Development of a Method for the Fast Detection of Extended-Spectrum β-Lactamase- and Plasmid-Mediated AmpC β-Lactamase-Producing Escherichia coli and Klebsiella pneumoniae from Dogs and Cats in the USA

Simple Summary Extended-spectrum β-lactamase (ESBL) and plasmid-mediated AmpC (pAmpC) β-lactamase-producing Escherichia coli and Klebsiella pneumoniae cause treatment failures in veterinary medicine. Many methods have been recommended for the detection of ESBL and pAmpC β-lactamase production but they are very subjective and the appropriate facilities are not available in most laboratories, especially not in clinics. We report the development of a method that can detect ESBL- and pAmpC β-lactamase-producing bacteria and this method is a fast, and low-cost tool for the screening of frequently encountered ESBL- and pAmpC β-lactamase-producing bacteria and would assist in diagnosis and improve therapeutic treatment in animal hospitals. Abstract Antibiotic resistance, such as resistance to beta-lactams and the development of resistance mechanisms, is associated with multifactorial phenomena and not only with the use of third-generation cephalosporins. Many methods have been recommended for the detection of ESBL and pAmpC β-lactamase production but they are very subjective and the appropriate facilities are not available in most laboratories, especially not in clinics. Therefore, for fast clinical antimicrobial selection, we need to rapidly detect ESBL- and pAmpC β-lactamase-producing bacteria using a simple method with samples containing large amounts of bacteria. For the detection of ESBL- and pAmpC phenotypes and genes, the disk diffusion test, DDST and multiplex PCR were conducted. Of the 109 samples, 99 (90.8%) samples were grown in MacConkey broth containing cephalothin, and 71 samples were grown on MacConkey agar containing ceftiofur. Of the 71 samples grown on MacConkey agar containing ceftiofur, 58 Escherichia coli and 19 Klebsiella pneumoniae isolates, in particular, harbored β-lactamase genes. Of the 38 samples that did not grow in MacConkey broth containing cephalothin or on MacConkey agar containing ceftiofur, 32 isolates were identified as E. coli, and 10 isolates were identified as K. pneumoniae; β-lactamase genes were not detected in these E. coli and K. pneumoniae isolates. Of the 78 ESBL- and pAmpC β-lactamase-producing E. coli and K. pneumoniae, 55 (70.5%) isolates carried one or more ESBL genes and 56 (71.8%) isolates carried one or more pAmpC β-lactamase genes. Our method is a fast, and low-cost tool for the screening of frequently encountered ESBL- and pAmpC β-lactamase-producing bacteria and it would assist in diagnosis and improve therapeutic treatment in animal hospitals.


Introduction
Escherichia coli and Klebsiella pneumoniae are members of the Enterobacteriaceae family, which mostly act as commensals in the intestinal tract of animals and humans. In particular, these bacteria can cause community-onset infections in animals and humans and are the common bacteria associated with urinary tract infections [1][2][3][4][5][6]. β-lactams are preferred for treating these infections in humans and veterinary medicine [7,8].
Extended-spectrum β-lactamase (ESBL) and plasmid-mediated AmpC (pAmpC) βlactamases are plasmid-encoded enzymes which are capable of inactivating a large number of β-lactam antibiotics, including extended-spectrum and very-broad-spectrum cephalosporins [9]. The most common bacteria that carry ESBL and pAmpC β-lactamase genes include E. coli and K. pneumoniae [10]. The emergence of ESBL-and pAmpC βlactamase-producing E. coli and K. pneumoniae in healthy and diseased animals constitutes an increasing challenge to infection management in veterinary therapy [11]. Moreover, the resistance caused by ESBL and pAmpC β-lactamases is usually multidrug resistance, which leads to critical therapeutic limitations [12,13].
Many tests have been recommended for the detection of ESBL and pAmpC β-lactamase production according to phenotype [14][15][16]. The most commonly used methods include the double disc synergy test (DDST) and Clinical and Laboratory Standards Institute (CLSI) confirmatory test. However, these methods are very subjective and can only be used for a single colony isolated from a sample. In addition, molecular methods are key tools in detection; however, the appropriate facilities are not available in most laboratories, especially not in clinics, and they are used for a single colony. Therefore, for fast clinical antimicrobial selection, we need to rapidly detect ESBL-and pAmpC β-lactamase-producing bacteria using a simple method with samples containing large amounts of bacteria. The objective of this study was to develop and evaluate a new detection method for ESBL-and pAmpC β-lactamase-producing bacteria and to evaluate the isolation and characterization of ESBLand pAmpC β-lactamase-producing E. coli and K. pneumoniae from shelter dogs and cats using this method.

Study Design
For the development of an isolation method for ESBL-and pAmpC β-lactamaseproducing E. coli and K. pneumoniae, we used a sample, isolated in 2014, from an animal hospital in Mississippi State University. For the detection of ESBL-and pAmpC β-lactamaseproducing bacteria by phenotype, ESBL production was confirmed using the DDST by using a disc of amoxicillin-clavulanate (AMC, 20/10 µg) along with four cephalosporins; cefotaxime (30 µg), ceftriaxone (30 µg), cefpodoxime (10 µg) and cefepime (50 µg), and pAmpC β-lactamase production was evaluated using cefoxitin (30 µg) as an inhibitor of pAmpC enzymes [10]. In addition, multiplex PCR of ESBL and pAmpC β-lactamase genes and the disk diffusion test were conducted to identify β-lactamases genes and cephalosporin resistance patterns [17]. Overall, 26 E. coli (each 13 ESBL-and pAmpC β-lactamase-producing bacteria or not) and 46 K. pneumoniae (each 23 ESBL-and pAmpC β-lactamase-producing bacteria or not) samples were isolated (Table 1).   Based on information about the use of cephalosporin in animal hospitals, we chose cephalosporin antimicrobials (1 first-generation cephalosporin and 2 third-generation cephalosporins). A total of 26 E. coli and 46 K. pneumoniae isolates were inoculated into tryptic soy broth (Sigma, St. Louis, MO, USA), and these inoculated cultures were incubated at 37 • C for 4 h. The pre-enriched TSB with bacteria was inoculated into MacConkey broth (Sigma) containing first-generation cephalosporin (cephalothin (128 µg/mL; Sigma)) and incubated at 37 • C for 24 h. After enrichment, only growth-positive MacConkey broth containing cephalothin was streaked on MacConkey agar (Sigma) containing different concentrations of each third-generation cephalosporin (ceftiofur (16, 32, 64 µg/mL; Sigma) and ceftriaxone (16,32, 64 µg/mL; Sigma)). The results of growth under different concentrations of each cephalosporin are shown in Table 2. Based on Table 2, we determined the type and concentration of cephalosporins (cephalothin (128 µg/mL) and ceftiofur (32 µg/mL)) for the isolation of ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae.

Sampling
We sampled dogs and cats in 6 animal shelters in Mississippi between May and August 2019. Dogs and cats eligible for sample collection were those that appeared healthy and caged individually. The feces, oral, and ear samples were collected using sterile cotton swabs and maintained at approximately 4 • C during transport to the research laboratory for processing.

Isolation of ESBL-and pAmpC β-Lactamase-Producing E. coli and K. pneumoniae
All samples were analyzed following a specific process ( Figure 1). After streaking on MacConkey agar plates, we selected the colony which appeared E. coli or K. pneumoniae.
To identify E. coli and K. pneumoniae, PCR was carried out as previously described [18,19]. For the isolation of ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae, all E. coli and K. pneumoniae samples were analyzed by multiplex PCR of ESBL and pAmpC β-lactamase genes as described above.

Isolation of ESBL and pAmpC-producing Escherichia coli and Klebsiella pneumoniae
A total of 109 samples were analyzed in this study: 77 samples from dogs and 32 samples from cats in 6 shelters in Mississippi. Among the 109 samples, 99 (90.8%) samples were grown in MacConkey broth containing cephalothin, and 71 samples were grown on MacConkey agar containing ceftiofur (Table 3). To identify ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae, we conducted multiplex PCR of ESBL and pAmpC β-lactamase genes in samples that grew or did not grow on MacConkey agar containing ceftiofur. Of the 71 samples grown on MacConkey agar containing ceftiofur, 58 isolates were identified as E. coli, and 20 isolates were identified as K. pneumoniae. All E. coli isolates harbored β-lactamase genes, and 19 K. pneumoniae isolates harbored β-lactamase genes. Among the 38 samples that did not grow in MacConkey broth containing cephalothin or on MacConkey agar containing ceftiofur, 32 isolates were identified as E.
coli, and 10 isolates were identified as K. pneumoniae; β-lactamase genes were not detected in these E. coli and K. pneumoniae isolates (Table 4). Only one K. pneumoniae isolate, which was isolated from MacConkey agar containing ceftiofur, did not carry ESBL and pAmpC β-lactamase genes. Dallenne et al. and Pimenta et al. reported that the absence of ESBL and pAmpC β-lactamase genes may be explained by the presence of a new enzyme due to the high rate of mutations of β-lactamase genes [17,21]. By using this method, we successfully isolated ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae from all the samples from dogs and cats in just 3 days.

Isolation of ESBL and pAmpC-producing Escherichia coli and Klebsiella pneumoniae
A total of 109 samples were analyzed in this study: 77 samples from dogs and 32 samples from cats in 6 shelters in Mississippi. Among the 109 samples, 99 (90.8%) samples were grown in MacConkey broth containing cephalothin, and 71 samples were grown on MacConkey agar containing ceftiofur (Table 3). To identify ESBL-and pAmpC β-lactamaseproducing E. coli and K. pneumoniae, we conducted multiplex PCR of ESBL and pAmpC β-lactamase genes in samples that grew or did not grow on MacConkey agar containing ceftiofur. Of the 71 samples grown on MacConkey agar containing ceftiofur, 58 isolates were identified as E. coli, and 20 isolates were identified as K. pneumoniae. All E. coli isolates harbored β-lactamase genes, and 19 K. pneumoniae isolates harbored β-lactamase genes. Among the 38 samples that did not grow in MacConkey broth containing cephalothin or on MacConkey agar containing ceftiofur, 32 isolates were identified as E. coli, and 10 isolates were identified as K. pneumoniae; β-lactamase genes were not detected in these E. coli and K. pneumoniae isolates (Table 4). Only one K. pneumoniae isolate, which was isolated from MacConkey agar containing ceftiofur, did not carry ESBL and pAmpC β-lactamase genes. Dallenne et al. and Pimenta et al. reported that the absence of ESBL and pAmpC β-lactamase genes may be explained by the presence of a new enzyme due to the high rate of mutations of β-lactamase genes [17,21]. By using this method, we successfully isolated ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae from all the samples from dogs and cats in just 3 days.

Characterization of ESBL and pAmpC-Producing Escherichia coli and Klebsiella pneumoniae
Of the 78 ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae, 55 (70.5%) isolates carried 1 or more ESBL genes: CTX-M-1 (24 E. coli isolates), CTX-M-2 (4 E. coli isolates and 2 K. pneumoniae isolates), CTX-M-9 (4 E. coli isolates and 3 K. pneumoniae isolates), TEM (25 E. coli isolates and 4 K. pneumoniae isolates), and OXA-1 (7 E. coli isolates). In addition, of the 78 ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae isolates, 56 (71.8%) isolates carried 1 or more pAmpC β-lactamase genes: CIT (43 E. coli isolates and 7 K. pneumoniae isolates), EBC (4 E. coli isolates and 1 K. pneumoniae isolate), ACC (1 E. coli isolate and 3 K. pneumoniae isolates), FOX (1 E. coli isolate and 3 K. pneumoniae isolates), and DHA (1 E. coli isolate) ( Table 5). Table 5. Distribution of ESBL and pAmpC β-lactamase genes in ESBL-and pAmpC β-lactamaseproducing E. coli and K. pneumoniae isolates from dogs and cats. The major ESBL genes were the CTX-M type, and the major pAmpC β-lactamase genes were the CIT type. These findings are consistent with those of a previous study showing that the occurrence of the CTX-M-type and CIT-type genes was the highest in animals in various countries [22][23][24][25]. In Korea, the distribution of CTX-M type genes in E. coli and K. pneumoniae isolated from dogs and cats has been reported, and pAmpC β-lactamase genes, especially the CIT type, have also been detected [26]. In Europe, CTX-M type genes as well as plasmid-mediated CIT type genes have been detected in K. pneumoniae and E. coli from healthy and sick animals including food-producing animals [27][28][29][30][31][32][33][34]. These results indicated that CTX-M-and CIT-type genes have been disseminated throughout the dog and cat populations in the USA.
The other ESBL genes conferring the β-lactam resistance detected in 52.7 and 12.7% of isolates in this study were the TEM and OXA genes, respectively. The TEM and OXA genes were previously identified in clinical E. coli and K. pneumoniae isolates from companion animals in Europe, which have been found to possess resistance genes against β-lactamase inhibitors (e.g., IRT genes), making such species more of a threat [2,3,32,[35][36][37][38].
We also detected various pAmpC β-lactamase genes such as EBC, ACC, FOX, and DHA. These enzymes are pAmpC β-lactamases developed through the transfer of chromosomal genes for inducible AmpC β-lactamases onto plasmids and confer a resistance pattern to most β-lactam antibiotics. Recent reports on E. coli and K. pneumoniae isolates have shown the prevalence of EBC, ACC, FOX, and DHA-like pAmpC β-lactamases in both human and animal hospitals. The existence of pAmpC β-lactamase genes poses a great challenge to infection control because they can be expressed in larger amounts and have high transmissibility to other bacterial species [39].

Antimicrobial Resistance Phenotypes
Among the 58 ESBL-and pAmpC β-lactamase-producing E. coli isolates, the rates of resistance to various antimicrobials were as follows: cefpodoxime (58/58, 100.0%), ceftazidime (58/58, 100.0%), cefotaxime (58/58, 100.  (Figure 3). In particular, the rate of resistance to cephalosporins, AM, AMC, CT, and NA was more than 60% for both E. coli and K. pneumoniae. In addition to the resistance to most β-lactam antibiotics, ESBL and pAmpC β-lactamase producers are also often resistant to quinolones and CT. This is because genes conferring resistance to quinolones and CT have been extensively reported in the same plasmid harboring β-lactamase genes [40]. In addition, ESBL and pAmpC β-lactamase cause multidrug resistance, thus limiting therapeutic choices [41]. There is no resistance to IMP for both ESBL-and pAmpC βlactamase-producing E. coli and K. pneumoniae, and the resistance to G was less than 20% for ESBL-and pAmpC β-lactamase-producing E. coli and K. pneumoniae. Therefore, IMP and G antimicrobials may be potential treatment options for ESBL-and pAmpC β-lactamaseassociated infections. The results of the present study could contribute to the improvement of therapeutic guidelines for treating dogs in veterinary hospitals in the USA.

Conclusions
In summary, we report the development of a method that can detect ESBL-and pAmpC β-lactamase-producing bacteria. This method is a fast, and low-cost tool for the screening of frequently encountered ESBL-and pAmpC β-lactamase-producing bacteria. It would assist in diagnosis and improve therapeutic treatment in animal hospitals. Our study showed that bacteria isolated by this method were identified as ESBL and pAmpC β-lactamase producing bacteria in most cases.

Conclusions
In summary, we report the development of a method that can detect ESBL-and pAmpC β-lactamase-producing bacteria. This method is a fast, and low-cost tool for the screening of frequently encountered ESBL-and pAmpC β-lactamase-producing bacteria. It would assist in diagnosis and improve therapeutic treatment in animal hospitals. Our study showed that bacteria isolated by this method were identified as ESBL and pAmpC β-lactamase producing bacteria in most cases.