Plasmid-Mediated Ampicillin, Quinolone, and Heavy Metal Co-Resistance among ESBL-Producing Isolates from the Yamuna River, New Delhi, India

Antibiotic resistance is one of the major current global health crises. Because of increasing contamination with antimicrobials, pesticides, and heavy metals, the aquatic environment has become a hotspot for emergence, maintenance, and dissemination of antibiotic and heavy metal resistance genes among bacteria. The aim of the present study was to determine the co-resistance to quinolones, ampicillin, and heavy metals among the bacterial isolates harboring extended-spectrum β-lactamases (ESBLs) genes. Among 73 bacterial strains isolated from a highly polluted stretch of the Yamuna River in Delhi, those carrying blaCTX-M, blaTEM, or blaSHV genes were analyzed to detect the genetic determinants of resistance to quinolones, ampicillin, mercury, and arsenic. The plasmid-mediated quinolone resistance (PMQR) gene qnrS was found in 22 isolates; however, the qnrA, B, C, and qnrD genes could not be detected in any of the bacteria. Two variants of CMY, blaCMY-2 and blaCMY-42, were identified among eight and seven strains, respectively. Furthermore, merB, merP, merT, and arsC genes were detected in 40, 40, 44, and 24 bacterial strains, respectively. Co-transfer of different resistance genes was also investigated in a transconjugation experiment. Successful transconjugants had antibiotic and heavy metal resistance genes with similar tolerance toward antibiotics and heavy metals as did their donors. This study indicates that the aquatic environment is a major reservoir of bacteria harboring resistance genes to antibiotics and heavy metals and emphasizes the need to study the genetic basis of resistant microorganisms and their public health implications.


Introduction
Ever since their discovery, antibiotics have provided many benefits in fields ranging from healthcare to agriculture. However, the emergence of antibiotic resistance has significantly reduced their efficacy and usefulness in treating humans and animals. Emergence of antibiotic resistance among microorganisms is associated with many complex metabolic and genetic factors. Among them, extended spectrum β-lactamases (ESBLs), AmpC β-lactamases, and plasmid-mediated quinolone resistance (PMQR) genes are major factors in resistance to a variety of drugs [1][2][3]. ESBLs are one of the most important mechanisms providing resistance against large numbers of life-saving drugs.

Detection and Characterization of blaCMY and PMQR Genes
Alkaline lysis method was used to extract the plasmid DNA from all the tested bacterial strains. In order to confirm the presence of plasmid, extracted genetic material from all the strains were subjected to 16S rRNA gene amplification, since this gene is known to be located on chromosomal DNA, it does not amplify by the plasmid as a template. Non-amplification of 16S rRNA gene confirmed the presence of plasmid DNA among 73 tested bacterial isolates.
Of 73 bacterial strains, 15 (20.5%) harbored the blaCMY gene. Nucleotide sequence analysis revealed the presence of two variants, blaCMY-2 and blaCMY-42. blaCMY-2 was found and characterized among eight strains, whereas blaCMY-42 was identified in seven bacteria (Table 2). Furthermore, PMQR genes (qnrA, qnrB, qnrC, qnrD, qnrS, and aac(6′)-Ib) were amplified and only qnrS was detected among 22 bacterial strains. Nucleotide sequences of all the 22 amplified products were analyzed for their homology with previously submitted sequences and submitted to NCBI GenBank database. The qnrA, qnrB, qnrC, qnrD, and aac(6′)-Ib genes could not be detected among any of the bacterial strains. The distribution pattern of antibiotic resistance genes is presented in Table 2.

Detection of Heavy Metal Resistance Genes
All the 73 bacterial strains were checked for the presence of heavy metal resistance genes on their plasmid DNA. Mercury resistance genes merB, merP, and merT were detected among 40, 40, and 43 strains, respectively (Table 2). Furthermore, the arsenic resistance gene arsC was successfully amplified among 24 strains (Table 2). Distribution patterns of heavy metal resistance genes among test strains are shown in Table 2. Combinations of different antibiotics and heavy metals resistance genes are presented in Figure 2. Antibiotic susceptibility pattern of extended-spectrum β-lactamases (ESBL) producing bacterial strains used in this study. Note 1: AMP, ampicillin; CAZ, ceftazidime; CTX, cefotaxime; CZ, cefazolin; CTR, ceftriaxone; AT, aztreonam; CX, cefoxitin; TR, trimethoprim; P/T-piperacillin/tazobactam, PB, polymyxin B; CIP, ciprofloxacin; A/S, ampicillin/sulbactam; LE, levofloxacin; AK, amikacin; TE, tetracycline; C, chloramphenicol; IPM-imipenem. Note 2: "Resistant"-Bacteria which could resist the effect of prescribed dosage concentration of antibiotic. "Intermediate"-Bacteria on its course to be resistant to prescribed concentration of antibiotic. "Sensitive"-Bacteria, which could not resist the prescribed dosage concentration of antibiotic.

Antibiotic Susceptibility Testing and Multiple Antibiotic Resistance (MAR) Index
Antibiotic susceptibility testing of all the 73 test strains was repeated toward different classes of antibiotics using the Kirby Bauer disk diffusion technique as per Clinical and Laboratory Standards Institute guidelines (CLSI 2016). Seventeen different antibiotics used in the study included: amikacin (30 µg), ampicillin (10 µg), ampicillin/sulbactam (10/10 µg), piperacillin/tazobactam (100/10 µg), cefazolin (30 µg), cefoxitin (30 µg), ciprofloxacin (5 µg), levofloxacin (5 µg), imipenem (10 µg), polymyxin-B (300 units), aztreonam (30 µg), trimethoprim (5 µg), tetracycline (30 µg), chloramphenicol (30 µg), cefotaxime (30 µg), ceftazidime (30 µg), and ceftriaxone (30 µg) (all from HiMedia, India). After overnight incubation at 37 • C, zone of inhibition was measured, susceptibility pattern was determined, and bacterial strains were classified into resistant, intermediate, and sensitive with the help of CLSI guidelines. The multiple antibiotic resistance (MAR) index of each isolate was calculated as described by Krumperman [33]. The overall resistance of a single isolate towards 17 tested antibiotics is represented as multiple antibiotic resistance index (MAR Index). The MAR index is risk assessment tool that helps in differentiation of high-and low-risk region where antimicrobial drugs has been overused. An MAR index > 0.2 indicates the overuse of antimicrobial drugs and high selection pressure in the river, whereas MAR index < 0.2, indicates the low risk of source contamination in the river.

Plasmid DNA Isolation and Amplification
Plasmid DNA was extracted by the alkaline lysis method [34]. Extracted plasmid DNA was used as a template and PCR amplification was carried for all the genes using specific sets of primers (Table 1). PCR amplification of antibiotics and heavy metal resistance genes was carried out in a 50 µL reaction volume, comprising 5 µL of 10× buffer, 3 µL of 25 mM MgCl2, 1 µL of 10 mM dNTP mix, 1 µM of each forward and reverse primer, 2U of Taq-polymerase (Promega, Madison, WI, USA) and 2 µL (30 ng/µL) of template DNA. For each PCR, initial denaturation at 95 • C for 5 min was done, followed by 35 cycles of: denaturation (94 • C for 1 min), annealing (respective Tm of primer for 1 min), extension (72 • C for 1 min), and a final extension at 72 • C for 10 min. In order to check the positive amplification, PCR products were subjected to electrophoresis on 1% agarose gel (GENEI, Bengaluru, India). Gels were further visualized and photographed in a gel documentation system (BIO-RAD, Hercules, CA, USA). Amplicon sizes of different genes were checked using a 100 bp DNA molecular weight marker as reference (Fermentas, Waltham, MA, USA).

Detection and Characterization of blaCMY and PMQR Genes
All the bacterial strains were screened for the existence of plasmid-mediated blaCMY and PMQR (qnrA, qnrB, qnrC, qnrD, qnrS, and aac(6 )-Ib) genes with the help of specific sets of primers (Table 1). Purified PCR products were sequenced with corresponding gene primers (SciGenom Labs, Kochi, Kerala, India). Nucleotide sequences were analyzed by FinchTV software (Geospiza, Inc., Seattle, WA, USA), and the BioEdit program (http://www.mbio.ncsu.edu/bioedit/bioedit.html) was used to check homology of gene sequences with already available GenBank sequences. Variants of different genes were determined using the NCBI Basic Local Alignment Search Tool (BLAST) and homology analysis.

Detection of Heavy Metal Resistance Genes
All the test isolates were checked for the possible co-occurrence of heavy metal (mercury and arsenic) resistance genes. Occurrence of heavy metal resistance determinants were checked with plasmid DNA as a template and specific sets of primers (Table 1). Mercury resistance genes, e.g., merB, merP, and merT, were amplified by the protocol described by Azam et al. [40]. Similarly, the arsenic resistance gene arsC was amplified as per the method of Sunita et al. [25]. PCR products were run on a 1% agarose gel to determine positive amplification.

Susceptibility towards Heavy Metals
Heavy metal tolerance was recorded among bacterial strains harboring any mercury and/or arsenic resistance determinants. Minimum inhibitory concentration (MIC) of mercuric chloride (HgCl 2 ) was determined against the strains carrying mer genes, whereas the MIC of sodium arsenate (Na 2 HAsO 4 ) was recorded in bacteria positive for the arsC gene. Methods described by Sunita et al. and Figueiredo et al. were used to check the tolerance of bacteria against heavy metal concentrations [25,41]. Briefly, Serial dilutions (1 mg/L to 512 mg/L) of heavy metal salts were prepared in 96 well microtiter plate containing 200 µL of Luria broth. Medium was inoculated with 100 µL of 10,000 times diluted 0.1 OD 600 culture of each isolate and incubated at 37 • C at 120 rpm for 14-18 h in an automated incubator shaker. The OD of samples was measured in a microplate reader (Thermo Scientific MultiskanGo, Waltham, MA, USA) at 600 nm wavelength, and MIC was determined. E. coli ATCC 25922 strains was used as negative control. Media without heavy metal salt was also served as control.

Conjugation Experiment
A conjugation experiment was carried out to determine the transferability of antibiotic and metal resistance genes (blaCTX-M, blaTEM, blaSHV, blaCMY, PMQR, merB, merP, merT, and arsC). Seven different isolates, namely, E. coli MKT3, Acinetobacter junii IBT29, Enterobacter cloacae BVT29, E. coli MBT29, K. pneumoniae MBT50, E. coli CCT42, and Acinetobacter sp. BOT40, harboring antibiotic and heavy metal resistance determinants were chosen as donors. Plasmid-free Escherichia coli J53 strain resistant to sodium azide and sensitive to antibiotics and heavy metal used in this study was selected as recipient. Transconjugation experiment was performed by the help of protocol described by Zurfluh et al. [31]. In brief, secondary cultures of donor and recipient bacterial isolates were mixed in a 1:1 ratio in Luria-Bertani broth (HiMedia, Mumbai, India) and incubated at 37 • C for 24 h without any shaking. Serial dilutions of transconjugant bacteria were streaked on LA plates containing cefotaxime (2 µg/mL) and sodium azide (100 µg/mL). After overnight incubation, transconjugant colonies were selected and revived for the plasmid isolation. PCR-based amplification using transconjugant plasmids as templates was carried out to detect the successful transfer of antibiotic and heavy metal resistance determinants.

Results
The present study was aimed to study the co-occurrence of PMQR, blaCMY, and heavy metal resistance genes among bacterial strains harboring ESBL genes collected from the highly polluted Yamuna River. Detailed results of this work are discussed below.

Susceptibility Profile of Test Strains
All the 73 bacterial strains were exposed to 17 different antibiotics to analyze their susceptibility patterns by the help of CLSI-2016 guidelines and classified into resistant, intermediate, and susceptible strains. Terms resistant, intermediate, and sensitive have been recommended by the CLSI (clinical laboratory standard institute) for the classification of bacteria on the basis of antibiotic susceptibility. "Resistant" are the organisms, which could resist the effect of prescribed dosage concentration of antibiotics, "intermediate" are those bacteria, which are on the course to be resistant to prescribed drug concentration, and "sensitive" are those, which could not resist the prescribed concentration of antibiotics.
Analysis of antibiotic profiling found that the majority of test strains were resistant to ceftazidime (62/73), followed by cefotaxime ( Table S2). Antimicrobial susceptibility patterns of all the tested strains are shown in Figure 1. The multiple antibiotic resistance (MAR) index of different bacteria was found to be in the range of 0.941-0.058. The MAR index of all tested bacterial strains is shown in Table 2.

Detection and Characterization of blaCMY and PMQR Genes
Alkaline lysis method was used to extract the plasmid DNA from all the tested bacterial strains. In order to confirm the presence of plasmid, extracted genetic material from all the strains were subjected to 16S rRNA gene amplification, since this gene is known to be located on chromosomal DNA, it does not amplify by the plasmid as a template. Non-amplification of 16S rRNA gene confirmed the presence of plasmid DNA among 73 tested bacterial isolates.
Of 73 bacterial strains, 15 (20.5%) harbored the blaCMY gene. Nucleotide sequence analysis revealed the presence of two variants, blaCMY-2 and blaCMY-42. blaCMY-2 was found and characterized among eight strains, whereas blaCMY-42 was identified in seven bacteria (Table 2). Furthermore, PMQR genes (qnrA, qnrB, qnrC, qnrD, qnrS, and aac(6 )-Ib) were amplified and only qnrS was detected among 22 bacterial strains. Nucleotide sequences of all the 22 amplified products were analyzed for their homology with previously submitted sequences and submitted to NCBI GenBank database. The qnrA, qnrB, qnrC, qnrD, and aac(6 )-Ib genes could not be detected among any of the bacterial strains. The distribution pattern of antibiotic resistance genes is presented in Table 2.

Detection of Heavy Metal Resistance Genes
All the 73 bacterial strains were checked for the presence of heavy metal resistance genes on their plasmid DNA. Mercury resistance genes merB, merP, and merT were detected among 40, 40, and 43 strains, respectively (Table 2). Furthermore, the arsenic resistance gene arsC was successfully amplified among 24 strains (Table 2). Distribution patterns of heavy metal resistance genes among test strains are shown in Table 2. Combinations of different antibiotics and heavy metals resistance genes are presented in Figure 2.

Susceptibility to Heavy Metals
All the bacterial strains carrying any of the respective metal resistance genes were studied to determine their tolerance toward HgCl2 or Na2HAsO4.Isolates harboring any of the mer genes showed tolerance to HgCl2 in the range of <1 to 8 mg/L. The MIC of Na2HAsO4 was in the range of 16 to >512 mg/L against the strains carrying the arsC gene (Table 2).

Susceptibility to Heavy Metals
All the bacterial strains carrying any of the respective metal resistance genes were studied to determine their tolerance toward HgCl 2 or Na 2 HAsO 4 .Isolates harboring any of the mer genes showed tolerance to HgCl 2 in the range of <1 to 8 mg/L. The MIC of Na 2 HAsO 4 was in the range of 16 to >512 mg/L against the strains carrying the arsC gene (Table 2).
The MIC of ampicillin, cefotaxime, ciprofloxacin, HgCl 2, and Na 2 HAsO 4 was determined against transconjugants and was similar to that of their donor bacterial strains (Table 3).
Acinetobacter sp. BOT40 Note: Escherichia coli J53 strain used as recipient for transconjugation was sensitive to all the tested antibiotics and heavy metal in this study.

Discussion
In the present study, we investigated the presence of PMQR, blaCMY, and heavy metal resistance genes among bacterial strains harboring ESBL genes collected from the highly polluted Yamuna River [32]. The most important finding of this work was co-occurrence of antibiotic and heavy metal resistance determinants among bacterial strains and their co-transfer to the host strain.
Our work highlights the importance of the resistance gene pool and its interaction with diverse groups of bacteria in the urban aquatic environment. In the present work, we studied the co-existence of PMQR, AmpC, and heavy metal resistance determinants among bacterial strains of aquatic origin harboring ESBL genes. The horizontal transfer of different resistance genes was also determined with conjugation experiments. One of the very important aspects of this study was the co-occurrence of different antibiotic and heavy metal resistance determinants among tested bacterial strains of aquatic origin.
Domestic, hospital, municipal, industrial and agricultural wastes continuously increase the pollution level of urban aquatic bodies. This high degree of pollution creates selection pressure for the emergence, dissemination, and maintenance of antibiotic and heavy metal resistance among aquatic microorganisms. Studies on water and soil quality of the Delhi stretch of the Yamuna River showed a significantly high concentration of various antibiotics and heavy metals in river water [20,21]. Furthermore, the incidence of antibiotic and heavy metal resistance determinants among microorganisms from polluted aquatic environments is significant [9,32,[42][43][44].
Plasmid-mediated CIT type AmpC β-lactamase was successfully amplified among 20.5% of bacterial strains. Bajaj et al., Martinez-Medina et al., and Loncaric et al. also reported the similar occurrence of AmpC-type β-lactamase among bacterial isolates from river water and patient samples, respectively [9,52,55]. Furthermore, sequence analysis revealed the occurrence of blaCMY-2 and blaCMY-42 among 8 and 7 strains, respectively. Among CIT type AmpC β-lactamases, blaCMY-2 is one of the most common variants reported in various studies [56,57]. blaCMY-42 is one of the variants of blaCMY-2 evolved by substitution of val231Ser [7,9]. Presence of the blaCMY-42 gene among aquatic isolates is a serious issue because it confers a high level of resistance against third-generation cephalosporins and monobactams.
Quinolone and β-lactam antibiotics are very common drugs for the treatment of various infections. Thus, the co-occurrence of resistance toward both of these drugs could pose a serious public health concern [14,15]. PMQR genes are major contributors in conferring resistance toward fluoroquinolone drugs [10,11]. Six different PMQR genes (qnrA, qnrB, qnrC, qnrD, qnrS, and aac(6 )-Ib) were checked for their occurrence among bacterial strains with ESBL genes. Successful amplification of the qnrS gene was observed among 30.1% of tested strains. None of the isolates was positive for qnrA, qnrB, qnrC, qnrD, and aac(6 )-Ib. Sequence homology analysis of partial genes revealed 100% similarity with the already available qnrS gene in the NCBI GenBank database. Our results are in accordance with the study of Bajaj et al., where 50% of CTX-M-producing isolates were positive for the qnrS gene [14]. Co-occurrence of PMQR and other antibiotic resistance genes has been reported in similar studies from China, Iran, and Austria [55,58,59]. The presence of PMQR genes among bacteria is a cause for concern, because these resistance genes are thought to create an optimum environment for selection of chromosomal resistance determinants [60,61]. This could pose a further serious health risk to human and animal populations. It is imperative to increase the scope of surveillance studies of drug-resistant bacteria to the genotypic level so that identification of the cause of resistance can be determined accurately.
Occurrence of mercury resistance genes (merB, merP and merT) and the arsenic resistance gene (arsC) was determined among ESBL-positive strains. In the present study, we observed the presence of merB, merP, and merT genes among 54.7%, 54.7%, and 58.95% of tested bacteria, respectively. The genes merB, merP, and merT are known to play a crucial role in the emergence of mercury resistance, but the combined action of all the genes of the mer operon exerts broad-spectrum resistance among microorganisms [62]. Studies among environmental bacteria from the United States and Brazil also reported the presence of mer operon genes in significantly high percentages [63,64]. Co-occurrence of antibiotic and mercury resistance gene was also reported by many studies carried out in India and Australia [40,65]. All the strains positive for at least one mer gene were tested for their tolerance toward mercuric chloride. The MIC of mercuric chloride was found in the range of <1 to 8 mg/L. Azam et al. and Rebello et al. also reported a similar tolerance pattern among bacterial isolates of river origin [40,64]. Various genes are known to be responsible for conferring arsenic resistance among bacteria. arsC is one of the key genes determining arsenic resistance by reducing arsenate to arsenite [66]. In this study, we observed the occurrence of the arsC gene on the plasmid DNA of 32.8% of bacteria. Di Cesare et al. reported the occurrence of arsC genes among antibiotic-resistant bacteria obtained from wastewater in Italy [43]. Other studies among bacterial isolates of environmental origin also showed the presence of the arsC gene in abundance [25,43,67]. In determining the MIC of sodium arsenate against the isolates harboring the arsC gene, we found that bacterial isolates were able to tolerate 16 to > 512 mg/L concentration of sodium arsenate. Studies on arsenic-tolerant bacteria carried out in India and Indonesia also showed the similar MICs of sodium arsenate [68,69]. Our results are in agreement with the previous studies on characterization of the arsC gene among bacteria in India and Italy [25,67,70].
We performed a conjugation experiment to check the possible horizontal transfer of antibiotic and heavy metal resistance genes. An in vitro conjugation experiment in this study revealed that bacterial strains E. coli MKT3, Acinetobacter junii IBT29, Enterobacter cloacae BVT29, E. coli MBT29, K. pneumoniae MBT50, E. coli CCT42, and Acinetobacter sp. BOT40 successfully transferred antibiotic and heavy metal-resistant genes to recipient E. coli J53AZ R . We did not find the presence of arsC gene in any of the transconjugants, which showed the possibility of presence of this gene on other non-conjugative plasmid. This result of conjugation experiment suggests the possibility of resistance gene transfer between different bacteria in environmental settings. The transfer of resistance determinants in the aquatic environment could significantly enhance the opportunity for pathogens to acquire resistance [71]. Our data corroborate those in the study conducted by Liu et al., in which successful transfer of various drug resistance genes was carried out through transconjugation [72]. A study conducted among bacterial isolates from wastewater in Italy has also suggested a co-selection of resistance genes of antibiotics and heavy metals [43]. Because antibiotic and heavy metal resistance genes are known to be located on a similar gene cluster, and also on the same conjugative plasmids, it is imperative to further explore the possible mechanisms of heavy metal-driven emergence and dissemination of drug resistance.

Conclusions
Waterborne bacterial strains carrying ESBL genes were found to harbor resistance determinants for AmpC β-lactamase, PMQR, mercury resistance, and arsenic resistance. The presence of resistance genes for such a large number of antimicrobial agents is a serious cause for concern. Detection of blaCMY-2 and blaCMY-42 is also very significant because these determinants confer increased resistance against cephalosporins and monobactams. Furthermore, coexistence and plasmid-mediated co-transfer of antibiotic and heavy metal resistance genes is a very serious issue. It warrants the deeper and large-scale study on molecular mechanisms of these resistance genes in different settings. Moreover, it is also very important to unravel the role of environmental factors influencing the emergence and dissemination resistance among microorganisms. Study also emphasizes the stern need of regulation on the use of various antimicrobial and heavy metal substances, which helps in the creation of selection pressure responsible for the rise of resistance.
Supplementary Materials: The following are available online at http://www.mdpi.com/2079-6382/9/11/826/s1, Table S1: Water samples collection sites of river Yamuna from where bacterial strains were isolated, Table S2: Bacterial strains used in this study with corresponding ESBLs genes and their antibiotics profiling.