Prevalence and fluoroquinolone resistance of Campylobacter spp. isolated from beef cattle in Japan

Abstract

Beef is a source of human Campylobacter infections. Antimicrobial treatment is needed when patients are immunocompromised or have other comorbidities. Therefore, we investigated the prevalence and antimicrobial resistance of Campylobacter spp. in beef cattle in Japan. Rectal swab samples were collected from 164 beef cattle at an abattoir between March 2021 and August 2021, and Campylobacter spp. were isolated from 94 (57.3%) cattle. C. jejuni and C. coli were isolated from 68 and 26 cattle, respectively. For Campylobacter jejuni, the resistant rates against ampicillin, tetracycline and ciprofloxacin were 20.6, 75.0 and 64.7%, respectively. For C. coli, the resistant rates against ampicillin, tetracycline and ciprofloxacin were 53.8, 76.9 and 88.5%, respectively. No Campylobacter isolates were resistant to erythromycin. By multilocus sequence typing, C. jejuni and C. coli isolates were classified into 22 and 2 sequence types (STs). The top three STs of C. jejuni were ST806 (12 isolates), ST21 (nine isolates), and ST459 (eight isolates). The most frequent ST of C. coli was ST1068 (23 isolates). The results suggest that Campylobacter spp. are prevalent in the gastrointestinal tract of beef cattle slaughtered at abattoirs. Furthermore, the administration of erythromycin is effective against human campylobacteriosis caused by beef consumption. Monitoring the prevalence and antimicrobial resistance of Campylobacter spp. in beef cattle could be useful for managing the risk of human campylobacteriosis.

Main text

Campylobacter spp. are one of four key global causes of diarrheal diseases in humans worldwide, and the two most frequent species isolated from human campylobacteriosis are C. jejuni and C. coli, according to the World Health Organization (WHO 2020). Antimicrobial therapy is effective in patients with severe symptoms (Yang et al. 2019; WHO 2020). Erythromycin is often used as first-line treatment for human campylobacteriosis; however, in the absence of a microbiological diagnosis, fluoroquinolones, which are classified as one of “critically important antimicrobials” by WHO (2019), are used to treat human campylobacteriosis (Japanese Association for Infectious Disease/Japanese Society of Chemotherapy 2018). Therefore, fluoroquinolone-resistance in Campylobacter spp. poses an important issue in the chemotherapy treatment of humans. In Campylobacter spp., fluoroquinolone resistance is mainly caused by chromosomal mutation in the gyrA gene, which results in a Thr-86-Ile substitution (Zirnstein et al. 1999; Zirnstein et al. 2000).

In Japan, Kumagai et al. (2020) investigated foodborne outbreaks caused by Campylobacter spp. between 2007 and 2018 and estimated that the first and second most important sources of the disease were chicken (80.3%) and beef products (10.5%). The Japanese Ministry of Agriculture, Forestry and Fisheries has identified chicken meat as the principal source of human campylobacteriosis and monitors the prevalence of Campylobacter spp. in broiler flocks (Sasaki et al. 2011; Haruna et al. 2012). However, there are only few reports on the prevalence of Campylobacter spp. in beef cattle in Japan. Therefore, we investigated the prevalence and antimicrobial resistance of Campylobacter spp. in beef cattle. The results of our study could be useful for managing the risk of human campylobacteriosis caused by the consumption of beef.

Rectal swabs were collected from 164 beef cattle shipped from 34 farms in seven regions (Hokkaido, Tohoku, Kanto, Tokai, Kinki, Chugoku, and Kyushu) between March 2021 and August 2021. Of the 164 cattle, 130 were Japanese Black (JB) and 33 were crossbred (JB × Holstein Friesian) (HF). The average ages of the 130 JB and 33 crossbred cattle were 30.4 (26.7 to 44.4) and 26.9 (24.0 to 32.1) months, respectively, with significant differences between the ages of JB and JB × HF crossbred cattle (Mann-Whitney U test: P < 0.01). The remaining cattle aged 28.6 months old was another crossbred (JB × Japanese Shorthorn) (JS). Of the 130 JB cattle, 17 and 4 were born from HF and JB × HF crossbred cows, respectively, by embryo transferring. Whereas 16 (9.8%) cattle (15 JB cattle and one JB × JS crossbred cattle) were shipped to an abattoir directly from beef farms where they were born, the remaining 148 (90.2%) were raised on at least two farms (average of 2.5 farms, 2 to 6 farms).

Campylobacter spp. were isolated from 94 (57.3%) cattle shipped from 29 (85.3%) beef farms in all seven regions. The prevalence of Campylobacter spp. was statistically higher (Fisher’s exact test: P < 0.05) in JB × HF crossbred cattle (72.7%, 24/33) than in JB cattle (53.1%, 69/130). The prevalence of Campylobacter spp. in JB cattle born from HF cows by embryo transferring was 52.9% (9/17) and was reflected similarly (53.1%, 60/113) in other JB cattle. Thirty-two JB cattle were older than the eldest JB × HF crossbred cattle (32.1 months old). The prevalence of Campylobacter spp. in them was 40.6% (13/32) and tended to be lower than that (57.1%. 56/98) of JB cattle younger than 32.1 months.

C. jejuni was isolated from 68 cattle and C. coli was isolated from 26 cattle (Table 1). No animals carried both C. jejuni and C. coli. C. jejuni resistant rates against ampicillin, streptomycin, tetracycline, nalidixic acid, and ciprofloxacin were 20.6, 5.9, 75.0, 64.7 and 64.7%, respectively. Resistance to both tetracycline and ciprofloxacin was observed in 54.4% (37/68) of the C. jejuni isolates in this study. All the C. jejuni isolates were susceptible to erythromycin, chloramphenicol and gentamicin. C. coli resistant rates against ampicillin, tetracycline, chloramphenicol, nalidixic acid, and ciprofloxacin were 53.8, 76.9, 11.5, 88.5, and 88.5%, respectively. Resistance to both tetracycline and ciprofloxacin was observed in 73.1% (19/26) of the C. coli isolates. All the C. coli isolates were susceptible to erythromycin and gentamicin. By multilocus sequence typing (MLST), C. jejuni was classified into 22 sequence types (STs) and C. coli isolates were classified into two STs (Table 2). For C. jejuni, the top three STs were ST806 (12 isolates), ST21 (nine isolates), and ST459 (eight isolates). These STs were obtained from 10 farms located in four regions (Hokkaido, Tohoku, Kanto and Chugoku), eight farms in four regions (Hokkaido, Tohoku, Kanto and Kyushu), and six farms in four regions (Hokkaido, Tohoku, Kanto and Chugoku), respectively. The C. jejuni isolates classified into these top three STs accounted for 45.3% (29/64) of the C. jejuni isolates, and resistance to both tetracycline and ciprofloxacin was observed in 79.3% (23/29). For C. coli, ST1068 isolates accounted for 88.5% (23/26) of the C. coli isolates and were obtained from nine farms in all regions. Resistance to both tetracycline and ciprofloxacin was observed in 78.3% (18/23) of ST1068 C. coli isolates. Ciprofloxacin resistance was observed in 16 (72.7%, ST21, ST487, ST596, ST656, ST806, ST982, ST4526, ST11029, ST1739, ST9078, ST459, ST6532, ST9681, ST61, ST929 and ST922) and two (100%, ST827 and ST1068) of 22 STs of C. jejuni and two STs of C. coli, respectively.

Table 1 Antimicrobial resistance in C. jejuni and C. coli isolates

Table 2 Sequence types and antimicrobial resistance profiles of C. jejuni and C. coli isolates

The Thr-86-Ile change (mediated by the C257T mutation in the gyrA genes) in all ciprofloxacin-resistant isolates was detected using two mismatch amplification mutation assay PCR methods. Moreover, in one ciprofloxacin-resistant strain per sequence type, the mutation was confirmed by directly sequencing of partial gyrA gene.

In the present study, the prevalence of Campylobacter spp. was statistically higher in JB × HF crossbred cattle than in JB cattle. One of the differences between the two was the age at the time of slaughter. Several studies conducted at dairy and beef farms and abattoirs have shown that young cattle had a higher prevalence of Campylobacter spp. than older cattle (Giacoboni et al. 1993; Beach et al. 2002; Mielsen 2002; Ramonaitė et al. 2013). Therefore, the higher prevalence of Campylobacter spp. in JB × HF crossbred cattle compared to JB cattle, might be attributed to the fact that the crossbred cattle were slaughtered at a younger age than the JB cattle.

A decade ago, we investigated the prevalence of Campylobacter spp. in beef and dairy farms and reported that the prevalence of both Campylobacter spp. in beef and dairy farms were approximately 90% (Haruna et al. 2013; Sasaki et al. 2013). In Japan, beef cattle are rarely transported directly from the farms where they were born to abattoirs. Since JB × HF crossbred cattle and JB cattle by embryo transferring are generally born in dairy farms, they are transported to beef farms to fatten. In the present study, 90.2% of cattle was raised on at least two farms (maximum 6 farms). In some cases, the distance between farms where cattle were transported was 900 km or more. The long-term shedding of Campylobacter spp. in beef and dairy cattle has been reported (Inglis et al. 2004; Hakkinen and Hänninen 2009). As a result of the transportation of cattle colonized with Campylobacter spp. between farms, it could spread to beef farms throughout Japan and the  the prevalence in beef farms could remain high.

The top three frequent STs of C. jejuni STs (ST806, ST21 and ST459) and the most frequent C. coli ST (ST1068) were obtained from at least six farms. It was previously reported that prevalent STs of C. jejuni and C. coli isolated from cattle in Japan were ST806 and ST21, and ST1068 (Asakura et al. 2012; Asakura et al. 2019; Sasaki et al. 2020). ST21 of C. jejuni and ST1068 of C. coli are also prevalent in cattle in Italy, USA., and the UK, demonstrated that the host associations of Campylobacter genotypes are more robust than their geographic variations (Kwan et al. 2008; Sanad et al. 2011; Roux et al. 2013; Bianchini et al. 2014; Jonas et al. 2015; Cha et al. 2017; Sheppard et al. 2010). The top three STs of C. jejuni and the most frequent ST (ST1068) of C. coli would be adaptable to beef cattle and widely spread to beef farms in Japan. Moreover, these prevalent STs might be selected by use of tetracyclines and fluoroquinolones, since approximately 80% of them were resistant to both tetracycline and ciprofloxacin. In Japan, tetracyclines and fluoroquinolones are used for treating beef cattle with bacterial diarrhea or pneumonia.

According to the report of the Japanese Veterinary Antimicrobial Resistance Monitoring System (JVARM) (National Veterinary Assay Laboratory of Japan 2020), resistance rates against ciprofloxacin in C. jejuni and C. coli isolated from cattle slaughtered at abattoirs in 2017 were 50.5 and 81.4%, respectively, slightly lower than those found in this study. However, no erythromycin resistance was observed in C. jejuni and C. coli isolated from cattle, which is reflected in the results of this study. Erythromycin is often the first drug of choice for the treatment of human campylobacteriosis (JAID/JSC 2018). This study indicate that erythromycin administration remains effective in treating human campylobacteriosis caused by the consumption of beef.

Conclusion

The Approximately 60% of beef cattle slaughtered at abattoirs carry Campylobacter spp. in the gastrointestinal tract and approximately 70% of the isolates are resistant to fluoroquinolones. Thus, it is necessary to monitor the prevalence and antimicrobial resistance of Campylobacter spp. in beef cattle.

Methods

Sampling was conducted at an abattoir in Tokyo between March 2021 and August 2021. We sampled rectal swabs from 164 beef cattle diagnosed as healthy upon visual inspection by veterinarians. The individual cattle identification number was recorded to obtain information about the production history of the cattle. After evisceration, a rectal sample was taken of the opened rectal lumen of the cattle using a cotton swab (SEEDSWAB No. 1; Eiken Chemical, Tokyo, Japan) and transported to the National Institute of Health Sciences at 4°C. At the laboratory, the sample was refrigerated at 4°C until examination, which was performed within 4 h of sampling. Each swab head was rubbed over the surface of modified charcoal cefoperazone deoxycholate agar (Oxoid, Basingstoke, United Kingdom) and incubated at 42°C for 48 h. A maximum of two suspected isolates per sample were identified using a multiplex PCR (Kamei et al. 2016). The primers used for the multiplex PCR are shown in Table 3. One Campylobacter species per sample was subjected to antimicrobial susceptibility testing and MLST (http://pubmlst.org/campylobacter/). Antimicrobial susceptibility testing was conducted against ampicillin (0.12–256 mg/L), streptomycin (0.12–128 mg/L), erythromycin (0.12–128 mg/L), tetracycline (0.12–128 mg/L), chloramphenicol (0.12–256 mg/L), nalidixic acid (0.12–128 mg/L), ciprofloxacin (0.03–64 mg/L), and gentamicin (0.12–256 mg/L). Antimicrobial minimal inhibitory concentrations were determined using the broth microdilution method with dried plates (Eiken Chemical), as per the guidelines of the Clinical and Laboratory Standards Institute (2016). The antimicrobial susceptibility breakpoints (ampicillin: 32 mg/L, streptomycin: 32 mg/L, erythromycin: 32 mg/L, tetracycline: 16 mg/L, chloramphenicol: 16 mg/L, nalidixic acid: 32 mg/L, and ciprofloxacin: 4 mg/L) adopted by Clinical and Laboratory Standards Institute 2020) and JVARM (2020) were used; except for the tests involving gentamicin (2 mg/L), which was specified by the Danish Integrated Antimicrobial Resistance Monitoring and Research Programme 2000). Campylobacter jejuni ATCC33560 were used as the quality control strain. To detect the Thr-86-Ile mutation (C. jejuni: ACA → ATA, C. coli: ACT → ATT) in gyrA genes of ciprofloxacin-resistant isolates, two mismatch amplification mutation assay PCR methods (Zirnstein et al. 1999; Zirnstein et al. 2000) were used. Moreover, one strain per sequence type was determined by MLST, partial gyrA genes were amplified by PCR (Zirnstein et al. 1999; Zirnstein et al. 2000), and the PCR products were directly sequenced to detect the Thr-86-Ile mutation. The primers used for the detection of the mutation of gyrA are shown in Table 3. All statistical analyses were performed using R V. 4.1.2.

Table 3 List of primers used in this study