Figures
Abstract
Antimicrobial resistance represents a major global threat to modern medicine. In vitro studies have shown that very low concentrations of drugs, as frequently identified in the environment, and in foods and water for human and animal consumption, can select for resistant bacteria. However, limited information is currently available on the in vivo impact of ingested drug residues. The objective of our study was to evaluate the effect of feeding preweaned calves milk containing antimicrobial drug residues (below the minimum inhibitory concentration), similar to concentrations detected in milk commonly fed to dairy calves, on selection of resistant fecal E. coli in calves from birth to weaning. At birth, thirty calves were randomly assigned to a controlled feeding trial where: 15 calves were fed raw milk with no drug residues (NR), and 15 calves were fed raw milk with drug residues (DR) by adding ceftiofur, penicillin, ampicillin, and oxytetracycline at final concentrations in the milk of 0.1, 0.005, 0.01, and 0.3 µg/ml, respectively. Fecal samples were rectally collected from each calf once a week starting at birth prior to the first feeding in the trial (pre-treatment) until 6 weeks of age. A significantly greater proportion of E. coli resistant to ampicillin, cefoxitin, ceftiofur, streptomycin and tetracycline was observed in DR calves when compared to NR calves. Additionally, isolates from DR calves had a significant decrease in susceptibility to ceftriaxone and ceftiofur when compared to isolates from NR calves. A greater proportion of E. coli isolates from calves in the DR group were resistant to 3 or more antimicrobial drugs when compared to calves in the ND group. These findings highlight the role that low concentrations of antimicrobial drugs have on the evolution and selection of resistance to multiple antimicrobial drugs in vivo.
Citation: Pereira RVV, Siler JD, Bicalho RC, Warnick LD (2014) In Vivo Selection of Resistant E. coli after Ingestion of Milk with Added Drug Residues. PLoS ONE 9(12): e115223. https://doi.org/10.1371/journal.pone.0115223
Editor: Vladimir N. Uversky, University of South Florida College of Medicine, United States of America
Received: September 24, 2014; Accepted: November 20, 2014; Published: December 15, 2014
Copyright: © 2014 Pereira et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper.
Funding: Research reported in this publication was supported by the Office of the Director, National Institutes of Health of the National Institutions of Health under Award Number T32ODO011000. Research reported here was also supported the United States Department of Agriculture (USDA), under Hatch award number NYC-478400. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health, the National Institute of Food and Agriculture (NIFA), or the USDA. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The rapid development of antimicrobial resistance in the past few decades is considered one of the greatest global threats to modern medicine [1], [2]. Acceleration in the rise of antimicrobial resistance has widely been blamed on misuse and overuse of drugs in humans and food-producing animals, emphasizing the need for the judicious use of antimicrobials [3], [4]. In the dairy industry diseases have a great impact, resulting in treatment expenses and production losses such as lower milk yield and withholding of milk due to presence of drug residues [5]. Most antimicrobial drugs used to treat cows result in the milk from these animals being withheld from sale (waste milk) because of the presence of drug residues above the tolerance concentration established by the U.S. Food and Drug Administration (FDA). A tolerance level is a concentration determined by the FDA at which residues of a substance present in a food will have no harmful effects on the human consumers of the food product [6]. To decrease production losses due to waste milk, 33% of dairy farms in the United States feed preweaned calves waste milk [7]. Feeding pasteurized waste milk instead of milk replacer to preweaned calves has been shown to result in an estimated saving of $0.69 per calf per day [8].
In a recent study conducted by our research group, waste milk fed to calves was collected from several dairy farms in central New York and screened for drug residues [9]. The three most common drug residues identified were the following β-lactams: ceftiofur, penicillin G, and ampicillin. Although these β-lactams drugs were observed at concentrations above the tolerance levels for human consumption of milk established by the FDA, their concentrations were still below the minimal inhibitory concentration established for E. coli by the Clinical Laboratory Standard Institute (CLSI) [10]. Minimum inhibitory concentrations (MICs) are defined as the lowest concentration of an antimicrobial that will inhibit the visible growth of a microorganism after overnight incubation. Because selection of resistant bacteria has traditionally been assumed to occur at concentrations between the MIC of the susceptible wild type population and that of the resistant bacteria, the potential impact of feeding waste milk to calves on drug resistance has not been evaluated extensively. However, recent studies have shown that selection of resistant bacteria can occur at low antimicrobial drug concentrations by selecting for resistant bacteria with compensatory mutations that counterbalance the decreased fitness cost caused by resistance [11], [12].
Studies have reported that antimicrobial drugs at concentrations below the minimal inhibitory concentration (sub-MIC) can stimulate mutagenesis and recombination, leading to bacterial adaptation to various stresses, including antimicrobial pressure [13], [14]. An increase in mutagenesis can also result in heterogeneous increases in MIC of the bacteria across a range of antimicrobials, irrespective of the drug target [15]–[17]. Exposure of bacteria to sub-MICs has also been shown to significantly increase the frequency of transfer on mobile genetic elements (MGE), which can result in increased dissemination of antimicrobial resistance genes between bacteria. In a recent study, E. coli carrying the multidrug resistance plasmid pB10 was exposed to concentrations of tetracycline as low as 0.01 µg/mL, what was shown to significantly increase the transfer rate of this plasmid to enteric bacteria present in activated sludge from a wastewater treatment plant [18]. Furthermore sub-MICs of antimicrobial drugs could potentially increase the spread of antibiotic resistance genes (ARG) in the environment, which are currently considered an emerging environmental contaminant in lakes, rivers, and at drinking water treatment plants, representing a health hazard for human and animal health [19].
The development of a commensal microbiota carrying resistance to antimicrobials represents an important reservoir for resistance genes (resistome) that could be transferred to pathogenic bacteria, leading to the emergence of clinically problematic strains [20]. Moreover, antimicrobial treatment failure in a disease caused by pathogenic bacteria susceptible to a drug used for treatment could occur as a consequence of a resistant commensal microbiota through an umbrella effect. This umbrella effect was shown in a study by Perlin et al. (2009) while examining if β-lactamase-producing E. coli could protect ampicillin-sensitive cohorts of other species, particularly species that can cause human disease [21]. They observed that β-lactamase-secreting E. coli allowed for survival of a large number of ampicillin-sensitive cohorts of Salmonella enterica serovar Typhimurium. The Salmonella survivors remained sensitive to ampicillin when re-plated onto solid medium and there was no evidence of gene transfer, suggesting a protective effect by β-lactamase-secreting E. coli.
Few studies have evaluated the impacts of ingesting sub-MICs of antimicrobial drugs in an in vivo model, and many questions remain about the effects on the enteric microbiota and selection of antimicrobial resistance [15]. The objective of our study was to evaluate the effect of feeding preweaned dairy calves raw milk with residual concentrations (sub-MICs) of ampicillin, ceftiofur, penicillin, and oxytetracycline from birth to weaning on the selection of resistant E. coli in the feces.
Material and Methods
Ethics statement
Fecal samples were collected from calves (Bos taurus) that were housed on Cornell University facilities. The research protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Cornell University (Protocol number: 2012–0090).
Study design and sample collection
Randomized controlled feeding trials were conducted at the College of Veterinary Medicine, Cornell University (Ithaca, NY, USA) from June 2013 to March 2014. Three feeding trials were completed with a total of 10 male calves in each trial, with 5 calves belonging to each treatment group. All thirty calves enrolled in the trials were purchased from a local dairy farm and enrolled in the study on their date of birth. Control calves (n = 15) were enrolled in the trials concurrently with test calves (n = 15). At least one author in the study was involved in all calf collection at the farm. Upon collection, a physical examination was performed and calves were weighed. Additionally, the assignment of calves to study groups was done at the farm by pairing calves born on the same day by weight and using a coin toss to randomly allocate a calf as either a test or control. Once a calf was assigned to a treatment group it received an identification tag, which was placed in the right ear. All calves were fed 2–4 liters of maternal colostrum from individual cows within the first 4 h of life. Colostrum fed to calves in both milk feeding treatments originated from the dairy farm where calves were collected. At this farm cows were under the same management. After feeding of colostrum, calves were transported from the source farm to Cornell University.
Calves were individually housed in concrete box stalls to prevent contact between calves. Blood samples were collected from each calf in the first 24–48 hours of life, and the serum total protein was measured to assess adequacy of passive transfer. Control calves were fed raw milk without the addition of antimicrobial drugs (NR), and test calves were fed raw milk with the addition of low concentrations of ceftiofur, penicillin, ampicillin and oxytetracycline (DR). All calves were bucket fed one gallon of raw whole milk twice a day from birth to 6 weeks of age. Feedings occurred once in the morning and once in the afternoon with approximately 12 hour intervals between feedings. A non-medicated pelleted calf starter (18% crude protein, 3% crude fat, 8% crude fiber; DuMOR Calf Starter, Tractor Supply Co., Brentwood, TN) was offered from day 7 until day 42 of life up to a maximum of 1 kg/day. To prevent cross-contamination between calves, each calf stall had dedicated equipment and supplies, and all study personnel used personal protective equipment when entering each calf stall, which was changed after exiting the calf stall.
Single-use gloves were used to collect rectal fecal samples from each calf once a week starting at birth, prior to the first feeding in the trial (pre-treatment), until 6 weeks of age. No calves required treatment with therapeutic antimicrobial drugs during the trial.
Spiking milk with drug residues
Raw milk used to feed calves was collected daily from the Cornell University College of Veterinary Medicine Dairy Farm. Antimicrobial stock solutions used to spike milk were prepared one week prior to each calf trial. Stocks were prepared by diluting powdered drugs in distilled water to a concentration of 100 µg/mL for ampicillin sodium salt, 1,000 µg/mL for ceftiofur sodium, 50 µg/mL for penicillin G sodium, and 3,000 µg/mL for oxytetracycline hydrochloride. Individual sterile cryovials with 2.28 ml of each stock solution for each antimicrobial drug were stored at −80°C until used [22]–[24]. Individual antimicrobial stocks (instead of a cocktail of drugs) for each antimicrobial drug were prepared to avoid potential interactions between different antimicrobials before adding it to the milk.
Milk containing drug residues fed to DR calves was prepared twice a day 10 to 20 minutes prior to feeding. For each drug, a tube containing 2.28 ml of antimicrobial stock solution was thawed at room temperature and added to a batch of 22.8 liters of raw milk which was stirred for 1 minute at approximately 400 RPM prior to feeding to calves. The final concentration of each antimicrobial drug in the milk fed to DR calves was calculated to be: 0.01 µg/ml of ampicillin sodium, 0.1 µg/ml of ceftiofur sodium, 0.005 µg/ml of penicillin G sodium, and 0.3 µg/ml of oxytetracycline hydrochloride. Because limited information is currently available on the impacts of waste milk on selection of resistance, drugs were pooled in the milk (instead of tested individually) to simulate what is expected to be observed on dairy farms. Milk fed to NR calves was also stirred for 1 minute at approximately 400 RPM prior to feeding, but without the addition of any antimicrobial and by using clean dedicated equipment to avoid drug contamination from milk fed to DR calves. After each feeding, equipment was thoroughly cleaned separately by milk feeding treatment group using a dedicated brush and hot water, liquid soap, and sodium hypochlorite (The Clorox Co., Oakland, CA). The selection of drugs and the concentration added to the milk was based on an article published by our research group where we screened waste milk withheld for sale at dairy farms in central New York [9]. In that study the most prevalent drugs detected by LC-MS/MS were ceftiofur (mean ± SE concentration = 0.151±0.042 µg/mL), penicillin G (mean ± SE concentration = 0.008±0.001 µg/mL), and ampicillin (mean ± SE concentration = 0.472±0.43 µg/mL). In addition, one sample had detectable concentrations of oxytetracycline (0.01 µg/mL). Because of the high frequency use of this drug in dairy cattle, it was also added to the milk fed to DR calves [25]. Moreover, the tolerance and safe levels for drug residues in raw milk used for human consumption determined by the Federal Department of Agriculture (FDA) were also used to determine at what concentration the selected antimicrobial drugs should be added to the milk.
Raw milk used in the trial was screened daily for drug residues prior to its use in the study. Two commercial tests were used: New SNAP Beta-lactam Test Kit (IDEXX Laboratories Inc., Westbrook, ME) which detects penicillin (LOD = 0.005 ppm), ampicillin (LOD = 0.01), amoxicillin (LOD = 0.01 ppm), cephapirin (LOD = 0.02 ppm), and ceftiofur (LOD = 0.1 ppm) residues in raw milk; and SNAP Tetracycline Test Kit (IDEXX Laboratories Inc., Westbrook, ME) which detects tetracycline (LOD = 0.05 ppm), oxytetracycline (LOD = 0.05 ppm), and chlortetracycline (LOD = 0.1 ppm) residues in raw milk. During the entire trial all milk samples from the dairy farm tested negative for drug residues.
Milk spiked with antimicrobial drugs fed to DR calves was screened weekly for drug residues using the New SNAP Beta-lactam Test Kit and the SNAP Tetracycline Test Kit. As expected, all spiked milk samples tested positive. There are no current commercial test kits for detecting drug residues in cow colostrum. Additionally the commercial kits available for milk are not appropriate for colostrum because the high fat percentage and viscosity of colostrum compared to milk can interfere in the results of the test. Therefore colostrum fed to calves was not screened for drug residues. However, colostrum fed to calves in both milk feeding treatments originated from one dairy farm where cows were under the same management.
Bacterial isolation, culture, and antimicrobial susceptibility testing
Individual fecal samples were streaked onto MacConkey agar plates on the day of collection and incubated overnight at 37°C. Up to three E. coli colonies were selected and stored in Luria-Bertani broth containing 20% glycerol at −80°C, as previously described [26]. All fecal samples had three E. coli colonies, except for a few samples collected at week 0, which were meconium samples. All samples collected at week 0 had at least one E. coli isolate isolated from feces. A total of 609 E. coli isolates were cultured from rectal fecal samples and tested for antimicrobial susceptibility.
Antimicrobial susceptibility of E. coli isolates was tested against a modified National Antimicrobial Resistance Monitoring System (NARMS) panel of 12 antimicrobial drugs. The susceptibility testing was done using a Kirby–Bauer disk diffusion agar assay in accordance with the guidelines published by the Clinical and Laboratory Standards Institute (CLSI) and methodology previously described [26]–[28]. Internal quality control was performed by inclusion of E. coli ATCC 25922, which was previously determined to be pansusceptible, and a previously characterized in-house E coli isolate known to have a blaCMY-2 gene and to be resistant to 9 of the antimicrobials tested. Antimicrobial susceptibility for all isolates was assessed using the following panel: ampicillin 10 µg, cefoxitin 30 µg, ceftiofur 30 µg, ceftriaxone 30 µg, chloramphenicol 30 µg, ciprofloxacin 5 µg, gentamicin 10 µg, nalidixic acid 30 µg, neomycin 30 µg, streptomycin 10 µg, tetracycline 30 µg and trimethoprim-sulfamethoxazole 23.75/1.25 µg. Results of the disk diffusion test for the internal quality control strains were within the anticipated standards. Susceptibility of the isolates to antimicrobial drugs was categorized as susceptible, intermediate, or resistant (SIR) by measuring the inhibition zone diameter according to interpretive criteria and breakpoints established by the CLSI guidelines [10].
Statistical analyses
Descriptive analysis for the SIR distribution of E. coli isolates by antimicrobial drug for each feeding treatment group was done using PROC FREQ in SAS (SAS Institute Inc., Cary, NC). Descriptive analysis of E. coli resistance phenotypes and the proportion of isolates pansusceptible or resistant to 3 or more antimicrobial drugs for each weekly sampling was also done using PROC FREQ in SAS. In this study, phenotypic multidrug resistance was defined as isolates having resistance to ≥3 antimicrobial agents. Descriptive analysis for the comparisons of the average weight gain during the 6 weeks of the calf trial between NR calves and DR calves was conducted in SAS using PROC GLIMMIX.
To evaluate the effects of calf milk feeding treatment over time in weeks on the proportion of resistant E. coli per calf for each of the 12 antimicrobials tested, multivariable mixed logistic regression models were fitted to the data using the GLIMMIX procedure of SAS. The independent variables milk feeding treatment, time in weeks of sampling, and interactions were included in all models. The effect of animal identification nested within trial number was controlled in the models as a random effect. Adjusted probabilities (for all variables and interactions offered to each model) for antimicrobial resistance were obtained using the LSMEANS statement. This statistical model was also used to evaluate the effect of milk feeding treatment over time in weeks on the proportions of multidrug resistant isolates, where the only difference was that the dependent variable was the binary variable for classification of E. coli as resistant to 3 or more antimicrobial drugs or not.
For antimicrobial drugs with more than 10% of E. coli isolates within each treatment group classified as intermediate according to CLSI breakpoints, an additional analysis using the GLIMMIX procedure of SAS was conducted. In this model, the dependent variable was a categorical variable that classified isolates as susceptible or nonsusceptible (intermediate or resistant). This binomial variable was used in the analysis for two main reasons: to reduce erroneous analysis of the data caused by a high number of isolates classified as intermediate and where the CLSI breakpoints may not correlate well to acquisition of resistant genes, and to focus the analysis on the presence or absence of isolates that are fully susceptible to the antimicrobial drugs tested. The independent variables milk feeding treatment, time in weeks of sampling, and interactions were included in all models. The effect of animal identification nested within trial number was controlled in the models as a random effect. Adjusted probabilities (for all variables and interactions offered to each model) for antimicrobial resistance were obtained using the LSMEANS statement.
To determine if there was a statistical difference between the diameter of the inhibition zones of E. coli at the calf level for each treatment group, generalized linear models were fitted to the data using the GLM procedure of SAS. For each antimicrobial drug, the calf level mean zone diameter value was calculated for each week using Proc Means in SAS. For the linear regression analysis, the independent variables milk feeding treatment, time in weeks of sampling, and interactions were included in all models. The effect of trial number was controlled in the models as a random effect. Adjusted means (for all variables and interactions offered to each model) for inhibition zone diameter were obtained using the LSMEANS statement.
Results
Descriptive data
The mean birth weight for the NR calves was 45.3 kg (range: 31–53 kg) and for the DR calves the mean was 44 kg (range: 34–53 kg). Mean serum total protein (g/dL) for blood samples collected 24–48 hrs after birth was 5.5 (range: 5.0–6.2 g/dL) for the DR calves and 5.4 (range: 5.1–6.0 g/dL) for the NR calves. No calf in the study had failure of passive transfer (serum total protein <5 g/dL). Average weight gain during the 6 weeks of the calf trials was 1.1lbs (95% C.I. 0.9–1.4) for NR calves and 1.35lbs (95% C.I. 1.1–1.6) for DR calves. There was no significant difference in the average weight gain during the 6 weeks of the calf trials between NR calves and DR calves (P-value = 0.3).
Antimicrobial Resistance in E. coli isolates
SIR distribution of E. coli and mean disk diffusion zone diameter by milk feeding treatment group are shown on table 1. Additionally, fig. 1 has a histogram for the zone diameter of all E. coli cultured from fecal samples and tested for antimicrobial susceptibility to three of the four drugs added to the milk of DR calves (ampicillin, ceftiofur and tetracycline). No difference was observed in the proportion of nonsusceptible isolates of NR and DR for the pre-treatment sample (week 0) for any of the antimicrobial drugs tested. For ampicillin (weeks 1 to 6), cefoxitin (weeks 1 to 6), ceftiofur (weeks 2, 5 and 6), streptomycin (weeks 3 and 5) and tetracycline (weeks 1 to 6), there was a significantly greater proportion of resistant isolates from DR calves when compared to NR calves (Fig. 2). Additionally for ceftiofur and ceftriaxone, which had more than 10% of isolated E. coli within each treatment group classified as intermediate, we observed a significantly greater proportion of non-susceptible isolates from DR calves when compared to NR calves (Fig. 3).
Data is stratified by milk feeding treatment group. Labeled shaded areas on graphs correspond to the zone diameter SIR classification according to Clinical and Laboratory Standards Institute (CLSI) breakpoints. The CLSI minimum inhibitory concentration breakpoints for E. coli classified as susceptible (MICs) and resistant (MICR) are displayed in graphs. 1. Calves fed raw milk without the addition of sub-MICs of antimicrobial drugs. 2. Calves fed raw milk with the addition of sub-MICs of ceftiofur, penicillin, ampicillin, and oxytetracycline from birth to 6 weeks of age.
Error bars represent 95% confidence interval of the least square mean. 1. Calves fed raw milk without the addition of sub-MICs of antimicrobial drugs. 2. Calves fed raw milk with the addition of sub-MICs of ceftiofur, penicillin, ampicillin, and oxytetracycline from birth to 6 weeks of age. *Sampling weeks where the proportion of resistance was significantly different between NR and DR.
Error bars represent 95% confidence interval of the least square mean. 1. Calves fed raw milk without the addition of sub-MICs of antimicrobial drugs. 2. Calves fed raw milk with the addition of sub-MICs of ceftiofur, penicillin, ampicillin, and oxytetracycline from birth to 6 weeks of age. *Sampling weeks where the proportion of resistance was significantly different between NR and DR.
For both treatment groups, calves in the study had a peak in the proportion of resistant E. coli at one to two weeks of age for the majority of the antimicrobial drugs tested (Fig. 2). The inhibition zone diameter for ampicillin, cefoxitin, ceftiofur, ceftriaxone, and tetracycline was significantly wider for these drugs in isolates from DR calves when compared to NR calves (Fig. 4).
Error bars represent 95% confidence interval of the least square mean. 1. Calves fed raw milk without the addition of sub-MICs of antimicrobial drugs. 2. Calves fed raw milk with the addition of sub-MICs of ceftiofur, penicillin, ampicillin, and oxytetracycline from birth to 6 weeks of age. *Sampling weeks where the mean zone diameter was significantly different between NR and DR.
Distribution of multidrug resistant E. coli and resistance phenotypes
Of the 270 E. coli isolates obtained from DR fecal samples collected during weeks 1 to 6, 6% were pansusceptible (isolates non-resistant to all antimicrobial drugs tested) and 84% were resistant to three or more antimicrobials. The most common resistance phenotype from DR calves was ampicillin-cefoxitin-ceftiofur-tetracycline, which was present in 17% of isolates (Table 2). Of the 270 E. coli isolates from NR fecal samples collected during weeks 1 to 6, 46% were pansusceptible and 37% were resistant to three or more antimicrobials. The most common resistance phenotype from NR calves was ampicillin-cefoxitin-ceftiofur-neomycin-streptomycin-tetracycline, which was present in 10.3% of isolates (Table 2).
From week 1 to week 6, a significantly greater proportion of isolates from DR calves was resistant to 3 or more antimicrobial drugs compared to isolates from NR calves (Fig. 5). No significant difference was observed in the proportion of isolates resistant to 3 or more antimicrobial drug between NR calves and DR calves for week 0 (the pre-treatment samples).
Error bars represent 95% confidence interval of the least square mean. 1. Calves fed raw milk without the addition of sub-MICs of antimicrobial drugs. 2. Calves fed raw milk with the addition of sub-MICs of ceftiofur, penicillin, ampicillin, and oxytetracycline from birth to 6 weeks of age. *Sampling weeks where the the proportion of resistant isolates resistance to 3 or more antimicrobial drugs was significantly different between NR and DR.
Discussion
The presence of low concentrations of antimicrobials and of antimicrobial resistance genes in natural environments such as rivers and soils, and in foods and water for human and animal consumption have been widely reported [9], . Our understanding of the impacts of low concentrations of drugs on antimicrobial resistance is still limited, although in vitro studies have indicated that selection and maintenance of resistant bacteria can occur [32]. The present study observed that calves fed milk with added antimicrobial drugs at very low concentrations had a significantly higher proportion of isolates resistant to three β-lactams drugs, an aminoglycoside drug, and a tetracycline drug when compared to calves fed milk without the addition of drugs (Fig. 2). The traditional concept of antimicrobial resistance has been suggested to be the selection of resistant bacteria when exposed to drugs at concentrations above the minimum inhibitory concentration (MIC) of the susceptible bacteria [33]. Above the MIC the susceptible strain will be out-competed by strains with phenotypic resistance to that drug. According to the CLSI, the MIC breakpoint for resistant bacteria in the Enterobacteriaceae family (which includes E. coli) for tetracycline, ceftiofur, and ampicillin is 16 µg/mL, 8 µg/mL, and 32 µg/mL, respectively (Fig. 1). The final concentration of tetracycline, ceftiofur, and ampicillin in the milk fed to DR calves was 0.3 µg/mL, 0.1 µg/mL, and 0.01 µg/mL, respectively. Therefore the concentration of tetracycline, ceftiofur, and ampicillin in the milk fed to DR calves was respectively 53, 80, and 3200 times below the MIC breakpoint for resistant E. coli [34]. Our findings challenged the conventional concentration window for selection of resistance and indicate that exposure of commensal E. coli to sub-MIC of antimicrobials drugs in preweaned dairy calves can cause selection of resistant bacteria.
In a study by Gullberg et al. (2011), isogenic pairs of susceptible and tetracycline resistant (Tn10dTet) S. enterica serovar Typhimurium were grown in the presence of tetracycline at concentrations many times below the MIC [32]. Their findings were similar to ours, observing that tetracycline concentrations up to 60 times below the MIC for the susceptible bacteria reduced the growth rate in the susceptible strain without any apparent effect on the resistant strain. These findings suggest that for most antimicrobial drugs, susceptible bacteria will experience a reduction in growth even when exposed to concentrations many times below the MIC. This sub-MIC selection window is commonly called the minimum selective concentration (MSC), and is defined as the lowest concentration of an antimicrobial drug that still selects for a given resistance determinant [35]. The MSC is directly associated with the fitness cost of the resistance determinant conferring competitive advantage to the resistant strain in relation to the susceptible strains. This means that if the fitness cost of resistance is too elevated, its presence will not necessarily result in a lower MSC for the resistant bacteria. For example, in the same study by Gullberg et al. (2011), antimicrobial sensitivity tests were conducted for ciprofloxacin using a pair of isogenic E. coli with either the mutation gyA(D87N) or gyA(S83L) [32]. Although both mutations select for resistance to ciprofloxacin, the MSC for E. coli with gyA(D87N) and gyA(S83L) was 0.0025 ug/ml and 0.00001 ug/ml, respectively. The cause of is difference is that the fitness cost for the mutation gyA(D87N) is 3%, while for gyA(S83L) it is 0.2%. These findings reinforce the argument that the effect of resistance determinants on fitness is a major factor affecting the MSC [35]. Moreover, this raises concerns about the exposure of a microbiota to antimicrobial drugs at sub-MIC, as it could cause selection of a sub-population of strains carrying resistance with low fitness cost that could contribute to the persistence of these bacteria in the commensal microbiota, even after exposure to antimicrobial drugs had ceased.
More than 35 percent of isolates in both feeding treatment groups were classified as intermediate for the third-generation cephalosporin ceftriaxone (Table 1). The intermediate zone is considered a buffer zone that should prevent small, uncontrolled, technical factors from causing major discrepancies in interpretation, and typically only a limited proportion of isolates are expected to be allocated in this category. Classification of many isolates as intermediate has also been reported by other researchers who suggested that species-specific breakpoints may be needed for these drugs to allow accurate prediction of the MIC equivalent of the zone diameters [36]. Although DR calves did not have a significantly higher proportion of isolates resistant to ceftriaxone when compared to ND calves (Fig. 2), they did have a significantly higher proportion of isolates nonsusceptible (isolates classified as intermediate or resistant) to ceftriaxone when compared to isolates from ND calves (Fig. 3). A lack of a species-specific clinical breakpoint for ceftriaxone, resulting in a high number of isolates classified as intermediate, could have been the reason for this discrepancy. The decrease in susceptibility to ceftriaxone could also be noted by a significantly narrower zone diameter for this drug in isolates from DR calves when compared to ND calves (Fig. 4). In our study, we compared the average zone diameter (ZD) of isolates at the calf level for each antimicrobial to allow detection of changes in susceptibility within a range that would not lead to changes in SIR classification. In this analysis, significantly larger zone diameters for an antimicrobial drug did not necessarily mean that one group had more clinically resistant isolates, but instead showed a reduced susceptibility to that drug. Resistance to third-generation cephalosporins is mainly conferred by the AmpC-like CMY β-lactamases and by the CTX-M β-lactamases [37], [38]. The selection of a sub-population of E. coli isolates less susceptible to third generation cephalosporins from feeding calves milk with sub-MICs is of concern to human health as drugs in this class have been labeled as critically important to human medicine by the World Health Organization (WHO) [39].
A higher proportion of resistant isolates from DR calves for three β-lactams was not unexpected, since three out of four of the drugs added to the milk of DR calves belonged to this drug class. This also applies to tetracycline resistance, because oxytetracycline was added to the milk with drug residues. Mechanism of resistance to β-lactams and tetracycline is commonly due to plasmid-mediated acquisition of resistance genes [40]–[43]. In addition to the selection pressure for resistant bacteria, sub-MIC concentrations of antimicrobials can increase rates of homologous recombination and horizontal gene transfer of resistant genes [44]–[46]. For example, exposure of Staphylococcus aureus to sub-MICs of β-lactams was shown to stimulate a 1000-fold increase in the transfer of a plasmid conveying tetracycline resistance [47]. Moreover sub-MICs of antimicrobials in DR calves could have increased resistance in isolates from DR calves by favoring the transfer rate of plasmids between commensal E. coli, amplifying the spread of resistance to β-lactams and tetracyclines between microbes in the gut.
Although no aminoglycoside was added to their milk, DR calves had a higher proportion of E. coli resistant to streptomycin at weeks 3 and 5 when compared to NR calves (Fig. 2). Target modification by mutation of the 16S rRNA or ribosomal proteins can cause resistance to aminoglycosides [48], [49]. Additionally, mutation in genes controlling efflux pumps for aminoglycoside drugs can result in overexpression, reducing susceptibility to aminoglycosides [50]. Similarly, resistance to β-lactams can also occur by mutations that increase the chromosomal expression level of β-lactamase (bla) genes, which can extend the substrate specificity to important 3rd and 4th generation cephalosporins [51]. Sub-MICs of antimicrobial drugs have been shown to increase mutagenesis, which has been correlated with an increase in reactive oxygen species and induction of the SOS response in bacteria [13], [17]. A recent study by Gutierrez et al. (2013) that exposed E. coli to sub-MICs of ampicillin elucidated some of the mechanisms that result in increased mutagenesis [52]. This study showed that sub-MICs of ampicillin induced mutagenesis by the combined activities of both the error-prone DNA polymerase and the normal replicative DNA polymerase in the absence of an adequate mismatch repair system. The mismatch repair system is a DNA-repair system present in most organisms that recognizes and repairs erroneous DNA replication and recombination, and DNA damage [53]. Moreover the higher proportion of isolates resistant to streptomycin in DR calves could have been caused by higher mutagenesis induced by sub-MICs of antimicrobial drugs. Alternatively, selection of resistance to streptomycin could have been caused by a higher occurrence of a class 1 integron in E. coli from DR calves. Integrons are mobile DNA elements capable of encoding multiple resistance genes at once [54]. Characterization of class 1 integron-mediated antimicrobial resistance among E. coli from calves has revealed its capability to integrate multiple antimicrobial resistance genes including β-lactams and aminoglycosides [55]. Sub-MICs of antimicrobials could result in the selection of E. coli with class 1 integrons in the DR group, causing co-selection of resistance to streptomycin.
DR calves had a significantly greater proportion of E. coli resistant to 3 or more antimicrobial drugs when compared to NR calves (Fig. 5). Multidrug resistance caused by sub-MICs of antimicrobials has been shown to be conferred through many of the mechanisms already discussed such as increase in horizontal gene transfer (e.g. mediated through plasmids), mutagenesis, co-selection through mobilization of integrons with cassettes carrying antimicrobial resistant genes, and activation and overexpression of bacteria efflux systems [47], . Additionally, an in vitro study by Kohanski et al. (2010) observed that exposure of E. coli to sub-MICs of bactericidal antimicrobials could lead to multidrug resistance via radical-induced mutagenesis [15]. Moreover, a study by Braoudaki et al. (2004) observed a high degree of cross-resistance to a range of antimicrobials and biocides in Salmonella serovar Virchow and E. coli O157 when strains were repeatedly exposed to sub-MICs of antimicrobials agents [57]. Although their study could not determine the specific mechanism contributing to this adaptive resistance, they suggested that the most likely cause was the presence of active multidrug efflux pumps. Resistance to multiple antimicrobials has been linked to mutations in drug-efflux systems such as the AcrAB multidrug efflux pump observed in E. coli [58], [59], as well as to mutations in the transcription factors controlling these systems, such as SoxS [60], MarA [61] and ROB [62]. These findings support the important role that sub-MICs of antimicrobial drugs have on the selection of multidrug resistant bacteria. Additionally, the small percentage of pansusceptible isolates in DR calves (6%) compared to NR calves (46%) is a reflection of the effect that sub-MICs of antimicrobials has on the promotion of resistance genes between commensal E. coli.
Independent of feeding treatment group, calves in the study had a peak in the proportion of resistant E. coli at one to two weeks of age for the majority of the antimicrobial drugs tested (Fig. 2). Similar findings were observed in a study by Berge et al. (2005), where calves 2 weeks of age were more likely to have multidrug resistant E. coli compared to calves 4 and 6 weeks old (OR = 29.8 and OR = 16.4, respectively). The introduction of resistant bacteria into the enteric microbiota depends on their ability to effectively compete with the indigenous microbiota [63]. Young calves lack a developed and diverse intestinal microflora, which could reduce the degree of protection against colonization by bacteria with a higher fitness cost, such as antimicrobial-resistant and pathogenic enteric bacteria [64]. Exposure of this tenuous enteric microbiota to sub-MICs of antimicrobials could increase the vulnerability of this microbiota to colonization by resistant bacteria. This was observed in our study, where from weeks one to six, calves in the DR group had a significantly higher proportion of isolates resistant to ampicillin, cefoxitin, and tetracycline compared to NR calves. However, because our study did not used quantitative methods, we could not determine whether this resulted primarily from decreased numbers of susceptible E. coli, an increase in the absolute number of resistant strains, or a combination of the two. Feeding calves milk with added sub-MICs of drugs since birth could have resulted in stress or perturbation beyond what the calves gastro intestinal microbiota could tolerate before its trajectory was changed towards a different equilibrium state [65]. This may result in a longitudinal change in the microbiota composition with an increase in the proportion of microbes carrying antimicrobial resistance genes that allow them to survive and reproduce under this continuous antimicrobial drug challenge. To test this hypothesis high throughput metagenomic methods would be needed. However, results from commensal E. coli suggest this as observed by a higher proportion of E. coli resistant to 3 or more antimicrobial drugs in the DR calves when compared to NR calves (Fig. 5).
Although at different rates, for most antimicrobial drugs in both milk feeding treatment groups the proportion of resistant E. coli decreased over the weeks after week 2. Calves are known to have a higher proportion of resistant E. coli when compared to older animals such as heifers and cows, which have an mature enteric microbiota [66], [67]. Although there were a greater proportion of resistant E. coli until the last week sampled in DR calves compared to NR calves, it is uncertain if as calves mature, animals from DR will continue to be hosts for a greater proportion of resistant E. coli when compared to animals from NR. Our study did not monitor animals beyond weaning and because of that we are unable to answer this question. However, studies in humans have revealed that treatment with antimicrobials can result in permanent changes to the microbiota that may result in the persistence of antimicrobial resistant bacteria [53], [68], [69]. Additional studies are needed to evaluate the impacts of feeding waste milk on the composition of the developing enteric microbiota of preweaned calves as they mature.
Based on the findings of our study, feeding waste milk to preweaned dairy calves is a practice that has the potential to at least temporarily increase the selection of resistance to drug of critical importance to both animal and human health. To eliminate the risk originating from feeding calves waste milk, an initial option would be to discontinue this practice at dairy farms. Some disadvantages include the possibility of polluting the environment with drug residues by disposing of the waste milk, disposal of a nutritious food source, and increase in the costs of feeding preweaned dairy calves. A reasonable assessment to determine whether to continue or discontinue feeding waste milk to dairy calves must take into account the health hazards to animals and humans, and economic advantages and disadvantages. An alternative to reduce the negative impacts of drug residues present in waste milk fed to calves is to use methods to degrade these drugs to concentrations below the minimum selective concentration (MSC) which may decrease the selection for a given resistance determinant. Some potential methods that could be used to reduce the concentration of drug residues in waste milk include heat treatment, time of storage, ion catalysis, and electrochemical methods. Heat treatment (120°C for 20 min) of milk containing beta-lactam drugs has shown to degrade 47% of amoxicillin, 84% of ampicillin, 53% of cloxacillin and 61% of penicillin G [70]. Biodegradation of ceftiofur has been shown to increase with the increase of temperature, with optimal degradation temperatures between 35°C and 45°C [71]. A study by Riediker et al. (2004) showed that penicillin G, amoxicillin, and ampicillin spiked in milk at a concentration of 0.01 ppm and stored for 6 days at 4°C suffered degradation in most cases of more than 50% of initial concentration [72]. Raw milk spiked with ceftiofur at a concentration of 0.01 µg/mL and stored for 14 days at 4°C retained 90 to 100% of the initial concentration, indicating that storage may not be the best option for reducing concentrations of cephalosporins in the milk [73]. Degradation of beta-lactams has also been shown in the presence of various metal ions where the ions catalyze the inactivation of the hydrolytic opening of β-lactams [74]–[76]. Electrochemical oxidation of raw milk with an initial concentration of oxytetracycline of 100 mg/ml has resulted in an 83% reduction of this drug after a 6h treatment [77]. Pasteurization of waste milk used to feed preweaned calves has been a recommended strategy to reduce bacterial contamination and limit the spread of disease, and furthermore may result in reduced concentration of certain antibiotics [78]. However additional studies must be conducted to measure the effectiveness of methods to reduce the concentration of drugs in the milk and take into account factors such as initial concentration of drug residues usually present in waste milk, final concentration of drugs after method application, and feasibility and costs of various methods.
In conclusion, feeding calves milk with residual concentrations (sub-MICs) of antimicrobial drugs from birth to weaning resulted in a higher proportion of E. coli isolates resistant to multiple antimicrobial drugs. Treatment of calves with sub-MICs of ampicillin, penicillin, ceftiofur, and tetracycline resulted in a greater proportion of E. coli resistant to ampicillin, cefoxitin, ceftiofur, streptomycin, and tetracycline. Additionally, isolates from DR calves had a significant decrease in susceptibility to ceftriaxone and ceftiofur when compared to isolates from NR calves. A significantly greater proportion of E. coli in DR calves was multidrug resistant when compared to isolates in NR calves. These findings stress that the exposure of E. coli to sub-MICs of antimicrobials in the milk has an important effect on the frequency of antimicrobial resistance in preweaned dairy calves.
Acknowledgments
For their invaluable contribution to the study, the authors thank the following team members: Alison Keggan, Carrie Harkins, Danique Wortel, Elizabeth Lennon, Jamie Ng, Nicole Irizarry, Stephanie Ruback, Theresa Hubbell, and Thomas Cully.
Author Contributions
Conceived and designed the experiments: RVP LDW RCB. Performed the experiments: RVP JDS. Analyzed the data: RVP LDW RCB. Contributed reagents/materials/analysis tools: RVP LDW. Wrote the paper: RVP JDS LDW RCB.
References
- 1. Huttner A, Harbarth S, Carlet J, Cosgrove S, Goossens H, et al. (2013) Antimicrobial resistance: A global view from the 2013 world healthcare-associated infections forum. Antimicrob Resist Infect Control 2 31-2994-2-31. eCollection 2013.
- 2.
World Health Organization (WHO) (2014) Antimicrobial resistance: Global report on surveillance 2014.
- 3. Flanders SA, Saint S (2014) Why does antimicrobial overuse in hospitalized patients persist? JAMA Intern Med 174:661–662.
- 4. Maron DF, Smith TJ, Nachman KE (2013) Restrictions on antimicrobial use in food animal production: An international regulatory and economic survey. Global Health 9 48-8603-9-48.
- 5. Cha E, Bar D, Hertl JA, Tauer LW, Bennett G, et al. (2011) The cost and management of different types of clinical mastitis in dairy cows estimated by dynamic programming. J Dairy Sci 94:4476–4487.
- 6.
Arkin RL (2005) FDA newsletter: CVM researchers use latest science to develop methods for detecting animal drug residues. 2014.
- 7.
USDA (2008) United States department of agriculture (USDA), National Animal Health Monitoring System (NAHMS), Fort Collins, CO; part I: Reference of dairy cattle health and management practices in the United States.
- 8. Godden SM, Fetrow JP, Feirtag JM, Green LR, Wells SJ (2005) Economic analysis of feeding pasteurized nonsaleable milk versus conventional milk replacer to dairy calves. J Am Vet Med Assoc 226:1547–54.
- 9. Pereira RV, Siler JD, Bicalho RC, Warnick LD (2014) Multiresidue screening of milk withheld for sale at dairy farms in Central New York State. J Dairy Sci
- 10.
CLSI, editor (2008) Performance standards for antimicrobial disk and dilution susceptibility tests for bacteria collected from animals; approved standards. Wayne, PA.
- 11. Davies J, Spiegelman GB, Yim G (2006) The world of subinhibitory antibiotic concentrations. Curr Opin Microbiol 9:445–453.
- 12. Andersson DI, Hughes D (2010) Antibiotic resistance and its cost: Is it possible to reverse resistance? Nat Rev Microbiol 8:260–271.
- 13. Thi TD, Lopez E, Rodriguez-Rojas A, Rodriguez-Beltran J, Couce A, et al. (2011) Effect of recA inactivation on mutagenesis of Escherichia coli exposed to sublethal concentrations of antimicrobials. J Antimicrob Chemother 66:531–538.
- 14. Lopez E, Blazquez J (2009) Effect of subinhibitory concentrations of antibiotics on intrachromosomal homologous recombination in Escherichia coli. Antimicrob Agents Chemother 53:3411–3415.
- 15. Kohanski MA, DePristo MA, Collins JJ (2010) Sublethal antibiotic treatment leads to multidrug resistance via radical-induced mutagenesis. Mol Cell 37:311–320.
- 16. Thi TD, Lopez E, Rodriguez-Rojas A, Rodriguez-Beltran J, Couce A, et al. (2011) Effect of recA inactivation on mutagenesis of Escherichia coli exposed to sublethal concentrations of antimicrobials. J Antimicrob Chemother 66:531–538.
- 17. Kohanski MA, Dwyer DJ, Hayete B, Lawrence CA, Collins JJ (2007) A common mechanism of cellular death induced by bactericidal antibiotics. Cell 130:797–810.
- 18. Kim S, Yun Z, Ha UH, Lee S, Park H, et al. (2014) Transfer of antibiotic resistance plasmids in pure and activated sludge cultures in the presence of environmentally representative micro-contaminant concentrations. Sci Total Environ 468–469:813–820.
- 19. Pruden A, Pei R, Storteboom H, Carlson KH (2006) Antibiotic resistance genes as emerging contaminants: Studies in Northern Colorado. Environ Sci Technol 40:7445–7450.
- 20. Wardwell LH, Huttenhower C, Garrett WS (2011) Current concepts of the intestinal microbiota and the pathogenesis of infection. Curr Infect Dis Rep 13:28–34.
- 21. Perlin MH, Clark DR, McKenzie C, Patel H, Jackson N, et al. (2009) Protection of Salmonella by ampicillin-resistant Escherichia coli in the presence of otherwise lethal drug concentrations. Proceedings of the Royal Society B: Biological Sciences Proceedings of the Royal Society B: Biological Sciences 276:3759–3768.
- 22. Berendsen BJ, Elbers IJ, Stolker AA (2011) Determination of the stability of antibiotics in matrix and reference solutions using a straightforward procedure applying mass spectrometric detection. Food Addit Contam Part A Chem Anal Control Expo Risk Assess 28:1657–1666.
- 23. Karageorgou EG, Samanidou VF (2010) Application of ultrasound-assisted matrix solid-phase dispersion extraction to the HPLC confirmatory determination of cephalosporin residues in milk. J Sep Sci 33:2862–2871.
- 24. Schenck FJ, Friedman SL (2000) The effect of storage at 4 degrees C on the stability of ampicillin residues in raw milk. Food Addit Contam 17:675–677.
- 25. Zwald AG, Ruegg PL, Kaneene JB, Warnick LD, Wells SJ, et al. (2004) Management practices and reported antimicrobial usage on conventional and organic dairy farms. J Dairy Sci 87:191–201.
- 26. Pereira RV, Santos TM, Bicalho ML, Caixeta LS, Machado VS, et al. (2011) Antimicrobial resistance and prevalence of virulence factor genes in fecal Escherichia coli of Holstein calves fed milk with and without antimicrobials. J Dairy Sci 94:4556–4565.
- 27.
CLSI (2008) Performance standards for antimicrobial disk and dilution susceptibility tests for bacteria isolated from animals: Approved standard. Clinical and Laboratory Standards Institute (CLSI). Wayne, PA: Clinical and Laboratory Standards Institute.
- 28. Hoelzer K, Cummings KJ, Warnick LD, Schukken YH, Siler JD, et al. (2011) Agar disk diffusion and automated microbroth dilution produce similar antimicrobial susceptibility testing results for Salmonella serotypes Newport, Typhimurium, and 4,5,12:I-, but differ in economic cost. Foodborne Pathog Dis 8:1281–1288.
- 29. Hernando MD, Mezcua M, Fernandez-Alba AR, Barcelo D (2006) Environmental risk assessment of pharmaceutical residues in wastewater effluents, surface waters and sediments. Talanta 69:334–342.
- 30. Bendz D, Paxeus NA, Ginn TR, Loge FJ (2005) Occurrence and fate of pharmaceutically active compounds in the environment, a case study: Hoje river in sweden. J Hazard Mater 122:195–204.
- 31. Schriks M, Heringa MB, van der Kooi MM, de Voogt P, van Wezel AP (2010) Toxicological relevance of emerging contaminants for drinking water quality. Water Res 44:461–476.
- 32. Gullberg E, Cao S, Berg OG, Ilback C, Sandegren L, et al. (2011) Selection of resistant bacteria at very low antibiotic concentrations. PLoS Pathog 7:e1002158.
- 33. Drlica K, Zhao X (2007) Mutant selection window hypothesis updated. Clin Infect Dis 44:681–688.
- 34.
CLSI, editor (2012) Performance standards for antimicrobial susceptibility testing; twenty-second informational supplement. M100-D22. Wayne, PA, USA: Clinical Laboratory Standard Institute.
- 35. Sandegren L (2014) Selection of antibiotic resistance at very low antibiotic concentrations. Ups J Med Sci 119:103–107.
- 36. Silley P (2012) Susceptibility testing methods, resistance and breakpoints: What do these terms really mean? Rev Sci Tech 31:33–41.
- 37. Daniels JB, Call DR, Hancock D, Sischo WM, Baker K, et al. (2009) Role of ceftiofur in selection and dissemination of blaCMY-2-mediated cephalosporin resistance in Salmonella enterica and commensal Escherichia coli isolates from cattle. Appl Environ Microbiol 75:3648–3655.
- 38. Shaheen BW, Nayak R, Foley SL, Kweon O, Deck J, et al. (2011) Molecular characterization of resistance to extended-spectrum cephalosporins in clinical Escherichia coli isolates from companion animals in the United States. Antimicrob Agents Chemother 55:5666–5675.
- 39. Collignon P, Powers JH, Chiller TM, Aidara-Kane A, Aarestrup FM (2009) World health organization ranking of antimicrobials according to their importance in human medicine: A critical step for developing risk management strategies for the use of antimicrobials in food production animals. Clin Infect Dis 49:132–141.
- 40. Roberts MC (2005) Update on acquired tetracycline resistance genes. FEMS Microbiol Lett 245:195–203.
- 41. Chopra I, Roberts M (2001) Tetracycline antibiotics: Mode of action, applications, molecular biology, and epidemiology of bacterial resistance. Microbiol Mol Biol Rev 65:232–60 second page, table of contents.
- 42. Batchelor M, Threlfall EJ, Liebana E (2005) Cephalosporin resistance among animal-associated Enterobacteria: A current perspective. Expert Rev Anti Infect Ther 3:403–417.
- 43. D'Andrea MM, Arena F, Pallecchi L, Rossolini GM (2013) CTX-M-type beta-lactamases: A successful story of antibiotic resistance. Int J Med Microbiol 303:305–317.
- 44. Canton R, Morosini MI (2011) Emergence and spread of antibiotic resistance following exposure to antibiotics. FEMS Microbiol Rev 35:977–991.
- 45. Lopez E, Elez M, Matic I, Blazquez J (2007) Antibiotic-mediated recombination: Ciprofloxacin stimulates SOS-independent recombination of divergent sequences in Escherichia coli. Mol Microbiol 64:83–93.
- 46. Couce A, Blazquez J (2009) Side effects of antibiotics on genetic variability. FEMS Microbiol Rev 33:531–538.
- 47. Barr V, Barr K, Millar MR, Lacey RW (1986) Beta-lactam antibiotics increase the frequency of plasmid transfer in Staphylococcus aureus. J Antimicrob Chemother 17:409–413.
- 48. Ramirez MS, Tolmasky ME (2010) Aminoglycoside modifying enzymes. Drug Resist Updat 13:151–171.
- 49. Galimand M, Sabtcheva S, Courvalin P, Lambert T (2005) Worldwide disseminated armA aminoglycoside resistance methylase gene is borne by composite transposon Tn1548. Antimicrob Agents Chemother 49:2949–2953.
- 50. Magnet S, Blanchard JS (2005) Molecular insights into aminoglycoside action and resistance. Chem Rev 105:477–498.
- 51. Pfeifer Y, Cullik A, Witte W (2010) Resistance to cephalosporins and carbapenems in Gram-negative bacterial pathogens. Int J Med Microbiol 300:371–379.
- 52. Gutierrez A, Laureti L, Crussard S, Abida H, Rodriguez-Rojas A, et al. (2013) Beta-lactam antibiotics promote bacterial mutagenesis via an RpoS-mediated reduction in replication fidelity. Nat Commun 4:1610.
- 53. Andersson DI, Hughes D (2014) Microbiological effects of sublethal levels of antibiotics. Nat Rev Microbiol 12:465–478.
- 54. Gillings M, Boucher Y, Labbate M, Holmes A, Krishnan S, et al. (2008) The evolution of class 1 integrons and the rise of antibiotic resistance. J Bacteriol 190:5095–5100.
- 55. Du X, Shen Z, Wu B, Xia S, Shen J (2005) Characterization of class 1 integrons-mediated antibiotic resistance among calf pathogenic Escherichia coli. FEMS Microbiol Lett 245:295–298.
- 56. Wu K, Wang F, Sun J, Wang Q, Chen Q, et al. (2012) Class 1 integron gene cassettes in multidrug-resistant Gram-negative bacteria in Southern China. Int J Antimicrob Agents 40:264–267.
- 57. Braoudaki M, Hilton AC (2004) Adaptive resistance to biocides in Salmonella enterica and Escherichia coli O157 and cross-resistance to antimicrobial agents. J Clin Microbiol 42:73–78.
- 58. George AM, Levy SB (1983) Amplifiable resistance to tetracycline, chloramphenicol, and other antibiotics in Escherichia coli: Involvement of a non-plasmid-determined efflux of tetracycline. J Bacteriol 155:531–540.
- 59. Ma D, Cook DN, Alberti M, Pon NG, Nikaido H, et al. (1993) Molecular cloning and characterization of acrA and acrE genes of Escherichia coli. J Bacteriol 175:6299–6313.
- 60. Amabile-Cuevas CF, Demple B (1991) Molecular characterization of the soxRS genes of Escherichia coli: Two genes control a superoxide stress regulon. Nucleic Acids Res 19:4479–4484.
- 61. Maneewannakul K, Levy SB (1996) Identification for mar mutants among quinolone-resistant clinical isolates of Escherichia coli. Antimicrob Agents Chemother 40:1695–1698.
- 62. Ariza RR, Li Z, Ringstad N, Demple B (1995) Activation of multiple antibiotic resistance and binding of stress-inducible promoters by Escherichia coli rob protein. J Bacteriol 177:1655–1661.
- 63. Jernberg C, Lofmark S, Edlund C, Jansson JK (2010) Long-term impacts of antibiotic exposure on the human intestinal microbiota. Microbiology 156:3216–3223.
- 64. Oikonomou G, Teixeira AG, Foditsch C, Bicalho ML, Machado VS, et al. (2013) Fecal microbial diversity in pre-weaned dairy calves as described by pyrosequencing of metagenomic 16S rDNA. associations of Faecalibacterium species with health and growth. PLoS One 8:e63157.
- 65. Lozupone CA, Stombaugh JI, Gordon JI, Jansson JK, Knight R (2012) Diversity, stability and resilience of the human gut microbiota. Nature 489:220–230.
- 66. Berge AC, Hancock DD, Sischo WM, Besser TE (2010) Geographic, farm, and animal factors associated with multiple antimicrobial resistance in fecal Escherichia coli isolates from cattle in the Western United States. J Am Vet Med Assoc 236:1338–1344.
- 67. Khachatryan AR, Hancock DD, Besser TE, Call DR (2004) Role of calf-adapted Escherichia coli in maintenance of antimicrobial drug resistance in dairy calves. Appl Environ Microbiol 70:752–757.
- 68. Jakobsson HE, Jernberg C, Andersson AF, Sjolund-Karlsson M, Jansson JK, et al. (2010) Short-term antibiotic treatment has differing long-term impacts on the human throat and gut microbiome. PLoS One 5:e9836.
- 69. Lindgren M, Lofmark S, Edlund C, Huovinen P, Jalava J (2009) Prolonged impact of a one-week course of clindamycin on Enterococcus spp. in human normal microbiota. Scand J Infect Dis 41:215–219.
- 70. Roca M, Villegas L, Kortabitarte ML, Althaus RL, Molina MP (2011) Effect of heat treatments on stability of beta-lactams in milk. J Dairy Sci 94:1155–1164.
- 71. Li X, Zheng W, Machesky ML, Yates SR, Katterhenry M (2011) Degradation kinetics and mechanism of antibiotic ceftiofur in recycled water derived from a beef farm. J Agric Food Chem 59:10176–81.
- 72. Riediker S, Rytz A, Stadler RH (2004) Cold-temperature stability of five beta-lactam antibiotics in bovine milk and milk extracts prepared for liquid chromatography-electrospray ionization tandem mass spectrometry analysis. J Chromatogr A 1054:359–363.
- 73. Karageorgou EG, Samanidou VF (2010) Application of ultrasound-assisted matrix solid-phase dispersion extraction to the HPLC confirmatory determination of cephalosporin residues in milk. Journal of separation science 33:17–18.
- 74. Alekseev V, Larin S, Shigina O, Shcherbakova E (2006) Interaction of ampicillin with zinc and cadmium ions in aqueous solutions. Russian Journal of General Chemistry 76:317–320.
- 75. Michnik A, Michalik K, Marcoin W (2004) Influence of magnesium glutamate on stability of penicillin G aqueous solution. Int J Pharm 273:1–2.
- 76. Navarro PG, Blázquez IH, Osso BQ, Martínez de las Parras PJ, Puentedura MI, et al. (2003) Penicillin degradation catalysed by zn(II) ions in methanol. Int J Biol Macromol 33:4–5.
- 77. Kitazono Y, Ihara I, Yoshida G, Toyoda K, Umetsu K (2012) Selective degradation of tetracycline antibiotics present in raw milk by electrochemical method. J Hazard Mater 243:112–116.
- 78. Elizondo-Salazar JA, Jones CM, Heinrichs AJ (2010) Evaluation of calf milk pasteurization systems on 6 Pennsylvania dairy farms. J Dairy Sci 93:5509–5513.