The expression profile of Escherichia coli K-12 in response to minimal, optimal and excess copper concentrations

Christopher J. Kershaw, Nigel L. Brown{dagger}, Chrystala Constantinidou, Mala D. Patel and Jon L. Hobman

School of Biosciences, The University of Birmingham, Edgbaston, Birmingham B15 2TT, UK

Correspondence
Jon L. Hobman
J.L.Hobman{at}bham.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
The gene expression profile of Escherichia coli K-12 MG1655 grown in minimal medium supplemented with elevated copper concentrations (as copper-glycine) has been analysed using whole-genome oligonucleotide microarrays. At 750 µM copper-glycine, the expression of both the cue and cus copper-export systems is evident. At near-lethal copper concentrations (2 mM copper-glycine), the expression of these two regulons increases significantly. Other regulons with increased transcription in response to elevated concentrations of copper-glycine include those for the superoxide stress response, iron homeostasis, and envelope stress. Furthermore, a variety of ORFs with decreased expression in response to increased copper-glycine has been identified, including the zinc ABC transporter and genes involved in the chemotactic response.


The GEO accession number for the microarray data reported in this paper is GSE1780.

{dagger}Present address: BBSRC, Polaris House, North Star Avenue, Swindon SN2 1VH, UK.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Copper is a transition metal required by all organisms living in oxic environments. It is an essential cofactor for enzymes involved in electron transfer utilizing dioxygen, and for bacterial nitrate reductases, which reduce nitrate to dinitrogen under anaerobic conditions. At high concentrations copper is toxic to bacteria, eukarya and archaea; therefore the amount of copper in a cell must be precisely controlled (Brown et al., 1994). The toxic effects of copper are twofold. Firstly, excess copper competes with other metals for their binding sites in proteins, resulting in a perturbation of protein function. Secondly, increased cytoplasmic copper concentrations can lead to the generation of highly toxic hydroxyl radicals during redox cycling between Cu(I) and Cu(II):

{mic1511187E001}

The production of these radicals leads to damage to lipids, proteins, DNA and other molecules (Harrison et al., 2000). Copper tolerance is conferred by two genetic systems encoded on the Escherichia coli chromosome, the Cu-sensing (cus) locus (Munson et al., 2000) and the Cu-efflux (cue) regulon (Outten et al., 2000; Rensing et al., 1999). The cus locus contains cusRS, encoding a copper-responsive two-component system that responds to an increase in periplasmic copper concentration. The sensor–regulator pair is responsible for activating the expression of the cusCFBA genes, which encode a copper-efflux protein complex (Munson et al., 2000; Franke et al., 2003). CueR, a copper-activated MerR homologue, regulates the Cu-efflux (cue) system. CueR senses cytoplasmic copper concentrations, and activates transcription of at least two promoters, PcopA and PcueO (Outten et al., 2000; Peterson & Moller, 2000; Stoyanov et al., 2001), in response to increasing copper stress. CopA is a Cu(I)-translocating P-type ATPase, which is the principal copper-efflux ATPase in E. coli, and CueO is a periplasmic multicopper oxidase, which is essential for full copper tolerance (Outten et al., 2000) and has been shown to oxidize enterobactin (Grass et al., 2004) and Cu(I) (Singh et al., 2004).

The cus and cue systems in E. coli both have the ability to confer copper tolerance, but have differing expression profiles. The cue regulon is expressed at basal levels under both aerobic and anaerobic conditions, as is the cus locus, yet at near-lethal copper concentrations the cue regulon is expressed at reduced levels (Outten et al., 2001). Under aerobic conditions, the cue system is activated to confer copper tolerance at moderate and high copper concentrations, whilst the cus locus is only activated at near-lethal copper concentrations (Outten et al., 2001). The cus locus plays an important role in protection against copper stress under anaerobic conditions. During anaerobic growth, the oxidizing power of the periplasmic compartment is decreased. This decrease, coupled with the loss of activity of oxygen-dependent CueO, leads to an increase in the periplasmic Cu(I) concentration, which is proposed to trigger the expression of the cus locus at lower external copper concentrations under anaerobic conditions (Outten et al., 2001).

Although the mechanisms of copper export in E. coli have been elucidated, as yet there is no known copper import mechanism, and previous attempts to define the nature of the import process have been inconclusive (Cotter & Trevors, 1988; C. Doering-Saad & N. L. Brown, unpublished).

Here, we describe the total expression profile of E. coli K-12 MG1655 (CGSC 7740) grown in defined minimal medium at both 750 µM and 2 mM added copper, compared to the expression profile of the cells with no added copper. We show that both the cue and cus regulons are expressed at copper concentrations optimized for growth of E. coli MG1655 in defined medium. We also demonstrate differential expression of other genes in response to secondary effects, such as oxidative stress, caused by elevated copper concentrations.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Strains and growth conditions.
E. coli K-12 MG1655 Seq (CGSC 7740) (Blattner et al., 1997) was used in this study, as this is the strain from which the genome sequence was determined, and to which the oligonucleotide array had been designed. E. coli was cultured in Defined Medium A (DMA: 11·3 g K2PO4, 5·4 g NaH2PO4, 200 mg MgSO4, 10 mg CaCl2, 5 mg FeSO4, 0·5 mg ZnSO4, 0·5 mg MnSO4, 0·1 mg CuSO4, 0·1 mg CoCl2, 0·1 mg sodium borate, 0·1 mg sodium molybdate, 0·26 g EDTA, 2 g NH4Cl per litre; Pirt, 1967). In our experiments the CuSO4 was not included in the medium formulation, but was added as copper-glycine to the cultures grown in DMA supplemented with copper. To limit the amount of copper added as a contaminant of the media components, Aristar-grade potassium and sodium phosphates were used. DMA was supplemented with 0·4 % (w/v) glucose and either no added CuSO4, or CuSO4 to a final concentration of 750 µM and 2 mM. CuSO4 was buffered at pH 7 using 4 mole equivalents of glycine to avoid precipitation of Cu(II) salts. Bacterial cultures were grown in sterile plasticware for small volume cultures, or acid-washed sterile glassware for cultures with volumes greater than 5 ml.

Growth experiments in DMA were performed in triplicate. A 5 µl inoculum from a stationary-phase culture, grown in DMA without copper supplementation, was added to 5 ml growth medium in a sterile plastic universal bottle. These cultures were grown for 10 h at 37 °C with shaking at 150 r.p.m.; 300 µl samples were withdrawn at 2 h intervals and the OD595 was measured in a 96-well plate using a Labsystems Multiskan MS plate reader. Luria–Bertani Medium (LB) (Sambrook et al., 1989) was used when creating electrocompetent cells.

RNA isolation and quantification.
‘Pool’ RNA was isolated from five independently grown 50 ml DMA cultures of E. coli K-12 MG1655. ‘Test’ RNA was isolated from three independently grown 50 ml DMA cultures of E. coli K-12 MG1655, for each growth condition (no added CuSO4, or supplemented with either 750 µM or 2 mM copper, buffered with 3 mM or 8 mM glycine, respectively). A separate control experiment was performed with glycine supplementation equivalent to that used for the copper-glycine conditions (no added glycine, or supplemented with 3 mM or 8 mM glycine). A sample of each culture was taken at the mid- to late-exponential growth phase at OD600 0·8 (measured using a Pharmacia Biotech Ultrospec 2000 in a cuvette with a path length of 10 mm). RNA was isolated using the Qiagen RNAprotect, RNeasy Mini Kit and RNase-free DNase sets according to the manufacturer's instructions. The total RNA isolated from the cultures was quantified using an Agilent Bioanalyser 2100 RNA NanoLabchip, where the quality of the RNA was indicated from the ratio of absorbance of 23S to 16S rRNA, which was between 1·6 and 2.

Oligonucleotide microarray construction and transcriptomic experiments.
The Operon Array Ready E. coli 1.0 microarray oligonucleotide set (Qiagen-Operon) was printed and treated as described by Zhang et al. (2004) and Zheng et al. (2004). For each experiment, 20 µg total RNA was converted to Cy3- and Cy5-labelled cDNA using the CyScribe Post-Labelling Kit (Amersham Biosciences/GE Healthcare) according to manufacturer's instructions. Amino-allyl-labelled nucleotides were incorporated into the cDNA by reverse transcription of the total RNA, followed by a direct chemical coupling of Cy3 and Cy5 NHS-dye esters to the amino-allyl-labelled cDNAs. Prehybridization of the array slides was performed for 4 h in filtered prehybridization solution [25 % formamide, 5x SSC, 10 mg l–1 BSA (fraction V), 0·1 % SDS] at 42 °C. Slides were briefly washed in ethanol and dried by centrifugation at 1000 g for 5 min. Hybridization of the probe was performed using hybridization solution (25 % formamide, 5x SSC, 0·1 % SDS, 0·1 µg poly(A) ml–1, 1x Denhardt's solution and 80 pmol Cy3 and Cy5 combined probe). The hybridization solution containing the Cy-Dye-labelled cDNA was heated to 95 °C for 3 min, Lifter slips (Erie Scientific) were placed on the array slides and the hybridization solution was applied between the Lifter slip and the array slide. The slides were incubated overnight at 42 °C in the dark in a Corning hybridization chamber (Corning Scientific).

After hybridization, slides were washed in wash buffer I (2x SSC/0·1 % SDS) for 2 min at 42 °C, in wash buffer II (0·2x SSC) for 2 min at room temperature and then twice in wash buffer III (0·05x SSC) for 2 min at room temperature. All washes were done with vigorous shaking of the microarray slides. The slides were dried by centrifugation at 1000 g for 5 min, and then analysed using an Axon Genepix 4000A microarray scanner, and Genepix 3.0 software (Axon Scientific).

Semi-quantitative real-time reverse transcriptase PCR.
Real-time PCR (Heid et al., 1996) was performed on 1 ng cDNA template, using SYBR Green master mix (Applied Biosystems) to which primers were added to a final concentration of 50 nM. The cDNA template was produced using Superscript II RNaseH reverse transcriptase (Invitrogen) from the RNA samples used in the DNA microarray experiment. Primers used were designed using Primer Express software supplied with the ABI PRISM 7000 (Applied Biosystems). The primer pairs were as follows. For copA: forward (F), 5'-CAAGCCAGAAATCGGTCAGC-3'; reverse (R), 5'-CAAAGAAATACCAGATTGCCGC-3'. For cueO: F, 5'-TCGCTCACCTTCACTGGTTCAG-3'; R, 5'-TGTAATGCCCGTTCGCTCAA-3'. For cusF: F, 5'-ATGAAACCATGAGCGAAGCACA-3'; R, 5'-CGGATCGTGATGGATGGTGAT-3'. For znuA: F, 5'-TCGCTTCGCTTAAACCCGTT-3'; R, 5'-ATAATCATGTTCTGAAGCGCCG-3'. The melting temperatures of the primers were designed to be 60±1 °C and the amplicon was designed to be 101 bp.

Construction of mutant strains.
Deletion mutations were made in E. coli MG1655 CGSC 7740 as described by Datsenko & Wanner (2000). All mutants were produced by replacing the gene of interest with a copy of the chloramphenicol acetyltransferase (cat) gene amplified from the plasmid pKD3 using the following primers. For copA: F, 5'-TTAAACGCGTGAAAGAAAGTCTTGAACAGCGTCCGGATGTTGAGCATGTAGGCTGGAGCTGCTTCG-3'; R, 5'-GCAGCAACCGGTTGGCGTTACTCACTACGGTAATCGACGAGAGCGCCATTGCATATGAATATCCTCCTTAGT-3'. For cusOP: F, 5'-TTAAGCGGGTAATGTGATAACAAACCTTGTCCCCCGCGCATCCGACGTTATGTAGGCTGGAGCTGCTTCG-3'; R, 5'-ATGTCGGTGCAGCCACATCAGCTTATACGCCGCCGGGATAATAAACAGCGATATGAATATCCTCCTTAGT-3'. For soxS: F 5'-GTCCCATCAGAAAATTATTCAGGATCTTATCGCATGGATTGACGAGCATATGAATATCCTCCTTAGT-3'; R, 5'-TTACAGGCGGTGGCGATAATCGCTGGGAGTGCGATCAAACTGTGTAGGCTGGAGCTGCTTCG-3'. PCR products were made using the Reddy Mix 1.1x PCR master mix (Abgene). Colony PCR using primers flanking each gene that was being replaced, and DIG-labelled cat gene (Roche) Southern blotting, were used to confirm the mutants.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
The expression profile of E. coli in response to copper
To determine the appropriate concentrations of copper to use in this study, E. coli MG1655 was grown in DMA with different concentrations of added CuSO4. Optimal growth occurred at 750 µM copper-glycine, and virtually no growth was observed above 4 mM copper-glycine (Fig. 1). Even with no added CuSO4, MG1655 cells still grew in DMA, albeit poorly, despite the use of Aristar-grade (Merck) phosphates and other Analar-grade chemical components. This may reflect either copper carry-over from the inoculum culture, or that even very low levels of copper present in the media components were sufficient to permit growth. Growth of the cultures supplemented with 0 µM and 2 mM copper-glycine was significantly slower than that of those supplemented with 750 µM copper-glycine (Fig. 2). RNA isolated from E. coli grown in DMA supplemented with 0, 750 µM and 2 mM copper-glycine was hybridized against a ‘pooled’ RNA sample, prepared by pooling RNA isolated from five independent cultures grown in DMA without added copper-glycine. To confirm that the pooled sample was a true representation of the gene expression of E. coli grown without additional copper, RNA isolated in triplicate under the same growth conditions (no added copper-glycine) was reverse-transcribed, labelled and hybridized on the DNA microarray against the pooled sample. The pooled sample was labelled with Cy3 and the test samples with Cy5. Results were expressed as a difference in expression from the no added copper pooled sample; thus a gene whose expression did not alter was defined as having a fold expression change of 1. Performing the experiment in this manner allowed comparison between all three copper concentrations.



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 1. Effect of increasing copper-glycine concentration in DMA upon the growth of E. coli MG1655 (CGSC 7740). Growth is shown as total growth after 10 h at 37 °C with shaking at 150 r.p.m. The experiment was performed in triplicate and error is shown as standard error from the mean (some of the error bars are smaller than the symbols).

 


View larger version (10K):
[in this window]
[in a new window]
 
Fig. 2. Effect of copper concentration on the growth rate of E. coli MG1655 (CGSC 7740). Growth rates were determined in DMA, either unsupplemented ({triangleup}) or supplemented with 750 µM ({square}) or 2 mM copper-glycine ({blacksquare}). The experiment was performed in triplicate and error is shown as standard error from the mean (some of the error bars are smaller than the symbols).

 
To remove the genes whose expression altered in response to added glycine and not copper-glycine, a separate experiment was conducted. Comparison of expression was made between a pooled RNA sample isolated from E. coli cultured without glycine supplementation, and triplicate cultures supplemented with 3 mM and 8 mM glycine. Only 14 genes altered in response to 3 mM or 8 mM glycine (agaD, b1030, b1578, b1825, cydA, entC, gcvH, gcvP, gldA, ilvB, nrdI, osmY, pinO and recJ). Where present, these have been removed from the lists of genes changing in response to copper-glycine, with the exception of entC and cydA. The expression of these two genes increased under elevated copper concentrations, but decreased in response to elevated glycine concentrations.

The data generated by the DNA microarray experiments were analysed using Genespring version 6.1 software (Silicon Genetics). The normalizations performed included Lowess per spot normalization, used to adjust for intensity, biases, and normalization against the pooled RNA sample. ORFs were filtered on expression level (removing genes whose expression failed to alter more than twofold), error (normalized signal intensity from oligonucleotides on the array with a standard deviation from the mean of greater than one were removed), and confidence [using Student's t-test with the multiple test correction Benjamini–Hochberg false discovery rate (Benjamini & Hochberg, 1995) and one-way ANOVA]. All statistical analyses were performed using the Genespring Cross Gene error model. In total, 340 reading frames passed the filtering with a P value less than 0·05. Of the 340 genes with a P value less than 0·05, 70 currently have no known function. The majority of the other genes identified could be grouped into operons and regulons. These microarray data, including the glycine response experiment, are deposited in the NCBI GEO database with the accession number GSE1780.

Copper homeostasis genes
At 750 µM copper-glycine, the optimal copper concentration for growth under the conditions tested (50 ml DMA in a 250 ml conical flask at 37 °C with shaking), both cue- and cus-regulated copper-efflux mechanisms were expressed (Table 1). The expression of the P-type ATPase transporter gene copA and of cusF (the gene encoding the periplasmic copper-binding protein of the cus operon) increased with increasing copper-glycine concentration. Not all cus genes passed the stringent statistical analysis applied to the microarray data, although these genes were observed as increasing in expression in the initial raw data. These data confirm previous studies (Outten et al., 2001) showing that both cusC and copA were expressed at maximal levels under aerobic conditions, in media containing 500 µM copper sulphate. The microarray data described here also show that the expression of copA was 2·5 times greater than cusF at 750 µM copper-glycine, yet only 1·1 times greater than cusF at 2 mM copper-glycine, suggesting that the cus system is expressed when the cue system cannot control the cellular copper levels effectively. These data agree with previously published work (Outten et al., 2001) which showed low-level, constitutive copA expression in the absence of copper and, as the copper concentration increased from 1 µM to 1 mM, copA expression increased 12-fold. In contrast, the cus system was not induced until the external copper concentration increased to 100 µM, with expression of cusC increasing 800-fold over its expression level in the absence of added copper (Outten et al., 2001).


View this table:
[in this window]
[in a new window]
 
Table 1. The putative primary and secondary effects of increased copper on transcription

Primary effects of elevated copper concentrations resulted in increases in the transcription of the two copper-responsive loci. Secondary effects resulted in increases in soxS transcription and an increase in transciption of SoxS-regulated genes. Transcription is expressed as a fold increase in hybridization signal relative to the ‘no added copper’ pooled sample.

 
Our real-time PCR data show that the fold induction of transcription of cusF is greater than that of copA at both 750 µM and 2 mM copper-glycine when compared to the no-copper pooled sample (Fig. 3). This is in agreement with previous studies, where the fold increase in expression from PcusC was greater than the fold increase in expression from PcopA (Outten et al., 2001).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 3. Semi-quantitative real-time PCR analysis of the expression profile of cheA, copA, cueO, cusF, entC and znuA in response to no added copper, 750 µM copper-glycine and 2 mM copper-glycine. Data are shown as relative expression compared to the pooled RNA sample (no added copper) used as the reference for the microarray experiment. The experiment was performed in triplicate and error is shown as standard error from the mean.

 
The CpxRA regulon
Our microarray data indicated that increasing copper concentrations resulted in increased expression of genes involved in ameliorating the deleterious effects of accumulation of misfolded proteins in the periplasm. The cpxP gene, encoding a periplasmic protein, was highly transcribed at 2 mM copper-glycine, suggesting an increase in expression of the cpx regulon (Table 2). The expression of the genes encoding the CpxRA regulator pair, however, showed no significant change. CpxR senses protein misfolding and positively regulates the expression of a number of genes, including degP, encoding a periplasmic protease (Danese et al., 1995), htpX, encoding a membrane protease (Shimohata et al., 2002), ompC, encoding an outer-membrane protein (De Wulf et al., 2002), and ppiD, encoding a peptidyl-prolyl isomerase (Dartigalongue & Raina, 1998). CpxRA also regulates motility and chemotaxis, where CpxR is thought to act directly as a repressor of the transcription of motAB, cheAW and tsr (De Wulf et al., 1999). Our microarray data support these findings (Table 2), with expression of motB, cheAW and tsr decreasing at 2 mM copper. CpxRA may also effect the expression of copA, as there is a potential CpxR-binding site within the copA promoter region, although the effect of CpxR on expression of copA is not known (De Wulf et al., 2002).


View this table:
[in this window]
[in a new window]
 
Table 2. Genes involved in the envelope stress response regulated by CpxR

Expression of genes regulated (or presumably regulated) by CpxR in response to envelope stress. Data are expressed as in Table 1. ORFs that have a predicted CpxR-binding site upstream of their start codon but with no in vivo data are marked with an asterisk (*) (De Wulf et al., 2002).

 
CpxRA also acts through negative regulation of the csg operon. This operon encodes curlin biosynthesis and export (Dorel et al., 1999), expression of which decreased in our microarray data (Table 2). Although the transcription profile of csgF and csgG in response to copper mirrored the predicted regulation of the csg operon, this operon is also under the regulation of OmpR–EnvZ, a two-component regulator pair that did not alter expression under any conditions in the microarray data. It therefore seems likely that increased copper was causing protein misfolding in the periplasm, possibly by producing oxygen radicals, resulting in the increase in CpxR regulation.

Although the regulatory effects of OmpR–EnvZ (an osmolarity response sensor–regulator) upon the csg operon were not apparent, the influence of OmpR–EnvZ upon expression of other ORFs was observed. The influence of copper on OmpR–EnvZ resulted in an increase in ompC expression with increased copper, and OmpR regulation may explain the transcription profile of the flagella biosynthetic operon. The expression of flhC and flhD, genes encoding the transcriptional activators of the flagellar regulon (Liu & Matsumura, 1994), are affected by two of the transcriptional regulators, H-NS (Bertin et al., 1994) and OmpR-EnvZ (Shin & Park, 1995), whose regulatory effect was suggested by our data. Previous studies have shown that a mutation in hns resulted in a complete lack of flagella (Bertin et al., 1994). Although the expression of flhC and flhD did not alter in response to copper, their regulatory effects can be seen in the expression of genes known to be influenced by these two proteins (Table 3). Therefore we predict that at 750 µM copper-glycine flhC and flhD are expressed, possibly through H-NS, and, at the higher copper concentration of 2 mM, OmpR competes with H-NS for the promoter region upstream of flhD, and represses the expression of flhD and flhC, therefore reducing the expression of the flagella biosynthetic operon. Other genes whose regulation is also known to be influenced by H-NS were present within the final gene list (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 3. Flagellar biosynthesis genes known to be under FlhC and FlhD regulation

Expression of genes regulated by the master flagellar biosynthesis regulators FlhD and FlhC in response to optimal and excess copper concentrations. Data are expressed as in Table 1.

 

View this table:
[in this window]
[in a new window]
 
Table 4. Genes known to be under H-NS regulation

These genes have been categorized into up-regulated and down-regulated genes from previous studies. The regulation of genes marked with an asterisk (*) does not correlate with previous data; however, other regulators of these genes may override or modify the effect of H-NS.

 
One of the few regulator-encoding genes that exhibited a change in expression in our microarray data is hns, encoding a nucleoid-associated DNA-binding protein (Schroder & Wagner, 2002). The expression of hns increased with increasing copper concentration. The effects of this increase of H-NS in the cell could be seen in the alteration in expression of several genes when cells were exposed to increasing copper (Table 4). Although the expression of some of the H-NS-regulated genes failed to correlate with previous studies, the raw data of H-NS-regulated genes did appear to correlate with previous data, with 17 of the 54 known H-NS-regulated genes following previously predicted transcription profiles, 27 genes whose expression did not alter and 9 whose expression was different from previous predictions. The known effects of other regulators upon the expression of these genes may explain their unexpected transcription profiles. The apparent downregulation of fliA and fliZ was possibly a consequence of a decrease in expression of flhD and flhC, mediated by OmpR. This decrease in FlhC and FlhD, the positive regulators of the flagellar biosynthesis operon, could have had a greater impact upon the expression of fliA and fliZ than the increase in H-NS. Another ORF that has an expression profile different from the predicted H-NS regulation, fimE, is positively induced by the leucine-responsive regulatory protein (Lrp) (Roesch & Blomfield, 1998). Other genes affected by Lrp expression were present in this gene list; however, their expression profile did not follow the predicted effects of Lrp.

Superoxide-response genes
A secondary effect of excess copper is oxygen radical production caused by the redox cycling of Cu(I) to Cu(II). Previous work (Kimura & Nishioka, 1997) has shown that the addition of CuSO4 to the growth medium causes superoxide-induced redox stress in E. coli, which was confirmed by our microarray data. Superoxide stress is sensed by the SoxR protein, which, when activated, functions as a potent activator of soxS transcription.

Our data show that the transcription of members of the SoxRS regulon also increased in response to 2 mM copper-glycine (Table 1). The regulatory effects of SoxS could be seen by an increase in expression of fldA, encoding flavodoxin 1, fur (Zheng et al., 1999), encoding an iron regulator, and zwf (Fawcett & Wolf, 1995), encoding glucose-6-phosphate-1-dehydrogenase, all of which occurred in response to 2 mM copper-glycine. Other members of the SoxRS regulon were not present in the final gene list, due to the stringent statistical analysis applied to the raw data. However, the raw data of SoxRS-regulated genes appeared to correlate with previous data, with 13 of the 22 known SoxRS-regulated genes increasing in response to 2 mM copper glycine.

Copper and iron-regulated gene expression
The expression of the iron regulator gene, fur, is also known to be induced by SoxS (Zheng et al., 1999) and fur expression increased in response to 2 mM copper-glycine (Table 5). However, the regulatory effects of Fur were not as clear as those for SoxS or CpxRA in our microarray data (Table 5). It is known that the ferric-uptake regulator, Fur, represses the expression of various genes in response to an increase in ferric iron (McHugh et al., 2003). Our data show that expression of fur increased at 2 mM copper-glycine compared to 750 µM copper-glycine, and no added copper. The increase in Fur concentration in the cell may have caused the observed decrease in expression of cheZ, cydA, cyoBC, nikC, nuoA–M and ycdO, all of which are known to be directly regulated by Fur.


View this table:
[in this window]
[in a new window]
 
Table 5. The expression of genes known to be involved in iron homeostasis

Iron- and Fur-regulated genes have been categorized into three separate functional pathways: Fe metabolism, energy metabolism, and miscellaneous function (McHugh et al., 2003). Genes without a Fur-binding site, suggesting indirect regulation by Fur, are marked with an asterisk (*); data from McHugh et al. (2003). Data are expressed as in Table 1.

 
The main physiological function of Fur is to repress iron-acquisition genes, although in our microarray data as fur expression increased, so did enterobactin expression. Previously it has been shown that Fur represses the transcription of the enterobactin operon (Brickman et al., 1990). However, expression of genes involved in the synthesis and uptake of enterobactin increased when E. coli was exposed to 2 mM copper-glycine (Table 5), suggesting that there may be another layer of regulation of these genes. The increase in expression of genes involved in iron import may be due to one of the following. First, there is a requirement for increased levels of ferric iron in E. coli, possibly due to competition between copper and iron for iron-binding sites on proteins, overriding signals of iron sufficiency. Second, the increase in the expression of the enterobactin biosynthesis and uptake operon may be due to increased copper-dependent expression of CueO, which oxidizes enterobactin and decreases iron uptake (Grass et al., 2004). The increased expression of the enterobactin biosynthetic and uptake genes in our microarray data would thus reflect the requirement of the cell for increased levels of functional enterobactin. Third, the increase in the transcription of the enterobactin synthesis and uptake genes may imply a more direct role in copper tolerance. The product of enterobactin oxidation, 2-carboxymuconate, sequesters copper and may act as an extracellular copper sink (Grass et al., 2004).

The suf operon, represented by sufA and sufD in our microarray data, is also regulated by Fur and is involved in iron–sulphur cluster formation (Table 5). As fur increased in expression in response to 2 mM copper, the same response profile would be expected by the Fur-regulated genes, sufA and sufD. However, our data showed that the transcription of these two genes increased greatly in response to 2 mM copper, which is the opposite of the anticipated result. This increase in expression may have been due to the increase in superoxide stress caused by the redox cycling between Cu(I) and Cu(II); the cysteine desulfurase, SufS, is required under oxidative stress for the activity of proteins containing labile iron–sulphur clusters (Nachin et al., 2003). It has been suggested that copper and iron homeostasis may be linked in yeast, where iron import is linked to copper homeostasis (Gross et al., 2000; Roberts et al., 2002), and our data indicated that this may also be the case in E. coli.

Copper and zinc homeostasis
The expression of another metal-responsive operon, znuABC, decreased in response to increasing copper concentrations in the microarray experiments (data not shown). znuABC encodes a zinc ABC transporter which imports zinc into E. coli (Patzer & Hantke, 1998). The decreased expression of this operon in response to excess copper may be due to cross-talk with the zinc-sensing systems, as copper may out-compete zinc for the metal-binding site of Zur, the negative zinc regulator shown to represses the expression of znuB and znuC (Patzer & Hantke, 2000). Alternatively, under limiting copper concentrations ZnuABC may act as a copper-import system by the non-specific import of copper in tandem with zinc import. A specific copper-uptake system has not been identified in E. coli (C. Doering-Saad & N. L. Brown, unpublished).

Real-time PCR of selected ORFs
The microarray data showed that in E. coli, there were different whole-genome transcriptional profiles in response to the three different experimental copper concentrations. Real-time PCR was performed on the cheA, copA, cusF, entC and znuA reading frames (Fig. 3), representing each of the different types of response profile detected, in order to check the DNA microarray data. The real-time PCR expression profiles of these genes confirmed the data from the DNA microarray experiment, in that they showed the same response profile.

The cueO gene is known to be regulated by CueR but was not one of the genes induced by increased external Cu(II) exposure of E. coli in the microarray experiment. Real-time PCR was performed on this gene, which showed that cueO and copA have similar expression profiles, as expected from their regulation by CueR. The increase in expression of cueO in response to an increase in copper concentration seen in real-time PCR supports previous findings, which suggest a role for CueO in enterobactin oxidation (Grass et al., 2004). The microarray data did not show induction of cueO, because the raw data for this gene failed to pass the stringent statistical analysis applied to the array data.

Growth of mutants in DMA supplemented with copper-glycine
To further validate the microarray data, mutations were made in E. coli MG1655, in which copA, genes from the cus operon and soxS were replaced with a cat gene cassette. The growth of each of these gene-replacement strains in 5 ml DMA supplemented with no added copper, or with 750 µM or 2 mM copper-glycine, was investigated (Fig. 4). The copA : : cat strain failed to grow at both 750 µM and 2 mM copper-glycine, indicating that the interpretation of the microarray data was correct. The drastic growth effect of a deletion of copA at both 750 µM and 2 mM copper-glycine can be explained by the pivotal role of this gene in the protection of the cytoplasmic compartment from copper toxicity. Similarly the soxS : : cat strain failed to grow at 2 mM copper-glycine, although this strain showed the same growth phenotype as the parental strain when exposed to 750 µM copper. This suggests that, if both the cus and cue systems are functional, then the intracellular level of copper, and hence copper-induced superoxide stress, is limited to concentrations that the cell can tolerate when no superoxide dismutases and other SoxS regulon genes are expressed. At both 750 µM and 2 mM copper, the cusRSCFBA : : cat strain had a similar growth profile to the parental strain. This is probably due to the cue system being the primary determinant of copper tolerance in aerobic growth, which would mask the effect of disruption of the cus operon.



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 4. Effect of copper-glycine concentration in DMA upon growth profiles of E. coli MG1655 ({triangleup}) {Delta}copA ({square}) {Delta}cusRSCFBA ({blacktriangleup}) and {Delta}soxS ({blacksquare}) with no added copper (a), 750 µM copper-glycine (b) and 2 mM copper-glycine (c). The experiment was performed in triplicate and error is shown as standard error from the mean (some of the error bars are smaller than the symbols).

 
The apparent difference between the growth observed in Fig. 1 and Fig. 4 is due to different inoculum sizes. The inocula were standardized within an experiment but not between experiments. As growth was determined as total growth after 10 h this difference in inoculum size resulted in an observed effect.

Concluding remarks
This study, which is believed to represent the first total genome expression profile of E. coli at three different copper concentrations, shows that excess copper leads to transcriptional responses of genes relating to direct copper export, CpxRA-regulated response to protein misfolding or envelope stress, and superoxide stress. Our data also suggest that cellular responses to copper stress within E. coli are linked with other metal ion import regulons, specifically those involved in iron and zinc homeostasis. Iron, copper and zinc are three biologically important metal ions required for growth. However, all three are toxic in excess and must be precisely controlled. Our data indicate that there is cross-talk between the homeostatic mechanisms for these three metal ions. This may be in order to reduce competition between these ions for essential metal-binding sites in proteins or may be due to lack of complete discrimination. The benefit of absolute discrimination between these metals may be too expensive metabolically. Furthermore, co-homeostasis may provide a mode of import for copper ions, as no candidate specific copper-uptake system has been identified in our microarray data or in proteomics data (C. Doering-Saad & N. L. Brown, unpublished). The possibility that copper import may occur non-specifically through the zinc ABC transport system (ZnuABC) requires further investigation.

These transcriptional data show the expression profile of E. coli in response to limiting, optimal and excess copper concentrations. The experimental design was such that the limiting and excess copper concentrations were determined by an alteration in the growth rate of E. coli, and this will inevitably have an effect upon the transcriptional data. However, these alterations in growth rate are a direct result of limiting and excess copper concentrations, and the transcriptional data rightly reflect this as a response to growth under these conditions. This alteration in growth rate could also affect oxygen transfer rates and other growth parameters, which may be reflected in the gene expression data.

This study has shown that continuous exposure to an increased extracellular concentration of one metal ion results in altered expression of other metal-homeostatic genes. Investigation of the transcriptional response to excess concentrations of other metal ions may also result in similar perturbations; future investigations into the homeostatic control of a single essential metal ion in isolation should take this into consideration.

The effect of adaptation of the transcriptome to near-toxic and insufficient copper concentrations has led to the identification of many ORFs whose functions are as yet unknown. Further gene-disruption studies and subsequent microarray analysis and other molecular biological approaches need to be performed to assign functions to these genes and to determine whether they have primary or secondary roles in the copper response.


   ACKNOWLEDGEMENTS
 
This work was supported by BBSRC grants EGA16107and JIF13209to the University of Birmingham and G14071 to N. L. B. C. J. K. was supported by a BBSRC Studentship. We would also like to thank staff of the Functional Genomics Laboratory, School of Biosciences, University of Birmingham, for technical support.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Barth, M., Marschall, C., Muffler, A., Fischer, D. & Hengge-Aronis, R. (1995). Role for the histone-like protein H-NS in growth phase-dependent and osmotic regulation of {sigma}s and many {sigma}s-dependent genes in Escherichia coli. J Bacteriol 177, 3455–3464.[Abstract/Free Full Text]

Benjamini, Y. & Hochberg, Y. (1995). Controlling the false discovery rate – a practical and powerful approach to multiple testing. J R Stat Soc Ser B Methodol 57, 289–300.

Bertin, P., Terao, E., Lee, E. H., Lejeune, P., Colson, C., Danchin, A. & Collatz, E. (1994). The H-NS protein is involved in the biogenesis of flagella in Escherichia coli. J Bacteriol 176, 5537–5540.[Abstract]

Blattner, F. R., Plunkett, G., Bloch, C. A. & 14 other authors (1997). The complete genome sequence of Escherichia coli K-12. Science 277, 1453–1469.[Abstract/Free Full Text]

Brickman, T. J., Ozenberger, B. A. & McIntosh, M. A. (1990). Regulation of divergent transcription from the iron-responsive fepB-entC promoter-operator regions in Escherichia coli. J Mol Biol 212, 669–682.[CrossRef][Medline]

Brown, N. L., Lee, B. T. O. & Silver, S. (1994). Bacterial transport of and resistance to copper. In Metal Ions in Biological Systems, pp. 405–434. Edited by H. Sigel & A. Sigel. New York: Marcel Dekker.

Browning, D. F., Cole, J. A. & Busby, S. J. W. (2000). Supression of FNR-dependent transcription activation at the Escherichia coli nir promoter by Fis, IHF and H-NS: modulation of transcription initiation by a complex nucleo-protein assembly. Mol Microbiol 37, 1258–1269.[CrossRef][Medline]

Cotter, C. & Trevors, J. T. (1988). Copper adsorption by Escherichia coli. Syst Appl Microbiol 10, 313–317.

Danese, P. N. & Silhavy, T. J. (1998). CpxP, a stress-combative member of the Cpx regulon. J Bacteriol 180, 831–839.[Abstract/Free Full Text]

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

Dartigalongue, C. & Raina, S. (1998). A new heat-shock gene, ppiD, encodes a peptidyl-prolyl isomerase required for folding of outer membrane proteins in Escherichia coli. EMBO J 17, 3968–3980.[Abstract/Free Full Text]

Datsenko, K. A. & Wanner, B. L. (2000). One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97, 6640–6645.[Abstract/Free Full Text]

De Wulf, P., Kwon, O. & Lin, E. C. C. (1999). The CpxRA signal transduction system of Escherichia coli: growth related autoactivation and control of unanticipated target operons. J Bacteriol 181, 6772–6778.[Abstract/Free Full Text]

De Wulf, P., McGuire, A. M., Liu, X. & Lin, E. C. C. (2002). Genome-wide profiling of promoter recognition by the two-component response regulator CpxR-P in Escherichia coli. J Biol Chem 277, 26652–26661.[Abstract/Free Full Text]

Dorel, C., Vidal, O., Prigent-Combaret, C., Vallet, I. & Lejeune, P. (1999). Involvement of the Cpx signal transduction pathway of E. coli in biofilm formation. FEMS Microbiol Lett 178, 169–175.[CrossRef][Medline]

Fawcett, W. P. & Wolf, R. E., Jr (1995). Genetic definition of the Escherichia coli zwf "Soxbox", the DNA binding site for SoxS-mediated induction of glucose 6-phosphate dehydrogenase in response to superoxide. J Bacteriol 177, 1742–1750.[Abstract/Free Full Text]

Franke, S., Grass, G., Rensing, C. & Nies, D. H. (2003). Molecular analysis of the copper-transporting efflux system CusCFBA of Escherichia coli. J Bacteriol 185, 3804–3812.[Abstract/Free Full Text]

Free, A. & Dorman, C. J. (1997). The Escherichia coli stpA gene is transiently expressed during growth in rich medium and is induced in minimal medium and by stress conditions. J Bacteriol 179, 909–918.[Abstract/Free Full Text]

Govantes, F., Orjalo, A. V. & Gunsalus, R. P. (2000). Interplay between three global regulatory proteins mediates oxygen regulation of the Escherichia coli cytochrome d oxidase (cydAB) operon. Mol Microbiol 38, 1061–1073.[CrossRef][Medline]

Grass, G., Thakali, K., Klebba, P. E., Thieme, D., Muller, A., Wildner, G. F. & Rensing, C. (2004). Linkage between catecholate siderophores and the multicopper oxidase CueO in Escherichia coli. J Bacteriol 186, 5826–5833.[Abstract/Free Full Text]

Gross, C., Kelleher, M., Iyer, V. R., Brown, P. O. & Winge, D. R. (2000). Identification of the copper regulon in Saccharomyces cerevisiae by DNA microarrays. J Biol Chem 275, 32310–32316.[Abstract/Free Full Text]

Harrison, M. D., Jones, C. E., Solioz, M. & Dameron, C. T. (2000). Intracellular copper routing: the role of copper chaperones. Trends Biochem Sci 25, 29–32.[CrossRef][Medline]

Heid, C. A., Stevens, J., Livak, K. J. & Williams, P. M. (1996). Real time quantitative PCR. Genome Research 6, 986–994.[Abstract]

Kimura, T. & Nishioka, H. (1997). Intracellular generation of superoxide by copper sulphate in Escherichia coli. Mutat Res 389, 237–242.[Medline]

Liu, X. & Matsumura, P. (1994). The FlhD/FlhC complex, a transcriptional activator of the Escherichia coli flagellar class II operons. J Bacteriol 176, 7345–7351.[Abstract]

Lucht, J. M., Dersch, P., Kempf, B. & Bremer, E. (1994). Interactions of the nucleoid-associated DNA-binding protein H-NS with the regulatory region of the osmotically controlled proU operon of Escherichia coli. J Biol Chem 269, 6578–6586.[Abstract/Free Full Text]

McHugh, J. P., Rodriguez-Quinones, F., Abdul-Tehrani, H., Svistunenko, D. A., Poole, R. K., Cooper, C. E. & Andrews, S. C. (2003). Global iron-dependent gene regulation in Escherichia coli – a new mechanism for iron homeostasis. J Biol Chem 278, 29478–29486.[Abstract/Free Full Text]

Munson, G. P., Lam, D. L., Outten, F. W. & O'Halloran, T. V. (2000). Identification of a copper-responsive two-component system on the chromosome of Escherichia coli K-12. J Bacteriol 182, 5864–5871.[Abstract/Free Full Text]

Nachin, L., Loiseau, L., Expert, D. & Barras, F. (2003). SufC: an unorthodox cytoplasmic ABC/ATPase required for [Fe–S] biogenesis under oxidative stress. EMBO J 22, 427–437.[Abstract/Free Full Text]

Olsen, P. B. & Klemm, P. (1994). Localization of promoters in the fim gene cluster and the efffect of H-NS on the transcription of fimB and fimE. FEMS Microbiol Lett 116, 95–100.[CrossRef][Medline]

Outten, F. W., Outten, C. E., Hale, J. & O'Halloran, T. (2000). Transcriptional activation of an Escherichia coli copper efflux regulon by the chromosomal MerR homologue, CueR. J Biol Chem 275, 31024–31029.[Abstract/Free Full Text]

Outten, F. W., Huffman, D. L., Hale, J. A. & O'Halloran, T. V. (2001). The independent cue and cus systems confer copper tolerance during aerobic and anaerobic growth in Escherichia coli. J Biol Chem 276, 30670–30677.[Abstract/Free Full Text]

Patzer, S. I. & Hantke, K. (1998). The ZnuABC high-affinity zinc uptake system and its regulator Zur in Escherichia coli. Mol Microbiol 28, 1199–1210.[CrossRef][Medline]

Patzer, S. I. & Hantke, K. (2000). The Zinc-responsive regulator Zur and its control of the znu gene cluster encoding the ZnuABC zinc uptake system in Escherichia coli. J Biol Chem 275, 24321–24332.[Abstract/Free Full Text]

Peterson, C. & Moller, L. B. (2000). Control of copper homeostsis in Escherichia coli by a P-type ATPase, CopA, and a MerR-like transcriptional activator, CopR. Gene 261, 289–298.[CrossRef][Medline]

Pirt, S. J. (1967). A kinetic study of the mode of growth of surface colonies of bacteria and fungi. J Gen Microbiol 47, 181–197.[Medline]

Rensing, C., Fan, B., Sharma, R., Mitra, B. & Rosen, B. P. (1999). CopA: an Escherichia coli Cu(I)-translocating P-type ATPase. Proc Natl Acad Sci U S A 97, 652–656.[CrossRef]

Roberts, S. A., Weichsel, A., Grass, G., Thakali, K., Hazzard, J. T., Tollin, G., Rensing, C. & Montfort, W. R. (2002). Crystal structure and electron transfer kenetics of CueO, a multicopper oxidase required for copper homeostasis in Escherichia coli. Proc Natl Acad Sci U S A 99, 2766–2771.[Abstract/Free Full Text]

Roesch, P. L. & Blomfield, I. C. (1998). Leucine alters the interaction of the leucine-responsive regulatory protein (Lrp) with the fim switch to stimulate site-specific recombination in Escherichia coli. Mol Microbiol 27, 751–761.[CrossRef][Medline]

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

Schembri, M. A., Olsen, P. B. & Klemm, P. (1998). Orientation-dependent enhancement by H-NS of the activity of the type 1 fimbrial phase switch promoter in Escherichia coli. Mol Gen Genet 259, 336–344.[CrossRef][Medline]

Schroder, O. & Wagner, R. (2002). The bacterial regulatory protein H-NS – a versatile modulator of nucleic acid structures. Biol Chem 383, 945–960.[CrossRef][Medline]

Shimohata, N., Chiba, S., Saikawa, N., Ito, K. & Akiyama, Y. (2002). The Cpx stress response system of Escherichia coli senses plasma membrane proteins and controls HtpX, a membrane protease with a cytosolic active site. Genes Cells 7, 653–662.[Abstract/Free Full Text]

Shin, S. & Park, C. (1995). Modulation of flagellar expression in Escherichia coli by acetyl phosphate and the osmoregulator OmpR. J Bacteriol 177, 4696–4702.[Abstract/Free Full Text]

Singh, S. K., Grass, G., Rensing, C. & Montfort, W. R. (2004). Cuprous oxidase activity of CueO from Escherichia coli. J Bacteriol 186, 7815–7817.[Abstract/Free Full Text]

Stoyanov, J. V., Hobman, J. L. & Brown, N. L. (2001). CueR (YbbI) of Escherichia coli is a MerR family regulator controlling expression of the copper exporter CopA. Mol Microbiol 39, 502–511.[CrossRef][Medline]

Tramonti, A., Visca, P., De Canio, M., Falconi, M. & De Biase, D. (2002). Functional characterization and regulation of gadX, a gene encoding an AraC/XylS-like transcriptional activator of the Escherichia coli glutamic acid decarboxylase system. J Bacteriol 184, 2603–2613.[Abstract/Free Full Text]

Wu, J. & Weiss, B. (1992). Two-stage induction of the soxRS (superoxide response) regulon of Escherichia coli. J Bacteriol 174, 3915–3920.[Abstract]

Zhang, L., Chaudhuri, R. R., Constantinidou, C. & 12 other authors (2004). Regulators encoded in the ETT2 gene cluster influence expression of genes within the locus of enterocyte effacement in enterohemorrhagic O157 : H7. Infect Immun 72, 7282–7293.[Abstract/Free Full Text]

Zheng, M., Doan, B., Schneider, T. D. & Storz, G. (1999). OxyR and SoxRS regulation of fur. J Bacteriol 181, 4639–4643.[Abstract/Free Full Text]

Zheng, D., Constantinidou, C., Hobman, J. L. & Minchin, S. D. (2004). Identification of the CRP regulon using in vitro and in vivo transcriptional profiling. Nucleic Acids Res 32, 5874–5893.[Abstract/Free Full Text]

Received 23 September 2004; revised 8 December 2004; accepted 22 December 2004.



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 Kershaw, C. J.
Articles by Hobman, J. L.
Articles citing this Article
PubMed
PubMed Citation
Articles by Kershaw, C. J.
Articles by Hobman, J. L.
Agricola
Articles by Kershaw, C. J.
Articles by Hobman, J. L.


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 © 2005 Society for General Microbiology.