A Novel Two-Component Signaling System That Activates Transcription of an Enterohemorrhagic Escherichia coli Effector Involved in Remodeling of Host Actin

Abstract

Enterohemorrhagic Escherichia coli (EHEC) O157:H7 is responsible for worldwide outbreaks of bloody diarrhea, hemorrhagic colitis, and life-threatening hemolytic uremic syndrome. After colonizing the large intestine, EHEC forms attaching and effacing (AE) lesions on intestinal epithelial cells. These lesions cause destruction of the microvilli and elicit actin rearrangement to form pedestals that cup each bacterium individually. EHEC responds to a signal produced by the intestinal microbial flora, autoinducer-3 (AI-3), and the host hormones epinephrine and norepinephrine to activate transcription of the genes involved in AE lesion formation. These three signals, involved in interkingdom communication, are sensed by bacterial sensor kinases. Here we describe a novel two-component system, QseEF (quorum-sensing E. coli regulators E and F), which is part of the AI-3/epinephrine/norepinephrine signaling system. QseE is the sensor kinase and QseF the response regulator. The qseEF genes are cotranscribed, and transcription of qseEF is activated by epinephrine through the QseC sensor. A qseF mutant does not form AE lesions. QseF activates transcription of the gene encoding EspFu, an effector protein translocated to the host cell by the EHEC, which mimics a eukaryotic SH2/SH3 adapter protein to engender actin polymerization during pedestal formation. Expression of the espFu gene from a plasmid restored AE lesion formation to the qseF mutant, suggesting that lack of espFu expression in this mutant was responsible for the loss of pedestal formation. These findings suggest the QseEF is a two-component system involved in the regulation of AE lesion formation by EHEC.

Enterohemorrhagic Escherichia coli (EHEC) has caused numerous outbreaks of bloody diarrhea and hemolytic uremic syndrome throughout the world. In North America, the United Kingdom, and Japan, the main serotype associated with these outbreaks was O157:H7. EHEC strains are part of a group of enteric pathogens that includes enteropathogenic Escherichia coli, rabbit enteropathogenic E. coli, Citrobacter rodentium, and Hafnia alvei, all of which are able to cause a lesion on intestinal epithelial cells termed attaching and effacing (AE). This lesion is characterized by the destruction of the microvilli and the rearrangement of the cytoskeleton to form a pedestal-like structure that cups the bacterium individually (21).

The genes involved in AE lesion formation are contained within the locus of enterocyte effacement (LEE). This pathogenicity island is composed of 41 genes arranged in five operons that encode structural and secondary proteins required for the formation of the type III secretion system (TTSS), a bacterial adhesin (intimin), and several effector proteins, which are translocated into host cells through the TTSS (10, 12, 26). Aside from effector proteins, this TTSS also secretes proteins that constitute its translocon, such as EspA, which forms a filament that creates a sheath around the TTSS needle, and EspB and EspD, located at the distal end of the TTSS, which form a pore in the host cell membrane (see Fig. 5A) (18, 25). When EHEC is in close proximity to the epithelial lining, an effector protein, translocated intimin receptor (Tir), is secreted through the TTSS and into the host cell. Once inside the host cell, it embeds itself into the eukaryotic membrane in a hairpin loop formation, and its extracellular domain serves as a bacterial receptor that binds the adhesin intimin, allowing the bacteria to attach tightly to the eukaryotic cell (23).

FIG. 5.

QseF regulates transcription of espFu. (A) Schematic of the TTSS machinery in EHEC and injection of effectors, Tir and EspFu, into host cells, enabling tight adherence to the host cell and AE lesion formation. (B) Western blots were performed on whole-cell lysates and secreted protein preparations from mutant and WT cultures grown to an OD₆₀₀ of 1.0 in DMEM with antibodies specific to EspA and B, intimin, and Tir. The mutant qseE strain is complemented with qseE in pBADMycHisA. Production and secretion, where applicable, were not deficient in mutant strains. (C) β-Galactosidase assay using an espFu::lacZ transcriptional reporter construct in the WT strain, the qseF mutant, and the qseF mutant containing qseF in pACYC177. espFu activation in the mutant is at background levels (third bar). (D) FAS assay on the qseF mutant strain containing an empty vector (left) or espFu under an inducible promoter (right). In the presence of espFu, pedestal formation is restored. Cells were viewed at a magnification of ×640.

Inside the host cell, the cytoplasmic portion of Tir initiates a signaling cascade that leads to recruitment of Arp2/3 and N-WASP, eliciting actin nucleation to form pedestals characteristic of the AE lesion (1, 3, 20). In enteropathogenic Escherichia coli, Tir is phosphorylated on Tyr⁴⁷⁴ by the eukaryotic proteins Src, Arg, and Abl (30, 38). This creates a binding site for the actin SH2/SH3 adapter protein Nck, which activates N-WASP and recruits the Arp2/3 complex to initiate actin nucleation. EHEC Tir does not contain a corresponding tyrosine residue and forms pedestals in a Nck-independent manner. While Tir is the only LEE-encoded effector necessary for pedestal formation in EHEC, an additional non-LEE-encoded effector is also required. This prophage-encoded and primarily EHEC-specific protein, EspFu/TccP, binds to the (GTP-binding domain) GBD domain of N-WASP and is recruited to Tir, activating actin polymerization (4, 15).

Transcription of the LEE genes is activated through the autoinducer-3 (AI-3)/epinephrine/norepinephrine interkingdom signaling system. AI-3 is a bacterial quorum-sensing signal produced by the resident human intestinal microbial flora (36), whose production under defined environmental conditions requires a functional luxS gene in EHEC (39). Both epinephrine and norepinephrine are present in the gastrointestinal tract. Norepinephrine is synthesized within the adrenergic neurons present in the enteric nervous system (13). Although epinephrine is not synthesized in the enteric nervous system, it is synthesized in the central nervous system and in the adrenal medulla, where it is released into the bloodstream and acts in a systemic manner, eventually reaching the intestine (31). Both hormones modulate intestinal smooth muscle contraction, submucosal blood flow, and chloride and potassium secretion in the intestine, all important during bacterial infection (17). The EHEC sensor kinase, QseC, autophosphorylates in response to each of these signals (6) and initiates a complex signaling cascade that regulates the expression of genes encoding proteins necessary for AE lesion formation and the flagellar regulon (6-8, 34-36). QseC is the sensor of the QseBC two-component system. In two-component signaling, a sensor histidine kinase autophosphorylates in response to environmental cues. It then transfers this phosphate to an aspartate residue on a response regulator, which is often a DNA binding protein. Only when phosphorylated are the response regulators able to bind DNA and alter downstream transcriptional activities (40).

Here we report the identification of an additional two-component regulatory system, YfhK and YfhA, herein renamed quorum-sensing regulators E and F (QseEF). QseE is putative sensor kinase, while QseF is a putative response regulator. QseEF are part of the AI-3/epinephrine/norepinephrine signaling cascade and activate transcription of espFu to drive actin polymerization during AE lesion formation.

MATERIALS AND METHODS

Strains and plasmids.

All bacterial strains and plasmids utilized in this study are listed in Table 1. E. coli strains were grown aerobically in LB medium or Dulbecco's modified Eagle medium (DMEM) at 37°C unless otherwise specified. Antibiotics were added at the following concentrations: 100 μg ml⁻¹ for ampicillin, 30 μg ml⁻¹ chloramphenicol, 50 μg ml⁻¹ kanamycin, and 25 μg ml⁻¹ tetracycline.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant genotypea	Reference or source
Strains
86-24	Stx2+ EHEC strain (serotype O157:H7)	16
VS-94	luxS isogenic mutant of 86-24	35
NR01	86-24 qseE nonpolar mutant	This study
NR02	86-24 qseF nonpolar mutant	This study
NR04	NR01 complemented with plasmid pNR01	This study
NR06	NR02 complemented with plasmid pNR02	This study
KRL7	86-24 qseE polar mutant	This study
NR26	pK187 in NR02	This study
NR27	pKC471 in NR02	This study
NR28	pKH35-4 in BL21 DE3 cells	40
NR30	pNR15 in NR02	This study
NR31	pET16 in NR02	This study
NR32	p635 in NR02	This study
Plasmids
pBadMycHis	C-terminal Myc-His tag cloning vector	Invitrogen
pNR01	qseE in pBadMycHis	This study
pNR02	qseF in pBadMycHis	This study
pCVD442	Suicide vector	11
pBluescript	Cloning vector	Invitrogen
pKD3	pANTSγ derivative containing FRT-flanked chloramphenicol resistance	9
pKM201	λ red recombinase expression plasmid	29
pCP20	TS replication and thermal induction of FLP synthesis	5
pRS551	lacZ reporter gene fusion vector	33
pNR10	espFu regulatory region in pRS551	This study
pK187	Low-copy-no. KanR vector	4
pKC471	espFu-myc in pK187	4
pKH35-4	yfhA in pET21a(+)	40
pNR15	qseF in pACYC177	This study
pACYC177	Cloning vector	New England Biolabs
pVS262	espFu regulatory region in TOPO	This study
Topo	Commercial blunt-end cloning vector	Invitrogen
p635	espFu in pET16	Kind gift from John M. Leong
pET16	Expression vector	Novagen

^aTS, temperature sensitive; FRT, FLP recombination target.

Recombinant DNA techniques.

Standard methods were used to perform restriction digestions, plasmid purification, PCR, ligation, transformation, and gel electrophoresis (32). All oligonucleotide primers are listed in Table 2. Plasmid pNR01 was constructed by amplifying the qseE gene from the EHEC strain 86-24 by use of Pfx DNA polymerase (Invitrogen) with primers YfhKFBAD and YfhKRBAD and cloning the resulting PCR product into the EcoRI-KpnI cloning site of vector pBADMycHisA (Invitrogen). Plasmid pNR02 was constructed by amplifying the qseF gene from the EHEC strain 86-24 by use of Pfx DNA polymerase (Invitrogen) with primers QseFFBAD and QseFRBAD and cloning the resulting PCR product into the KpnI-HindIII cloning site of vector pBADMycHisA (Invitrogen).

TABLE 2.

Oligonucleotides used in this study

Primer name	Sequence	Description
yfhKP1	GGCAAAGCCTGAATGCGCCTTAGCGACCAGGCGGCGCTGGTCAACCGCACCACGCTTATCGATGCCCGGCGCAGCGAAGCAATGACCAACGCGGCGCTGGATGTAGGCTGGAGCTGCTTC	Construction of the QseE isogenic mutant
yfhKP2	TTGCCCGCTCTCGTCGACCAGATACAGTTCCCCTTGCATACGGCGAATACAATCCCTGGCAATGCTTAATCCCAGACCGCTGCCCTTCACCGCCCCTTTTATATGAATATCCTCCTTA	Construction of the QseE isogenic mutant
yfhARedP1	GAAACTGCTTGGCCTGCGCCTGACCAGCGAAGGCTACAGCGTGGTCACGGGTGTAGGCTGGAGCTGCTTC	Construction of the QseF isogenic mutant
yfhARedP2	AATTCTGTCCGGTTGCGCCCCGCCATTCTCGCCGCGTGGGTGACGTTGCCCATATGAATATCCTCCTTA	Construction of the QseF isogenic mutant
QseFFBAD	GGTACCGTGAAAAGCCCGCGCCATCCA	Construction of pNR02
QseFRBAD	AAGCTTATCGTTTGCATCCAGCTCGTGT	Construction of pNR02
YfhKFBAD	GGTACCTATCTGAACTTCCCCTCGGTT	Construction of pNR01
YfhKRBAD	GAATTCCCTTTCGTGTTTTTCGACGACGG	Construction of pNR01
EspFuF	CGCGGATCCCTGTCGGCTCTCTTCTAGAT	espFu activity assay, EMSA, RT-PCR
EspFuR	CGC GGA TCC ATA TTG CGG TTG ACG GTT GG	espFu activity assay
EspFuSEQ	ATA TTG CGG TTG ACG GTT GG	Primer extension, sequencing ladder
QseE RTR	CGC GCC ATG ATC TTC GA	Real-time analysis
QseE RTF	CCC TTC ACC GCC CCT TT	Real-time analysis
EspFuCInv	TGC GGC GAT GTA TAA ATG AC	Primer extension
EspFuBInv	ACA GCC ATT CCT CCT GTG T	EMSA
EspFuCInv	TGC GGC GAT GTA TAA ATG AC	EMSA
EspFuRN	GTTCATTTTGTACTGGCGGC	RT-PCR
EspFuRT2	GCATCCTATTTATTGCTCACGTTA	RT-PCR
EspFuProbeR	CGAGCGCTTAGATGTATTAATGCC	RT-PCR
EspFuProbeF	ATGATTAACAATGTTTCTTCACTTTTTCC	RT-PCR
QseF RT R	CGT AAG CTG CTG CAA ATT ACC A	Real-time analysis
QseF RT F	CGC CCC GCC ATT CTC	Real-time analysis
3a	GAA ATG GTA CAG GCC ACG CGC ACT	RT-PCR
3b	ATC AGC ACA ATC CAT CGC CCG	RT-PCR
6a	CGG GCG ATG GAT TGT GCT GAT	RT-PCR
6b	TAT GTC GCT CTG CCG CCT GGC	RT-PCR

Construction of isogenic mutants.

Construction of the qseE polar mutant (KRL7) was carried out by replacing qseE with the tet gene amplified from plasmid pACYC184 by use of suicide vector pCVD442, as previously described (11). Construction of nonpolar isogenic qseE (NR01) and qseF (NR02) mutants was carried out as previously described (9). Briefly, 86-24 cells containing pKM201 were prepared for electroporation. PCR products of a chloramphenicol resistance cassette with 50 to 80 overhangs homologous to either qseE or qseF were generated using primers YfhKP1 and YfhKP2 (qseE) and YfhARedP1 and YfhARedP2 (qseF) and with pKD3 as a template. Electroporation of the PCR product into these cells was performed; cells were incubated at 37°C for 2 h and plated on media containing 30 μg ml⁻¹ chloramphenicol overnight at 37°C. Resulting colonies were patched for chloramphenicol resistance and ampicillin sensitivity and PCR verified for the absence of the gene. The chloramphenicol resistance cassette was then resolved from the mutants in order to create nonpolar, isogenic qseE (NR01) and qseF (NR02) mutants. Plasmid pCP20, encoding a resolvase, was electroporated into the mutant strain, and resulting colonies were patched for chloramphenicol sensitivity. Both mutants were complemented with plasmids pNR01 for NR01-generating strain NR04 and with pNR02 for NR02-generating strain NR06.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotting.

For blots using whole-cell lysates, total proteins were extracted from strains 86-24, NR01, NR02, and NR04 grown in DMEM to an optical density at 600 nm (OD₆₀₀) of 1.0. Briefly, 3 ml of culture was pelleted (13,000 rpm for 5 min at 4°C), resuspended in 300 μl lysis buffer (50 mM Tris-HCl [pH 7.5], 50 mM NaCl, 5% glycerol, 1 mM dithiothreitol, and 30 mM phenylmethylsulfonyl fluoride), subjected to lysozyme addition to a final concentration of 300 μg/ml, incubated at 4°C for 4 h, and DNase I treated for 45 min at 4°C; cell debris was then pelleted (13,000 rpm for 10 min at 4°C), and supernatant containing whole-cell protein was removed. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotting procedures were performed as previously described (32), and products were probed with polyclonal antisera against either EspA, EspB, intimin, or Tir (kindly provided by James Kaper). Proteins were detected using enhanced chemiluminescence (Bio-Rad). Equal amounts of whole-cell lysate protein were determined using the Bradford assay (32).

Secreted proteins.

Secreted proteins from 86-24, NR01, NR02, and NR04 were harvested as previously described by Jarvis et al. (19). Briefly, bacteria were grown aerobically in DMEM at 37°C and collected at late exponential growth phase (OD₆₀₀, 1.0). Total secreted protein from culture supernatants was separated by removing bacteria by use of centrifugation and filtration and then precipitating the secreted proteins present in the supernatant with trichloroacetic acid. The samples were then subjected to immunoblotting with rabbit polyclonal antisera to EspA, EspB, and Tir (kindly provided by James Kaper) and visualized with enhanced chemiluminescence (Bio-Rad).

Reporter gene assays.

Plasmid pNR10 was generated by amplifying the regulatory region of espFu with primers EspFuF and EspFuR and cloning the resulting fragments into the BamHI restriction sites of plasmid pRS551 (33). Because of plasmid compatibility issues, to perform these assays, the qseF mutant was complemented by cloning the qseF gene into the SmaI restriction site of pACYC177, generating plasmid pNR15. Bacteria containing the lacZ fusions were grown overnight at 37°C in LB containing the appropriate selective antibiotic. Cultures were diluted 1:100 and grown in DMEM to an OD₆₀₀ of 1.0 at 37°C. These cultures were then assayed for β-galactosidase activity by use of o-nitrophenyl-β-d-galactopyranoside as a substrate as described previously (28).

RNA extraction and real-time RT-PCR studies.

Overnight cultures grown aerobically in LB at 37°C of 86-24 and VS94 (luxS mutant) were diluted 1:100 in DMEM and grown aerobically at 37°C. RNA from three biological replicate cultures of each strain was extracted at mid-exponential growth phase (OD₆₀₀, 0.5) and late exponential growth phase (OD₆₀₀, 1.0) by use of a RiboPure bacterial RNA isolation kit (Ambion) following the manufacturer's guidelines. The primers used in the real-time assays were designed using Primer Express v1.5 (Applied Biosystems) (Table 2). Real-time reverse transcription-PCR (RT-PCR) was performed in a one-step reaction using an ABI 7500 sequence detection system (Applied Biosystems).

For each 20-μl reaction volume, 10 μl 2× SYBR master mix, 0.1 μl Multi-scribe reverse transcriptase (Applied Biosystems), and 0.1 μl RNase inhibitor (Applied Biosystems) were added. The amplification efficiency of each of the primer pairs was verified using standard curves of known RNA concentrations. Melting curve analysis was used to ensure template specificity by heating products to 95°C for 15 s, followed by cooling to 60°C and heating to 95°C while monitoring fluorescence. Once amplification efficiency and template specificity were determined for each primer pair, relative quantification analysis was used to analyze the unknown samples by use of the following conditions for cDNA generation and amplification: 48°C for 30 min, 95°C for 10 min, and 40 cycles at 95°C for 15 s and 60°C for 1 min. The rpoA (RNA polymerase subunit A) gene was used as the endogenous control.

Detection, quantification, and statistical analysis.

Data collection was performed using the ABI Sequence Detection 1.3 software (Applied Biosystems). Data were normalized to levels of rpoA and analyzed using the comparative critical threshold method previously described (2). The expression levels of the target genes at the different growth phases were compared using the relative quantification method (2). Real-time data are presented as change (n-fold) compared to wild-type (WT) levels at the early exponential growth phase and change (n-fold) compared to WT levels at late exponential growth phase. Error bars represent the standard deviation of the ΔΔC_T value, where C_T is the critical threshold (2). Statistical significance was determined by use of Student's t test. A P value of <0.05 was considered significant.

FAS test.

Fluorescein actin staining (FAS) assays were performed as previously described by Knutton et al. (24). In brief, overnight bacterial cultures grown aerobically in LB at 37°C were diluted 1:100 and used to infect confluent monolayers of HeLa cells grown on glass coverslips at 37°C and 5% CO₂. Cells were grown for 6 h at 37°C and 5% CO₂. The coverslips were then washed, permeabilized with 0.2% Triton X-100, and treated with fluorescein isothiocyanate-labeled phalloidin to visualize actin accumulation, and propidium iodide was added to stain bacteria. Samples were visualized by immunofluorescence with a Zeiss Axiovert microscope. The entire field of at least six coverslips from each strain was examined, and images of AE lesions were taken.

Purification of QseF.

QseF was purified after being expressed from plasmid pKH35-4 (40). E. coli BL21(DE3) containing plasmid pKH35-4 was grown at 37°C in LB to an OD₆₀₀ of 0.7, at which point isopropyl-β-d-thiogalactopyranoside was added to a final volume of 0.5 mM and allowed to induce for 3 hours. His-tagged QseF protein was then purified under native conditions by use of a nickel column according to the manufacturer's instructions (QIAGEN).

EMSAs.

In order to study the binding of QseF to the espFu promoter, electrophoretic mobility shift assays (EMSAs) were performed using purified QseF-His and PCR-amplified DNA probes. Probes were end labeled with [γ-³²P]ATP (NEB) by use of T4 polynucleotide kinase according to standard procedures (32) and gel purified using a QIAGEN PCR purification kit. EMSAs were performed by adding increasing amounts of purified QseF protein (0 to 3 μg) to end-labeled probe (10 ng) in binding buffer [500 μg ml⁻¹ bovine serum albumin (NEB), 50 ng μl⁻¹ poly(dI-dC), 60 mM HEPES (pH 7.5), 5 mM EDTA, 3 mM dithiothreitol, 300 mM KCl, 25 mM MgCl₂] with or without 0.1 M acetyl phosphate for 20 min at 4°C. A 5% Ficoll solution was added to the mixtures immediately before loading. Reaction mixtures were then electrophoresed on a 6% polyacrylamide gel, dried, and exposed to KODAK X-OMAT film.

RNA purification and primer extension analysis.

RNA purification was performed according to the manufacturer's instructions using the TRIzol reagent (Invitrogen). RNA was isolated from strain 86-24 grown in DMEM aerobically at 37°C to an OD₆₀₀ of 1.0. Primer extension analysis was then performed as described previously (27). Briefly, EspFuSEQ, approximately 40 bp downstream of the ATG start site, and EspFuCInv, approximately 100 bp upstream of the ATG start site (Table 2), were end labeled using [γ-³²P]dATP. A total of 35 μg of RNA was incubated with the end-labeled primer and reverse transcribed using a SuperScript first-strand synthesis system for RT-PCR (Invitrogen) according to the manufacturer's instructions. A sequencing ladder was generated using the Sequenase version 2.0 DNA sequencing kit (USB) according to the manufacturer's instructions. The sequencing ladder was generated using primer EspFuSEQ and plasmid pVS262. Plasmid pVS262 was created by amplifying the espFu regulatory region and cloning the product into blunt-end TOPO (Invitrogen). These experiments were repeated three times, with two different primers to ensure the correct mapping of this promoter.

RESULTS

QseEF constitute a novel putative two-component system required for pedestal formation.

The gene yfhK encodes a 55-kDa putative sensor kinase protein predicted to lie in the inner membrane. This protein contains two transmembrane domains, a signal histidine kinase domain, and an ATPase domain. The gene yfhA encodes a 49-kDa putative response regulator and contains a response regulator domain, a helix-turn-helix DNA binding domain, and a σ⁵⁴ activation domain (Fig. 1). These genes were originally identified in a microarray study comparing differential gene expression for a WT EHEC strain and that for an isogenic EHEC luxS knockout strain. Under defined environmental conditions, the luxS mutant does not produce the AI-3 bacterial signal (39). Transcription of yfhK was up-regulated twofold in the luxS mutant in late exponential growth compared to that for WT (35). To confirm the array data, we performed real-time RT-PCR analysis using cDNA synthesized from RNA extracted from WT and an isogenic luxS mutant during mid-exponential (OD₆₀₀, 0.5) and late exponential (OD₆₀₀, 1.0) growth. Transcription of yfhK was mildly decreased 0.5-fold for the luxS mutant during mid-exponential growth (P ≤ 0.0075), whereas it was increased twofold for the luxS mutant during late exponential growth (P ≤ 0.0051) (Fig. 2A). In the same microarray study, transcription of qseBC and qseA was altered in the luxS mutant compared to WT (34, 37), leading to the subsequent description of these regulators as being part of the AI-3/epinephrine/norepinephrine signaling cascade in EHEC (34, 36, 37). These results suggested that yfhK and yfhA could also be part of the AI-3/epinephrine/norepinephrine signaling cascade. QseA, a LysR family regulator (35), is activated by the AI-3/epinephrine/norepinephrine signaling cascade and subsequently activates the LEE region (36). Transcription of yfhK in a qseA mutant during mid-exponential growth is increased twofold (P ≤ 0.0028) and decreased 0.5-fold at late exponential growth (P ≤ 0.0047) (Fig. 2A), providing additional evidence that yfhK and yfhA are part of this signaling pathway. Finally, transcription of yfhK and yfhA is increased by epinephrine in WT EHEC in late exponential growth, while it is decreased in a qseC mutant either in the presence of epinephrine or in its absence (Fig. 2B) (QseC is a sensor for AI-3, epinephrine, and norepinephrine, and an EHEC qseC mutant is unable to sense these three signals [6]). Hence, we renamed these genes quorum-sensing regulators E and F (qseEF), in which QseE is a putative sensor kinase and QseF is a putative response regulator.

FIG. 1.

QseEF is a predicted two-component signaling system that is a part of the EHEC cell-to-cell signaling cascade. Shown are a schematic of QseE and QseF and their respective domains and the expected localization of QseE and QseF within the bacterial cell and proposed signaling mechanism.

FIG. 2.

(A) Real-time analysis of qseE expression in a luxS mutant and a qseA mutant compared to WT. qseE expression is down at mid-log growth and up at late log growth for the luxS mutant. Conversely, qseE expression is up at mid-log phase and down at late log phase for the qseA mutant. (B) Real-time analysis of qseF and qseE expression in a qseC mutant compared to WT in the presence and absence of epinephrine (Epi). qseF and qseE expression is increased in the presence of epinephrine and decreased in the qseC mutant with or without epinephrine.

To test QseEF's role in this pathway, a polar mutation in qseE was constructed utilizing a tetracycline resistance cassette. We then tested whether any of the phenotypes regulated by the AI-3/epinephrine/norepinephrine signaling cascade, namely, AE lesion (pedestal) formation and flagellation and motility, were altered. We used FAS (24) to visualize pedestal formation in the WT and mutant strains. Actin was stained using fluorescein isothiocyanate-phalloidin, and HeLa cell nuclei and bacteria were stained with propidium iodide. Pedestal formation was visualized as brilliant patches of stained actin (green) localized underneath a red bacterium. Wild-type EHEC formed pedestals; however, the qseE mutant was abrogated for pedestal formation (see Fig. 4A). There were no differences in flagellar expression as seen by Western blotting and no differences in motility between the WT strain and the qseE mutant (data not shown), indicating that QseEF is involved in the regulation of the genes required for pedestal formation but not motility.

FIG. 4.

(A) Detection of AE lesion formation using the FAS test on HeLa cells comparing WT and a qseE polar (P) mutant. Green shows the HeLa cell actin cytoskeleton, and red shows the bacteria and cell nuclei. WT EHEC forms pedestals (top left); the qseE mutant does not (top right). Nonpolar mutant strains were tested on HeLa cells by use of a FAS assay. Nonpolar mutations reveal that only the qseF mutant is unable to form AE lesions. The qseE mutant formed AE lesions, while the qseF mutant could not (left bottom two panels). Each mutant strain was complemented in a pBADMycHis vector. Cells were viewed at a magnification of ×640. (B) QseF is cross-phosphorylated by several E. coli noncognate sensors. It is not known if QseC is able to phosphorylate QseF.

The qseE and qseF genes are contained in a cluster of several genes including yfhG, an uncharacterized gene, and glnB, which encodes the PII protein involved in nitrogen regulation (Fig. 3). To determine if these genes are transcribed in one operon, we designed primers flanking qseE to qseF and qseF to glnB and performed RT-PCR. RT-PCR was performed using cDNA synthesized from RNA isolated from the WT strain. RT-PCR indicated that transcriptional read-through occurred across the intergenic region between qseE and qseF and qseF and glnB (Fig. 3), suggesting that these four genes are transcriptionally linked. Because qseE, yfhG, qseF, and glnB are cotranscribed, we constructed nonpolar mutations in qseE and qseF. Each mutant was tested for its ability to form pedestals on HeLa cells by use of the FAS assay. The nonpolar qseE mutant was still able to form pedestals; however, the qseF mutant was abrogated for pedestal formation. Pedestal formation was restored in the qseF mutant upon complementation with the QseF gene on a plasmid (Fig. 4A).

FIG. 3.

RT-PCR analysis showing that qseE, yfhG, qseF, and glnB are cotranscribed. In lanes 1 to 3, primers flanking qseE to qseF show no product when no RT is added (lane 1) and show product when either a genomic DNA control or cDNA is used (lanes 2 and 3); in lanes 4 to 6, primers flanking qseF to glnB show no product when no RT is added (lane 4) and show product when either the genomic DNA control or cDNA is used (lanes 5 and 6). Neg. Con., negative control.

QseE and QseF are predicted to be a cognate two-component system. This is supported by the fact that the genes encoding QseEF are cotranscribed and that QseE phosphorylates QseF (40). However, QseF is also phosphorylated by multiple noncognate sensor kinases (UhpB, BaeS, EnvZ, RstB) (Fig. 4B) (40), which may account for the contrasting phenotypes of the qseE and qseF mutants.

QseF activates expression of espFu to induce pedestal formation.

Several factors are required for pedestal formation in EHEC, including components of the TTSS, intimin, and the effector protein Tir (21) (Fig. 5A). We suspected that one of the genes encoding proteins that comprise different portions of this system might be activated by QseF. We tested the expression of EspA, EspB, Tir, and intimin by Western blotting of whole-cell lysates and found no defect in production of any of these proteins in a qseF mutant (Fig. 5B). In addition, we tested secretion of the TTSS proteins Tir (data not shown), EspA, and EspB, and again no defect was seen in secretion of these proteins (Fig. 5B).

EspFu was recently identified as a non-LEE-encoded effector protein necessary for pedestal formation in EHEC (4, 15), possibly serving as a Nck-like protein to recruit N-WASP and Arp-2/3 to Tir. We tested whether QseF activated espFu transcription using an espFu::lacZ transcriptional reporter fusion. In the qseF mutant, transcription of espFu was abolished. Transcription of espFu in this mutant was restored to WT levels upon complementation with a functional copy of qseF (Fig. 5C). Transcription of espFu was only mildly decreased in the qseE mutant (twofold) (Fig. 5C). This mild decrease in the qseE mutant, in contrast to the striking decrease in the qseF mutant, can again be attributed to the fact that QseF can be phosphorylated by multiple kinases (Fig. 4B). To further investigate whether the lack of espFu expression was the reason for the defect on pedestal formation in the qseF mutant, we performed FAS assays using the qseF mutant containing a plasmid expressing espFu. Under these conditions, we were able to reconstitute pedestal formation (Fig. 5D).

A recent report suggested that espFu is cotranscribed in an operon with espJ. The authors of this study created gfp transcriptional fusions with the upstream regions of these genes and found low levels of expression of espFu::gfp. However, no primer extensions, Northern blots, or RT-PCR analysis was performed in these studies (14). Given that there are 324 bp in between the espJ and espFu genes (AE005419), it is possible that espFu is a stand-alone gene. To examine the operon structure of espFu and upstream genes (espJ and Z3069), we used RT-PCR. We detected transcription of espFu using an internal primer set to this gene (Fig. 6). No amplification was observed with primers flanking espFu and either of the two upstream genes (Fig. 6). These data indicate that espFu and espJ are not transcriptionally linked. The contrast between the data obtained by our group and those obtained by Garmendia et al. (14) may be due to the low sensitivity of gfp reporters. To identify the espFu promoter, we performed primer extensions. These primer extension studies were performed three times using two different primers to ensure the correct mapping of the transcriptional start site of espFu. The espFu gene contains a conserved extended σ⁷⁰ consensus sequence (Fig. 7). The combined RT-PCR and primer extension data led us to conclude that espFu is encoded by itself and driven by its own promoter. In silico analysis indicates that QseF contains a σ⁵⁴ activator domain (www.promscan.uklinux.net/; Fig. 1). Because espFu contains a σ⁷⁰ rather than the alternative σ⁵⁴ promoter, it is likely that QseF indirectly activates transcription of this gene, possibly through an additional regulatory protein. Consistent with this hypothesis, we were unable to observe direct binding of QseF to the espFu regulatory region by use of EMSAs, even after addition of acetyl phosphate as a phosphate source for activation of QseF (data not shown). This suggests that QseF does not directly bind to the promoter region of espFu and that QseF regulation of espFu is indirect, involving an intermediary factor transcribed in a σ⁵⁴-dependent fashion.

FIG. 6.

Reverse transcription analysis was performed to evaluate the operon structure of espFu by use of espFu internal primers (1., EspFuProbeR and EspFuProbeF), primers flanking espFu and the two upstream genes espJ and z3069 (2., EspFuRN and EspFuRT2), and primers flanking espFu and one gene upstream, espJ (3., EspFuF and EspFuRN). Only the primers to amplify espFu alone from WT cDNA showed a product, indicating that espFu is a stand-alone gene. For each primer set, PCR was performed using genomic DNA as a positive control and RNA with no reverse transcriptase added as a negative control.

FIG. 7.

The promoter of espFu was mapped using primer extension with primers downstream and upstream of the ATG start site. The promoter region of espFu contains an extended σ⁷⁰ consensus sequence corresponding to −10 and −35 from the +1 transcriptional start site that was mapped.

DISCUSSION

The cell-to-cell signaling system utilized by EHEC to control virulence is highly complex, involving many levels of gene regulation. Although we know that multiple EHEC virulence genes are activated in response to the host epinephrine/norepinephrine and the bacterial signal AI-3, numerous aspects of this regulation are still unresolved, including how the signals are sensed and transduced by the bacterial cell as well as the global hierarchy of signaling that leads to gene activation. With respect to this signaling, EHEC virulence can be divided into two main pathways: (i) flagellation and motility and (ii) the formation of AE lesions (Fig. 8). The two-component system QseBC activates the flagellar regulon in response to AI-3 and epinephrine/norepinephrine. QseC's autophosphorylation in response to these signals can be blocked only by α-adrenergic antagonists (6). AE lesion formation, however, can also be blocked by a β-adrenergic antagonist (36). This suggests that more than one sensor kinase responds to AI-3 and epinephrine/norepinephrine. QseE could function as an additional sensor, transducing environmental signals in the direction of AE lesion formation rather than in that of flagellar regulation.

FIG. 8.

Schematic of AI-3, epinephrine, and norepinephrine signaling in EHEC. AI-3, epinephrine, and norepinephrine are sensed by sensor kinases in the EHEC membrane. QseBC regulate flagella and motility, while QseEF may be a second two-component system that regulates AE lesion formation in conjunction with other EHEC regulators such as QseA.

Here we describe QseEF, a two-component system regulated by cell-to-cell signaling in EHEC. These proteins regulate pedestal formation through indirect activation of espFu. The response regulator QseF does not bind directly to the σ⁷⁰ promoter region of espFu. Instead, it is likely that QseF regulates espFu through an unidentified σ⁵⁴-dependent intermediate. Based on our data, QseEF do not appear to regulate the LEE genes. These data reinforce the complexity of EHEC virulence signaling and indicate that another two-component system, along with intermediates such as QseA, is involved in AE lesion formation by EHEC (Fig. 8).

The timing of flagellar production and formation of AE lesions is crucial for successful EHEC colonization of the human intestine, thus making the kinetics of activation of these genes equally important. Interaction of environmental signals with multiple sensors would allow a precise timing mechanism enabling successive rather than simultaneous production of these systems and, ultimately, more efficiency. We suspect that upon initial colonization of the intestine, EHEC receives signals from the host and commensal flora, allowing the bacteria to activate the flagellar genes and swim across the mucus layer of the intestine. Once in proximity to the epithelium of the intestine, the TTSS genes are activated for AE lesion formation. At this point, the flagellar genes must be down-regulated. It is further intriguing that there is yet an additional level of fine tuning through QseEF regulation of espFu transcription but not LEE expression. Determining how these two systems are connected and how they use cell-to-cell signaling to orchestrate regulation in response to the environment will give us further insight not only into cell-to-cell signaling but also into the biology of the intestine and interaction between pathogenic and commensal organisms.

Only a small number of signals for two-component systems are currently known. This along with the difficulty in purifying membrane proteins prompted Yamamoto et al. (40) to use only the cytoplasmic region of the sensor kinases in their study of all two-component systems in E. coli. Although QseE has been shown to phosphorylate QseF (40), the physiological signal that activates QseE is still unknown. We have attempted to purify QseE and insert it into liposomes in order to test possible chemical and environmental signals to which it responds. However, despite attempting to express QseE in several different vector contexts, this protein remains insoluble and difficult to purify in large quantities. In addition, we have attempted to examine potential interaction between the two two-component systems, QseBC and QseEF, by conducting cross-phosphorylation experiments in which we assess whether phosphorylated QseC can phosphorylate QseF. However, the similar sizes of QseF (49 kDa) and QseC (56 kDa), as well as their similar isoelectric points (QseE, 5.6; QseC, 6.0), have made separation of these proteins after phosphotransfer assays challenging. Alternative methods to resolve these problems are under investigation.

Two-component signaling systems are found primarily in prokaryotes and have not yet been identified in animals and humans, making them an ideal target for drug inhibitors and therapeutics (39). This is important in an age where antimicrobial resistance is becoming increasingly prevalent. In particular, for EHEC infection, no current treatment exists (22), and antibiotics can worsen the infection and lead to hemolytic uremic syndrome. Blocking EHEC's signaling mechanisms for expressing virulence genes could potentially render these bacteria harmless. Elucidating the EHEC hierarchy of virulence expression is crucial in order to utilize this pathway in treatment.