OUP user menu

New members of the Escherichia coli σE regulon identified by a two-plasmid system

Bronislava Rezuchova, Henrieta Miticka, Dagmar Homerova, Mark Roberts, Jan Kormanec
DOI: http://dx.doi.org/10.1016/S0378-1097(03)00480-4 1-7 First published online: 1 August 2003


A previously established method, based on a two-plasmid system, was used to identify promoters recognized by RNA polymerase containing the extracytoplasmic stress response sigma factor σE in Escherichia coli. In addition to previously identified rpoE-dependent promoters, 11 new promoters potentially directing the expression of 15 genes were identified that were active only after over-expression of rpoE. The promoters were confirmed and transcriptional start points of the promoters were determined by primer extension analysis and S1-nuclease mapping. All the promoters contained sequences similar to the consensus sequence of rpoE-dependent promoters. The new rpoE-dependent promoters governed expression of genes encoding proteins involved in primary metabolism (fusA, tufA, recR), phospholipid and lipopolysaccharide biosynthesis (psd, lpxP), signal transduction (sixA), proposed inner or outer membrane proteins (bacA, sbmA, smpA, yeaY), and proteins with unknown function (ybaB, yaiW, yiiS, yiiT, yfeY).

  • Promoter
  • RNA polymerase
  • Sigma factor
  • rpoE
  • Stress response

1 Introduction

Escherichia coli has to adapt to a variety of extracytoplasmic stresses in the bacterial envelope (periplasm, inner and outer membranes). The extracytoplasmic function sigma factor σE together with two other signal transduction pathways, CpxRA and BaeSR, govern expression of genes involved in the envelope stress responses [1,2]. The rpoE gene, encoding σE, is located in the σE-autoregulated operon including three downstream genes, rseA, rseB and rseC, encoding proteins involved in the post-translational regulation of σE[3,4]. The activity of σE is negatively controlled by a membrane-bound anti-sigma factor RseA, which sequesters σE in unstressed cells. In response to outer membrane protein folding perturbations, RseA is proteolytically inactivated by the sequential action of the proteases DegS and YaeL. This releases σE into cytoplasm where it can interact with RNA polymerase core enzyme [1,5]. The resulting holoenzyme (EσE) directs expression of genes in the σE regulon. Originally, two-dimensional gel electrophoresis suggested at least 10 members of the σE regulon, of which four were identified. They are: rpoH, encoding heat-shock sigma factor σ32; rpoE, encoding σE; degP (htrA), encoding a periplasmic protease; and fkpA, encoding a periplasmic peptidyl prolyl isomerase [1,3,4]. Recently, new members of the σE regulon have been identified, increasing its number to 43, including regulatory proteins, proteins involved in lipopolysaccharide (LPS) biosynthesis/transport, periplasmic proteases and folding factors, and several proteins with unknown function [6].

In order to identify missing members of the σE regulon, we used a recently developed method for the identification of promoters recognized by a particular sigma factor of RNA polymerase. The method was based on two compatible E. coli plasmids. The first plasmid, pAC-rpoE4, has the E. coli rpoE gene under the control of the arabinose-inducible PBAD promoter. Following induction with arabinose, σE accumulates in the cytoplasm and interacts with RNA polymerase core enzyme to form EσE, which is able to recognize a σE-cognate promoter present in a library of chromosomal fragments cloned upstream of a promoter-less lacZα reporter gene in the second compatible plasmid, pSB40 [7]. The system was verified using two known rpoE-dependent promoters, the rpoEp2 directing the rpoE gene and the degPp directing the degP gene encoding a periplasmic protease [7].

The present paper describes the use of this new system for identification of new members of the E. coliσE regulon. In addition to the recently identified rpoE-dependent promoters [6], 11 new promoters dependent upon σE in E. coli were identified and characterized. The promoters can potentially direct the expression of 15 genes and the function or proposed function of these genes is discussed.

2 Materials and methods

2.1 Bacterial strains, plasmids and culture conditions

Wild-type E. coli W3100 [8] was used for chromosomal DNA preparation. The E. coli promoter-probe plasmid pSB40 [9] was kindly provided by Dr. M.K. Winson, University of Nottingham, UK. The E. coli expression plasmid pAC7 is described in [7]. Construction of the plasmid pAC-rpoE4, containing the E. coli rpoE gene under the control of the arabinose-inducible PBAD promoter, is described in [7]. E. coli XL1Blue (Stratagene) was used as a host for cloning experiments. Conditions for E. coli growth and transformation were as described in [10].

2.2 DNA manipulations

DNA manipulations in E. coli were performed as described in [10]. Nucleotide sequencing was performed by the chemical method [11] and by the dideoxy chain termination method [12], using the TaqTrack™ kit (Promega).

2.3 Detection of E. coli clones containing the rpoE-dependent promoter fragment

An E. coli W3100 genomic library (prepared by cloning 0.6–1.2-kb partial Sau3AI chromosomal fragments into the BamHI site of the promoter-probe plasmid pSB40) was transformed into E. coli XL1Blue containing the compatible plasmid pAC-rpoE4 [7]. The clones were selected on LBACX plates (LB medium with 5 g l−1 lactose, 100 µg ml−1 ampicillin, 40 µg ml−1 chloramphenicol, 20 µg ml−1 X-gal) with 2 µg ml−1 arabinose as described previously [7]. The colonies were screened after 24 h growth at 37°C. Blue clones were inoculated in parallel onto two LBACX plates containing either 2 µg ml−1 arabinose (LBACX-ARA) or 2 mg ml−1 glucose (LBACX-GLU). Clones that were blue on LBACX-ARA and white on LBACX-GLU were inoculated into 1 ml LB+Ap (100 µg ml−1) liquid medium and grown overnight at 37°C. Cells were pelleted, suspended in 200 µl STE buffer (0.1 M NaCl, 10 mM Tris–HCl pH 8, 1 mM EDTA) with 0.5 mg ml−1 lysozyme, incubated for 5 min at room temperature, boiled for 1.5 min, and centrifuged for 10 min at 13 000 rpm. 1 µl of supernatant was transformed in parallel into E. coli XL1 Blue strains harboring either pAC-rpoE4 or pAC7, and plated on LBACX-ARA.

2.4 Isolation of RNA and S1 nuclease mapping

Total RNA was prepared from cultures of E. coli and high-resolution S1 nuclease mapping was performed as previously described [13]. Samples (40 µg) of RNA (estimated spectrophotometrically) were hybridized to approximately 0.02 pmol of a suitable DNA probe labelled at the 5′ end with [γ-32P]ATP (approximately 3×106 cpm pmol−1 of a probe). The probes used were prepared by polymerase chain reaction amplification from the corresponding pE plasmids using the 5′ end-labelled universal oligonucleotide primer −47 (5′-CGCCAGGGTTTTCCCAGTCACGAC-3′) from the lacZα coding region, and primer mut80 (5′-GGGTTCCGCGCACATTTCCCCG-3′) from the 5′ region flanking the polylinker of pSB40. The protected DNA fragments were analyzed on DNA sequencing gels together with G+A and T+C sequencing ladders derived from the end-labelled fragments [11].

2.5 Primer extension analysis

80 µg of total RNA was dissolved in 60 µl of hybridization buffer (40 mM PIPES pH 6.4, 1 mM EDTA, 0.4 M NaCl, 80% (v/v) formamide) at 65°C, denatured together with 0.5 pmol primer DNA (universal oligonucleotide primer −47) for 5 min at 95°C, and annealed for 4 h at 45°C. DNA sample was ethanol-precipitated, dissolved in 7 µl of water and following components were added: 0.75 µl of RNasin (Promega), 3 µl of 5×AMV-RT buffer (Promega), 0.75 µl 5 mM each of dATP, dGTP, dTTP and 0.5 mM dCTP, 0.75 µl of 4 mg ml−1 actinomycin, 1 µl [α-35S]dCTP (1200 Ci mmol−1, ICN), and the mixture was incubated for 2 min at 42°C. The primer extension was initiated by adding 2 µl (18 U) of AMV-RT (Promega) and incubated for 2 h at 42°C. The reaction was terminated with 25 µl of RNase mix (100 µg ml−1 DNase-free RNase A, 30 µg ml−1 sonicated salmon sperm DNA, 10 mM TE, pH 8) and incubated for 30 min at 37°C. After addition of 20 µl of 1 M NaCl, the mixture was extracted with alkaline phenol/chloroform and DNA precipitated with ethanol. The pellet was dissolved in 5 µl of loading buffer (80% (v/v) formamide, 10 mM NaOH, 1 mM EDTA, 0.05% xylene cyanol, 0.05% bromophenol blue), heated for 2 min at 95°C, and an aliquot was loaded on a 6% denaturing gel and separated together with the sequencing ladders generated using the same primer used for the primer extension reaction [12].

3 Results and discussion

3.1 Using the two-plasmid system to identify promoters recognized by EσE

In order to identify new E. coliσE-dependent promoters, we recently described an optimized E. coli two-plasmid screening system. The system was successfully verified using two well-characterized E. coli rpoE-dependent promoters, rpoEp2 and degPp, that were expressed only after arabinose induction of rpoE [7]. Moreover, the transcription start points (tsp) of these two rpoE-dependent promoters in the E. coli two-plasmid system after induced expression of rpoE were identical to that published previously (Fig. 2A). Therefore this system could be used for identification of other rpoE-dependent promoters in E. coli. After screening of ∼140 000 colonies of the E. coli W3100 library (Section 2) on LBACX-ARA plates, 2400 blue colonies that represented all active promoters (including rpoE-dependent promoters) were selected for further study. To identify the clones possessing rpoE-dependent promoters, the clones were inoculated onto both LBACX-ARA and LBACX-GLU plates. Three hundred and twenty-five clones, blue colonies on LBACX-ARA and white colonies on LBACX-GLU, were identified. However, these clones may also include rpoE-independent promoters that are repressed by glucose. To eliminate such clones, plasmids were isolated from the 325 clones and transformed in parallel into E. coli XL1Blue strains containing pAC7 or pAC-rpoE4 and colonies were screened on LBACX-ARA plates. Clones containing plasmids with rpoE-dependent promoters were blue in E. coli XL1Blue with pAC-rpoE4 and white in E. coli XL1Blue containing pAC7. Clones with rpoE-independent constitutive promoters remained blue in both strains. Using this screen, we identified 146 positive clones containing rpoE-dependent promoters. Restriction mapping, dot blot hybridization and sequence analysis of the DNA fragments cloned in the plasmids isolated from these positive clones revealed 27 representative DNA fragments (plasmids pE1–27). Employing restriction sites, deletion analyses of these fragments were performed to locate the rpoE-dependent promoters to DNA fragments of minimal size (150–300 bp).

Figure 2

A: Nucleotide sequence alignment of the new E. coli rpoE-dependent promoters and their comparison to the two canonical rpoE-dependent promoters, rpoEp2 and degPp. The corresponding −10 and −35 regions are depicted in bold. The tsp is in bold and underlined. The consensus sequence is shown below the alignment. Bases of the consensus sequence conserved in more than 80% of the sequences are in capital letters. B: DNA sequences surrounding the putative CpxR-P binding sites upstream of three rpoE-dependent genes (psd, bacA, smpA) [14]. Proposed CpxR-P DNA recognition boxes [14] are in bold and underlined.

3.2 Characterization of the new rpoE-dependent promoters

While this work was in progress, a different genetic approach was used to identify the E. coliσE regulon, and 20 rpoE-dependent promoters were identified [6]. Therefore, we compared the 27 promoters identified by our screen to those previously characterized rpoE-dependent promoters and found that, out of these 27 rpoE-dependent promoters, 11 corresponded to new promoters. In order to identify tsp of these new rpoE-dependent promoters, high-resolution S1 nuclease mapping and primer extension analysis were performed using RNA isolated from E. coli containing a pE plasmid bearing a particular rpoE-dependent promoter and pAC-rpoE4 grown to exponential phase and induced by arabinose. The 5′-labelled −47 primer for primer extension analysis and 5′-labelled DNA probes for S1 nuclease mapping are described in Section 2. These experiments are documented in Fig. 1, and the results are summarized in Fig. 2A. As shown in Fig. 1, no RNA-protected fragment was identified with a control RNA from E. coli containing a particular pE plasmid and pAC7 grown in similar conditions (Fig. 1, lane 2). In all cases, positions of the RNA-protected fragments or primer extension products located tsp downstream of sequences highly similar to the consensus sequence of rpoE-dependent promoters [6] (Fig. 2A). In all cases, the strictly conserved spacing (16 bp) between the −10 and −35 recognition sites was found in these new rpoE-dependent promoters. This spacing was identical to that of the two E. coli canonical rpoE-dependent promoters, rpoEp2 and degPp, that were used for optimization of our two-plasmid system (Fig. 2A). Interestingly, the sequences of the new rpoE-dependent promoters matched more closely the two E. coli canonical rpoE-dependent promoters, rpoEp2 and degPp (Fig. 2A), than the previously characterized rpoE-dependent promoters [6]. All the new promoters shared the 16-bp spacer region, whereas the previously characterized promoters varied greatly in the length of the spacer region [6]. Among the 11 new rpoE-dependent promoters, five promoters were located within the coding region of the upstream convergent gene (Table 1). As with the previously identified rpoE-dependent genes [6], some of the new rpoE-dependent gene set (psd, bacA and smpA) may also be regulated by the CpxA/R envelope stress response system, as sequences similar to the CpxR-P recognition sequence (GTAAA-N5-GTAAA) have been found in the promoter regions of the genes (Fig. 2B) [14].

Figure 1

Example of tsp determination of the E. coli rpoE-dependent promoters (lpxPp and smpAp) by high-resolution S1 nuclease mapping (A) and primer extension analysis (B). A: For S1 nuclease mapping, the 5′-labelled DNA fragment (Section 2) was hybridized with 40 µg RNA, and treated with 100 U of S1 nuclease. RNA was isolated from exponentially grown E. coli, containing the corresponding pE plasmid (pE1 for lpxPp and pE3 for smpAp) and pAC-rpoE4 (lane 1) or pAC7 (lane 2), respectively, in LB supplemented with 100 µg ml−1 ampicillin, 40 µg ml−1 chloramphenicol, and 2 µg ml−1 arabinose. The RNA-protected DNA fragments were analyzed on DNA sequencing gels together with G+A (lane A) and T+C (lane T) sequencing ladders derived from the end-labelled fragments [11]. Thin horizontal arrows indicate the positions of RNA-protected fragments and thick bent vertical arrows indicate the nucleotide corresponding to tsp. Before assigning the tsp, 1.5 nucleotides were subtracted from the length of the protected fragment to account for the difference in the 3′ ends resulting from S1 nuclease digestion and the chemical sequencing reactions. B: The same RNA (80 µg) was used for primer extension analysis using primer −47 (Section 2). The extension products were analyzed on DNA sequencing gels together with GATC chain termination sequencing ladders generated from the corresponding plasmids (pE1 or pE3) using the same −47 primer. The thick bent vertical arrows denote the nucleotide corresponding to tsp.

View this table:
Table 1

Function and genetic organization of new E. coliσE-dependent genes

Gene name (synonym)Operon structureFunctionE. coli K12 genomic sequence (accession number)
Primary metabolism functions
fusA (b3340)rpsG* fusA tufATranslation elongation factors EF-G, EF-TuAE000410
Phospholipid and LPS biogenesis
lpxP (ddg, b2378)lpxPCold shock-induced palmitoleoyl transferaseAE000326
psd (b4160)yjeQ* psdPhosphatidylserine decarboxylaseAE000488
Sensory proteins
sixA (b2340)yfcX* sixAHPt-specific phosphohistidine phosphataseAE000322
Unknown functions
bacA (b3057)bacAPutative inner membrane protein — undecaprenol kinaseAE000387
sbmA (b0377)sbmA yaiWPutative transport inner membrane proteinAE000144
smpA (b2617)smpASmall outer membrane proteinAE000347
yeaY (b1806)yeaZ* yeaYPutative outer membrane lipoproteinAE000275
ybaB (b0471)dnaX* ybaB recRUnknown, DNA repairAE000153
yiiS (b3922)yiiS yiiT (uspD)Unknown, universal stress protein paralogueAE000467
yfeY (b2432)yfeYUnknownAE000330
  • Asterisks indicate the presence of an internal σE-dependent promoter; in all other cases promoters lie to the left of the leftmost gene.

3.3 Identity and function of the new genes directed by EσE

The sequences of the new rpoE-dependent promoters were compared to the E. coli K12 genome to identify 15 genes governed by the promoters. The promoters directed expression of the fusA tufA, lpxP, psd, sixA, bacA, sbmA yaiW, smpA, yeaY, ybaB recR, yiiS, yiiT, and yfeY genes (Table 1). All the identified genes were also present in other sequenced E. coli genomes, including several enterohemorrhagic and uropathogenic strains. The previously characterized members of the σE regulon included regulatory proteins, periplasmic proteases and folding factors, proteins involved in LPS biogenesis, sensory proteins, and proteins with unknown function [6]. Thus, the main role suggested for the σE regulon is to control folding of polypeptides in the bacterial envelope and biosynthesis/transport of LPS [6]. The other proposed role of the σE regulon involves protection of E. coli from potentially lethal envelope stresses [1].

The inferred functions of almost all the 15 new members of the σE regulon fell broadly in the same categories as previously described [6]. Two new members, psd and lpxP, encode proteins involved in the synthesis of phospholipids and LPS, respectively. The psd gene encodes phosphatidylserine decarboxylase, which catalyzes the formation of phosphatidylethanolamine. E. coli lacking phosphatidylethanolamine exhibited several alterations in cellular physiology at high temperatures, including hypersensitivity to certain antibiotics and defects in electron transport, and it is also required for motility and chemotaxis ([15] and references therein). The lpxP (ddg) gene encodes the cold shock-induced palmitoleoyl transferase, which specifically incorporates palmitoleate into nascent lipid A of LPS at low temperatures to insure better membrane fluidity [16]. Recently, an E. coli lpxP mutant has been shown to be more sensitive to several antibiotics at low temperature, indicating that LpxP confers a selective advantage upon E. coli growing at low temperatures by making the outer membrane a more effective barrier to harmful chemicals [17]. Therefore, both rpoE-dependent genes, psd and lpxP, have a role in the integrity of bacterial membranes and resistance against harmful chemicals, thus imposing a new role for the σE regulon in cell envelope integrity. Moreover, other identified rpoE-dependent genes were also inferred to have a similar function. The smpA gene encodes a protein of unknown function which belongs to a family of novel outer membrane lipoproteins. One representative of this family, OmlA of Pseudomonas aeruginosa, has been suggested to have a structural role in maintaining cell envelope integrity. An P. aeruginosa omlA mutant was hypersensitive to anionic detergents and some antibiotics [18]. Two other newly identified rpoE-dependent genes, bacA and sbmA, encode inner membrane proteins conferring resistance in E. coli to the peptide antibiotics bacitracin A and microcins B17 and J25, respectively [19,20]. The bacA gene has been predicted to encode a membrane-bound undecaprenol kinase that phosphorylates undecaprenol, thus overcoming bacitracin inhibition [19]. SbmA is thought to be an inner membrane transport protein involved in the uptake of microcins B17 and J25 and bleomycin in E. coli [20]. Homologues of SbmA in Sinorhizobium meliloti and Brucella abortus are important for long-term survival within plant and mammalian cells respectively [21]. An S. meliloti sbmA mutant exhibited increased sensitivity to agents such as hydrophobic dyes and detergents, which indicates that SbmA is also important in maintaining envelope integrity [22].

The sensory role of the σE regulon is represented by the sixA gene. It encodes an HPt-specific phospho-histidine phosphatase that inhibits the transfer of phosphate groups between the histidine kinase ArcB and its non-cognate response regulator OmpR, and it modulates the response of the ArcBA signalling pathway under certain anaerobic respiratory growth conditions [23].

The five new members of the σE regulon, yeaY, yfeY, yiiS, yaiW, and ybaB, encode proteins with unknown functions. The yeaY gene encodes a putative outer membrane lipoprotein YeaY containing a signal sequence typical of bacterial lipoproteins followed by a characteristic Cys residue at position 22 which could serve as the lipid attachment site. A similar signal sequence is predicted to be present in the product of yfeY. Interestingly, YeaY has significant homology (42% amino acid identity) to an outer membrane lipoprotein encoded by the carbon starvation-inducible and stationary phase-inducible slp gene in E. coli [24]. For the yiiS, yaiW, and ybaB genes no function could be predicted, though ybaB is translationally coupled with the recR gene encoding a protein involved in DNA recombination and repair. However, the yiiS gene is translationally coupled with the yiiT (uspD) gene encoding the universal stress protein paralogue of UspA in E. coli. Similar to uspA, uspD was also induced in stationary phase and by a variety of stresses, and this induction was independent of stationary phase sigma factor σS. Also, an E. coli uspD mutant was found to be sensitive to UV exposure [25]. Thus, this protein likely falls into the category of rpoE-dependent proteins having a function in protection of E. coli from envelope stresses. Interestingly, one of the new rpoE-dependent promoters controlled expression of the fusA tufA operon, which encodes proteins involved in primary metabolism, translation elongation factors G and Tu, respectively. The presence of the σE-dependent promoter might provide growth advantages for E. coli, for example, by increasing or maintaining sufficient levels of elongation factors under adverse conditions.

In conclusion, we identified 11 new rpoE-dependent promoters that can potentially direct the expression of 15 genes. In concordance with the proposed major function of the σE regulon, a number of the newly identified members of the regulon are involved in envelope homeostasis. However, the known or hypothesized function of some of the genes identified in this study extends the role of the σE regulon to include primary metabolic functions and modulation of respiratory pathways. Further work will be needed to characterize the role of these genes and the importance of σE control of their expression for E. coli physiology.


We are grateful to M.K. Winson for plasmid pSB40. This work was supported by Grant 2/3010/23 from the Slovak Academy of Sciences and Wellcome Trust Grant 065027/Z/01/Z.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
View Abstract