Transcriptional regulation of drug efflux genes by EvgAS, a two-component system in Escherichia coli

Yoko Eguchi1, Taku Oshima2, Hirotada Mori2, Rikizo Aono3, Kaneyoshi Yamamoto1, Akira Ishihama4 and Ryutaro Utsumi1

1 Department of Bioscience and Biotechnology, Graduate School of Agriculture of Kinki University, 3327-204, Nakamachi, Nara 631-8505, Japan
2 Research and Education Center for Genetic Information, Nara Institute of Science and Technology, Ikoma 630-0101, Japan
3 Department of Bioengineering, Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuta 4259, Midori-ku, Yokohama 226-0027, Japan
4 Division of Molecular Biology, Nippon Institute for Biological Science, Ome, Tokyo 190-0024, Japan

Correspondence
Ryutaro Utsumi
utsumi{at}nara.kindai.ac.jp


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A constitutively active mutant of histidine kinase sensor EvgS was found to confer multi-drug resistance (MDR) to an acrA-deficient Escherichia coli, indicating the relationship between the two-component system EvgAS and the expression of the MDR system. The observed MDR also depended on an outer-membrane channel, TolC. Microarray and S1 mapping assays indicated that, in the presence of this constitutive mutant EvgS, the level of transcription increased for some MDR genes, including the drug efflux genes emrKY, yhiUV, acrAB, mdfA and tolC. Transcription in vitro of emrK increased by the addition of phosphorylated EvgA. Transcription activation of tolC by the activated EvgS was, however, dependent on both EvgAS and PhoPQ (Mg2+-responsive two-component system), in agreement with the presence of the binding site (PhoP box) for the regulator PhoP in the tolC promoter region. Transcription in vitro of yhiUV also appears to require an as-yet-unidentified additional transcriptional factor besides EvgA. Taken together we propose that the expression of the MDR system is under a complex regulatory network, including the phosphorylated EvgA serving as the master regulator.


Abbreviations: MDR, multi-drug resistance


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The EvgAS two-component signal transduction system in Escherichia coli is highly homologous to the virulence-related BvgAS system of Bordetella pertussis (Utsumi et al., 1994). A BLAST search indicates the presence of homologous systems in Klebsiella pneumoniae, Vibrio cholerae and Pseudomonas aeruginosa. EvgS is a histidine kinase hybrid sensor, which is composed of an N-terminal periplasmic region and a C-terminal cytoplasmic region that is divided into four domains: linker, transmitter, receiver and output (Hpt) (Perraud et al., 1998). The periplasmic domain of EvgS is involved in signal recognition, ultimately transducing the signal into a transcriptional regulation network via a cascade of phosphorylation (Utsumi, 2002; Utsumi et al., 1994). To date, however, the target genes under the control of EvgAS are still poorly characterized. Auto-phosphorylation of the cytoplasmic region of EvgS (without the transmembrane and periplasmic region) has been reported to be inhibited in vitro by an oxidized ubiquinone-0 (Bock & Gross, 2002), as is also the case with the anaerobic sensor ArcB (Georgellis et al., 2001). However, the inhibition of EvgS autophosphorylation by ubiquinone has not been confirmed in in vivo experiments, and the environmental signal(s) sensed by EvgS also remains to be determined.

There are two methods of analysing the function of the EvgAS system: deletion of evgAS and overexpression of the regulator EvgA. Ohshima et al. (2002b) have recently reported the genome-wide transcriptional changes caused by the deletion of each of the two-component systems of E. coli (http://ecoli.aist-nara.ac.jp/xp_analysis/2_components). According to the report, only subtle changes were observed when some of the two-component systems investigated were deleted. The EvgAS system is one such system and its deletion decreased the transcription of as few as three ORFs. From this result, the EvgAS system is considered to be not activated under the conditions employed and deleting such a non-activated system will give little information. Overexpression of the regulator to artificially activate the system is another method for analysing the EvgAS system. However, artifacts arising from the overexpression may occur (Albright et al., 1989; Utsumi et al., 1994). Moreover, it is difficult to interpret the transcriptional changes resulting from the overexpression: did it occur from the increased level of non-phosphorylated EvgA or from EvgA phosphorylated by sources other than its cognate sensor EvgS, such as acetyl phosphate?

By using a spontaneous E. coli mutant (evgS1 mutant; Kato et al., 2000), which produces a constitutively active EvgS with an F577S mutation within the linker region (also a predicted PAS domain; Bock & Gross, 2002), these authors (Kato et al., 2000) first found that transcription of emrKY, a multi-component drug efflux pump of the major facilitator superfamily (MFS), is positively regulated by the EvgAS system. We analysed in this study transcription patterns for some drug resistance genes in this evgS1 mutant to clarify the relationship between the EvgAS system and multi-drug resistance (MDR) expression.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and growth conditions.
The E. coli strains and plasmids used in this study are listed in Table 1. Construction of acrA, phoP and phoQ mutants derived from KMY1 (wild-type) and KMY2001 (evgS1 mutant) was performed by P1 transduction with JA300A (acrA : : cat), WP3022 (phoP : : cat) and WQ3007 (phoQ : : cat) as the donor strains, respectively. Construction of tolC mutants of strains KMA1 and KMA2001 was also performed by P1 transduction with W4573tolC : : Tn10. Strains were grown at 37 °C with aeration in Luria–Bertani (LB) medium (pH 7·5), containing 1 % (w/v) Bacto Tryptone (Difco), 0·5 % (w/v) Bacto Yeast Extract (Difco) and 1 % (w/v) NaCl. When necessary, selective antibiotics were added to the medium as follows: 100 µg ampicillin ml-1, 25 µg chloramphenicol ml-1, 12·5 µg tetracycline ml-1 or 25 µg kanamycin ml-1.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids used in this study

 
Plasmid construction.
Plasmids containing the emrK and yhiU promoter regions were constructed as follows. The 545 bp emrK promoter fragment and 405 bp yhiU promoter fragment were amplified by PCR with genomic DNA from strain W3110 and the primers: 5'-GCTGAACAGATCTTCCGCCTTCAGT-3' and 5'-AATCTGATGCATTATTATCTCTCATTTCTC-3' for emrK, and 5'-GAAACTAAGATCTATGAGCGACATCGTCAC-3' and 5'-TCTTCTATGCATTTTAGTCCCTGAAAATTC-3' for yhiU. These primers contain EcoT22I and BglII sites (underlined) suitable for cloning. The PCR products were digested with restriction enzymes and then ligated into the EcoT22I and BglII sites of pGMI301.

Drug susceptibility tests.
The MICs of drugs were determined on LB agar plates containing serial dilutions of the following compounds: doxorubicin hydrochloride (Wako Pure Chemicals), crystal violet (Tokyo Kasei Kogyo), rhodamine 6G (Sigma–Aldrich), ethidium bromide (Sigma–Aldrich), acriflavine (Tokyo Kasei Kogyo), benzalkonium chloride (Tokyo Kasei Kogyo), SDS (Nacalai Tesque), deoxycholate (Nacalai Tesque), cholate (Nacalai Tesque), norfloxacin (Wako Pure Chemicals) and trimethoprim (Wako Pure Chemicals). Experiments were repeated at least twice. The inoculum E. coli cultures were prepared by diluting overnight cultures to 106 c.f.u. ml-1 and 5 µl of the diluted culture was spotted. Inoculated plates were incubated for 16–20 h at 37 °C. The lowest concentration of a drug that completely inhibited growth was identified as the MIC.

RNA preparation and S1 nuclease mapping.
Total RNA was extracted with hot phenol from a mid-exponential-phase culture (OD600 0·7–0·9) at 60 °C (Aiba, 1983). S1 mapping was carried out as described by Kato et al. (2000). In brief, 32P-end-labelled probe was prepared by PCR using primers listed in Table 2, MC4100 genomic DNA as the template and ExTaq DNA polymerase (Takara). A mixture of 32P-end-labelled probe and 100 µg total RNA was incubated for 10 min at 75 °C, then gradually cooled to 37 °C and incubated overnight for hybridization, followed by S1 nuclease (Takara) digestion for 10 min at 37 °C. Undigested RNA-probe DNA was extracted with phenol, precipitated with ethanol and subjected to electrophoresis on a 6 % (w/v) polyacrylamide sequencing gel.


View this table:
[in this window]
[in a new window]
 
Table 2. PCR primers

 
DNA microarray analysis.
Total RNA from strains KMY1 and KMY2001 was extracted from mid-exponential-phase cultures as described above (S1 mapping) and was further treated with RNase-free DNase I (Takara). Preparation of fluorescent labelled cDNA, microarray hybridization and data analysis were performed as described by Oshima et al. (2002a). In brief, cDNA labelled with Cy3-dUTP (KMY1) and Cy5-dUTP (KMY2001) was synthesized from each total RNA by random priming. Labelled cDNA probes were purified and hybridized to a custom glass slide microarray (Takara), which was spotted in duplicates with 4097 PCR products corresponding to full-length E. coli ORFs and the human {beta}-actin gene as a negative control. The slides were scanned for fluorescence intensity using a GMS418 array scanner (Genetic Microsystems) and recorded as 16 bit image files. The signal density of each spot in the array was quantified using Imagene software (BioDiscovery). A normalized relative Cy5/Cy3 ratio of 2 and above was considered as a significant increase in expression, and 0·5 and below as a significant decrease in expression.

Purification of EvgA.
A plasmid expressing EvgA, His-tagged at its amino terminus (pEvgA), was transformed into E. coli M15(pREP4). For expression of EvgA, the transformant was grown at 37 °C with aeration in 200 ml 2x YT broth, containing 1·6 % (w/v) Bacto Tryptone (Difco), 1 % (w/v) Bacto Yeast Extract (Difco) and 0·5 % (w/v) NaCl, to a cell density of OD600=0·8, followed by addition of IPTG at a final concentration of 1 mM. After 4 h incubation at 37 °C with aeration, cells were harvested by centrifugation, washed with sonication buffer (25 mM NaH2PO4, 25 mM Na2HPO4, 1 M NaCl, 5 % (w/v) glycerol) and stored at -80 °C until use. For protein purification, frozen cells were resuspended in 4 ml sonication buffer with 1 mM PMSF, lysed by sonication and centrifuged at 5000 r.p.m. for 30 min at 4 °C. The supernatant was mixed with 2 ml 50 % Ni(II)-NTA agarose suspension (Qiagen) and loaded onto a column. EvgA was eluted with sonication buffer containing 100 mM imidazole. EvgA of this fraction was concentrated and the buffer was exchanged for storage buffer (50 mM Tris/HCl, pH 7·5, 50 mM KCl, 10 mM MgCl2, 1 mM DTT, 50 %, w/v, glycerol) with centrifugal filtration (Ultrafree-MC; Millipore). The protein concentration was determined using the Bradford method (Protein Assay Kit; Bio-Rad) and the purity was checked by SDS-PAGE.

In vitro transcription assays.
pGMIemrK and pGMIyhiU were digested with EcoT22I and BglII to prepare linear DNAs, including the emrK and yhiU promoter regions, respectively, as the template of in vitro transcription. RNA polymerase containing {sigma}D was used for the assay, which was performed under standard reaction conditions (Kajitani & Ishihama, 1983). In brief, 0·1 pmol of the template was incubated with 0, 25, 50 or 100 pmol His-tagged EvgA for 10 min at 37 °C in the presence or absence of 10 mM acetylphosphate in a total volume of 33 µl. Into this reaction mixture, 1 pmol RNA polymerase was added and incubated for 20 min at 37 °C to form an open complex. Then a substrate/heparin mixture containing [{alpha}-32P]UTP was added and further incubated for 10 min at 37 °C. The transcripts were precipitated with ethanol and subjected to electrophoresis on 6 % (w/v) PAGE with 7 M urea. The radioactivity of the transcripts was measured by BAS1000 Mac (Fuji Film) and analysed by MacBAS2.2 (Fuji Film).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Role of EvgAS two-component system in expression in vivo of MDR genes: MDR of evgS1 mutant
The MDR phenotype of E. coli is determined by a number of genes, including multi-drug efflux genes. To test possible involvement of the EvgAS two-component system in expression of the MDR in vivo, the sensitivity to various drugs was tested for evgS1 mutant (KMY2001) which produces a constitutively active mutant of EvgS. This mutant strain, KMY2001, showed a twofold increase in the MICs of crystal violet, rhodamine 6G, norfloxacin and trimethoprim, and a twofold or more increase in deoxycholate compared to the wild-type parent KMY1 (Table 3). To eliminate the possible influence of the AcrAB pump, the primary multi-drug transport pump (Ma et al., 1993), we then checked the drug sensitivity of these strains in the absence of acrA (Table 3). Mutant KMA2001 (acrA evgS1) showed higher resistance to at least nine drugs: doxorubicin (64-fold increase in MIC), crystal violet (2-fold), rhodamine 6G (128-fold), ethidium bromide (4-fold), acriflavin (2-fold), benzalkonium chloride (2-fold), SDS (256-fold), deoxycholate (>16-fold) and cholate (>8-fold) than KMA1 (acrA) (Table 3), suggesting that the EvgAS system is involved in the expression of genes conferring resistance to these drugs. Increased resistance to doxorubicin, crystal violet, ethidium bromide, acriflavin, benzalkonium chloride and cholate has been observed after overexpression of EvgA (Nishino & Yamaguchi, 2001a). Resistance to two drugs, norfloxacin and trimethoprim, was essentially the same between KMA1 (acrA) and KMA2001 (acrA evgS1), but a twofold increase in MICs of these two drugs was found in KMY2001, suggesting that this modest level of increase in resistance is determined by the AcrAB pump.


View this table:
[in this window]
[in a new window]
 
Table 3. MDR of evgS1 acrA mutant

 
To assess whether the outer-membrane channel, TolC, is required for the MDR phenotype, the drug sensitivity was then compared between KMAT1 (acrA tolC) and KMAT2001 (acrA tolC evgS1). Both KMAT1 and KMAT2001 were highly sensitive to all the drugs tested and no difference was observed in the drug sensitivity between these two strains (Table 3), indicating that TolC is essential for the MDR conferred by the evgS1 mutation.

Role of the EvgAS two-component system in expression in vivo of multi-drug efflux genes: enhanced transcription by the constitutive evgS1 mutant
Positive autoregulation is found in some two-component systems of E. coli such as PhoPQ (Kato et al., 1999). Along this line, the intracellular level of the response regulator EvgA could be influenced in the constitutive evgS1 mutant. To test this possibility, S1 mapping was performed for evgA mRNA that was prepared in the exponential-phase cells grown in LB medium at 37 °C. As shown in Fig. 1, the levels of evgA mRNAs from downstream of both P1 and P2 promoters (Tanabe et al., 1998) in the evgS1 mutant were as high as those in wild-type E. coli, indicating that no strong autoregulation operates in transcription of the evgAS operon.



View larger version (28K):
[in this window]
[in a new window]
 
Fig. 1. S1 mapping of the evgA promoter. Lane A+G represents the Maxam–Gilbert sequencing ladder, while lanes 1 and 2 represent S1 mapping of evgA for KMY1 (parent) and KMY2001 (evgS1 mutant), respectively. Two transcripts of evgA are indicated by arrows.

 
Next, we examined transcriptional levels of multi-drug efflux genes in an E. coli mutant expressing the constitutive mutant EvgS1 using an E. coli DNA microarray (available as supplementary data in Microbiology online at http://mic.sgmjournals.org). Most of the putative multi-drug efflux-related genes (selected on the basis of sequence similarities; Paulsen, 1998; Nishino & Yamaguchi, 2001b; Sulavik et al., 2001) were below the detection level under the steady-state of the exponential growth phase, but at least five genes or operons, which were emrKY, yhiUV, acrA, mdfA and tolC, showed increased expression in the presence of constitutive mutant EvgS1 (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4. Expression ratio of multi-drug efflux genes

MFS, major facilitator superfamily; RND, resistance-nodulation-cell division family; SMR, small MDR family; ABC, ATP-binding cassette family; MATE, multi-drug and toxic compound extrusion family; OMF, outer-membrane factor.

 
For precise estimation of transcription activation of the multi-drug efflux genes in the presence of the constitutive evgS1 mutant, S1 mapping was performed for the five genes identified by the microarray assay and the results are shown in Fig. 2(a). For the emrK gene, a single site of transcription initiation was identified in the evgS1 mutant at 109 nt upstream of the GTG start codon, as reported previously (Kato et al., 2000). For yhiU, three major protection bands, designated yhiUP1, yhiUP2 and yhiUP3, were detected at 148, 92 and 87 nt upstream of the ATG start codon, respectively, only in the evgS1 mutant strain. Likewise, the mdfA mRNA was detected only in the evgS1 mutant when a three times greater amount of RNA was used for S1 mapping. The transcription initiation site for mdfA is located at 104 nt upstream of the initiation codon. Induction of mdfA by the evgS1 mutation was also confirmed by Northern analysis (data not shown). On the other hand, transcripts of acrA and tolC were detected in both wild-type and the evgS1 mutant, even though the level of the major acrA mRNA and the tolCP1 and tolCP2 mRNAs increased in the evgS1 mutant. The transcription initiation site for the major acrA mRNA is located at 79 nt upstream of the start codon in agreement with Martin et al. (1999). As for tolC, the two transcription start sites, tolCP1 and tolCP2, were located at 105 and 97 nt upstream of the start codon, respectively.



View larger version (43K):
[in this window]
[in a new window]
 
Fig. 2. (a) S1 mapping of the multi-drug efflux genes. S1 mapping was performed using 100 µg (emrK, yhiU, acrA and tolC) or 300 µg (mdfA) RNA from wild-type KMY1 (lanes 1) or evgS1 mutant KMY2001 (lanes 2). Transcription start sites are marked with asterisks. Transcripts increased by evgS1 mutation are indicated by arrows. (b) Promoter analysis of emrK and tolC. Transcription start sites determined by S1 mapping are indicated in capital letters. The putative PhoP box (Kato et al., 1999) and the proposed 18 bp consensus sequence for EvgA binding (Masuda & Church, 2002) are boxed. Open arrows indicate the proposed inverted repeat for EvgA binding (Kato et al., 2000).

 
Previously, Kato et al. (2000) proposed an inverted repeat located in the emrK promoter region between 175 and 200 nt upstream of the translation start codon as the binding site for EvgA (Fig. 2b). A similar repeat sequence can be detected in the tolC and yhiU promoter regions, but not in acrA and mdfA. On the other hand, Masuda & Church (2002) proposed an 18 bp inverted repeat motif for EvgA binding located between 174 and 191 nt upstream of the ATG start codon of emrK. This conserved motif can be detected in the upstream regions of several highly induced genes (ydeP, b1500, yegR, emrK, yfdX and yfdW) by overexpression of EvgA. Although this motif is present in the emrK promoter region, it is not present in tolC, yhiU, acrA or mdfA.

Role of PhoPQ two-component system in tolC transcription
A close examination of the tolC promoter region revealed the presence of a sequence similar to the PhoP box consisting of a direct repeat of (T/G)GTTTA (Kato et al., 1999) between positions -44 and -29 upstream of the P1 promoter (Fig. 2b). The regulator PhoP of the PhoPQ two-component system, controlling the response to the availability of external Mg2+, is known to bind directly to this box and regulates transcription of Mg2+-responsive genes (Yamamoto et al., 2002; Minagawa et al., 2003). We then analysed the possible influence of PhoP on tolC transcription. EvgAS-dependent up-regulation from the P2 promoter of tolC (Fig. 3a, lanes 1 and 2) was markedly decreased in the absence of PhoP or PhoQ (Fig. 3a, lanes 4 and 6), indicating that the PhoPQ system is also involved in transcription activation of the tolC gene. Deleting phoPQ from the wild-type strain also resulted in a decrease in the level of tolC transcription (Fig. 3a, lanes 1, 3 and 5).



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 3. S1 mapping of tolC. (a) Transcriptional level of tolC in the phoP- and phoQ-deleted evgS1 mutant. Lanes: 1, KMY1; 2, KMY2001 (evgS1); 3, KMP1 (phoP); 4, KMP2001 (phoP evgS1); 5, KMQ1 (phoQ); 6, KMQ2001 (phoQ evgS1). (b) Transcriptional level of tolC in an evgA-overexpressed strain. Lanes: 1, KMY1/pA191 (induced with arabinose); 2, KMY1/pA191 (not induced with arabinose). (c) Transcriptional level of tolC in phoP-overexpressed strain. Lanes: 1, KMY1/pMW119; 2, KMY1/pHO119.

 
We then suspected a signal transduction from the activated sensor EvgS to the regulator PhoP, referred to as ‘cross-talk’ between the two systems. To check this possibility, evgA was overexpressed from an arabinose promoter. Even in the wild-type evgS background, up-regulation of tolC was observed after expression of an excess amount of EvgA (Fig. 3b), indicating that cross-talk between EvgS and PhoP was not the major pathway of regulation. However, a very modest increase in transcription from the P2 promoter was found when the phoQ and evgS1 phoQ mutants were compared (Fig. 3a, lanes 5 and 6). This increase may derive from cross-talk between EvgS and PhoP, although we were not able to detect in vitro phosphotransfer from EvgS to PhoP (data not shown). An increase in transcription of tolC from overproduction of EvgA has also been mentioned by Masuda & Church (2002) and Nishino et al. (2003).

In our array results, the evgS1 mutant resulted in increased transcription of phoP and phoQ, as well as other known PhoP regulons such as mgtA and mgrB (Minagawa et al., 2003). Therefore, we suspected a simple cascade of signal transduction where the increased level of PhoP by the EvgAS system enhances the transcription of tolC. Transformation of the wild-type strain with a plasmid containing the phoPQ genes increased the transcriptional level of phoP to the level found in the evgS1 mutant (data not shown), but did not increase the transcriptional level of tolC (Fig. 3c), denying the involvement of the simple cascade. Signal transduction from the EvgAS to the PhoPQ system is currently under investigation. In addition, direct binding of EvgA to the tolC promoter was not detected in our DNase I footprinting analysis (data not shown).

Activation by EvgA of in vitro transcription of emrK
To confirm the transcriptional activation of multi-drug efflux genes emrKY and yhiUV by phosphorylated EvgA, in vitro transcription of truncated DNA templates containing either the emrK or yhiU promoter region was carried out in the presence or absence of EvgA, which was phosphorylated by acetylphosphate. As shown in Fig. 4(a), transcription of emrK was activated 2·3-fold by the addition of 100 pmol EvgA without prior treatment with acetylphosphate. In the presence of acetylphosphate, however, the transcription was activated 4·3-fold by 100 pmol EvgA, indicating that the phosphorylated EvgA directly activates emrK transcription. On the other hand, the increased amount of EvgA did not enhance the transcriptional level of evgA, which is transcribed divergently from emrK, neither in the absence nor in the presence of acetylphosphate. This result is consistent with our S1 mapping (Fig. 1) and array results (expression ratio of evgA; 0·82, 0·66) indicating once again that transcription of evgA is not enhanced by the activation of EvgS.



View larger version (75K):
[in this window]
[in a new window]
 
Fig. 4. Effect of phosphorylated EvgA on in vitro transcription of emrK and yhiU. DNA templates containing promoter regions of emrK (a) and yhiU (b) were incubated with EvgA, RNA polymerase and with or without the presence of acetylphosphate. Samples were subjected to electrophoresis on a 6 % polyacrylamide sequencing gel. Transcripts originating from each promoter(s) are indicated by arrows.

 
Transcription in vitro was also carried out for the yhiU promoter. Addition of both EvgA and phosphorylated EvgA did not show any effect on transcription of yhiU (Fig. 4b). We have also performed a DNaseI footprinting analysis of EvgA against the promoter region of yhiU, but failed to detect direct binding of EvgA (data not shown). These results suggested that yhiU transcription either is indirectly regulated by EvgA or requires an additional factor.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
It has been reported that overexpression of the regulator EvgA confers MDR to a drug-hypersusceptible E. coli strain which lacks the major multi-drug efflux pump gene, acrB (Nishino & Yamaguchi, 2001a, 2002), and acid resistance to exponentially growing cells (Masuda & Church, 2002). In addition to the up-regulation of the multi-drug efflux genes confirmed in the present study, activation of EvgS also up-regulated the acid-resistance-related genes, gadABC, hdeAB, ydeP, ydeO and yhiE (available as supplementary data in Microbiology online at http://mic.sgmjournals.org), as in the EvgA-overexpressed strain (Masuda & Church, 2002). We have found that acid resistance is also conferred to an exponentially grown evgS1 mutant (data not shown). Although as many as 225 ORFs were positively regulated by the EvgS activation from our array data, only 79 ORFs were identified from the array data reported for an EvgA-overexpressing strain (Masuda & Church, 2002). This difference comes mostly from how the positive ORFs were selected: we chose ORFs with changes of twofold or more, Masuda & Church (2002) chose those with changes of more than fourfold (133 ORFs showed changes of more than fourfold in our array). Another reason may be that the activation of the EvgAS system regulates the EvgAS-regulated genes more efficiently than by the overproduction of EvgA. In fact, transcriptional levels of tolC were higher in the evgS1 mutant than in the EvgA-overexpressed strains in our S1 mapping (Fig. 3a, lane 2, and Fig. 3b, lane 1). A difference in the microarray employed may also contribute to the difference in sensitivity. Whereas Masuda & Church (2002) used an array of short oligonucleotides, we used an array of full-length ORFs. Longer nucleotides on the array will result in a higher intensity of signals, resulting in better detection of ORFs with low expression levels. However, when the genes most affected were compared, the two array data were highly similar, in spite of the differences in the strains and arrays used. This confirms that the overproduction method is valid for analysing the EvgAS system and the overproduction method can probably be employed for other non-activated two-component systems.

MDR was conferred by EvgS activation in the present study, clearly indicating the involvement of the EvgAS system to MDR. From our drug susceptibility tests (Table 3), the MDR phenotype conferred by the evgS1 mutation required TolC. Although EmrKY, YhiUV and AcrAB are three-component efflux pumps requiring TolC for their activity (Nishino & Yamaguchi, 2002; Fralick, 1996), MdfA is a single-component efflux pump not requiring TolC (Edgar & Bibi, 1997, 1999). Thus, the contribution of the increased expression of MdfA to the MDR found under the conditions in this study was considered to be small. However, since single-component pumps such as MdfA can cooperate with three-component pumps, with the latter presumably pumping out the drugs from the periplasm (Lee et al., 2000), the possibility of MdfA cooperating with the YhiUV and EmrKY pumps still remains. Furthermore, since KMY2001 did not show prominent resistance compared with KMY1, the contribution of AcrAB was considered as negligible. Therefore, the MDR conferred by the evgS1 mutation was suggested to be mainly caused by the orchestrated expression of emrKY, yhiUV and tolC. Overexpression of EmrKY results in resistance against deoxycholate and overexpression of YhiUV to resistance against rhodamine 6G, erythromycin, doxorubicin, ethidium bromide, tetraphenylphosphonium bromide, SDS, deoxycholate, crystal violet and benzalkonium (Nishino & Yamaguchi, 2001b). The variety of drugs to which the evgS1 mutant showed resistance fits well with the proposed drugs extruded by EmrKY and YhiUV.

We have also shown that the phosphorylated form of EvgA increased the in vitro transcription of emrKY (Fig. 4a), indicating a direct regulation of emrKY by the EvgAS system. The other pump, yhiUV, was not up-regulated solely by phosphorylated EvgA (Fig. 4b), thus suggesting the participation of other transcriptional factors. From our microarray analysis, a number of transcriptional factors such as ydeO, yhiF and yhiE were up-regulated in the evgS1 mutant (available as supplementary data in Microbiology online at http://mic.sgmjournals.org). These factors may be involved in the direct regulation of yhiUV.

It has been reported that TolC is a member of the Mar-Sox-Rob regulon and contains a consensus sequence of the Mar-Sox-Rob box at a location upstream (-96 to -73) of the structural tolC gene (Aono et al., 1998). The putative Mar-Sox-Rob box is located just downstream of tolCP2 (Fig. 2b). Together with our array results, indicating a decreased level of marA transcripts by the evgS1 mutation (expression ratio 0·18, 0·25), unchanged rob level (1·3, 1·4) and non-detectable soxS level, the up-regulation of tolC by EvgAS is considered to be independent of the global regulators MarA, SoxS or Rob. In this study, we have found that the PhoPQ system, involved in the adaptation to Mg2+-limiting environments by regulation of genes related to Mg2+ transport and LPS modification (Groisman, 2001), was essential to the regulation of tolC by EvgAS. This regulation did not involve cross-talk between EvgS and PhoP, nor a simple cascade due to an increase in the PhoP level. Another cascade of signal transduction, in which the phosphorylated EvgA activates the PhoPQ system followed by the up-regulation of tolC, was suggested (Fig. 3). Such a cascade between different two-component systems has been reported between PhoPQ and PmrAB systems via the pmrD gene in Salmonella (Groisman, 2001).

In the present study, our data suggested that the activation of the sensor EvgS initiated a transcriptional network with the phosphorylated EvgA serving as the master regulator and led to MDR by a coordinated transcription of emrKY, yhiUV and tolC. Further investigation of this network will clarify the biological function of the EvgAS system and identify the environmental signal to which EvgS responds. These findings will contribute to our knowledge of the mechanism underlying the occurrence of MDR.


   ACKNOWLEDGEMENTS
 
We thank Joe A. Fralick for kindly supplying strains W4573 and W4573tolC : : Tn10. This work was supported in part by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan, CREST of JST (Japan Science and Technology), the Sasakawa Research Grant from the Japan Science Society, and the Hayashi Memorial Foundation for Female Natural Scientists.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Aiba, H. (1983). Autoregulation of the Escherichia coli crp gene: CRP is a transcriptional repressor of its own gene. Cell 32, 141–149.[Medline]

Albright, L. M., Huala, E. & Ausubel, E. M. (1989). Prokaryotic signal transduction mediated by sensor and regulator protein pairs. Annu Rev Genet 23, 11–36.

Aono, R., Aibe, K., Inoue, A. & Horikoshi, K. (1991). Preparation of organic solvent tolerant mutants from Escherichia coli K-12. Agric Biol Chem 55, 1935–1938.

Aono, R., Tsukagoshi, N. & Yamamoto, M. (1998). Involvement of outer membrane protein TolC, a possible member of the mar-sox regulon, in maintenance and improvement of organic solvent tolerance of Escherichia coli K-12. J Bacteriol 180, 938–944.[Abstract/Free Full Text]

Bock, A. & Gross, R. (2002). The unorthodox histidine kinases BvgS and EvgS are responsive to the oxidation status of a quinone electron carrier. Eur J Biochem 269, 3479–3484.[Abstract/Free Full Text]

Casadaban, M. J. (1976). Transposition and fusion of lac genes to selected promoters in Escherichia coli using bacteriophage lambda and Mu. J Mol Biol 104, 541–555.[Medline]

Edgar, R. & Bibi, E. (1997). MdfA, an Escherichia coli multidrug resistance protein with an extraordinarily broad spectrum of drug recognition. J Bacteriol 179, 2274–2280.[Abstract]

Edgar, R. & Bibi, E. (1999). A single membrane-embedded negative charge is critical for recognizing positively charged drugs by the Escherichia coli multidrug resistance protein MdfA. EMBO J 18, 822–832.[Abstract/Free Full Text]

Fralick, J. A. (1996). Evidence that TolC is required for functioning of the Mar/AcrAB efflux pump of Escherichia coli. J Bacteriol 178, 5803–5805.[Abstract/Free Full Text]

Georgellis, D., Kwon, O. & Lin, E. C. (2001). Quinones as the redox signal for the arc two-component systems of bacteria. Science 292, 2314–2316.[Abstract/Free Full Text]

Groisman, E. A. (2001). The pleiotropic two-component regulatory system PhoP-PhoQ. J Bacteriol 183, 1835–1842.[Free Full Text]

Kajitani, M. & Ishihama, A. (1983). Determination of the promoter strength in the mixed transcription system: promoters of lactose, tryptophan and ribosomal protein L10 operons from Escherichia coli. Nucleic Acids Res 11, 671–686.[Abstract]

Kato, A., Tanabe, H. & Utsumi, R. (1999). Molecular characterization of the PhoP-PhoQ two-component system in Escherichia coli K-12: identification of extracellular Mg2+-responsive promoters. J Bacteriol 181, 5516–5520.[Abstract/Free Full Text]

Kato, A., Ohnishi, H., Yamamoto, K., Furuta, E., Tanabe, H. & Utsumi, R. (2000). Transcription of emrKY is regulated by the EvgA-EvgS two-component system in Escherichia coli K-12. Biosci Biotechnol Biochem 64, 1203–1209.[Medline]

Lee, A., Mao, W., Warren, M. S., Mistry, A., Hoshino, K., Okumura, R., Ishida, H. & Lomovskaya, O. (2000). Interplay between efflux pumps may provide either additive or multiplicative effects on drug resistance. J Bacteriol 182, 3142–3150.[Abstract/Free Full Text]

Ma, D., Cook, D. N., Alberti, M., Pon, N. G., Nikaido, H. & Hearst, J. E. (1993). Molecular cloning and characterization of acrA and acrE genes of Escherichia coli. J Bacteriol 175, 6299–6313.[Abstract]

Makinoshima, H., Nishimura, A. & Ishihama, A. (2002). Fractionation of Escherichia coli cell populations at different stages during growth transition to stationary phase. Mol Microbiol 43, 269–279.[CrossRef][Medline]

Martin, R. G., Gillette, W. K., Rhee, S. & Rosner, J. L. (1999). Structural requirements for marbox function in transcriptional activation of mar/sox/rob regulon promoters in Escherichia coli: sequence, orientation and spatial relationship to the core promoter. Mol Microbiol 34, 431–441.[CrossRef][Medline]

Masuda, N. & Church, G. M. (2002). Escherichia coli gene expression responsive to levels of the response regulator EvgA. J Bacteriol 184, 6225–6234.[Abstract/Free Full Text]

Minagawa, S., Ogasawara, H., Kato, A., Yamamoto, K., Eguchi, Y., Oshima, T., Mori, H., Ishihama, A. & Utsumi, R. (2003). Identification and molecular characterization of the Mg2+-stimulon of Escherichia coli. J Bacteriol 185, 3696–3702.[Abstract/Free Full Text]

Nishino, K. & Yamaguchi, A. (2001a). Overexpression of the response regulator evgA of the two-component signal transduction system modulates multidrug resistance conferred by multidrug resistance transporters. J Bacteriol 183, 1455–1458.[Abstract/Free Full Text]

Nishino, K. & Yamaguchi, A. (2001b). Analysis of a complete library of putative drug transporter genes in Escherichia coli. J Bacteriol 183, 5803–5812.[Abstract/Free Full Text]

Nishino, K. & Yamaguchi, A. (2002). EvgA of the two-component signal transduction system modulates production of the YhiUV multidrug transporter in Escherichia coli. J Bacteriol 184, 2319–2323.[Abstract/Free Full Text]

Nishino, K., Inazumi, Y. & Yamaguchi, A. (2003). Global analysis of genes regulated by EvgA of the two-component regulatory system in Escherichia coli. J Bacteriol 185, 2667–2672.[Abstract/Free Full Text]

Oshima, T., Wada, C., Kawagoe, Y., Ara, T., Maeda, M., Masuda, Y., Hiraga, S. & Mori, H. (2002a). Genome-wide analysis of deoxyadenosine methyltransferase- mediated control of gene expression in Escherichia coli. Mol Microbiol 45, 673–695.[CrossRef][Medline]

Oshima, T., Aiba, H., Masuda, Y., Kanaya, S., Sugiura, M., Wanner, B. L., Mori, H. & Mizuno, T. (2002b). Transcriptome analysis of all two-component regulatory system mutants of Escherichia coli K-12. Mol Microbiol 46, 281–291.[CrossRef][Medline]

Paulsen, I. T., Sliwinski, M. K. & Saier, M. H. Jr (1998). Microbial genome analyses: global comparisons of transport capabilities based on phylogenies, bioenergetics and substrate specificities. J Mol Biol 277, 573–593.[CrossRef][Medline]

Perraud, A.-L., Weiss, V. & Gross, R. (1999). Signaling pathways in two-component phosphorelay systems. Trends Microbiol 7, 115–120.[CrossRef][Medline]

Sulavik, M. C., Houseweart, C., Cramer, C. & 8 other authors (2001). Antibiotic susceptibility profiles of Escherichia coli strains lacking multidrug efflux pump genes. Antimicrob Agents Chemother 45, 1126–1136.[Abstract/Free Full Text]

Tanabe, H., Yamasaki, K., Kato, A., Yoshioka, S. & Utsumi, R. (1998). Identification of the promoter region and the transcriptional regulatory sequence of the evgAS operon of Escherichia coli. Biosci Biotechnol Biochem 62, 286–270.[Medline]

Tsukagoshi, N. & Aono, R. (2000). Entry into and release of solvents by Escherichia coli in an organic aqueous two-liquid-phase system and substrate specificity of the AcrAB-TolC solvent extruding pump. J Bacteriol 182, 4803–4810.[Abstract/Free Full Text]

Utsumi, R. (2002). Bacterial signal transduction systems for environmental adaptation and their inhibitors. Microbes and Environment 17, 1–5.[CrossRef]

Utsumi, R., Katayama, S., Taniguchi, M. & 7 other authors (1994). Newly identified genes involved in the signal transduction of Escherichia coli K-12. Gene 140, 73–77.[CrossRef][Medline]

Yamamoto, K., Ogasawara, H., Fujita, N., Utsumi, R. & Ishihama, A. (2002). Novel mode of transcription regulation of divergently overlapping promoters by PhoP, the regulator of two-component system sensing external magnesium availability. Mol Microbiol 45, 423–438.[CrossRef][Medline]

Received 6 May 2003; revised 20 June 2003; accepted 25 June 2003.