Emergence of Ertapenem Resistance in an Escherichia coli Clinical Isolate Producing Extended-Spectrum β-Lactamase AmpC

Abstract

Escherichia coli isolate MEV, responsible for a bloodstream infection, was resistant to penicillins, cephalosporins, and ertapenem. Molecular and biochemical characterization revealed the production of a novel, chromosome-borne, extended-spectrum AmpC (ESAC) β-lactamase with a Ser-282 duplication and increased carbapenemase activity. This study demonstrates for the first time that chromosome-borne ESAC β-lactamases can contribute to the emergence of ertapenem resistance in E. coli clinical isolates.

TEXT

Escherichia coli produces a chromosomal AmpC β-lactamase at a very low level because of a weak promoter (4). Spontaneous mutations affecting the promoter region can induce overproduction of this enzyme, conferring resistance to narrow-spectrum cephalosporins (20). Moreover, structural alterations in the R2 binding site, which accommodates the R2 lateral side chain of β-lactams, can broaden the hydrolysis spectrum of AmpC β-lactamases toward extended-spectrum cephalosporins (ESCs), including cefepime (14, 28). Extended-spectrum AmpC (ESAC) β-lactamases, which constitute novel group 1e in the updated functional classification of Bush and Jacoby (3), also exhibited increased catalytic efficiency against carbapenems compared to that of the parental enzymes (28). However, at that time, the phenotypic expression of this weak carbapenemase activity was detected only in porin-deficient E. coli recombinant clones (24).

We described here for the first time the emergence of ertapenem resistance in an ESAC-producing E. coli clinical isolate. E. coli MEV was recovered in two blood cultures from a 50-year-old patient suffering from myeloma at the Morvan hospital (Brest, France) in April 2008. The patient had been hospitalized multiple times in the previous 6 months and had received multiple courses of antibacterials. On admission, the patient was empirically treated with vancomycin, ceftazidime, and ciprofloxacin. Once the bacteriological documentation became available, the treatment was changed to imipenem and vancomycin for 13 days combined with amikacin for the first 5 days. The course of the infection was favorable.

Susceptibility testing, which was performed as previously described (26), showed that E. coli MEV was resistant to penicillins, cephalosporins, and ertapenem but remained susceptible to imipenem according to the revised CLSI criteria (7 [June 2010 update]) (Table 1). The resistance was not antagonized by clavulanic acid, whereas the susceptibility to ESCs and ertapenem was partially restored by cloxacillin and phenylalanine arginine-β-naphthylamide (PAβN; Sigma-Aldrich), in agreement with AmpC β-lactamase production and efflux overexpression, respectively (3, 5). Furthermore, E. coli MEV was resistant to fluoroquinolones, gentamicin, and tobramycin.

Table 1.

MICs of β-lactams for E. coli MEV, the E. coli TOP10(pMEV) and TOP10(pAmpC-A) recombinant clones, and the E. coli TOP10 recipient strain

β-Lactam	E. coli MEVa	E. coli MEVa + CLOXAc	E. coli MEVa + PAβNd	E. coli TOP10(pMEV)a	E. coli TOP10(pAmpC-A)a	E. coli TOP10
Amoxicillin	>512	16	>512	>512	512	2
Amoxicillin-CLAb	512	8	512	512	512	2
Ticarcillin	256	4	128	256	8	1
Ticarcillin-CLAb	128	4	64	128	8	1
Piperacillin	256	4	128	256	8	1
Piperacillin-TZBe	128	4	64	64	8	1
Cefotaxime	64	0.5	16	32	1	0.06
Ceftazidime	256	1	128	256	2	0.06
Cefepime	16	0.5	2	2	0.06	0.06
Imipenem	0.5	0.25	0.5	0.25	0.06	0.06
Ertapenem	1	0.06	0.5	0.125	0.016	0.008
Meropenem	0.25	0.125	0.25	0.032	0.032	0.032

^aE. coli MEV and TOP10(pMEV) produced the extended-spectrum AmpC-MEV β-lactamase, E. coli TOP10(pAmpC-A) produced the narrow-spectrum AmpC-A β-lactamase, and the E. coli TOP10 recipient strain did not produce a β-lactamase.

^bCLA, clavulanic acid at 2 μg/ml.

^cCloxacillin (CLOXA) MICs were obtained using 250 μg/ml cloxacillin-containing Mueller-Hinton agar plates.

^dPAβN was used at 26.3 μg/ml (5).

^eTZB, tazobactam at 4 μg/ml.

PCR screening revealed that E. coli MEV did not harbor plasmid-mediated ampC genes (29). Moreover, isoelectric focusing analysis of the culture extract of E. coli MEV, which was performed as previously described (23), gave one β-lactamase activity with a pI value of 9.0. The entire chromosome-borne ampC gene was amplified and analyzed as described previously (20). Compared to the wild-type promoter of E. coli K-12, DNA sequence analysis of the bla_AmpC-MEV gene promoter showed a C→T transition at position −42 that made a perfect TTGACA box upstream of the native −35 sequence. Moreover, the T→A transversion at position −18 resulted in a new −10 box separated by 17 bp from the new −42 box, giving rise to a strong promoter (4). The analysis of the coding sequence of the bla_AmpC-MEV gene revealed a 3-bp insertion compared to the bla_AmpC-A gene that codes for the wild-type cephalosporinase of E. coli strains belonging to phylogenetic group A (6, 21, 25). This insertion led to a duplication of the serine residue located at position 282 in the H-9 helix (Fig. 1) (9, 20, 28). This structural alteration is reported here for the first time.

Fig. 1.

Alignment of the amino acid sequences of wild-type AmpC and ESAC β-lactamases of E. coli clinical isolates. AmpC-A is a narrow-spectrum enzyme from E. coli EC2 (21, 25), whereas the other β-lactamases are variants exhibiting extended hydrolysis spectra. The AmpC-MEV, AmpC-BER, and AmpC-ECB33 enzymes were produced by E. coli MEV (this study), BER (26), and ECB33 isolates (20), respectively. The AmpC-EC14, AmpC-EC15, AmpC-EC16, and AmpC-EC18 β-lactamases were produced by E. coli isolates EC14, EC15, EC16, and EC18, respectively (25). AmpC^D was produced by E. coli isolate HKY28 (9). The AmpC-7014517, AmpC-8009162, and AmpC-EC80 β-lactamases were produced by E. coli isolates 7014517, 8009162, and EC80, respectively (1, 8). The amino acids in the R2 loop are shaded light gray, whereas those in the H-9 helix are shaded dark gray.

The coding regions of the bla_AmpC-MEV and bla_AmpC-A genes were cloned into pCR-BluntII-Topo (Invitrogen, Cergy-Pontoise, France), and the recombinant plasmids were subsequently transformed into E. coli strain TOP10, giving rise to E. coli TOP10(pMEV) and TOP10(pAmpC-A) recombinant clones, respectively (26). In all of the recombinant plasmids, the orientation of the clone insert was the same, with the ampC gene under the transcriptional control of the lacZ promoter flanking the cloning site.

The AmpC-MEV and AmpC-A β-lactamases were extracted from the E. coli TOP10 recombinant strains and purified as described previously (26), yielding two extracts containing the proteins at 0.38 mg/ml and 1.52 mg/ml, respectively. The homogeneity of the purified extracts (>99%) was assessed by SDS-PAGE analysis (15). The specific activities, determined with 100 μM cephalothin as the substrate, were 60 and 490 μmol/min/mg of protein, respectively. The kinetic parameters, which were determined as previously described (26), are presented in Table 2. The k_cat and K_m values of AmpC-MEV β-lactamase were, respectively, increased and decreased with respect to those of AmpC-A for all ESCs and imipenem, accounting for enhanced catalytic efficiencies (k_cat/K_m).

Table 2.

Kinetic parameters of AmpC-MEV β-lactamase and the wild-type AmpC-A enzyme

ß-Lactams	AmpC-MEV			AmpC-A
	kcat (s−1)	Km (μM)	kcat/Km (μM−1.s−1)	kcat (s−1)	Km (μM)	kcat/Km (μM−1.s−1)
Cephalothin	135 ± 15c	18 ± 2	7.5	500 ± 20	30 ± 2	16
Cephaloridine	160 ± 12	38 ± 4	4.2	230 ± 15	650 ± 30	0.35
Cefoxitina	0.056 ± 0.01	0.08 ± 0.01	0.7	0.3 ± 0.02	1 ± 0.05	0.3
Cefotaximea	0.25 ± 0.005	0.05 ± 0.01	5	0.2 ± 0.005	25 ± 2	0.008
Ceftazidimea	0.78 ± 0.03	0.4 ± 0.08	1.95	0.5 ± 0.03	70 ± 3	0.007
Cefepime	1.5 ± 0.2	60 ± 4	0.025	NDb	>1,000	>0.01
Imipenema	0.01 ± 0.005	0.3 ± 0.08	0.03	0.1 ± 0.005	10 ± 0.4	0.01

^aFor those compounds with a K_m value of <5 μM, the K_i was determined, instead of the K_m, with cephaloridine as the substrate.

^bND, not determinable.

^cThe values shown are the means ± the standard deviations of at least three independent experiments.

Surprisingly, hydrolysis of ertapenem by AmpC-MEV and AmpC-A extracts was not detectable. Such a discrepancy between phenotypic and biochemical results has already been reported (22) and might be attributable to the low but nonzero deacylation rate of AmpC β-lactamases for this compound (16).

MICs for the E. coli TOP10(pMEV) and TOP10(pAmpC-A) recombinant strains are presented in Table 1. AmpC-MEV β-lactamase conferred higher levels of resistance to all oxyiminocephalosporins than the AmpC-A enzyme, thus confirming that the Ser-282 insertion extended the hydrolysis spectrum. In contrast, E. coli strain TOP10(pMEV) remained susceptible to ertapenem, although the MIC was slightly increased compared to that for wild-type E. coli strain TOP10 (4-fold increase). This result suggested that the E. coli MEV isolate expressed an additional mechanism of resistance affecting its susceptibility to ertapenem.

The AcrAB efflux transporter and the marA gene, which encodes a transcriptional activator of the acrRAB operon, were determined by quantitative reverse transcription (RT)-PCR as previously described (11, 13). This showed that the expression of the AcrAB transporter was significantly increased (12-fold) in E. coli MEV whereas the expression of the MarA transcriptional activator was decreased (11-fold) (Fig. 2).

Fig. 2.

RT-PCR analysis of marA and acrA expression. Total bacterial RNA was isolated from mid-log-phase cultures of E. coli isolate MEV and wild-type E. coli strain TOP10. The normalized expression ratios of the acrA and marA genes are shown in panels A and B, respectively. The error bars represent the standard deviations of the means of triplicate samples.

Amplification of the acrR gene, which encodes a transcriptional repressor of the acrA gene (19), was performed with primers AcrR-F (5′-GCTGCGTTTATATTATCGTCGTGC-3′) and AcrR-R (5′-GTCAAACCGCAAGAATATCACGACG-3′) and a standard PCR protocol (denaturation for 10 min at 94°C; 35 cycles of 1 min at 94°C, 1 min at 55°C, and 1 min at 72°C; and a final extension step of 10 min at 72°C). This yielded an 815-bp amplicon for E. coli strain TOP10, whereas it was negative for E. coli MEV, suggesting that the AcrAB efflux overexpression seen in E. coli MEV is related to the deletion of the acrR gene. Whole-cell DNA of E. coli MEV was extracted and transferred onto a nylon membrane (31). Hybridization of the membrane with a fluorescein-labeled probe that was made of the PCR product of the acrR gene of E. coli strain TOP10 (23) failed, thus confirming the deletion of the acrR gene in E. coli MEV. Nevertheless, further studies, such as homologous recombination, are needed to confirm whether AcrAB efflux transporter overexpression contributes to ertapenem resistance in E. coli.

The outer membrane protein (OMP) profiles of E. coli clinical isolate MEV and control strains expressing OmpC and/or OmpF porins (27) were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described previously (30). Comparison of the OMP profiles showed weak expression of OmpC and lack of OmpF protein in E. coli MEV (Fig. 3), which might explain the additional resistance to ertapenem seen in E. coli clinical isolate MEV (10, 30).

Fig. 3.

OMP profiles of E. coli strains. OMPs were profiled by SDS-PAGE. The amounts of protein added to the gel lanes were normalized. Lanes: 1, E. coli TOP10 wild-type strain; 2, E. coli isolate JF703 expressing OmpC alone (27); 3, E. coli clinical isolate MEV. The horizontal arrows on the left indicate the positions of OMPs OmpC and OmpF.

Carbapenem resistance is an emerging phenomenon among E. coli clinical isolates. To date, it has been related to transmissible β-lactamases with strong carbapenemase activity, such as metallo-β-lactamases and KPC-type and OXA-48-type β-lactamases or transmissible β-lactamases with weak carbapenemase activity, such as CTX-M-type β-lactamases or some plasmid-mediated AmpC β-lactamases, in combination with a lack of outer membrane permeability (2, 12, 17, 18, 22, 30, 32). This study demonstrates for the first time that chromosome-borne ESAC β-lactamases can also contribute to the emergence of ertapenem resistance in E. coli clinical isolates.

Nucleotide sequence accession number.

The nucleotide sequence of the bla_AMPC-MEV gene of E. coli MEV has been deposited in the EMBL nucleotide sequence database under accession number HQ419012.