An inducible tellurite-resistance operon in Proteus mirabilis

Anna Toptchieva1, Gary Sisson1, Louis J. Bryden1, Diane E. Taylor2 and Paul S. Hoffman1,3

1 Department of Microbiology and Immunology, Dalhousie University, Sir Charles Tupper Medical Building, College Street, Halifax, Nova Scotia, Canada B3H 4H7
2 Department of Medical Microbiology and Immunology, 1-28 Medical Science Building, University of Alberta, Edmonton, Alberta, Canada T6G 2H7
3 Division of Infectious Diseases, Department of Medicine, Dalhousie University, Sir Charles Tupper Medical Building, College Street, Halifax, Nova Scotia, Canada B3H 4H7

Correspondence
Paul S. Hoffman
phoffman{at}tupdean2.med.dal.ca


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Tellurite resistance (Ter) is widespread in nature and it is shown here that the natural resistance of Proteus mirabilis to tellurite is due to a chromosomally located orthologue of plasmid-borne ter genes found in enteric bacteria. The P. mirabilis ter locus (terZABCDE) was identified in a screen of Tn5lacZ-generated mutants of which one contained an insertion in terC. The P. mirabilis terC mutant displayed increased susceptibility to tellurite (Tes) and complementation with terC carried on a multicopy plasmid restored high-level Ter. Primer extension analysis revealed a single transcriptional start site upstream of terZ, but only with RNA harvested from bacteria grown in the presence of tellurite. Northern blotting and reverse transcriptase-PCR (RT-PCR) analyses confirmed that the ter operon was inducible by tellurite and to a lesser extent by oxidative stress inducers such as hydrogen peroxide and methyl viologen (paraquat). Direct and inverted repeat sequences were identified in the ter promoter region as well as motifs upstream of the -35 hexamer that resembled OxyR-binding sequences. Finally, the 390 bp intergenic promoter region located between orf3 and terZ showed no DNA sequence identity with any other published ter sequences, whereas terZABCDE genes exhibited 73–85 % DNA sequence identity. The ter operon was present in all clinical isolates of P. mirabilis and Proteus vulgaris tested and is inferred for Morganella and Providencia spp. based on screening for high level Ter and preliminary PCR analysis. Thus, a chromosomally located inducible tellurite resistance operon appears to be a common feature of the genus Proteus.


Abbreviations: Ter, tellurite resistance

The GenBank accession number for the sequence reported in this paper is AF168355.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Tellurite has a long history as an antimicrobial agent and is often employed in selective media for the isolation of a wide range of pathogens, including Corynebacterium diphtheriae, Shigella sp. and verocytotoxigenic Escherichia coli O157 : H7 (‘hamburger disease’ bacterium) (Rahaman et al., 1986; Shimada et al., 1990; Zadic et al., 1993; Kormutakova, 2000). The oxyanions of tellurium, tellurite (K2TeO3) and tellurate, are highly toxic for most micro-organisms, causing direct oxidation of cellular thiols (Taylor, 1999; Turner et al., 1999) or, following reduction to telluride, is inappropriately incorporated in place of sulfur in amino acids (Garberg et al., 1999; Taylor, 1999). Tellurite can also be reduced to the metal by nitrate reductase (Avazeri et al., 1997) or the terminal oxidases of most Gram-negative bacteria (Trutko et al., 2000; Di Tomaso et al., 2002). Resistant bacteria produce jet-black colonies on solid medium supplemented with K2TeO3 as the result of internal deposition of elemental tellurium (Hill et al., 1993; Whelan et al., 1995). The mechanisms associated with tellurium sequestration by the Ter proteins are not known and their study is complicated by the intrinsic toxicity of these proteins for Escherichia coli (Taylor, 1999; Whelan et al., 1997).

At least five genetically distinct chromosomal and plasmid-borne bacterial tellurite resistance systems have been described (Taylor, 1999; Turner et al., 1999; Taylor et al., 2002). The emergence of several unrelated Ter determinants among a wide range of bacterial species, including human pathogens, suggests that these determinants provide some selective advantage in natural environments, which may be unrelated to the Ter phenotype (Hill et al., 1993; Walter & Taylor, 1992). For example, IncHI2 and IncHII conjugative plasmids carrying the ter locus (terZABCDEF) confer constitutive, high level Ter to E. coli, as well as resistance to bacteriophage and colicins (Taylor, 1999; Whelan et al., 1995; Taylor et al., 2002). Homology searches have found putative ter genes in other bacteria, including Yersinia pestis and Deinococcus radiodurans, and variability in gene complement and number of copies among clinical isolates of E. coli O157 (Taylor, 1999; Taylor et al., 2002).

In this study we have characterized a chromosomal cluster of six genes from the urinary tract pathogen Proteus mirabilis that are highly homologous to members of the ter-gene clusters found on large conjugative IncHI2 plasmids (Jobling & Ritchie, 1987, 1988; Whelan et al., 1995) and in the chromosome of E. coli O157 : H7 strains (Taylor et al., 2002). The transposon insertion into the carboxyl coding sequences of the terC gene resulted in increased susceptibility to K2TeO3 (Tes), which could be reversed (Ter) in trans by a multicopy plasmid carrying terC. In contrast to the plasmid-borne ter loci, the ter operon of P. mirabilis was induced in response to K2TeO3 and to a lesser extent by hydrogen peroxide and superoxide anions, suggesting that the mechanism of activation might involve thiol oxidation. Finally, unlike the well-characterized plasmid-borne ter loci, the inducible ter operon of P. mirabilis is contained in the chromosome, an apparent feature of this genus.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids and media.
The P. mirabilis and E. coli strains, plasmids, cosmids, phagemids and oligonucleotides used in this study are described in Table 1. P. mirabilis, P. vulgaris, Morganella morganii and Providencia sp. clinical isolates were obtained from Victoria General Hospital, Halifax, Nova Scotia, Canada. Plasmid construction and DNA manipulations were carried out with E. coli strains JF626 and LE392 (Silhavy et al., 1984), plasmid vectors pBR322 (Bolivar et al., 1977), pBOC20 (O'Connell et al., 1996), pBluescript (Short et al., 1988) and cosmid pHC79 (Hohn & Collins, 1980). Antibacterial agents were used at the following concentrations: ampicillin, 100 µg ml-1; kanamycin, 50 µg ml-1; K2TeO3, 20–300 µg ml-1; rifampicin, 50 µg ml-1; chloramphenicol, 25 µg ml-1. LB medium (BDH) was used in all experiments, unless otherwise indicated. LB lacking sodium chloride (5 g yeast extract plus 10 g tryptone per litre, 2 % agar) was used as the non-swarming medium.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains, plasmids, bacteriophages and oligonucleotides

 
Transposon mutagenesis.
The mini-Tn5-lacZ1 promoter–probe mini-transposon carried on the suicide vector pUT (De Lorenzo et al., 1990) was used for random mutagenesis in P. mirabilis S2R. The reporter lacks terminator sequences, but contains an endogenous promoter upstream of the Kmr determinant that should minimize polar effects of the insertion on downstream genes (De Lorenzo et al., 1990). The conjugation matings between recipient strain P. mirabilis S2R and the donor strain E. coli SM10 ({lambda}pir) containing the delivery vector with the mini-transposon were performed as described by De Lorenzo et al. (1990). Individual colonies of transconjugants were transferred with sterile toothpicks into 96-well tissue-culture plates with covers containing freezing medium (20 % glycerol in LB) and stored at -70 °C. All mutants were screened for {beta}-galactosidase activity (X-Gal) on LB medium permissive for swarmer cell differentiation. Non-swarming medium, also containing X-Gal, was prepared by leaving out NaCl from LB medium or from Nutrient agar. Swarmer and short cells were harvested and differentiated as described previously (Hoffman & Falkinham, 1981).

DNA manipulations.
Recombinant DNA techniques employed standard protocols (Sambrook et al., 1989). Plasmid DNA was isolated by alkaline extraction (Birnboim & Doly, 1979) and by centrifugation of total DNA in caesium chloride/ethidium bromide gradients (Sambrook et al., 1989). Restriction endonucleases, T4 DNA ligase, RNase, T4 polynucleotide kinase and DNA polymerase I were purchased from Gibco-BRL or New England Biolabs and used according to the manufacturer's directions.

Construction of P. mirabilis genomic libraries.
Genomic libraries from P. mirabilis strains S2R and S2R-6/39 were constructed following partial cleavage of chromosomal DNA with Sau3AI and ligation into cosmid pHC79 as described previously (Hoffman et al., 1989). A library of ~3 kbp EcoRI–PstI fragments of P. mirabilis chromosomal DNA was created in E. coli JF626 following separation of the DNA fragments by agarose gel electrophoresis and ligation into pBluescript. Transformants were selected for appropriate antibiotic resistance on solid medium and colonies were picked to 96-well plates for screening. Libraries were screened for the desired genes by colony Southern blot procedures as described previously (Hoffman et al., 1989).

DNA sequencing and sequence analysis.
DNA sequencing was performed manually with a Sequenase (version 2.0) kit from USB and [35S]dATP from NEN/Life Science Products (Sanger et al., 1977), or automated on a LiCor 4000L sequencer by the Dalhousie University–NRC Institute for Marine Biosciences Joint Laboratory. Sequence assembly and analysis were performed using the Genetics Computer Group suite of software (Devereux et al., 1984). Relatedness between predicted polypeptides was detected with the BLAST (Altschul et al., 1990) and FASTA (Pearson, 1990) programs of the National Center for Biological Information (NCBI), Bethesda, MD, USA, and CLUSTAL W (Thompson et al., 1994) was used for multiple alignments of putative amino acid sequences.

Complementation of the terC mutant.
A XhoI/BamHI DNA fragment containing terC was subcloned from pBSK-HP into pBOC20 and the resulting plasmid (pATC12) was introduced into the terC : : lacZ mutant strain S2R-6/39 by conjugation from E. coli S17-1 {lambda}pir. Exconjugants were selected on non-swarming LB agar supplemented with Km and Cm to maintain pATC12 : terC in trans or following selection on the non-swarming LB agar. Ter was determined by viable count on non-swarming medium containing various concentrations of K2TeO3 (0–350 µg ml-1).

RNA manipulations.
Total RNA was extracted by the hot SDS/acid phenol method (Hoffman et al., 1992) from P. mirabilis strains grown in the presence or absence of K2TeO3 or following treatments with hydrogen peroxide or methyl viologen. Northern blot hybridization assays were optimized as generally described (Kroczek & Siebert, 1990). [32P]dCTP-radiolabelled DNA probes were generated by random priming of restriction fragments obtained from cloned DNA sequences from regions of the ter operon, including terZA, terCD or from a 1·6 kb BamHI fragment of ureR, a gene constitutively expressed in the absence of urea (unpublished data). Primer extension was performed as described previously (Hoffman et al., 1992) by hybridizing a 5'-end-labelled oligonucleotide primer TERZ2R (see Table 1) to P. mirabilis total RNA. End-labelling of TERZ2R was done using T4 polynucleotide kinase with subsequent purification using a Sep-Pack C18 cartridge (Millipore). RT-PCR was also used to monitor mRNA levels of the ter operon. Total RNA was treated with RQ1 DNase (Promega) and then purified by standard protocols. cDNA was generated with M-MLV reverse transcriptase (Gibco-BRL), according to the manufacturer's instructions, in a 20 µl reaction containing 1·5 µg RNA and 1 µg random primers, pd(N)6 (Amersham Pharmacia Biotech). Standard PCR procedures were used to generate amplicons from 1 µl of the RT reaction using primer pairs (see Table 1) designed for the 5' region of terZ, the first gene in the putative operon, and ureR. The minimum number of cycles required for detection of amplicons was determined empirically (generally 22–27 cycles). PCR samples were examined by agarose gel electrophoresis and photographed under UV light using a Fotodyne MP-4 photographic system (Fotodyne). Digital images were then created using an Epson ES-1200C scanner (Seiko Epson) and Adobe PhotoShop 4.0 software. Analysis of the gel bands was done with Gel-Pro Analyser software (Media Cybernetics).

Determination of plating efficiency on K2TeO3 medium following induction of the ter operon.
P. mirabilis strain S2R was grown overnight in LB broth and following adjustment of the OD660 to 0·5, either 50 µg K2TeO3 ml-1 or 100 µM H2O2 was added and the cultures (including a control) were shaken at 37 °C. At 30 and 60 min, samples were diluted and plated on LB agar (-NaCl) with or without 200 µg K2TeO3 ml-1. The plating efficiency was determined for each treatment by comparing the number of c.f.u. ml-1 of the treated or untreated groups relative to the number of c.f.u. obtained on K2TeO3-containing medium. The means and standard deviations for triplicate platings were computed.

{beta}-Galactosidase assays and inhibitory effect of K2TeO3.
P. mirabilis mutant S2R-6/39 containing mini-Tn5lacZ was grown in LB broth and aliquots were removed at intervals and assayed for {beta}-galactosidase activity by the method of Miller (1972). All assays were done in triplicate and the mean and standard deviation were computed. The inhibitory effect of K2TeO3 on {beta}-galactosidase activity was determined by measuring the activity of the terC : lacZ fusion in the absence and presence of K2TeO3 (50–200 µg ml-1). As a control for these experiments, a ureR : lacZ fusion (mini-Tn5) inserted at a locus other than the urease locus in P. mirabilis strain S2UR1 was similarly challenged with K2TeO3. UreR is a positive activator of the urease locus and in the absence of the inducer urea, is expressed constitutively at low level (unpublished data). The percentage inhibition was determined from the mean of three determinations.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Distribution of Ter in the genus Proteus and related genera
Clinical isolates of P. mirabilis are a nuisance in the primary isolation of other pathogens due to their ability to swarm over solid media that often contain antibiotics or other selective agents. In studies aimed at identifying anti-swarm agents, we noted that our laboratory strains were resistant to tellurite. Moreover, these strains could be induced to high-level Ter by prior growth of the bacteria on medium containing tellurite. To more fully assess this trait, the lab strain S2R (wild-type) and additional clinical isolates of P. mirabilis and P. vulgaris were screened for Ter. All tested strains of P. mirabilis (28 of 28) and P. vulgaris (6 of 6) displayed high-level Ter (MIC >250 µg ml-1; criteria developed for analysis of plasmid-borne tellurite resistance genes by Whelan et al., 1995). A wider examination of related organisms showed that, while all clinical isolates of Morganella morganii (16 of 16) exhibited Ter, the trait was more variable in the Providencia group (4 of 6).

Isolation of a mini-Tn5-lacZ insertion in terC
The genetic nature of the Ter phenotype of P. mirabilis was investigated by transposon mutagenesis by screening for {beta}-galactosidase-positive mutants exhibiting a Tes phenotype. One clone (S2R-6/39) exhibited decreased levels of {beta}-galactosidase activity on LB agar (X-Gal) containing tellurite and on LB agar supplemented with X-Gal the concentric rings of swarming bacteria appeared white while the non-swarming vegetative bacteria (central inoculum) appeared blue. Swarmer cells and vegetative cells were harvested from LB agar (Hoffman & Falkinham, 1981) and assayed for {beta}-galactosidase activity (Miller, 1972). Vegetative S2R-6/39 bacteria contained 345±50 Miller units while swarm cells contained 73±48 units of activity, suggesting that the transposon had inserted into a locus that was downregulated during swarmer differentiation. This mutant was more susceptible to K2TeO3 (MIC ~100 µg ml-1) than was the wild-type S2R strain (>300 µg ml-1) and colonies of S2R-6/39 required an additional day to appear on tellurite-containing medium.

The mini-transposon and flanking DNA from this mutant (P. mirabilis S2R-6/39) was cloned into pBR322 (pAT9501) and DNA sequence analysis revealed the transposon to have inserted into the C-terminal coding region of terC, an orthologue of the Ter systems of plasmids R478 (Whelan et al., 1995) and pMER610 (Jobling & Ritchie, 1988). Western blot analysis with antibody against {beta}-galactosidase eliminated the possibility that the insertion produced a TerC-LacZ chimera (data not shown). Since Ter genes are often located on plasmids (Jobling & Ritchie, 1987, 1988; Whelan et al., 1995), we used caesium chloride gradient centrifugation (plasmid R478 as a control) to examine P. mirabilis for IncHI related plasmids. No evidence for plasmid sequences could be demonstrated in these studies (data not presented), suggesting that the ter genes of P. mirabilis were located in the chromosome [confirmed in the unpublished genomic sequence of P. mirabilis strain 210 (P. S. Hoffman, SmithKline Beecham)].

The S2R-6/39 mutant displayed Tes (efficiency of plating at 100 µg ml-1=10-2 compared with wild-type at 300 µg ml-1, efficiency of plating=1), which could be restored to high-level Ter following introduction of terC harboured on pATC12 (data not presented). Interestingly, the complemented mutant displayed higher resistance to K2TeO3 (efficiency of plating=1 at 350 µg ml-1) than the wild-type strain, perhaps resulting from increased expression of terC from a multicopy plasmid. These studies suggest that mutation in terC is responsible for the increased susceptibility of S2R-6/39 to K2TeO3, consistent with findings reported for terC of E. coli O157 (Kormutakova et al., 2000).

Sequence analysis of the ter gene cluster
Fig. 1 displays the physical map of the P. mirabilis Ter locus and for comparison, those from other well characterized Ter systems. The nucleotide sequence of the Ter locus revealed seven ORFs within a 6120 bp region that corresponded to orf3, terZ, -A, -B, -C, -D and -E of the plasmid-borne ter locus (Whelan et al., 1995). Interestingly, no terF homologue was detected within the P. mirabilis genome either by sequence analysis of over 1 kb of chromosomal DNA extending 3' of the terE gene, or by PCR amplification using TEREF-F and TEREF-R primers (Table 1) designed to amplify a putative terEF intergenic region (terF sequences are not present in the P. mirabilis strain 210 genome sequence). Also absent from the P. mirabilis Ter locus was an ORF of unknown function between orf3 and terZ that was found in other Ter systems. The highest degree of relatedness between the P. mirabilis-encoded and plasmid-borne Ter gene clusters was detected for the TerD polypeptides, with 92 and 90 % of 192 aa being identical with the orthologues of R478 and pMER610 plasmids, respectively. The lowest degree of relatedness was detected for the predicted TerA polypeptides with 66 % identity (aa) with both R478- and pMER610-encoded polypeptides. The predicted TerA polypeptide was larger by 44 aa at the N-terminal region than its plasmid-encoded counterparts, which begin at position 45 in the P. mirabilis gene (Whelan et al., 1995). A BLAST search of the TerA N-terminal stretch of 44 aa residues showed high similarity of this polypeptide to TerD proteins, suggesting an intergenic recombination (duplication/deletion) might have occurred between these highly conserved genes.



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 1. Physical map of the Ter locus of P. mirabilis S2R and comparison of the deduced ORFs with those of E. coli O157 (Perna et al., 2001), plasmid R478 (Whelan et al., 1995) and Y. pestis (Parkhill et al., 2001). The site of mini-transposon insertion in the chromosome of the P. mirabilis S2R-6/39 is indicated by the ‘lollipop’. Restriction enzyme sites: B, BamHI; Bg, BglII; C, ClaI; E, EcoRI; Ev, EcoRV; H, HindIII; K, KpnI; P, PstI; S, SacI; X, XmnI.

 
With the exception of the terC gene, all of the other ORF-encoded polypeptides are predicted to function in the bacterial cytoplasm (Klein et al., 1985). The orf3 gene of unknown function is divergently transcribed and located 400 bp upstream of terZ (Fig. 1). Other ter-related genes described in the plasmid-borne ter systems (Whelan et al., 1997) are either absent in the P. mirabilis strain 210 genome (terX and terY) or highly diverged (terW) (data not presented).

TerC sequence analysis
The PSORT algorithm (Nakai & Kanehisa, 1991) predicted that TerC is associated with the bacterial inner membrane. TMpred hydropathy analysis (Hofmann & Stoffel, 1993) identified nine potential transmembrane-spanning hydrophobic helices within TerC with lengths of 18–23 aa. The TerC N terminus is predicted to extend into the periplasmic space, while the C terminus protrudes into the cytoplasm. In addition, a leucine-zipper motif (Kouzarides & Ziff, 1989) was identified in the C-terminal region (residues 268–289) of this protein, though no function has been ascribed. This motif was also present in the C-terminal regions of three TerC homologues: two plasmid-encoded TerC proteins and in an M. tuberculosis hypothetical protein. Sequence analysis revealed that the mini-transposon had inserted in the carboxyl region of mutant strain S2R-6/39 (codon position 295) of terC, thereby creating a truncated TerC protein. Together with the complementation studies, these results suggest that the decreased level of Ter noted with this mutant is most likely due to attenuation of function of TerC.

Regulation of the ter operon by K2TeO3
Northern blot hybridization analysis, with different DNA probes internal to the ter operon, revealed transcripts of various sizes in RNA extracted from the wild-type and S2R-6/39 mutant strains of P. mirabilis grown in the presence of K2TeO3 (Fig. 2a). Five transcripts of 5·0, 3·0, 2·7, 2·0 and 1·5 kb were detected (K2TeO3-induced strains) when a partial terC-terD coding region was used as a probe. Similar patterns were obtained when blots were probed with a 0·9 kbp BamHI–EcoRI fragment containing DNA sequences of the terZ-terA coding region (data not presented), suggesting that the multiple bands might represent mRNA degradation products rather than distinct transcripts originating from internal promoters. The 5 kb band detected in the wild-type strain is likely to correspond to the multicistronic product of the predicted ter operon, but the identities of the other bands are more difficult to establish, because of the extensive internal sequence similarity among individual ter genes. In the absence of K2TeO3, no hybridization signals were detected in RNA, suggesting that the ter operon of P. mirabilis, unlike the locus found on IncHI plasmids, may be regulated.



View larger version (74K):
[in this window]
[in a new window]
 
Fig. 2. Northern blot analysis of ter operon (a) and urease operon (ureR) (b) of P. mirabilis strains grown in the presence or absence of tellurite. Total RNA was isolated from P. mirabilis strains S2R (wild-type) and S2R-6/39 (mutant) grown either in the presence or absence of 100 µg K2TeO3 ml-1 (indicated by the plus and minus signs, respectively). The bottom panels show rRNA in the gel before transfer, to indicate amounts loaded. Positions of molecular mass standards are shown on the left of each panel. In (a) a 1·3 kb HindIII fragment containing sequences for the terC-terD coding region was used as probe; in (b) a 1·6 kb BamHI fragment encoding the P. mirabilis ureR sequences was used as a probe.

 
The intensities of all bands (especially that corresponding to the 5 kb transcript), appeared to be significantly lower in the S2R-6/39 mutant grown in the presence of K2TeO3 (see Fig. 2a). The mini-Tn5lacZ operon fusion would likely maintain the relative size of the transcript, while transcription initiation from the Kmr cassette promoter would produce an approximately 3 kb transcript (Kmr gene, terD and terE). To evaluate the possibility that K2TeO3 exerts a more generalized effect on transcription, we compared message levels of ureR, whose expression in the absence of the inducer urea is constitutive, to those of the ter operon. As seen in Fig. 2(b), the ureR transcript level was more dramatically affected in the mutant strain than in the wild-type (6-fold versus <2-fold, respectively, by densitometry) in the presence of K2TeO3, which most likely reflects a greater toxic effect on the mutant strain. It is noteworthy that the mutant strain also exhibited slower growth in liquid media containing K2TeO3, and colony formation required an additional day over the time taken for the wild-type strain (data not presented).

Tellurite inhibits {beta}-galactosidase activity in P. mirabilis
To address the possibility that the low {beta}-galactosidase activity measured for the fully K2TeO3-induced P. mirabilis mutant strain S2R-6/39 might be due to assay interference by K2TeO3, we measured {beta}-galactosidase activity of S2R-6/39 and P. mirabilis strain S2UR1 containing an operon fusion with the ca 500 bp upstream ureR promoter region (ureR : lacZ) and inserted in single copy elsewhere from the urease gene cluster (unpublished data). As seen in Table 2, K2TeO3 decreased {beta}-galactosidase activity by 39 % for S2UR1 (fully Ter) containing a ureR : lacZ fusion. In contrast, the {beta}-galactosidase activity of S2R-6/39 was decreased by 70 % over controls following growth in K2TeO3-containing medium. The more dramatic inhibition noted with the mutant is most likely due to both decreased production of mRNA as determined by Northern blotting (Fig. 2) and to assay inhibition as demonstrated with S2UR1. Thus, intrinsic tellurite toxicity prevented the use of the lacZ fusion as a reporter for regulatory studies.


View this table:
[in this window]
[in a new window]
 
Table 2. Effect of K2TeO3 on {beta}-galactosidase levels in P. mirabilis

P. mirabilis strains were grown in the indicated concentrations of K2TeO3 as described in the text. P. mirabilis strain S2UR1 contains a single chromosomal insertion of a mini-Tn5lacZ reporter containing ~500 bp upstream region of ureR that contains promoter and regulatory sequences (unpublished observations).

 
Primer-extension analysis
Primer-extension studies identified a transcription start site upstream of terZ, the first gene in the putative operon, with RNA obtained from bacteria grown in the presence of K2TeO3 (Fig. 3a). No extension products were detected with RNA from bacteria grown in the absence of K2TeO3. A doublet corresponding to positions -51 and -52 relative to the terZ translation start site was observed, though the somewhat greater intensity of the signal at the -52 position would suggest that transcription might initiate more frequently at this base. While the location of the presumptive -10 sequence (TATAAT) was identical to the E. coli E{sigma}70 consensus sequence (Hoopes & McClure, 1987) (Fig. 3b), a consensus -35 sequence was not detected within the normal 16–21 nt spacing between the -35 and -10 sites of E. coli (Hoopes & McClure, 1987). In addition, 7 bp inverted repeats (TCATCAT) between the -35 and -10 sequences, and 7 bp direct repeats (ATTTATA) overlapping the -10 sequence and transcriptional start site might be associated with regulation of the locus. Upstream of the -35 region were sequences resembling those for OxyR-regulated genes, though this sequence showed some degeneracy from consensus OxyR regulatory motifs (Ochsner et al., 2000). The other ter genes lacked similar promoter or regulatory sequences, which together with RT-PCR analysis of terZ and terC supported an operon regulatory structure. BLAST analysis of published upstream DNA sequences of several ter operons (plasmid and bacterial) did not reveal any putative regulatory sequences or any DNA sequence similarity with the 390 bp regulatory region of P. mirabilis (data not presented). Similar analyses of nucleotide sequences of the flanking genes (orf3 and terZ) showed sequence identities ranging from 79 to 85 %, suggesting that the regulatory region and observed regulation of Ter may be unique to members of the genus Proteus.



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 3. Primer-extension analysis of the transcription start sites for the terZ locus. (a) Total RNA was extracted from P. mirabilis S2R grown in the presence or absence of tellurite and hybridized to the radiolabelled TERZ2R oligonucleotide. The arrowheads mark the transcription start sites and the corresponding major extension products. The putative -10 promoter region is underlined. The sequencing ladder was generated using the same oligonucleotide and pBSK-C1.6 as a template. (b) The terZ promoter region. The arrow above the promoter region sequence indicates the transcription start site (TT). The arrows below the sequence indicate direct and inverted repeats identified within this region. A consensus OxyR-binding motif with a 7 bp span between motifs is presented above the DNA sequence upstream of the -35 region.

 
Induction of the ter operon and response to oxidative stress
Analysis of the sequences upstream of the putative -35 hexamer (promoter sequence of terZ) revealed motifs resembling the OxyR-binding motifs associated with oxidative stress regulation (Ochsner et al., 2000). To examine the possibility that oxidative stress plays a role in regulation of Ter, P. mirabilis bacteria were incubated for 30–60 min with various concentrations of hydrogen peroxide, methyl viologen or K2TeO3, and total RNA was isolated and subjected to RT-PCR. Fig. 4 shows that K2TeO3 was a strong inducer of the ter operon (>eightfold at 26 cycles relative to uninduced controls), while hydrogen peroxide (OxyR) and methyl viologen (superoxide/SoxR) were weak inducers (approx. twofold induction).



View larger version (61K):
[in this window]
[in a new window]
 
Fig. 4. RT-PCR analysis of K2TeO3, H2O2 and superoxide induction of the ter operon. Total RNA was obtained from P. mirabilis S2R grown in broth for 30 min with either 50 µg K2TeO3 ml-1, 100 µM H2O2 or 200 mM methyl viologen (superoxide generation). Transcripts of ureR, a gene whose expression is constitutive in the absence of urea as inducter, served as a control. The relative increase in expression at 25 cycles (indicated below each lane) was normalized to the ureR amplicons.

 
To determine if prior exposure to K2TeO3 or hydrogen peroxide might increase survivability of P. mirabilis on medium containing 200 µg K2TeO3 ml-1, we compared viable counts of bacteria exposed to 50 µg K2TeO3 ml-1 or 100 µM H2O2 for 30 and 60 min to induce the ter operon with those of untreated bacteria. As seen in Table 3, in the absence of treatment, there was a 3-log decrease in plating efficiency on K2TeO3-containing medium, whereas prior induction of the ter operon by K2TeO3 was protective (<twofold decrease in efficiency of plating). Similarly, treatment of P. mirabilis with H2O2, afforded some protection against K2TeO3 lethality (~1 log). Taken together, these results demonstrate that the ter operon is activated in response to K2TeO3 and that oxidative stress partly contributes to activation of the ter operon.


View this table:
[in this window]
[in a new window]
 
Table 3. Prior treatment with K2TeO3 or hydrogen peroxide is protective against K2TeO3 toxicity

K2TeO3 or H2O2 was added to exponential-phase cultures of P. mirabilis S2/6R for the indicated times, and then decimal dilutions were spread-plated onto medium (±K2TeO3) and the numbers of c.f.u. were determined. Toxicity of K2TeO3 was measured as a decrease in c.f.u. compared to plating efficiency on medium lacking K2TeO3.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In this study we have characterized the genetic basis for Ter in P. mirabilis, a phenotype first described for this genus by Alexander Fleming in 1932 (Fleming, 1932). While Ter has typically been associated with plasmids (IncHI2 or IncHII) of enteric pathogens (Taylor 1999), our studies show that this locus is chromosomal in P. mirabilis, a feature supported by the unpublished whole-genome sequence from P. mirabilis strain 210 (SmithKline Beecham). The ter operon was also present in all recent clinical isolates of Proteus spp. examined in this study, as well as in some isolates of Morganella and Providencia. Furthermore, the ter operon was not located on large IncJ or IncT plasmids found in some strains of P. vulgaris and Providencia spp. (Boltner et al., 2002; Murata et al., 2002). A phylogenetic analysis of Ter proteins suggests a rather recent dissemination among the enteric bacteria, most likely accelerated by transmissible plasmids (Taylor et al., 2002). This conclusion is also supported by the high level of nucleotide identity (70–80 %) among terBCDE genes from the IncHI plasmids and genomic sequences from E. coli O157 : H7, Y. pestis and P. mirabilis.

The P. mirabilis ter operon differed from both its plasmid- and E. coli O157-borne counterparts, including (i) a larger terA product (44 aa larger) which may have resulted from a fusion of terD sequences with terA; (ii) the absence of terF, which is also absent from plasmids pMJ606 and pTE53 (Kormutakova et al., 2000; Taylor et al., 2002); and (iii) inducibility by tellurite and to a lesser extent by agents associated with oxidative stress. Alignment of the intervening DNA sequences between orf3 and terZ of P. mirabilis (~390 bp region containing putative regulatory motifs) with those from the published sequences of other ter operons revealed no nucleotide sequence identity, while that of the flanking genes was ~80 %, suggesting that the regulatory region may be unique to the P. mirabilis ter operon. A consequence of this regulation for P. mirabilis is that uninduced bacteria are susceptible to the toxicity of tellurite when plated onto medium containing sub-MIC levels of K2TeO3. In contrast, fully induced bacteria exhibited no loss in viability when plated onto K2TeO3-containing medium and by RT-PCR, ter mRNA transcripts from induced bacteria were increased >eightfold over those from the uninduced bacteria. The selective advantage, if any, for a regulated ter operon in natural environments, such as the urinary tract of humans, remains to be investigated.

Until recently, Ter was considered a constitutive trait of IncHI2 plasmids or the closely related chromosomal counterparts in E. coli O157 : H7 strains. Moreover, since these loci are also associated with resistance to colicins and phage, constant insults in environments occupied by enteric pathogens, constitutive expression of Ter was viewed as essential for survival. Our finding that the ter operon of P. mirabilis was inducible by tellurite has since led to evaluation of Ter in E. coli O157 : H7 strains (Taylor et al., 2002). There is considerable variability in ter loci among clinical isolates with some strains containing duplications of ter genes while others show variability in the complement of ter genes (Taylor et al., 2002). In some E. coli O157 : H7 strains, prior exposure to tellurite led to increased resistance of strains when plated on K2TeO3-containing medium and analysis of the ter genes by RT-PCR showed that mRNA of terBC and F genes was increased while that for terZD and E was not induced (Taylor et al., 2002). Interestingly, the study found that terA was not expressed, suggesting that a potential regulatory region might exist in this sequence. In contrast, an operon structure was inferred for the P. mirabilis ter locus based on primer extension and Northern blotting analysis using probes for terZA and terCD regions. Northern blots and primer extensions were negative in the absence of prior induction by tellurite and transcript size was consistent with production of a polycistronic message originating from a {sigma}70 promoter upstream of terZ. However, we cannot rule out the possibility of internal promoters, message processing or degradation contributing to the smaller transcript sizes observed by Northern blotting.

Analysis of upstream DNA promoter sequences of the ter operon revealed several motifs that might be associated with regulation. Both inverted and direct repeats were found (see Fig. 3) in the -10 and -35 promoter region and upstream of the -35 hexamer were sequence motifs resembling OxyR-binding sites (Ochsner et al., 2000). While the sequences were not identical to canonical OxyR-binding sites of catalase and alkylhydroperoxide reductase – other genes putatively regulated by OxyR in P. mirabilis (unpublished results) – the similarity may be sufficient to account for the low level stimulation of Ter by reduced oxygen intermediates that were observed in this study.

Our initial observation that non-swarming bacteria contained higher levels of {beta}-galactosidase than swarming bacteria could not be pursued further due to the intrinsic toxicity of tellurite for {beta}-galactosidase activity. Thus, the eightfold increase in terC mRNA was not accompanied by a similar increase in {beta}-galactosidase activity. In general, tellurite was toxic even for wild-type P. mirabilis as shown by decreased levels of mRNA of ter genes as well as for ureR.

Tellurite resistance is a biological puzzle
It seems unlikely that tellurite resistance mechanisms, which are both diverse and widespread in the microbial world, exist solely for the purpose of protection from tellurite toxicity, as tellurite is not found in appreciable levels in the intestinal and urinary habitats of many of these bacteria. Furthermore, since the ter operon of P. mirabilis is not expressed in the absence of tellurite, its participation in colicin and phage resistance also seems unlikely since constitutive expression, as shown for the plasmid-borne Ter operon (Whelan et al., 1997), would be required. In contrast to other enteric bacteria, colicin (proticine) production and susceptibility of the genus Proteus occurs during swarming as demonstrated by lysis of swarmer cells of incompatible strains at the intersection of two colliding swarming bands (Dienes test) (Senior, 1977). Our finding that the ter operon was even further depressed during swarmer cell development might explain the inability of this locus to provide adequate protection from the proticines of incompatible strains (data not presented). We presume that the ter operon must contribute to virulence or fitness since all Proteus strains tested contained this operon and displayed high-level Ter. Perhaps the Ter genes are associated with protection from other forms of oxidative stress or agents causing membrane damage as suggested in this study.

Finally, inducible Ter appears to be an integral feature of the genus Proteus and perhaps of related genera (Morganella and Providencia spp.), suggesting that Ter may be a distinguishing characteristic and a selective advantage in the ecology of the urinary tract.


   ACKNOWLEDGEMENTS
 
The authors would like to thank Michelle Rooker for her excellent technical assistance during the early stages of this research, David Faguy for all his invaluable help and patience with the construction of homology trees. We also wish to thank Angela Lizama, Risini Weeratna, Rafael Garduño and Avery Goodwin for their input and advice during the different stages of this research. A. T. acknowledges Dalhousie Faculty of Graduate Studies Fellowship and George Mattar Summer Studentship. D. E. T. acknowledges funding from the Canadian Institutes for Health Research (Grant MT6200) and an Alberta Heritage Scientist award. P. S. H. acknowledges funding from the Canadian Institutes for Health Research (Grant MT12668).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Altschul, S. F., Gish, W., Miller, W., Myers, E. F. & Lipman, D. J. (1990). Basic local alignment search tool. J Mol Biol 215, 403–410.[CrossRef][Medline]

Avazeri, C., Turner, R. J., Pommier, J., Weiner, J. H., Giordano, G. & Vermeglio, A. (1997). Tellurite reductase activity of nitrate reductase is responsible for the basal resistance of Escherichia coli to tellurite. Microbiology 143, 1181–1189.[Abstract]

Birnboim, H. C. & Doly, J. (1979). A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res 7, 1513–1523.[Abstract]

Bolivar, F., Rodriguez, R. L., Greene, P. J., Betlach, M. C., Heyneker, H. L., Boyer, H. W., Crosa, J. H. & Falkow, S. (1977). Construction and characterization of new cloning vehicles. II. A multipurpose cloning system. Gene 2, 95–103.[Medline]

Boltner, D., MacMahon, C., Pembroke, J. T., Strike, P. & Osborn, A. M. (2002). R391: a conjugative integrating mosaic comprised of phage, plasmid and transposon elements. J Bacteriol 184, 5158–5169.[Abstract/Free Full Text]

De Lorenzo, V., Herrero, M., Jakkubzik, U. & Timmis, K. N. (1990). Mini-Tn5 transposon derivatives for insertion mutagenesis, promoter probing, and chromosomal insertion of cloned DNA in gram-negative bacteria. J Bacteriol 172, 6568–6572.[Medline]

Devereux, J., Haeberli, P. & Smithies, O. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res 12, 387–395.[Abstract]

Di Tomaso, G., Fedi, S., Carnevali, M., Manegatti, M., Taddei, C. & Zannoni, D. (2002). The membrane-bound respiratory chain of Pseudomonas pseudoalcaligenes KF707 cells grown in the presence or absence of potassium tellurite. Microbiology 148, 1699–1708.[Abstract/Free Full Text]

Fleming, A. (1932). On the specific antibacterial properties of penicillin and potassium tellurite. J Pathol Bacteriol 35, 831–842.

Garberg, P., Engman, L., Tolmachev, V., Lundqvist, H., Gerdes, R. G. & Cotgreave, I. A. (1999). Binding of tellurium to hepatocellular selenoproteins during incubation with inorganic tellurite: consequences for the activity of selenium-dependent glutathione peroxidase. Inter J Biochem Cell Biol 31, 291–301.[CrossRef][Medline]

Hill, S. M., Jobling, M. G., Lloyd, B. H., Strike, P. & Ritchie, D. A. (1993). Functional expression of the tellurite resistance determinant from the IncHI-2 plasmid pMER610. Mol Gen Genet 241, 203–212.[Medline]

Hofmann, K. & Stoffel, W. (1993). TMBASE – A database of membrane spanning protein segments. Biol Chem Hoppe-Seyler 374, 166–173.

Hoffman, P. S. & Falkinham, J. O., III (1981). Induction of tryptophanase in short cells and swarm cells of Proteus vulgaris. J Bacteriol 148, 736–738.[Medline]

Hoffman, P. S., Butler, C. A. & Quinn, F. D. (1989). Cloning and temperature-dependent expression in Escherichia coli of a Legionella pneumophila gene coding for a genus common 60 kilodalton antigen. Infect Immun 57, 1731–1739.[Medline]

Hoffman, P. S., Ripley, M. & Weeratna, R. (1992). Cloning and nucleotide sequence of a gene (ompS) encoding the major outer membrane protein of Legionella pneumophila. J Bacteriol 174, 914–920.[Abstract]

Hohn, B. & Collins, J. (1980). A small cosmid for efficient cloning of large DNA fragments. Gene 11, 291–298.[CrossRef][Medline]

Hoopes, B. C. & McClure, W. R. (1987). Strategies in regulation of transcription initiation. In Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, pp. 1231–1240. Edited by J. L. Ingraham, K. B. Low, B. Magasanik, F. C. Neidhardt, M. Schaechter & H. E. Umbarger. Washington, DC: American Society for Microbiology.

Jobling, M. G. & Ritchie, D. A. (1987). Genetic and physical analysis of plasmid genes expressing inducible resistance of tellurite in Escherichia coli. Mol Gen Genet 208, 288–293.[Medline]

Jobling, M. G. & Ritchie, D. A. (1988). The nucleotide sequence of a plasmid determinant for resistance to tellurium anions. Gene 66, 245–258.[CrossRef][Medline]

Kieny, M. P., Lathe, R. & Lecocq, J. P. (1983). New versatile cloning and sequencing vectors based on bacteriophage M13. Gene 26, 91–99.[CrossRef][Medline]

Klein, P., Kanehisa, M. & DeLisi, C. (1985). The detection and classification of membrane-spanning proteins. Biochim Biophys Acta 815, 468–476.[Medline]

Kormutakova, R., Klucar, L. & Turna, J. (2000). DNA sequence analysis of the tellurite-resistance determinant from clinical strain of Escherichia coli and identification of essential genes. BioMetals 13, 135–139.[CrossRef][Medline]

Kouzarides, T. & Ziff, E. (1989). Leucine zipper of fos, jun and GCN4 dictate dimerization specificity and thereby control DNA binding. Nature 340, 568–571.[CrossRef][Medline]

Kroczek, R. A. & Siebert, E. (1990). Optimization of northern blot analysis by vacuum-blotting, RNA transfer, and ultraviolet fixation. Anal Biochem 184, 90–95.[Medline]

Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Murata, M., Ohnishi, T., Ara J. & 13 other authors (2002). Complete nucleotide sequence of plasmid Rts1: implications for evolution of large plasmid genomes. J Bacteriol 184, 3194–3202.[Abstract/Free Full Text]

Nakai, K. & Kanehisa, M. (1991). Expert system for predicting protein localization sites in Gram-negative bacteria. Prot Struct Funct Genet 11, 95–110.

Ochsner, U. A., Vasil, M. L., Alsabbagh, E., Parvatiyar, K. & Hassett, D. J. (2000). Role of the Pseudomonas aeruginosa oxyR-regG operon in oxidative stress defense and DNA repair: OxyR-dependent regulation of katB-ankB, ahpB and ahpC-ahpF. J Bacteriol 182, 4533–4544.[Abstract/Free Full Text]

O'Connell, W. A., Hickey, E. K. & Cianciotto, N. P. (1996). A Legionella pneumophila gene that promotes hemin binding. Infect Immun 64, 842–848.[Abstract]

Parkhill, J., Wren, B. W., Thomson, N. R. & 32 other authors (2001). Genome sequence of Yersinia pestis, the causative agent of plague. Nature 413, 523–527.[CrossRef][Medline]

Pearson, W. R. (1990). Rapid and sensitive sequence comparison with FASTP and FASTA. Methods Enzymol 183, 63–98.[Medline]

Perna, N. T., Plunkett, G., III, Burland, V. & 25 other authors (2001). Genome sequence of enterohaemorrhagic Escherichia coli O157 : H7. Nature 409, 529–533.[CrossRef][Medline]

Rahaman, M. M., Morshed, M. G., Sultanul Aziz, K. M. & Munshi, M. M. (1986). Improved medium for isolating Shigella. Lancet 1, 271–272.

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74, 5463–5467.[Abstract]

Senior, B. W. (1977). The Dienes phenomenon: identification of the determinants of compatibility. J Gen Microbiol 102, 235–244.[Medline]

Shimada, T., Sakazaki, R., Fujimura, S., Niwano, K., Mishina, M. & Takizawa, K. (1990). A new selective, differential agar medium for isolation of Vibrio cholerae O1: PMT (polymyxin-mannose-tellurite) agar. Jpn J Med Sci Biol 43, 37–41.[Medline]

Short, J. M., Fernandez, J. M., Sorge, J. A. & Huse, W. D. (1988). {lambda} ZAP: A bacteriophage {lambda} expression vector with in vivo excision properties. Nucleic Acids Res 16, 7583–7590.[Abstract]

Silhavy, T. J., Berman, M. L. & Enquist, L. W. (1984). Experiments with Gene Fusions. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Taylor, D. E. (1999). Bacterial tellurite resistance. Trends Microbiol 7, 111–115.[CrossRef][Medline]

Taylor, D. E., Rooker, M., Keelan, M., Ng, L.-K., Martin, I., Perna, N. T., Burland, N. T. V. & Blattner, F. R. (2002). Genomic variability of O islands encoding tellurite resistance in enterohemorrhagic Escherichia coli O157 : H7 isolates. J Bacteriol 184, 4690–4698.[Abstract/Free Full Text]

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.[Abstract]

Trutko, S. M., Akimenko, V. K., Suzina, N. E., Anisimova, L. A., Shlyapnikov, M., Baskunov, B. P., Duda, V. I. & Boronin, A. M. (2000). Involvement of the respiratory chain of Gram-negative bacteria in the reduction of tellurite. Arch Microbiol 173, 178–186.[CrossRef][Medline]

Turner, R. J., Weiner, J. H. & Taylor, D. E. (1999). Tellurite-mediated thiol oxidation in Escherichia coli. Microbiology 145, 2549–2557.[Abstract/Free Full Text]

Walter, E. G. & Taylor, D. E. (1992). Plasmid-mediated resistance to tellurite: expressed and cryptic. Plasmid 27, 52–64.[Medline]

Whelan, K. F., Colleran, E. & Taylor, D. E. (1995). Phage inhibition, colicin resistance, and tellurite resistance are encoded by a single cluster of genes on the IncH12 plasmid R478. J Bacteriol 177, 5016–5027.[Abstract]

Whelan, K. F., Sherburne, R. K. & Taylor, D. E. (1997). Characterisation of a region of the IncH12 plasmid R478 which protects Escherichia coli from toxic effects specified by components of the tellurite, phage and colicin resistance cluster. J Bacteriol 179, 63–71.[Abstract]

Zadic, P. M., Chapman, P. A. & Siddons, C. A. (1993). Use of tellurite for the selection of verocytotoxigenic Escherichia coli O157. J Med Microbiol 39, 155–158.[Abstract]

Received 6 September 2002; revised 20 January 2003; accepted 24 January 2003.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Toptchieva, A.
Articles by Hoffman, P. S.
Articles citing this Article
PubMed
PubMed Citation
Articles by Toptchieva, A.
Articles by Hoffman, P. S.
Agricola
Articles by Toptchieva, A.
Articles by Hoffman, P. S.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2003 Society for General Microbiology.