INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Many enterococcal isolates are resistant to the majority of antimicrobial agents that are in clinical use. This causes therapeutic problems, especially for patients with serious, multidrug-resistant enterococcal infections. Enterococci are equipped with a variety of antibiotic resistance genes, some of which are inherent and some of which are acquired. The intrinsic resistance of enterococci to certain antibiotics may be partially due to genes that encode multidrug resistance efflux pumps (MDRs).

MDRs are found in a variety of organisms ranging from bacteria to humans, suggesting their universal importance. Although the primary functions and mechanisms of action for most MDRs remain unclear, these proteins are known to actively transport toxic compounds out of the cell. Antibiotic-specific efflux pumps ([e.g, Tet(K) and CmlA)] are often encoded on transmissible plasmids and transposons in bacteria. However, most genes for MDRs are encoded on the bacterial chromosome (6).

In a number of other bacteria, MDRs have also been implicated as important contributors to multidrug resistance (3, 4, 5, 7). Because enterococci are intrinsically resistant to a variety of antibiotics, it seems reasonable to assume that MDRs could be contributing to the drug resistance. However, there has been only one study on MDRs in enterococci prior to the present study. Enterococci were shown to extrude chloramphenicol, tetracycline, and norfloxacin; but the efflux pump(s) responsible for extruding these drugs was not identified (9). The need to identify drug efflux pumps, to find suitable drug targets, and to understand drug resistance has led to our investigation of MDRs in enterococci. Previously, we used a genomics approach to identify 34 potential enterococcal MDR-encoding genes (2). We have now characterized an enterococcal MDR, EmeA, which was identified because of its sequence homology to Staphylococcus aureus NorA, and have demonstrated that EmeA is involved in the resistance of enterococci to various unrelated compounds. Known MDR inhibitors were also effective at inhibiting drug efflux in enterococci. Our analysis demonstrates that EmeA is an MDR that belongs to the major facilitator superfamily (MFS).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Bacterial strains, plasmids, and culture conditions. The bacterial strains and plasmids used in the study are listed in Table 1. All Escherichia coli strains were grown by using standard conditions (15). All Enterococcus faecalis strains were grown with aeration at 37°C in brain heart infusion (BHI), Todd-Hewitt (TH; Difco Laboratories, Detroit, Mich.), or Mueller-Hinton cation-adjusted (MH II; Becton Dickinson) broth or agar supplemented with appropriate antibiotics. The antibiotics used for E. coli included ampicillin at 100 µg/ml, kanamycin (KAN) at 25 µg/ml, chloramphenicol at 25 µg/ml, and erythromycin (ERY) at 300 µg/ml. The antibiotics used for E. faecalis included KAN at 1 to 2 mg/ml and ERY at 10 µg/ml.

Chemicals and antibiotics. Antibiotics, carbonyl cyanide m-chlorophenylhydrazone (CCCP), ethidium bromide (EtBr), verapamil, lansoprazole, and reserpine were purchased from Sigma-Aldrich Chemical Co. (St. Louis, Mo.). CCCP was dissolved in dimethyl sulfoxide (100 mM); verapamil was dissolved in distilled H₂O (dH₂O; 10 mg/ml); reserpine was dissolved in CHCl₃ (10 mg/ml) and then diluted in MH II broth, as needed; lansoprazole (125 mg) was first dissolved in 5 ml of methanol and was then diluted to 5 mg/ml in dH₂O; and norfloxacin was suspended in dH₂O (one-half of the total volume needed), and 100 mM NaOH was added (in 100-µl aliquots) until the solution became clear and then the solution was diluted to the desired concentration with dH₂O.

MIC testing. MIC testing was performed with MH II medium in microtiter plates and on agar dilution plates by the protocols provided by NCCLS (11). The compounds tested were diluted and tested at several different concentrations in order to determine more precise MICs. MICs were recorded as the lowest concentration of antibiotic that totally inhibited growth after 20 h of incubation at 35°C. The compounds tested included EtBr, norfloxacin, verapamil, lansoprazole, and reserpine. Each test was performed at least in triplicate. As a control, E. faecalis ATCC 29212 was included in tests for MICs. Strains tested on E-test strips (PDM Epsilometer test; AB Biodisk North America, Inc., Piscataway, N.J.) were suspended in 0.8% saline to a 0.5 McFarland standard (~10⁷ CFU/ml); the bacteria were then swabbed onto Mueller-Hinton agar plates and allowed to dry for about 10 min. After the plates were dry, the E-test strips were placed in the middle of the plate with the labeled side up. The plates were incubated overnight at 37°C. The MICs were determined as the lowest concentration of drug that totally inhibited growth.

EtBr efflux assay. Cultures of bacteria were grown overnight in BHI broth at 37°C. The cells were then pelleted and washed two times with 20 mM HEPES buffer (pH 7.0). The cells were resuspended in HEPES buffer to 30 Klett units (~2 × 10⁸ CFU/ml). The cells were then loaded with EtBr by shaking of the cells at 37°C and the addition of CCCP (final concentration, 40 µM) to dissipate the membrane potential and EtBr (final concentration, 2.5 µM). The cells were incubated for 1 h. The cells were then washed three times with HEPES buffer containing EtBr (2.5 µM) and were resuspended in the same buffer to 15 Klett units (~2 × 10⁷ CFU/ml). The cells were stored on ice until initiation of efflux. Efflux was initiated by the addition of 80 µl of TH broth to 3 ml of the cell suspension. In some cases, efflux pump inhibitors (reserpine at 20 µg/ml, verapamil at 100 µg/ml, lansoprazole at 100 µg/ml) were added before the addition of TH broth. The concentrations of the efflux pump inhibitors were chosen on the basis of previous research (1). Fluorescence was measured with a spectrophotomer (Photon Technology International, Lawrenceville, N.J.) every 30 s for 30 to 35 min with an excitation wavelength of 500 nm and an emission wavelength of 590 nm. All measurements were performed in triplicate.

DNA preparations and transformation. DNA preparation, purification, restriction digestion, agarose gel electrophoresis, and ligation were performed by standard methods or following the manufacturer's instructions (15). E. coli transformations and cloning and transformations with the TA and TOPO cloning vectors were performed by the protocol provided by Invitrogen (Carlsbad, Calif.). Electroporation of E. faecalis was performed as described previously (15).

DNA sequencing and sequence analysis. Sequencing reactions were performed by the Taq dye-deoxy terminator method with a 377 DNA sequencing system (PE Applied Biosystems, Foster City, Calif.). Transmembrane sequences were identified by using TMHMM or SOSUI at ExPASy. Also, SignalP was used to identify signal sequences. BLAST program searches were performed by using the National Center for Biotechnology Information (NCBI) website, the Baylor College of Medicine Search Launcher website, and GCG (Genetics Computer Group, Madison, Wis.) programs. The ORF Finder at the NCBI website was used to identify open reading frames. Amino acid sequences were obtained by using Entrez at the NCBI website. Multiple sequence alignment was performed with the ClustalW (version 1.8) program at the Baylor College of Medicine Search Launcher website.

PCR and protocols. The sequences of the primers used in the PCRs are shown in Table 2. PCR reagents and enzymes were purchased from Perkin-Elmer. Standard and colony PCRs were performed as described previously (15). Long-range PCR was performed with a GeneAmp XL PCR kit according to the manufacturer's recommendations.

Crossover PCR. Crossover PCR was performed as described previously (8). Primers 511 and 502 were used to amplify the left flanking region (~534 bp) of emeA. Primers 499 and 500 were used to amplify the right flanking region (~1,073 bp). The flanking regions were amplified from cosmid template BO1C6(2), which contains emeA on an ~40-kb insert of E. faecalis strain OG1RF chromosomal DNA (Table 1). The crossover PCR product of the flanking regions (1.607 kb) was gel purified and extracted and was then ligated into the TOPO-TA cloning vector (Invitrogen) and transformed into E. coli TOP10F' by the TOPO cloning protocol. The correct transformant strain, GW4475, which contained the construct pGW4475, was verified by restriction enzyme analysis and DNA sequencing.

Insertional mutagenesis. A disruption mutant of emeA was obtained by standard methods (15). An internal fragment of emeA (~603 bp) was amplified with primers 455 and 456. The PCR product was cloned between the BamHI and EcoRI sites of the pTEX4577 vector. This construct (pGW4427) was electroporated into E. faecalis OG1RF competent cells.

Deletion of emeA from chromosome of OG1RF and complementation. The 1.6-kb insert of the flanking regions was released from pGW4475 with EcoRI and was ligated into the EcoRI site of pTEX4577 (resulting in plasmid pGW4476), transformed into E. coli DH5 cells, and plated onto Luria-Bertani agar containing KAN (25 µg/ml), isopropyl--D-thiogalactopyranoside, and 5-bromo-4-chloro-3-indolyl--D-galactopyranoside. Plasmid DNA from transformant colonies was verified by digestion with EcoRI, agarose gel electrophoresis, and DNA sequencing. The pGW4476 construct was then electroporated into OG1RF competent cells, and the cells were plated onto TH agar with KAN (2 mg/ml). After overnight incubation, several colonies were obtained and restreaked. Integration of pGW4476 into the chromosome of OG1RF was verified by colony PCR with primers 513 and 516. Transformant GW4477 (OG1RF::pGW4476) was used to obtain the deletion mutant. First, an overnight culture was diluted 100-fold and was grown in BHI broth at 37°C for 90 min. Aliquots were plated onto TH agar plates, and colonies were then replica plated onto TH agar plates with and without KAN (2 mg/ml). Colonies that were no longer resistant to KAN were then tested for possible deletion of emeA by colony PCR with primers 500 and 511.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Characterization of emeA. emeA was identified by searching the E. faecalis V583 database with the sequence of the S. aureus NorA protein (accession number M80252). Once the ORF was identified, the enterococcal protein sequence was used to search the nonredundant protein database (NCBI website). EmeA shows 32% identity to NorA from S. aureus and 39 to 40% identities to Bmr2 and Bmr1 efflux pumps, respectively, from Bacillus subtilis. EmeA consists of 380 amino acids and has 12 predicted transmembrane regions and a predicted signal peptide. These characteristics are similar to those of other MFS MDRs (13).

View larger version (86K): [in this window] [in a new window]   FIG. 1.   Multiple sequence alignment. ClustalW was used to analyze sequences from the NCBI website. The translocase consensus sequence (GXXXDR/KXGRR/K) and the DE consensus sequence (GXhyhyGPXXGG [where hy represents hydrophobic residues and X is any residue]) are labeled and highlighted. Hyphens indicate that no residues are present, residues on a black background indicate conserved residues, and residues on a shaded background indicate related residues.

Mutagenesis of emeA in E. faecalis and MIC testing. In order to determine if the emeA ORF encodes a product involved in resistance to toxic compounds, we initially used insertional mutagenesis to inactivate the gene. Following electroporation of strain OG1RF with pGW4427, 14 KAN-resistant colonies were obtained and their structure was verified by long-range colony PCR with primers 479 and 463. All 14 colonies produced 9.0-kb products, indicating that integration of pGW4427 into the chromosome of OG1RF had occurred. Therefore, the enterococcal emeA had been disrupted. Southern hybridization also confirmed the integration (data not shown).

Deletion of emeA from OG1RF chromosome, complementation, and susceptibility testing of resulting strains. In order to eliminate the effects of truncated products or the possibility of polar effects on other genes, we sought to delete the gene from the chromosome. Colonies obtained through screening of GW4477 for the loss of KAN resistance were tested for deletion of emeA by PCR. Colony PCR yielded a 2.5-kb product for OG1RF and a 1.6-kb product for GW4481, suggesting deletion of emeA from the OG1RF chromosome. Sequencing showed that emeA was deleted from base pairs 7 through 1079 in strain GW4481.

EtBr efflux assay. EtBr is highly fluorescent when it is bound to DNA inside the cell. When the EtBr is effluxed from the cell by way of an efflux pump, the fluorescence will decrease and this can be measured with a fluorescence spectrophotometer. The fluorescence of EtBr was used to measure the efflux of the compound from bacterial cells. The bacterial cells were starved and loaded with EtBr after dissipation of the membrane potential with the protonophore CCCP. Without an energy source the MDRs cannot function; therefore, the EtBr cannot be effluxed and the cells become loaded. After the cells were loaded, they were washed to remove the CCCP and were resuspended in buffer containing EtBr. Efflux of EtBr was initiated by the addition of a carbon source that allows reconstitution of the membrane potential. Figure 2A shows that in the presence of TH broth, wild-type strain OG1RF was able to efflux EtBr. However, without TH broth or the EmeA MDR, EtBr was not pumped out of the bacterial cells. After 35 min there was a 30% difference in fluorescence between wild-type OG1RF (with TH broth) and OG1RFemeA (with TH broth). No decrease in fluorescence was seen in the mutant or the wild type suspended in buffer only. These results indicate that EmeA is an energy-dependent MDR.

View larger version (29K): [in this window] [in a new window]   FIG. 2.   EtBr efflux. (A) Efflux was initiated by the addition of TH broth. Fluorescence units are expressed as the percentage of the starting value for wild-type strain OG1RF. GW4481 is OG1RFemeA (deletion mutant). (B) Each inhibitor was added before the addition of TH broth. Fluorescence units are expressed as the percentage of the starting value for wild-type strain OG1RF. Reserpine was used at 20 µg/ml, verapamil was used at 100 µg/ml, and lansoprazole was used at 100 µg/ml.