- Email: [email protected]
TbtABM, a multidrug efflux pump associated with tributyltin resistance in Pseudomonas stutzeri
FEMS Microbiology Letters 232 (2004) 7^14
www.fems-microbiology.org
TbtABM, a multidrug e¥ux pump associated with tributyltin resistance in Pseudomonas stutzeri Florence Jude a
a;b
, Corinne Arpin a; , Ce¤line Brachet-Castang a , Miche'le Capdepuy Pierre Caumette c , Claudine Quentin a
a;b
,
Laboratoire de Microbiologie, Universite¤ Victor Segalen Bordeaux 2, 146 rue Le¤o Saignat, 33076 Bordeaux Cedex, France b Laboratoire d’Oce¤anographie Biologique, Universite¤ de Bordeaux 1, 33120 Arcachon, France c Laboratoire d’Ecologie Mole¤culaire, Universite¤ de Pau, 64000 Pau, France Received 25 November 2003; received in revised form 19 December 2003; accepted 21 December 2003 First published online 30 January 2004
Abstract Tributyltin (TBT) is a toxic agent used in marine antifouling paints. Among the bacterial flora of a polluted harbor, TBT-resistant strains of Pseudomonas stutzeri have been isolated. In the strain 5MP1 (TBT minimum inhibitory concentration (MIC) v 1000 mg l31 ), TBT resistance was found to be associated with the presence of the operon tbtABM, homologous to the resistance^nodulation^cell division (RND) efflux pump family, as demonstrated by cloning in Escherichia coli. TbtABM exhibited the greatest homology (60.9^ 84.9%) with the TtgDEF and SrpABC systems, both involved in aromatic compound tolerance in P. putida. TbtABM conferred multidrug resistance (MDR) including to n-hexane, nalidixic acid, chloramphenicol, and sulfamethoxazole (antibiotic MICsU4 for the E. coli host strain carrying the operon). By polymerase chain reaction amplification and hybridization experiments, the presence of tbtABM was detected in the TBT-sensitive P. stutzeri 3MP1 (TBT MIC 25 mg l31 ). However, the latter strain did not seem to express TbtABM. This is the first description of a MDR efflux pump in P. stutzeri, and of a new kind of substrate, TBT, for the RND family of transporters. = 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords : Tributyltin resistance; Pseudomonas stutzeri ; E¥ux pump; RND family transporter; Multidrug resistance
1. Introduction Tributyltin (TBT) is an organostannic compound, used industrially as a stabilizer in plastics, wood preservative, and antifouling agent in boat paints [1]. However, in the latter case, TBT has induced high larval malformations of oysters [2]. Actually, TBT is toxic for both eukaryotes and prokaryotes due to a partition coe⁄cient in octanol^water mixture (log Pow at 25‡C) of 3.7, allowing its dissolution into biological membranes, disturbing their integrity, and ultimately compromising their physiological functions [1,3]. Although the use of TBT has been restricted for 15 years in many countries, this compound persists in freshwater and marine sediments and remains a potential source of pollution [4,5]. Bacterial populations might help to detoxify TBT. In-
* Corresponding author. Tel. : +33 (5) 57 57 10 75; Fax : +33 (5) 56 90 90 72. E-mail address : [email protected] (C. Arpin).
deed, pure bacterial cultures as well as natural aquatic sediments are able to biomethylate the inorganic SnCl2 [5^8]. They might also perform the reverse reaction on another tin compound, i.e. demethylation of TBT that leads to less toxic derivatives. Thus, TBT-resistant strains might serve as cleaning agents, justifying their screening in polluted environments. However, enzymic modi¢cations may not be involved. Indeed, very few studies have investigated the mechanism of TBT bacterial resistance. Wuertz et al. [9] have described TBT-resistant bacteria that contained plasmids and had multiple heavy metal and antibiotic resistances. Miller et al. [10] have reported TBT together with chromium resistance, mediated by a conjugative plasmid in Pseudomonas aeruginosa. Finally, Fukugawa and Suzuki have characterized a chromosomal gene involved in TBT resistance in a strain of Alteromonas sp., encoding a protein related to Naþ /Hþ antiporters and various Ca2þ transport systems [11] and/or glycosylase [12]. The harbor of Arcachon (France) contains relatively high TBT concentrations, reaching 150 ng g31 of sediment
0378-1097 / 04 / $22.00 = 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/S0378-1097(04)00012-6
FEMSLE 11399 3-3-04
8
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
[4]. In a previous work, 103 bacterial strains, including 20 Pseudomonas (eight P. stutzeri, two P. £uorescens and 10 P. putida), have been isolated from this polluted environment, and their TBT susceptibilities have been evaluated [4]. Six P. stutzeri were found to have a high level of TBT resistance with minimum inhibitory concentrations (MICs) of 500 to v 1000 mg l31 compared with two sensitive P. stutzeri strains exhibiting a MICTBT of 25 mg l31 . The TBT-resistant strains did not carry any plasmid and the TBT resistance was not transferable by conjugation, suggesting that the genetic determinant was located on the chromosomal DNA [4]. The aim of this study was to characterize the TBT resistance mechanism(s) in P. stutzeri. After constructing a DNA library in Escherichia coli, and sequencing a subcloned insert conferring TBT resistance, the analysis of the deduced amino acid sequence showed that the involved proteins, named TbtA-TbtB-TbtM (Tbt, for tributyltin), were homologous to the components of the multidrug resistance (MDR) e¥ux pumps belonging to the resistance^ nodulation^division (RND) family [13,14].
2. Materials and methods 2.1. Bacterial strains, plasmids and growth conditions P. stutzeri strains [4] and E. coli HB101 (supE44 3 hsdS20(r3 B mB ) recA13 ara-14 proA2 lacY1 galK12 rpsL20 xyl-5 mtl-1) (Gibco-Bethesda Research Laboratories) used in this study are listed in Table 1. The other Pseudomonas sp. strains analyzed by polymerase chain reaction (PCR) and hybridization experiments have been described elsewhere [4]. Bacterial strains were grown with shaking (200 strokes min31 ) at 30‡C (P. stutzeri) or 37‡C (E. coli) in Mueller^Hinton medium (MH, BioMe¤rieux) or in Luria broth (Gibco-BRL). Selecting agents for resident plasmids were added to growth media when appropriate : ticarcillin (P. stutzeri) or ampicillin (E. coli) at 200 mg l31 , and tributyltin chloride (TBTCl, Sigma-Aldrich) from 10 to 1000 mg l31 .
2.2. DNA extraction and analysis Total DNA was prepared by a previously described method [15]. Midi-preparations of plasmid DNA were performed using a Qiagen puri¢cation kit, whereas mini-preparations were achieved by an alkaline lysis procedure [15]. Restriction digests were carried out according to the supplier’s instructions (Gibco BRL) and analyzed by electrophoresis on agarose gels in Tris-borate bu¡er [15]. 2.3. Cloning experiments and transformation of E. coli and P. stutzeri strains After ligation using a T4 DNA ligase (Gibco BRL) between 1 Wg of HindIII alkaline phosphatase-treated plasmid pUC18 (Gibco-BRL) and 2 Wg of HindIII genomic DNA from P. stutzeri 5MP1, the mixture was used to transform the E. coli HB101 strain by electroporation, as indicated by the manufacturer (Bio-Rad). Screening of TBT-resistant clones required a two-step selection: ¢rstly, colonies were collected on MH agar plates with ampicillin; then, ampicillin-resistant clones were replicated on MH agar plates containing both ampicillin and TBTCl. One TBT-resistant clone was selected that carried a recombinant plasmid, pTBT10, containing a 10.6-kb insert. After subcloning using the endonucleases XbaI and MunI, the 8.0-kb generated fragment was puri¢ed from agarose gel using the protocol of the QIAEXII kit (Qiagen), and its ends were completed by the Klenow enzyme (Gibco BRL) before ligation into the SmaI-linearized pUC18 vector, and transformation experiment. 2.4. DNA sequencing and analysis Automated £uorescent sequencing was done using the ABI Prism Dye Terminator Cycle Sequencing Ready Reaction kit and the Applied Biosystem model 377 (Biosystems Division, Perkin Elmer). Overlapping sequences from both strands were obtained using a series of custom-synthesized primers (Eurogentec). Nucleotide and deduced amino acid sequences were analyzed using the PC Gene
Table 1 Antimicrobial susceptibilities of TBT-sensitive and -resistant strains used in this study Strain
Presence of tbtABMb
P. P. P. P. E. E.
+ + + 3 3 +
MIC (mg l31 ) of: TBT
stutzeri 3MP1 and 3MP2 stutzeri 5MP1 stutzeri 3FP1 and 3FP2 stutzeri ATCC 17588 coli HB101 coli HB101+pTBT10
25 v 1000 500 750 50 v 1000
a
Growth on n-hexanec
NAL
CHL
SMX
8 16 16 8 1 4
8 32 16 16 2 8
4 4 4 4 0.05 0.2
a
NAL, nalidixic acid ; CHL, chloramphenicol ; SMX, sulfamethoxazole. Presence (+) or absence (3) of the tbtABM operon, as detected by PCR and hybridization experiments. c +, tolerance is assessed by a visible lawn of growth; or 3, sensitive strains produced no visible lawn. b
FEMSLE 11399 3-3-04
+ + + 3 3 +
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
Software Package (Intelligenetics) and the BLAST program available from the National Institute for Biotechnology Information server (http://www.blast.genome.ad.jp). 2.5. Antimicrobial susceptibility testing The MICs of 15 antimicrobial agents including 11 antibiotics (nalidixic acid, cipro£oxacin, tetracycline, doxycycline, chloramphenicol, sulfamethoxazole, erythromycin, rifampicin, fusidic acid, vancomycin, ticarcillin, ethidium bromide, sodium dodecyl sulfate (SDS), crystal violet and acri£avine) were determined by the standard serial twofold dilution technique in MH agar with a ¢nal inoculum of about 104 CFU per spot (http://www.sfm.asso.fr). Tolerance to organic solvents (cyclohexane, n-hexane, p-xylene, toluene) and alcohol (butanol) was measured by plating a 5-Wl aliquot of a bacterial suspension (107 cells ml31 ) on MH agar [16]. After drying, the organic solvent or alcohol was overlaid to a depth of 2^3 mm. Then, the plates were sealed and incubated overnight.
9
separated by electrophoresis on a 0.8% (w/v) agarose gel before being transferred to a nylon N membrane (Nytran, Schleicher and Schuell). The Southern blot hybridizations were performed using as a probe a digoxigenin (DIG)labelled 1.423-kb fragment, which consisted of the tbtB3tbtB4 amplicon (see above). Stringent hybridizations at 42‡C in the presence of 50% formamide and subsequent developments of the blot were carried out using the DIG DNA kit according to the supplier’s instructions (Boehringer Mannheim). 2.8. Outer membrane protein analysis The outer membrane proteins from P. stutzeri and E. coli strains were prepared by a di¡erential 1% sarcosyl solubilization of isolated cell envelopes [17]. An identical quantity of proteins per sample (35 Wg) was separated by SDS^polyacrylamide gel electrophoresis (PAGE) (10% w/v acrylamide and 0.27% w/v N,NP-methylene bisacrylamide), before being stained with a Coomassie brilliant blue solution [15].
2.6. PCR ampli¢cation 2.9. Reverse transcription (RT) PCR experiments An internal sequence of tbtB was ampli¢ed using primers tbtB3 (5P-CGCCGGCGCGTTATCGCTGG-3P, starting 60 bp downstream of the ATG initiation codon of the tbtB open reading frame (ORF)) and tbtB4 (5P-GGTGGCGCACAGCGCCGGGG-3P, starting 1.423 kb downstream of the tbtB3 primer) (Fig. 1). The tbtR coding region and the intergenic tbtR-tbtA region were ampli¢ed with oligonucleotides tbtR4 (5P-GATGCCACAGACTTTTAAC-3P, starting 711 bp upstream of the end of tbtR) and tbtA1 (5P-CAGCGGAACGGAATGCCATGC-3P, beginning 45 bp downstream of the GTG initiation codon of tbtA) (Fig. 1).
RNA from 1 ml late culture was prepared with the SV Total RNA Isolation kit that includes a DNase treatment, according to the supplier’s instructions (Promega). The RT-PCR analysis was undertaken with and without RT enzyme (as control) in a single tube RT-PCR system, using the AccessQuick0 RT-PCR kit (Promega). RT-PCR experiments were carried out with the oligonucleotides tbtBD (5P-GGCCATCGGCTTGCTGGTGG-3P) and tbtB5 (5P-CAGCGCCACCAGCACCGAGAG-3P), starting at nucleotides 3371 and 3629 of the tbtB gene, according to the numbering of the sequence deposited in GenBank.
2.7. DNA hybridization experiments 2.10. Accession number Chromosomal DNA (1 Wg) of bacterial strains was digested with HindIII, and the generated fragments were
The DNA sequences of the genes described in this study have been deposited in the GenBank1 sequence database under accession number AF041464.
3. Results 3.1. TBT resistance is associated with the tbtABM operon
Fig. 1. Partial restriction map of the HindIII insert carrying the tbtABM operon. The translation orientation of ORFs is indicated. The position and orientation of primers used for PCR and RT-PCR ampli¢cations are indicated.
After construction of an HindIII DNA library in E. coli HB101, a single clone of the 1000 replicated was resistant to both ampicillin and TBT (MICTBT v 1000 mg l31 vs. 50 mg l31 for the host strain). This clone contained the plasmid pTBT10, consisting of the pUC18 vector with a 10.6-kb insert. The partial restriction map of this fragment is shown in Fig. 1. To con¢rm that TBT-resistant phenotype was due to the presence of this recombinant plasmid,
FEMSLE 11399 3-3-04
10
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
pTBT10 was transferred by electroporation to a new set of competent cells of E. coli HB101. After the two-step selection, 100% of the replicated ampicillin-resistant clones were found to be TBT-resistant, in contrast with none of the clones harboring the native plasmid pUC18. After subcloning, an insert of 8.0 kb obtained by XbaI-MunI double digestion still conveyed TBT resistance. A part of the XbaI-NunI fragment was sequenced on both strands and the analysis of the resulting 7.311-kb sequence revealed the presence of three successive ORFs, called tbtA (1146 bases), tbtB (3138 bases) and tbtM (1428 bases) in the same orientation, as indicated in Fig. 1. All three ORFs exhibited a putative ribosome binding site sequence -GAGG- (tbtA and tbtB) or -AGGA- (tbtM), 8^12 nucleotides upstream of their GTG (tbtA) or ATG (tbtB and tbtM) initiation codon. These three genes potentially formed an operon since a stretch of only 14 bp bridged the ¢rst and the second ORFs, and the second gene stop codon overlapped the ATG of the third gene. 3.2. TbtABM is a member of the e¥ux pump RND family Scan of the GenBank sequence databases revealed that TbtA-TbtB-TbtM showed striking resemblances with proteins known to be involved in proton-dependent multidrug e¥ux pumps, such as the MexA-MexB-OprM pump (56.764.3-62.6%, respectively with TbtA-TbtB-TbtM) in P. aeruginosa [18], and AcrA-AcrB-TolC (49.6-58.2-15.8%, respectively) in E. coli [19]. Nevertheless, the highest degrees of homology were obtained with TtgD-TtgE-TtgF (72.5-84.9-65.8%, respectively) and SrpA-SrpB-SrpC (73.3-84.4-60.9%, respectively) which are e¥ux systems conferring organic solvent tolerance in P. putida [20,21]. The hydropathy pro¢les of TbtA (41.6 kDa) and TbtM (52.4 kDa) were consistent with their putative functions of fusion membrane protein and outer membrane protein, respectively [22]. Hydropathy analysis of the putative transporter TbtB (113.1 kDa) showed that it comprised 12 potential transmembrane segments (TMS) and two large loops, between TMS1 and 2 and between TMS7 and 8. TbtB possessed four highly conserved sequences, motif A (residues spanning positions 86^131), motif B (positions 445^494), motif C (amino acids 966^1010) and motif D (residues 389^424) which exhibited 93, 72, 75 and 72% homology, respectively, with previously described consensus regions of the RND transporters [13,14]. 3.3. tbtABM is speci¢c to P. stutzeri, but irregularly found in this species The presence of tbtABM was searched by PCR using primers tbtB3 and tbtB4 (Fig. 1). Among the 20 Pseudomonas sp. strains (eight P. stutzeri, two P. £uorescens and 10 P. putida) previously isolated from the TBT-polluted sediments of Arcachon harbor [4], and two collection strains, E. coli HB101 and P. stutzeri ATCC 17588, a
Fig. 2. Southern blot hybridization of the HindIII-restricted genomic DNA. Only strains for which DNA gave a positive hybridization are shown in this ¢gure. The probe used for the hybridization experiments was a 1.423-kb internal tbtB fragment. Lanes 1 and 2, TBT-sensitive P. stutzeri 3MP1 and 3MP2, respectively. Lanes 3^5, TBT-resistant P. stutzeri 3FP1, 3FP2 and 5MP1, respectively.
1.423-kb PCR product was obtained in only three of the six TBT-resistant P. stutzeri (5MP1, 3FP1, and 3FP2), and in the two TBT-sensitive P. stutzeri (3MP1 and 3MP2) (data not shown). The EcoRI restriction pro¢les of the amplicon con¢rmed the presence of the tbtB sequence, since two fragments of the expected sizes (760 and 663 bp) were obtained. Reproducible hybridization experiments con¢rmed the previous PCR results (Fig. 2). Indeed, the ¢ve P. stutzeri strains (5MP1, 3FP1, 3FP2, 3MP1 and 3MP2) carried tbtABM on a HindIII fragment of ca. 11.0 kb. 3.4. TbtABM is a multidrug e¥ux pump in E. coli P. stutzeri strains did not exhibit signi¢cant di¡erences in susceptibility to the 15 tested agents, except for nalidixic acid and chloramphenicol. MICs of these antimicrobials were slightly increased (two- to four-fold) in the three TBT-resistant strains harboring the tbtABM operon (5MP1, 3FP1 and, 3FP2) compared with the two TBTsensitive strains (3MP1 and 3MP2) (Table 1). More convincingly, the presence of pTBT10 in E. coli HB101 led to a similar increase in MICs of nalidixic acid and chloramphenicol, but also of sulfamethoxazole (Table 1). The MIC increases were low (four-fold), but reproducible through three independent experiments. Furthermore, the recombinant E. coli clone was able to grow in the presence of nhexane, whereas the host strain did not. In addition, ¢ve of eight P. stutzeri strains (5MP1, 3FP1, 3FP2, 3MP1 and 3MP2) produced a visible lawn of growth on a medium containing n-hexane (Table 1), and the MIC of TBT in 3MP1 was increased from 25 to 500 mg l31 in the presence of this organic solvent. The other tested solvents or alcohol with a higher toxicity (log Pow lower than 3.5) gave either negative or non-reproducible results by the method used.
FEMSLE 11399 3-3-04
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
11
3.5. Di¡erential TbtABM expression between the P. stutzeri strains The content of the outer membrane proteins of the TBT-sensitive 3MP1 and TBT-resistant 5MP1 strains of P. stutzeri was compared. The results, shown in Fig. 3, indicated the presence of an additional protein in 5MP1 (lane 2), which was absent in 3MP1 (lane 1); the size of 52 kDa of this protein was consistent with the theoretical molecular mass of the outer membrane protein, TbtM (52.4 kDa). In addition, RT-PCR analysis (Fig. 4), performed with and without RT enzyme (as control for DNA contamination) and using the tbtBD and tbtB5 primers, revealed a 259-bp RT fragment in the TBT-resistant strain (lane 4) although no ampli¢cations were observed with the sensitive strain (lane 2). 3.6. TbtR is a potential negative regulator Sequencing of the 1008-bp region located upstream of tbtA revealed that the locus contained an additional ORF named tbtR, of small size (777 bp) and transcribed divergently from tbtABM (Fig. 1). The deduced TbtR sequence (27.7 kDa) showed no signi¢cant identities with those compiled in the GenBank databases, except for TtgV (58%) and SrpS (56%), the putative regulatory proteins of the ttgGHI and srpABC operons of P. putida, respec-
Fig. 4. RT-PCR analysis of TBT-sensitive and -resistant strains. Lanes 1 and 2, P. stutzeri 3MP1 strain; lanes 3 and 4, P. stutzeri 5MP1 strain; lanes 5 and 6, E. coli HB101 strain; lanes 7 and 8, E. coli HB101 strain with the recombinant plasmid pTBT10; lane M, pGEM0 size marker (Promega). Lanes 3 and + were experiments performed without and with reverse transcriptase, respectively.
tively [23,24]. Using the primers tbtR4 and tbtA1 (Fig. 1), the region upstream of tbtABM in 3MP1 was ampli¢ed and sequenced. Sequence comparison between 5MP1 and 3MP1 revealed 14 substitutions in the tbtR coding region, four of which led to amino acid changes: Pro76Leu, Cys103Ser, Met165Leu, Lys185Glu. Furthermore, 24 mutations, including two nucleotide insertions, were observed in the 211-bp (5MP1) and 213-bp (3MP1) region situated between the tbtR and tbtA genes (data not shown). 3.7. Spontaneous TBT-resistant mutants of 3MP1 exhibit di¡erent antibiotic resistance phenotypes Spontaneous TBT-resistant mutants from the TBT-sensitive 3MP1 strain (MIC 25 mg l31 ) were obtained at a high frequency (2U1037 ), after plating on agar medium containing 50 mg l31 of TBT. MICs of TBT were determined for 100 clones selected at random, and the results indicated that 75% of them gave MICs v 500 mg l31 (M500 mutants) and the remaining divided into 5% with MIC 250 mg l31 (M-250 mutants), 10% with MIC 100 mg l31 and 10% with MIC 50 mg l31 . Only the high-level TBTresistant mutants (i.e. with MICs 250 and v 500 mg l31 ) showed an increased resistance to nalidixic acid (128 mg l31 vs. 8 mg l31 for the 3MP1 strain), chloramphenicol (64 mg l31 vs. 8 mg l31 ) and sulfamethoxazole (512 mg l31 vs. 4 mg l31 ). The analysis of the tbtR gene and the intergenic region tbtR-tbtA from two high-level TBT-resistant mutants (M-500 and M-250) selected at random showed no nucleotide di¡erences with the corresponding sequence from the 3MP1 native strain.
4. Discussion
Fig. 3. SDS^PAGE of outer membrane proteins from TBT-sensitive and -resistant strains. Lane 1, P. stutzeri 3MP1 strain; lane 2, P. stutzeri 5MP1 strain; lane M, SDS standard proteins (Sigma-Aldrich).
TBT resistance in P. stutzeri 5MP1 was found to be associated with the presence of an operon, called tbtABM, as demonstrated by cloning in E. coli, where it led to a more than 20-fold increase in the level of TBT MIC.
FEMSLE 11399 3-3-04
12
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
TbtABM presented extensive similarities with proton-dependent e¥ux pumps belonging to the RND (analogous to TbtB), membrane fusion (analogous to TbtA) and outer membrane (analogous to TbtM) proteins which function together to extrude substrates across membranes of the Gram-negative bacteria [13]. In particular, the TbtB protein possessed the four characteristic motifs (A, B, C and D) which serve to identify new members of this family in bacterial genomes [14]. Based on the high degrees of homology with the TtgDEF and SrpABC systems and given that TBT acts similarly as organic solvents by dissolving into biological membranes [1], we conclude that tbtABM is responsible for the TBT resistance in E. coli due to an active e¥ux of this toxic compound out of the cells. Despite the absence of direct evidence, we also suggest that TbtABM removes TBT out of P. stutzeri 5MP1. Indeed, other RND e¥ux pumps of Pseudomonas sp., such as the MexABOprM system of P. aeruginosa, have been cloned in E. coli where they accommodate a similar range of substrates as in their original host [25]. As shown by PCR and hybridization experiments, tbtABM appeared to be speci¢c to P. stutzeri, although the operon was not systematically present in this species. Indeed, only half of the environmental TBT-resistant P. stutzeri carried tbtABM. P. stutzeri is a species divided into seven genomovars which exhibit great genomic diversity, even within the same genomovar [26]. Thus, the three TBT-resistant strains lacking tbtABM might be phylogenetically distinct from 5MP1 and 3MP1, and harbor other TBT e¥ux pumps not directly related to tbtABM. In the same way, this operon has not been detected in the 10 environmental P. putida tested, despite the presence of closely related RND e¥ux systems since the 1.423-kb internal tbtB fragment used as a probe and the corresponding fragment on ttgE or srpB genes shared 76% of DNA homology [20,21]. TBT, like the majority of substrates of the MDR systems, is a hydrophobic positively charged compound. To verify that the TbtABM system was able to export other structurally di¡erent drugs, antimicrobial agents previously described as MDR pump substrates have been tested. Susceptibility testing of TBT-resistant (e.g. 5MP1) compared with TBT-sensitive (e.g. 3MP1) P. stutzeri strains showed a limited increase in nalidixic acid and chloramphenicol MICs. However, the di¡erences observed cannot be considered signi¢cant since these strains may contain other e¥ux systems which can interfere with the substrate speci¢city of TbtABM, as described for the TtgGHI e¥ux pump in P. putida [23]. To assess the range of substrates of RND e¥ux pumps, structural gene knockout mutants can be selected [21]. Nevertheless, cloning the operon in E. coli provides analogous information. Indeed, as indicated above, e¥ux pumps of Pseudomonas introduced in E. coli retain their substrate speci¢city [25]. MIC comparison between the E. coli host strain with and without the operon revealed that tbtABM expelled
not only TBT, but also antibiotics such as nalidixic acid, chloramphenicol, and sulfamethoxazole. This ability to ef£ux some antibiotics distinguishes TbtABM from the most closely related TtgDEF and SrpABC systems which expel aromatic hydrocarbons, but not antibiotics [21,27]. However, other P. putida e¥ux pumps such as TtgABC (63.4% homology between TbtB and TtgB) a¡ord resistance to organic solvents as well as several antibiotics, including nalidixic acid and chloramphenicol [28]. The increases in antibiotic MICs (four-fold) induced by both TbtABM and SrpABC systems are similar and consistent with the data of the literature on this type of pump [29,30]. As expected, with regard to high homologies with the P. putida e¥ux systems, tbtABM also conferred tolerance to n-hexane, an organic solvent. In P. stutzeri, all strains either TBT-resistant or -sensitive but containing tbtABM were tolerant to this compound, and the TBT MIC in 3MP1 was 20-fold increased in the presence of the organic solvent. These results suggest that n-hexane could act as an inducer of tbtABM or other TBT e¥ux system(s) present in this P. stutzeri strain. Such induction by organic solvents has also been described for the previously cited e¥ux systems of P. putida [20,21]. Both experiments of outer membrane protein analysis and RT-PCR assays tended to indicate that TbtABM was expressed in 5MP1 but not, or to a much lesser degree, in 3MP1. This di¡erence in the level of expression could re£ect TbtABM overproduction in 5MP1, due to an impaired negative regulation. Indeed, the RND e¥ux systems are often controlled by a transcriptional repressor [13]. Actually, a potential regulatory gene of tbtABM, named tbtR, was identi¢ed upstream of the operon. TbtR exhibited many similarities with the previously described repressors of the RND e¥ux systems, i.e. small size and divergent gene orientation, presence of three conserved amino acids (Asn179, Leu187 and Gly193) in the NH2 -terminal region [31], as well as 56% identity with SrpS, one of the regulatory genes of SrpABC [24]. Comparison of the TbtR gene of 3MP1 and 5MP1 showed that of the four substitutions found, two amino acids (Ser103 and Leu165) in 3MP1 were replaced by two residues with SH-grouping (Cys103 and Met165) in 5MP1, and a negatively charged amino acid (Glu185) was changed to a positively charged one (Lys185). These modi¢cations might have modi¢ed the structure of TbtR accounting for its inactivation in 5MP1. In the literature a single substitution in repressor genes has often been reported to induce the overproduction of RND-type systems [31,32]. However, other mutations situated in the intergenic region between the regulatory and the membrane fusion protein genes have also been described as resulting in the overexpression of e¥ux pumps [33]. In agreement with this observation, as many as 24 nucleotide di¡erences have been detected in this area between 5MP1 and 3MP1. Multidrug-resistant mutants with increased expression of RND pumps due to mutations in regulatory systems can
FEMSLE 11399 3-3-04
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
easily be isolated [17,34]. Accordingly, spontaneous TBTresistant mutants from the 3MP1 strain were selected at a frequency of 2U1037 . These mutants harbored di¡erent TBT resistance levels (ranged from 50 to v 500 mg l31 ) associated or not with antibiotic resistances. These results suggest the existence of di¡erent TBT resistance mechanisms and/or di¡erent e¥ux systems in this P. stutzeri strain. Moreover, in two high-level TBT-resistant mutants, the tbtR gene remained unchanged in comparison with the parental strain, suggesting the presence of mutations in other regulatory systems. Similarly, Wery et al. [24] have shown that a cluster of two genes, srpR and srpS, is needed for e¡ecting the repression of srpABC. Furthermore, mutations outside the RND locus, such as in global and cognate regulators, also enable the expression of these systems to be in£uenced [35]. In conclusion, this study describes a new multidrug RND-type operon, tbtABM, in P. stutzeri, a species where such an e¥ux pump has not been reported before. This observation supports the previous suggestion that RND e¥ux pumps are broadly distributed among Pseudomonas species and related organisms [36]. Our data also allowed us to ¢nd another kind of substrate, i.e. TBT, for a member of this RND-type e¥ux pump family.
[10]
[11]
[12] [13] [14]
[15]
[16]
[17]
[18]
[19]
Acknowledgements We thank Laure Coulange and Ce¤cile Frigo for their technical assistance. This work was supported by grants of the Foundation for Medical Research (FI003197/1) and by the French Ministry of Research (Ea525).
[20]
[21]
[22]
References [1] Cooney, J.J. and Wuertz, S. (1989) Toxic e¡ects of tin compounds on microorganisms. J. Ind. Microbiol. 4, 375^402. [2] Alzieu, C., Sanjuan, J., Michel, P., Borel, M. and Dreno, J.P. (1989) Monitoring and assessment of butyltins in Atlantic coastal waters. Mar. Pollut. Bull. 20, 22^26. [3] Laughlin, R.B., Guard, H.E. and Coleman, W.M. (1986) Tributyltin in seawater: speciation and octanol-water partition coe⁄cient. Environ. Sci. Technol. 20, 201^204. [4] Jude, F., Lascourreges, J.F., Capdepuy, M., Quentin, C. and Caumette, P. (1996) Evaluation of tributyltin resistance in marine sediment bacteria. Can. J. Microbiol. 42, 525^532. [5] White, J.S., Tobin, J.M. and Cooney, J.J. (1999) Organotin compounds and their interactions with microorganisms. Can. J. Microbiol. 45, 541^554. [6] Gadd, G.M. (2000) Microbial interactions with tributyltin compounds: detoxi¢cation, accumulation, and environmental fate. Sci. Total Environ. 258, 119^127. [7] Gilmour, C.C., Tuttle, J.H. and Means, J.C. (1987) Anaerobic microbial methylation of inorganic tin in estuarine sediment slurries. Microb. Ecol. 14, 233^242. [8] Hallas, L.E., Means, J.C. and Cooney, J.J. (1982) Methylation of tin by estuarine microorganisms. Science 215, 1505^1507. [9] Wuertz, S., Miller, C.E., P¢ster, R.M. and Cooney, J.J. (1991) Trib-
[23]
[24]
[25]
[26]
[27]
[28]
[29]
13
utyltin-resistant bacteria from estuarine and freshwater sediments. Appl. Environ. Microbiol. 57, 2783^2789. Miller, C.E., Wuertz, S., Cooney, J.J. and P¢ster, R.M. (1995) Plasmids in tributyltin-resistant bacteria from fresh and estuarine waters. J. Ind. Microbiol. 14, 337^342. Fukagawa, T. and Suzuki, S. (1993) Cloning of gene responsible for tributyltin chloride (TBTCl) resistance in TBTCl-resistant marine bacterium, Alteromonas sp. M-1. Biochem. Biophys. Res. Commun. 194, 733^740. Suzuki, S. and Fukagawa, T. (1995) Tributyltin-resistant marine bacteria: a summary of recent work. J. Ind. Microbiol. 14, 154^158. Paulsen, I.T., Brown, M.H. and Skurray, R.A. (1996) Proton-dependent multidrug e¥ux systems. Microbiol. Rev. 60, 575^608. Putman, M., van Veen, H.W. and Konings, W.N. (2000) Molecular properties of bacterial multidrug transporters. Microbiol. Mol. Biol. Rev. 64, 672^693. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. White, D.G., Goldman, J.D., Demple, B. and Levy, S.B. (1997) Role of the acrAB locus in organic solvent tolerance mediated by expression of marA, soxS, or robA in Escherichia coli. J. Bacteriol. 179, 6122^6126. Masuda, N., Sakagawa, E. and Ohya, S. (1995) Outer membrane proteins responsible for multiple drug resistance in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 39, 645^649. Poole, K., Heinrichs, D.E. and Neshat, S. (1993) Cloning and sequence analysis of an EnvCD homologue in Pseudomonas aeruginosa : regulation by iron and possible involvement in the secretion of the siderophore pyoverdine. Mol. Microbiol. 10, 529^544. Ma, D., Cook, D.N., Alberti, M., Pon, N.G., Nikaido, H. and Hearst, J.E. (1993) Molecular cloning and characterization of acrA and acrE genes of Escherichia coli. J. Bacteriol. 175, 6299^6313. Kieboom, J., Dennis, J.J., de Bont, J.A. and Zylstra, G.J. (1998) Identi¢cation and molecular characterization of an e¥ux pump involved in Pseudomonas putida S12 solvent tolerance. J. Biol. Chem. 273, 85^91. Mosqueda, G. and Ramos, J.L. (2000) A set of genes encoding a second toluene e¥ux system in Pseudomonas putida DOT-T1E is linked to the tod genes for toluene metabolism. J. Bacteriol. 182, 937^943. Johnson, J.M. and Church, G.M. (1999) Alignment and structure prediction of divergent protein families: periplasmic and outer membrane proteins of bacterial e¥ux pumps. J. Mol. Biol. 287, 695^715. Rojas, A., Duque, E., Mosqueda, G., Golden, G., Hurtado, A., Ramos, J.L. and Segura, A. (2001) Three e¥ux pumps are required to provide e⁄cient tolerance to toluene in Pseudomonas putida DOTT1E. J. Bacteriol. 183, 3967^3973. Wery, J., Hidayat, B., Kieboom, J. and de Bont, J.A. (2001) An insertion sequence prepares Pseudomonas putida S12 for severe solvent stress. J. Biol. Chem. 276, 5700^5706. Srikumar, R., Kon, T., Gotoh, N. and Poole, K. (1998) Expression of Pseudomonas aeruginosa multidrug e¥ux pumps MexA-MexB-OprM and MexC-MexD-OprJ in a multidrug-sensitive Escherichia coli strain. Antimicrob. Agents Chemother. 42, 65^71. Ginard, M., Lalucat, J., Tummler, B. and Romling, U. (1997) Genome organization of Pseudomonas stutzeri and resulting taxonomic and evolutionary considerations. Int. J. Syst. Bacteriol. 47, 132^143. Isken, S. and De Bont, J.A. (2000) The solvent e¥ux system of Pseudomonas putida S12 is not involved in antibiotic resistance. Appl. Microbiol. Biotechnol. 54, 711^714. Ramos, J.L., Duque, E., Godoy, P. and Segura, A. (1998) E¥ux pumps involved in toluene tolerance in Pseudomonas putida DOTT1E. J. Bacteriol. 180, 3323^3329. Li, X.Z., Nikaido, H. and Poole, K. (1995) Role of MexA-MexBOprM in antibiotic e¥ux in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 39, 1948^1953.
FEMSLE 11399 3-3-04
14
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
[30] Fukumori, F., Hirayama, H., Takami, H., Inoue, A. and Horikoshi, K. (1998) Isolation and transposon mutagenesis of a Pseudomonas putida KT2442 toluene-resistant variant: involvement of an e¥ux system in solvent resistance. Extremophiles 2, 395^400. [31] Poole, K., Tetro, K., Zhao, Q., Neshat, S., Heinrichs, D.E. and Bianco, N. (1996) Expression of the multidrug resistance operon mexA-mexB-oprM in Pseudomonas aeruginosa : mexR encodes a regulator of operon expression. Antimicrob. Agents Chemother. 40, 2021^2028. [32] Srikumar, R., Paul, C.J. and Poole, K. (2000) In£uence of mutations in the mexR repressor gene on expression of the MexA-MexB-OprM multidrug e¥ux system of Pseudomonas aeruginosa. J. Bacteriol. 182, 1410^1414. [33] Evans, K., Adewoye, L. and Poole, K. (2001) MexR repressor of the
mexAB-oprM multidrug e¥ux operon of Pseudomonas aeruginosa : identi¢cation of MexR binding sites in the mexA-mexR intergenic region. J. Bacteriol. 183, 807^812. [34] Ziha-Zari¢, I., Llanes, C., Kohler, T., Pechere, J.C. and Plesiat, P. (1999) In vivo emergence of multidrug-resistant mutants of Pseudomonas aeruginosa overexpressing the active e¥ux system MexAMexB-OprM. Antimicrob. Agents Chemother. 43, 287^291. [35] Duque, E., Segura, A., Mosqueda, G. and Ramos, J.L. (2001) Global and cognate regulators control the expression of the organic solvent e¥ux pumps TtgABC and TtgDEF of Pseudomonas putida. Mol. Microbiol. 39, 1100^1106. [36] Bianco, N., Neshat, S. and Poole, K. (1997) Conservation of the multidrug resistance e¥ux gene oprM in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 41, 853^856.
FEMSLE 11399 3-3-04
www.fems-microbiology.org
TbtABM, a multidrug e¥ux pump associated with tributyltin resistance in Pseudomonas stutzeri Florence Jude a
a;b
, Corinne Arpin a; , Ce¤line Brachet-Castang a , Miche'le Capdepuy Pierre Caumette c , Claudine Quentin a
a;b
,
Laboratoire de Microbiologie, Universite¤ Victor Segalen Bordeaux 2, 146 rue Le¤o Saignat, 33076 Bordeaux Cedex, France b Laboratoire d’Oce¤anographie Biologique, Universite¤ de Bordeaux 1, 33120 Arcachon, France c Laboratoire d’Ecologie Mole¤culaire, Universite¤ de Pau, 64000 Pau, France Received 25 November 2003; received in revised form 19 December 2003; accepted 21 December 2003 First published online 30 January 2004
Abstract Tributyltin (TBT) is a toxic agent used in marine antifouling paints. Among the bacterial flora of a polluted harbor, TBT-resistant strains of Pseudomonas stutzeri have been isolated. In the strain 5MP1 (TBT minimum inhibitory concentration (MIC) v 1000 mg l31 ), TBT resistance was found to be associated with the presence of the operon tbtABM, homologous to the resistance^nodulation^cell division (RND) efflux pump family, as demonstrated by cloning in Escherichia coli. TbtABM exhibited the greatest homology (60.9^ 84.9%) with the TtgDEF and SrpABC systems, both involved in aromatic compound tolerance in P. putida. TbtABM conferred multidrug resistance (MDR) including to n-hexane, nalidixic acid, chloramphenicol, and sulfamethoxazole (antibiotic MICsU4 for the E. coli host strain carrying the operon). By polymerase chain reaction amplification and hybridization experiments, the presence of tbtABM was detected in the TBT-sensitive P. stutzeri 3MP1 (TBT MIC 25 mg l31 ). However, the latter strain did not seem to express TbtABM. This is the first description of a MDR efflux pump in P. stutzeri, and of a new kind of substrate, TBT, for the RND family of transporters. = 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords : Tributyltin resistance; Pseudomonas stutzeri ; E¥ux pump; RND family transporter; Multidrug resistance
1. Introduction Tributyltin (TBT) is an organostannic compound, used industrially as a stabilizer in plastics, wood preservative, and antifouling agent in boat paints [1]. However, in the latter case, TBT has induced high larval malformations of oysters [2]. Actually, TBT is toxic for both eukaryotes and prokaryotes due to a partition coe⁄cient in octanol^water mixture (log Pow at 25‡C) of 3.7, allowing its dissolution into biological membranes, disturbing their integrity, and ultimately compromising their physiological functions [1,3]. Although the use of TBT has been restricted for 15 years in many countries, this compound persists in freshwater and marine sediments and remains a potential source of pollution [4,5]. Bacterial populations might help to detoxify TBT. In-
* Corresponding author. Tel. : +33 (5) 57 57 10 75; Fax : +33 (5) 56 90 90 72. E-mail address : [email protected] (C. Arpin).
deed, pure bacterial cultures as well as natural aquatic sediments are able to biomethylate the inorganic SnCl2 [5^8]. They might also perform the reverse reaction on another tin compound, i.e. demethylation of TBT that leads to less toxic derivatives. Thus, TBT-resistant strains might serve as cleaning agents, justifying their screening in polluted environments. However, enzymic modi¢cations may not be involved. Indeed, very few studies have investigated the mechanism of TBT bacterial resistance. Wuertz et al. [9] have described TBT-resistant bacteria that contained plasmids and had multiple heavy metal and antibiotic resistances. Miller et al. [10] have reported TBT together with chromium resistance, mediated by a conjugative plasmid in Pseudomonas aeruginosa. Finally, Fukugawa and Suzuki have characterized a chromosomal gene involved in TBT resistance in a strain of Alteromonas sp., encoding a protein related to Naþ /Hþ antiporters and various Ca2þ transport systems [11] and/or glycosylase [12]. The harbor of Arcachon (France) contains relatively high TBT concentrations, reaching 150 ng g31 of sediment
0378-1097 / 04 / $22.00 = 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/S0378-1097(04)00012-6
FEMSLE 11399 3-3-04
8
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
[4]. In a previous work, 103 bacterial strains, including 20 Pseudomonas (eight P. stutzeri, two P. £uorescens and 10 P. putida), have been isolated from this polluted environment, and their TBT susceptibilities have been evaluated [4]. Six P. stutzeri were found to have a high level of TBT resistance with minimum inhibitory concentrations (MICs) of 500 to v 1000 mg l31 compared with two sensitive P. stutzeri strains exhibiting a MICTBT of 25 mg l31 . The TBT-resistant strains did not carry any plasmid and the TBT resistance was not transferable by conjugation, suggesting that the genetic determinant was located on the chromosomal DNA [4]. The aim of this study was to characterize the TBT resistance mechanism(s) in P. stutzeri. After constructing a DNA library in Escherichia coli, and sequencing a subcloned insert conferring TBT resistance, the analysis of the deduced amino acid sequence showed that the involved proteins, named TbtA-TbtB-TbtM (Tbt, for tributyltin), were homologous to the components of the multidrug resistance (MDR) e¥ux pumps belonging to the resistance^ nodulation^division (RND) family [13,14].
2. Materials and methods 2.1. Bacterial strains, plasmids and growth conditions P. stutzeri strains [4] and E. coli HB101 (supE44 3 hsdS20(r3 B mB ) recA13 ara-14 proA2 lacY1 galK12 rpsL20 xyl-5 mtl-1) (Gibco-Bethesda Research Laboratories) used in this study are listed in Table 1. The other Pseudomonas sp. strains analyzed by polymerase chain reaction (PCR) and hybridization experiments have been described elsewhere [4]. Bacterial strains were grown with shaking (200 strokes min31 ) at 30‡C (P. stutzeri) or 37‡C (E. coli) in Mueller^Hinton medium (MH, BioMe¤rieux) or in Luria broth (Gibco-BRL). Selecting agents for resident plasmids were added to growth media when appropriate : ticarcillin (P. stutzeri) or ampicillin (E. coli) at 200 mg l31 , and tributyltin chloride (TBTCl, Sigma-Aldrich) from 10 to 1000 mg l31 .
2.2. DNA extraction and analysis Total DNA was prepared by a previously described method [15]. Midi-preparations of plasmid DNA were performed using a Qiagen puri¢cation kit, whereas mini-preparations were achieved by an alkaline lysis procedure [15]. Restriction digests were carried out according to the supplier’s instructions (Gibco BRL) and analyzed by electrophoresis on agarose gels in Tris-borate bu¡er [15]. 2.3. Cloning experiments and transformation of E. coli and P. stutzeri strains After ligation using a T4 DNA ligase (Gibco BRL) between 1 Wg of HindIII alkaline phosphatase-treated plasmid pUC18 (Gibco-BRL) and 2 Wg of HindIII genomic DNA from P. stutzeri 5MP1, the mixture was used to transform the E. coli HB101 strain by electroporation, as indicated by the manufacturer (Bio-Rad). Screening of TBT-resistant clones required a two-step selection: ¢rstly, colonies were collected on MH agar plates with ampicillin; then, ampicillin-resistant clones were replicated on MH agar plates containing both ampicillin and TBTCl. One TBT-resistant clone was selected that carried a recombinant plasmid, pTBT10, containing a 10.6-kb insert. After subcloning using the endonucleases XbaI and MunI, the 8.0-kb generated fragment was puri¢ed from agarose gel using the protocol of the QIAEXII kit (Qiagen), and its ends were completed by the Klenow enzyme (Gibco BRL) before ligation into the SmaI-linearized pUC18 vector, and transformation experiment. 2.4. DNA sequencing and analysis Automated £uorescent sequencing was done using the ABI Prism Dye Terminator Cycle Sequencing Ready Reaction kit and the Applied Biosystem model 377 (Biosystems Division, Perkin Elmer). Overlapping sequences from both strands were obtained using a series of custom-synthesized primers (Eurogentec). Nucleotide and deduced amino acid sequences were analyzed using the PC Gene
Table 1 Antimicrobial susceptibilities of TBT-sensitive and -resistant strains used in this study Strain
Presence of tbtABMb
P. P. P. P. E. E.
+ + + 3 3 +
MIC (mg l31 ) of: TBT
stutzeri 3MP1 and 3MP2 stutzeri 5MP1 stutzeri 3FP1 and 3FP2 stutzeri ATCC 17588 coli HB101 coli HB101+pTBT10
25 v 1000 500 750 50 v 1000
a
Growth on n-hexanec
NAL
CHL
SMX
8 16 16 8 1 4
8 32 16 16 2 8
4 4 4 4 0.05 0.2
a
NAL, nalidixic acid ; CHL, chloramphenicol ; SMX, sulfamethoxazole. Presence (+) or absence (3) of the tbtABM operon, as detected by PCR and hybridization experiments. c +, tolerance is assessed by a visible lawn of growth; or 3, sensitive strains produced no visible lawn. b
FEMSLE 11399 3-3-04
+ + + 3 3 +
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
Software Package (Intelligenetics) and the BLAST program available from the National Institute for Biotechnology Information server (http://www.blast.genome.ad.jp). 2.5. Antimicrobial susceptibility testing The MICs of 15 antimicrobial agents including 11 antibiotics (nalidixic acid, cipro£oxacin, tetracycline, doxycycline, chloramphenicol, sulfamethoxazole, erythromycin, rifampicin, fusidic acid, vancomycin, ticarcillin, ethidium bromide, sodium dodecyl sulfate (SDS), crystal violet and acri£avine) were determined by the standard serial twofold dilution technique in MH agar with a ¢nal inoculum of about 104 CFU per spot (http://www.sfm.asso.fr). Tolerance to organic solvents (cyclohexane, n-hexane, p-xylene, toluene) and alcohol (butanol) was measured by plating a 5-Wl aliquot of a bacterial suspension (107 cells ml31 ) on MH agar [16]. After drying, the organic solvent or alcohol was overlaid to a depth of 2^3 mm. Then, the plates were sealed and incubated overnight.
9
separated by electrophoresis on a 0.8% (w/v) agarose gel before being transferred to a nylon N membrane (Nytran, Schleicher and Schuell). The Southern blot hybridizations were performed using as a probe a digoxigenin (DIG)labelled 1.423-kb fragment, which consisted of the tbtB3tbtB4 amplicon (see above). Stringent hybridizations at 42‡C in the presence of 50% formamide and subsequent developments of the blot were carried out using the DIG DNA kit according to the supplier’s instructions (Boehringer Mannheim). 2.8. Outer membrane protein analysis The outer membrane proteins from P. stutzeri and E. coli strains were prepared by a di¡erential 1% sarcosyl solubilization of isolated cell envelopes [17]. An identical quantity of proteins per sample (35 Wg) was separated by SDS^polyacrylamide gel electrophoresis (PAGE) (10% w/v acrylamide and 0.27% w/v N,NP-methylene bisacrylamide), before being stained with a Coomassie brilliant blue solution [15].
2.6. PCR ampli¢cation 2.9. Reverse transcription (RT) PCR experiments An internal sequence of tbtB was ampli¢ed using primers tbtB3 (5P-CGCCGGCGCGTTATCGCTGG-3P, starting 60 bp downstream of the ATG initiation codon of the tbtB open reading frame (ORF)) and tbtB4 (5P-GGTGGCGCACAGCGCCGGGG-3P, starting 1.423 kb downstream of the tbtB3 primer) (Fig. 1). The tbtR coding region and the intergenic tbtR-tbtA region were ampli¢ed with oligonucleotides tbtR4 (5P-GATGCCACAGACTTTTAAC-3P, starting 711 bp upstream of the end of tbtR) and tbtA1 (5P-CAGCGGAACGGAATGCCATGC-3P, beginning 45 bp downstream of the GTG initiation codon of tbtA) (Fig. 1).
RNA from 1 ml late culture was prepared with the SV Total RNA Isolation kit that includes a DNase treatment, according to the supplier’s instructions (Promega). The RT-PCR analysis was undertaken with and without RT enzyme (as control) in a single tube RT-PCR system, using the AccessQuick0 RT-PCR kit (Promega). RT-PCR experiments were carried out with the oligonucleotides tbtBD (5P-GGCCATCGGCTTGCTGGTGG-3P) and tbtB5 (5P-CAGCGCCACCAGCACCGAGAG-3P), starting at nucleotides 3371 and 3629 of the tbtB gene, according to the numbering of the sequence deposited in GenBank.
2.7. DNA hybridization experiments 2.10. Accession number Chromosomal DNA (1 Wg) of bacterial strains was digested with HindIII, and the generated fragments were
The DNA sequences of the genes described in this study have been deposited in the GenBank1 sequence database under accession number AF041464.
3. Results 3.1. TBT resistance is associated with the tbtABM operon
Fig. 1. Partial restriction map of the HindIII insert carrying the tbtABM operon. The translation orientation of ORFs is indicated. The position and orientation of primers used for PCR and RT-PCR ampli¢cations are indicated.
After construction of an HindIII DNA library in E. coli HB101, a single clone of the 1000 replicated was resistant to both ampicillin and TBT (MICTBT v 1000 mg l31 vs. 50 mg l31 for the host strain). This clone contained the plasmid pTBT10, consisting of the pUC18 vector with a 10.6-kb insert. The partial restriction map of this fragment is shown in Fig. 1. To con¢rm that TBT-resistant phenotype was due to the presence of this recombinant plasmid,
FEMSLE 11399 3-3-04
10
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
pTBT10 was transferred by electroporation to a new set of competent cells of E. coli HB101. After the two-step selection, 100% of the replicated ampicillin-resistant clones were found to be TBT-resistant, in contrast with none of the clones harboring the native plasmid pUC18. After subcloning, an insert of 8.0 kb obtained by XbaI-MunI double digestion still conveyed TBT resistance. A part of the XbaI-NunI fragment was sequenced on both strands and the analysis of the resulting 7.311-kb sequence revealed the presence of three successive ORFs, called tbtA (1146 bases), tbtB (3138 bases) and tbtM (1428 bases) in the same orientation, as indicated in Fig. 1. All three ORFs exhibited a putative ribosome binding site sequence -GAGG- (tbtA and tbtB) or -AGGA- (tbtM), 8^12 nucleotides upstream of their GTG (tbtA) or ATG (tbtB and tbtM) initiation codon. These three genes potentially formed an operon since a stretch of only 14 bp bridged the ¢rst and the second ORFs, and the second gene stop codon overlapped the ATG of the third gene. 3.2. TbtABM is a member of the e¥ux pump RND family Scan of the GenBank sequence databases revealed that TbtA-TbtB-TbtM showed striking resemblances with proteins known to be involved in proton-dependent multidrug e¥ux pumps, such as the MexA-MexB-OprM pump (56.764.3-62.6%, respectively with TbtA-TbtB-TbtM) in P. aeruginosa [18], and AcrA-AcrB-TolC (49.6-58.2-15.8%, respectively) in E. coli [19]. Nevertheless, the highest degrees of homology were obtained with TtgD-TtgE-TtgF (72.5-84.9-65.8%, respectively) and SrpA-SrpB-SrpC (73.3-84.4-60.9%, respectively) which are e¥ux systems conferring organic solvent tolerance in P. putida [20,21]. The hydropathy pro¢les of TbtA (41.6 kDa) and TbtM (52.4 kDa) were consistent with their putative functions of fusion membrane protein and outer membrane protein, respectively [22]. Hydropathy analysis of the putative transporter TbtB (113.1 kDa) showed that it comprised 12 potential transmembrane segments (TMS) and two large loops, between TMS1 and 2 and between TMS7 and 8. TbtB possessed four highly conserved sequences, motif A (residues spanning positions 86^131), motif B (positions 445^494), motif C (amino acids 966^1010) and motif D (residues 389^424) which exhibited 93, 72, 75 and 72% homology, respectively, with previously described consensus regions of the RND transporters [13,14]. 3.3. tbtABM is speci¢c to P. stutzeri, but irregularly found in this species The presence of tbtABM was searched by PCR using primers tbtB3 and tbtB4 (Fig. 1). Among the 20 Pseudomonas sp. strains (eight P. stutzeri, two P. £uorescens and 10 P. putida) previously isolated from the TBT-polluted sediments of Arcachon harbor [4], and two collection strains, E. coli HB101 and P. stutzeri ATCC 17588, a
Fig. 2. Southern blot hybridization of the HindIII-restricted genomic DNA. Only strains for which DNA gave a positive hybridization are shown in this ¢gure. The probe used for the hybridization experiments was a 1.423-kb internal tbtB fragment. Lanes 1 and 2, TBT-sensitive P. stutzeri 3MP1 and 3MP2, respectively. Lanes 3^5, TBT-resistant P. stutzeri 3FP1, 3FP2 and 5MP1, respectively.
1.423-kb PCR product was obtained in only three of the six TBT-resistant P. stutzeri (5MP1, 3FP1, and 3FP2), and in the two TBT-sensitive P. stutzeri (3MP1 and 3MP2) (data not shown). The EcoRI restriction pro¢les of the amplicon con¢rmed the presence of the tbtB sequence, since two fragments of the expected sizes (760 and 663 bp) were obtained. Reproducible hybridization experiments con¢rmed the previous PCR results (Fig. 2). Indeed, the ¢ve P. stutzeri strains (5MP1, 3FP1, 3FP2, 3MP1 and 3MP2) carried tbtABM on a HindIII fragment of ca. 11.0 kb. 3.4. TbtABM is a multidrug e¥ux pump in E. coli P. stutzeri strains did not exhibit signi¢cant di¡erences in susceptibility to the 15 tested agents, except for nalidixic acid and chloramphenicol. MICs of these antimicrobials were slightly increased (two- to four-fold) in the three TBT-resistant strains harboring the tbtABM operon (5MP1, 3FP1 and, 3FP2) compared with the two TBTsensitive strains (3MP1 and 3MP2) (Table 1). More convincingly, the presence of pTBT10 in E. coli HB101 led to a similar increase in MICs of nalidixic acid and chloramphenicol, but also of sulfamethoxazole (Table 1). The MIC increases were low (four-fold), but reproducible through three independent experiments. Furthermore, the recombinant E. coli clone was able to grow in the presence of nhexane, whereas the host strain did not. In addition, ¢ve of eight P. stutzeri strains (5MP1, 3FP1, 3FP2, 3MP1 and 3MP2) produced a visible lawn of growth on a medium containing n-hexane (Table 1), and the MIC of TBT in 3MP1 was increased from 25 to 500 mg l31 in the presence of this organic solvent. The other tested solvents or alcohol with a higher toxicity (log Pow lower than 3.5) gave either negative or non-reproducible results by the method used.
FEMSLE 11399 3-3-04
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
11
3.5. Di¡erential TbtABM expression between the P. stutzeri strains The content of the outer membrane proteins of the TBT-sensitive 3MP1 and TBT-resistant 5MP1 strains of P. stutzeri was compared. The results, shown in Fig. 3, indicated the presence of an additional protein in 5MP1 (lane 2), which was absent in 3MP1 (lane 1); the size of 52 kDa of this protein was consistent with the theoretical molecular mass of the outer membrane protein, TbtM (52.4 kDa). In addition, RT-PCR analysis (Fig. 4), performed with and without RT enzyme (as control for DNA contamination) and using the tbtBD and tbtB5 primers, revealed a 259-bp RT fragment in the TBT-resistant strain (lane 4) although no ampli¢cations were observed with the sensitive strain (lane 2). 3.6. TbtR is a potential negative regulator Sequencing of the 1008-bp region located upstream of tbtA revealed that the locus contained an additional ORF named tbtR, of small size (777 bp) and transcribed divergently from tbtABM (Fig. 1). The deduced TbtR sequence (27.7 kDa) showed no signi¢cant identities with those compiled in the GenBank databases, except for TtgV (58%) and SrpS (56%), the putative regulatory proteins of the ttgGHI and srpABC operons of P. putida, respec-
Fig. 4. RT-PCR analysis of TBT-sensitive and -resistant strains. Lanes 1 and 2, P. stutzeri 3MP1 strain; lanes 3 and 4, P. stutzeri 5MP1 strain; lanes 5 and 6, E. coli HB101 strain; lanes 7 and 8, E. coli HB101 strain with the recombinant plasmid pTBT10; lane M, pGEM0 size marker (Promega). Lanes 3 and + were experiments performed without and with reverse transcriptase, respectively.
tively [23,24]. Using the primers tbtR4 and tbtA1 (Fig. 1), the region upstream of tbtABM in 3MP1 was ampli¢ed and sequenced. Sequence comparison between 5MP1 and 3MP1 revealed 14 substitutions in the tbtR coding region, four of which led to amino acid changes: Pro76Leu, Cys103Ser, Met165Leu, Lys185Glu. Furthermore, 24 mutations, including two nucleotide insertions, were observed in the 211-bp (5MP1) and 213-bp (3MP1) region situated between the tbtR and tbtA genes (data not shown). 3.7. Spontaneous TBT-resistant mutants of 3MP1 exhibit di¡erent antibiotic resistance phenotypes Spontaneous TBT-resistant mutants from the TBT-sensitive 3MP1 strain (MIC 25 mg l31 ) were obtained at a high frequency (2U1037 ), after plating on agar medium containing 50 mg l31 of TBT. MICs of TBT were determined for 100 clones selected at random, and the results indicated that 75% of them gave MICs v 500 mg l31 (M500 mutants) and the remaining divided into 5% with MIC 250 mg l31 (M-250 mutants), 10% with MIC 100 mg l31 and 10% with MIC 50 mg l31 . Only the high-level TBTresistant mutants (i.e. with MICs 250 and v 500 mg l31 ) showed an increased resistance to nalidixic acid (128 mg l31 vs. 8 mg l31 for the 3MP1 strain), chloramphenicol (64 mg l31 vs. 8 mg l31 ) and sulfamethoxazole (512 mg l31 vs. 4 mg l31 ). The analysis of the tbtR gene and the intergenic region tbtR-tbtA from two high-level TBT-resistant mutants (M-500 and M-250) selected at random showed no nucleotide di¡erences with the corresponding sequence from the 3MP1 native strain.
4. Discussion
Fig. 3. SDS^PAGE of outer membrane proteins from TBT-sensitive and -resistant strains. Lane 1, P. stutzeri 3MP1 strain; lane 2, P. stutzeri 5MP1 strain; lane M, SDS standard proteins (Sigma-Aldrich).
TBT resistance in P. stutzeri 5MP1 was found to be associated with the presence of an operon, called tbtABM, as demonstrated by cloning in E. coli, where it led to a more than 20-fold increase in the level of TBT MIC.
FEMSLE 11399 3-3-04
12
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
TbtABM presented extensive similarities with proton-dependent e¥ux pumps belonging to the RND (analogous to TbtB), membrane fusion (analogous to TbtA) and outer membrane (analogous to TbtM) proteins which function together to extrude substrates across membranes of the Gram-negative bacteria [13]. In particular, the TbtB protein possessed the four characteristic motifs (A, B, C and D) which serve to identify new members of this family in bacterial genomes [14]. Based on the high degrees of homology with the TtgDEF and SrpABC systems and given that TBT acts similarly as organic solvents by dissolving into biological membranes [1], we conclude that tbtABM is responsible for the TBT resistance in E. coli due to an active e¥ux of this toxic compound out of the cells. Despite the absence of direct evidence, we also suggest that TbtABM removes TBT out of P. stutzeri 5MP1. Indeed, other RND e¥ux pumps of Pseudomonas sp., such as the MexABOprM system of P. aeruginosa, have been cloned in E. coli where they accommodate a similar range of substrates as in their original host [25]. As shown by PCR and hybridization experiments, tbtABM appeared to be speci¢c to P. stutzeri, although the operon was not systematically present in this species. Indeed, only half of the environmental TBT-resistant P. stutzeri carried tbtABM. P. stutzeri is a species divided into seven genomovars which exhibit great genomic diversity, even within the same genomovar [26]. Thus, the three TBT-resistant strains lacking tbtABM might be phylogenetically distinct from 5MP1 and 3MP1, and harbor other TBT e¥ux pumps not directly related to tbtABM. In the same way, this operon has not been detected in the 10 environmental P. putida tested, despite the presence of closely related RND e¥ux systems since the 1.423-kb internal tbtB fragment used as a probe and the corresponding fragment on ttgE or srpB genes shared 76% of DNA homology [20,21]. TBT, like the majority of substrates of the MDR systems, is a hydrophobic positively charged compound. To verify that the TbtABM system was able to export other structurally di¡erent drugs, antimicrobial agents previously described as MDR pump substrates have been tested. Susceptibility testing of TBT-resistant (e.g. 5MP1) compared with TBT-sensitive (e.g. 3MP1) P. stutzeri strains showed a limited increase in nalidixic acid and chloramphenicol MICs. However, the di¡erences observed cannot be considered signi¢cant since these strains may contain other e¥ux systems which can interfere with the substrate speci¢city of TbtABM, as described for the TtgGHI e¥ux pump in P. putida [23]. To assess the range of substrates of RND e¥ux pumps, structural gene knockout mutants can be selected [21]. Nevertheless, cloning the operon in E. coli provides analogous information. Indeed, as indicated above, e¥ux pumps of Pseudomonas introduced in E. coli retain their substrate speci¢city [25]. MIC comparison between the E. coli host strain with and without the operon revealed that tbtABM expelled
not only TBT, but also antibiotics such as nalidixic acid, chloramphenicol, and sulfamethoxazole. This ability to ef£ux some antibiotics distinguishes TbtABM from the most closely related TtgDEF and SrpABC systems which expel aromatic hydrocarbons, but not antibiotics [21,27]. However, other P. putida e¥ux pumps such as TtgABC (63.4% homology between TbtB and TtgB) a¡ord resistance to organic solvents as well as several antibiotics, including nalidixic acid and chloramphenicol [28]. The increases in antibiotic MICs (four-fold) induced by both TbtABM and SrpABC systems are similar and consistent with the data of the literature on this type of pump [29,30]. As expected, with regard to high homologies with the P. putida e¥ux systems, tbtABM also conferred tolerance to n-hexane, an organic solvent. In P. stutzeri, all strains either TBT-resistant or -sensitive but containing tbtABM were tolerant to this compound, and the TBT MIC in 3MP1 was 20-fold increased in the presence of the organic solvent. These results suggest that n-hexane could act as an inducer of tbtABM or other TBT e¥ux system(s) present in this P. stutzeri strain. Such induction by organic solvents has also been described for the previously cited e¥ux systems of P. putida [20,21]. Both experiments of outer membrane protein analysis and RT-PCR assays tended to indicate that TbtABM was expressed in 5MP1 but not, or to a much lesser degree, in 3MP1. This di¡erence in the level of expression could re£ect TbtABM overproduction in 5MP1, due to an impaired negative regulation. Indeed, the RND e¥ux systems are often controlled by a transcriptional repressor [13]. Actually, a potential regulatory gene of tbtABM, named tbtR, was identi¢ed upstream of the operon. TbtR exhibited many similarities with the previously described repressors of the RND e¥ux systems, i.e. small size and divergent gene orientation, presence of three conserved amino acids (Asn179, Leu187 and Gly193) in the NH2 -terminal region [31], as well as 56% identity with SrpS, one of the regulatory genes of SrpABC [24]. Comparison of the TbtR gene of 3MP1 and 5MP1 showed that of the four substitutions found, two amino acids (Ser103 and Leu165) in 3MP1 were replaced by two residues with SH-grouping (Cys103 and Met165) in 5MP1, and a negatively charged amino acid (Glu185) was changed to a positively charged one (Lys185). These modi¢cations might have modi¢ed the structure of TbtR accounting for its inactivation in 5MP1. In the literature a single substitution in repressor genes has often been reported to induce the overproduction of RND-type systems [31,32]. However, other mutations situated in the intergenic region between the regulatory and the membrane fusion protein genes have also been described as resulting in the overexpression of e¥ux pumps [33]. In agreement with this observation, as many as 24 nucleotide di¡erences have been detected in this area between 5MP1 and 3MP1. Multidrug-resistant mutants with increased expression of RND pumps due to mutations in regulatory systems can
FEMSLE 11399 3-3-04
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
easily be isolated [17,34]. Accordingly, spontaneous TBTresistant mutants from the 3MP1 strain were selected at a frequency of 2U1037 . These mutants harbored di¡erent TBT resistance levels (ranged from 50 to v 500 mg l31 ) associated or not with antibiotic resistances. These results suggest the existence of di¡erent TBT resistance mechanisms and/or di¡erent e¥ux systems in this P. stutzeri strain. Moreover, in two high-level TBT-resistant mutants, the tbtR gene remained unchanged in comparison with the parental strain, suggesting the presence of mutations in other regulatory systems. Similarly, Wery et al. [24] have shown that a cluster of two genes, srpR and srpS, is needed for e¡ecting the repression of srpABC. Furthermore, mutations outside the RND locus, such as in global and cognate regulators, also enable the expression of these systems to be in£uenced [35]. In conclusion, this study describes a new multidrug RND-type operon, tbtABM, in P. stutzeri, a species where such an e¥ux pump has not been reported before. This observation supports the previous suggestion that RND e¥ux pumps are broadly distributed among Pseudomonas species and related organisms [36]. Our data also allowed us to ¢nd another kind of substrate, i.e. TBT, for a member of this RND-type e¥ux pump family.
[10]
[11]
[12] [13] [14]
[15]
[16]
[17]
[18]
[19]
Acknowledgements We thank Laure Coulange and Ce¤cile Frigo for their technical assistance. This work was supported by grants of the Foundation for Medical Research (FI003197/1) and by the French Ministry of Research (Ea525).
[20]
[21]
[22]
References [1] Cooney, J.J. and Wuertz, S. (1989) Toxic e¡ects of tin compounds on microorganisms. J. Ind. Microbiol. 4, 375^402. [2] Alzieu, C., Sanjuan, J., Michel, P., Borel, M. and Dreno, J.P. (1989) Monitoring and assessment of butyltins in Atlantic coastal waters. Mar. Pollut. Bull. 20, 22^26. [3] Laughlin, R.B., Guard, H.E. and Coleman, W.M. (1986) Tributyltin in seawater: speciation and octanol-water partition coe⁄cient. Environ. Sci. Technol. 20, 201^204. [4] Jude, F., Lascourreges, J.F., Capdepuy, M., Quentin, C. and Caumette, P. (1996) Evaluation of tributyltin resistance in marine sediment bacteria. Can. J. Microbiol. 42, 525^532. [5] White, J.S., Tobin, J.M. and Cooney, J.J. (1999) Organotin compounds and their interactions with microorganisms. Can. J. Microbiol. 45, 541^554. [6] Gadd, G.M. (2000) Microbial interactions with tributyltin compounds: detoxi¢cation, accumulation, and environmental fate. Sci. Total Environ. 258, 119^127. [7] Gilmour, C.C., Tuttle, J.H. and Means, J.C. (1987) Anaerobic microbial methylation of inorganic tin in estuarine sediment slurries. Microb. Ecol. 14, 233^242. [8] Hallas, L.E., Means, J.C. and Cooney, J.J. (1982) Methylation of tin by estuarine microorganisms. Science 215, 1505^1507. [9] Wuertz, S., Miller, C.E., P¢ster, R.M. and Cooney, J.J. (1991) Trib-
[23]
[24]
[25]
[26]
[27]
[28]
[29]
13
utyltin-resistant bacteria from estuarine and freshwater sediments. Appl. Environ. Microbiol. 57, 2783^2789. Miller, C.E., Wuertz, S., Cooney, J.J. and P¢ster, R.M. (1995) Plasmids in tributyltin-resistant bacteria from fresh and estuarine waters. J. Ind. Microbiol. 14, 337^342. Fukagawa, T. and Suzuki, S. (1993) Cloning of gene responsible for tributyltin chloride (TBTCl) resistance in TBTCl-resistant marine bacterium, Alteromonas sp. M-1. Biochem. Biophys. Res. Commun. 194, 733^740. Suzuki, S. and Fukagawa, T. (1995) Tributyltin-resistant marine bacteria: a summary of recent work. J. Ind. Microbiol. 14, 154^158. Paulsen, I.T., Brown, M.H. and Skurray, R.A. (1996) Proton-dependent multidrug e¥ux systems. Microbiol. Rev. 60, 575^608. Putman, M., van Veen, H.W. and Konings, W.N. (2000) Molecular properties of bacterial multidrug transporters. Microbiol. Mol. Biol. Rev. 64, 672^693. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. White, D.G., Goldman, J.D., Demple, B. and Levy, S.B. (1997) Role of the acrAB locus in organic solvent tolerance mediated by expression of marA, soxS, or robA in Escherichia coli. J. Bacteriol. 179, 6122^6126. Masuda, N., Sakagawa, E. and Ohya, S. (1995) Outer membrane proteins responsible for multiple drug resistance in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 39, 645^649. Poole, K., Heinrichs, D.E. and Neshat, S. (1993) Cloning and sequence analysis of an EnvCD homologue in Pseudomonas aeruginosa : regulation by iron and possible involvement in the secretion of the siderophore pyoverdine. Mol. Microbiol. 10, 529^544. Ma, D., Cook, D.N., Alberti, M., Pon, N.G., Nikaido, H. and Hearst, J.E. (1993) Molecular cloning and characterization of acrA and acrE genes of Escherichia coli. J. Bacteriol. 175, 6299^6313. Kieboom, J., Dennis, J.J., de Bont, J.A. and Zylstra, G.J. (1998) Identi¢cation and molecular characterization of an e¥ux pump involved in Pseudomonas putida S12 solvent tolerance. J. Biol. Chem. 273, 85^91. Mosqueda, G. and Ramos, J.L. (2000) A set of genes encoding a second toluene e¥ux system in Pseudomonas putida DOT-T1E is linked to the tod genes for toluene metabolism. J. Bacteriol. 182, 937^943. Johnson, J.M. and Church, G.M. (1999) Alignment and structure prediction of divergent protein families: periplasmic and outer membrane proteins of bacterial e¥ux pumps. J. Mol. Biol. 287, 695^715. Rojas, A., Duque, E., Mosqueda, G., Golden, G., Hurtado, A., Ramos, J.L. and Segura, A. (2001) Three e¥ux pumps are required to provide e⁄cient tolerance to toluene in Pseudomonas putida DOTT1E. J. Bacteriol. 183, 3967^3973. Wery, J., Hidayat, B., Kieboom, J. and de Bont, J.A. (2001) An insertion sequence prepares Pseudomonas putida S12 for severe solvent stress. J. Biol. Chem. 276, 5700^5706. Srikumar, R., Kon, T., Gotoh, N. and Poole, K. (1998) Expression of Pseudomonas aeruginosa multidrug e¥ux pumps MexA-MexB-OprM and MexC-MexD-OprJ in a multidrug-sensitive Escherichia coli strain. Antimicrob. Agents Chemother. 42, 65^71. Ginard, M., Lalucat, J., Tummler, B. and Romling, U. (1997) Genome organization of Pseudomonas stutzeri and resulting taxonomic and evolutionary considerations. Int. J. Syst. Bacteriol. 47, 132^143. Isken, S. and De Bont, J.A. (2000) The solvent e¥ux system of Pseudomonas putida S12 is not involved in antibiotic resistance. Appl. Microbiol. Biotechnol. 54, 711^714. Ramos, J.L., Duque, E., Godoy, P. and Segura, A. (1998) E¥ux pumps involved in toluene tolerance in Pseudomonas putida DOTT1E. J. Bacteriol. 180, 3323^3329. Li, X.Z., Nikaido, H. and Poole, K. (1995) Role of MexA-MexBOprM in antibiotic e¥ux in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 39, 1948^1953.
FEMSLE 11399 3-3-04
14
F. Jude et al. / FEMS Microbiology Letters 232 (2004) 7^14
[30] Fukumori, F., Hirayama, H., Takami, H., Inoue, A. and Horikoshi, K. (1998) Isolation and transposon mutagenesis of a Pseudomonas putida KT2442 toluene-resistant variant: involvement of an e¥ux system in solvent resistance. Extremophiles 2, 395^400. [31] Poole, K., Tetro, K., Zhao, Q., Neshat, S., Heinrichs, D.E. and Bianco, N. (1996) Expression of the multidrug resistance operon mexA-mexB-oprM in Pseudomonas aeruginosa : mexR encodes a regulator of operon expression. Antimicrob. Agents Chemother. 40, 2021^2028. [32] Srikumar, R., Paul, C.J. and Poole, K. (2000) In£uence of mutations in the mexR repressor gene on expression of the MexA-MexB-OprM multidrug e¥ux system of Pseudomonas aeruginosa. J. Bacteriol. 182, 1410^1414. [33] Evans, K., Adewoye, L. and Poole, K. (2001) MexR repressor of the
mexAB-oprM multidrug e¥ux operon of Pseudomonas aeruginosa : identi¢cation of MexR binding sites in the mexA-mexR intergenic region. J. Bacteriol. 183, 807^812. [34] Ziha-Zari¢, I., Llanes, C., Kohler, T., Pechere, J.C. and Plesiat, P. (1999) In vivo emergence of multidrug-resistant mutants of Pseudomonas aeruginosa overexpressing the active e¥ux system MexAMexB-OprM. Antimicrob. Agents Chemother. 43, 287^291. [35] Duque, E., Segura, A., Mosqueda, G. and Ramos, J.L. (2001) Global and cognate regulators control the expression of the organic solvent e¥ux pumps TtgABC and TtgDEF of Pseudomonas putida. Mol. Microbiol. 39, 1100^1106. [36] Bianco, N., Neshat, S. and Poole, K. (1997) Conservation of the multidrug resistance e¥ux gene oprM in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 41, 853^856.
FEMSLE 11399 3-3-04