Synergistic Antibacterial Effects of Meropenem in Combination with Aminoglycosides against Carbapenem-Resistant Escherichia coli Harboring blaNDM-1 and blaNDM-5

Infections due to carbapenem-resistant Escherichia coli (CREC) are problematic due to limitation in treatment options. Combination therapies of existing antimicrobial agents have become a reliable strategy to control these infections. In this study, the synergistic effects of meropenem in combination with aminoglycosides were assessed by checkerboard and time-kill assays. Of the 35 isolates, 19 isolates (54.3%) were resistant to carbapenems (imipenem and meropenem) with the MIC ranges from 16 to 128 µg/mL. These isolates were resistant to almost all antibiotic classes. Molecular characteristics revealed co-harboring of carbapenemase (blaNDM-1, blaNDM-5 and blaOXA-48) and extended-spectrum β-lactamases (ESBL) genes (blaCTX-M, blaSHV and blaTEM). The checkerboard assay displayed synergistic effects of meropenem and several aminoglycosides against most CREC isolates. Time-kill assays further demonstrated strong synergistic effects of meropenem in combination with either amikacin, gentamicin, kanamycin, streptomycin, and tobramycin. The results suggested that meropenem in combination with aminoglycoside therapy might be an efficient optional treatment for infections cause by CREC.


Introduction
Infections due to carbapenem-resistant Escherichia coli (CREC), particularly the New Delhi metallo-β-lactamases (NDM)-producing isolates, are critically problematic to global health care [1]. These infections usually yield unfavorable clinical outcomes, prolonged length of hospitalization and high hospital costs [2]. The national antimicrobial resistance surveillance data reported by the Thailand National Institute of Health (2016-2018), indicated a high prevalence of carbapenem-resistant Enterobacteriaceae (CRE) (93%) among hospitalized patients in Thailand [3]. In the past, carbapenems were the most reliable antimicrobial agents against hospital-acquired infections caused by extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae [4]. However, extensive usage as both empirical and definitive regimens [5], resulted in the emergence of CRE [4].
Enterobacteriaceae resistance to carbapenems is mainly associated with the production of several kinds of carbapenemases, which are enzymes capable of hydrolyzing carbapenems and other β-lactams [6]. In addition, the lack of porin proteins by alteration in the permeability of the bacterial cell membrane, and overexpression of efflux pumps are additive carbapenem resistance mechanisms [7]. Numerous epidemiological studies have suggested that the acquisition of carbapenemase-encoding genes might lead to a rapid outbreak mostly in the hospital-setting and sometimes in the community-setting [8][9][10]. Moreover, the specific class of the carbapenemase should be considered during the development of novel antimicrobial agents as each class possesses a unique mechanism and spectrum of activity [11]. Previous studies have reported that ceftazidime-avibactam binds reversibly to class A, C, and some D β-lactamases [12,13], whereas imipenem-cilastatin-relebactam and meropenem-vaborbactam reversibly and competitively inhibited class A and C βlactamases [14,15]. However, these antibiotics did not inhibit metallo-β-lactamases such as NDM carbapenemases [12,14,15]. Globally, the predominant carbapenemases include NDM, Klebsiella pneumoniae carbapenemase (KPC), Verona integrin-encoded metallo-βlactamase (VIM), imipenemase (IMP), and oxacillinases (OXA)-type enzymes, which are encoded by bla NDM , bla KPC , bla VIM , bla IMP , and bla OXA genes, respectively [6]. However, bla NDM has gained relevance due to the high-level of resistance to many clinically available β-lactams and ease of horizontal transfer between different isolates. To date, several variants of NDM enzymes have been identified [16] with amino acid substitutions at different positions. NDM-5 differed from NDM-1 by substitutions at positions 88 (Val→Leu) and 154 (Met→Leu), and several studies have showed that bla NDM-5 is carried by conjugatable IncX3 plasmids responsible for the rapid spread [17][18][19].
Currently, therapeutic options for the management of infections caused by CREC are limited [20]. Moreover, the development of new antimicrobial agents are costly, timeconsuming, and require various stages of toxicological evaluations to ensure safety [11]. Hence, combining existing antimicrobial agents has become a strategy against several kinds of infections caused by multi-drug resistant (MDR) organisms [21]. Previous studies have supported the use of combination therapy as an effective treatment option for infections caused by several MDR Gram-negative bacteria [22][23][24]. A recent study demonstrated the synergistic effect of meropenem and aminoglycosides against KPC-2 and NDM-1producing carbapenem-resistant Klebsiella pneumoniae [25]. Additionally, the ability of meropenem to potentiate aminoglycoside activity, largely dependent on the MexXY-OprM multidrug efflux system, has been shown [26]. However, data for combinations between meropenem and several aminoglycosides against CREC harboring bla NDM genes is lacking. This study evaluated the effects of meropenem in combination with several commonly used aminoglycosides (amikacin, gentamicin, kanamycin, streptomycin, and tobramycin) on CREC isolates harboring bla NDM genes.

Bacterial Isolates
A total of 35 suspected CREC isolates were collected from eight hospitals located in Southern Thailand. The isolates were obtained from various clinical specimens, including blood (n = 11), rectal (n = 19), throat (n = 3) and environment (n = 2). Data of isolates and antimicrobial response to imipenem and meropenem are shown in Supplementary Materials Table S1. The results indicated that 19 isolates were resistant to carbapenems. Demographic information, clinical data and outcomes of the patients infected with CREC are presented in Table S2. Similar to previous reports of risk factors associated with CRE acquisition or infection [27,28], most of patients in this study had previous exposure to various antimicrobial agents, particularly carbapenems. The results support previous observation that exposure to antibiotics including β-lactams such as carbapenems and cephalosporins, as well as fluoroquinolones were associated with CRE [23]. Patient information indicated that most of the patients were admitted in intensive care units (ICU), which are in consonance with observations of a previous study that showed high prevalence of carbapenemase producing Enterobacteriaceae in the ICU [29].
To date, polymyxins, fosfomycin, aminoglycosides, and tigecycline are considered choice drugs for the management of infections caused by carbapenem-resistant Gramnegative bacteria [30]. However, resistance to these antibiotics is increasing rapidly with high chance of toxicity due to the relative high doses required for monotherapy medications. Results of this study revealed that approximately 79% of CREC isolates were resistant to tigecycline, contrary to previous reports of 0.7% and 11.2% [31,32]. In addition, the low plasma levels of tigecycline [33] constitutes a clinical concern for mono-therapeutic administration. Polymyxin on the other hand showed excellent antimicrobial effects against CREC with a 100% susceptibility. However, the nephrotoxicity and poor tissue perfusion of polymyxins [34] are limiting factors hindering extensive therapeutic usage. The rapid acquisition of resistance and sodium overload with intravenous fosfomycin [35] are also of clinical concern.

Antimicrobial Susceptibility to Carbapenem and Aminoglycosides
The MIC of carbapenems and aminoglycosides on 19 CREC isolates were determined by the broth microdilution method (   Aminoglycosides are an important class of bactericidal antibiotics that are frequently used for the treatment of severe infections caused by Gram-negative bacteria. The major resistance mechanism to aminoglycosides in Gram-negative bacteria is the production of aminoglycoside-modifying enzymes (AMEs) or the modification of ribosome by acquired 16S rRNA methyltransferases (RMTases) [36,37]. AMEs modify select to specific aminoglycosides, hence bacterial isolates show discordant susceptibility among different aminoglycosides.
A previous study demonstrated the co-occurrence of aminoglycoside and β-lactam resistance mechanisms in E. coli isolates [38]. In addition, co-harboring of ESBLs, carbapenemases, and 16S rRNA methylase genes within a plasmid have been noted to result in multidrug-resistance in Enterobacteriaceae [39].

Genotypic Resistance Mechanism in Carbapenem-Resistant E. coli Isolates
The 19 CREC isolates were screened for antimicrobial resistance genes including carbapenemase genes (bla KPC , bla IMP , bla VIM , bla NDM , and bla OXA-48 ) and ESBL genes (bla TEM , bla SHV , and bla CTX-M ) using PCR ( Table 3). The results for carbapenemase genes, demonstrated high prevalence of bla NDM-1 and bla NDM-5 . However, bla OXA-48 was observed in one of the tested isolates. Furthermore, co-harboring of carbapenemase and ESBL genes were represented in almost all isolates. The results showed that six isolates with bla NDM-1 co-harbored bla CTX-M and bla TEM (Table 2). Additionally, CREC 18 carrying bla NDM-1 and bla OXA-48, co-harbored ESBL genes (bla CTX-M and bla TEM ). bla NDM-5 was found in nine isolates co-harboring ESBL genes (bla CTX-M and bla TEM ). However, two out of the nine isolates that harbored bla NDM-5 had only bla TEM . The results further showed that three of the isolates had no carbapenemase genes but carried ESBL genes. According to the Ambler classification method, carbapenemase-produced by Enterobacteriaceae can be classified into three classes including class A, class B, and class D β-lactamases [6]. However, the clinical relevance of Ambler class C is still unknown [40]. The most widely spread carbapenemase in E. coli include class A; KPC, class B; NDM-1, NDM-5, NDM-9, and VIM, class D; OXA-48, OXA-181, and OXA-244 [41,42]. Class A, B and D β-lactamases enzymes are plasmid-mediated and are responsible for the high levels of antimicrobial resistance and rapid dissemination by horizontal transfer [43]. Epidemiological studies have revealed the diversity of carbapenemases predominate in several regions and countries [43]. In the United States, Argentina, Columbia, Greece, Israel, and Italy, KPC-producing Enterobacteriaceae, are mostly endemic among nosocomial isolates [1]. NDM was reported as the main carbapenemase-mediating resistance in E. coli isolates in India, Pakistan, and Sri Lanka, whereas OXA-48 was reported in North Africa, Malta, and Turkey [44]. NDM and OXA-48 were identified in both nosocomial and community-acquired pathogens [43,45]. A recent study done in Thailand reported a high prevalence (99%) of CREC isolates having at least one carbapenemase-producing gene (CP-gene) [3]. The most common CP-gene among CREC isolates in Thailand were bla NDM (94%) and a bla OXA-48-like (18%) gene [3]. In this study, bla NDM was found in 16 isolates, including seven isolates harboring bla NDM-1 and nine isolates harboring bla NDM-5 . Similar results were reported in a recent study with a high prevalence of NDM-1 in E. coli [46]. The increased usage of antibiotics maybe driving the evolution of NDM-1 variants. M154L amino acid substitution in NDM-5 was the most common substitution in all NDMs variants leading to increase carbapenemase activity [47]. However, a previous study reported that the difference in the activity of NDM-5 and NDM-1 is due to variations in the affinity for zinc [48]. Moreover, V88L amino acid substitution in NDM-5 contribute to lower catalytic activity on imipenem and meropenem [49]. Several studies showed that bla NDM-5 was carried by IncX3 plasmids which have been shown to be conjugatable and could explain the rapid spread of bla NDM-5carrying isolates [50]. However, bla KPC which is the most commonly found in the United States [1], was not presented in this study. So far, the prevalence of bla KPC in Thailand has remained very low. A previous report indicated a 0.02% (n = 12,741) prevalence of bla KPC-13 among Enterobacteriaceae and 1.7% (n = 181) among CRE isolates [51], whereas a separate report showed that the prevalence rate of bla KPC-2 in CRE isolates was 0.13% (n = 2245) [52]. Furthermore, the study illustrated the co-existence of carbapenemase and ESBL genes in CREC isolates. Carbapenems were used as first-line antibiotic for treatment of infection caused by extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae. Thus, the co-harboring of multiple antibiotic resistance genes will promote multi-resistance, which might amount to significant therapeutic concerns.

The Combined Effect of Meropenem and Aminoglycosides
The results of antimicrobial combinations against the 19 CREC isolates are shown in Table 4 and summarized in Table S4. Synergistic effects were observed for meropenem plus gentamicin and meropenem plus streptomycin in 16 (84.2%) isolates, followed by meropenem plus kanamycin and meropenem plus tobramycin in 15 (79%) isolates. Furthermore, synergistic activity was observed in 13 (68.4%) isolates for meropenem plus amikacin. The isolate CREC 11 (bla CTX-M and bla TEM ), with high resistance to aminoglycosides, was resistant to all combinations, while isolate CREC 12 (bla NDM-5 , bla CTX-M and bla TEM ) was susceptible to meropenem plus amikacin, or gentamicin, or streptomycin combinations. Combination of meropenem plus gentamicin and meropenem plus tobramycin exhibited synergism against CREC 14 (bla NDM-5 and bla TEM ). The cross resistance of CREC 11 to all the combinations might be due to the cumulative effects of other resistance mechanisms such as overexpression of efflux pump and/or porin with the β-lactamases leading to high level of resistance. However, the results did not reveal an antagonistic effect for the tested combinations.
The results revealed that addition of aminoglycosides as adjunctive therapy to meropenem could restore meropenem activity against CREC isolate harboring bla NDM . Combination of meropenem and aminoglycosides might promote membrane disruption since aminoglycosides exert disruptive effects on the outer membrane structure by binding with the negatively charged lipopolysaccharides in the outer membrane of Gram-negative bacteria. Thus, the aminoglycoside promotes the permeabilizing effect and enhances the periplasmic target site penetration of other antibiotics such as carbapenems used in combination [55,56]. Meropenem is a safe, well-tolerated, and commonly used as monotherapy or as combination regimens for hospital-acquired infection due to several MDR Gram-negative bacteria [57][58][59]. Similarly, aminoglycosides are effective against Gram-negative aerobic bacteria including resistant Enterobacteriaceae [60]. However, aminoglycosides monothera-pies can lead to unfavorable clinical outcomes due to rapid emergence of resistance, and nephrotoxicity among patients with prolonged usage of aminoglycosides [61,62].

Bacterial Collection and Identification
A total of 35 suspected CREC isolates were collected from eight hospitals located in Southern Thailand. The isolates grew on MacConkey agar supplemented with imipenem at 6 µg/mL. All isolates were identified to species level using standard biochemical tests

Bacterial Collection and Identification
A total of 35 suspected CREC isolates were collected from eight hospitals located in Southern Thailand. The isolates grew on MacConkey agar supplemented with imipenem at 6 µg/mL. All isolates were identified to species level using standard biochemical tests and MALDI-TOF-MS. E. coli ATCC 25922 was used as quality control. The samples were kept in tryptic soy broth supplemented with 20% glycerol and stored at −80 • C.

Screening for Carbapenem Resistance
Resistance of the 35 suspected CREC isolates was assessed by the broth microdilution method according to the Clinical and Laboratory Standards Institute [65]. Briefly, the isolates were grown in cation-adjusted Mueller-Hinton broth (CAMHB). Bacterial cultures were adjusted with sterile 0.85% NaCl to McFarland 0.5 turbidity standard. Aliquot of 100 µL diluted bacterial suspension (1 × 10 6 CFU/mL) was mixed with 100 µL antibiotic in a 96-well plate and incubated at 37 • C for 18 h. The minimum inhibitory concentration (MIC) was expressed as the lowest concentration of the antibiotic that inhibits visible growth after incubation as indicated by the resazurin test.

Genotypic Determination of Carbapenemase and ESBL
Genomic DNA from E. coli was prepared using Presto TM Mini gDNA Bacteria Kit. Quantification of the extracted DNA was determined by spectroscopy at 260 nm. Antimicrobial resistance genes, including carbapenemase (bla IMP , bla KPC , bla NDM , bla OXA-48 , and bla VIM ) and ESBL (bla CTX-M , bla SHV , and bla TEM ) were detected by PCR using the primers shown in Table 3. The amplification conditions for detecting IMP, KPC, and OXA-48 genes were initial denaturation at 94 • C for 10 m, 36 cycles of 94 • C for 30 s, 52 • C for 40 s, and 72 • C for 50 s, and final elongation at 72 • C for 5 m. The amplification condition for NDM and VIM genes were initial denaturation at 94 • C for 10 m, 36 cycles of 94 • C for 30 s, 56 • C for 40 s, and 72 • C for 50 s, and final elongation at 72 • C for 5 m. The amplification conditions for detecting ESBL genes included CTX-M, SHV, and TEM genes were initial denaturation at 95 • C for 15 m, 30 cycles of 94 • C for 30 s, 60 • C for 30 s, and 72 • C for 2 m, and final elongation at 72 • C for 10 m.

Checkerboard Technique
The synergistic activities of meropenem combined with five aminoglycosides (amikacin, gentamicin, kanamycin, streptomycin, and tobramycin) on CREC were determined by the checkerboard technique. Briefly, 100 µL of 1 × 10 6 CFU/mL bacterial suspension was added to wells containing 50 µL of each subinhibitory concentrations of meropenem and aminoglycosides. The plates were incubated for 18 h at 37 • C. Inhibitory concentrations were determined as concentrations without bacterial growth as indicated by the resazurin test. The experiments were performed in triplicate for three independent repeats. The activity of the antimicrobial combinations was defined by the fractional inhibitory concentration index (FICI), as follows:

Time-Kill Assay
The activity of meropenem and aminoglycosides combinations were confirmed by the time-kill assay. Antibiotics were tested alone and in combination at 1/4 MIC. An inoculum size of 1 × 10 6 CFU/mL was added and incubated at 37 • C. Bacterial growth controls were maintained throughout the experiment. Bacterial growth was assessed at 0, 2, 4, 8, 12 and 18 h by plating 10-fold serially diluted suspensions on Mueller-Hinton agar plates. Plates were incubated overnight at 37 • C, and the number of colonies were counted. The experiments were performed in triplicate and recorded as mean averages. Bactericidal activity was defined as a ≥3 log 10 CFU/mL reduction when compare the number of viable cells at time zero (0 h). Antibiotic combination synergism was defined as a ≥2 log 10 CFU/mL at 18 h for the antimicrobial combination, compared with the most active agent. Indifferent was defined as <2 log 10 CFU/mL increase or decrease at 18 h for the drug combination when compare with the most active drug and antagonism was defined as ≥2 log 10 CFU/mL increase between the combination and the most active single drug [67].

Conclusions
Combination therapies have been highlighted as a possible treatment option for the management of infections caused by drug resistant bacterial isolates. This study demonstrated that combinations of meropenem with aminoglycoside might still be an efficient therapeutic option for the treatment of CREC harboring bla NDM-1 and bla NDM-5 . However, due to indifferent results observed with the FICI, it is important to consider other mechanisms of aminoglycoside and carbapenem co-resistance. In addition, further studies on toxicology, pharmacokinetics and pharmacodynamics of these combination regimens are required prior to clinical trials.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/antibiotics10081023/s1, Table S1: Screening for carbapenem resistance in 35 suspected carbapenem-resistant Escherichia coli isolates, Table S2: Clinical information and outcome of patients in 19 carbapenem-resistant Escherichia coli (CREC) isolates, Table S3: Minimum inhibitory concentrations of antimicrobial agents against the 19 carbapenem-resistant Escherichia coli isolates, Table S4: Summary of the synergistic effects of meropenem in combination with aminoglycosides against 19 carbapenemresistant Escherichia coli.
Author Contributions: P.T., experimental design, experimentation, manuscript drafting, and data analysis; O.F.N., experimentation, writing, editing, and revision; T.S., experimental design, supervision; S.P.V., providing funding, supervision, and editing; S.C., conceptualization, providing funding, supervision, and editing. All authors have read and agreed to the published version of the manuscript. Institutional Review Board Statement: This retrospective study was approved by the Institutional Review Board (IRB) of the Faculty of Medicine, Prince of Songkla University, Thailand (EC: 54-080-14-1-2). The researchers were granted permission to extract the data from the database with a waiver of consent because of the observational nature of the study. All data were fully anonymized before being accessed and analyzed. Data Availability Statement: Data is contained within the article or Supplementary Material.

Acknowledgments:
The authors wish to thank the Postdoctoral Fellowship, Prince of Songkla University.

Conflicts of Interest:
The authors declare no conflict of interest.