Pet is a cytotoxic autotransporter protein secreted by the pathogenic enteroaggregative Escherichia coli strain 042. Expression of Pet is co-dependent on two global transcription regulators: CRP (cyclic AMP receptor protein) and Fis (factor for inversion stimulation). At the pet promoter CRP binds to a single site centred at position -40.5 upstream of the start site for transcription. Due to the suboptimal positioning of this site, CRP alone activates transcription poorly and requires Fis to bind upstream to promote full activation. Here, we show that CRP and Fis control the expression of other important autotransporter toxins, namely Sat from uropathogenic E. coli (UPEC) and SigA from Shigella sonnei, and that this regulation has been conserved in different pathogens. Furthermore, we investigate the mechanism of Fis-mediated co-activation, exploiting a series of semi-synthetic promoters, with similar architecture to the pet promoter. We show that, when bound at position -40.5, CRP recruits RNA polymerase inefficiently and that Fis compensates by aiding polymerase recruitment through a direct protein–protein interaction. We demonstrate that other suitably positioned upstream transcription factors, which directly recruit RNA polymerase, can also compensate for the inappropriate positioning of CRP. We propose that this is a simple ‘shared-recruitment’ mechanism, by which co-dependence of promoters on two transcription factors could evolve.

INTRODUCTION

Autotransporters are important virulence determinants for many Gram-negative bacterial pathogens and belong to a large family of bacterial outer membrane proteins. They have been shown to function as toxins, enzymes and adhesins, and much is known regarding their assembly and biogenesis [1,2]. In spite of their importance, however, few studies have investigated how the genes that encode autotransporter proteins are regulated. Thus, there is a need to understand the environmental cues and transcription factors that control their production, and at what stage during the infection process these important proteins are expressed.

Enteroaggregative Escherichia coli (EAEC) are a diverse group of bacterial pathogens that can cause persistent diarrhoea in humans [3,4]. EAEC cause disease by adhering to epithelial cells to form an extensive biofilm and damage the epithelium by secreting protein toxins [3]. The plasmid-encoded toxin (Pet) from EAEC strain 042 is a cytotoxic autotransporter protein that is secreted during EAEC infection [5,6]. Pet is encoded on a large virulence plasmid, pAA2, and it is an important virulence determinant for EAEC strain 042 [68]. It functions as a serine protease, which, once internalized by gut epithelial cells, leads to cleavage of the cytoskeletal component fodrin and results in cell rounding and detachment [6,9]. Previously, we demonstrated that Pet gene transcription initiated from a single promoter, and that transcript initiation was controlled by two well-studied transcription factors: the cyclic AMP receptor protein (CRP) and the factor for inversion stimulation (Fis) [10].

CRP is a global transcription factor which is required by E. coli for the expression of many genes that are necessary for the utilization of non-preferred carbon sources under poor growth conditions [11,12]. CRP binds as a homodimer to a 16-bp DNA binding site and its DNA-binding activity is allosterically controlled by the binding of cAMP [12,13]. Depending on the promoter context to which it binds, CRP can function as either a transcriptional activator or repressor [11]. At many CRP-activated promoters, the CRP binding site is centred at position -41.5 bp, upstream of the transcriptional start site (+1), such that the distance from its binding site to the -10 hexamer element is 21 bp. This is often referred to as class II CRP-dependent promoter organization [14]. In this situation, CRP activates transcription directly by interacting with the C- and N-terminal domains of the α subunit of RNA polymerase (αCTD and αNTD) using two surface-exposed determinants, termed activating region 1 (AR1) and activating region 2 (AR2), respectively [14].

Fis is a versatile global transcription factor, which is involved in site-specific recombination events, organization of local DNA topology in bacterial chromosomes, as well as functioning as a global transcription factor [1517]. It binds DNA as a homodimer, recognizing a degenerate 15-bp binding sequence that is found in many promoter regions [18,19]. The activity of Fis is controlled by its level of expression, which changes according to growth conditions, and the protein can act as an activator and repressor at many promoters, often switching on genes that are required for growth in favourable conditions [17,1921]. Fis can directly activate transcription at promoters by interacting with RNA polymerase, as is the case at the ribosomal RNA operon promoters [22,23]. However, it can also stimulate transcription indirectly by altering DNA topology and promoter structure [16,24].

At the pet gene promoter, Fis binds to two DNA sites. The Fis I site is centred at position -91 and Fis binding here activates transcription, while occupancy of the lower-affinity Fis II site, centred at position -122, marginally represses promoter activity [10]. Pet expression is completely dependent on CRP, binding to a single site centred at position -40.5. As the positioning of this CRP binding site is suboptimal for a class II promoter, CRP alone activates transcription poorly and requires the help of Fis, bound to Fis I, for maximal pet expression [10]. Thus, the suboptimal positioning of CRP ensures that the promoter is co-dependent on both transcription factors.

Here, we report that similar co-activation operates at promoters that control the regulation of other important autotransporter-encoding genes, namely sat from uropathogenic E. coli (UPEC) and sigA from Shigella sonnei [25,26]. Thus, the utilization of CRP and Fis to control the expression of these virulence determinants has been maintained. Previously, we constructed a semi-synthetic promoter, which mimics the pet promoter architecture and behaves in a manner similar to the wild-type promoter [10]. Here, revisiting this system in more detail, we define the mechanism by which Fis facilitates CRP-mediated activation, when CRP is inappropriately positioned. We show that for Fis to co-activate, it must be bound on the same face of the DNA helix as CRP, and that only certain positions allow productive co-activation. Using genetic and biochemical methods, we demonstrate that Fis activates transcription by aiding the recruitment of RNA polymerase to the promoter region.

EXPERIMENTAL

Bacterial strains, plasmids and growth conditions

The bacterial strains and plasmids used in this work are listed in Table 1 and DNA primers are detailed in Supplementary Table S1. Strains were cultured in lysogeny broth (LB) [27] or Lennox broth (2% (w/v) peptone (Merck), 1% (w/v) yeast extract (Fisher Scientific) and 170 mM NaCl) [28]. For routine DNA manipulations, in vitro transcript analysis, and as a source of DNA fragments for gel retardation assays, DNA fragments carrying promoters were cloned into plasmid pSR [29]. To measure promoter activities, fragments were cloned into the low copy number lac expression vector pRW50 [30]. Derivatives of pSR and plasmid pKK223-3 (Pharmacia), containing the wild-type fis gene and positive control mutants Arg-71-Lys and Gly-72-Ala [31], were maintained in host cells using medium supplemented with 100 μg/ml ampicillin, while pRW50 derivatives were maintained in medium with 35 μg/ml tetracycline.

Table 1
Strains and plasmids
Strain or plasmid Relevant genotype or sequence Reference 
Bacterial strains   
E. coli K-12 strains   
  BW25113 E. coli K-12 lacIqrrnBT14 ΔlacZWJ16hsdR514 ΔaraBADAH33 D rhaBADLD78 [64
  BW25113 Δcrp BW25113 crp::aph [64
  BW25113 Δfis BW25113 fis::aph [64
  M182 E. coli K-12 Δlac [65
  M182 Δcrp M182 Δlac Δcrp [65
  JCB3884 E. coli K-12 Δnir Δlac narL narP253::Tn10dCm [66
  JCB3884 ΔihfA JCB3884 himA452::Tn10dTc TetS [36
 Other bacterial strains   
  EAEC Wild-type E. coli strain 042 [7
  UPEC Wild-type E. coli strain CFT073 [32
  S. sonnei Ss046 Wild-type S. sonnei strain Ss046 [26
Bacterial plasmids   
 pSR pBR322 derivative containing a λ oop transcription terminator [29
 pRW50 Broad-host-range lacZ fusion vector containing the RK2 origin of replication [30
 pRW50/CC(-41.5) pRW50 carrying the E. coli melR promoter derivative with a consensus DNA site for CRP centred at -41.5 [42
 pRW50/CC(-40.5) pRW50 carrying the E. coli melR promoter derivative with a consensus DNA site for CRP centred at -40.5 [34
 pRW50/ML1 pRW50 carrying the E. coli melR promoter derivative with the consensus DNA sites for CRP centred at -41.5 and -90.5 [35
 pKK223-3 E. coli over-expression vector Pharmacia 
 pKK223-3 fis pKK223-3 carrying the E. coli fis gene [31
 pKK223-3 fisR71K pKK223-3 carrying the fis Arg-71-Lys positive control mutant [31
 pKK223-3 fisG72A pKK223-3 carrying the fis Gly-72-Ala positive control mutant [31
Strain or plasmid Relevant genotype or sequence Reference 
Bacterial strains   
E. coli K-12 strains   
  BW25113 E. coli K-12 lacIqrrnBT14 ΔlacZWJ16hsdR514 ΔaraBADAH33 D rhaBADLD78 [64
  BW25113 Δcrp BW25113 crp::aph [64
  BW25113 Δfis BW25113 fis::aph [64
  M182 E. coli K-12 Δlac [65
  M182 Δcrp M182 Δlac Δcrp [65
  JCB3884 E. coli K-12 Δnir Δlac narL narP253::Tn10dCm [66
  JCB3884 ΔihfA JCB3884 himA452::Tn10dTc TetS [36
 Other bacterial strains   
  EAEC Wild-type E. coli strain 042 [7
  UPEC Wild-type E. coli strain CFT073 [32
  S. sonnei Ss046 Wild-type S. sonnei strain Ss046 [26
Bacterial plasmids   
 pSR pBR322 derivative containing a λ oop transcription terminator [29
 pRW50 Broad-host-range lacZ fusion vector containing the RK2 origin of replication [30
 pRW50/CC(-41.5) pRW50 carrying the E. coli melR promoter derivative with a consensus DNA site for CRP centred at -41.5 [42
 pRW50/CC(-40.5) pRW50 carrying the E. coli melR promoter derivative with a consensus DNA site for CRP centred at -40.5 [34
 pRW50/ML1 pRW50 carrying the E. coli melR promoter derivative with the consensus DNA sites for CRP centred at -41.5 and -90.5 [35
 pKK223-3 E. coli over-expression vector Pharmacia 
 pKK223-3 fis pKK223-3 carrying the E. coli fis gene [31
 pKK223-3 fisR71K pKK223-3 carrying the fis Arg-71-Lys positive control mutant [31
 pKK223-3 fisG72A pKK223-3 carrying the fis Gly-72-Ala positive control mutant [31

Promoter fragments

Promoter fragments, containing the UPEC CFT073 sat and S. sonnei Ss046 sigA regulatory regions were amplified by PCR with primer pairs SatFw/SatRev and SigAFw/SigARev, using genomic DNA as template [26,32]. PCR products were restricted with EcoRI and HindIII and cloned into pSR and pRW50. The sat+1 promoter fragment was generated using megaprimer PCR [33]. Primers Sat+1 and SatRev were used with pSR/sat template to generate the first PCR product, which was then used with SatFw and pSR/sat to generate the full-length product. PCR fragments were cloned into pSR using EcoRI and HindIII sites.

The F(-x)CC(-40.5) series of promoter fragments were constructed using standard molecular biology techniques [27]. The F(-91)CC(-40.5) promoter fragment was generated by PCR, using primers F(-91)Fw and D10527 with pRW50/CC(-40.5) as template [34]. PCR product was restricted with EcoRI and HindIII and cloned into pSR and pRW50. To generate the other F(-x)CC(-40.5) constructs promoter DNA was amplified using primer pSRRev and the corresponding F(-x)Fw primer (Supplementary Table S1) with pSR/F(-91)CC(-40.5) as template. The PCR products were restricted with SacI and HindIII and cloned into pSR/F(-91)CC(-40.5). To introduce point mutations into the Fis site of F(-x)CC(-40.5) promoters, DNA was PCR amplified using primers D10527 and FisMut with the relevant pRW50/F(-x)CC(-40.5) plasmid as template. Products were restricted with EcoRI and HindIII and cloned into pRW50.

The CC(-89.5)CC(-40.5) promoter fragment was synthesized using megaprimer PCR [33]. In the first round of PCR primers CRP(-40.5) and D10527 were used with pRW50/ML1 as template [35]. The product of this reaction was then used with primer D10520 and pRW50/ML1 to generate the final PCR product, which was restricted with EcoRI and HindIII and cloned into pRW50. To generate the other CC(-x)CC(-40.5) constructs promoter DNA was amplified using primer D10520 and the corresponding CC(-x)Fw primer (Supplementary Table S1) with pRW50/CC(-89.5)CC(-40.5) as template. PCR products were used in a second round of PCR with pRW50/CC(-89.5)CC(-40.5) as template and primer D10527. Products were restricted with EcoRI and HindIII and cloned into pRW50.

The IHF (integration host factor) I site from the E. coli K-12 nir promoter [36] was introduced into our promoter constructs using PCR. DNA fragments were amplified using primer pSRRev and IHFFw with different pSR/F(-x)CC(-40.5) derivatives as template. Each PCR product was cut with EcoRI and HindIII and cloned into pRW50.

The disruption of the -10 element of various promoters was achieved using megaprimer PCR [33]. Primers CCp9G or CCp11G were used with primer pSRRev and various pSR templates to generate the first PCR product, which was then used with pSRUp and the same pSR template to generate the final product. The disruption of the CRP binding site centred at position -40.5 was also achieved by megaprimer PCR [33], using either primer CRPMut or CRPMutIHF with D10527 and the relevant pRW50 template. Primer D10520 was then used with the same DNA template to generate the final PCR products. Mutant PCR fragments were restricted with EcoRI and HindIII and cloned into pRW50. All DNA constructs were verified by DNA sequencing.

β-Galactosidase assays

To assay promoter activity, pRW50 reporter plasmids, carrying the desired promoter fragments, were transformed into appropriate host strains and β-galactosidase levels were determined using o-nitrophenyl-β-D-galactopyranose (ONPG) and the Miller protocol as previously described [37]. In all cases, cells were grown aerobically with shaking in LB or Lennox broth at 37°C. β-Galactosidase activities are reported as nmol of ONPG hydrolysed/min/mg of dry cell mass and, where indicated, expressed as a percentage of the appropriate strain. Each value is the average for three independent experiments.

Purified proteins

E. coli RNA polymerase holoenzyme containing σ70 was purchased from Epicentre Technologies. Fis and CRP protein were purified as described in [38] and IHF protein as in [39].

In vitro transcription assays

Multiple-round in vitro transcription assays were performed on preparations of pSR plasmid containing the indicated promoter fragments as template as described in [40]. For each experiment, 8 nM of template DNA was incubated in 40 mM Tris/HCl (pH 7.9), 10 mM MgCl2, 50 mM KCl, 0.1 mM DTT, 500 μM GTP, 500 μM ATP, 500 μM CTP, 5 μM UTP and 5 μCi of [α-32P]UTP. Where indicated, Fis was added to a final concentration of 200 nM and CRP was added to a final concentration of 400 nM, supplemented with 200 nM cAMP. Transcription was initiated by the addition of RNA polymerase σ70 holoenzyme. RNA products were visualized on a denaturing 6% polyacrylamide gel and quantified using a Bio-Rad Molecular Imager FX and Quantity One software.

Gel retardation assays

Purified EcoRI–HindIII promoter fragments from pSR derivatives were end-labelled with [γ-32P]ATP using polynucleotide kinase (NEB). Approximately 0.5 nM of each fragment was incubated with the indicated concentrations of purified proteins in the final reaction volume of 10 μl. The reaction buffer contained 20 mM Hepes (pH 8), 5 mM MgCl2, 50 mM potassium glutamate, 1 mM DTT, 5% (v/v) glycerol, 0.5 mg/ml BSA, 3 μg/ml herring sperm DNA and 200 nM cAMP. In all experiments Fis and/or CRP proteins were incubated with DNA fragments for 15 min at 37°C, after which RNA polymerase was added and incubated for a further 15 min. Samples were loaded directly on to 5% polyacrylamide gels containing 2% (v/v) glycerol, while running under tension (12 V/cm) in 0.25× Tris/borate/EDTA (TBE) buffer [36]. All gels contained 200 nM cAMP to prevent dissociation of CRP from the promoter DNA. Where indicated, RNA polymerase promoter complexes were challenged by the addition of heparin to a final concentration of 52 μg/ml. Data from experiments were visualized and quantified using a Bio-Rad Molecular Imager FX and Quantity One software.

RESULTS

Fis and CRP co-activate transcription at both the sat and sigA promoters

Previously, using bioinformatics, we identified a number of promoters with similar architecture to the EAEC 042 pet promoter [10]. The sat promoter from UPEC and the sigA promoter from S. sonnei control the expression of the cytotoxic autotransporters Sat and SigA, which are both Pet homologues [25,26,41]. Sequence alignments identified potential CRP and Fis binding sites at both promoters, centred at positions -40.5 and -91 upstream of the transcription start site (Figure 1A). Importantly, each CRP site was positioned 20-bp upstream of the predicted -10 element, suggesting that both promoters may be co-dependent on CRP and Fis. Hence, we cloned the regulatory regions upstream of sat and sigA into pRW50 to create lacZ transcriptional fusions. The recombinant plasmids, including pRW50/pet, which carries the wild-type pet promoter cloned into pRW50 [10], were transformed into the E. coli K-12 Δlac strain BW25113 and its Δcrp and Δfis derivatives. β-Galactosidase activities were then measured in exponentially growing cells. Results in Figure 1B demonstrate that expression from the pet, sat and sigA promoters is dependent on CRP, as activities in the Δcrp mutant strain were 10%, 20% and 3%, respectively, of the parental strain. Similarly, β-galactosidase activities in the Δfis strain were also decreased for the pet, sat and sigA promoters, indicating that Fis is required for full activation at the sat and sigA promoters, although the dependence on Fis is not as great for the sat and sigA promoters.

Regulation of transcription by CRP and Fis at the pet, sat and sigA promoters

Figure 1
Regulation of transcription by CRP and Fis at the pet, sat and sigA promoters

(A) DNA sequence alignment of the pet, sat and sigA regulatory regions. The figure shows the alignment of the pet promoter from EAEC 042 [10] with the upstream regulatory regions of the UPEC sat and S. sonnei sigA genes [25,26]. The transcription start site for the pet promoter (+1) is in lower case and the -10 element is bold and underlined. The locations of the CRP binding sites are indicated by shaded arrows and the positions of the Fis sites are shown by a shaded rectangle. The centre of each site is numbered with respect to the transcript start of pet and the DNA sites for CRP, Fis and the -10 element are aligned with the relevant consensus sequences. The site where 1 bp was inserted into the sat promoter to generate the sat+1 promoter fragment is shown by a triangle. (B) The figure shows β-galactosidase activities measured in E. coli BW25113, BW25113 Δcrp and BW25113 Δfis cells, carrying pRW50 containing the pet, sat or sigA promoters. Cells were grown aerobically in LB broth and measurements were taken at early exponential phase. For each promoter, activities are expressed as a percentage of the activity measured in E. coli BW25113 cells and each value is the average of three independent determinations and the error bars represent the standard deviation of values.

Figure 1
Regulation of transcription by CRP and Fis at the pet, sat and sigA promoters

(A) DNA sequence alignment of the pet, sat and sigA regulatory regions. The figure shows the alignment of the pet promoter from EAEC 042 [10] with the upstream regulatory regions of the UPEC sat and S. sonnei sigA genes [25,26]. The transcription start site for the pet promoter (+1) is in lower case and the -10 element is bold and underlined. The locations of the CRP binding sites are indicated by shaded arrows and the positions of the Fis sites are shown by a shaded rectangle. The centre of each site is numbered with respect to the transcript start of pet and the DNA sites for CRP, Fis and the -10 element are aligned with the relevant consensus sequences. The site where 1 bp was inserted into the sat promoter to generate the sat+1 promoter fragment is shown by a triangle. (B) The figure shows β-galactosidase activities measured in E. coli BW25113, BW25113 Δcrp and BW25113 Δfis cells, carrying pRW50 containing the pet, sat or sigA promoters. Cells were grown aerobically in LB broth and measurements were taken at early exponential phase. For each promoter, activities are expressed as a percentage of the activity measured in E. coli BW25113 cells and each value is the average of three independent determinations and the error bars represent the standard deviation of values.

Fis compensates for the suboptimal positioning of the CRP site at the sat promoter

To confirm that CRP and Fis co-activate transcription at the sat promoter and to investigate whether Fis compensates for the suboptimal positioning of CRP, we examined the regulation of the sat promoter using in vitro transcription assays. The wild-type sat promoter fragment, along with the sat+1 fragment, which carries a single G:C base pair insertion between positions -20 and -21 (Figure 1A), was cloned into plasmid pSR, placing each promoter upstream of the λ oop transcriptional terminator. In vitro transcription assays were initiated by incubating each pSR derivative with purified CRP, Fis and RNA polymerase and transcripts were separated by denaturing gel electrophoresis. Results illustrated in Figure 2 show that transcription from the wild-type sat promoter was co-dependent on both CRP and Fis, as the sat transcript was maximally produced when both Fis and CRP were present (lane 3). In contrast, high level transcription from the sat+1 promoter required only CRP (lanes 5 and 6). Therefore, we conclude that expression from the UPEC sat promoter requires both Fis and CRP and that the mechanism of co-dependence is conserved at different autotransporter promoters.

Optimal spacing between the promoter -10 element and the CRP binding site relieves the requirement for Fis at the sat promoter

Figure 2
Optimal spacing between the promoter -10 element and the CRP binding site relieves the requirement for Fis at the sat promoter

The figure illustrates the results of an in vitro transcription experiment examining transcription from the wild-type sat promoter and the sat+1 promoter. Multi-round in vitro transcription assays were performed in the presence of 50 nM RNA polymerase holoenzyme with 400 nM CRP and 200 nM Fis present, as indicated. Both the sat and rna1 control transcripts are indicated by arrows. The dashed black line between lanes 3 and 4 indicates where non-adjacent lanes have been spliced together.

Figure 2
Optimal spacing between the promoter -10 element and the CRP binding site relieves the requirement for Fis at the sat promoter

The figure illustrates the results of an in vitro transcription experiment examining transcription from the wild-type sat promoter and the sat+1 promoter. Multi-round in vitro transcription assays were performed in the presence of 50 nM RNA polymerase holoenzyme with 400 nM CRP and 200 nM Fis present, as indicated. Both the sat and rna1 control transcripts are indicated by arrows. The dashed black line between lanes 3 and 4 indicates where non-adjacent lanes have been spliced together.

Co-activation of transcription is optimal when Fis and CRP are bound in helical register

To simplify studies of the pet promoter, we previously generated a series of semi-synthetic promoters to mimic this system [10]. The starting point of this work was the CRP-dependent promoter, CC(-41.5), and its derivative, CC(-40.5), which contain the DNA binding site for CRP positioned at -41.5 and -40.5 bp, respectively [34,42]. To examine Fis-mediated co-activation, we introduced DNA encoding the Fis I site from the pet promoter upstream of the CRP site in CC(-40.5) to generate the F(-x)CC(-40.5) series of promoters, where x denotes the centre of the 15-bp Fis site relative to the transcript start (Figure 3A). Thus, to understand better the rules of Fis-mediated co-activation, we moved the location of the Fis site. All promoter fragments were cloned into the low copy number lacZ expression vector pRW50 to generate promoter lacZ transcription fusions and recombinant plasmids were transformed into strains BW25113 and BW25113 Δfis. Promoter activities were then determined by measuring β-galactosidase expression in mid-exponential phase cells. Results in Figure 3B show that moving the CRP site from position -41.5 to -40.5 led to a large decrease in the expression for the CC(-40.5) promoter, as previously observed [10]. Furthermore, placing the pet promoter Fis I binding site at position -91 in the F(-91)CC(-40.5) promoter, to replicate the pet promoter architecture, increased promoter activity in strain BW25113, and this was dependent on Fis, as expression was decreased in BW25113 Δfis cells. This confirms our previous observation that pet regulation can be recapitulated using a semi-synthetic promoter [10]. When the Fis site was on the same face of the DNA helix at position -80 and -81 (i.e. the F(-80)CC(-40.5) and F(-81)CC(-40.5) promoters) Fis-mediated activation was still observed. However, when the Fis site was repositioned by deleting or inserting 5 bp of DNA (i.e. the F(-86)CC(-40.5) and F(-96)CC(-40.5) promoter constructs) Fis-dependent activation was lost. Thus, for Fis to co-activate, its binding site must be correctly positioned on the same face of the helix as CRP. Interestingly, the effects of Fis were minimal at the F(-71)CC(-40.5) and F(-101)CC(-40.5) promoters, even though Fis sites in these derivatives should be suitably positioned on the DNA helix. Thus, we conclude that not all upstream sites are productive for Fis-mediated co-activation. In extra control experiments, illustrated in Supplementary Figure S1, mutation of Fis binding sites in the F(-x)CC(-40.5) promoters confirmed that Fis activates transcription at these promoters, and the disruption of the -10 element in the F(-81)CC(-40.5) promoter construct demonstrated that additional promoter elements had not been generated by our promoter manipulations.

Regulation of semi-synthetic promoters by Fis

Figure 3
Regulation of semi-synthetic promoters by Fis

(A) The panel shows the DNA sequence of the CC(-41.5), CC(-40.5) and F(-x)CC(-40.5) promoter constructs used in this study. The transcription start site (+1) is indicated by lower case text and the -10 element is underlined. The location of the CRP binding site is represented by inverted arrows and the upstream Fis binding site for each construct is underlined. The site where 1 bp was deleted from the CC(-41.5) promoter to generate the CC(-40.5) promoter is indicated by a triangle. (B) The panel illustrates measured β-galactosidase activities in BW25113 and BW25113 Δfis cells carrying pRW50 containing the different promoter fragments from (A). (C) The panel shows the crystal structure of Fis bound to the F2-binding sequence (FJR9) [67] with amino acid residues Arg-71 and Gly-72, within the BC loop, highlighted. The image was generated using PyMOL (http:://www.pymol.org). (D) The panel details measured β-galactosidase activities of E. coli BW25113 Δfis cells carrying pRW50/F(-81)CC(-40.5) and pKK223-3 plasmids encoding wild-type fis and the Arg-71-Lys and Gly-72-Ala fis positive control mutants. In (B and D) β-galactosidase activities are expressed as nmol of ONPG hydrolysed/min/mg of dry cell mass, each activity is the average of three independent determinations and the error bars represent the standard deviation of values.

Figure 3
Regulation of semi-synthetic promoters by Fis

(A) The panel shows the DNA sequence of the CC(-41.5), CC(-40.5) and F(-x)CC(-40.5) promoter constructs used in this study. The transcription start site (+1) is indicated by lower case text and the -10 element is underlined. The location of the CRP binding site is represented by inverted arrows and the upstream Fis binding site for each construct is underlined. The site where 1 bp was deleted from the CC(-41.5) promoter to generate the CC(-40.5) promoter is indicated by a triangle. (B) The panel illustrates measured β-galactosidase activities in BW25113 and BW25113 Δfis cells carrying pRW50 containing the different promoter fragments from (A). (C) The panel shows the crystal structure of Fis bound to the F2-binding sequence (FJR9) [67] with amino acid residues Arg-71 and Gly-72, within the BC loop, highlighted. The image was generated using PyMOL (http:://www.pymol.org). (D) The panel details measured β-galactosidase activities of E. coli BW25113 Δfis cells carrying pRW50/F(-81)CC(-40.5) and pKK223-3 plasmids encoding wild-type fis and the Arg-71-Lys and Gly-72-Ala fis positive control mutants. In (B and D) β-galactosidase activities are expressed as nmol of ONPG hydrolysed/min/mg of dry cell mass, each activity is the average of three independent determinations and the error bars represent the standard deviation of values.

Fis-mediated transcription activation is dependent on residues within the Fis BC loop

At the E. coli rrnB P1 promoter, Fis activates transcription from an upstream site centred at position -71 by directly contacting the αCTD of RNA polymerase, using a surface-exposed loop, termed the BC loop (Fis residues Thr-70 to Gln-74) (Figure 3C) [23,31]. Our analysis of the F(-x)CC(-40.5) promoters indicated that Fis must be positioned on the same face of the helix as CRP to enhance transcription and suggests that Fis activates by making similar interactions. To investigate this, we utilized Fis mutants that carried single amino acid substitutions in the BC loop that bind to DNA sites normally, but are unable to activate transcription at the rrnB promoter [31,43]. As the F(-81)CC(-40.5) promoter displayed the highest promoter activity (Figure 3B), we examined this construct further and, hence, E. coli strain BW25113 Δfis, containing pRW50/F(-81)CC(-40.5), was transformed with plasmid pKK223-3 expressing either wild-type Fis or Fis positive control mutants carrying the Arg-71-Lys or Gly-72-Ala substitutions within the Fis BC loop. β-Galactosidase activities were determined in exponentially growing cells and results in Figure 3D show that wild-type Fis triggered a 2-fold increase in β-galactosidase activity. In contrast, neither of the altered Fis proteins increased promoter activity when compared with cells carrying the empty pKK223-3 vector. Thus, we conclude that the Fis BC loop is essential for Fis-mediated activation at the F(-81)CC(-40.5) promoter and that Fis likely enhances transcription at this promoter by interacting directly with RNA polymerase.

Fis co-activates transcription by recruiting RNA polymerase

Our results indicate that, when CRP is positioned at -40.5, it is compromised in its ability to activate transcription and that Fis co-activates by interacting with RNA polymerase. To investigate this further, we used gel retardation assays to examine the binding of purified CRP and RNA polymerase to 32P end-labelled CC(-41.5) and CC(-40.5) promoter fragments. When CRP was incubated with each fragment a single shifted species was observed, consistent with each fragment possessing a single CRP binding site of similar affinity (Figure 4A and Supplementary Figure S2). When increasing concentrations of RNA polymerase were incubated with CRP and the promoter DNA, a supershifted complex was observed for both fragments, indicating that RNA polymerase and CRP can bind together to both the CC(-41.5) and CC(-40.5) promoters (Figure 4A, lanes 3–5 and 11–13). However, it is clear that RNA polymerase binds poorly to the CC(-40.5) promoter fragment, in comparison with CC(-41.5). Thus, when CRP is positioned at -40.5, it is unable to recruit RNA polymerase effectively.

Fis aids the recruitment of RNA polymerase

Figure 4
Fis aids the recruitment of RNA polymerase

The figure shows gel retardation assays with purified RNA polymerase, CRP and Fis proteins. (A) End-labelled CC(-41.5) and CC(-40.5) EcoRI–HindIII fragments were incubated with RNA polymerase and CRP. The concentration of RNA polymerase was: lanes 1, 2, 9 and 10, no protein; lanes 3, 6, 11 and 14, 25 nM; lanes 4, 7, 12 and 15, 50 nM; lanes 5, 8, 13 and 16, 100 nM. The concentration of CRP was: lanes 1, 6–9 and 14–16, no protein; lanes 2–5 and 10–13, 100 nM. (B) End-labelled F(-81)CC(-40.5) EcoRI–HindIII fragment was incubated with RNA polymerase, CRP and Fis. The concentration of RNA polymerase was: lanes 1–6, no protein; lanes 7–12, 100 nM. The concentration of CRP was: lanes 1–3 and 10–12, no protein; lanes 4–9, 100 nM. The concentration of Fis was: lanes 1, 4, 7 and 10, no protein; lanes 2, 5, 8 and 11, 50 nM; lanes 3, 6, 9 and 12, 100 nM. (C) End-labelled F(-81)CC(-40.5) EcoRI–HindIII fragment was incubated with either RNA polymerase and CRP (lanes 1–8) or RNA polymerase, CRP and Fis (lanes 9–16). At time zero, heparin was added to disrupt RNA polymerase closed complexes that had formed on the promoter DNA, and samples were withdrawn at 2 min intervals and immediately loaded on to a running polyacrylamide gel. The concentration of both RNA polymerase and CRP was 100 nM in all lanes and in lanes 9–16 Fis was included at 50 nM.

Figure 4
Fis aids the recruitment of RNA polymerase

The figure shows gel retardation assays with purified RNA polymerase, CRP and Fis proteins. (A) End-labelled CC(-41.5) and CC(-40.5) EcoRI–HindIII fragments were incubated with RNA polymerase and CRP. The concentration of RNA polymerase was: lanes 1, 2, 9 and 10, no protein; lanes 3, 6, 11 and 14, 25 nM; lanes 4, 7, 12 and 15, 50 nM; lanes 5, 8, 13 and 16, 100 nM. The concentration of CRP was: lanes 1, 6–9 and 14–16, no protein; lanes 2–5 and 10–13, 100 nM. (B) End-labelled F(-81)CC(-40.5) EcoRI–HindIII fragment was incubated with RNA polymerase, CRP and Fis. The concentration of RNA polymerase was: lanes 1–6, no protein; lanes 7–12, 100 nM. The concentration of CRP was: lanes 1–3 and 10–12, no protein; lanes 4–9, 100 nM. The concentration of Fis was: lanes 1, 4, 7 and 10, no protein; lanes 2, 5, 8 and 11, 50 nM; lanes 3, 6, 9 and 12, 100 nM. (C) End-labelled F(-81)CC(-40.5) EcoRI–HindIII fragment was incubated with either RNA polymerase and CRP (lanes 1–8) or RNA polymerase, CRP and Fis (lanes 9–16). At time zero, heparin was added to disrupt RNA polymerase closed complexes that had formed on the promoter DNA, and samples were withdrawn at 2 min intervals and immediately loaded on to a running polyacrylamide gel. The concentration of both RNA polymerase and CRP was 100 nM in all lanes and in lanes 9–16 Fis was included at 50 nM.

To study Fis-mediated co-activation, we examined the binding of purified Fis, CRP and RNA polymerase to the F(-81)CC(-40.5) promoter fragment (Figure 4B). When Fis and CRP were incubated separately with end-labelled DNA, a single shifted species was observed for both proteins, consistent with the promoter containing a single site for each factor (Figure 4B, and Supplementary Figures S2 and S3). When both proteins were present, a Fis/CRP complex was detected (Figure 4B), confirming that CRP and Fis can bind F(-81)CC(-40.5) together. Incubating CRP and RNA polymerase together with promoter DNA produced a supershifted complex (lane 7), the intensity of which increased 2.4-fold by the addition of Fis (lanes 8 and 9). This shows directly that Fis aids RNA polymerase recruitment to promoters where CRP is suboptimally positioned.

When RNA polymerase binds promoter DNA, it forms a closed complex, before undergoing an isomerization event to unwind the -10 region and generate the open complex [44]. As it is unclear which complex predominates when RNA polymerase binds to the F(-81)CC(-40.5) promoter fragment, we examined the effect of challenging samples with heparin. Note that closed complexes are easily destabilized by heparin, while open complexes are heparin-resistant. Figure 4C shows a gel retardation experiment in which the end-labelled F(-81)CC(-40.5) fragment was mixed with CRP and RNA polymerase in the presence or absence of Fis. Samples were then challenged with heparin to destabilize RNA polymerase closed complexes, and aliquots were withdrawn at 2 min intervals and immediately subjected to gel electrophoresis. As before, in the presence of RNA polymerase and CRP alone a number of diffuse supershifted complexes, containing RNA polymerase, were detected (lane 1). On the addition of heparin, a single shifted species remained, which stayed at a constant level for the duration of the experiment (lanes 2–8), indicating that CRP alone can promote the formation of a stable RNA polymerase open complex on the F(-81)CC(-40.5) promoter. The inclusion of Fis in the reaction mix increased the amount of RNA polymerase supershifted complex (lane 9) and this major species was unaffected by the addition of heparin (lanes 10–16). Thus, we conclude that Fis co-activates at the F(-81)CC(-40.5) promoter by aiding RNA polymerase recruitment and subsequent open complex formation.

Other transcription factors can compensate for the suboptimal positioning of CRP

Our results argue that Fis compensates for the inappropriate positioning of CRP by interacting with the C-terminal domain of the RNA polymerase α subunit. This suggests that other regulators could be commandeered to promote transcription. In addition to activating via a class II mechanism, CRP can also activate from upstream sites (at class I promoters), recruiting RNA polymerase by a direct interaction between AR1 and αCTD [14]. To test whether CRP could functionally replace Fis, we introduced a second CRP binding site upstream of that centred at position -40.5, to generate the CC(-x)CC(-40.5) promoter series (Figure 5A). Promoter fragments were cloned into pRW50 and transformed into the E. coli K-12 ∆lac strains M182 and M182 Δcrp. Results in Figure 5B show that placing a second CRP site at positions -91.5 or -89.5 (i.e. the CC(-91.5)CC(-40.5) and CC(-89.5)CC(-40.5) promoters) considerably increased expression levels in the E. coli K-12 strain M182, confirming that CRP can activate from an upstream site at these promoters. Furthermore, activation was dependent on the face of the helix, as, when the second CRP site was centred at positions -95.5, -93.5, -87.5 or -85.5 (i.e. the CC(-95.5)CC(-40.5), CC(-93.5)CC(-40.5), CC(-87.5)CC(-40.5), and CC(-85.5)CC(-40.5) promoters) promoter activity was similar to that of CC(-40.5). Thus, we conclude that transcription factors, other than Fis, can compensate for the defect observed at our semi-synthetic promoters. Note that control experiments indicated that promoter expression was completely dependent on CRP (Figure 5B) and still required both the CRP binding site centred at position -40.5 and the -10 promoter element sequence (Supplementary Figure S4).

Regulation of semi-synthetic promoters by CRP from an upstream site

Figure 5
Regulation of semi-synthetic promoters by CRP from an upstream site

(A) The panel shows the upstream DNA sequence of the CC(-x)CC(-40.5) promoter constructs used in this study. The location of the CRP binding sites is represented by inverted arrows and the CRP binding sites for each construct are underlined. (B) The panel illustrates measured β-galactosidase activities in M182 and M182 Δcrp cells carrying pRW50 containing the different promoter fragments from (A). β-Galactosidase activities are expressed as nmol of ONPG hydrolysed/min/mg of dry cell mass, each activity is the average of three independent determinations and the error bars represent the standard deviation of values.

Figure 5
Regulation of semi-synthetic promoters by CRP from an upstream site

(A) The panel shows the upstream DNA sequence of the CC(-x)CC(-40.5) promoter constructs used in this study. The location of the CRP binding sites is represented by inverted arrows and the CRP binding sites for each construct are underlined. (B) The panel illustrates measured β-galactosidase activities in M182 and M182 Δcrp cells carrying pRW50 containing the different promoter fragments from (A). β-Galactosidase activities are expressed as nmol of ONPG hydrolysed/min/mg of dry cell mass, each activity is the average of three independent determinations and the error bars represent the standard deviation of values.

Upstream DNA sequences can stimulate promoter expression

In addition to Fis, other nucleoid-associated proteins have been found to regulate gene expression [17]. One such factor is IHF, a small sequence-specific DNA-binding protein, which binds to specific 27-bp sites and bends DNA >140° [45]. In most instances IHF activates transcription due to the conformation changes it induces in promoter DNA, rather than interacting directly with RNA polymerase [46,47]. To test whether IHF could activate transcription in our synthetic system, we introduced one of the upstream IHF sites from the E. coli nir operon regulatory region into our promoters, generating the I(-x)CC(-40.5) promoter constructs (Figure 6A) [36]. Promoter fragments were cloned into pRW50 and β-galactosidase expression was measured in the E. coli K-12 ∆lac strains JCB3884 and JCB3884 ΔihfA. Results in Figure 6B show that, for all constructs, expression in JCB3884 never exceeds that observed with the ihfA-null mutant, indicating that IHF does not activate in any of our constructs, even though IHF bound to these promoters in vitro (Supplementary Figure S5). Intriguingly, promoters which carried an upstream IHF site centred at positions -87, -88 and -89 (the I(-87)CC(-40.5), I(-88)CC(-40.5) and I(-89)CC(-40.5) promoters) had expression levels considerably higher than CC(-40.5) in the absence of IHF (Figure 6B). As expression from these promoters was still dependent on CRP and the -10 promoter element (Supplementary Figure S6), this suggests that the DNA site for IHF itself is capable of stimulating transcription initiation. Many IHF binding sites contain A/T-rich DNA sequences, which resemble bacterial UP elements that stimulate transcription by docking the αCTD of RNA polymerase on to the DNA [4850]. As the nirB IHF site used in this study closely matches the consensus sequence for an UP element [51], the most likely explanation for our results is that in the absence of IHF, sequences within this site are bound by αCTD, stimulating transcription by RNA polymerase recruitment. Thus, we propose that upstream cis-acting DNA sequences can also compensate for the incorrect positioning of CRP.

Regulation of semi-synthetic promoters by IHF bound to an upstream site

Figure 6
Regulation of semi-synthetic promoters by IHF bound to an upstream site

(A) The panel shows the upstream DNA sequence of the I(-x)CC(-40.5) promoter constructs used in this study. The location of the CRP binding site is represented by inverted arrows and the upstream IHF binding site for each construct is underlined. (B) The panel illustrates measured β-galactosidase activities in JCB3884 and JCB3884 ΔihfA cells carrying pRW50 containing the different promoter fragments from (A). β-Galactosidase activities are expressed as nmol of ONPG hydrolysed/min/mg of dry cell mass, each activity is the average of three independent determinations and the error bars represent the standard deviation of values.

Figure 6
Regulation of semi-synthetic promoters by IHF bound to an upstream site

(A) The panel shows the upstream DNA sequence of the I(-x)CC(-40.5) promoter constructs used in this study. The location of the CRP binding site is represented by inverted arrows and the upstream IHF binding site for each construct is underlined. (B) The panel illustrates measured β-galactosidase activities in JCB3884 and JCB3884 ΔihfA cells carrying pRW50 containing the different promoter fragments from (A). β-Galactosidase activities are expressed as nmol of ONPG hydrolysed/min/mg of dry cell mass, each activity is the average of three independent determinations and the error bars represent the standard deviation of values.

DISCUSSION

Fis and CRP are two important global regulators, which control the expression of hundreds of genes in enteric bacteria [11,12,21,52]. At the EAEC 042 pet promoter both Fis and CRP are required for full activation [10]. Here, using semi-synthetic promoters which mimic the pet promoter organization, we show that when CRP is bound to a suboptimal site positioned at -40.5, it recruits RNA polymerase inefficiently, and that Fis-mediated activation compensates for this (Figures 3 and 4). For Fis to activate, it must be positioned on the same face of the DNA helix as CRP and requires the surface-exposed BC loop, which is involved in direct activation at a number of E. coli promoters (Figure 3) [23,53]. Thus, we propose that Fis rescues the inability of CRP to activate by directly facilitating RNA polymerase recruitment, and Fis and CRP co-activate here by a ‘shared-recruitment’ mechanism.

When CRP binds to a class II CRP-dependent promoter, it interacts with RNA polymerase using AR1 and AR2. The AR1 interaction ensures that one αCTD of RNA polymerase is positioned upstream of CRP, leaving the second αCTD free to dock with upstream DNA or interact with a second transcription activator [14,54]. At our synthetic CC(-40.5) promoters, AR1 and AR2 will be misaligned, due to the positioning of the CRP site, and this impedes the ability of CRP to recruit RNA polymerase to the promoter DNA (Figure 4). Placing a Fis binding site at position -81 or -91 compensates for this (Figures 3 and 4). At the E. coli rrnB P1 promoter Fis activates transcription from position -71 by interacting with the αCTD of RNA polymerase, using the BC loop [23]. In this instance, the αCTD binds to the DNA immediately downstream of Fis [23]. Our results suggest that a similar interaction takes place at our Fis-activated synthetic promoters (Figure 7). Interestingly, only certain locations allow productive activation, as Fis sites at -71 or -101 have little or no effect, even though Fis should be bound on the correct face of the DNA helix (Figure 3). At class II CRP-dependent promoters the second αCTD of RNA polymerase is free to interact with DNA at positions -73 and -83 [14,54] and this would allow Fis, when bound at -81 or -91, respectively, to interact with αCTD (Figure 7). However, when Fis is positioned at -71, there is insufficient space for the second αCTD to dock correctly, and it is likely that the αCTD cannot reach Fis when it is positioned far upstream at position -101 [54].

A model of Fis- and CRP-dependent co-activation

Figure 7
A model of Fis- and CRP-dependent co-activation

The figure illustrates a model of RNA polymerase, CRP and Fis bound to the F(-81)CC(-40.5) and F(-91)CC(-40.5) promoters. The upstream locations, to which CRP, Fis and the αCTDs of RNA polymerase bind, are indicated [14,54].

Figure 7
A model of Fis- and CRP-dependent co-activation

The figure illustrates a model of RNA polymerase, CRP and Fis bound to the F(-81)CC(-40.5) and F(-91)CC(-40.5) promoters. The upstream locations, to which CRP, Fis and the αCTDs of RNA polymerase bind, are indicated [14,54].

The observation that suboptimal positioning of CRP can be rescued by a transcription factor that interacts with αCTD, suggests that other transcription factors might similarly promote transcription. Hence, this could provide a cell with a simple strategy to integrate additional regulatory responses to environmental signals, and provides a mechanism by which co-dependence could evolve. Consistent with this, the introduction of a second upstream CRP site activated transcription, rescuing expression levels to near those of CC(-41.5) (Figure 5), while IHF, which does not activate by interacting directly with RNA polymerase, could not (Figure 6). However, our experiments with the I(-x)CC(-40.5) constructs, demonstrated that upstream cis-acting sequences, which resemble bacterial UP elements, can enhance transcription (Figure 6B), and indicates that there are a variety of mechanisms by which the inappropriate positioning of CRP can be counteracted at our promoters.

UPEC strains are the most common cause of urinary tract infections [55], while Shigella species (including S. sonnei) induce the devastating gastrointestinal infection shigellosis [56]. The Sat and SigA autotransporters are important virulence determinants and both are cytopathic serine proteases, homologous with Pet [5,26,41]. Alignment of their upstream regulatory sequences indicated that their promoter regions are conserved and our in vivo analysis confirmed that expression from each promoter was co-activated by CRP and Fis (Figure 1). Hence, we demonstrate that, for the sat promoter, maximal activity in vitro required both transcription factors and the requirement for Fis was due to suboptimal positioning of CRP (Figure 2). This argues that the mechanism of co-activation, found at the EAEC 042 pet promoter, has been maintained at the sat promoter.

Both CRP and Fis play important roles in the virulence of many bacterial pathogens, for example in UPEC the disruption of crp leads to attenuation [57], while Fis activates the expression of virF in Shigella flexneri and aggR in EAEC strain 042, both of which encode master virulence regulators, essential for pathogenesis [58,59]. In E. coli K-12, CRP activates many genes required for utilization of alternative carbon sources under poor growth conditions [11], while Fis induces the expression of genes required for rapid growth when conditions are favourable [19,21]. Thus, promoters activated by both proteins are rare, since each transcription factor regulates responses to opposing growth conditions. As Pet expression ‘spikes’ during exponential growth phase [9], we have suggested this is due to the co-dependence on CRP and Fis, and that this is an adaptation of EAEC 042 to its environment in the intestinal lumen [10].

Many transcription factors that control bacterial virulence respond to host-specific signals, such as temperature, the presence of bile salts or bicarbonate ions [5961]. However, the expression of pet, sat and sigA is regulated by two global transcription factors that reflect the homoeostasis of the cell, suggesting that maximal expression occurs when it is physiologically suitable to mount an attack. The Pet serine protease has wide substrate specificity, cleaving fodrin, pepsin, casein and gelatin among others [5,62]. As well as being an enterotoxin and cytopathic to human epithelial cells, Pet enhances killing of the nematode Caenorhabditis elegans, reflecting the toxin's virulence in very different systems [5,8]. Thus, we propose that pet expression, rather than depending on host-specific signals, is linked to growth-specific regulation to ensure that toxin expression takes place under as many situations as possible. EAEC, UPEC and S. sonnei colonize very different habitats within the human body and we note that the in vivo dependence of both sat and sigA promoters on Fis is less pronounced than observed at the pet promoter, suggesting that a degree of fine-tuning has taken place. Interestingly, sigA in Shigella flexneri is up-regulated within human epithelial cells and production of SigA protein is also induced by growth at 37°C, unlike Pet [5,41,63]. This suggests that although sigA expression is still dependent on CRP and Fis, the regulation of sigA may have evolved as well. Such subtle differences suggest the regulation of these genes may have evolved to integrate niche-specific environmental signals and may be influenced by other regulators.

Abbreviations

     
  • AR1

    CRP activating region 1

  •  
  • AR2

    CRP activating region 2

  •  
  • CRP

    cyclic AMP receptor protein

  •  
  • αCTD

    C-terminal domain of the α subunit of RNA polymerase

  •  
  • EAEC

    enteroaggregative E. coli

  •  
  • Fis

    factor for inversion stimulation

  •  
  • IHF

    integration host factor

  •  
  • LB

    lysogeny broth

  •  
  • αNTD

    N-terminal domain of the α subunit of RNA polymerase

  •  
  • ONPG

    o-nitrophenyl-β-D-galactopyranose

  •  
  • Pet

    plasmid-encoded toxin

  •  
  • TBE

    Tris/borate/EDTA

  •  
  • UPEC

    uropathogenic E. coli

AUTHOR CONTRIBUTION

Amanda Rossiter, Stephen Busby, Ian Henderson and Douglas Browning conceived and designed the research programme. Amanda Rossiter, Rita Godfrey, Jack Connolly and Douglas Browning performed the experiments. Amanda Rossiter and Douglas Browning wrote the manuscript with input from all authors.

We thank David Lee and David Grainger for purified proteins and Reid Johnson for donating plasmid pKK223-3 and its derivatives.

FUNDING

This work was generously supported by Medical Research Council and Biotechnology and Biological Science Research Council grants to I.R.H. and a Biotechnology and Biological Science Research Council grant [grant number BB/J006076/1] to D.F.B. and S.J.W.B.

References

References
1
Leyton
D.L.
Rossiter
A.E.
Henderson
I.R.
From self sufficiency to dependence: mechanisms and factors important for autotransporter biogenesis
Nat. Rev. Microbiol.
2012
, vol. 
10
 (pg. 
213
-
225
)
[PubMed]
2
Henderson
I.R.
Navarro-Garcia
F.
Desvaux
M.
Fernandez
R.C.
Ala’Aldeen
D.
Type V protein secretion pathway: the autotransporter story
Microbiol. Mol. Biol. Rev.
2004
, vol. 
68
 (pg. 
692
-
744
)
[PubMed]
3
Harrington
S.M.
Sheikh
J.
Henderson
I.R.
Ruiz-Perez
F.
Cohen
P.S.
Nataro
J.P.
The Pic protease of enteroaggregative Escherichia coli promotes intestinal colonization and growth in the presence of mucin
Infect. Immun.
2009
, vol. 
77
 (pg. 
2465
-
2473
)
[PubMed]
4
Nataro
J.P.
Deng
Y.
Cookson
S.
Cravioto
A.
Savarino
S.J.
Guers
L.D.
Levine
M.M.
Tacket
C.O.
Heterogeneity of enteroaggregative Escherichia coli virulence demonstrated in volunteers
J. Infect. Dis.
1995
, vol. 
171
 (pg. 
465
-
468
)
[PubMed]
5
Eslava
C.
Navarro-Garcia
F.
Czeczulin
J.R.
Henderson
I.R.
Cravioto
A.
Nataro
J.P.
Pet, an autotransporter enterotoxin from enteroaggregative Escherichia coli
Infect. Immun.
1998
, vol. 
66
 (pg. 
3155
-
3163
)
[PubMed]
6
Henderson
I.R.
Hicks
S.
Navarro-Garcia
F.
Elias
W.P.
Philips
A.D.
Nataro
J.P.
Involvement of the enteroaggregative Escherichia coli plasmid-encoded toxin in causing human intestinal damage
Infect. Immun.
1999
, vol. 
67
 (pg. 
5338
-
5344
)
[PubMed]
7
Chaudhuri
R.R.
Sebaihia
M.
Hobman
J.L.
Webber
M.A.
Leyton
D.L.
Goldberg
M.D.
Cunningham
A.F.
Scott-Tucker
A.
Ferguson
P.R.
Thomas
C.M.
, et al. 
Complete genome sequence and comparative metabolic profiling of the prototypical enteroaggregative Escherichia coli strain 042
PLoS One
2010
, vol. 
5
 pg. 
e8801
 
[PubMed]
8
Franca
F.L.
Wells
T.J.
Browning
D.F.
Nogueira
R.T.
Sarges
F.S.
Pereira
A.C.
Cunningham
A.F.
Lucheze
K.
Rosa
A.C.
Henderson
I.R.
de Luna
M.
Genotypic and phenotypic characterisation of enteroaggregative Escherichia coli from children in Rio de Janeiro
Brazil. PLoS One
2013
, vol. 
8
 pg. 
e69971
 
9
Betancourt-Sanchez
M.
Navarro-Garcia
F.
Pet secretion, internalization and induction of cell death during infection of epithelial cells by enteroaggregative Escherichia coli
Microbiology
2009
, vol. 
155
 (pg. 
2895
-
2906
)
[PubMed]
10
Rossiter
A.E.
Browning
D.F.
Leyton
D.L.
Johnson
M.D.
Godfrey
R.E.
Wardius
C.A.
Desvaux
M.
Cunningham
A.F.
Ruiz-Perez
F.
Nataro
J.P.
, et al. 
Transcription of the plasmid-encoded toxin gene from enteroaggregative Escherichia coli is regulated by a novel co-activation mechanism involving CRP and Fis
Mol. Microbiol.
2011
, vol. 
81
 (pg. 
179
-
191
)
[PubMed]
11
Gosset
G.
Zhang
Z.
Nayyar
S.
Cuevas
W.A.
Saier
M.H.
Jr.
Transcriptome analysis of Crp-dependent catabolite control of gene expression in Escherichia coli
J. Bacteriol.
2004
, vol. 
186
 (pg. 
3516
-
3524
)
[PubMed]
12
Green
J.
Stapleton
M.R.
Smith
L.J.
Artymiuk
P.J.
Kahramanoglou
C.
Hunt
D.M.
Buxton
R.S.
Cyclic-AMP and bacterial cyclic-AMP receptor proteins revisited: adaptation for different ecological niches
Curr. Opin. Microbiol.
2014
, vol. 
18C
 (pg. 
1
-
7
)
13
Won
H.S.
Lee
Y.S.
Lee
S.H.
Lee
B.J.
Structural overview on the allosteric activation of cyclic AMP receptor protein
Biochim. Biophys. Acta
2009
, vol. 
1794
 (pg. 
1299
-
1308
)
[PubMed]
14
Lawson
C.L.
Swigon
D.
Murakami
K.S.
Darst
S.A.
Berman
H.M.
Ebright
R.H.
Catabolite activator protein: DNA binding and transcription activation
Curr. Opin. Struct. Biol.
2004
, vol. 
14
 (pg. 
10
-
20
)
[PubMed]
15
Koch
C.
Kahmann
R.
Purification and properties of the Escherichia coli host factor required for inversion of the G segment in bacteriophage Mu
J. Biol. Chem.
1986
, vol. 
261
 (pg. 
15673
-
15678
)
[PubMed]
16
Travers
A.
Schneider
R.
Muskhelishvili
G.
DNA supercoiling and transcription in Escherichia coli: the FIS connection
Biochimie
2001
, vol. 
83
 (pg. 
213
-
217
)
[PubMed]
17
Browning
D.F.
Grainger
D.C.
Busby
S.J.
Effects of nucleoid-associated proteins on bacterial chromosome structure and gene expression
Curr. Opin. Microbiol.
2010
, vol. 
13
 (pg. 
773
-
780
)
[PubMed]
18
Grainger
D.C.
Hurd
D.
Goldberg
M.D.
Busby
S.J.
Association of nucleoid proteins with coding and non-coding segments of the Escherichia coli genome
Nucleic Acids Res.
2006
, vol. 
34
 (pg. 
4642
-
4652
)
[PubMed]
19
Kahramanoglou
C.
Seshasayee
A.S.
Prieto
A.I.
Ibberson
D.
Schmidt
S.
Zimmermann
J.
Benes
V.
Fraser
G.M.
Luscombe
N.M.
Direct and indirect effects of H-NS and Fis on global gene expression control in Escherichia coli
Nucleic Acids Res.
2011
, vol. 
39
 (pg. 
2073
-
2091
)
[PubMed]
20
Ball
C.A.
Osuna
R.
Ferguson
K.C.
Johnson
R.C.
Dramatic changes in Fis levels upon nutrient upshift in Escherichia coli
J. Bacteriol.
1992
, vol. 
174
 (pg. 
8043
-
8056
)
[PubMed]
21
Bradley
M.D.
Beach
M.B.
de Koning
A.P.
Pratt
T.S.
Osuna
R.
Effects of Fis on Escherichia coli gene expression during different growth stages
Microbiology
2007
, vol. 
153
 (pg. 
2922
-
2940
)
[PubMed]
22
Hirvonen
C.A.
Ross
W.
Wozniak
C.E.
Marasco
E.
Anthony
J.R.
Aiyar
S.E.
Newburn
V.H.
Gourse
R.L.
Contributions of UP elements and the transcription factor FIS to expression from the seven rrn P1 promoters in Escherichia coli
J. Bacteriol.
2001
, vol. 
183
 (pg. 
6305
-
6314
)
[PubMed]
23
Aiyar
S.E.
McLeod
S.M.
Ross
W.
Hirvonen
C.A.
Thomas
M.S.
Johnson
R.C.
Gourse
R.L.
Architecture of Fis-activated transcription complexes at the Escherichia coli rrnB P1 and rrnE P1 promoters
J. Mol. Biol.
2002
, vol. 
316
 (pg. 
501
-
516
)
[PubMed]
24
Opel
M.L.
Aeling
K.A.
Holmes
W.M.
Johnson
R.C.
Benham
C.J.
Hatfield
G.W.
Activation of transcription initiation from a stable RNA promoter by a Fis protein-mediated DNA structural transmission mechanism
Mol. Microbiol.
2004
, vol. 
53
 (pg. 
665
-
674
)
[PubMed]
25
Guyer
D.M.
Henderson
I.R.
Nataro
J.P.
Mobley
H.L.
Identification of sat, an autotransporter toxin produced by uropathogenic Escherichia coli
Mol. Microbiol.
2000
, vol. 
38
 (pg. 
53
-
66
)
[PubMed]
26
Yang
F.
Yang
J.
Zhang
X.
Chen
L.
Jiang
Y.
Yan
Y.
Tang
X.
Wang
J.
Xiong
Z.
Dong
J.
, et al. 
Genome dynamics and diversity of Shigella species, the etiologic agents of bacillary dysentery
Nucleic Acids Res.
2005
, vol. 
33
 (pg. 
6445
-
6458
)
[PubMed]
27
Sambrook
J.
Russell
D.W.
Molecular Cloning: a Laboratory Manual
2001
Cold Spring Harbor
Cold Spring Harbor Laboratory Press
28
Squire
D.J.
Xu
M.
Cole
J.A.
Busby
S.J.
Browning
D.F.
Competition between NarL-dependent activation and Fis-dependent repression controls expression from the Escherichia coli yeaR and ogt promoters
Biochem. J.
2009
, vol. 
420
 (pg. 
249
-
257
)
[PubMed]
29
Kolb
A.
Kotlarz
D.
Kusano
S.
Ishihama
A.
Selectivity of the Escherichia coli RNA polymerase E sigma 38 for overlapping promoters and ability to support CRP activation
Nucleic Acids Res.
1995
, vol. 
23
 (pg. 
819
-
826
)
[PubMed]
30
Lodge
J.
Fear
J.
Busby
S.
Gunasekaran
P.
Kamini
N.R.
Broad host range plasmids carrying the Escherichia coli lactose and galactose operons
FEMS Microbiol. Lett.
1992
, vol. 
74
 (pg. 
271
-
276
)
[PubMed]
31
Bokal
A.J.
Ross
W.
Gaal
T.
Johnson
R.C.
Gourse
R.L.
Molecular anatomy of a transcription activation patch: FIS-RNA polymerase interactions at the Escherichia coli rrnB P1 promoter
EMBO J.
1997
, vol. 
16
 (pg. 
154
-
162
)
[PubMed]
32
Luo
C.
Hu
G.Q.
Zhu
H.
Genome reannotation of Escherichia coli CFT073 with new insights into virulence
BMC Genomics
2009
, vol. 
10
 pg. 
552
 
33
Sarkar
G.
Sommer
S.S.
The “megaprimer” method of site-directed mutagenesis
Biotechniques
1990
, vol. 
8
 (pg. 
404
-
407
)
[PubMed]
34
West
D.
Williams
R.
Rhodius
V.
Bell
A.
Sharma
N.
Zou
C.
Fujita
N.
Ishihama
A.
Busby
S.
Interactions between the Escherichia coli cyclic AMP receptor protein and RNA polymerase at class II promoters
Mol. Microbiol.
1993
, vol. 
10
 (pg. 
789
-
797
)
[PubMed]
35
Busby
S.
West
D.
Lawes
M.
Webster
C.
Ishihama
A.
Kolb
A.
Transcription activation by the Escherichia coli cyclic AMP receptor protein. Receptors bound in tandem at promoters can interact synergistically
J. Mol. Biol.
1994
, vol. 
241
 (pg. 
341
-
352
)
[PubMed]
36
Browning
D.F.
Cole
J.A.
Busby
S.J.
Regulation by nucleoid-associated proteins at the Escherichia coli nir operon promoter
J. Bacteriol.
2008
, vol. 
190
 (pg. 
7258
-
7267
)
[PubMed]
37
Jayaraman
P.S.
Peakman
T.C.
Busby
S.J.
Quincey
R.V.
Cole
J.A.
Location and sequence of the promoter of the gene for the NADH-dependent nitrite reductase of Escherichia coli and its regulation by oxygen, the Fnr protein and nitrite
J. Mol. Biol.
1987
, vol. 
196
 (pg. 
781
-
788
)
[PubMed]
38
Grainger
D.C.
Goldberg
M.D.
Lee
D.J.
Busby
S.J.
Selective repression by Fis and H-NS at the Escherichia coli dps promoter
Mol. Microbiol.
2008
, vol. 
68
 (pg. 
1366
-
1377
)
[PubMed]
39
Nash
H.A.
Robertson
C.A.
Flamm
E.
Weisberg
R.A.
Miller
H.I.
Overproduction of Escherichia coli integration host factor, a protein with nonidentical subunits
J. Bacteriol.
1987
, vol. 
169
 (pg. 
4124
-
4127
)
[PubMed]
40
Browning
D.
Savery
N.
Kolb
A.
Busby
S.
Assays for transcription factor activity
Methods Mol. Biol.
2009
, vol. 
543
 (pg. 
369
-
387
)
[PubMed]
41
Al-Hasani
K.
Henderson
I.R.
Sakellaris
H.
Rajakumar
K.
Grant
T.
Nataro
J.P.
Robins-Browne
R.
Adler
B.
The sigA gene which is borne on the she pathogenicity island of Shigella flexneri 2a encodes an exported cytopathic protease involved in intestinal fluid accumulation
Infect. Immun.
2000
, vol. 
68
 (pg. 
2457
-
2463
)
[PubMed]
42
Gaston
K.
Bell
A.
Kolb
A.
Buc
H.
Busby
S.
Stringent spacing requirements for transcription activation by CRP
Cell
1990
, vol. 
62
 (pg. 
733
-
743
)
[PubMed]
43
McLeod
S.M.
Xu
J.
Cramton
S.E.
Gaal
T.
Gourse
R.L.
Johnson
R.C.
Localization of amino acids required for Fis to function as a class II transcriptional activator at the RpoS-dependent proP P2 promoter
J. Mol. Biol.
1999
, vol. 
294
 (pg. 
333
-
346
)
[PubMed]
44
Browning
D.F.
Busby
S.J.
The regulation of bacterial transcription initiation
Nat. Rev. Microbiol.
2004
, vol. 
2
 (pg. 
57
-
65
)
[PubMed]
45
Rice
P.A.
Yang
S.
Mizuuchi
K.
Nash
H.A.
Crystal structure of an IHF-DNA complex: a protein-induced DNA U-turn
Cell
1996
, vol. 
87
 (pg. 
1295
-
1306
)
[PubMed]
46
Sheridan
S.D.
Benham
C.J.
Hatfield
G.W.
Activation of gene expression by a novel DNA structural transmission mechanism that requires supercoiling-induced DNA duplex destabilization in an upstream activating sequence
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
21298
-
21308
)
[PubMed]
47
Shingler
V.
Signal sensory systems that impact σ54-dependent transcription
FEMS Microbiol. Rev.
2011
, vol. 
35
 (pg. 
425
-
440
)
[PubMed]
48
Giladi
H.
Koby
S.
Prag
G.
Engelhorn
M.
Geiselmann
J.
Oppenheim
A.B.
Participation of IHF and a distant UP element in the stimulation of the phage lambda PL promoter
Mol. Microbiol.
1998
, vol. 
30
 (pg. 
443
-
451
)
[PubMed]
49
Giladi
H.
Murakami
K.
Ishihama
A.
Oppenheim
A.B.
Identification of an UP element within the IHF binding site at the PL1-PL2 tandem promoter of bacteriophage lambda
J. Mol. Biol.
1996
, vol. 
260
 (pg. 
484
-
491
)
[PubMed]
50
Fyfe
J.A.
Davies
J.K.
An AT-rich tract containing an integration host factor-binding domain and two UP-like elements enhances transcription from the pilEp1 promoter of Neisseria gonorrhoeae
J. Bacteriol.
1998
, vol. 
180
 (pg. 
2152
-
2159
)
[PubMed]
51
Estrem
S.T.
Gaal
T.
Ross
W.
Gourse
R.L.
Identification of an UP element consensus sequence for bacterial promoters
Proc. Nat. Acad. Sci. U.S.A.
1998
, vol. 
95
 (pg. 
9761
-
9766
)
52
Kelly
A.
Goldberg
M.D.
Carroll
R.K.
Danino
V.
Hinton
J.C.
Dorman
C.J.
A global role for Fis in the transcriptional control of metabolism and type III secretion in Salmonella enterica serovar Typhimurium
Microbiology
2004
, vol. 
150
 (pg. 
2037
-
2053
)
[PubMed]
53
McLeod
S.M.
Aiyar
S.E.
Gourse
R.L.
Johnson
R.C.
The C-terminal domains of the RNA polymerase alpha subunits: contact site with Fis and localization during co-activation with CRP at the Escherichia coli proP P2 promoter
J. Mol. Biol.
2002
, vol. 
316
 (pg. 
517
-
529
)
[PubMed]
54
Belyaeva
T.A.
Rhodius
V.A.
Webster
C.L.
Busby
S.J.
Transcription activation at promoters carrying tandem DNA sites for the Escherichia coli cyclic AMP receptor protein: organisation of the RNA polymerase alpha subunits
J. Mol. Biol.
1998
, vol. 
277
 (pg. 
789
-
804
)
[PubMed]
55
Hooton
T.M.
Clinical practice. Uncomplicated urinary tract infection
N. Engl. J. Med.
2012
, vol. 
366
 (pg. 
1028
-
1037
)
[PubMed]
56
Kosek
M.
Yori
P.P.
Olortegui
M.P.
Shigellosis update: advancing antibiotic resistance, investment empowered vaccine development, and green bananas
Curr. Opin. Infect. Dis.
2010
, vol. 
23
 (pg. 
475
-
480
)
[PubMed]
57
Donovan
G.T.
Norton
J.P.
Bower
J.M.
Mulvey
M.A.
Adenylate cyclase and the cyclic AMP receptor protein modulate stress resistance and virulence capacity of uropathogenic Escherichia coli
Infect. Immun.
2013
, vol. 
81
 (pg. 
249
-
258
)
[PubMed]
58
Morin
N.
Tirling
C.
Ivison
S.M.
Kaur
A.P.
Nataro
J.P.
Steiner
T.S.
Autoactivation of the AggR regulator of enteroaggregative Escherichia coli in vitro and in vivo
FEMS Immunol. Med. Microbiol.
2010
, vol. 
58
 (pg. 
344
-
355
)
[PubMed]
59
Falconi
M.
Prosseda
G.
Giangrossi
M.
Beghetto
E.
Colonna
B.
Involvement of FIS in the H-NS-mediated regulation of virF gene of Shigella and enteroinvasive Escherichia coli
Mol. Microbiol.
2001
, vol. 
42
 (pg. 
439
-
452
)
[PubMed]
60
Childers
B.M.
Cao
X.
Weber
G.G.
Demeler
B.
Hart
P.J.
Klose
K.E.
N-terminal residues of the Vibrio cholerae virulence regulatory protein ToxT involved in dimerization and modulation by fatty acids
J. Biol. Chem.
2011
, vol. 
286
 (pg. 
28644
-
28655
)
[PubMed]
61
Yang
J.
Dogovski
C.
Hocking
D.
Tauschek
M.
Perugini
M.
Robins-Browne
R.M.
Bicarbonate-mediated stimulation of RegA, the global virulence regulator from Citrobacter rodentium
J. Mol. Biol.
2009
, vol. 
394
 (pg. 
591
-
599
)
[PubMed]
62
Dutta
P.R.
Cappello
R.
Navarro-Garcia
F.
Nataro
J.P.
Functional comparison of serine protease autotransporters of enterobacteriaceae
Infect. Immun.
2002
, vol. 
70
 (pg. 
7105
-
7113
)
[PubMed]
63
Morris
C.R.
Grassel
C.L.
Redman
J.C.
Sahl
J.W.
Barry
E.M.
Rasko
D.A.
Characterization of intracellular growth regulator icgR by utilizing transcriptomics to identify mediators of pathogenesis in Shigella flexneri
Infect. Immun.
2013
, vol. 
81
 (pg. 
3068
-
3076
)
[PubMed]
64
Baba
T.
Ara
T.
Hasegawa
M.
Takai
Y.
Okumura
Y.
Baba
M.
Datsenko
K.A.
Tomita
M.
Wanner
B.L.
Mori
H.
Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection
Mol. Syst. Biol.
2006
, vol. 
2
  
2006. 0008
65
Busby
S.
Kotlarz
D.
Buc
H.
Deletion mutagenesis of the Escherichia coli galactose operon promoter region
J. Mol. Biol.
1983
, vol. 
167
 (pg. 
259
-
274
)
[PubMed]
66
Tyson
K.L.
Cole
J.A.
Busby
S.J.
Nitrite and nitrate regulation at the promoters of two Escherichia coli operons encoding nitrite reductase: identification of common target heptamers for both NarP- and NarL-dependent regulation
Mol. Microbiol.
1994
, vol. 
13
 (pg. 
1045
-
1055
)
[PubMed]
67
Stella
S.
Cascio
D.
Johnson
R.C.
The shape of the DNA minor groove directs binding by the DNA-bending protein Fis
Genes Dev.
2010
, vol. 
24
 (pg. 
814
-
826
)
[PubMed]

Author notes

1

These authors contributed equally to this work.

Supplementary data