Campylobacter jejuni from Canine and Bovine Cases of Campylobacteriosis Express High Antimicrobial Resistance Rates against (Fluoro)quinolones and Tetracyclines

Campylobacter (C.) spp. from poultry is the main source of foodborne human campylobacteriosis, but diseased pets and cattle shedding Campylobacter spp. may contribute sporadically as a source of human infection. As fluoroquinolones are one of the drugs of choice for the treatment of severe human campylobacteriosis, the resistance rates of C. jejuni and C. coli from poultry against antibiotics, including fluoroquinolones, are monitored within the European program on antimicrobial resistance (AMR) in livestock. However, much less is published on the AMR rates of C. jejuni and C. coli from pets and cattle. Therefore, C. jejuni and C. coli isolated from diseased animals were tested phenotypically for AMR, and associated AMR genes or mutations were identified by whole genome sequencing. High rates of resistance to (fluoro)quinolones (41%) and tetracyclines (61.1%) were found in C. jejuni (n = 29/66). (Fluoro)quinolone resistance was associated with the known point mutation in the quinolone resistance-determining region (QRDR) of gyrA, and tetracycline resistance was mostly caused by the tet(O) gene. These high rates of resistance, especially to critically important antibiotics in C. jejuni and C. coli, are worrisome not only in veterinary medicine. Efforts to preserve the efficacy of important antimicrobial treatment options in human and veterinary medicine have to be strengthened in the future.


Introduction
Human campylobacteriosis is the most common cause of bacterial gastroenteritis worldwide. Campylobacter (C.) jejuni and C. coli are the most frequently isolated species in humans with diarrhea [1]. In most cases, the infection is foodborne, from handling or eating undercooked poultry, raw milk, or contaminated water. In Europe, more than 240,000 cases were reported in 2018 [1]. Source attribution studies identified C. jejuni and C. coli from poultry as the main source of human campylobacteriosis [2]. Therefore, antimicrobial resistance (AMR) in C. jejuni and C. coli is a key element of European AMR monitoring. High to very high rates of resistance of Campylobacter spp. from poultry resistant to antimicrobials that are critically important for humans, such as (fluoro)quinolones, have been detected with an increasing trend over the last 10 years [3]. The results from a Swiss AMR monitoring program using cecum samples from broilers showed that 51.4% of C. jejuni were resistant to fluoroquinolones and 40.0% to tetracycline. Moreover, 66.7% of C. coli in these broilers were resistant to fluoroquinolones and 40.0% to tetracycline. A similar situation exists for porcine Campylobacter spp. About half of the tested isolates were resistant to quinolones (50.3% to ciprofloxacin and 52.2% to nalidixic acid). Furthermore, 62.1% of these strains were resistant to tetracycline [3].
On the other hand, much less is known about the Campylobacter spp. strains isolated from clinical cases of cattle and pets. Although not identified as a main source of human campylobacteriosis, Campylobacter spp. from these animal species could contribute to human cases, and transmission to humans is possible. Campylobacteriosis in pets and cattle occurs only sporadically, and it is not recommended to treat infected animals even with symptoms like diarrhea with antibiotics, unless there is a risk of sepsis [4,5]. Therefore, data on AMR for these pathogens are rare. In studies analyzing antimicrobial susceptibility to ciprofloxacin of C. jejuni from dogs, major differences depending on the country could be observed [6][7][8][9][10]. It must be considered that different methods of antimicrobial susceptibility testing (AST; E-test, agar dilution, disk diffusion, microbroth dilution, or multilocus sequence typing) and different breakpoints and cut-off values are used.
In Campylobacter, point mutations in the quinolone resistance-determining region (QRDR) of gyrA are most often responsible for resistance to fluoroquinolones [11]. Specific single point mutations in the QRDR can reduce the susceptibility of Campylobacter spp. to fluoroquinolones considerably [12], while avoiding fitness costs [13]. The resistance of Campylobacter spp. to tetracycline is normally mediated by the acquisition of the tet(O) gene, which can be located on the chromosome or on transferable plasmids [14,15]. In a study from Italy with C. coli from pigs, they found a strong correlation between phenotypic and genotypic resistance to fluoroquinolones, on the one hand, and tetracycline on the other. In particular, point mutations in gyrA and the presence of tet(O) were associated with such resistance [16].
The aim of this study was to determine the AMR rates of C. jejuni and C. coli isolated from diseased dogs, cats, and cattle for critically important antimicrobials using the European-wide harmonized method, and to compare these data to the known resistance rates of regularly monitored livestock animals. Furthermore, the molecular mechanisms responsible for phenotypic AMR were analyzed using whole genome sequencing.
Note: * Number of resistant isolates/number of isolates tested (percentage of resistance). Calculations were only performed when ≥10 isolates were present. CI, confidence interval.
One canine C. coli isolate was resistant to CIP, NAL, and STR. All eight tested isolates of bovine C. coli showed resistance to STR. Six of the eight tested bovine C. coli isolates were resistant to both (fluoro)quinolones and TET. Five isolates showed an additional resistance to ERY. No C. coli isolates from cats were available.
Of the 75 tested C. jejuni/coli isolates, 25 isolates showed phenotypical microbiological resistance against only one antimicrobial class (Table 2). Eighteen isolates were only resistant to CIP-NAL, five strains only to TET, and two strains exclusively to STR. The most often found resistance pattern in C. jejuni was CIP-NAL (18/36), and CIP-NAL-ERY-TET-STR in bovine C. coli (5/9). Of the 66 tested C. jejuni, 13 were fully susceptible to all the antimicrobials tested, whereas none of the nine C. coli were fully susceptible.
Because of the limited number of isolates, the calculated confidence intervals overlap to a great extent. Therefore, further statistical analyses were not performed.

Antimicrobial Resistance Genes
Whole genome sequencing was performed on all the strains with phenotypic AMR (n = 45). There were 36 C. jejuni strains (canine n = 21, feline n = 3, bovine n = 12) and nine C. coli strains (canine n = 1, bovine n = 8). The phenotypic AMR pattern and the presence of corresponding AMR genes and point mutations are listed in Table 2. Resistance to both NAL and CIP (n = 36) was found to be caused by a point mutation resulting in T86I in the QRDR of gyrA. This mutation was not found in the one C. jejuni isolate resistant to NAL but not to CIP. Three TET resistance ribosomal protection proteins were identified in isolates showing non-wild-type TET minimum inhibitory concentration (MIC) (n = 24). tet(O) was the most common gene and was detected in 16 isolates. Furthermore, the mosaic tet(O/32/O) was identified in two C. jejuni strains, and tet(W) was found in one bovine C. coli isolate. Phenotypic TET resistance in five isolates could not be assigned to a known resistance gene. Various resistance genes and one single point mutation were associated with resistance to STR. In one C. jejuni isolate, a point mutation in rpsL leading to an amino acid change K88R was detected. The aadE-Cc gene was discovered in four C. coli isolates and one C. jejuni isolate. Three bovine C. coli strains harbored the cassette aadE-Cc, ant(6)-Ia, aph(3 )-IIIa, sat4. These isolates were resistant to all tested antimicrobials, except GEN. Additionally, cfr(C) was present in these three isolates, which could be linked to ERY resistance. In the two other isolates resistant to ERY, the point mutation A2075G in the 23S rRNA gene was observed.

Discussion
In our study, AMR to (fluoro)quinolones was slightly higher in canine C. jejuni isolates (n = 15/39; 38.5% (95% CI: 24.9-54.1) than in a study from 2014 in Switzerland (n = 28/134; 20.9% (95% CI: 14.9-28.5)), which typed the isolates with MLST and fla-typing [10]. This could be an indication of increasing resistance to quinolones in companion animals. In Switzerland, a statistically significant increase in the resistance of C. jejuni to CIP in broilers has been observed over the past 10 years [3]. Because (fluoro)quinolone resistance is not associated with a loss of fitness, a rapid development of resistance occurs [13]. The resistance rate of C. jejuni and C. coli to CIP in diseased cattle (61.1% (95% CI: 38.6-79.7) in our study is comparably high to that in slaughter pigs (50.3% (95% CI: 42.7-57.9)) and slaughter broilers (54.1% (95% CI: 46.2-61.4)), which are regularly monitored in Switzerland [3]. It is interesting, as one might expect, that the resistance rates are higher in diseased and therefore possibly treated animals than in healthy slaughter animals. Moreover, antimicrobial treatment regimens for livestock species differ markedly-e.g., for broilers in particular, only oral treatment is applied, as opposed to cattle, where other treatment routes are common. The resistance of C. jejuni to TET is lower in the isolates of this study than in the broilers of AMR monitoring. In contrast to the results of the AMR monitoring, no isolate of C. jejuni resistant to GEN or ERY was found in our study, whereas the ERY and GEN resistance of C. jejuni in broilers is very low but does occur [3]. The resistance of C. jejuni to STR is comparable to the resistance of Swiss broilers harboring C. jejuni as a commensal. All the tested C. coli were resistant to STR. In other studies, high resistance rates of C. coli to STR are described [3,17]. The resistance rate to STR in this study was higher compared to other aminoglycosides such as GEN, because STR binds to a single site on the 30S subunit of the ribosome and high-level resistance can be selected through a single mutation [18]. All the resistance rates in cats and all the resistance rates for C. coli can only be evaluated to a limited extent due to the low number of cases.
Fluoroquinolone resistance often arises from point mutations in the QRDR of gyrA, with the resulting amino acid change C257T being the most common [11,19]. Mutations in gyrA leading to T86I were found in all the isolates resistant to CIP and NAL, except for one strain, which was only resistant to NAL and not to CIP. This phenomenon was described by Jesse et al. (2006), who detected a single Thr86Ala mutation in Campylobacter, leading to phenotypic resistance to NAL but not to CIP [20]. However, this mutation was not detected in our strain. Three tetracycline ribosomal protection proteins were detected in our study; tet(O) was detected most frequently, followed by tet(O/32/O) and tet(W). The tet(O) gene widely occurs in C. coli and C. jejuni and can be located either on the chromosome or on a plasmid [21][22][23].  [25,26]. It is likely that Campylobacter spp. acquired tet(O/32/O) from other intestinal bacteria.
The third ribosomal protection protein, tet(W), was detected in a bovine isolate of C. coli. Previous studies found tet(W) widely distributed in the genome of ruminal bacteria of cattle, and the authors assumed that antimicrobial treatment is the driver of positive selection for tetracycline-resistant bacteria [27]. Holman et al. found that the use of oxytetracycline increased the proportion of tet(W) in fecal samples [28]. These studies show that the tet(W) gene can be present in the microbiome of the digestive tract and can be transferred to other bacteria.
All five ERY-resistant C. coli isolates showed very high MICs (>128 mg/L) and were isolated from cattle. For two of these resistant strains, the causal mutation is in domain V of the 23S rRNA gene and occurs at position 2075, which leads to a substitution from A to G. This point mutation is the predominant mutation in clinical and field isolates [11,19]. Probably, the cause of ERY resistance in the other three strains is the presence of cfr(C), which could only be detected in these isolates. Cfr(C) is an rRNA methyltransferase that is normally linked with resistance to phenicols, lincosamides, oxazolidines, pleuromutilins, and streptogramin A [29,30]. A Chinese study showed that cfr(C)-harboring C. coli showed an increased MIC to erythromycin, but not to phenicols. The authors suggested that cfr(C) was suppressed by unknown mechanisms. Moreover, they observed that cfr(C) was located next to the gene cluster aadE-aphA3-sat4 [31]. The three strains in our study carrying cfr(C) also harbored the gene cluster and were multidrug-resistant to CIP, NAL, TET, ERY, and STR. Resistance to STR can be caused by several genes and mutations. A mutation causing K88R in rpsL was detected in one of our bovine C. jejuni isolates; that this causes a high MIC with STR in Campylobacter spp. has been confirmed in other studies [32]. The most frequently discovered acquired resistance gene in STR-resistant isolates in our study was aadE-Cc, either alone or as cassette with ant(6)-Ia, aph(3 )-IIIa, and sat4. The most prevalent mechanism of aminoglycoside resistance in Campylobacter spp. and other bacteria is the modification of the aminoglycoside structure by enzymes such as aminoglycoside acetyltransferases, aminoglycoside phosphotransferases, and aminoglycoside nucleotidyltransferases [33,34]. The presence of the aadE-aphA-3-sat4 cluster was observed earlier in both C. jejuni and C. coli in several publications [35,36]. In contrast to other studies, no resistance to GEN was found in our isolates [37].
In general, a strong correlation of detected AMR genes and point mutations and phenotypic resistance could be observed for all the groups of antimicrobials. In three C. jejuni strains and one C. coli strain, phenotypic resistance against TET was observed, but no corresponding gene was found. For two phenotypically STR-resistant C. coli, no responsible mutation or AMR gene could be confirmed. The overall low number of isolates from clinical cases of canine, feline, and bovine camplyobacteriosis limits our findings to some extent, and hampered statistical analysis in general. Hence, the results are presented solely descriptively.

Materials and Methods
Between 2015 and 2018, a total of 75 C. jejuni/coli strains isolated from canine, feline, and bovine clinical cases by 6 diagnostic laboratories were sent to the Swiss national reference laboratory for species identification (Table 3). There were 9 strains of C. coli (cattle n = 8, dog n = 1) and 66 strains of C. jejuni (dog n = 39, cat n = 9, cattle n = 18). Species identification was performed using matrix-assisted laser desorption/ionization time-of-flight mass spectroscopy (MALDI-TOF MS) (Biotyper 3.0, Bruker Daltonics, Bremen, Germany). Isolates were stored at -80 • C in tryptone soy medium containing 30% glycerol until AST was performed. The strains were recovered on trypticase soy agar plates with 5% sheep blood (TSA SB; Becton Dickinson, Franklin Lakes, NJ, USA) and incubated microaerobically at 37 ± 1 • C for 48 h. AST was performed by microbroth dilution according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI) against the following antimicrobials: erythromycin (ERY), ciprofloxacin (CIP), tetracycline (TET), gentamicin (GEN), nalidixic acid (NAL), and streptomycin (STR). For a standardized concentration of 5 × 10 5 CFU/mL, a McFarland of 0.5 (Densicheck, BioMérieux, Marcy-l'Etoile, France) was set and 50 µL of this suspension was transferred to 10 mL of cation-adjusted Mueller-Hinton broth with 5% lysed horse blood (TREK Diagnostic Systems, Thermo Fisher Scientific, UK). In each well of the plate, EUCAMP2 100 µL (TREK Diagnostic Systems) was inoculated using an auto-inoculator (Thermo Fisher Scientific). The plates were incubated in a 36 ± 1 • C microaerobic atmosphere for 48 h. The minimal inhibitory concentration (MIC) was defined as the lowest concentration of antibiotic showing no growth. The isolates were defined as microbiologically susceptible or resistant according to the epidemiological cut-off values (ECOFFs) issued by the European Committee on Antimicrobial Susceptibility Testing (www.eucast.org; MIC and zone diameter distributions and ECOFFs, version 5.26).
Whole genome sequencing was performed on isolates after DNA extraction on a Qiagen Qiacube using the QIAamp DNA mini kit, with Nextera XT or Nexteralexlibrary production and sequencing at 24-to 48-plex on a MiSeq with 2 × 300 bp or 96-plex on a NextSeq 500 with a 2 × 150 bp to minimum 40× mean coverage. Assemblies were performed using Unicycler v0.3.0b [38]. All the raw data have been deposited with the European Nucelotide Archive (ENA) under project number PRJEB39858.
The identification of determinants to quinolones, erythromycin, aminoglycosides, and tetracycline was performed with ABRicate (https://github.com/tseemann/abricate) using the National Center for Biotechnology Information (NCBI) database [39], ResFinder 3.0 [40], and CARD [41], with a threshold for the identification of acquired genes of 90% identity and 60% minimum length. ABRicate does not detect mutations, hence ResFinder 3.0 and CARD were used to detect point mutations. ResFinder 3.0 did not find point mutations in C. coli.

Conclusions
The known high resistance rate of C. jejuni to critically important (fluoro)quinolones in broilers in Europe is also seen in Swiss C. jejuni isolated from bovine campylobacteriosis cases (61.1%) and, to a lesser extent, in canine campylobacteriosis cases (38.5%). This (fluoro)quinolone resistance is caused by a known point mutation in the QRDR of gyrA. Moreover, tetracycline resistance as a result of the presence of tet(O), tet(W), and tet(O/32/O) was observed in bovine (33.3%) and canine (23.1%) C. jejuni isolates. In general, the number of C. coli isolates resistant to at least one antimicrobial tested was high, but it has to be considered that the number of isolates was very low. The finding of high resistance rates of Campylobacter spp. to critically important antimicrobials in our study emphasizes the need for regular AMR monitoring not only in zoonotic and commensal bacteria from healthy slaughter animals, but also in pathogens from diseased livestock and companion animals.