- stores and department stores were lower than those in tra- icant seasonal differences (P Ͼ 0.05). ditional markets throughout the year, except from Septem- Raw chicken samples purchased from various retail es- ber through October. Campylobacter in chicken samples tablishments were used to investigate whether they might obtained from traditional markets and department stores inﬂuence the isolation rates and cell populations of C. jejuni was isolated to a similar degree during the early fall season and C. coli. Prevalence of Campylobacter in raw chicken (September through October), while that from the large re- meat sold at three retail establishments is presented in Fig- tail market was lower by 30% than in the other two places ure 2. Seasonal effects on the prevalence of Campylobacter for the same period. Cell populations of Campylobacter J. Food Prot., Vol. 69, No. 12 RESISTANCE OF CAMPYLOBACTER SP. ISOLATED FROM MEAT AND STOOLS 2919

FIGURE 3. Antibiotic resistance of Cam- pylobacter isolated from raw chicken samples in 2000 (; n ϭ 234), 2001 (; n ϭ 129), and 2002 (□; n ϭ 231). Numbers in parenthesis indicate the number of samples tested in each year. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021

from the samples sold in the traditional market sharply in- low 400 CFU/g throughout the year from November creased after March through April, reaching the highest of through February, typical late fall and winter seasons in 1,258 CFU/g in September through October, while those Korea (Fig. 2B). Statistical analysis of enumeration data from samples sold at the other two locations remained be- showed no significant changes among those obtained throughout the year from large retail stores and department stores (P Ͼ 0.05). However, contamination levels of Cam- TABLE 2. Distribution of the MICs for Campylobacter jejuni and pylobacter in raw chickens purchased from traditional mar- Campylobacter coli isolates (n ϭ 594) from raw chicken meat kets in July through August and September through Octo- No. of isolates with MIC (␮g/ml) ber were statistically significant (P Ͻ 0.05). MIC Cipro- Chloram- Erythro- Kana- Nalidixic Tetra- Antimicrobial resistance of C. jejuni and C. coli iso- ␮ ( g/ml) floxacin phenicol mycin mycin acid cycline lates. Susceptibilities of the 594 Campylobacter isolates 0.016 4 collected from raw chicken samples from 2000 through 0.032 13 1 2002 to six antibiotics, including ciprofloxaxin, chloram- 0.047 7 3 2 phenicol, erythromycin, kanamycin, nalidixic acid, and tet- 0.064 8 8 3 racycline, are presented in Figure 3. Campylobacter isolates 0.094 3 15 0 tested were the most resistant to nalidixic acid (91.4%) fol- 0.125 1 52 4 lowed by ciprofloxaxin (87.9%), tetracycline (87.2%), 0.19 1 51 1 kanamycin (30.6%), erythromycin (19.4%), and chloram- 0.25 1 3 68 0 phenicol (1.3%), respectively. Likewise, Campylobacter 0.38 1 17 51 1 2 isolates from human stool samples showed a similar resis- 0.5 1 45 85 0 0 tance to the six antibiotics tested (data not shown). 0.75 1 77 54 2 2 1 1.0 1 167 52 16 13 2 MIC distributions for Campylobacter isolates from the 1.5 3 105 16 29 13 5 raw chicken meat are shown for the six antibiotics (Table 2.0 14 87 11 52 8 7 2). When tested against ciprofloxaxin, the majority of the 3.0 13 37 5 48 1 5 chicken isolates (87.9%) required a MIC of Ն4 ␮g/ml, the 4.0 26a 21 5 57 2 9 antibiotic resistance breakpoint, while a few isolates re- 6.0 22 10 3 71 1 7 quired lower MICs, indicating that most of the Campylo- 8.0 22 2 1 75 3 10 bacter isolates were resistant to the antibiotic. Campylo- 12 18 4 2 39 3 17 bacter isolates from raw chicken meat (98.7%) required 16 9 8 3 15 1 46 MICs of Ͻ32 ␮g/ml, the resistance breakpoint of chlor- 24 6 3 1 4 3 47 amphenicol, indicating that most isolates were susceptible 32 419 0 1 3 9 54 48 1 0 1 16 20 to the antibiotic. However, Campylobacter isolates suscep- 64 1 1 0 22 41 tible to chloramphenicol showed a wide distribution of 96 0 0 0 5 13 MICs, suggesting a possibility that some of them may be- 128 0 2 1 7 10 come resistant if exposed further to the antibiotics. When Ͼ256 6 104 181 484 287 tested against erythromycin and kanamycin, the chicken isolates showed a similar MIC distribution pattern with re- a Bold characters indicate the breakpoint used: ciprofloaxacin, Ն4 sistant isolates being 19.4 and 30.6%, respectively. Cam- ␮g/ml; chloramphenicol, Ն32 ␮g/ml; erythromycin, Ն8 ␮g/ml; kanamycin, Ն64 ␮g/ml; nalidixic acid, Ն32 ␮g/ml; tetracycline, pylobacter isolated from human stool samples also dis- Ն16 ␮g/ml. played the similar distribution profiles and resistance to the 2920 KANG ET AL. J. Food Prot., Vol. 69, No. 12

TABLE 3. Patterns of multidrug-resistance to six antibiotics of Campylobacter jejuni and Campylobacter coli isolated from raw chicken meat and human stool samples isolated from 2000 through 2002 No. of Campylobacter isolates (%)a

No. of Raw chicken meat Human stool samples resistant antibiotics 2000b 2001 2002 2000

0 6 (2.6) 1 (0.8) 0 (0.0) 1 (6.7) 1 15 (6.4) 5 (3.9) 8 (3.5) 3 (20.0) 2 31 (13.2) 11 (8.5) 7 (3.0) 2 (13.4) 3 107 (45.7) 56 (43.4) 149 (64.5) 7 (46.7) 4 48 (20.5) 37 (28.7) 54 (23.4) 1 (6.7)

5 26 (11.1) 18 (14.0) 13 (5.6) 1 (6.7) Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021 6 1 (0.4) 1 (0.8) 0 (0.0) 0 (0.0) Total 234 129 231 15 a Numbers in parentheses indicate the ratio of Campylobacter isolates (%) showing multidrug resistance with respective numbers of resistant antibiotics. b The year when Campylobacter isolates were obtained. two antibiotics, erythromycin and kanamycin, with 13.3% rate of Campylobacter spp. in raw chicken meat was 61.8% for both of them. Human isolates resistant to nalidixic acid from 2000 through 2002 (Table 1), which is similar to the and tetracycline were 66.7 and 86.7%, respectively (data result reported by Flynn et al. (8) where the isolation rate not shown). of Campylobacter spp. from fresh chicken wings was There have been reports on emerging microorganisms 64.7%. A high contamination level of Campylobacter spp. displaying resistance to more than one antibiotic (11). Mul- in broilers (82.9%) and raw chicken meat (81.3%) in Italy tidrug resistance of Campylobacter isolates of both raw from 2000 through 2001 was also reported (25). Remark- chicken meat and human stool samples was investigated in ably, these levels were even higher than those observed in this study (Table 3). Approximately 32% of the Campylo- this study. The results of this study strongly suggest that bacter isolates from raw chicken meat collected in 2000 prevention and control measures for Campylobacter con- showed simultaneous resistance to more than four antibi- tamination in raw chicken meat as well as for possible otics while 6.4% of the isolates were resistant to one anti- cross-contamination of the pathogen with other foods and biotic. There were even two chicken isolates obtained from food contact surfaces or humans should be firmly estab- 2000 through 2002 that were resistant to all six antibiotics lished, in spite of relatively low reported outbreaks of cam- tested. Approximately 2.6% of the chicken isolates collect- pylobacteriosis in Korea. For the effective prevention and ed in 2000 were susceptible to six antibiotics tested. Cam- control of cross-contamination of resistant Campylobacter pylobacter isolates with resistance to four or more antibi- spp., the measures on external packaging used for raw otics increased to 43.5% in 2001, but decreased to 29% in chicken meat may also be included as pointed out by Bur- 2002. On the other hand, the ratio of Campylobacter iso- gess et al. (4). According to the authors, approximately lates resistant to three antibiotics was 45.7% in 2000 and 1.1% of Campylobacter spp. on the external packaging of 43.4% in 2001, but increased to 64.5% in 2002. However, raw meat was observed, indicating thorough care is re- all the Campylobacter isolates obtained in 2002 were required to minimize cross-contamination of the pathogen sistant to at least one antibiotic. The human isolates col- with other food contact surfaces. Although Campylobacter lected in 2000 showed the severe multidrug resistance with is microaerophilic, it should be noted that the pathogen can 13.4% of the isolates being resistant to more than four an- survive outside of the animal host, possibly causing cross- tibiotics tested. Among the 15 human Campylobacter iso- contamination (4). The results of this study are also sug- lates, only one isolate was resistant to five antibiotics while gestive of this possibility, and thus, it is necessary to handle none of the human isolates were resistant to all six antibi- raw chicken meat with caution as evidenced by the high otics tested. Twenty percent of the human isolates were contamination rate of over 60%. Compared with raw chick- resistant to one antibiotic while only 6.7% of the isolates en samples, the prevalence of Campylobacter in human were susceptible to all six antibiotics tested. stool samples was relatively low. A possible explanation for this is that there were time lapses when collecting the DISCUSSION samples at hospitals. In fact, human stool samples were Campylobacter species have been reported as one of collected and stored for 1 to 2 weeks at hospitals until used the major microbial contaminants in meat and poultry for the study. Another possible explanation for the low worldwide (3, 7, 20). In Korea, however, a few cases of the prevalence of Campylobacter in human stool samples is campylobacteriosis have been reported only in recent years. attributed to the fact that people cook chicken meat and In fact, the first official report on the campylobacteriosis thus successfully kill the pathogen before eating it. outbreak in Korea was in 2001 and since then, one or two Contamination and prevalence of Campylobacter spp. outbreaks have been reported every year (14). The isolation in raw chicken meat samples are influenced by several fac- J. Food Prot., Vol. 69, No. 12 RESISTANCE OF CAMPYLOBACTER SP. ISOLATED FROM MEAT AND STOOLS 2921 tors, such as storage temperature (refrigeration, freezing, or conditions on chicken skin stored at 4ЊC for 7 days and at room temperature), relative humidity, pH of the meat, and/ ambient room temperatures for 3 days. Another report by or storage period (3, 12). In this study, we investigated how Chantarapanont et al. (7) showed that when C. jejuni of seasonal changes in atmospheric temperature of different less than 500 CFU/g was inoculated onto skin of raw chick- retail establishments affected the prevalence of Campylo- en and stored at 4 and 25ЊC, the pathogen could survive bacter in raw chicken meat. The climate of Korea is known for 3 days at 4ЊC, but the cell population decreased 1 log for four seasons with distinct temperature changes in each during 24 h of storage at 25ЊC. In contrast, when inoculated season. Prevalence of Campylobacter spp. in raw chicken to a 20- to 30-␮m depth beneath the skin, C. jejuni re- meat sold in Korea was monitored in 2002 in order to in- mained viable without decrease in cell population at 25ЊC, vestigate how seasonal factors influenced the prevalence of possibly since the environment beneath the skin of raw the pathogen (Fig. 2). Seasonal temperature, likely in com- chicken provided favorable conditions for the growth of bination with other environmental factors such as moisture this microaerobic bacteria. It may be hypothesized that, if content and pH, appeared to affect the prevalence of Cam- raw chicken meat is contaminated either during slaughter- Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021 pylobacter in raw chicken. Both the isolation rates and cell ing or postslaughter processes with a relatively high amount populations of C. jejuni and C. coli peaked from September of C. jejuni on the skin or beneath the skin through cuts, through October when the average temperature was lower the pathogen may have a chance for survival and growth than that in summer from July through August (25.2ЊC). It even at ambient temperatures. may be contributed to the compliance with the food safety Inappropriate handling and storage of raw chicken management guidelines by the retailers and strict guidance meat during slaughtering, processing, and/or transportation of the food safety authorities in Korea during the high- may result in a relatively high contamination level at the temperature season. When seasonal temperatures started to time of warehousing them, increasing the chance of surviv- drop steeply in early fall, September through October, the al and/or growth of Campylobacter in raw chicken meat. safety management system might be relaxed to some degree In the case of large retail stores and department stores causing the higher contamination level of Campylobacter where the raw chicken meat is handled under temperature- spp. in raw chicken meat. Meldrum et al. (20) reported that controlled conditions, C. jejuni may survive for an extended the positive rate of Campylobacter spp. contaminated in period without growing (7, 20). It may be speculated that raw chickens sold in Wales (United Kingdom) from 2001 a low initial contamination level of C. jejuni in the samples through 2002 was 70.8%. Contamination was the highest sold in these retail establishments, which supposedly pro- in June and the lowest in January, March, and December, vide more stringent hygienic management than a traditional typical winter months in Wales. Results of several studies market does, might have contributed to relatively un- strongly suggested that enforcement of and compliance changed cell population during the summer season. How- with an appropriate safety management system was impor- ever, further investigation is needed to support this hypoth- tant for the reduction of foodborne campylobateriosis (4, esis. 20). Distinct seasonal variation of Campylobacter preva- Antibiotic resistance of microorganisms has become a lence in poultry meat stresses the importance of long-term major health-threatening issue worldwide (2, 13, 25). Gupta surveillances on chickens, rather than snapshot ones which et al. (11) screened a total of 1,553 human Campylobacter may give an artificially high or low result, depending on isolates for antibiotic resistance in 2001 and found that 19% the season sampled. of the isolates were ciprofloxacin resistant while 2% were Prevalence of Campylobacter spp. in raw chicken meat erythromycin resistant. When multidrug resistance was in- varied among the market places. Typically in traditional vestigated, the percentages of human isolates resistant to markets, raw chicken meat is sold on ice-filled open con- more than one, two, and three antibiotics were 51, 18, and tainers. On the other hand, the large retail stores and de- 10%, respectively (11). The three major kinds of antibiotics partment stores sell the product under refrigerated condi- to which most of Campylobacter strains were resistant in tions. The isolation rate of C. jejuni and C. coli from raw the United States were ciprofloxacin, nalidixic acid, and chicken meat in the traditional markets was the highest at tetracycline. There are reports describing that the antibiotic 93.3% during July through August, but the cell population resistance of Campylobacter isolated from poultry meat and peaked from September through October (1,258 CFU/g). It human samples increased each year (10, 13, 17). The re- is generally known that C. jejuni can survive 2 to 4 weeks sistance to fluoroquinolone antibiotics has become a con- under moist, microaerobic conditions at 4ЊC, and even 2 to siderable problem worldwide, which was correlated with 5 months at Ϫ20ЊC, but only a few days at room temper- the increased use of the fluoroquinolones to food animals ature (12). However, it was shown in this study that both in many countries (1, 11, 22). Although antibiotic resistance the isolation rate and cell population of C. jejuni obtained of microorganisms has been recognized as a health-threat- from raw chicken meat sold in the traditional market were ening issue in Korea for the past several years, antibiotic relatively low from November through February when the resistance of Campylobacter spp. isolated in Korea had not average atmospheric temperature was similar to refrigerated been reported until now. To the best of our knowledge, this temperature, whereas they were the highest from July is the first report on the antibiotic resistance of Campylo- through October when the average temperature was 18 to bacter spp. isolated from raw chicken meat sold in Korea. 25ЊC. Lee et al. (18) also reported that C. jejuni was able When tested against antibiotics in our study, 87.9% of the to persist at high numbers or grow under normal packing chicken isolates collected from 2000 through 2002 dis- 2922 KANG ET AL. J. Food Prot., Vol. 69, No. 12 played resistance to ciprofloxacin indicating that contami- we suggest that the application of antibiotics to food ani- nation of ciprofloxacin-resistant Campylobacter in raw mals should be approached with caution and that it is pru- chicken meat is at a considerable level. Nalidixic acid, the dent to establish an appropriate antibiotic management sys- first generation drug of quinolone antibiotics, is a marker tem in order to effectively decrease antibiotic resistance of of ciprofloxacin susceptibility because the nalidixic acid– Campylobacter spp. in Korea. The results of this study pro- susceptible strains are also susceptible to ciprofloxacin, vide information that may be useful to develop a compre- whereas most of the resistant ones are resistant to cipro- hensive database as a sound scientific information source. floxacin (9). Likewise, in our study, 520 of the 543 chicken The database, in turn, may promote the establishment of isolates resistant to nalidixic acid were found to be cipro- prevention and control measures of Campylobacter spp. in floxacin resistant (95.8%), while 516 ciprofloxacin-resistant raw chicken meat and be used to reduce the number of chicken isolates of 522 strains were resistant to nalidixic antibiotic-resistant pathogens. acid (98.9%). The proportion of Campylobacter isolates obtained from 2000 through 2002 with multidrug resistance ACKNOWLEDGMENT Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021 to four or more antimicrobials ranged from 28 to 43.5%, This work was funded by the Ministry of Health and Welfare, Re- indicating that it could be a health-threatening factor. public of Korea (project no. HMP-99-F-06-0001). There have been a few reports on human campylobacteriosis in Korea, which makes it difficult to get information REFERENCES on the amount of antibiotics therapeutically used to treat 1. Angulo, F. J., N. L. Baker, S. J. Olsen, A. Anderson, and T. J. Barrett. the illness. Statistical information on antibiotics used in the 2004. Antimicrobial use in agriculture: controlling the transfer of Korean poultry industry has only been available since 2001 antimicrobial resistance to humans. Pediatr. Infect. Dis. J. 15:78–85. (15, 16). Tetracycline is one of the antibiotics to which 2. Angulo, F. J., V. N. Nargund, and T. C. Chiller. 2004. Evidence of resistance of many microorganisms has been reported in an association between use of anti-microbial agents in food animals other countries (10, 25). Tetracycline was also the most and anti-microbial resistance among bacteria isolated from humans and the human health consequences of such resistance. J. Vet. Med. widely used antibiotic in the Korean poultry industry in B51:374–379. 2001 and 2002 with total sales of 100,918 and 119,201 kg, 3. Boer, E. D., and M. Hahne. 1990. Cross-contamination with Cam- respectively (15, 16). On the other hand, the amount of pylobacter jejuni and Salmonella spp. from raw chicken products chloramphenicol used in the poultry industry was minimal during food preparation. J. Food Prot. 53:1067–1068. (i.e., 1 to 2 kg during 2001 and 2002 in Korea). When the 4. Burgess, F., C. L. Little, G. Allen, K. Williamson, and R. T. Mitchell. 2005. Prevalence of Campylobacter, Salmonella, and Escherichia resistance of Campylobacter to antibiotics was compared coli on the external packaging of raw meat. J. Food Prot. 68:469– with the amount of antibiotic sales in Korea, it appeared as 475. though the antibiotic resistance of the pathogen was in- 5. Centers for Disease Control and Prevention. 2003. National antimi- versely related to the sale amounts of the respective anti- crobial resistance monitoring system (NARMS): 2001 annual report. biotics. For example, increase in sales of tetracycline in U.S. Department of Health and Human Services, Atlanta. 6. Centers for Disease Control and Prevention. 2004. National antimi- 2002 was 18% and it was observed that Campylobacter crobial resistance monitoring system (NARMS): 2002 annual report. resistance to this antibiotic increased by 4.6% during the U.S. Department of Health and Human Services, Atlanta. same period. A similar trend on antibiotic use and resis- 7. Chantarapanont, W., M. Berrang, and J. F. Frank. 2003. Direct mi- tance of Campylobacter was observed with other antibiotics croscopic observation and viability determination of Campylobacter tested, suggesting that limiting the amount of antibiotic use jejuni on chicken skin. J. Food Prot. 66:2222–2230. in the poultry industry may be one control measure that 8. Flynn, O. M. J., I. S. Blair, and D. A. Mcdowell. 1994. Prevalence of Campylobacter species on fresh retail chicken wings in Northern reduces antibiotic resistance of Campylobacter spp. How- Ireland. J. Food Prot. 57:334–336. ever, further investigation is necessary to determine the cor- 9. Gaudreau, C., and H. Gilbert. 1997. Comparison of disc diffusion relation between antibiotic use and emergence of antibiotic- and agar dilution methods for antibiotic susceptibility testing of resistant Campylobacter spp., because official reports on Campylobacter jejuni subsp. jejuni and Campylobacter coli. J. An- antibiotic sale amounts only became available in 2000, lim- timicrob. Chemother. 39:707–712. iting a comprehensive statistical analysis. 10. Ge, B. G., D. G. White, P. F. McDermott, W. Girand, S. Zhao, S. Hubert, and J. Meng. 2003. Antimicrobial-resistant Campylobacter Resistance to nalidixic acid correlated closely with re- species from retail raw meat. Appl. Environ. Microbiol. 69:3005– sistance to fluoroquinolones, such as ciprofloxacin (1). Cur- 3007. rently, nalidixic acid is not used in the poultry industry in 11. Gupta, A., J. M. Nelson, T. J. Barrett, R. V. Tauxe, S. P. Rossiter, C. Korea because ciprofloxacin is widely used instead (15). R. Friedman, K. W. Joyce, K. E. Smith, T. F. Jones, M. A. Hawkins, Angulo et al. (1) reported that an effective decrease in an- B. Shiferaw, J. L. Beebe, D. J. Vugia, T. Rabatsky-Ehr, J. A. Benson, T. P. Root, and F. J. Angulo. 2004. Antimicrobial resistance among tibiotic-resistant microorganisms could result from a reduc- Campylobacter strains, United States 1997–2001. Emerg. Infect. Dis. tion in antibiotic use in the European Union without com- 10:1102–1109. promising animal health and significantly increasing the 12. Hunt, J. M., C. Abeyta, and T. Tran (ed.). 1998. Bacteriological an- cost of production. According to Ge et al. (10), when a alytical manual online, chap. 7, 8th ed. AOAC International, Gai- total of 378 Campylobacter strains isolated from raw meat thersburg, Md. sold in retail stores were investigated for antibiotic resis- 13. Kassenborg, H. D., K. E. Smith, D. J. Vugia, T. Rabasky-Ehr, M. R. Bate, M. A. Carter, N. B. Dumas, M. P. Cassidy, N. Marano, R. V. tance, the isolates showed the highest resistance to tetra- Tauxe, and F. J. Angulo. 2004. Fluoroquinolone-resistant Campylo- cycline (82%), followed by erythromycin (54%), nalidixic bacter infection: eating poultry outside of home and foreign travel acid (44%), and ciprofloxacin (35%). Based on our study, are risk factors. Clin. Infect. Dis. 38:S279–S284. J. Food Prot., Vol. 69, No. 12 RESISTANCE OF CAMPYLOBACTER SP. ISOLATED FROM MEAT AND STOOLS 2923

14. Korea Food and Drug Administration. 2004. Foodborne outbreaks: for antibiotic susceptibility testing of Campylobacter jejuni and current status and prevention strategies. Seoul, Korea. Campylobacter coli. J. Clin. Microbiol. 41:1062–1068. 15. Korea Food and Drug Administration, National Antimicrobial 20. Meldrum, R. J., D. Tucker, and C. Edwards. 2004. Baseline rates of Resistance Management Program. 2003. Establishment of control Campylobacter and Salmonella in raw chicken in Wales, United system of antibiotics for livestocks in 2003, p. 183–283. Seoul, Kingdom in 2002. J. Food Prot. 67:1226–1228. Korea. 21. Nachamkin I., and M. J. Blaser (ed.). 2000. Campylobacter, 2nd ed. 16. Korea Food and Drug Administration, National Antimicrobial Re- American Society for Microbiology Press, Washington, D.C. sistance Management Program. 2004. Establishment of control sys- 22. Nachamkin, I., M. J. Blaser, and L. S. Tompkins (ed.). 1992. Cam- tem of antibiotics for livestocks in 2004, p. 221–292. Seoul, Korea. pylobacter jejuni: current status and future trends. American Society for Microbiology Press, Washington, D.C. 17. Ledergerber, U., G. Regula, R. Steohan, J. Damuser, B. Bissigm, and 23. National Committee for Clinical Laboratory Standards. 2003. Per- K. D. C. Stark. 2003. Risk factors for antibiotic resistance in Cam- formance standard for antimicrobial disk susceptibility tests, vol. 23, pylobacter spp. isolated from raw poultry meat in Switzerland. BMC no. 1. M2-A8. NCCLS, Wayne, Pa. Public Health. 3:1–9. 24. National Committee for Clinical Laboratory Standards. 2004. Per- 18. Lee A., S. C. Smith, and P. J. Coloe. 1998. Survival and growth of formance standard for antimicrobial disk susceptibility tests, vol. 24, Campylobacter jejuni after artiﬁcial inoculation onto chicken skin as no. 1. M100-S14. NCCLS, Wayne, Pa. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/69/12/2915/2002244/0362-028x-69_12_2915.pdf by guest on 26 September 2021 a function of temperature and packaging conditions. J. Food Prot. 25. Pezzotti, G., A. Seraﬁn, I. Luzzi, R. Mioni, M. Milan, and R. Perin. 61:1609–1614. 2003. Occurrence and resistance to antibiotics of Campylobacter je- 19. Luber, P., E. Bartelt, E. Genschow, J. Wangner, and H. Hahn. 2003. juni and Campylobacter coli in animals and meat in northeastern Comparison of broth microdilution, E test and agar dilution methods Italy. Int. J. Food Microbiol. 82:281–287.