Regulation of Multidrug Efflux Systems Involved in Multidrug and Metal Resistance of Salmonella enterica Serovar Typhimurium

Abstract

Multidrug-resistant strains of Salmonella are now encountered frequently, and the rates of multidrug resistance have increased considerably in recent years. Here, we report that the two-component regulatory system BaeSR increases multidrug and metal resistance in Salmonella through the induction of drug efflux systems. Screening of random fragments of genomic DNA for the ability to increase β-lactam resistance in Salmonella enterica led to the isolation of a plasmid containing baeR, which codes for the response regulator of BaeSR. When overexpressed, baeR significantly increased the resistance of the ΔacrB strain to oxacillin, cloxacillin, and nafcillin. baeR overexpression conferred resistance to novobiocin and deoxycholate, as well as to β-lactams in Salmonella. The increase in drug resistance caused by baeR overexpression was completely suppressed by deletion of the multifunctional outer membrane channel gene tolC. TolC interacts with different drug efflux systems. Among the nine drug efflux systems in Salmonella, quantitative real-time PCR analysis showed that BaeR induced the expression of acrD and mdtABC. Double deletion of these two genes completely suppressed BaeR-mediated multidrug resistance, whereas single deletion of either gene did not. The promoter regions of acrD and mdtABC harbor binding sites for the response regulator BaeR, which activates acrD and mdtABC transcription in response to indole, copper, and zinc. In addition to their role in multidrug resistance, we found that BaeSR, AcrD, and MdtABC contribute to copper and zinc resistance in Salmonella. Our results indicate that the BaeSR system increases multidrug and metal resistance in Salmonella by inducing the AcrD and MdtABC drug efflux systems. We found a previously uncharacterized physiological role for the AcrD and MdtABC multidrug efflux systems in metal resistance.

Salmonella enterica is a pathogen that causes a variety of diseases in humans ranging from gastroenteritis to bacteremia and typhoid fever (65). In the 1990s, the prevalence of multidrug-resistant S. enterica increased dramatically in the United Kingdom (67, 68), the United States (20, 23), and Canada (41). Many countries have also documented outbreaks associated with drug-resistant Salmonella in poultry, beef, and swine (10, 12, 21, 34, 69). Emerging resistance to antibiotics in Salmonella has been found in both humans and animals and is thus a potentially serious public health problem (9, 58). High-level fluoroquinolone resistance in S. enterica serovar Typhimurium phage type DT204 has been shown to be largely attributable to multiple target gene mutations and to active efflux by the AcrAB-TolC system (4, 5).

Multidrug efflux pumps cause serious problems in cancer chemotherapy and the treatment of bacterial infections. In bacteria, resistance to drugs is often associated with multidrug transporters that function to decrease cellular drug accumulation (42, 72). Multidrug transporters in bacteria are classified into five families on the basis of sequence similarity: the major facilitator (MFS), resistance-nodulation-cell division (RND), small multidrug resistance, multidrug and toxic compound extrusion, and ATP-binding cassette families (8, 55, 60). In gram-negative bacteria, pumps belonging to the RND family are especially effective in generating resistance (35, 36, 42, 71). The sequencing of bacterial genomes enables us to trace putative drug-resistance genes (56, 57). There are many putative and proven drug transporters in the Salmonella genome. Recent studies have shown that S. enterica serovar Typhimurium has nine functional drug efflux pumps (46). Because many of these multidrug transporters have overlapping substrate spectra (46), it is intriguing that bacteria, with their economically organized genomes, harbor such large sets of multidrug efflux genes.

The key to understanding how bacteria utilize these multiple transporters lies in the regulation of transporter expression. Currently available data show that multidrug transporters are often expressed under precise and elaborate transcriptional control (2, 7, 22, 29). Expression of acrAB, which encodes the major AcrAB efflux pump, is subject to multiple levels of regulation. In Escherichia coli, it is modulated locally by the repressor AcrR (30). At a more global level, it is modulated by stress conditions and by global regulators such as MarA, SoxS, and Rob (62, 63). Olliver et al. (53) reported that mutation in acrR contributes to overexpression of acrAB in S. enterica and increases resistance to multiple drugs. Eaves et al. (14) reported that acrB, acrF, and acrD are coordinately regulated and that their expression influences the expression of the transcriptional activators marA and soxS. Furthermore, it has been reported that integration of an IS1 or an IS10 element in the upstream region of the acrEF operon causes increased expression of acrEF (52). These examples illustrate the complexity and diversity of the mechanisms regulating bacterial multidrug efflux pumps.

However, few data are available on the regulation of S. enterica multidrug transporter genes other than acrAB, acrD, and acrEF (14, 52, 53). In the present study, we report positive regulation of the multidrug efflux pump mdtABC and acrD genes by a response regulator, BaeR, of the two-component BaeSR signal transduction system (38), a phenomenon that leads to increased resistance to β-lactams, novobiocin and deoxycholate. Despite compelling evidence implicating multidrug efflux systems in multidrug resistance, the ongoing debate concerning the “natural” function of these efflux proteins has not abated (25, 40, 59, 70). The various considerations regarding the physiological role of multidrug transporters have been thoroughly reviewed (28, 39, 40). In the present study, we show that BaeSR, AcrD, and MdtABC contribute to copper and zinc resistance in Salmonella in addition to their role in multidrug resistance. The results suggest a previously uncharacterized physiological role for AcrD and MdtABC in metal resistance.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The bacterial strains and plasmids used in the present study are listed in Table 1. The S. enterica serovar Typhimurium strains used are derived from the wild-type strain ATCC 14028s (17). Phage P22-mediated transductions were performed as described previously (13). Bacterial strains were grown at 37°C in Luria-Bertani (LB) broth (64). Cells were rapidly collected for total RNA extraction when the cultures reached an optical density of 0.6 at 600 nm.

TABLE 1.

S. enterica strains and plasmids used in this study

Strain genotype or plasmid	Original name	Characteristicsa	Source or reference
Strain genotypes
Wild type	ATCC 14028s		17
ΔacrB	EG16565	ΔacrB::Kmr	46
ΔtolC	NKS144	ΔtolC::Cmr	This study
ΔtolC acrB	NKS198	ΔtolC ΔacrB::Kmr	This study
ΔacrD acrB	NKS199	ΔacrD ΔacrB::Kmr	This study
ΔmdtABC acrB	NKS200	ΔmdtABC ΔacrB::Kmr	This study
ΔmdtABC acrD acrB	NKS341	ΔmdtABC ΔacrDΔacrB	This study
ΔacrB/vector	NKS463	ΔacrB::Kmr/pHSG398	This study
ΔtolC acrB/vector	NKS464	ΔtolC ΔacrB::Kmr/pHSG398	This study
ΔacrD acrB/vector	NKS465	ΔacrD ΔacrB::Kmr/pHSG398	This study
ΔmdtABC acrB/vector	NKS466	ΔmdtABC ΔacrB::Kmr/pHSG398	This study
ΔmdtABC acrD acrB/vector	NKS469	ΔmdtABC ΔacrD ΔacrB/pHSG398	This study
ΔacrB/pbaeR	NKS460	ΔacrB::Kmr/pHSGbaeR	This study
ΔtolC acrB/pbaeR	NKS476	ΔtolC ΔacrB::Kmr/pHSGbaeR	This study
ΔacrD acrB/pbaeR	NKS477	ΔacrD ΔacrB::Kmr/pHSGbaeR	This study
ΔmdtABC acrB/pbaeR	NKS478	ΔmdtABC ΔacrB::Kmr/pHSGbaeR	This study
ΔmdtABC acrD acrB/pbaeR	NKS479	ΔmdtABC ΔacrD ΔacrB/pHSGbaeR	This study
acrD-lacZY	NKS512	ΔacrD-lacZY+; Kmr	This study
mdtA-lacZY	NKS514	ΔmdtABC-lacZY+; Kmr	This study
tolC-lacZY	NKS501	ΔtolC-lacZY+; Kmr	This study
acrD-lacZY/vector	NES3	ΔacrD-lacZY+; Kmr/pHSG398	This study
acrD-lacZY/pbaeR	NES4	ΔacrD-lacZY+; Kmr/pHSGbaeR	This study
mdtA-lacZY/vector	NES7	ΔmdtABC-lacZY+; Kmr/pHSG398	This study
mdtA-lacZY/pbaeR	NES8	ΔmdtABC-lacZY+; Kmr/pHSGbaeR	This study
tolC-lacZY/vector	NES61	ΔtolC-lacZY+; Kmr/pHSG398	This study
tolC-lacZY/pbaeR	NES62	ΔtolC-lacZY+; Kmr/pHSGbaeR	This study
ΔbaeSR acrD-lacZY	NES16	ΔbaeSR ΔacrD-lacZY+; Kmr	This study
ΔbaeSR mdtA-lacZY	NES18	ΔbaeSR ΔmdtA-lacZY+; Kmr	This study
ΔbaeSR	NKS206	ΔbaeSR	This study
ΔacrD mdtABC	NKS117	ΔacrD::Cmr ΔmdtABC	This study
Plasmids
pKD3		repR6Kg[r]; Apr FRT Cmr FRT	11
pKD4		repR6Kg[r]; Apr FRT Kmr FRT	11
pCP20		reppSC101ts; Apr CmrcI857lPRflp	11
pHSG398		reppMBI; Cmr	Takara Bio
pHSGbaeR		baeR gene cloned into pHSG398; Cmr	This study
pQEbaeRH6		baeR gene cloned into pQE30; Apr	This study

^aCm^r, chloramphenicol resistance; Km^r, kanamycin resistance; Ap^r, ampicillin resistance.

Screening for positive regulators of multidrug resistance.

DNA manipulation generally followed standard practice (64). A genomic library was made by partial Sau3AI digestion of the chromosomal DNA as follows. Chromosomal DNA prepared from an overnight culture of the wild-type strain ATCC 14028s (17) was digested with Sau3AI (1 U μl⁻¹) for 15, 20, 30, or 40 min. The digested DNA was separated on a 0.8% agarose gel, and fragments approximately 0.5 to 3 kb in size were purified and ligated into the BamHI site of vector pHSG398 (Takara Bio, Japan). The ligation products were transformed into E. coli DH5α (64), selecting for chloramphenicol-resistant transformants. Plasmid DNA was prepared from a pool of 16,000 transformants and used to transform the acrB deletion strain EG16565 (46). Cells were plated on LB agar medium (64) containing 15 μg of chloramphenicol/ml and inhibitory concentrations of various drugs.

Plasmid construction.

The baeR gene was amplified from ATCC 14028s genomic DNA by using the primers baeR-F_BamHI and baeR-R_SalI listed in Table 2, which introduced BamHI and SalI sites at the ends of the amplified fragment. This PCR product was cloned between the BamHI and SalI sites of the vector pHSG398 (Takara Bio) to produce the plasmid pHSGbaeR.

TABLE 2.

Primers used in this study

Use and primer	Sequence (5′-3′)
Gene deletion
tolC-P1	TACAAATTGATCAGCGCTAAATACTGCTTCACAACAAGGAGTGTAGGCTGGAGCTGCTTC
tolC-P2	TTTTGCGAAATGATGCGTGATGGATGGATTTTGTCCGTTTCATATGAATATCCTCCTTAG
acrB-P1	GCTCAGCCCAGGTCTTAACTTAAACAGGAGCCGTTAAGACGTGTAGGCTGGAGCTGCTTC
acrB-P2	TAAGCTGTGCTATATCATGTCTTTTGGGTGAGTATTCGTCCATATGAATATCCTCCTTAG
acrD-P1	AAATCTATAACGATATGTAGAAACACGAGGTTTCCCTTTAGTGTAGGCTGGAGCTGCTTC
acrD-P2	TTTTGATCGTGTCGCAATTCTTAATGCCATAGAGGGTAATCATATGAATATCCTCCTTAG
mdtA-P1	AACATTCCGCAAAACGTTTCAGGATGAGAAACTTATACCGGTGTAGGCTGGAGCTGCTTC
mdtC-P2	CAATCCAGAGTTGCCAGCGGGTGTTGTCAGGAAGTTCTGTCATATGAATATCCTCCTTAG
Gene cloning
baeR-F_BamHI	CGCGGATCCTACACGCGCATAACGGTCATC
baeR-R_SalI	CGCGTCGACAAAAATCCATGTATAATTAAG
Chromosomal lacZY fusion
acrD-P2forP2P1	AAATCTATAACGATATGTAGAAACACGAGGTTTCCCTTTACATATGAATATCCTCCTTAG
acrD-P1forP2P1	TTTTTTGTGCCCGACACCTCGTATCAGGCTGGCCGGGATCGTGTAGGCTGGAGCTGCTTC
mdtA-P2forP2P1	AACATTCCGCAAAACGTTTCAGGATGAGAAACTTATACCGCATATGAATATCCTCCTTAG
mdtC-P1forP2P1	CATGACCAAACCGCCGACAATGGTTATCCCCAGCGGCTGCGTGTAGGCTGGAGCTGCTTC
tolC-P2forP2P1	TACAAATTGATCAGCGCTAAATACTGCTTCACAACAAGGACATATGAATATCCTCCTTAG
tolC-P1forP2P1	GGGCACAGGTCTGATAAGCGCAGCGCCAGCGAATAACTTAGTGTAGGCTGGAGCTGCTTC
Quantitative PCR
rrs-F	CCAGCAGCCGCGGTAAT
rrs-R	TTTACGCCCAGTAATTCCGATT
macA-F	CGCGCCAGCAGCAGTTA
macA-R	CGCCGCGGTATCCAGAT
macB-F	ACAGCAGCAGCGTGTCAGTATT
macB-R	TCGGCTCATCTGCCAGAATC
tolC-F	GCCCGTGCGCAATATGAT
tolC-R	CCGCGTTATCCAGGTTGTTG
acrA-F	AAAACGGCAAAGCGAAGGT
acrA-R	GTACCGGACTGCGGGAATT
acrD-F	CGTTATTAAAACCGCTGCACAA
acrD-R	TACGGTTAAACCAGCCGAAAA
acrE-F	AAGCGGCTGCGGCTATC
acrE-R	TTGTGCCGACCAGTGGAA
acrF-F	GCTCTGTCGTCCATCTCAAAGA
acrF-R	CGCGCTACAACGTTATAGTTTTCA
mdtA-F	GAATGCGCGTCGTGATCTG
mdtA-R	TCCAGTTCCTGACGGGAAAC
mdtB-F	GATTAACGCAGCCACCAATTTAT
mdtB-R	GCCGGATTGACTTTGCTGTAA
mdtC-F	GAGGTAGAAGAGACACTGGCTATCTCT
mdtC-R	CGGAGCGCAGGAATAAAAAC
mdtD-F	TCGCTGGATACCACCATCGT
mdtD-R	TTCCCCCAGGCTTTTCG
mdsA-F	TGCTAAAGCCCTTAGCCGTACA
mdsA-R	GCGCGGCCAGAAAACC
mdsB-F	CCCGGTCTGTCGGTTAACG
mdsB-R	CGCTCGTCAAAGGGTTTCAG
mdsC-F	TCGTAACGCGCTGGAATTG
mdsC-R	TTTAGTCGACGCGACAGTTCA
emrA-F	GCGCAAAGCGACCTTAACC
emrA-R	CTTCCCGGCCAATAAGATTG
emrB-F	TCTGGTCAATGACCGTCATTG
emrB-R	TGTAGCCCCCCAGAATCG
mdfA-F	TTTGAGGAGGCGGTGTGTATAA
mdfA-R	AGCGGCGCGATTAACG
mdtK-F	TTTGTTTGGTCACGGCTTACTG
mdtK-R	GCCGGAGCCATTGAGTTG

Construction of gene deletion mutants.

Gene deletion was performed according to the method of Datsenko and Wanner, with recombination between short homologous DNA regions catalyzed by phage λ Red recombinase (11). A curable expression plasmid encoding Red recombinase (pKD46) was introduced into strain ATCC 14028s. The chloramphenicol resistance gene cat or the kanamycin resistance gene kan, flanked by Flp recognition target sites, was amplified by PCR using the primers listed in Table 2. Plasmid pKD3 or pKD4 was used as a template. This PCR product was used to transform the ATCC 14028s strain expressing Red recombinase, and recombinant clones were isolated as chloramphenicol- or kanamycin-resistant colonies. The pKD46 vector was eliminated by incubation at a nonpermissive temperature of 37°C, as confirmed by the loss of resistance to ampicillin. The chromosomal structure of the mutated loci was verified both by PCR as described previously (11) and by Southern hybridization with probes specific for (i) the antibiotic resistance genes used during the construction of the chromosomal deletions and (ii) sequences flanking the inactivated loci. The deletions were then transferred to the wild-type ATCC 14028s strain by P22 transduction. The cat and kan genes were eliminated by using plasmid pCP20 as described previously (11).

Construction of targeted single copy lac fusion strains.

A chromosomal lacZY fusion strain was constructed as described previously (16). A cat or kan cassette was amplified from pKD3 or pKD4 plasmid by using primers listed in Table 2, and drug transporter genes were disrupted by the one-step gene inactivation method (11). After removing the cat or kan cassette using plasmid pCP20, the lacZY transcriptional fusion plasmid pCE37 (16) was integrated into the FLP target sequence downstream of the promoter of the drug efflux genes by FLP-mediated recombination. Thus, the drug efflux genes were deleted in the resulting strains. To investigate the expression of drug transporter genes, the constructed strains listed in Table 1 were streaked onto L-agar plates containing 80 μg of X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside)/ml.

β-Galactosidase assay.

Single colonies of each bacterial strain to be assayed were inoculated into 2 ml of LB broth containing the appropriate antibiotics. After growth overnight at 37°C, the cultures were diluted 1:50 in the same medium and grown at 37°C to late log phase. To assay the effects of indole and metals on gene expression, 2 mM indole, 2 mM CuSO₄, or 1 mM ZnSO₄ was added to secondary cultures. β-Galactosidase was assayed by using ONPG (o-nitrophenyl-β-d-galactopyranoside) as described previously (33). All assays were performed in triplicate.

Determination of the MIC for toxic compounds.

The antibacterial activities of different agents were determined on L agar (1% tryptone, 0.5% yeast extract, 0.5% NaCl) plates containing the compounds (Sigma) indicated in Table 3 at various concentrations. Agar plates were made by the twofold agar dilution technique as described previously (51). To determine the MICs, bacteria were grown in LB broth at 37°C overnight, diluted into the same medium, and then tested at a final inoculum size of 10⁴ CFU μl⁻¹ by using a multipoint inoculator (Sakuma Seisakusyo, Tokyo, Japan) after incubation at 37°C for 20 h. The MIC was the lowest concentration of compound that inhibited cell growth.

TABLE 3.

Susceptibility of Salmonella strains to β-lactams and toxic compounds

Strain genotype	MIC (μg/ml)a
	OXA	MCIPC	NAF	NOV	DOC
Wild type	512	512	>512	256	>40,000
ΔacrB	2	4	8	2	20,000
ΔacrB/vector	2	4	8	2	20,000
ΔacrB/pbaeR	16	32	32	32	>40,000
ΔtolC acrB	0.5	0.5	1	0.5	156
ΔtolC acrB/vector	0.5	0.5	1	0.5	156
ΔtolC acrB/pbaeR	0.5	0.5	1	0.5	156
ΔacrD acrB	2	4	8	2	20,000
ΔacrD acrB/vector	2	4	8	2	20,000
ΔacrD acrB/pbaeR	8	8	16	16	40,000
ΔmdtABC acrB	2	4	8	2	20,000
ΔmdtABC acrB/vector	2	4	8	2	20,000
ΔmdtABC acrB/pbaeR	8	16	32	8	40,000
ΔmdtABC acrD acrB	2	4	8	2	20,000
ΔmdtABC acrD acrB/vector	2	4	8	2	20,000
ΔmdtABC acrD acrB/pbaeR	2	4	8	2	20,000

^aOXA, oxacillin; MCIPC, cloxacillin; NAF, nafcillin; NOV, novobiocin; DOC, deoxycholate. Values in boldface are larger than those of a corresponding parental strain harboring the pHSG398 vector. MIC determinations were repeated at least three times.

RNA extraction.

Total RNA was isolated from bacterial cultures by using an RNeasy Protect Bacteria minikit (QIAGEN) and RNase-free DNase (QIAGEN) as described previously (44). The absence of genomic DNA from DNase-treated RNA samples was confirmed by both nondenaturing agarose electrophoresis gels and PCR with primers known to target genomic DNA. The RNA concentration was determined spectrophotometrically (64).

Determination of specific transcript levels by quantitative real-time PCR after reverse transcription.

Bulk cDNA samples were synthesized from total Salmonella RNA by using TaqMan reverse transcription reagents (PE Applied Biosystems) and random hexamers as described previously (45, 50). The specific primer pairs listed in Table 2 were designed by using ABI Prism primer express software (PE Applied Biosystems). rrs of the 16S rRNA gene was chosen as the normalizing gene (46). Real-time PCR was performed with each specific primer pair by using SYBR green PCR Master Mix (PE Applied Biosystems). The reactions were run on an ABI Prism 7000 sequence detection system (PE Applied Biosystems); the fluorescence signal due to SYBR green intercalation was monitored to quantify the double-stranded DNA product formed in each PCR cycle.

Purification of the BaeR protein.

Histidine-tagged BaeR protein was expressed in E. coli M15[pREP4] (QIAGEN) containing plasmid pQEbaeRH6, and the protein was purified on Ni²⁺-nitrilotriacetic acid Superflow resin (QIAGEN) as follows. Cells were grown in 80 ml of LB medium containing 100 μg of ampicillin/ml and 25 μg of kanamycin/ml to an optical density at 600 nm of 0.6, at which time IPTG (isopropyl-β-d-thiogalactopyranoside) was added to a final concentration of 1 mM, and incubation was continued for 5 h. Cells were harvested by centrifugation, resuspended in 1 ml of the binding buffer (5 mM imidazole, 0.5 M NaCl, 20 mM Tris-Cl [pH 7.9]) containing 1 mg of lysozyme/ml, left on ice for 30 min, and disrupted by sonication. The lysate was clarified by centrifugation at 17,000 × g for 30 min at 4°C, and 2.5 ml of a 50% slurry of Superflow resin in the binding buffer was added. The mixture was left for 1 h at 4°C with slow stirring and then poured into a column, which was washed with 7 ml of binding buffer, followed by 10 ml of wash buffer (60 mM imidazole, 0.5 M NaCl, 20 mM Tris-Cl [pH 7.9]). The histidine-tagged BaeR protein was eluted from the column with 8 ml of 250 mM imidazole in 0.5 M NaCl-20 mM Tris-Cl (pH 7.9).

Electrophoretic mobility shift assay.

DNA fragments used for the electrophoretic mobility shift assay were amplified by the PCR using S. enterica serovar Typhimurium chromosomal DNA as a template. Fragments were end labeled with T4 polynucleotide kinase and [γ-³²P]ATP and purified by using a Qiaquick nucleotide removal kit (QIAGEN). Approximately 25 pmol of labeled DNA and 0, 50, 100, or 150 pmol of BaeR protein in a 100-μl volume was incubated at room temperature for 15 min. The binding buffer used for protein-DNA incubations was composed of 20 mM Tris (pH 7.4), 5 mM MgCl₂, 50 mM KCl, 50 μg of bovine serum albumin/ml, 2.5 μg of salmon sperm DNA/ml, and 10% glycerol. Samples (20 μl) were run on a 4% nondenaturing Tris-glycine polyacrylamide gel at 2°C. After electrophoresis the gel was dried and autoradiographed.

Cell growth assay.

Single colonies of each bacterial strain to be assayed were inoculated into 2 ml of LB. After growth overnight at 37°C, the cultures were diluted to an A₆₀₀ of 0.1 in the same medium. Spectrophotometric readings were then taken with a Versa Max microplate reader (Molecular Devices). To assay the effects of metals on cell growth, 600 μM CuSO₄ or 600 μM ZnSO₄ was added to the secondary cultures.

RESULTS

Overexpression of BaeR increases resistance to oxacillin.

Expression of multidrug efflux genes is often regulated in a complex manner, as described in the introduction. We therefore screened the Salmonella genomic library for genes that increased multidrug resistance levels in this organism. Our preliminary effort with an acrB⁺ strain, however, failed to isolate genes conferring multidrug resistance.

We therefore used a host strain lacking a functional acrB gene in the screening in order to identify regulatory elements involved in the expression of other multidrug resistance systems. The library was made from the chromosomal DNA of strain ATCC 14028s. Random fragments of chromosomal DNA were cloned into the BamHI site of vector pHSG398 as described in Materials and Methods. The recombinant plasmids were transformed into the ΔacrB strain EG16565, and the transformants were plated on LB agar containing 15 μg of chloramphenicol/ml and inhibitory concentrations of various drugs. In one experiment, the medium contained 8 μg of oxacillin, which had an MIC of 2 μg/ml against EG16565 (ΔacrB). When one of the transformant colonies that grew on this medium was purified and reexamined, we indeed found an eightfold increase in oxacillin MIC against the transformant (data not shown). Introduction of the plasmid isolated from this strain into fresh ΔacrB cells resulted in the same oxacillin resistance phenotype: the MIC was increased eightfold over the recipient strain (data not shown).

Sequencing of the plasmid revealed an insertion containing the complete coding sequence of baeR and a partial sequence of baeS. BaeR is the response regulator of a putative two-component system, which also includes the sensor kinase BaeS, encoded in a gene located upstream of baeR (38). It seemed likely that in cells carrying this plasmid, overexpressed BaeR was causing the transcriptional activation of genes involved in oxacillin resistance.

To test whether overexpression of baeR confers oxacillin resistance on the Salmonella ΔacrB strain, full-length wild-type baeR was amplified by PCR and cloned in the pHSG398 vector to obtain pbaeR (Table 1), which was then introduced into the ΔacrB strain EG16565. Oxacillin MICs for cells harboring pbaeR were eight times higher (16 versus 2 μg/ml) than for cells harboring the pHSG398 vector (Table 3), suggesting that the BaeR regulator produced by this plasmid conferred oxacillin resistance on Salmonella. Further experiments in the present study were therefore carried out with pbaeR. When pbaeR was used to transform ATCC 14028s cells, which express the AcrB pump (46), no effect on oxacillin resistance was observed (data not shown), presumably because the intrinsic multidrug efflux pump AcrB masks the effect of baeR overexpression and ATCC 14028s was already highly resistant to this drug.

Overexpression of BaeR increases resistance to β-lactams, novobiocin, and deoxycholate.

Our results had shown that overexpression of baeR increased Salmonella resistance to oxacillin, a β-lactam antibiotic. We therefore investigated the effect of baeR overexpression on the susceptibility of Salmonella to other β-lactams. pbaeR was also found to increase the resistance of ΔacrB cells to cloxacillin and nafcillin (Table 3). Various other drugs were tested, including common substrates of multidrug transporters, and we found that pbaeR increased the resistance of the ΔacrB strain to novobiocin and deoxycholate (Table 3). These results indicate that the overproduced BaeR regulator induces multidrug resistance in Salmonella. The sensor kinase of a typical two-component system monitors some environmental conditions and accordingly modulates the phosphorylation state of the response regulator. Overexpression of baeR is thought to mimic the physiological phosphorylation response as previously reported (1, 27). There was no change in the level of resistance to various other toxic compounds tested such as erythromycin, doxorubicin, acriflavin, benzalkonium chloride, carbonyl cyanide m-chlorophenylhydrazone, cefamandole, crystal violet, enoxacin, ethidium bromide, methylene blue, nalidixic acid, norfloxacin, ofloxacin, rhodamine 6G, tetracycline, and tetraphenylphosphonium bromide.

Effect of tolC deletion on the multidrug resistance modulated by the BaeR regulator.

The results described above indicate that the expression of a multidrug exporter(s) may be induced by the overexpression of baeR. In a previous study, it was revealed that at least nine intrinsic drug efflux transporters are encoded in the S. enterica chromosome (46). Among these, RND-family transporters play major roles in both intrinsic and elevated resistance of gram-negative bacteria to a wide range of noxious compounds, including β-lactams (31, 42, 43, 47, 48). RND transporters need two other proteins for their function: a membrane fusion protein and an outer membrane channel. It has been reported that many drug transporter systems in E. coli need TolC to function (15, 18, 47, 49). TolC is also present in S. enterica, and deletion of it increases susceptibility to many antimicrobial agents and chemical compounds (46).

In order to determine whether BaeR-mediated multidrug resistance is attributable to the TolC-dependent drug exporter(s), we investigated the effect of tolC deletion on the drug resistance of the baeR-overexpressing cells. Deletion of tolC from the strain ΔacrB increased susceptibility to oxacillin, cloxacillin, nafcillin, novobiocin, and deoxycholate (Table 3), which is in good agreement with recent reports (6, 46). The tolC deletion completely inhibited BaeR-mediated multidrug resistance (Table 3). This result indicates that BaeR-mediated multidrug resistance is attributable to increased expression of a TolC-dependent drug exporter(s).

Determination of the amounts of drug exporter transcripts by quantitative real-time reverse transcription-PCR (qRT-PCR).

In order to determine which drug exporters show increased expression when baeR is overexpressed, we used qRT-PCR to investigate changes in the amounts of drug exporter gene mRNAs dependent on baeR overexpression. Total RNAs were isolated from exponential-phase cultures of ΔacrB/vector and ΔacrB/pbaeR, and cDNA samples were synthesized by using TaqMan reverse transcription reagents (PE Applied Biosytsems) with random hexamers as primers. Real-time PCR of the cDNAs was performed with each specific primer pair using the SYBR green PCR Master Mix (PE Applied Biosystems). The expression levels of drug exporter genes (except for acrB) and tolC in ΔacrB/pbaeR were compared to those in ΔacrB/vector. The results are shown in Table 4. The expression levels of mdtABCD, acrD, and tolC were significantly increased (more than fivefold compared to basal levels) by baeR overexpression: 150-, 64-, 25-, 11-, 15-, and 5.4-fold increases were observed for mdtA, mdtB, mdtC, mdtD, acrD, and tolC, respectively. Overexpression of baeR did not increase the expression levels of other drug exporter genes (Table 4).

TABLE 4.

Fold induction of specific transcripts attributed to baeR overexpression as determined by qRT-PCR^a

Gene	Fold increase (mean ± SD)
tolC	5.4 ± 0.34
acrA	1.3 ± 0.15
acrE	0.82 ± 0.12
acrF	0.74 ± 0.091
mdtA	150 ± 2.3
mdtB	64 ± 2
mdtC	25 ± 0.23
mdtD	11 ± 0.88
mdsA	0.71 ± 0.053
mdsB	0.73 ± 0.041
mdsC	0.68 ± 0.11
emrA	1.3 ± 0.036
emrB	1.1 ± 0.13
mdfA	1.2 ± 0.028
mdtK	0.89 ± 0.039
macA	1.3 ± 0.095
macB	0.96 ± 0.12
acrD	15 ± 1.2

^aThe amount of transcript was determined by qRT-PCR as described in Materials and Methods. The fold change ratio was calculated by dividing the expression level of the gene in the ΔacrB/pbaeR strain by that in the ΔacrB/vector strain. Experiments were performed in triplicate, and the data are represented as mean values ± the standard deviation. The values in boldface indicate increases of more than twofold.

BaeR controls the expression of mdtABC, acrD, and tolC.

To investigate the effect of baeR overexpression on the expression levels of the mdt operon, acrD and tolC, we constructed strains in which E. coli lacZY replaced the chromosomal copies of the drug efflux system genes shown in Table 1. We streaked strains harboring the vector or the pbaeR plasmid onto X-Gal LB agar plates (Fig. 1A). The mdtA-lacZY and acrD-lacZY strains overexpressing baeR were blue on plates containing 80 μg of X-Gal/ml, whereas the strains harboring the pHSG398 vector were white (Fig. 1A). Both the tolC-lacZY strains harboring the pHSG398 vector and pbaeR plasmid were blue (Fig. 1A) because the tolC gene is constitutively expressed in complex laboratory media (46). Measurements of β-galactosidase activity showed that baeR overexpression conferred higher β-galactosidase activities on the mdtA-lacZY, acrD-lacZY, and tolC-lacZY strains than the pHSG398vector (Fig. 1B). These results indicate that BaeR protein induces the expression of the MdtABC and AcrD drug efflux systems and the TolC outer membrane channel.

FIG. 1.

Effect of the BaeR regulator on the expression levels of Salmonella drug transporter genes. (A) The mdtA-lacZY/vector (NES7), mdtA-lacZY/pbaeR (NES8), acrD-lacZY/vector (NES3), acrD-lacZY/pbaeR (NES4), tolC-lacZY/vector (NES61), and tolC-lacZY/pbaeR (NES62) strains were streaked onto LB agar plates containing X-Gal (80 μg/ml). The plates were incubated for 24 h at 37°C and photographed the following day. (B) Expression levels of the mdtA-lacZY, acrD-lacZY, and tolC-lacZY strains harboring the vector or the pbaeR plasmid were determined as described in Materials and Methods. The data correspond to mean values of three independent experiments. Error bars correspond to the standard deviation.

Effects of deletion of drug exporter genes on BaeR-mediated multidrug resistance.

In order to determine whether multidrug resistance mediated by baeR overexpression is due to increased expression of mdtABC and acrD, we investigated the effects of deleting these genes on drug resistance levels in ΔacrB/vector and ΔacrB/pbaeR (Table 3). When mdtABC and acrD were deleted separately or simultaneously from the ΔacrB strain, there was no change in drug resistance in the resulting strains. In the ΔacrD acrB strain, overexpression of baeR conferred resistance to oxacillin, cloxacillin, nafcillin, novobiocin, and deoxycholate, but the resistance levels conferred by BaeR were lower than those in the ΔacrB/pbaeR strain, suggesting that the AcrD drug efflux system makes some contribution to the multidrug resistance modulated by the BaeR regulator. In the ΔmdtABC acrB strain, baeR overexpression conferred resistance to oxacillin, cloxacillin, nafcillin, novobiocin, and deoxycholate, but the levels of resistance to oxacillin, cloxacillin, and novobiocin were lower than those in the ΔacrB/pbaeR strain, suggesting that the MdtABC drug efflux system contributes to the multidrug resistance modulated by the BaeR regulator. On the other hand, overexpression of baeR conferred no drug resistance on the ΔmdtABC acrD acrB strain (Table 3). Together, these data indicate that the multidrug resistance conferred by the BaeR regulator is due to increased expression of both the mdtABC and acrD drug efflux system genes.

BaeR-binding site sequence motifs.

Our previous analysis using an E. coli DNA microarray and the motif-finding program revealed the BaeR-binding site sequence motif in E. coli. This motif has an 18-bp consensus sequence: 5′-TTTTTCTCCATDATTGGC-3′ (where D is G, A, or T) (44). To investigate whether the BaeR-binding motif exists in the upstream region of the mdt operon and acrD in S. enterica, we used the program Gene Promoter Scan (GPS) (73, 74). We identified sequences resembling the BaeR box in the region upstream of the mdt operon and acrD (Fig. 2A). A search of the regions upstream (100 bp) of tolC did not reveal the consensus sequence resembling the BaeR box. To examine the ability of the BaeR protein to bind the promoter regions of mdt operon and acrD, we first constructed a derivative of the baeR gene encoding a protein with six histidine residues at its amino terminus, and we showed that a plasmid encoding this baeR derivative (pQEbaeRH6 shown in Table 1) could activate the transcription of the mdtABC and acrD genes (data not shown). An electrophoretic mobility shift assay with the purified BaeR-His₆ protein showed binding to the regions upstream of the mdt operon and acrD (Fig. 2B). As a control, an assay showed no interaction between BaeR and the fragment including the acrA promoter (Fig. 2B). These results indicate that the mdtABC and acrD drug efflux genes are directly regulated by BaeR protein (Fig. 2C).

FIG. 2.

BaeR-binding site sequence motifs and model for the control of drug efflux system genes by the two-component BaeS/BaeR regulatory system. (A) Consensus sequence in the upstream regions of BaeR-regulated genes. Consensus sequences were found upstream of mdtA and acrD using the GPS method. The numbering is relative to the start codon of the genes. (B) The BaeR protein binds to the promoter region of the mdtA and acrD genes. The PCR fragment includes the region upstream of mdtA, acrD, or acrA. The concentration of BaeR-H6 added to each reaction is indicated at the top of each lane. (C) The BaeS/BaeR system activates expression of the mdt operon, acrD, and tolC. The induced efflux systems confer multidrug resistance and metal resistance on Salmonella. The baeSR genes are located immediately downstream of the mdtABCD genes and make an operon and positive feedback loop for the mdt operon.

Role of BaeSR in the induction of the mdtABC and acrD drug efflux genes.

Because expression of mdtABC and acrD in E. coli is responsive to indole (44), we investigated the possibility that indole may induce expression of the acrD and mdtABC drug efflux systems in Salmonella. We also tested whether the two-component BaeSR signal transduction pathway was involved in this induction. First, the induction of mdtA and acrD expression in response to indole was examined. β-Galactosidase assays showed that indole leads to induction of mdtA and acrD expression in a wild-type strain (Fig. 3). When the BaeSR pathway is ablated, indole induction of mdtA and acrD is still observed, although at markedly reduced levels (Fig. 3A and B).

FIG. 3.

The BaeSR signal transduction pathway activates mdtA and acrD expression in response to indole, copper, and zinc. (A and B) β-Galactosidase levels were assayed in strains with the mdtA-lacZY transcriptional fusion in the wild-type (mdtA-lac) and ΔbaeSR genetic background (ΔbaeSR mdtA-lac) (B) or with the acrD-lacZY transcriptional fusion in wild-type (acrD-lac) and ΔbaeSR genetic background (ΔbaeSR acrD-lac) in either the absence or the presence of 2 mM indole, 2 mM copper, or 1 mM zinc (B). Asterisks indicate statistically significant differences (*, P < 0.05; **, P < 0.01) in the paired Student t test.

Because expression of mdtABC and acrD in E. coli is also responsive to zinc (26), we sought to determine whether metals induce mdtABC and acrD. We found that copper and zinc induce the expression of mdtA and acrD in a wild-type background (Fig. 3). Removal of the BaeSR signal transduction pathway eliminated copper and zinc induction of mdtA and acrD expression (Fig. 3A and B). These data suggest that the BaeSR system contributes to the induction of mdtA and acrD expression in response to copper and zinc.

Role of MdtABC, AcrD, and BaeSR in metal resistance.

Because our results above indicate that expression of mdtABC and acrD is induced by metals (Fig. 3), and because there was no change in drug resistance in strains when mdtABC and acrD were deleted (Table 3), we investigated the possibility that the MdtABC and AcrD efflux systems may have physiological roles in metal homeostasis in addition to multidrug resistance. To examine the physiological relevance of MdtABC and AcrD in metal homeostasis, we characterized a strain that is completely devoid of the chromosomal mdtABC and acrD genes. We tested the sensitivity of ΔacrD mdtABC to copper and zinc, since our results showed that these metals are signals that activate both efflux systems. Wild-type and the ΔacrD mdtABC double mutant were grown in LB broth in the absence or the presence of copper (600 μM) or zinc (600 μM), and growth was measured (Fig. 4). In the absence of metals, both strains grew at approximately the same rate. In contrast, in the presence of copper or zinc, the ΔacrD mdtABC double mutant grew more poorly than the wild-type strain (Fig. 4). These data suggest that AcrD and MdtABC contribute to metal resistance in Salmonella. TolC was also required for normal cell growth in the presence of cooper or zinc (Fig. 4). Further, we performed identical growth experiments with the ΔbaeSR mutant and found that it grew more poorly than the wild-type strain in the presence of copper or zinc (Fig. 4) even though its growth in LB medium was normal. These observations indicate that BaeSR must regulate the expression of the mdtABC and acrD efflux genes, which contribute to copper and zinc adaptation. Together, the results indicate that induction of the MdtABC and AcrD efflux systems by the BaeSR signal transduction system contribute to metal resistance in Salmonella.

FIG. 4.

ΔacrD mdtABC, ΔbaeSR, and ΔtolC mutants are more sensitive to copper and zinc. Growth of wild-type, ΔacrD mdtABC, ΔbaeSR, and ΔtolC strains was monitored by measuring the A₆₀₀ of cultures in either the absence or the presence of 600 μM copper (A) or 600 μM zinc (B). Copper and zinc inhibit the growth of ΔacrD mdtABC, ΔbaeSR, and ΔtolC more than the wild type. The strains used were ATCC 14028s, NKS117, NKS206, and NKS144.

DISCUSSION

In this study, we performed a genome-wide search for a regulator of multidrug resistance in S. enterica serovar Typhimurium by random shotgun cloning and discovered BaeR, which upregulates the mdtABCD locus and acrD, thereby increasing resistance to β-lactams, novobiocin, and deoxycholate. We initially found by random shotgun cloning that the plasmid carrying baeR conferred oxacillin resistance on the ΔacrB strain. However, when this plasmid was used to transform ATCC 14028s cells, which express the AcrB pump (46), no effect on oxacillin resistance was observed (data not shown), presumably because ATCC 14028s was already highly resistant to this drug; the MIC against oxacillin in this strain was 512 μg/ml (Table 3). Then we investigated the susceptibility of the ΔacrB strain overexpressing baeR to various drugs, including common substrates of multidrug transporters, and found that BaeR stimulates Salmonella resistance to oxacillin, cloxacillin, nafcillin, novobiocin, and deoxycholate (Table 3).

BaeR exhibits sequence similarity to the response regulators and is likely to constitute a two-component regulatory system with a sensor kinase, BaeS. The genes baeS and baeR were first identified during random screening for the two-component signal transduction genes in E. coli on the basis of their ability to suppress mutational lesions of the sensor kinase genes envZ and phoR/creC phenotypically (38). BaeR and BaeS exhibit an in vitro phosphotransfer reaction in the presence of ATP (38). baeS and baeR are found immediately downstream of mdtABCD (Fig. 2C) (3, 37), and the start codon of each successive member of this cluster of six genes is immediately adjacent to or overlaps the stop codon of the preceding one (32). Thus, these genes probably form an operon. In the present study, we found that overexpression of BaeR strongly stimulates the expression of mdtABCD and acrD in Salmonella. Sequences resembling the BaeR box were found in the region upstream of the mdt-bae operon and acrD (Fig. 2A), and an electrophoretic mobility shift assay showed that the BaeR protein binds to these regions (Fig. 2B). BaeR thus regulates the transcription of the mdt-bae operon and acrD directly. It should regulate the expression of baeS and baeR positively because these genes form an operon with mdtABCD. This indicates that the BaeS/BaeR two-component system generates a positive feedback loop by regulating the mdt-bae operon (Fig. 2C).

Recently, it has been reported that Salmonella BaeR controls the expression of the outer membrane proteins OmpW and STM3031, which might be associated with ceftriaxone resistance (24). In the present study, we found that BaeR activates expression of the multifunctional outer membrane channel gene tolC in S. enterica. Search of the regions upstream (100 bp) of tolC did not reveal the consensus sequence resembling the BaeR box, and an electrophoretic mobility shift assay did not show significant BaeR binding to this region (data not shown), suggesting that BaeR may indirectly activates the expression of tolC. In E. coli, TolC is required for the function of the MdtABC and AcrD multidrug efflux systems (47). TolC induction by the two-component BaeS/BaeR system may be required to form complexes with BaeR-induced AcrD or MdtABC. Indeed, tolC is required for the multidrug resistance stimulated by baeR overexpression (Table 3). Both AcrD and MdtABC are required for the Salmonella multidrug resistance induced by BaeR (Table 3), indicating that this resistance is due to the increased expression of acrD and mdtABC. BaeR also activates the expression of mdtD, which encodes a putative MFS-type transporter, but it has been reported that mdtD is not related to multidrug resistance (3, 37, 48).

Raffa and Raivio showed that the BaeSR system controls an envelope stress response in E. coli that induces expression of a distinct set of adaptive genes including spy, which is a member of the Cpx stress response regulon (61). Genome-wide transcriptional analysis also revealed that spy is a member of the BaeSR regulon, like mdtABC and acrD (44). Indole, a toxic compound expected to disrupt the bacterial envelope, induces the expression of spy (61), mdtABC, and acrD (44). We found that indole also induces expression of MdtABC and AcrD in Salmonella (Fig. 3), and the two-component BaeSR signal transduction pathway is involved in this induction. These findings therefore have potentially important implications for the influence of envelope stress on bacterial resistance to antibiotics.

In addition to the roles of MdtABC, AcrD, and BaeSR in multidrug resistance, we found that they contribute to copper and zinc resistance in Salmonella. Both copper and zinc are essential for organisms but can be toxic, and microorganisms express diverse resistance mechanisms. We showed that the expression of mdtABC and acrD is induced by copper or zinc, and BaeSR is involved in this induction (Fig. 3). Some bacterial efflux pumps export not only antibiotics and other substances, such as dyes and detergents (35, 59), but also host-derived antimicrobial agents (66). This finding has led to the suggestion that the physiological role of these systems is evasion of such naturally produced molecules, thereby allowing the bacterium to survive in its ecological niche. In the present study, we found that the MdtABC, AcrD drug efflux systems and the two-component BaeSR signal transduction system contribute to copper and zinc resistance in Salmonella (Fig. 4). The baeSR mutant was more sensitive to zinc than the acrD mdtABC double mutant (Fig. 4B), indicating that BaeSR may regulate an additional zinc resistance determinant. Our results suggests that the MdtABC and AcrD efflux systems may have physiological roles in metal homeostasis in addition to multidrug resistance, because there was a change in metal resistance in the strains when mdtABC and acrD were deleted (Fig. 4) but there was no change in drug resistance (Table 3), and because the contributions of MdtABC, AcrD, and BaeSR to metal resistance were observed in the wild-type background (Fig. 4), but their contributions to drug resistance were observed only in the ΔacrB background. The MdtABC and AcrD systems may be related to bacterial metal homeostasis by transporting metals directly. This is reminiscent of the copper and silver resistance mechanism by cation efflux of the CusABC system belonging to the RND protein superfamily (19, 54). Our results suggest a previously uncharacterized physiological role for AcrD and MdtABC in metal resistance.

Further investigation of the regulation of multidrug efflux systems in several natural environments, such as inside hosts or soil, is needed in order to understand the biological significance of their regulatory networks. Such investigation may provide further insights into the role of multidrug efflux systems in the physiology of the cell.