Campylobacteriosis is an essentially foodborne zoonotic disease worldwide (Allos, 2001). Most of these human illnesses involve C. jejuni (81%) or C. coli (8.4%) (Sifré et al., 2015). Enteric diseases caused by the thermophilic species C. jejuni, C. coli, C. lari, and C. upsaliensis range from asymptomatic infections to severe inflammatory bloody diarrhea (Ketley, 1997). In addition, C. jejuni is often associated with the Guillain-Barre syndrome (Dingle et al., 2001). Campylobacter spp. widely and asymptomatically colonize the intestinal tracts of food producing animals (Altekruse et al., 1999; Harvey et al., 1999; Stanley and Jones, 2003), and C. jejuni is the main species affecting chickens, whereas C. coli is mainly found in pigs (Sáenz et al., 2000; Haruna et al., 2013). Poultry with no obvious clinical symptoms are the main reservoirs of these pathogens (Luangtongkum et al., 2006; Humphrey et al., 2007). Pigs are also usually subclinically infected with Campylobacter spp. During the slaughter process, intestinal content spillage could contaminate the meat and meat products (Weijtens et al., 1997; Gill et al., 1999).

Although Campylobacter infection is usually self-limiting, severe, long-lasting or systemic Campylobacter infections can be treated with antibacterial drugs (Allos, 2001). Tetracyclines, macrolides, and fluoroquinolones are used to control Campylobacter-mediated infections (Sifré et al., 2015). In food-producing animal industries, ciprofloxacin (a fluoroquinolone) is frequently used because it has a broad antibacterial spectrum. In recent decades, fluoroquinolone resistance in animal and human Campylobacter isolates have increased worldwide (Engberg et al., 2001; Moore et al., 2006; Wang et al., 2016). With this increase in fluoroquinolone-resistant Campylobacter, macrolides have become common drugs in treatment of campylobacteriosis (Luangtongkum et al., 2009). Macrolides, including tylosin, tilmicosin, tulathromycin, and tildipirosin, have been widely used in animal production in many countries (Poehlsgaard et al., 2012; McEwen and Collignon, 2018), and macrolide-resistant Campylobacter strains have been isolated in many regions (Kaakoush et al., 2015; Zhang et al., 2016) and from animals (Wang et al., 2016; Lim et al., 2017; Hafez and Attia, 2020). Tetracycline has also been extremely widely used in animal farming because it has broad-spectrum antibacterial activity, low cost, and excellent efficacy (Chopra and Roberts, 2001), and it may have cause high tetracycline resistance in Campylobacter from food-producing animals (Premarathne et al., 2017; Woźniak-Biel et al., 2018). The proportion of Campylobacter spp. strains resistant to ciprofloxacin, tetracycline, and erythromycin is increasing (García-Fernández et al., 2018). A tendency to detect increasing multidrug-resistant (MDR) strains has been reported in both humans (Feizabadi et al., 2007) and food-producing animals (Li et al., 2016; Neogi et al., 2020). The potential risk of MDR Campylobacter spread from animals to human is of great concern.

In Campylobacter, the C257T mutation in the gyrase gene (gyrA) is the most frequent mechanism causing quinolone and fluoroquinolone resistance (Iovine, 2013). Efflux pumps and ribosomal protection genes are two major mechanisms of tetracycline resistance in various bacteria (Roberts, 2005). The tetO gene encodes a ribosomal protection protein that is primarily responsible for tetracycline-resistance. It is carried on the chromosome (Chopra and Roberts, 2001) or transferred on plasmids (Gibreel et al., 2004). The efflux genes, tetA and tetB, encode membrane-bound efflux proteins that export tetracycline from the cell (Chopra and Roberts, 2001), and the presence of tetA in Campylobacter isolates was first detected by Abdi-Hachesoo et al. (2014). Campylobacter has three main mechanisms that confer resistance to macrolides: modifications to ribosome-associated genes (V region of 23S rRNA and the rplD and rplV ribosomal-protein-encoding genes) (Luangtongkum et al., 2009), multidrug efflux pumps mediated by CmeABC (Lin et al., 2002), and a ribosomal methylase encoded by ermB, which was first reported in China in 2014 (Qin et al., 2014). Since then, ermB-positive Campylobacter has been isolated from broilers in Spain and the United States (Florez-Cuadrado et al., 2016).

Several studies have been published in China that report relatively high levels of antimicrobial resistance in Campylobacter from animals in different areas (Ma et al., 2014, 2017). Furthermore, in a campylobacteriosis outbreak in China in 2018, and the domain isolate was resistant to nalidixic acid, tetracycline and ciprofloxacin (Qu et al., 2019). In China, a national annual antimicrobial-resistance surveillance system for Campylobacter in food-producing animals has been in place from 2017 to 2020 to strengthen the monitoring of antimicrobial resistance and ensure the safety of feed animals and public health. To obtain epidemiological data on Campylobacter from food-producing animals in Jiangsu Province, China, the resistance rates of Campylobacter strains in chicken and pig feces on farms and the relevant molecular mechanisms of their resistance to quinolones, tetracycline, and macrolides have been investigated.

Thermophilic Campylobacter are zoonotic pathogens, and antimicrobial resistance in Campylobacter has gradually become a major public health problem in both developed and developing countries (Wieczorek and Osek, 2013). In the present study, 45 C. jejuni and 25 C. coli were isolated from chickens, and 23 C. coli were isolated from pigs. The C. jejuni was the predominant Campylobacter species in chicken isolates (64.3%), followed by C. coli (35.7%), and C. coli was the dominant Campylobacter species in pigs, accounting for 100% of the pig isolates. The isolation rates of Campylobacter on pig farms was 23.0%, which was within the range of 13.4%–26.1% reported by Wang et al. (2016). In China, laying hens are used as stewed chicken. Hence, Campylobacter in layers can also pose a risk for consumers. The isolation rates of Campylobacter on layer farms (34.3%) was lower than the rate on broiler farms (57.5%). The total isolation rates of Campylobacter on chicken farms in Jiangsu Province was 46.7%, which was higher than the isolation rates in other regions of China (Wang et al., 2016). The isolation rate of Campylobacter differed in different regions, which may be attributable to the different prevalence of Campylobacter in different regions or to different animal breeding conditions, breeding scales, sample numbers, and the isolation methods used by various laboratories.

Telithromycin belongs to a class of semisynthetic macrolides, the ketolides, and has been designed to treat respiratory infections. In this study, resistance to telithromycin was highly prevalent among the Campylobacter isolates: 42.2% of C. jejuni isolates from chickens, 44.0% of C. coli isolates from chickens, and 60.9% of C. coli isolates from pigs were resistant to telithromycin. We noted a difference in telithromycin resistance between C. coli from chickens and those from pigs, but the relationship between the emergence of ketolide resistance among Campylobacter and the macrolides used in pig production is not yet clear. The resistance rate to gentamicin in the C. coli isolates from chickens (40.0%) was similar to that for C. jejuni isolates (31.1%). These findings are inconsistent with the results of Li et al. (2016), who reported that the prevalence of gentamicin resistance among C. coli isolates was 93.6%, whereas that in C. jejuni isolates was 17.9%. Overall, our results indicate that the drug resistance rates in C. coli for the antimicrobial classes tested were higher than those in C. jejuni from chickens. In recent years, MDR Campylobacter strains have been a growing global public health problem. In this study, a high prevalence of MDR Campylobacter isolates was detected (up to 64.5%). However, the MDR rate was lower than in other countries, such as Italy (100%) (Fraqueza et al., 2014). The prevalence of MDR isolates of C. coli (68%) was similar to that of C. jejuni (64.4%) in chickens, which is inconsistent with the results of Wang et al. (2016). The reason for this discrepancy may be the small number of samples examined in the present study.

High rates of fluoroquinolone resistance were observed among the 93 Campylobacter isolates. We found that the rates of fluoroquinolone resistance in C. coli (92.0% and 100%) and C. jejuni isolates (80.0% and 66.7%) from chickens were higher than the rates of resistance in C. coli isolates from pigs (65.2% and 21.7%). This was attributable to the use of fluoroquinolones as the preferred drugs against bacterial infections in poultry production, and has been caused by the unreasonable and uncontrolled use of antimicrobial agents in poultry production in China. These results are consistent with previous reports in China, in which the Campylobacter strains isolated from chickens were resistant to quinolones (Ma et al., 2014; Li et al., 2016). Therefore, the high prevalence of fluoroquinolone-resistant Campylobacter isolates in chickens may cause serious public health problems.

Quinolone resistance in Campylobacter spp. is usually caused by the point mutation T86I in the gyrase protein, and is the most frequently detected mechanism (Hormeño et al., 2016). In this study, 79 Campylobacter isolates (39 C. jejuni and 40 C. coli) with the quinolone (ciprofloxacin or nalidixic acid) resistance phenotype were positive for the T86I mutation. Other authors have reported similar results, where all ciprofloxacin-resistant Campylobacter strains carried the same mutation (Woźniak-Biel et al., 2018; Elhadidy et al., 2020).

Given the long-term use of tetracyclines in feed additives for livestock and poultry, large numbers of tetracycline-resistant isolates are found in animal reservoirs (Premarathne et al., 2017; Hafez and Attia, 2020). It has previously been reported that 95.6% of C. jejuni isolates from chickens, 97.5% of C. coli isolates from chickens and 97.5% of C. coli isolates from pigs in China were resistant to tetracycline (Wang et al., 2016). In the present study, a relatively high level of resistance to tetracycline was observed in the C. jejuni (71.1%) and C. coli (88.0%) isolates from chickens.

The tetO gene is the most commonly reported tetracycline resistance gene in Campylobacter spp. (Luangtongkum et al., 2009), and is located on the chromosome or transmissible plasmids (Abdi-Hachesoo et al., 2014; Narvaez-Bravo et al., 2017). In this study, all 64 tetracycline-resistant Campylobacter isolates (32 C. jejuni and 32 C. coli) carried the tetO gene. The high prevalence of the tetO gene reflects the high rate of tetracycline resistance in Campylobacter spp. isolated from chickens and pigs in China.

Chopra and Roberts (2001) reported that the efflux proteins encoded by tetA and tetB can export tetracycline from the cell. In this study, the tetA detection rate in the Campylobacter isolates was 6.5%, but there was no evidence of tetB. Abdi-Hachesoo et al. (2014) reported detecting the tetA resistance gene in 18% of Campylobacter spp. isolated from poultry carcasses in Iran. Nguyen et al. (2016) reported that the detection rates of tetA gene in C. jejuni and C. coli isolates from chickens in Kenya were 90.3% and 100%, respectively. Previous studies have proved that the tetA gene can be located on mobile elements, such as plasmids, and can be horizontally transferred among bacterial strains (Szczepanowski et al., 2004). Campylobacter spp. with the tetA gene may be involved in the distribution of this resistance gene to other food-borne bacteria in the animal industry.

The prevalence of Campylobacter resistance to macrolides varies according to geographic region (Rożynek et al., 2013). In the present study, erythromycin and azithromycin resistance were detected in 62.2%–100.0% and 60.0%–82.6% of Campylobacter isolates from pigs and chickens. In Korea in 2010, Lim et al. reported that 15.5% of Campylobacter isolates from chicken and swine feces or carcasses showed phenotypic resistance to erythromycin (Lim et al., 2017). Rozynek et al. reported a low rate of erythromycin resistance (4.7%), which was similar to that reported in chicken-origin isolates in other European countries (Rożynek et al., 2013). In the present study, the rates of erythromycin and azithromycin resistance were much higher in C. coli from chickens and pigs than in C. jejuni from chickens, as has been observed in previous studies (Chen et al., 2010; Wang et al., 2016). The rate of erythromycin-resistant C. coli isolates in pigs (100%) was higher than the rate in chickens (72.0%), which was much higher than has been reported in previous studies in China (Qin et al., 2011). Notably, in the present study, the erythromycin resistance rate in C. jejuni from chickens (62.2%) was higher than those reported in most previous studies undertaken in China (Chen et al., 2010; Li et al., 2016; Wang et al., 2016; Zhang et al., 2016). High levels of macrolide resistance were attributable to the widespread and increasing use of macrolides in animal production, including tylosin, tilmicosin, erythromycin, and kitasamycin.

When the macrolide-resistant isolates were screened, a range of different molecular resistance mechanisms were identified, including mutations in the 23S rRNA, rplD, and rplV genes and the presence of ermB. Our results show that the A2075G mutation in the 23S rRNA gene in the Campylobacter isolates might be the cause of their high resistance to azithromycin and erythromycin. The rate of the A2075G mutation in the present study was 98.4% (62/63) in the Campylobacter isolates with high level resistance to erythromycin, which is similar to the frequencies reported in a previous study (Luangtongkum et al., 2009). The mutation was not present in Campylobacter isolates that were susceptible to azithromycin or erythromycin or that displayed low level or intermediate resistance to these agents. No other mutations in the 23S rRNA gene were detected in this study.

Several studies have reported that ribosomal protein L4 mutations at amino acid positions 192 and 121, which conferred no significant difference between macrolide-resistant and -susceptible strains (Lehtopolku et al., 2011; Qin et al., 2011). In the present study, several mutations, including V65I, G74A, A103V, T109A, T109S, A111E, A114T, V130A, and A132V, in ribosomal protein L22 were detected in macrolide-susceptible and macrolide-resistant strains. This result is similar to a previous report by Lim et al. (2017). However, in the present study, several mutations are reported for the first time, including G69A, P120T, V121A, X125T, X126S, and E133K, which were found in both macrolide-susceptible and -resistant isolates. We noted one amino acid insertion (118APAAKK119) in the L22 ribosomal protein in nine Campylobacter isolates and another amino acid insertion (120TTKAP121) in L22 in two macrolide-susceptible C. jejuni isolates. The L22 protein of C. jejuni 81-176 also contains amino acid insertion 118APAAKK119 (Hao et al., 2013). Our results suggested that these changes were unlikely to contribute directly to macrolide resistance.

The mechanism of resistance mediated by ermB is particularly noteworthy because this gene can transfer macrolide resistance horizontally among Campylobacter strains and confer high-level resistance (Qin et al., 2014). In the present study, the detection rate of ermB was 52.7%, which was higher than those reported in other areas of China (Wang et al., 2014; Li et al., 2016). We observed that the rate of ermB-positive C. jejuni isolates (30/45) was significantly higher than that of ermB-positive C. coli isolates (19/48, P = 0.0089), which is inconsistent with previous reports (Zhang et al., 2016; Liu et al., 2017). The prevalence of the ermB gene in the Campylobacter from chickens (65.7%, n = 46) was significantly higher than in those from pigs (13.0%, n = 3, P < 0.0001). Notably, 27 (43.5%, 27/62) ermB-positive isolates also carried the A2075G mutation in 23S rRNA, which is a higher rate than in previous reports by Wang et al. (38%, 22/58) (Wang et al., 2014) and Zhang et al. (1.9%, 3/157) (Zhang et al., 2016). These results confirmed that ermB was markedly more prevalent among Campylobacter strains isolated from chickens than in those from pigs. The ermB gene must be monitored in Campylobacter because it is so highly transmittable. In general, the A2075G point mutation in domain V of the 23S rRNA and the ermB gene were the main causes of macrolide resistance in the Campylobacter isolates from Jiangsu Province, China.

In conclusion, this study has demonstrated that Campylobacter spp. isolated from chickens and pigs had high drug resistance rates to fluoroquinolones, tetracyclines and macrolides. The emergence of MDR Campylobacter strains is attributable to the widespread use of antibiotics in poultry and pig production. Monitoring antimicrobial resistance in Campylobacter and the prudent use of antimicrobials in animal-food production are essential to reducing antimicrobial resistance in bacterial pathogens.

All datasets generated for this study are included in the article/supplementary material.

The animal study was reviewed and approved by Animal Welfare and Ethical Censor Committee at Jiangsu Institute of Poultry Science. Written informed consent was obtained from the owners for the participation of their animals in this study.

QZ, JZ, XZ, and MT performed antibiotic susceptibility tests. XT and JL collected the samples. XT, MT, and SZ did PCR amplification and DNA sequence analysis. QZ conducted the statistical analysis of data. MT completed the manuscript. YG directed the study and assisted in the completion of the manuscript. All authors read and agreed to submit the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.