Campylobacter is one of the most prevalent causes of bacterial foodborne gastroenteritis worldwide (1, 2). In the United States, Campylobacter causes an estimated 1.3 million illnesses and costs ~$1.7 billion yearly for medical treatment and lost productivity (3, 4). Human Campylobacter infections are primarily caused by the consumption of contaminated poultry meat (5, 6). In addition to chickens, Campylobacter is prevalent in both beef and dairy cattle (7–9). Humans can acquire Campylobacter from cattle through direct contact, ingestion of unpasteurized milk, and water contamination (10–15). Although most individuals infected with Campylobacter may not require antibiotic treatment, severe and systemic infections necessitate antimicrobial therapy, including macrolides (e.g., erythromycin) and fluoroquinolones (FQs) (e.g., ciprofloxacin) (16–19). Unfortunately, both classes of antibiotics are becoming less effective in treating campylobacteriosis due to increasing rates of resistance to these drugs in Campylobacter (20–22). The fact that Campylobacter is a zoonotic pathogen exposed to FQs used in both animal production (e.g., beef cattle and non-lactating dairy cattle) and human medicine may contribute to the development of FQ-resistant (FQ-R) Campylobacter. In counties like the United States, Australia, and Canada, FQ antibiotics such as enrofloxacin and danofloxacin have indications for subcutaneous use in both sick (therapeutic treatment) and healthy cattle (metaphylaxis) at high risk of bovine respiratory disease (BRD) development (23–29).

Fluoroquinolone-resistant mutant can spontaneously develop in Campylobacter (30, 31), and the use of FQ antibiotics selects and enriches these mutants (32). In Campylobacter, FQ resistance is mostly caused by point mutations in the quinolone resistance-determining regions (QRDR) of DNA gyrase (gyrA) (33, 34), most commonly with the Thr-86-Ile amino acid substitution (C257T mutation), in conjunction with the function multidrug efflux pump CmeABC (34–37). Interestingly, FQ resistance caused by gyrA mutations can be maintained in Campylobacter without antibiotic selection pressure, suggesting that FQ-R mutants do not carry a fitness burden (38, 39). For example, a previous study conducted by our group revealed a significant fitness advantage of FQ-R over FQ-susceptible (FQ-S) Campylobacter jejuni without antibiotic selection pressure when co-inoculated into chickens (40). Interestingly, the fitness change in FQ-R C. jejuni could not be attributed to compensatory mutations because no mutations other than the resistance-conferring C257T mutation were found in the gyrA and gyrB genes of the resistant strains (40).

Because FQ-R Campylobacter may still maintain fitness in the absence of antibiotic selection pressure, the reduced or discontinued antimicrobial use in food-producing animals may not necessarily result in an immediate decline in the frequency of FQ-R Campylobacter. For example, FQ-R Campylobacter was found in 40% of chicken products in two United States companies that had not used FQs for at least 1 year (41). Likewise, FQ-R Campylobacter remained for many rotations on Danish broiler farms that had stopped using FQ antibiotics for 4 years (42). In a recent study conducted by our group, it was found that the vast majority of dairy calves (26/30; 87%) were colonized by FQ-R C. jejuni even though they had no known previous exposure to FQ antibiotics (32). Similar findings were noted in a study with beef calves in which more than 60% of the Campylobacter isolates were resistant to at least one FQ antibiotic (e.g., nalidixic acid or ciprofloxacin) before treatment (43). A study conducted at commercial beef cattle confined feeding operations in Alberta, Canada found a relatively low level of resistance to FQs (^~5–7%, ciprofloxacin and nalidixic acid) in C. jejuni isolates upon feedlot arrival, but the resistance rate significantly increased (to ^~10–15%) after 60 days of maintenance period at some operations that did not use any FQ antibiotics (44). Interestingly, the same study showed a correlation between FQ resistance and genotype as certain subtypes of C. jejuni had higher rates of resistant isolates (44). Intriguingly, a longitudinal research on the incidence of antimicrobial-resistant Campylobacter in swine raised without antibiotics discovered a ciprofloxacin resistance rate of 17.1% in Campylobacter coli (45). These studies suggest that the fitness of FQ-R Campylobacter may contribute to the persistence of FQ resistance in the farm environment of various food-producing animals regardless of antimicrobial usage.

Very recently, we conducted a study with commercial dairy calves to evaluate the effect of subcutaneous (s.c.) administration of a single dose danofloxacin on the development of FQ resistance in C. jejuni in both healthy and BRD-induced calves (32). Data from that study showed that most of the calves were naturally colonized by a mixture of FQ-R and FQ-S C. jejuni strains (^~50% of each population) even though these animals were known not to be exposed to FQs previously per the farm records, suggesting that FQ-R strains may have a fitness advantage over FQ-S strains that allowed them to thrive in the gastrointestinal tract of cattle in the absence antibiotic selection pressure. To test this hypothesis, here we performed a series of in vitro experiments using both Mueller-Hinton (MH) broth and bovine fecal extract (in an attempt to mimic cattle intestinal tract), including the growth kinetics and competition as well as resistance development, using the FQ-R and FQ-S C. jejuni strains collected from the same study (32). It should be noted that natural carriage of Campylobacter in the intestine of healthy cattle is common and the organism is usually not associated with any overt disease in cattle (32).

Over the past decades, Campylobacter has developed perpetual resistance to clinically important antibiotics that are used for the treatment of severe cases of human infections, in particular to FQs, posing a threat to treatment efficacy in clinical cases (21, 38, 53). The global predominance of FQ-R Campylobacter may have been directly influenced by the frequency with which resistant mutants emerged in response to the selection pressure imposed by the use of antibiotics in both human medicine and veterinary settings (34, 54–59). Notoriously, the transmission and spread of antibiotic-resistant pathogens is not only affected by the emergence of resistant mutants in response to the selection pressure, but also influenced by the relative fitness of the drug-resistant organisms in the absence of selection pressure (39, 40, 60, 61). Cattle are a significant source of human Campylobacter infections, and there is a clear trend that FQ-R Campylobacter from cattle has become highly prevalent in recent years (7, 15, 62, 63). Even though the selection pressure (use of FQs in cattle) may have contributed to the expansion of FQ-R Campylobacter, the actual impact of this factor appears to be rather low (32, 44, 46, 64). In the current study, we examined the hypothesis that the fitness of FQ-R Campylobacter may have also played a role in the rise seen in FQ-R Campylobacter isolates of cattle origin. By using the FQ-R and FQ-S C. jejuni strains collected from calves from our recent study (32), we determined (a) in vitro growth kinetics of FQ-R and FQ-S strains in mono-cultures, (b) fitness of FQ-R C. jejuni without antibiotic selection pressure, and (c) examined the FQ resistance development in FQ-S C. jejuni by using different ciprofloxacin concentrations and initial bacterial cell densities.

Quinolone resistance typically develops at an average rate of 5 × 10⁻⁹ in Campylobacter, with this rate being as high as 5 × 10⁻⁷ in some strains (65, 66)_. When Campylobacter is exposed to FQs, ciprofloxacin-resistant mutants will likely arise if the cell population is large enough (>10⁶ CFU) (38), suggesting that Campylobacter possess a high mutation rate to FQ resistance. Our results are in line with an in vitro study conducted previously by our research group (50), in which FQ-resistance emerged readily from FQ-S C. jejuni at high (10⁷ and 10⁶ CFU/mL) initial bacterial cell densities when cultured in broth medium containing 4 μg/mL ciprofloxacin though no resistance developed when the initial concentration was10³ CFU/mL. Similar findings were also observed in the present study, as FQ-S C. jejuni developed resistance to ciprofloxacin (4 μg/mL) within 24 h of in vitro exposure at a relatively high initial bacterial cell density (10⁷ CFU/mL; Figure 4), while no colonies of resistant C. jejuni strain was detected at a low initial bacterial cell density (10⁵ CFU/mL). In agreement with these in vitro data, observations made in animals also indicate the importance of bacterial cell density in the development of FQ resistance in Campylobacter. For example, FQ resistance in Campylobacter emerges rapidly in chickens but not in cattle under FQ selection pressure, which can, at least in part, be explained by the fact that the organism typically colonizes the chicken intestine at a much higher magnitude (10^8–9 CFU/g feces) (32, 67) than it does the cattle intestinal tract (10^3–5 CFU/g feces) (67, 68). In chickens, as soon as 24 h after treatment with FQ antibiotics (enrofloxacin, sarafloxacin, or difloxacin; typically given in drinking water for 5 days), FQ-R Campylobacter mutants were found in the feces of treated birds and gradually colonized the intestinal tract at high densities (34, 56, 57). In big contrast, our recent study with calves showed that a single dose s.c. enrofloxacin treatment (7.5 or 12.5 mg/kg) did not result in any detectable level of FQ resistance development from FQ-S C. jejuni inhabiting the intestine (^~10^4–5 CFU/g feces) of calves (46). Similarly, therapeutic administration of neither oral (20 mg/kg daily for 7 days) nor subcutaneous (20 mg/kg daily for 1–7 days) enrofloxacin resulted in development FQ-resistance in C. jejuni NCTC 11168 following experimental inoculation of mice via oral gavage (69).

In the current study, FQ-S C. jejuni strains developed resistance to ciprofloxacin more readily when exposed to low levels of ciprofloxacin (2 and 4 μg/mL) compared with exposure to a high level of ciprofloxacin (20 μg/mL). Our data suggest that a high dose of ciprofloxacin is lethal to Campylobacter, whereas a low dose may favor the emergence of FQ-R C. jejuni from the susceptible strains. Even though it can be rather speculative and cannot be stated with a high degree of certainity, the notion of the mutant selection window (MSW) theory could provide a reasonable explanation for this observation. The range of antimicrobial concentrations known as the MSW ranges from the lowest concentration required to block the growth of wild-type bacteria (MIC) to the highest concentration needed to inhibit the growth of the least susceptible mutant (70). The upper boundary is also known as the mutant prevention concentration (MPC) (71). According to previous publications, the typical MIC of ciprofloxacin in FQ-R C. jejuni ranges from 4 to 16 μg/mL (34, 38, 51, 56, 72). Under this theory, the antibiotic becomes lethal to bacteria at concentrations over the MSW, and could no longer select for resistant strains. In the present study, the high level of antibiotic selection pressure might have reached/exceeded the MSW, and thus greatly reducing the emergence of FQ-R mutants in bovine fecal extract and MH broth (Figure 6). In line with this finding, a recent study conducted by our group in which calves were treated with a single dose s.c. enrofloxacin (7.5 and 12.5 mg/kg) found that the drug concentration in the rectal feces of calves had a median of 38–54 μg/g feces for enrofloxacin and 18–21 μg/g feces for ciprofloxacin within 12 h of the injection (73). Notably, in the same study, no FQ-R C. jejuni was detected in any of the calves that received enrofloxacin independent of the drug dose used (46). Similarly, we also showed that single dose s.c. danofloxacin treatment in calves colonized with both FQ-R and FQ-S C. jejuni resulted in high drug concentration in the rectal feces (median of 382–236 μg/g feces), but did not appear to lead to the development of de novo FQ resistance from susceptible strains (32). In contrast to cattle, in a study conducted with broiler chickens, the peak concentration of enrofloxacin was only around 2–4 μg/mL in the intestines of the birds during a standard multi-dose enrofloxacin water treatment, in which FQ-R C. jejuni developed soon after the treatment (72). Altogether, these results suggest that the low ciprofloxacin concentrations used in the current study and observed in the intestine of chickens (67, 68) may well have been within the MSW, while the high ciprofloxacin concentrations employed in this study and detected in calf feces (32, 46, 74) may have reached very close to or even exceeded the MPC.

The persistence of antibiotic-resistant Campylobacter is influenced by its ability to compete with antibiotic-susceptible strains; this competition dictates whether antibiotic-resistant Campylobacter prevails or declines in the absence of antibiotic selection pressure (38). In our study, FQ-R and FQ-S C. jejuni had comparable growth rates when individually cultured in either bovine fecal extract or MH broth (Figure 1). Next, we performed pairwise competition experiments to assess the fitness of FQ-R Campylobacter by co-culturing several FQ-R C. jejuni and FQ-S C. jejuni strains of cattle origin in either bovine fecal extract or MH broth containing no antibiotic. Interestingly, FQ-R strains did not have any fitness defect in mixed cultures in the absence of antibiotic selection pressure, but rather displayed a small, albeit significant, growth advantage over the FQ-S strains (Figures 2, 3). Importantly, similar observations were made in calves (from which the C. jejuni isolates used here were derived) in our recent study (32), where FQ-R resistant strains were found to coexist with FQ-S strains approximately in equal proportions in the intestinal tract with no antibiotic selection pressure present. Collectively, the findings from both in vivo and in vitro studies clearly indicate the overall fit nature of FQ-R C. jejuni of cattle origin and provide a plausible explanation, at least in part, for the rising trend seen in the prevalence of FQ-R Campylobacter in cattle over the past decade.

Our study has some limitations. For example, bovine fecal samples collected from calves in our previous study (46) were stored at −80°C for about 3 years before being used as a bovine fecal extract in the present study. Thus, the storage may have impacted the composition and microbiological properties of the fecal samples. Moreover, the freeze-thawing process (fecal samples were thawed to prepare the fecal extract and then frozen back until further use) may have caused some degree of degradation of the bovine fecal extract. Finaly, even though the bovine fecal extract may be a relevant growth medium to be employed in the experiments performed in the current study, it is important to emphasize that the degree to which it actually mimicked the gastrointestinal tract of cattle is likely to be quite small. Use of digesta instead of fecal extract could have offered more relevant results as it would better mimic the anaerobic conditions in the intestinal lumen. It also should be underlined that caution must be used when extrapolating from in vitro results to in vivo results and attempting to explain the data with unproven scientific concepts (e.g., the MSW theory).

Findings from the current study indicate that FQ-R and FQ-S C. jejuni strains of cattle origin had comparable growth kinetics and fitness in mono- and co-cultures, respectively. Moreover, FQ-S C. jejuni were shown to develop resistance to FQs more readily when exposed to low levels of ciprofloxacin and at a high initial bacterial cell density compared with exposure to a high level of ciprofloxacin and at a low level of initial bacterial cell density. The latter finding suggests that emergence of FQ-R C. jejuni mutants from susceptible strains in cattle is likely hampered by both the relatively low level (CFU/g feces) of bacterial colonization and the high level of antibiotic selection pressure in the intestinal tract following the FQ treatment. Altogether, FQ-R C. jejuni derived from cattle is found to compete well with FQ-S C. jejuni and does not display any fitness defect in the absence of antibiotic selection pressure, providing a plausible explanation for the high prevalence of FQ-R Campylobacter in cattle production.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving animals were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC-18-372) at Iowa State University. Written informed consent from the owners for the participation of their animals in this study was not required in accordance with the national legislation and the institutional requirements.

DG: methodology, data collection, statistical analysis, preparation of the manuscript. QZ: methodology, funding acquisition, supervision, and preparation of the manuscript. OS: methodology, funding acquisition, supervision, and preparation of the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by the Agriculture and Food Research Initiative’s Competitive Grant 2017-68003-26499 from the USDA National Institute of Food and Agriculture.

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.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.