Comparative occurrence and antibiogram of extended-spectrum β-lactamase-producing Escherichia coli among post-weaned calves and lactating cows from smallholder dairy farms in a parallel animal husbandry area

Background and Aim: Inappropriate overuse of antimicrobials might be associated with the spreading of antimicrobial-resistant bacteria in animal-based food products. Extended-spectrum β-lactamase (ESBL)-producing Escherichia coli have been recognized as an emerging global problem in a One Health approach. This study aimed to assess the occurrence and antimicrobial-susceptible profiles of ESBL-producing E. coli among post-weaned calves and lactating cows in a parallel animal husbandry area. Materials and Methods: Seventy-two pool fecal samples were collected from 36 smallholder dairy farms registered in Ban Hong Dairy Cooperatives, Lamphun Province, Thailand. Pre-enriched fecal samples were cultured in MacConkey agar supplemented with cefotaxime. The potential E. coli isolates were identified by not only biochemical tests but also polymerase chain reaction assay of the 16S rRNA gene. ESBL production was confirmed by the combination disk test. Antimicrobial susceptibility testing was performed by the Kirby–Bauer disk diffusion method. Results: The occurrence of ESBL-producing E. coli at the farm level was 80.56%. The different phenotypic antibiogram of ESBL-producing E. coli was observed among post-weaned calf and lactating cow specimens. The most frequent resistance patterns of ESBL-producing isolates from both groups were amoxicillin-ceftiofur-cephalexin-cephalothin-cloxacillin-streptomycin-oxytetracycline-sulfamethoxazole/trimethoprim. For the median zone diameter, enrofloxacin-resistant isolates with narrow zone diameter values from lactating cow specimens were particularly more than post-weaned calf specimens (p<0.05). Conclusion: These findings revealed the dynamic changes in ESBL-producing E. coli from calves and lactating cows in Lamphun Province, posing the inevitability to prevent bacterial transmission and optimize antimicrobial therapy in dairy farming.


Introduction
Antimicrobial resistance (AMR) has occurred worldwide, with consequent economic impact and public health concerns [1,2]. AMR is a comprehensive problem related to humans, animals, and the environment. Because the global situation is fast progressing into a post-antibiotic era, urgent strategic plans are essential for surveillance and prevention to avert the AMR crisis. At present, the AMR mechanism of bacterial enzyme-mediated resistance to antimicrobial agents is well documented. In particular, β-lactamase enzymes (EC 3.5.2.6) produced by Gramnegative bacteria that hydrolyze the covalent bond of the β-lactam ring structure are the primary resistance mechanism to β-lactam antimicrobials [3]. Extendedspectrum β-lactamase (ESBL)-producing Escherichia coli are the commensal and pathogenic bacteria in the critical priority group classified according to the greatest threat to human health and the urgent need for new antimicrobials by the World Health Organization (WHO) [4]. ESBL enzymes can confer bacterial resistance to penicillin, first-to third-generation cephalosporin, and monobactam that are inhibited by β-lactamase inhibitors, such as clavulanic acid, Available at www.veterinaryworld.org/Vol.14/May-2021/33.pdf sulbactam, and tazobactam [5]. In 2017, the Centers for Disease Control and Prevention have reported that ESBL-producing Enterobacteriaceae threats in the United States had approximately 197,000 morbidity cases and 9100 deaths [6]. Moreover, many reports have shown that ESBL-producing E. coli infections lead to increased morbidity and mortality rates, treatment failure, requiring more complex treatments, prolonged hospital stays, and limited therapeutic options [4][5][6]. Nevertheless, ESBL-producing E. coli occur in healthcare settings and are present in the communities. During the past decade, these problems have recently raised significant concerns in the human food chain. Several reports have published ESBLproducing E. coli detected in farm animals, including pigs, poultry, and cattle [7][8][9].
Cattle meat and dairy products are the primary food protein sources unquestionably required in the human diet [10,11]. Unfortunately, animal fecal mass is the major source of gut microflora as a reservoir and disseminates ESBL-producing E. coli [12]. The environmental contamination of ESBL-producing E. coli creates a consensual concern in the scientific community and the One Health approach [13,14]. Of additional importance, E. coli producing CTX-M-2 β-lactamase in cattle has been previously described in Japan [15]. From the first description to the present, ESBL-producing Enterobacteriaceae have recently been reported in cattle production in 40 countries [16][17][18][19]. Despite this significant problem, information on the occurrence and dissemination of ESBL among calves and cows is lacking, whereas ongoing surveillance has previously provided some information on antimicrobial use [20] in bulls, cows, and calves [21]. In addition, the excessive and inappropriate antimicrobial use in both calves and cows has been considered one of the main contributors to the selection of ESBL-producing E. coli.
The Thailand situation report on ESBL-producing E. coli in dairy cattle is seldom published. Moreover, the presence of ESBL-producing E. coli in the development status of cattle farming is still very limited. Consequently, this study aimed to assess the occurrence of ESBL-producing E. coli from pooled fecal samples from healthy calves and cows in smallholder dairy farms in Lamphun Province, Thailand. No data are available on the antibiograms of multidrug-resistant (MDR) bacteria or ESBL-producing E. coli in healthy cattle from smallholder dairy farms in a parallel animal husbandry area. Therefore, the antimicrobial susceptibility profile of ESBL-producing isolates was also evaluated.

Ethical approval
The research study was ethically approved by the Animal Care and Use Committee, Faculty of Veterinary Medicine, Chiang Mai University (approval number: S3/2562). Holstein Friesian cows were housed in free-stall barn dairy farms and milked twice a day. Cows with good udder health were required for this study. The Institute Biosafety Committee, Chiang Mai University, also granted permission to test the pathogens (approval number: CMU A-0762019).

Study period and location
The study was conducted from April to May 2020 in 36 smallholder dairy farms registered in Ban Hong Dairy Cooperatives, Lamphun Province, Thailand.

Sample population and collection
The mean herd sizes were 30 lactating cows (max=80, min=6) and 25 post-weaned calves (max=86, min=6). On each farm, the fecal samples were taken through rectal palpation of all dairy cattle by the veterinarian. In addition, 10 fecal samples from lactating dairy cows were included in one pooled fecal sample and five fecal samples from calves were collected for one pooled fecal sample. Both pooled fecal samples from healthy calves and lactating cows were assembled on the same farm. All samples were kept at 5°C and transported to the laboratory within 6 h. On the same day of sample collection, the dairy farmers answered a questionnaire on general information, such as demographic data, antimicrobial use, and calf feeding.

Microbiological identification of ESBL-producing E. coli
Pooled fecal samples (5 g) were cultured in 45 mL pre-enrichment Luria-Bertani (LB) [22] broth (HiMedia, India). After incubation at 37°C for 24 h, a loop full of the pre-enrichment cultures was streaked onto MacConkey agar (HiMedia, India) supplemented with 1 mg/L cefotaxime. The suspected pink colonies with precipitated bile on MacConkey agar containing cefotaxime were identified to be E. coli using standard IMIViC biochemical tests, including indole production test, methyl red test, Voges-Proskauer test, and citrate utilization test. These tests also included triple sugar iron test and urease test. The 16S rRNA gene of strains of E. coli was also determined using previously published primers [23]. Then, E. coli isolates were measured as ESBL production by the combination disk test (CDT) using cefotaxime (30 µg), cefotaxime/clavulanic acid (30/10 µg), ceftazidime (30 µg), and ceftazidime/clavulanic acid (30/10 µg) [24]. The phenotypic ESBL isolates were confirmed when the inhibition zone of cephalosporins combined with clavulanic acid was ≥5 mm compared to cephalosporins alone. In addition, E. coli (ATCC 25922) and Klebsiella pneumoniae (ATCC 700603) were used as the quality negative control strain and positive ESBL control type strain, respectively.

Antimicrobial susceptibility testing
Antimicrobial susceptibility testing was performed by the Kirby-Bauer disk diffusion method according to CLSI [24]. ESBL-producing E. coli isolates were streaked on 5% sheep blood agar plates and incubated at 37°C for 24 h. Bacterial culture was adjusted to a concentration of 1.5 × 10 8 colony-forming units/mL in Mueller-Hinton broth using a McFarland densitometer (Grant Instruments, Cambridgeshire, UK) and swabbed on BBL TM Mueller-Hinton agar plates (Becton Dickinson & Co., Sparks, MD, USA). All ESBL-producing E. coli isolates were subjected to the antimicrobial susceptibility test by the Kirby-Bauer disk diffusion method with 14 antimicrobial agents (Table-1). After 24 h incubation at 30°C, the size of the bacterial growth inhibition zones was interpreted as sensitive (S), intermediately resistant (I), or resistant (R) according to the antimicrobial breakpoints for Enterobacteriaceae by the CLSI guidelines [24,25].

Data management and statistical analysis
The geographical distribution of selected smallholder dairy farms was demonstrated by mapping using Quantum Information System version 2.18.28. Descriptive statistics were used to describe data, including frequency, percentage, proportion to express the general characteristics, and basic information on the occurrence of ESBL-producing E. coli and its antibiogram, to compare the main outcome differences between the calf and cow groups. The antimicrobial susceptibility of ESBL-producing E. coli was interpreted qualitatively as resistant, intermediate, or susceptible. The isolates with resistant phenotype to three or more antimicrobial classes were identified as MDR. The AMR patterns of ESBL-producing E. coli isolates were summarized in terms of frequencies. The distribution of inhibition zone diameters of antimicrobials against ESBL-producing E. coli was plotted, and the trendlines of cumulative curves were performed in sixth-degree polynomial approximation. By the point of interception, the median (ZD 50 ) and 90 th percentile (ZD 90 ) were also calculated by polynomial regression equations. The Mann-Whitney test was also used for comparison between the calf and cow groups with a non-normal distribution. The differences between variables were considered statistically significant when the bicaudal probability of their occurrence due to chance (error type I) was lower than 5% (p<0.05). Statistical analysis was performed with R statistical software (RStudio, Boston, MA, USA).

Results
All pooled fecal samples from the calf and cow groups were pre-enriched in LB broth. Sixty-eight of the 72 samples (94.44%) were culture positive for the screening of cefotaxime-resistant Enterobacteriaceae. Subsequently, 40 ESBL-producing E. coli isolates were confirmed by CDT. At the sample level of the pooled fecal samples, ESBL-producing E. coli were found in 55.56% (40/72) of the total samples. At least one positive sample in either the calf group or the cow group for ESBL-producing E. coli was counted as the occurrence of values within the farm level. In addition, the occurrence frequency of ESBL-producing E. coli at the farm level was 80.56% (29/36 farms). Positive ESBL-producing E. coli in both calf and cow fecal samples were found on 11 farms (30.56%). Nine farms only detected positive ESBL-producing E. coli in calf fecal samples (25.00%), whereas nine other farms detected positive ESBL-producing E. coli in cow fecal samples (25.00%; Table-2). The geographical distribution of ESBL-producing E. coli isolates of dairy farms at the farm level and at the status level is shown in Figure-1.
In ESBL-producing E. coli from cow fecal samples results, 20 ESBL-producing E. coli isolates  were completely resistant (100%) to amoxicillin, cephalexin, cephalothin, and cloxacillin. A high level (50-99%) of drug resistance to oxytetracycline, ceftiofur, streptomycin, sulfamethoxazole/trimethoprim, gentamicin, and kanamycin was observed. In contrast, a low level (1-49%) of drug resistance to chloramphenicol was also found. No ESBL-producing E. coli isolates were resistant to amoxicillin/clavulanic acid, enrofloxacin, and imipenem (Figure-3). ESBL-producing E. coli isolates were tested against nine groups of antimicrobials, and resistance to at least three groups indicated that the isolates were MDR. All isolates collected from healthy calves and cows were resistant to at least three classes. In all, 30.0% of ESBL-producing E. coli isolates were resistant to eight and nine antimicrobial agents, and 10% showed resistance to seven antimicrobial agents. The most frequent resistance patterns of ESBL-producing E. coli isolated from calf groups (four isolates) and cow groups (three isolates) ( Table-3).
By the point of interception, the distribution of inhibition zone diameters of ESBL-producing E. coli isolates was described in the ZD 50 and ZD 90 values calculated by polynomial regression equations. In both calf and cow groups, the ZD 50 and ZD 90 values for amoxicillin, cephalexin, cephalothin, and cloxacillin were 0 mm. The resistance was eventually seen to almost all antimicrobial inhibition diameters that were tested (Table-4). However, most antimicrobial ZD 90 values in the cow group were broader than the cow group, except for kanamycin, streptomycin, oxytetracycline, and imipenem. For ZD 50 , most antimicrobial diameters in the calf group were broader than the cow group, except for chloramphenicol, gentamicin, oxytetracycline, and sulfamethoxazole/trimethoprim. Interestingly, various inhibition diameters for enrofloxacin against ESBL-producing E. coli isolates  Available at www.veterinaryworld.org/Vol.14/May-2021/33.pdf were observed among the calf group (16-26 mm) and the cow group (17-23 mm). Moreover, the median zone diameter for enrofloxacin against ESBLproducing E. coli isolates was 23 (interquartile range [IQR], 23-23) for the calf group and 20 (IQR, [19][20][21][22] for the cow group (P < 0.05, Mann-Whitney test) (Figure-4).

Discussion
This research was designed to determine the presence of ESBL-producing E. coli on dairy cattle from smallholder farms in Lamphun Province and elucidate the antibiograms from pooled fecal samples from healthy calves and cows in a parallel animal husbandry area. The ESBL-producing E. coli status was defined among calf and cow specimens at the herd level; therefore, individual animal aspects were not assessed. In a previous study, ESBL-producing E. coli strains could be isolated more than twice using enrichment procedures [19]. Consequently, in this study, the pooled fecal samples were pre-enriched in LB broth and cultured in MacConkey agar supplemented with cefotaxime to screen cefotaxime-resistant Enterobacteriaceae. Before ESBL phenotypic confirmation using the CDT, potential    AML=Amoxicillin, AMC=Amoxicillin/clavulanic acid, EFT=Ceftiofur, CL=Cephalexin, KF=Cephalothin, OB=Cloxacillin, C=Chloramphenicol, ENR=Enrofloxacin, CN=Gentamicin, S=Streptomycin, OT=Oxytetracycline, IPM=Imipenem, SXT=Sulfamethoxazole/trimethoprim, ESBL=Extended-spectrum β-lactamase E. coli isolates were identified by not only biochemical tests but also polymerase chain reaction (PCR) assay. Based on the approaches described in Materials and Methods, the molecular characterization of antimicrobial-resistant E. coli isolated from domestic and food-producing animals was widely tested for the presence of 16S rRNA gene using PCR methods [26,27]. In smallholder farms in Lamphun Province, a high occurrence of ESBL-producing E. coli at the farm level was observed (80.56%). These findings agree with previous reports on the percentage of ESBL-producing E. coli (86.7%)-positive samples in Bavarian dairy and beef cattle farms in Germany [19]. Subsequently, the presence of ESBL-producing E. coli strains in dairy cattle was evaluated in distinct countries, and the occurrence frequency of resistance varied in individual regions. In contrast, a low occurrence of ESBL-producing E. coli from cattle farming was reported in the Netherlands (59.6%) [28], China (43.6%) [29], Great Britain (37.5%) [30], Germany (18.0%) [7], and Finland (2.0%) [31]. Interestingly, this study demonstrated that one-half occurrence of ESBL-producing E. coli was found in either calf or cow group (55.56%). This result was in contrast to the previous epidemiological studies in Germany [19], Switzerland [18], and Great Britain [30] that the prevalence of ESBL-producing E. coli in calves was higher than in cows. In this study, the geographical distribution of ESBL-producing E. coli in 36 dairy farms located in Ban Hong Dairy Cooperatives was conducted. The global positioning system of this subregion coordinates is 18°19′35.1″N, 98°46′36.6″E. The average temperature range is 20.7-33.0°C (69.26-91.4°F), and the average humidity is 72.16%. Interestingly, this area is consistent with recently published studies on the occurrence of ESBL-producing E. coli in healthy pigs (76.7-98.06%) [9,32], healthy poultry (40.0%) [32], and healthy humans (77.30-96.52%) [9,32].
The Kirby-Bauer test for the antimicrobial susceptibility to ESBL producer isolates was interpreted qualitatively as resistant, intermediate, or susceptible. This study demonstrated that ESBL-producing E. coli isolates from both calf and cow specimens were predominantly inhibited by imipenem and amoxicillin/ clavulanic acid. However, stewardship efforts of preserving the imipenem are a prerequisite [33]. All ESBLproducing E. coli isolates showed a high rate of resistance against β-lactam antimicrobials, oxytetracycline, sulfamethoxazole/trimethoprim, and streptomycin. Moreover, the results demonstrated that all isolates of ESBL-producing E. coli were resistant to at least three classes from healthy calves and healthy cows. The previous studies have shown that all ESBL-producing E. coli isolates from healthy or sick dairy cattle (e.g. diarrhea and mastitis) were commonly present as MDR [34][35][36]. Besides, resistance to β-lactam antimicrobials and resistance to tetracycline, macrolide, sulfonamide, and diaminopyrimidine were the most common resistance pattern among ESBL-producing E. coli isolates. ESBL resistance genes selected by non-β-lactams were previously documented [8]. Paterson and Bonomo also revealed that using various antimicrobial classes, including sulfamethoxazole/trimethoprim, quinolones, and aminoglycosides, are associated with subsequent infections due to ESBL-producing bacteria [5]. These resistant antimicrobials in this study have been routinely used to treat and prevent disease in dairy cattle worldwide and also in Thailand. The alimentary tract of ruminants has a great quantity of bacteria as the normal commensal flora [20]. For Enterobacteriaceae generally considered as commensal alimentary inhabitants, AMR genes for ESBLs could potentially be horizontally transferred among acquired resistance bacteria [8,37,38]. Therefore, there is a higher probability for non-pathogenic commensal E. coli to become a reservoir of AMR in the food chain.
Interestingly, the interpretive results of enrofloxacin susceptibility testing of ESBL-producing E. coli from calf specimens with susceptible phenotype, in contrast to isolates from cow specimens with intermediate phenotype, were observed. Moreover, similar data were obtained with regard to the sensitivity of the tested isolates to enrofloxacin. The ZD 50 value from the calf group was entirely in the sensitivity zone, and the range of inhibition zone diameters was markedly broader than the cow group. Increasing enrofloxacin resistance may also pose a significant threat to animal health and food safety. In food-producing animals, mutations within the chromosomal target site of gyrA and parC in E. coli were described [39,40]. Moreover, the most common plasmid-borne resistance mechanism is also the mutation of the plasmid-mediated quinolone resistance gene of oqxAB [41]. Fluoroquinolones are recommended for use in non-lactating dairy calve for the treatment or control of diarrhea and respiratory disease. A previous study demonstrated the increased resistance of E. coli after using enrofloxacin in calves [42]. Treatment of pre-weaned calves at high risk for bovine respiratory disease with enrofloxacin resulted in a significant increase in the shedding of ciprofloxacin-resistant E. coli in feces for up to 7 days after medication [43]. These study findings may have differed from the current study due to the use of different pre-weaning and post-weaning groups. Additional differences could be particularly attributed to the bacterial strains of non-ESBL and ESBL producers.
In 2001, the WHO published that the inappropriate use of antimicrobials (dose, duration, and indication) might increase the overall selective pressure of AMR pathogens [44]. During the past decade, several studies in many countries, including Thailand, have generally reported that commensal E. coli in food-producing animals are becoming more resistant to antimicrobials (7)(8)(9). Therefore, it is important to continuously monitor the antimicrobial susceptibility profiles of E. coli isolated from food-producing animals in practices that result in resistant antimicrobials important in human and animal medicine.

Conclusion
The results highlighted the first study on ESBLproducing E. coli in herd status in dairy farms in Lamphun Province, Thailand. This research indicated that the different phenotypic antibiogram of ESBLproducing E. coli was observed among post-weaned calf and lactating cow specimens. Interestingly, enrofloxacin-resistant isolates with narrow zone diameter values from lactating cow specimens were particularly more than post-weaned calf specimens. Further studies are necessary to deepen the epidemiological knowledge on AMR in the dairy food chain. Continuous monitoring of antimicrobial susceptibility profiles should be performed to improve antimicrobial stewardship in animal farming and ensure final product safety.

Authors' Contributions
KNL, KK, and RM: Designed the study concept and the research experiments. CV: Performed the sample collection and laboratory testing. RM, KNL, PC, and VP: Conducted the formal analysis and the data curation. CV and RM: Prepared the original draft of the manuscript. RM: Contributed the scientific advice, review, and editing. All authors read and approved the final manuscript.

Acknowledgments
We are grateful to the veterinarian and staff of the Ban Hong dairy cooperatives, Lamphun Province, Thailand, for the facility and instrument support. In addition, we also thank the owners of dairy farms for participating in this study. This research work was partially funded by Chiang Mai University. The authors are thankful for the epidemiology research center of infectious disease, Chiang Mai University for providing the advice.