Physiological analysis of the role of truB in Escherichia coli: a role for tRNA modification in extreme temperature resistance

Seonag M. Kinghorn1, Conor P. O’Byrne1, Ian R. Booth1 and Ian Stansfield1

Department of Molecular & Cell Biology, University of Aberdeen, Institute of Medical Sciences, Foresterhill, Aberdeen AB25 2ZD, UK1

Author for correspondence: Ian Stansfield. Tel: +44 1224 273106. Fax: +44 1224 273144. e-mail: i.stansfield{at}abdn.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The truB gene of Escherichia coli encodes the pseudouridine-55 ({psi}55) synthase and is responsible for modifying all tRNA molecules in the cell at the U55 position. A truB null mutant grew normally on all growth media tested, but exhibited a competitive disadvantage in extended co-culture with its wild-type progenitor. The mutant phenotype could be complemented by both the cloned truB gene and by a D48C, catalytically inactive allele of truB. The truB mutant also exhibited a defect in survival of rapid transfer from 37 to 50 °C. This mutant phenotype could be complemented by the cloned truB gene but not by a D48C, catalytically inactive allele of truB. The temperature sensitivity of truB mutants could be enhanced by combination with a mutation in the trmA gene, encoding an m5U-methyltransferase, modifying the universal U54 tRNA nucleoside, but not by mutations in trmH, encoding the enzyme catalysing the formation of Gm18. The truB mutant proteome contained altered levels of intermediates involved in biogenesis of the outer-membrane proteins OmpA and OmpX. The truB mutation also reduced the basal expression from two {sigma}E promoters, degP and rpoHP3. Three novel aspects to the phenotype of truB mutants were identified. Importantly the data support the hypothesis that TruB-effected {psi}55 modification of tRNA is not essential, but contributes to thermal stress tolerance in E. coli, possibly by optimizing the stability of the tRNA population at high temperatures.

Keywords: pseudouridine-55 synthase, heat stress, trmA, sigma E

Abbreviations: Cm, chloramphenicol; Kan, kanamycin; OMP, outer-membrane protein; Tet, tetracycline; {psi}, pseudouridine


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The growth and survival of bacteria in diverse environments is achieved through homeostatic mechanisms that have evolved to maintain relatively constant cytoplasmic conditions. In parallel, gene expression is modulated so that the overall activity of metabolic pathways remains adequate for growth and survival. Despite these mechanisms, the cell has to cope with significant variations in the cytoplasmic constitution, e.g. solute composition and concentration, and pH, as well as operating at a range of temperatures. Evolution has built into the major systems of the cell, e.g. replication, transcription and translation, a degree of inherent stability that allows the cell to cope with a diversity of challenges that are imposed by changes in the cytoplasm. In the context of translation, it is now clear this stability is conferred in part by the broad range of known tRNA nucleoside modifications, which affect tRNA structure and function in a number of subtle ways. The degree of nucleoside modification can, in some cases, be varied in response to changes in the environment (Emilsson et al., 1992 ). Many examples of the dependency of specific patterns of gene expression on different components of the tRNA structural repertoire have come to light (Persson, 1993 ). Isolation of mutants with defective tRNA modification activity has frequently been associated with altered pathogenicity or metabolic activity phenotypes. The genes miaA and tgt, encoding tRNA isopentyladenosine synthase and tRNA-guanine transglycosylase, respectively, were identified as pathogenicity modulators in Shigella (Durand et al., 1997 , 1994 ). In Escherichia coli and Salmonella typhimurium, mutations at the trmD, miaE and miaA loci had associated specific metabolic deficiencies (Persson et al., 1998 ; Li & Bjork, 1995 ; Connolly & Winkler, 1991 ). At least two cases of tRNA modification regulating cellular responses to changes in the environment have been reported. The modification pathway of residue A37 to 2-methylthio-N-6-isopentenyl adenosine (ms2i6A37) is partly inhibited by iron limitation, triggering synthesis of the iron chelator enterobactin through attenuation mechanisms (Buck & Griffiths, 1982 ). The nuvA, nuvC-dependent 4-thio-U modification of U8 acts as a sensor for broad-spectrum UV irradiation; cross-linking of this modified residue following UV exposure triggers the stringent response in E. coli (Ramabhadran & Jagger, 1975 ; Persson, 1993 and references therein). Despite these discoveries, the functions of some of the most widely used and frequently found modifications, such as pseudouridine-55 ({psi}55) present in the tRNAs of eukaryotes, prokaryotes and archaea, have remained unclear.

Growth of bacterial cells across a range of osmolarity requires the modulation of the composition of the cytoplasm during maintenance of the turgor pressure. At low osmolarity the cytoplasm of most Gram-negative bacteria is predominantly composed of a mixed salts solution, principally K+ glutamate, but a number of other anions also contribute to ion balance (McLaggan et al., 1994 ; Roe et al., 1998 ). As the osmolarity is raised, the constituents of the cytoplasm are substantially determined by the composition of the growth medium, since the presence of compatible solutes, such as glycine betaine or proline, allows the cytoplasmic salt concentration to fall (Dinnbier et al., 1988 ). In the absence of compatible solutes, or their precursors, the salt concentration in the cytoplasm is very high. Recent work has suggested that tRNA modification may play a role in potentiating the translation of specific mRNA molecules in osmotically stressed cells (Sage et al., 1997 ). In Pseudomonas aeruginosa PAO1 the expression of the plcH-encoded phospholipase C (PlcH) in cells grown under phosphate-limiting conditions is dependent on the presence of glycine betaine or its precursor, choline. A transposon insertion in the orp ORF lowered this osmoprotectant-dependent induction of PlcH, and while not affecting the abundance of the mRNA, did affect PlcH protein production (Sage et al., 1997 ). This mutation also reduced the salt-tolerance of the cells and reduced the protection conferred by compatible solutes such as glycine betaine and choline under osmotic stress conditions. Subsequently, it has become clear that the orp gene is a homologue of the E. coli truB gene, which encodes tRNA {psi}55 synthase (Nurse et al., 1995 ). The enzyme is solely responsible for modification of U55 to {psi} in the T{psi}C loop of all E. coli tRNA molecules (Gutgsell et al., 2000 ). A role for this tRNA modification had not been previously identified, but the Pseudomonas data suggested a possible role in optimizing translation under conditions of osmotic stress. Thus, we sought to create a truB null mutant in E. coli and analyse its impact on cell physiology. The data show that in E. coli the gene is not essential for viability and that in this bacterium it appears to have no role in the response to osmotic stress. However, we show TruB modification of tRNA is important for optimizing cell survival at high temperatures, possibly by conferring tRNA thermal tolerance.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Strains.
The strains and plasmids used in this study are listed in Table 1. Derivatives carrying the truB deletion were constructed by co-transduction with sfsB using Tn10 insertions in this gene as the positive selection. P1 lysates were made from CAG12072 and CAG12127, which carry TetR and KanR markers in the sfsB gene close to truB. After selection of transductants on Kan- or Tet-containing LB medium and purification, putative {Delta}truB transductants were identified by PCR using primers truB11 (5'-TCCGCACTGGCGATGTGATC-3') and truB12 (5'-CAGTGAATTGCTGCCGTCAG-3') which are located upstream and downstream, respectively, of the flanking regions amplified to create the truB deletion. In the truB wild-type and mutant genomic DNA, PCR amplification with these primers yielded products of 2·0 and 1·3 kb, respectively.


View this table:
[in this window]
[in a new window]
 
Table 1. Strains and plasmids

 
Creation of truB deletion.
A chromosomal deletion in the truB gene was constructed by allelic exchange using a derivative of the suicide vector pDM4 carrying DNA fragments corresponding to truB flanking regions (Milton et al., 1996 ). PCR was used to generate these flanking regions and introduce appropriate restriction sites. The upstream flanking region, bounded by SacI and SphI restriction sites at 5' and 3' ends, respectively, included the last 13 bases of infB and the entire coding sequence of rbfA, as well as the first 75 nt of the truB gene ORF. It was amplified using primers truB1 (5'-TAGATGGAGCTCACGTACCATTGCTTAAGG-3'; SacI site underlined) and truB7 (5'-TAGATGGCATGCGCTGGACATACCCTGAGG-3'; SphI site underlined). The downstream flanking region contained the last 189 bases of the truB coding sequence and extended down to 13 bases past the stop codon of the rpsO gene downstream of truB. This fragment was bounded by SphI and XbaI sites at the 5' and 3' ends, respectively. It was amplified using primers truB8 (5'-TAGATGGCATGCCCGGTGGTGAATCTTCCG-3'; SphI site underlined) and truB4 (5'-TAGATGTCTAGAGAAACTCGCAAGAATTAG-3'; XbaI site underlined). Following PCR amplification, the upstream and downstream flanking regions were digested with SphI and ligated together to produce a single fragment containing the truB deletion. This fragment, restricted with SacI and XbaI, was then cloned into plasmid pDM4, cut using the same enzymes, to create plasmid p{Delta}truB. The plasmid was transformed into the maintenance strain S17-1, which carries the pir gene, required for replication of the suicide plasmid. Frag1 was conjugated with S17-1 with selection on chloramphenicol (Cm) minimal medium. Cm resistance was used to select for chromosomal integration in Frag1, since this strain cannot support pDM4 plasmid replication. To complete the allelic exchange, the integrated suicide plasmid, which carries the sucrose intolerance gene sacB (Milton et al., 1996 ), was induced to recombine out of the chromosome by growth on sucrose; CmS transconjugants were then selected to confirm pDM4 excision. Loss of the plasmid by recombination should occur so as to produce wild-type truB and {Delta}truB strains in an equal ratio. However, the majority of CmS clones carried the wild-type truB allele. To increase the chances of obtaining a deletion strain, 208 CmR transconjugants were cultured together in LB and plated onto 15% sucrose medium. Two hundred colonies were replica-plated onto Cm to check for plasmid loss. These were then screened by PCR for the presence of wild-type or {Delta}truB alleles. One out of the 200 colonies screened contained the deleted form of the gene. The deletion was confirmed by amplifying the deleted region using PCR and sequencing the cloned PCR products both to verify the presence of the deletion and to ensure that no PCR-generated base substitutions were introduced into the flanking genes.

The chromosomal truB deletion was combined by transduction with mutant trmA or trmH loci in the following way. P1 lysates were prepared from strains GB1701 and CF1499 (a gift from G. Björk, Ume, Sweden) and used to transduce the trmA and trmH insertions into strains MJF545 (TruB+ TetR) and MJF546 (TruB- TetR). The presence of wild-type or mutant truB, trmA and trmH alleles was confirmed by PCR. Strains MJF565 (TruB+ TrmA-), MJF566 (TruB- TrmA-), MJF569 (TruB+ TrmH- CmR) and MJF570 (TruB- TrmH- CmR) were generated in this way.

Cloning of truB.
To complement any phenotypes resulting from deletion of the truB gene, a truB expression vector was constructed. truB was cloned into the inducible expression vector pTrc99A (Amann et al., 1988 ) which carries the strong trc promoter. This promoter can be induced by the addition of 1–5 mM IPTG to give strong expression of the cloned gene. In the absence of IPTG the lacIq gene represses expression. However, it has been shown that even in the absence of inducer, some read-through occurs to give basal expression of the gene (Tötemeyer et al., 1998 ). truB was amplified by PCR in triplicate from a Frag1 template using primers truBF (5'-AGGAGGACCCATGGGTCGTC-3') and truB10 (5'-TAGATCGGATCCTATCGCAAGACGGTTAAC-3'). The forward primer, truBF, includes an NcoI site which overlaps the start codon of truB, and truB10 includes a BamHI site for cloning into pTrc99A. The PCR products were digested and cloned into ptrc99A and the recovered plasmid, pSK1, was sequenced on both strands. Initially, no expression of the truB gene could be observed in cell extracts obtained after induction with IPTG. The introduction of an NcoI site at the start codon of truB could have caused this problem with expression as it changes the second codon from AGT (Ser) to GGT (Gly). ‘Quickchange’ PCR site-directed mutagenesis (Stratagene) was performed to restore the normal codon and the resulting plasmid, pSK2, was sequenced. IPTG-induced expression of pSK2 resulted in synthesis of a 35 kDa protein in cells carrying pSK2, but not for those carrying pTrc99A. This 35 kDa protein was excised from the gel and identified as TruB by digestion with trypsin and characterization of the resulting peptides by MS (data not shown). In addition complementation studies were carried out with two plasmids ptruB and ptruBD48C (Gutgsell et al., 2001), which were supplied by Dr J. Ofengand (University of Florida, Miami, USA).

Growth conditions and media.
Growth of E. coli strains was carried out at 37 °C in LB (g l-1: tryptone, 10; yeast extract, 5; NaCl, 10) or defined medium based on citrate-phosphate buffer, pH 7·0, containing (l-1): Na2HPO4, 7·56 g; citric acid, 1·13 g; K2HPO4, 0·28 g; MgSO4, 0·1 g; (NH4)2SO4, 1 g; (NH4)2SO4.FeSO4, 0·02 mg; thiamine, 1 µg. Growth was monitored by measuring OD650 of samples. For the competition culture experiments, viable cells were determined by plating dilutions onto both LB plates and LB plates containing Tet at 25 µg ml-1. Relative percentage survival was calculated by expressing the number of TetR cells as a fraction of the total cell number and normalizing this against the fraction present in the original mixture. When plasmids were present during competition experiments, carbenicillin (0·1 mg ml-1) was included in the medium throughout and cells were also plated onto LB plates containing 0·1 mg carbenicillin ml-1 to check for maintenance of the plasmids. For growth to exponential phase in defined medium, a single colony was used as inoculum for overnight growth under limiting glucose conditions (0·04%, w/v). Cells were supplemented with glucose (0·2%, w/v) and allowed to go through one cell doubling before dilution to OD650 of 0·03 and growth to the appropriate optical density. Other than in the extended competition experiments, when plasmids were present in strains ampicillin (25 µg ml-1) was included in the overnight culture only. For growth under anaerobic conditions, cells were plated onto LB or minimal agar plates and incubated in an anaerobic cabinet for 3 days. Minimal plates were supplemented with 0·2% glucose, lactate or pyruvate as carbon source and with 20 mM NaNO3. Liquid cultures were also tested in the same media. Media (23 ml) in 25 ml tubes were inoculated and sealed. These were incubated in a static incubator over 2 days and growth was observed by monitoring medium turbidity.

Proteomic analysis.
MJF543 and MJF544 were grown to exponential phase in defined medium and harvested at an OD650 of 0·3. A 20 ml sample of culture was centrifuged at approximately 6000 g (4500 r.p.m.; Jouan MR22i) for 10 min. The pellet was washed twice in cold PBS (g l-1: NaCl, 10; KCl, 0·25; NaH2PO4, 1·44; K2HPO4, 0·25) and resuspended in 100 µl lysis buffer [8 M urea, 4% (v/v) Triton X-100, 40 mM Tris Base]. Cells were lysed by three 20 s pulses of sonication on ice (ARTEK 20 kHz sonic dismembranator). Unbroken cells were removed by centrifugation at 10000 g (13000 r.p.m.; Jouan A14) for 5 min. Samples to be compared were run on one-dimensional (1D) SDS-PAGE (Laemmli, 1970 ) to equalize loading before two-dimensional (2D) electrophoresis was carried out. Approximately 20 µg protein was loaded onto pH 4·0–7·0 iso-electric focusing (IEF) strips for separation in the first dimension (Pharmacia 11 cm Immobiline drystrips). Proteins were resolved in the second dimension on 12–14% acrylamide gels (Pharmacia ExelGel XL). Proteins were visualized by either silver staining (Pharmacia PlusOne kit) or with Gelcode Blue (Pierce and Warriner). Spots of interest were excised and analysed by tryptic fragment digestion and MS, using an Applied Biosystems Voyager DE-STR MALDI-TOF mass spectrometer. Outer-membrane samples of MJF543 and MJF544 were prepared during exponential growth in defined medium as described by Piddock et al. (1990).

ß-Galactosidase activity measurement.
ß-Galactosidase activity was measured for cells growing exponentially in defined medium at an OD650 of 0·3 or for cells during growth in LB. Assays for ß-galactosidase were performed according to the method of Miller (1972) , modified as follows. Bacterial culture (75 µl) was removed and incubated on ice for at least 30 min. Cells were then permeabilized by adding 675 µl Z-buffer (Miller, 1972 ), 75 µl CHCl3, 37·5 µl 0·1% SDS, and vortexing for 10 s. The assay tubes were incubated at 28 °C for 5 min before adding 200 µl O-nitrophenyl-ß-D-galactopyranoside (ONPG, 4 mg ml-1; made up in distilled H2O). The reaction was incubated at 28 °C until developed and stopped by the addition of 375 µl 1 M Na2CO3. Cell debris was removed by centrifugation for 4 min. One millilitre of the resulting supernatant was removed and A420 was determined.

Determination of viability at 50 °C.
MJF543 and MJF544 containing pTrc99A, ptruB or ptruBD48C were grown to exponential phase in defined medium to an OD650 of 0·4 before being diluted 1:20 into medium pre-warmed to 50 °C. Samples were taken over a time course and diluted into citrate-phosphate buffer lacking supplements. Serial dilutions were made and three 5 µl spots of each dilution were plated onto defined medium plates. Plates were incubated for 48 h at 37 °C and colonies were counted to measure survival.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Creation of a truB deletion strain
The influence of TruB-mediated modification of tRNA was investigated through the construction of a null mutant and through cloning of the structural gene. During the course of this study a further truB null mutant and complementing clones became available (Gutgsell et al., 2000 ). In view of the complex and subtle phenotypes observed in this study, which have not previously been observed, the important results have been replicated with these independently isolated strains and constructs. A truB mutant of Frag1 was created by allelic exchange using the suicide vector pDM4 (Milton et al., 1996 ). The resulting mutation removes 226 of 314 codons from the gene sequence, including a highly conserved region that encodes the invariant aspartate residue (D48), known to be essential for catalytic activity (Ramamurthy et al., 1999 ). Deletion of the truB gene results in complete lack of modification of U55 residues in tRNACys and tRNAPhe tested (Gutgsell et al., 2000 ). The mutant allele could be subsequently transduced into different strains of E. coli at a high frequency by linkage to the proximal transposon located in the sfsB gene [sfsB203::Tn10 and sfsB3198::Tn10(Kan) by transduction to TetR and KanR, respectively]. Linkage of approximately 29±9% was observed between {Delta}truB and the transposon insertion in all transductions, whether the strain already possessed the truB deletion or when {Delta}truB was co-transduced into a new genetic background. The isogenic {Delta}truB strains arising from transduction of Frag1 to either TetR (MJF546) or KanR (MJF544) were used for the analysis of the physiology of the truB mutant (Table 1). Since the Tn10(Kan) marker was created by allelic conversion of the Tn10 strain (Singer et al., 1989 ), the two markers are present in the same chromosomal location. This enabled direct comparison of the strains carrying {Delta}truB both with their isogenic parents and with their isoallelic, but differentially drug-resistant TruB+ derivatives.

Deletion of the truB gene produces a growth disadvantage
The importance of the truB gene product to osmotic stress responses in P. aeruginosa, where expression of the phospholipase plcH gene was compromised in an orp (truB) disruptant (Sage et al., 1997 ), prompted an examination of the osmotic stress responses of the E. coli TruB- strain, MJF544 (TruB- KanR). This strain was viable and grew normally in both rich and minimal glucose media. Growth was then tested under conditions of osmotic stress in the presence and absence of compatible solutes. Unlike a truB (orp) mutant in P. aeruginosa, the response of MJF544 (TruB-) to these particular osmotic stress conditions was indistinguishable from the isogenic control strain MJF543 (TruB+ KanR) under these conditions. The growth rates of MJF543 and MJF544 were equally inhibited in the presence of 0·6 M NaCl and partially restored to the same extent in the presence of 1 mM glycine betaine (data not shown). From these results it is clear that deletion of truB does not significantly affect growth under conditions of osmotic stress or the ability of osmoprotectants to restore growth. Growth in the presence of NaCl concentrations ranging from 0·1 to 0·8 M and the use of proline as a compatible solute were also indistinguishable between MJF543 (TruB+) and MJF544 (TruB-) (data not shown). Loss of truB likewise had no detectable effect on growth under conditions of limiting magnesium (0·4 nM–0·4 mM), weak acid stress (pH 6·0, 0–20 mM sodium acetate), anaerobic growth and a range of temperatures from 15 to 42 °C (data not shown).

Because the TruB- strain displayed no overt growth disadvantage under the conditions tested, mixed culture competition experiments were used, as they can reveal very small differences in growth and/or adaptation over repeated cycling through lag to exponential and stationary phases (Bjork & Neidhardt, 1975 ). Accordingly, the TruB- strain MJF546 (TetR) was co-cultured with the wild-type parent Frag1 and its ability to compete was observed by monitoring the relative proportions of the two strains after extended subculturing. Overnight cultures of wild-type and truB mutant strains were mixed in equal ratios, grown for 24 h at 37 °C (equivalent to 10 generations followed by entry into stationary phase) before dilution into fresh medium to begin a fresh cycle. The proportion of TruB- cells to wild-type cells present in the mixture was measured by plating dilutions of culture onto LB and LB Tet plates and counting the number of colonies formed. A decrease in the ability of MJF546 (TruB- TetR) to compete with Frag1 became apparent after about four cycles of competitive growth, i.e. about 40 generations (Fig. 1). Strain MJF545 (truB+) was not competitively disadvantaged in co-culture with Frag1 (Table 1), suggesting that the antibiotic resistance markers did not contribute to this difference (data not shown). The ability of truB to complement this loss in competitive fitness was tested by transforming MJF546 (TruB-) with ptruB (Gutgsell et al., 2000 ) or a plasmid carrying an inactive mutant, ptruBD48C (Gutgsell et al., 2000 ). Fig. 1 shows that both the wild-type gene and the D48C allele encoding inactive TruB can complement the loss in competitive fitness that results from deletion of chromosomal truB, while the vector pTrc99A alone cannot. These experiments, demonstrating complementation of a truB deletion phenotype by both wild-type and D48C truB alleles, confirm those reported previously using an alternative Kan disruption allele of truB (Gutgsell et al., 2000 ).



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 1. Loss of truB impairs competitive ability in co-culture with wild-type. TruB+ and TruB- strains transformed with appropriate truB-expressing plasmids or control vectors were grown overnight in LB medium containing carbenicillin. The cultures to be tested in competition with one another were mixed in a 1:1 ratio and diluted 103-fold into fresh LB containing carbenicillin. These were incubated at 37 °C for 24 h and grown for about 10 generations before reaching stationary phase. The stationary-phase culture was used as an inoculum for further cycles of growth and the composition of the mixed culture was determined by plating dilutions onto LB plates and LB plates containing Tet. Relative percentages of TetR cells are shown for MJF546 (TruB- TetR) containing the complementation plasmids pTrc99A ({bullet}), ptruB ({circ}) or ptruBD48C ({square}) in competition with Frag1 pTrc99A.

 
Deletion of the truB gene affects the levels of key outer-membrane proteins (OMPs)
The discovery that deletion of the truB gene produces detectable and reproducible growth defects in E. coli, but no alterations in osmotic sensitivity, prompted the question of its role in cellular physiology. The proteome of strain MJF544 (TruB-) was compared with that of strain MJF543 (TruB+) to determine global changes in gene expression consequent upon deletion of truB and resulting absence of {psi}55 in tRNA molecules. A proteomic approach rather than transcriptome analysis was employed as deleting truB might primarily affect post-transcriptional control of gene expression, as was found for the plcH gene in P. aeruginosa orp (truB) mutants (Sage et al., 1997 ). MJF544 (TruB- KanR) and MJF543 (TruB+ KanR) cultures growing exponentially in glucose minimal medium were harvested and total cell extracts were prepared. These were analysed by 2D PAGE. Reproducible (n=3) changes in the levels of three separate spots whose intensities exhibited major differences were observed in protein pattern of MJF544 compared to that of MJF543. In the mutant strains one protein increased in abundance (Fig. 2a) and two proteins decreased in abundance (Fig. 2a, b) relative to the parent. The affected proteins were identified by excision of the spot, digestion with trypsin and characterization of the resulting peptides by MS. The 40 and 33 kDa proteins were both found to be OmpA (13 of 23 peptides and 6 of 12 peptides, respectively, matched the OmpA fingerprint; Fig. 2a). While no tryptic fragment corresponding to the OmpA signal peptide was detected, a mass for OmpA of 40 kDa, and the more alkaline pI of this spot, are both consistent with the predicted properties of the signal-peptide-tagged form of the protein. The third change, a 17 kDa protein that has decreased in abundance in the TruB- proteome, was identified as OmpX (3 of 4 peptides matched the OmpX fingerprint). The major proteins analysed by proteomics are soluble components of the cell, but both proteins identified are components of the outer membrane. Therefore, OMP fractions were isolated and were analysed by 1D gel electrophoresis. No reproducible (n=3) differences in the abundance of the mature OMPs were observed between parent and mutant (Fig. 2c). These data suggest subtle changes in the expression of OmpA and OmpX in the TruB- mutant, with OmpA processing possibly also affected.



View larger version (97K):
[in this window]
[in a new window]
 
Fig. 2. truB deletion affects OmpA and OmpX processing and expression. (a, b) Total cell protein extracts were prepared for MJF543 (TruB+) and MJF544 (TruB-) growing exponentially in minimal medium at 37 °C. These were analysed by 2D PAGE and visualized by silver staining. Three reproducible changes are circled in enlarged representations of the 2D gel. These spots were identified by tryptic fragment digestion and MS. Panel (a) shows changes in OmpA: 40 and 32 kDa proteins exhibiting decreased and increased abundance, respectively; in the mutant these were both identified as OmpA. Panel (b) shows changes in OmpX: a 17 kDa protein exhibiting decreased abundance in the mutant was identified as OmpX. (c) Outer-membrane fractions were also prepared from exponentially growing cultures as described in Piddock et al. (1990) . Outer-membrane fractions from MJF543 (lane 1) and MJF544 (lane 2) were analysed by 1D SDS-PAGE. Slight differences in band intensity apparent in different OMP fractions were not seen reproducibly (n=3).

 
Basal activity of {sigma}E is reduced in a {Delta}truB strain
Changes in expression of OMPs have been shown to affect the activity of the alternative sigma factor, {sigma}E, which regulates extracytoplasmic stress responses (Mecsas et al., 1993 ). Among the genes regulated by {sigma}E are the heat shock sigma factor gene rpoH (Erickson et al., 1987 ), a periplasmic endopeptidase DegP (Lipinska et al., 1988 ; Lipinska et al., 1989 ; Erickson & Gross, 1989 ) and its own gene rpoE (Rouviere et al., 1995 ; Raina et al., 1995 ). Since deletion of truB causes altered accumulation of soluble forms of the OMPs OmpA and OmpX, the effect of the mutation on basal and induced levels of {sigma}E activity was investigated. Strains TR49 and TR71 containing the {sigma}E-regulated gene fusions degP-lacZ and rpoHP3-lacZ, respectively, (Danese et al., 1995 ) were used to monitor activity of {sigma}E within cells. The truB deletion was moved into these strain backgrounds by P1 transduction using the linked sfsB::Tn10Kan allele as a selectable marker (Table 1). ß-Galactosidase activities were then monitored in cells growing exponentially in minimal glucose medium (OD600 0·3). Deletion of truB reduced expression of the degP reporter gene to 39±5% (n=3) of that measured in the wild-type strain, and of the rpoH reporter gene to 55±6% (n=3) of that measured in the wild-type strain. In contrast, expression of lacZ from two control plasmids, pHSGS and pHSGL carrying mscS-lacZ and mscL-lacZ fusions, respectively, as well as the vector alone, pHSGlacZ, was unaffected by the presence or absence of a functional truB allele (data not shown). These data suggested that the basal level of {sigma}E activity is impaired in a TruB- background.

{sigma}E activity is induced by high temperatures, ethanol and overexpression of OMPs (Erickson & Gross, 1989 ; De Las Penas et al., 1997 ; Mecsas et al., 1993 ). Overexpression of the ompC gene was used to investigate whether or not deletion of truB affects activity of {sigma}E under inducing conditions. Cells carrying the lacZ gene fusions were transformed with either pMY150, which carries ompC under the control of its own promoter, or with the vector alone, pBR322, and ß-galactosidase activity was measured. Overexpression of ompC induces high expression of degP and rpoH, and similar levels of expression were seen in both the TruB- mutant and the parent strain (data not shown). Thus the effect of deleting truB on basal {sigma}E regulation can be overcome under {sigma}E-inducing conditions. Importantly, these data rule out the possibility that lack of TruB specifically impairs the translation of lacZ mRNA. The activity of {sigma}E also increases dramatically in stationary-phase cells in LB medium. This is probably due to stress imposed by the rise in the pH of the medium (up to pH 9) and by nutrient deprivation. The truB mutant stationary-phase response was indistinguishable from that of the isogenic parent, indicating again that {sigma}E activity can be induced to normal levels in such mutants (data not shown).

The ability of pSK2 and ptruB to complement the reduction of {sigma}E activity seen in the truB mutant was tested, but no complementation was observed even when truB expression from the trc promoter was induced using IPTG (data not shown). To rule out the possibility that the truncated protein encoded by the in-frame deletion created in the truB gene may generate unforeseen dominant-negative effects, a truB mutation created by a KanR cassette replacement of the truB gene (Gutgsell et al., 2000 ) was transduced into the degP-lacZ reporter strain. This truB mutant also exhibited reductions in activity of {sigma}E as reported by reduced basal activity of the degP promoter, but again this failed to be complemented by overexpression of truB (data not shown).

truB is important for survival at very high temperatures
One of the genes regulated by {sigma}E is rpoH, encoding the heat shock sigma factor, {sigma}32. Since the majority of rpoH transcription at temperatures above 50 °C is {sigma}E-dependent and {sigma}E activity is important for survival of lethal temperatures (Hiratsu et al., 1995 ), the ability of the truB mutant strain to survive a heat shock of 50 °C was studied. Survival of exponential-phase cells rapidly transferred from 37 to 50 °C was investigated using strains MJF544 (TruB-) and MJF543 (TruB+). TruB- cells suffered a consistently greater loss in viability than the TruB+ isogenic strain (Fig. 3). A similar pattern was also observed after transfer to 52 °C, with the TruB- mutant even less able to survive these conditions (data not shown). The ability of truB to complement this loss in viability was tested by transforming MJF544 (TruB-) with either ptruB or the inactive mutant, ptruBD48C (Gutgsell et al., 2000 ). Fig. 3 shows a representative graph of loss of viability over an 80 minute time course with and without complementation by these two plasmids. It is clear that the loss in viability of the TruB- strain is restored to wild-type levels when complemented with catalytically active truB, but not by the inactive D48C mutant. The data suggest that a truB mutant is impaired in its survival of high temperature, and that the cause of this phenotype is loss of {psi}55 in tRNAs.



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 3. Loss of truB impairs survival at 50 °C. MJF543(TruB+)(pTrc99A) ({bullet}), MJF544(TruB-)(pTrc99A) ({circ}), MJF544(TruB-)(ptruB) ({blacksquare}) and MJF544(TruB-)(ptruBD48C) ({square}) growing exponentially in defined medium at 37 °C were diluted 1:20 into medium preheated to 50 °C. Cultures were sampled for viability over a time-course. Error bars represent ±1 SD (n=2) from duplicates carried out on the same day. The experiment was repeated three times, each time giving similar relative survival profiles.

 
Combining mutations in trmA, encoding the tRNA m5U54 methyl-transferase, and truB enhances the truB extreme temperature sensitivity phenotype
Deletion of truB reduces the ability of E. coli to survive exposure to lethal temperatures and this phenotype cannot be complemented by a D48C catalytically inactive mutant. The inference from this finding is that the {psi}55 modification itself is required for survival of extreme thermal stresses. The {psi}55 nucleoside is adjacent to another universally conserved modified residue, T54 (where T denotes m5U), in the tRNA T{psi}C loop. The T54 modification is catalysed by trmA (Persson et al., 1992 ). The T{psi}C loop combines with the tRNA D loop in the tRNA tertiary structure to form the ‘elbow’ of the folded tRNA, bringing {psi}55 in close proximity to the conserved residue G18, which in many tRNAs is methylated by the trmH gene product (Sprinzl et al., 1996 ; Persson et al., 1997 ). To probe the role of tRNA ‘elbow’ nucleoside modifications in conferring tRNA thermal stability, the effect of combining trmA or trmH mutations with deletion of the truB gene on survival of a 50 °C challenge was investigated. The single and double mutants described above were grown to exponential phase in minimal medium before being diluted into minimal medium pre-warmed to 50 °C. Viability was sampled over an 80 min time course. The trmH mutant does not enhance truB mutant temperature sensitivity (Fig. 4b). However, in marked contrast, while the trmA single mutant has the same survival profile as its wild-type parent, the trmA truB double mutant shows a greater loss in viability than the truB mutant alone (Fig. 4a). The simultaneous elimination of both {psi}55 and T54 modifications from E. coli tRNAs therefore combines to produce an enhancement of the truB thermal sensitivity phenotype.



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 4. truB trmA double mutants exhibit increased sensitivity to 50 °C incubation compared to either mutant individually. MJF545 (TruB+; {bullet}), MJF546 (TruB-; {blacksquare}) and a series of mutant strains in which the truB disruption was combined with either trmA or trmH mutations (detailed below) were grown to exponential phase in defined medium at 37 °C and diluted 1:20 into medium preheated to 50 °C. Cultures were sampled for viability over a time course. Error bars represent ±1 SD (n=2) from duplicates carried out on the same day. The experiment was repeated three times, each time giving similar relative survival profiles. (a) MJF565 (TruB+ TrmA-, {circ}) and MJF566 (TruB- TrmA-, {square}). (b) MJF569 (TruB+ TrmH- CmR, {circ}) and MJF570 (TruB- TrmH- CmR, {square}).

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In the enteric bacteria, a large number of genes (almost 50) are involved in modification of tRNA. Such a genetic commitment to nucleoside modification poses the question why does the cell modify tRNA? In this study we have shown that loss of the truB gene in E. coli causes sensitivity to high temperature, alterations in the basal expression of components of the {sigma}E regulon, changes in the processing of OMPs and a subtle defect in growth that is only exposed by repeated cycling of the mutant in competition with the wild-type. The last of these is a well characterized phenotype associated with several tRNA modifying enzymes (Bjork & Neidhardt, 1975 ; Raychaudhuri et al., 1999 ) and has also been established for truB mutants in E. coli (Gutgsell et al., 2000 ). For truB it is clear, both from recent observations and from our own study, that it is the protein rather than its modifying activity that is required for successful competition with the wild-type (Gutgsell et al., 2000 ; Fig. 1). However, we have also established for the first time a role for the TruB-mediated tRNA modification itself, as well as for the TrmA-catalysed tRNA modification, T54 (Figs 3 and 4).

Growth of the truB mutant was normal in most circumstances, e.g. in the presence of 0·8 M NaCl, at pH 6 with 20 mM sodium acetate, and at 42 °C. The absence of a major growth phenotype suggests subtle modifications of the cell. In Saccharomyces cerevisiae, the truB-homologous gene, PUS4, which encodes the enzyme responsible for {psi}55 formation, is likewise non-essential and no significant mutant phenotypes have yet been detected (Becker et al., 1997 ). More significant changes are seen in similar mutants of P. aeruginosa, which exhibit temperature-sensitive growth on solid media and reduced salt tolerance (Sage et al., 1997 ). The lack of a comparable truB osmotic sensitivity phenotype in E. coli (this study) indicates that the roles of TruB in fine-tuning cell physiology are likely to be species-specific.

By analysing the proteome of truB mutants and wild-type cells in detail, our study has shown, however, that the E. coli truB mutant may exhibit altered processing of OMP precursors (Fig. 2). Another clear phenotype associated with the truB mutation connected with OMP processing was a reduced basal level of expression of the {sigma}E regulon (this study, see Results), which is required for survival at high temperature (Erickson & Gross, 1989 ). Induced expression of the {sigma}E regulon, either by constitutive high-level synthesis of OmpC or by entry into stationary phase in LB, was not affected, suggesting that this aspect of the phenotype is overcome by conditions inducing high-level expression of {sigma}E. Reduced levels of basal {sigma}E expression would be expected to generate precisely the extreme temperature sensitivity phenotype we observed (Figs 3 and 4). It is intriguing then that the {sigma}E expression phenotype could not be complemented by the wild-type truB+ gene on pTrc99A. Since the genomic truB allele generated was an in-frame deletion, it seems unlikely that the mutation would cause polar effects on downstream genes in the operon through premature termination of transcription. Furthermore, our studies have shown that an identical effect on {sigma}E basal expression is exhibited by a KanR gene disruption of truB, generated in a separate study (Gutgsell et al., 2000 ). It is possible that complementation of this phenotype requires a carefully regulated, and precise, level of TruB expression not achievable using a pTrc99A expression system either with or without IPTG induction. However, it is clear at least that the truB effect on {sigma}E basal expression is not responsible for the observed sensitivity to thermal shock, since of the two phenotypes, only the thermal sensitivity phenotype is complemented by the wild-type truB gene.

The major phenotypic change in E. coli truB mutants is the reduced survival upon rapid transfer from 37 to 50 °C. The reduction in viability of the TruB- strain at 50 °C was restored by overexpression of truB, but not by expression of the inactive form truBD48C, indicating that the tRNA {psi}55 modifying activity of TruB is required for optimal survival of lethal heat stress. Thus a TruB- phenotype identified in this work for the first time demonstrates a requirement for the {psi}55 tRNA modification. Using extreme temperature survival phenotypes, we therefore demonstrate that it is possible to uncouple the dependency on {psi}55 formation and a requirement for some other, as yet unidentified function of catalytically inactive TruB important in growth competition (Gutgsell et al., 2000 ). There are two models to explain our data. First, undermodification of the T{Psi}C loop at U55 and U54 causes physiological effects (e.g. reducing translation accuracy) consequently leading to temperature sensitivity. Previous studies have indicated, however, that in vitro tRNAs lacking {psi}55 are used in translational elongation with an efficiency indistinguishable from wild-type tRNAs (Nazarenko et al., 1994 ; Rudinger et al., 1994 ). Measurements of translational accuracy using ß-galactosidase thermal lability did not show any indication of misreading occurring in a truB mutant (data not shown), although subtle changes would not be detected with this assay. Second, it is possible that undermodification of tRNAs at position 55 directly causes extreme temperature sensitivity, since this truB phenotype can be enhanced by combination with a trmA mutant, which modifies the U54 residue immediately adjacent to {psi}55. Thus the modifications in the T{Psi}C loop of tRNAs may impart thermal stability to the tRNA molecule and, consequently, the whole cell. Interestingly, lack of modification of tRNA residue G18, methylated in many tRNAs and juxtaposed with residues 54 and 55 in the folded tRNA molecule, did not enhance the truB temperature sensitivity phenotype (Fig. 4b). This implies either that the Gm18 residue does not contribute to tRNA thermal stability via direct {psi}55 interaction or that its non-ubiquitous nature in E. coli tRNAs reduces its importance for overall fitness of the cell.

In conclusion, for the first time we have demonstrated a role for the enzymic activity of the TruB-mediated modification of tRNA in E. coli, and for that of the TrmA enzyme. Our work suggests the intriguing possibility that tRNA nucleoside modification in the elbow region of a tRNA molecule is important for tRNA thermal stability and thus cell survival potential at high temperatures.


   ACKNOWLEDGEMENTS
 
We are grateful to Dr James Ofengand, University of Florida, Miami, USA, for the generous gift of plasmids ptruB and ptruBD48C, to Professor Glenn Björk (University of Ume, Sweden), for the gift of trmA and trmH strains GB1701 and CF1499, and to Dr Tom Silhavy (Princeton University, USA) for the gift of the lacZ fusion strains TR49 and TR71.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Amann, E., Ochs, B. & Abel, K.-J. (1988). Tightly regulated tac promoter vectors useful for the expression of unfused and fused proteins in Escherichia coli. Gene 69, 301-315.[Medline]

Becker, H. F., Motorin, Y., Planta, R. J. & Grosjean, H. (1997). The yeast gene YNL292w encodes a pseudouridine synthase (Pus4) catalyzing the formation of psi55 in both mitochondrial and cytoplasmic tRNAs. Nucleic Acids Res 25, 4493-4499.[Abstract/Free Full Text]

Bjork, G. R. & Neidhardt, F. C. (1975). Physiological and biochemical studies on the function of 5-methyluridine in the transfer ribonucleic acid of Escherichia coli. J Bacteriol 124, 99-111.[Medline]

Buck, M. & Griffiths, E. (1982). Iron mediated methylation of tRNA as a regulator of operon expression in Escherichia coli. Nucleic Acids Res 10, 2609-2624.[Abstract]

Connolly, D. M. & Winkler, M. E. (1991). Structure of Escherichia coli K-12 miaA and characterization of the mutator phenotype caused by miaA insertion mutations. J Bacteriol 173, 1711-1721.[Medline]

Danese, P. N., Snyder, W. B., Cosma, C. L., Davis, L. J. B. & Silhavy, T. J. (1995). The Cpx two-component transduction pathway of Escherichia coli regulates transcription of the gene specifying the stress-inducible periplasmic protease, DepP. Genes Dev 9, 387-398.[Abstract]

De Las Penas, A., Connolly, L. & Gross, C. A. (1997). {sigma}E is an essential sigma factor in Escherichia coli. J Bacteriol 179, 6862-6864.[Abstract]

Dinnbier, U., Limpinsel, E., Schmid, R. & Bakker, E. P. (1988). Transient accumulation of potassium glutamate and its replacement by trehalose during adaptation of growing cells of Escherichia coli K-12 to elevated sodium chloride concentrations. Arch Microbiol 150, 348-357.[Medline]

Durand, J. M., Okada, N., Tobe, T. & 7 other authors (1994). vacC, a virulence-associated chromosomal locus of Shigella flexneri, is homologous to tgt, a gene encoding tRNA-guanine transglycolase (Tgt) of Escherichia coli K12. J Bacteriol 176, 4627–4634.[Abstract]

Durand, J. M., Bjork, G. R., Kuwae, A., Yoshikawa, M. & Sasakawa, C. (1997). The modified nucleoside 2-methylthio-N6-isopentyladenosine in tRNA of Shigella flexneri is required for expression of virulence genes. J Bacteriol 179, 5777-5782.[Abstract]

Emilsson, V., Naslund, A. K. & Kurland, C. G. (1992). Thiolation of transfer RNA in Escherichia coli varies with growth rate. Nucleic Acids Res 20, 4499-4505.[Abstract]

Erickson, J. W. & Gross, C. A. (1989). Identification of the sigma E subunit of Escherichia coli RNA polymerase: a second alternate sigma factor involved in high-temperature gene expression. Genes Dev 3, 1462-1471.[Abstract]

Erickson, J. W., Vaughn, V., Walter, W. A., Neidhardt, F. C. & Gross, C. A. (1987). Regulation of the promoters and transcripts of rpoH, the Escherichia coli heat shock regulatory gene. Genes Dev 1, 419-432.[Abstract]

Gutgsell, N., Englund, N., Niu, L., Kaya, Y., Lane, B. G. & Ofengand, J. (2000). Deletion of the Escherichia coli pseudouridine synthase gene truB blocks formation of pseudouridine 55 in tRNA in vivo, does not affect exponential growth, but confers a strong selective disadvantage in competition with wild-type cells. RNA 6, 1870-1881.[Abstract/Free Full Text]

Hiratsu, K., Amemura, M., Nashimoto, H., Shinagawa, H. & Makino, K. (1995). The rpoE gene of Escherichia coli, which encodes {sigma}E, is essential for bacterial growth at high temperature. J Bacteriol 177, 2918-2922.[Abstract]

Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685.[Medline]

Li, J. N. & Bjork, G. R. (1995). 1-Methylguanosine deficiency of tRNA influences cognate codon interaction and metabolism in Salmonella typhimurium. J Bacteriol 177.

Lipinska, B., Sharma, S. & Georgopoulos, C. (1988). Sequence analysis and regulation of the htrA gene of Escherichia coli: a sigma 32-independent mechanism of heat-inducible transcription. Nucleic Acids Res 16, 10053-10067.[Abstract]

Lipinska, B., Fayet, O., Baird, L. & Georgopoulos, C. (1989). Identification, characterization, and mapping of the Escherichia coli htrA gene, whose product is essential for bacterial growth only at elevated temperatures. J Bacteriol 171, 1574-1584.[Medline]

McLaggan, D., Naprstek, J., Buurman, E. T. & Epstein, W. (1994). Interdependence of K+ and glutamate accumulation during osmotic adaptation of Escherichia coli. J Biol Chem 269, 1911-1917.[Abstract/Free Full Text]

Mecsas, J., Rouviere, P. E., Erickson, J. W., Donohue, T. J. & Gross, C. A. (1993). The activity of {sigma}E, an Escherichia coli heat-inducible {sigma}-factor, is modulated by expression of outer membrane proteins. Genes Dev 7, 2618-2628.[Abstract]

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

Milton, D. L., O’Toole, R., Horstedt, P. & Wolf-Watz, H. (1996). Flagellin A is essential for the virulence of Vibrio anguillarum. J Bacteriol 178, 1310-1319.[Abstract]

Mizuno, T., Chou, M.-Y. & Inouye, M. (1984). A unique mechanism regulating gene expression: translational inhibition by a complementary RNA transcript (micRNA). Proc Natl Acad Sci USA 81, 1966-1970.[Abstract]

Nazarenko, I. A., Harrington, K. M. & Uhlenbeck, O. C. (1994). Many of the conserved nucleotides of tRNA(Phe) are not essential for ternary complex formation and peptide elongation. EMBO J 13, 2464-2471.[Abstract]

Nurse, K., Wrzesinski, J., Bakin, A., Lane, B. G. & Ofengand, J. (1995). Purification, cloning, and properties of the tRNA psi 55 synthase from Escherichia coli. RNA 1, 102-112.[Abstract]

Persson, B. C. (1993). Modification of tRNA as a regulatory device. Mol Microbiol 8, 1011-1016.[Medline]

Persson, B. C., Gustafsson, C., Berg, D. E. & Björk, G. R. (1992). The gene for a tRNA modifying enzyme, m5U54-methyltransferase, is essential for viability in Escherichia coli. Proc Natl Acad Sci USA 89, 3995-3998.[Abstract]

Persson, B. C., Jäger, G. & Gustafsson, C. (1997). The spoU gene of Escherichia coli, the fourth gene of the spoT operon, is essential for tRNA(Gm18) 2'-O-methyltransferase activity. Nucleic Acids Res 25, 4093-4097.[Abstract/Free Full Text]

Persson, B. C., Olafsson, O., Lundgren, H. K., Hederstedt, L. & Bjork, G. R. (1998). The ms2io6A37 modification of tRNA in Salmonella typhimurium regulates growth on citric acid cycle intermediates. J Bacteriol 180, 3144-3151.[Abstract]

Piddock, L. J. V., Traynor, E. A. & Wise, R. (1990). A comparison of the mechanisms of decreased susceptibility of aztreonam-resistant and ceftazidime-resistant Enterobacteriaceae. J Antimicrob Chemother 26, 749-762.[Abstract]

Raina, S., Missiakas, D. & Georgopoulos, C. (1995). The rpoE gene encoding the {sigma}E ({sigma}24) heat shock sigma factor of Escherichia coli. EMBO J 14, 1043-1055.[Abstract]

Ramabhadran, T. V. & Jagger, J. (1975). Evidence against DNA as the target for 334 nm-induced growth delay in Escherichia coli. Photochem Photobiol 21, 227-233.[Medline]

Ramamurthy, V., Swann, S. L., Paulson, J. L., Spedaliere, C. J. & Mueller, E. G. (1999). Critical aspartic acid residues in pseudouridine synthases. J Biol Chem 274, 22225-22230.[Abstract/Free Full Text]

Raychaudhuri, S., Niu, L., Conrad, J., Lane, B. G. & Ofengand, J. (1999). Functional effect of deletion and mutation of the Escherichia coli ribosomal RNA and tRNA pseudouridine synthase RluA. J Biol Chem 274, 18880-18886.[Abstract/Free Full Text]

Roe, A. J., McLaggan, D., Davidson, I., O’Byrne, C. & Booth, I. R. (1998). Perturbation of anion balance during inhibition of growth of Escherichia coli by weak acids. J Bacteriol 180, 767-772.[Abstract/Free Full Text]

Rouviere, P. E., De Las Penas, A., Mecsas, J., Lu, C. Z., Rudd, K. E. & Gross, C. A. (1995). rpoE, the gene encoding the second heat-shock sigma factor, {sigma}E, in Escherichia coli. EMBO J 14, 1032-1042.[Abstract]

Rudinger, J., Blechschmidt, B., Ribeiro, S. & Sprinzl, M. (1994). Minimalist aminoacylated RNAs as efficient substrates for elongation factor Tu. Biochemistry 33, 5682-5688.[Medline]

Sage, A. E., Vasil, A. I. & Vasil, M. L. (1997). Molecular characterisation of mutants affected in the osmoprotectant-dependent induction of phospholipase C in Pseudomonas aeruginosa PAO1. Mol Microbiol 23, 43-56.[Medline]

Singer, M., Baker, T. A., Schnitzler, G. & 7 other authors (1989). A collection of strains containing genetically linked alternating antibiotic resistance elements for genetic mapping of Escherichia coli. Microbiol Rev 53, 1–24.

Sprinzl, M., Steegborg, C., Hubel, F. & Steinberg, S. (1996). Compilation of tRNA sequences and sequences of tRNA genes. Nucleic Acids Res 24, 68-72.[Free Full Text]

Stokes, N. R. (2000). Analysis of the function and regulation of mechanosensitive channels in bacteria. PhD Thesis, University of Aberdeen.

Tötemeyer, S., Booth, N. A., Nichols, W. W., Dunbar, B. & Booth, I. R. (1998). From famine to feast: the role of methylglyoxal production in Escherichia coli. Mol Microbiol 27, 553-562.[Medline]

Received 11 March 2002; revised 24 June 2002; accepted 11 July 2002.