Monitoring of Non-β-Lactam Antibiotic Resistance-Associated Genes in ESBL Producing Enterobacterales Isolates

Genetic context of extended spectrum β-Lactamase (ESBL) producing Enterobacterales and its association with plasmid mediated quinolone resistance (PMQR), aminoglycoside modifying enzymes (AME) and Trimethoprim/Sulfamethoxazole (TMP-SMX) resistance is little known from North India. Therefore, the current study was aimed to investigate the frequency of Non-β-Lactam antibiotic resistance associated genes in extended spectrum β-Lactamase producing Enterobacterales. For this study, Non-Duplicate phenotypically confirmed ESBL producing Enterobacterales isolates (N = 186) were analyzed for ESBLs, PMQRs, AMEs and TMP-SMX resistance genes using polymerase chain reaction (PCR). PCR detected presence of PMQR genes in 81.29% (N = 139) of ESBL isolates (N = 171), AME genes in 60.82% and TMP-SMX resistance genes in 63.74% of the isolates. Molecular characterization of ESBL producing Enterobacterales showed 84.79% blaTEM followed by 73.68% blaCTX-M, 43.86% blaSHV, 19.88% blaPER and 9.94% blaVEB, respectively. Analysis of PMQR genes revealed 77.7% aac(6′)-lb-cr the most commonly detected gene followed by 67.63% oqxB, 62.59% oqxA, 43.17% qnrB, 19.42% qnrD, 18.7% qnrS, 9.35% qnrA, 3.6% qepA and 2.88% qnrC, respectively. Analysis of AMEs gene profile demonstrated 81.73% aac(6′)-Ib, the most frequently encountered gene followed by 46.15% aph(3′)-Ia, 44.23% ant(3”)-Ia, respectively. A 100% prevalence of sul1, followed by dfrA (54.63%) and sul2 (15.74%) was observed. In summary, prevalence of ESBL-Producing genes (particularly blaTEM and blaCTX-M) along with PMQR, AMEs, and TMP-SMX resistant genes may potentially aid in the transfer of antimicrobial resistance among these strains.

Literature indicates that synthesis of aminoglycoside modifying enzymes (AMEs) is an important AMR mechanism that produce high level of aminoglycoside resistance among Gram-negative bacteria [9,10]. These AMEs are grouped into three categories: (i) aminoglycoside N-acetyltransferases (AACs), (ii) aminoglycoside O-phosphotransferases (APHs), and (iii) aminoglycoside O-nucleotidyltransferases (ANTs) [10]. Further, the structural genes coding for AMEs are often located on plasmids that carry multiple resistance elements for ESBLs [11,12]. This type of association between ESBL and AMEs coding genes are of foremost apprehension in the treatment of bacterial infections. Besides resistance to β-lactams, fluoroquinolones, and aminoglycosides, members of Enterobacterales demonstrate resistance to TMP-SMX as well. However, little has been reported on the genetic context of various species of ESBL producing Enterobacterales and its association with, PMQR, AMEs TMP-SMX resistance exclusively from North India. Therefore, the current study was designed to analyze the distribution of non-β-lactam antibiotic resistance associated genes prevailing in ESBL producing Enterobacterales. To our knowledge, this is the first study that investigated the distribution of AME, PMQR and TMP-SMX resistance genes among ESBL producing Enterobacterales that are isolated from North India. Furthermore, this study included clinical isolates of different members of Enterobacterales (Escherichia coli, Klebsiella pneumonia, Citrobacter freundii, Klebsiella oxytoca, Morganella morganii Proteus mirabilis, Proteus vulgaris, and Enterobacter cloacae), while the earlier studies limited their molecular characterization in two species (E. coli and K. pneumoniae) [8,10,13]. In addition, this study explores prevalence of TMP-SMX resistance genes among ESBL producing clinical Enterobacterales, an observation that has been made only a few times earlier.
Of the total ESBL isolates that carrying sul1 gene (N = 108), majority of the isolates were having co-existence with bla TEM (95/108), bla CTX-M (81/108) and bla SHV (51/108) genes, respectively. Similarly, strains carrying dfrA1 gene (N = 59) demonstrated coexistence with bla TEM (53/59) and bla CTX-M (48/59) (Supplementary Materials Table S5). Further, Enterobacterales isolates that showed elevated MICs (4/76 ≥ 16/304 µg/mL) for TMP-SMX mainly harbored sul1 and dfrA1 genes. It was also observed that the frequency of TMP-SMX genes was higher in strains with high TMP-SMX MIC values (Supplementary Materials Table S8). Table 5 summarizes the comparison of origin of strain and type of resistance genes detected. The data analysis shows no direct correlation exists between origin of strain (wound, respiratory tract specimens, blood and body fluids) and the type of resistance genes (ESBL, PMQR, AME, and TMP-SMX resistance genes) detected. The number of types of resistance genes detected in wound specimens were significantly (p < 0.05) different from that found in the respiratory tract specimens (Table 5).

Discussion
The current study investigated the frequency of non-β-lactam antibiotic resistance associated genes among ESBL producing Enterobacterales. It was observed that the prevalence of ESBL genes detected in the current study (particularly among E. coli and K. pneumoniae) were comparable with earlier studies [2,22]. The most commonly encountered ESBL gene in this study was bla TEM (84.79%) followed by bla CTX-M (73.68%), bla SHV (43.86%), bla PER (18.71%) and bla SHV (9.94%), respectively. These findings were in accordance with an earlier study wherein most prevalent ESBL gene found was bla TEM (73%) followed by bla CTX-M (25-100%) and bla SHV (23%) [2]. The low prevalence of bla PER and bla VEB in the present study were comparable with the data reported by Khurana et al. [2]. Further, it was observed that most of the ESBL producing organisms were resistant to fluoroquinolones, aminoglycosides and TMP-SMX, respectively. This may be possibly due to the co-existence of PMQR, AME and TMP-SMX resistance genes in the same plasmids that also code for ESBL proteins [22].
In the present study, PCRs detected presence of PMQR genes in 81.29% (N = 139) of genotypically confirmed ESBL isolates (N = 171), indicating the presence of high frequency of PMQR genes among ESBL strains. Interestingly, K. pneumoniae (44.6%) was having higher number of PMQR genes detected followed by E. coli (31.65%). This observation is in accordance with an earlier study wherein, PMQR genes were more frequently encountered among E. coli, Klebsiella species, and Enterobacter species [3]. However, the frequency of occurrence of PMQR genes were low among Enterobacter cloacae in this study. This is probably due to the difference in the geographical location as the earlier study was conducted in the southern part of India (where the usage of antibiotics is different). Further, the widespread antibiotic resistance prevalent in India may be attributed to readily availability of antibiotics across the pharmacy counters. This could play a major role in increased distribution of antibiotic-resistance genes throughout the population. This study included isolates that obtained from wound, respiratory tract, body fluid and blood specimens while other studies were performed mostly using isolates that are collected from urine [4,9,20]. Of the total genotypically confirmed ESBL cases (N = 171), the most frequent PMQR gene detected was aac(6 )-lb-cr (77.7%) which is in agreement with an earlier report wherein the prevalence rate was found to be 64.5% [3]. In this study, relatively higher prevalence of qnrB (43.17%) was observed, which was in accordance with Yang et.al observation (prevalence rate~50%) [13]. Previous studies also reported the low prevalence of qnrD, qnrS, qnrA, and qnrC [4,5,22,23]. It was observed that the distribution of efflux pumps genes among ESBL producing Enterobacterales isolates were found to be oqxB (67.63%), oqxA (62.59%) and qepA (3.6%), respectively. However, this efflux pump mediated drug resistance mechanism of bacteria can be subdued using various efflux inhibitory molecules [24], for instance, the susceptibility of antibiotics against multidrug-resistant bacteria (that developed exclusively due to efflux pump mechanisms) can be enhanced in presence of efflux pump inhibitor such as omeprazole [25]. Further, the genome sequencing analysis of multidrug-resistant strains that might reveal the potential genes that are associated with multidrug efflux pumps and once the genes and gene products are identified, the molecular docking studies that may further help in developing appropriate efflux pump inhibitors. These efflux pump inhibitors can be incorporated with antibiotic molecules in order to overcome the efflux pump mediated drug resistance [26]. Further, to our knowledge, this is the first study that investigated the prevalence of PMQR among Enterobacterales in North India. However, the oqxA and oqxB prevalence rates were comparable with earlier reports wherein the prevalence rates were found to be 88% and 30% for oqxA and oqxB genes, respectively [27]. Further, the differences in the prevalence of oqxA and oqxB genes may be attributed to the geographical distribution and type of isolates studied, as most of the studies were conducted on E. coli and K. pneumonia isolates [5,19]. The low prevalence of qepA observed (3.6%) was similar with previous data (2%) in the literature [13,28]. This low prevalence of qepA in the current study may also indicate low incidence of qepA gene among different strains of Enterobacterales across the world [5]. In this study, the presence of PMQR genes were associated with ESBL genes, possibly due to the common carriage on the same plasmids [23]. The isolates that carrying a minimum of two β-lactamases coding genes (particularly, bla TEM and bla CTX-M ) were more likely to carry aac(6 )-Ib-cr and qnrB genes. The genes that code for both ESBL and PMQR proteins are usually located on same plasmids and consequently that may have higher chances of transfer among the members of Enterobacterales. Therefore, it is very pertinent to comprehend the drug resistance mechanisms prevalent among the members of medically important bacteria as it may be a major concern for patient safety and in determination of therapeutic strategies.
Genes encoding AMEs are prevalent in various groups of bacteria [10,[29][30][31][32]. In this study, prevalence of AME genes were found to be 60.82% among ESBL producing strains of Enterobacterales. This relatively higher prevalence rate of AMEs in ESBL producing strains may be due to co-existence of genes encoding ESBLs and AMEs in Gram-negative bacteria [9]. In this study, AMEs coding genes were most frequently isolated from K. pneumoniae (44.23%) and E. coli (38.46%). This was in accordance with an earlier study conducted by Haidar et al., wherein a higher prevalence of AME genes were reported among K. pneumoniae [9]. In the present study, the most frequently encountered AME gene was aac(6 )-Ib (81.73%) which is in agreement with the earlier studies (prevalence was 73%) [9]. The predominance and coexistence of aac(6 )-Ib with other AME genes observed may be attributed to the fact that the gene coding for aac(6 )-Ib enzyme is frequently located within class I integrons. Further, it is known that the gene cassettes that carry other genes coding for AMEs can be easily incorporated into class I integrons resulting in the development of resistance to currently used aminoglycosides [10]. The other prevalent genes were aph(3 )-Ia (46.15%), followed by ant(3")-Ia (44.23%), aac(3)-IIa (45.19%), and aph(3")-Ib (35.58%). The higher prevalence of aph(3 )-Ia and aph(3")-Ib is alarming as this type of resistance may usually produce high level of aminoglycoside resistance. The prevalence of other AME genes were found to be relatively low, which is comparable with earlier studies [9,23].
The TMP-SMX resistance genes such as sul1, sul2, or dfrA genes are likely to be present either on chromosome or on plasmids [33,34]. In the present study, TMP-SMX resistance genes were obtained from 63.74% of ESBL producing isolates. This comparatively low detection rate of these resistance genes may be attributed to the presence of alternative resistance mechanisms prevailing in TMP-SMX resistant isolates [33]. However, additional investigations are required to explore the genetic basis of TMP-SMX resistance mechanisms that prevailing in various strains of Enterobacterales. Further, among the genotypically confirmed TMP-SMX resistant strains (N = 108), 42.59% of K. pneumonia isolates were carrying TMP-SMX resistance genes, followed by E. coli (34.26%). However, due to the paucity of literatures, the comparison with earlier studies could not performed. To our knowledge, this new study report, the prevalence of TMP-SMX resistance genes among various clinical strains of ESBL producing Enterobacterales. A higher prevalence of sul1 (100%) followed by dfrA (54.63%) and sul2 (15.74%), genes were noted in this study. This higher prevalence of sul1 gene may be attributed to the fact that the sul1 genes are usually located within class I integrons and this particular characteristic (which is a horizontally transferable genetic element) might have further helped in its wide distribution [35]. The comparatively higher prevalence of dfrA1 (33.91%) gene in the present study was in agreement with an earlier study wherein in the prevalence rate was also found to be high [22]. Enterobacterales isolates that showed elevated MICs for TMP-SMX mainly harbored sul1 and dfrA1 genes indicating the likelihood of these genes in imparting resistance to TMP-SMX. However, no genes were detected from number of isolates that were having higher MIC values, suggesting the existence of alternative pathways of resistance in TMP-SMX resistant isolates.

Study Setting and Clinical Specimens
Between July 2018 and June 2019, a total of 2134 clinical samples (wound, respiratory tract specimens, blood and body fluids) received in the Microbiology laboratory, Government Medical College and Hospital, Badaun, India were analyzed for ESBL producing strains of Enterobacterales. This hospital laboratory receives samples from two civil hospitals and three primary health care centers that are attached to it. Among the total samples analyzed (N = 2134), a total of 186 non-repetitive phenotypically confirmed ESBL producing Enterobacterales isolates (one organism per patient was included to avoid duplication) were obtained. All the isolates were identified by manual API ® system (BioMérieux, Durham, NC, USA) and the results interpreted as recommended by the manufacturer. The strains included were K. pneumoniae (N = 74), E. coli (N = 58), P. mirabilis (N = 15), C. freundii (N = 13), K. oxytoca (N = 11), E. cloacae (N = 9), P. vulgaris (N = 3) and M. morganii (N = 3). All the clinical isolates were identified by standard laboratory procedure [11]. All the bacterial isolates were stored at −80 • C in glycerol for future use.
The MIC of fluoroquinolones (MIC determined for selected antibiotics that include; ciprofloxacin, levofloxacin, nalidixic acid, gatifloxacin, and moxifloxacin), aminoglycosides (MIC calculated for; gentamycin, tobramycin, amikacin and kanamycin) and TMP-SMX were determined by broth micro dilution method and the results were interpreted in accordance with the CLSI guidelines [36]. The MICs were calculated to determine the association between presence of antibiotic resistance genes and concentration of antibiotic tested.
The double-disk synergy diffusion test (phenotypic confirmatory test for ESBL) was carried out as per CLSI recommendations. Briefly, ceftazidime (30 µg) and cefotaxime (30 µg) alone and in combination with clavulanic acid (10 µg; Himedia, Mumbai, India) were used. The ESBL production is confirmed when there is an increase of zone diameter (≥5 mm) around disk with antibiotic-clavulanic acid combination [36].
The disc approximation test was used to confirm the ESBL production in E. cloacae, C. freundii, P. vulgaris, and M. morganii strains. This test was performed as disc diffusion assay on Mueller-Hinton agar (MHA). Briefly, antibiotic discs containing aztreonam (30 µg), ceftazidime (30 µg), ceftriaxone (30 µg), and cefotaxime (30 µg) were kept 30 mm apart (center to center) around amoxicillin-clavulanic acid (20 µg + 10 µg) disc on MHA plate inoculated with the organism to be tested. The MHA plates were incubated at 37 • C for 24 h. An increased zone of inhibition of any of the test antibiotic towards amoxicillin-clavulanic acid was considered as ESBL production [37]. The control strains used were E. coli ATCC 25922 (non-beta-lactamases producer) and K. pneumoniae ATCC 700603 (ESBL producer), respectively.

Extraction of Bacterial DNA
Bacterial DNA was obtained from the phenotypically confirmed ESBL strains of Enterobacterales using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). The DNA samples obtained by this procedure were segregated into two aliquots; first aliquot was used as template for the subsequent PCR reactions and the second aliquot was stored at −80 • C (Jindal Ultra Freezer (SMI-165E), Ghaziabad, India) for future use.

Molecular Detection of ESBL, PMQR, AME and TMP-SMX Resistance Genes
All the phenotypically confirmed ESBL producing Enterobacterales isolates were subjected to molecular characterization of the relevant encoding genes such as bla SHV , bla CTX-M , bla TEM , bla PER , and bla VEB by PCR using primers and PCR conditions shown in Table 1 [14,15].
All isolates phenotypically resistant to TMP-SMX were subjected to PCR for detection of sul1, sul2 and dfrA genes using primers and PCR conditions shown in Table 1 [22,38]. Briefly, 2 µL (~500 ng) of purified DNA was subjected to each multiplex PCR in a 100 µL reaction mixture containing 1 × PCR buffer (10 Tris-HCl pH 8.8, (NH 4 ) 2 SO 4 , 3 mM MgCl 2 , 0.2% Tween 20), 200 mM of each dNTPs, 0.5 µM of each primer, and 2.0 units of AURA Taq DNA polymerase. Amplification was carried out as follows: initial denaturation at 95 • C for 5 min; 30 cycles of denaturation at 95 • C for 30 s, annealing at 50-62 • C for 30 s, extension/elongation at 72 • C for 45 s; and a final elongation step at 72 • C for 5 min. PCR-generated products were detected by electrophoresis of 7 µL of each amplification mixture in 2% agarose gels in 1% Tris Borate-EDTA buffer and 0.5 µg/mL ethidium bromide.
To identify the ESBL, PMQR, AME, and TMP-SMX resistance genes detected in the PCR assays, automated DNA sequencing of the amplicons were conducted. More specifically, multiplex PCR-generated products were separated in 2% low melting agarose gel with 1% Tris-acetate-EDTA buffer. The PCR products were excised from the agarose gel and purified using the QIAquick PCR purification kit (Qiagen, Hilden, Germany) as recommended by the manufacturer. The nucleotide sequencing of amplicons was conducted using an ABI 3730xl DNA Analyzer (Applied Biosystems, Branch burg, NJ, USA). Basic Local Alignment Search Tool (BLAST) program was used to compare each ESBL, PMQR, AME, and TMP-SMX resistance gene sequences against those available in gene bank at the National Center of Biotechnology Information database.

Statistical Analysis
Chi-squared test was used to compare the association between the origin of strain and type of resistance genes detected. The null hypothesis will be accepted if the presence of genes in all the groups (wound, respiratory tract specimens, and blood body fluids) were similar. Dunn's multiple comparisons test was performed to compare the differences in the number of resistance genes obtained between two categories of samples. All statistical analyses were performed using Graph pad Prism (version 6, Graph-Pad Software, Inc., La Jolla, CA, USA). The statistical difference values showing p < 0.05 were considered as significant.

Conclusions
In summary, this is the first study that investigated the occurrence of genetic determinants prevailing in ESBL producing Enterobacterales and the association of these genetic determinants with PMQR, AME, and TMP-SMX resistance genes in the north India. The current study demonstrated widespread occurrence of PMQR, AME, and TMP-SMX drug resistant genetic determinants in the ESBL producing Enterobacterales strains. Screening of PMQR genetic elements in ESBL producing Enterobacterales strains revealed high prevalence of both aac(6 )-Ib-cr and qnrB. However, molecular analysis of AMEs producing Enterobacterales strains showed high prevalence of aac(6 )-Ib followed by aph (3 )-Ia. On examination of TMP-SMX resistant strains, sul1 was found to be the most frequently encountered gene followed by dfrA. The association of ESBL-producing genes with the PMQR, AMEs, and TMP-SMX resistance genes may potentially aid in transfer of drug resistance determinants among these strains. Therefore, a complete understanding of PMQRs, AMEs, and other drug resistance mechanisms will help in determining rationale of treatment and infection control measures in hospital settings.
Supplementary Materials: The following are available online at http://www.mdpi.com/2079-6382/9/12/884/s1, Table S1: Antibiotic resistance pattern of Enterobacterales isolates tested, Table S2: Distribution of Enterobacterales isolates in various specimens studied, Table S3: Distribution of plasmid mediated quinolone resistance (PMQR) genes in extended spectrum beta-lactamase producing Enterobacterales isolates, Table S4: Distribution of aminoglycoside modifying enzyme (AME) genes in extended spectrum beta-lactamase producing Enterobacterales, Table S5: Distribution of TMP-SMX resistant genes in extended spectrum beta-lactamase producing Enterobacterales, Table S6: Distribution plasmid mediated quinolone resistance genes in extended spectrum beta-lactamase producing Enterobacterales isolates and its comparison with minimum inhibitory concentrations of fluoroquinolones tested, Table S7: Distribution of aminoglycoside modifying genes in extended spectrum beta-lactamase producing Enterobacterales isolates and its comparison with minimum inhibitory concentrations of aminoglycosides tested, Table S8: Distribution of trimethoprim-sulfamethoxazole (TMP-SMX) resistance genes in extended spectrum beta-lactamase producing Enterobacterales isolates and its comparison with minimum inhibitory concentrations of TMP-SMX tested.