regulatory control of the escherichia coli o157:h7 lpf1 by...

JOURNAL OF BACTERIOLOGY, Apr. 2011, p. 1622–1632 Vol. 193, No. 70021-9193/11/$12.00 doi:10.1128/JB.01082-10Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Regulatory Control of the Escherichia coli O157:H7 lpf1 Operonby H-NS and Ler�

Maricarmen Rojas-Lopez,1,2 Margarita M. P. Arenas-Hernandez,2,3

Abraham Medrano-Lopez,4 Claudia F. Martínez de la Pena,3Jose Luis Puente,4 Ygnacio Martínez-Laguna,2,3

and Alfredo G. Torres1,5*Department of Microbiology and Immunology, Department of Pathology,5 and Sealy Center for Vaccine Developments, University of

Texas Medical Branch, Galveston, Texas 77555-10701; Posgrado en Microbiología,2 Centro de Investigaciones enCiencias Microbiologicas,3 B. Universidad Autonoma de Puebla, Apartado Postal 1622, Puebla,

Mexico; and Departamento de Microbiología Molecular, Instituto de Biotecnología,Universidad Nacional Autonoma de Mexico, Cuernavaca, Morelos 62260, Mexico4

Received 10 September 2010/Accepted 21 January 2011

Long polar fimbriae 1 (Lpf1) of Escherichia coli O157:H7 is a tightly regulated adhesin, with H-NS silencingthe transcriptional expression of the lpf1 operon while Ler (locus of enterocyte effacement-encoded regulator)acts as an antisilencer. We mapped the minimal regulatory region of lpf1 required for H-NS- and Ler-mediatedregulation and found that it is 79% AT rich. Three putative sites for H-NS binding were identified. Two of them,named silencer regulatory sequence 1 (SRS1) and SRS2, are located on a region that covers both of the lpf1promoters (P1 and P2). The third putative H-NS binding site is located within the lpfA1 gene in a regionextending from �258 bp to �545 bp downstream of ATG; however, this site does not seem to play a role in lpfA1regulation under the conditions tested in this work. Ler was also found to interact with Ler binding sites(LBSs). Ler binding site 1 (LBS1) and LBS2 are located upstream of the two promoters. LBS1 overlaps SRS1,while LBS3 overlaps the P1 promoter and SRS2. Based on the experimental data, we propose that H-NSsilences lpf1 expression by binding to both of the SRSs on the promoter region, forming an SRS–H-NS complexthat prevents RNA polymerase-mediated transcription. A model of the regulation of the lpfA1 operon of E. coliO157:H7 by H-NS and Ler is discussed.

Enterohemorrhagic Escherichia coli (EHEC) is an importantintestinal pathogen and causative agent of diarrheal diseasecommonly associated with the consumption of contaminatedfood. The disease that occurs in humans ranges from uncom-plicated diarrhea to hemorrhagic colitis and, in some cases,hemolytic uremic syndrome (21). EHEC belongs to a group ofpathogens that cause a histopathological lesion known as attach-ing and effacing (A/E), which is produced by the gene productsencoded within a pathogenicity island termed LEE (locus of en-terocyte effacement) (reviewed in references 15 and 21).

Although it is well documented that the EHEC serotypeO157:H7 ability to colonize the intestinal epithelia and causeA/E lesions is dependent on the expression of the LEE-en-coded protein intimin, the presence of other adhesins on thesurface of EHEC is also important for the colonization process(reviewed in reference 31). Analysis of sequenced genomes ofdifferent EHEC O157:H7 strains have revealed the presence ofat least 23 loci encoding putative adhesins, with 10 of whichcorresponding to potential fimbrial adhesin gene clusters and13 regions encoding nonfimbrial adhesins (13, 23). We haveshown that E. coli O157:H7 contains two nonidentical locus-encoding homologues of the long polar fimbriae (Lpf), firstdescribed in Salmonella enterica serovar Typhimurium (re-

viewed in reference 31). The maximum expression of the E.coli O157:H7 lpf1 loci occurs in late exponential growth phasein tissue culture media at pH 6.5 and 37°C (30). Expression oflpf1 has been shown to be driven by two putative sigma 70-dependent promoters (29). We also demonstrated that thechromosomal lpfABCC�DE fimbrial operon encodes Lpf1 andthat the expression of the E. coli O157:H7 lpf1 in E. coli K-12has been linked to increase adherence to tissue-cultured cellsand is associated with the appearance of long, fine fimbrialstructures (28). Further, E. coli O157:H7 strains harboringmutations in one or both of the lpf loci have diminished colo-nization abilities in different animal models of infection (14,30) and alter human intestinal tissue tropism (11).

Ler (LEE-encoded regulator), a protein closely related toH-NS, is not only involved as a key factor for the A/E pheno-type but also has a role as a positive regulator of virulencefactors outside the LEE. The global regulatory effect of Lerincludes genes such as espC in enteropathogenic E. coli(EPEC) or the EHEC virulence plasmid pO157-carried stcEgene (reviewed in reference 18). H-NS, a regulatory silencerprotein, exerts a more global effect that extends to more than400 genes, many of them involved in metabolism and patho-genesis (9, 12, 18). In the case of lpf1, we reported that itsexpression is regulated by H-NS, presumably binding to theregulatory sequence upstream of lpfA1 and silencing its tran-scription, while Ler acts as an antisilencer, outcompeting therepression exerted by H-NS (28, 29).

Analysis of the silencing/antisilencing mechanisms mediated

* Corresponding author. Mailing address: Department of Microbi-ology and Immunology, University of Texas Medical Branch, 301 Uni-versity Blvd., Galveston, TX 77555-1070. Phone: (409) 747-0189. Fax:(409) 747-6869. E-mail: [email protected].

� Published ahead of print on 28 January 2011.

1622

on June 1, 2018 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

by the H-NS and Ler proteins, for example, on the LEE2 andLEE3 divergent LEE-encoded operons in EPEC has revealedthat H-NS represses expression by binding to two negativeregulatory sequences denoted silencing regulatory sequence 1(SRS1) and SRS2. These SRS regions flank the LEE2 andLEE3 promoters and generate a nucleoprotein repressor com-plex. Under conditions that induce LEE-encoded gene expres-sion, Ler counteracts H-NS repression by preferentially bind-ing to one of the silencers and displacing H-NS from it eitherby direct competition for its binding sites or by altering thelocal DNA architecture. This in turn disrupts the repressorcomplex and permits transcription of both promoters to pro-ceed (2, 26). Further, Haack and colleagues (12) have demon-strated that cis-acting DNA sequences are necessary for Lerbinding at the LEE5 operon. LEE5 is also negatively regulatedby H-NS, forming a nucleoprotein complex similar to thatproposed for the LEE2-LEE3 operon. In this case, Ler binds

to one side of the LEE5 operon, increasing transcription atboth promoters and disrupting the repression exerted by theH-NS-DNA complex (12, 18).

Our prior studies have suggested that the regulatory controlof the lpf1 operon could share similarities with the Ler- andH-NS-dependent mechanism controlling the expression of LEE-encoded genes, perhaps following a strategy similar to that of thesilencing/antisilencing of the LEE2 and LEE3 operons (2, 29).Therefore, the aim of the present study was to investigate whetherthe regulation of the EHEC lpf1 operon requires binding of Lerand H-NS to AT-rich sequences and whether specific regions forbinding can be identified in the vicinity of the promoter region orwithin the structural lpfA1 gene.

MATERIALS AND METHODS

Bacterial strains, plasmids, media, and growth conditions. Bacterial strainsand plasmids used in this study are listed in Table 1. Strains were routinely grown

TABLE 1. Bacterial strains and plasmids used in this study

Strain or plasmid Characteristics Reference or source

E. coli strainsEDL933 Prototype EHEC O157:H7 A. G. Torres stock

collectionAC425 EDL933 hns::Tn10 Tcr 29CB49 EDL933 ler::kan Kmr 29SDP01 EDL933 hns::Tn10 ler::kan Tcr Kmr 29MC4100 E. coli K-12 araD139 �(argF-lac)169 lam-flhD5301 fruA25 relA1 rpsL150 (Strr)

rbsR22 deoC13

BL21(DE3)/pLys F� ompT hsdSB(rB� mB

�) gal dcm �(DE3) pLys Cmr Invitrogen

PlasmidspPLPFA pRS551 with EcoRI/BamHI 1,164-bp lpf promoter region 28pT3-HNS pMPM-T3 derivative carrying the hns gene, Kmr 2pK3-HNS pT3-HNS SalI and SacI fragment containing hns cloned into pMPM-K3, Tcr This studypMPM-K3 Low-copy-no. cloning vector and p15 derivative, Kmr 17pT6Ler pMPM-T6 derivative expressing Ler-His6 under the control of arabinose-

inducible promoter, Tcr1

pT6HNS pMPM-T6 derivative expressing H-NS–His6 under the control of arabinose-inducible promoter, Tcr

1

TABLE 2. Primers used in this study

Primer Sequence Characteristic Reference

PLPFA258F 5�-CTTATTATTTCCCGGGTGAATGTT-3� 129 bp from lpfAp This studyPLPFA129R 5�-CCAATTTCAAAGCTTAAAAAAATC-3�PLPFA197F 5�-CCTAAAAATAAACCCGGGCTTTTT-3� 138 bp from lpfAp This studyPLPFA60R 5�-ACTTTGAAGGAAGCTTTATACAAT-3�PLPFA147F 5�-AAATGTAATCCCGGGATTTTTTTA-3� 113 bp from lpfAp This studyPLPFA91R 5�-CGACAACTTTGAAAGAAGGA-3�PLPFA120F 5�-TAATATATTACCCGGGTAATTTTA-3� 129 bp from lpfAp This studyPLPFA9R 5�-CTTTTTCATGAAAAGCTTCATTTA-3�eLPFA25F 5�-CCGCACTGGCGTTAACTTCTGG-3� 141, 257 and 520 bp from lpfA gene This studyeLPFA166R 5�-GTACCTGACCCAACACAACTTC-3� 141 bp from lpfA gene This studyeLPFA282R 5�-AACACCACTAAAGCTCACATTG-3� 257 bp from lpfA gene This studyeLPFA545R 5�-GCCATTTGTAAAACGGACGATT-3� 520 bp from lpfA gene This studyORF1-A 5�-CTGGCTGTAGCTTATGTTCCG-3� Regulatory region of ler 2LERRDRB 5�-GTGAGATAACGTTATGTCCG-3�16SKI2232R 5�-GCCCAGATGGGATTAGCTAAGT-3� 16S rRNA 2916SL12iORF 5� GGAAAGTTCTGTGGATGTCAAG 3�5RTLPFA 5�-ACCCTGGTCGCTCTTAACGG-3� qRT-PCR lpfA gene This study3RTLPFA 5�-GGTCGTTACCGTTGCCAGAG-3�5RTRRSB 5�-TGCAAGTCGAACGGTAACAG-3� qRT-PCR rrsB gene 163RTRRSB 5�-AGTTATCCCCCTCCATCAGG-3�

VOL. 193, 2011 lpf1 REGULATION BY H-NS AND Ler 1623


.org/D

ownloaded from

http://jb.asm.org/

in Luria-Bertani (LB) broth or Dulbecco’s modified Eagle’s medium (DMEM;Gibco/Invitrogen) at 37°C. When required, antibiotics at the following concen-trations were added to the media: ampicillin, 100 �g/ml; tetracycline, 12.5 �g/ml;kanamycin, 50 �g/ml; and chloramphenicol, 30 �g/ml.

Construction of pK3-HNS. A SalI and SacI fragment containing the hns genewas obtained from plasmid pT3-HNS (2) and subcloned into the pMPM-K3digested with the same restriction enzymes, generating plasmid pK3-HNS.

RNA isolation and cDNA synthesis. Bacterial cultures grown in LB broth werediluted in DMEM and grown at 37°C until they reached an optical density at 600nm (OD600) of approximately 1.2. Cultures were treated with RNAprotect re-agent (Qiagen, Valencia, CA) to stabilize the RNA. Bacteria were harvested bycentrifugation at 12,000 rpm for 15 min at 4°C and resuspended in RNeasy lysisbuffer (Qiagen). RNA was purified in RNeasy columns (Qiagen), DNase treated(Ambion, Austin, TX), quantified, and qualitatively analyzed on agarose gels. Atotal of 5 �g of total RNA was used for cDNA synthesis by employing theSuperScript first-strand synthesis system (Invitrogen, Carlsbad, CA), accordingto the manufacturer’s instructions. A negative control with no reverse transcrip-

tase was also included. The resulting cDNA was utilized for real-time quantita-tive reverse transcriptase PCR (qRT-PCR).

Phenotype complementation of mutant strains. To genetically complementthe isogenic mutants, AC425 (hns::tet) and CB49 (ler::kan) were transformedwith low-copy-number plasmids pK3-HNS and pT6Ler, respectively. For Lerexpression using pT6Ler, EHEC strains were grown in DMEM at pH 7.2 at anOD600 of up to 0.6, followed by the addition of L-(�)-arabinose to a finalconcentration of 0.1%. Cultures were incubated until they reached an OD600 of1.2 before RNA was isolated.

Real-time qRT-PCR. Real-time quantitative RT-PCR was performed by usingthe SsoFast EvaGreen supermix, the CFX96 system test (Bio-Rad, Hercules,CA), and the primers listed in Table 2. We used the rrsB gene to normalize thedata, and a value of 1 was used to standardize lpfA1 expression in the wild-typestrain. For each reaction, 1 �l of reverse-transcribed cDNA was subjected toPCR amplification in a 25-�l final volume containing 500 nM each primer and12.5 �l of 2� EvaGreen master mix. The following conditions were used foramplification: 1 cycle at 95°C for 30 s and then 40 cycles at 95°C for 5 s and 60°C

FIG. 1. Quantitative real-time RT-PCR analysis of lpfA expression. The strains were grown in DMEM at 37°C without (A) or with (B) additionof 0.1% L-(�)-arabinose. The fold variation of gene expression was obtained by the threshold cycle (��CT) method. Wild-type EHEC expressionis represented by a value of 1, and variations are related to this value. (A) The lpfA1 expression increased 3.9-fold (*, P � 0.05) in the hns mutantcompared to that of the wild type. In the AC452/hns� strain, the lpfA1 expression was restored (*, P � 0.01 compared to that of the hns mutant).(B) The expression of lpfA1 was reduced in the ler mutant compared to that of the wild type (*, P � 0.001) and the CB49/ler� strain restoredexpression of lpfA1, which is significantly different from those of the ler mutant (*, P � 0.05).

1624 ROJAS-LOPEZ ET AL. J. BACTERIOL.


.org/D

ownloaded from

http://jb.asm.org/

FIG. 2. EMSAs of Ler and H-NS binding to lpfA1p. The EMSAs were performed with 50 ng of each PCR fragment and 150 nM (50 ng) ofpurified Ler-His6 or H-NS-His6. (A) Lane 1, free DNA containing the regulatory region of the ler gene (rLer); lane 2, Ler plus rLer; lane 3, freeDNA containing the 16S rRNA gene; lane 4, Ler plus the 16S rRNA gene; lane 5, free DNA F1 fragment (129 bp); lane 6, Ler plus F1; lane 7,free DNA F2 fragment (138 bp) from lpfA1p271; lane 8, Ler plus F2; lane 9, DNA F4 fragment (129 bp) from lpfA1p271; and lane 10, Ler plus F4.(B) Lane 1, free DNA containing rLer; lane 2, H-NS plus rLer; lane 3, free DNA containing the 16S rRNA gene; lane 4, H-NS plus the 16S rRNAgene; lane 5, free DNA F1 fragment (129 bp); lane 6, H-NS plus F1; lane 7, free F2 DNA fragment (138 bp) from lpfA1p271; lane 8, H-NS plusF2; lane 9, free F4 DNA fragment (129 bp) from lpfA1p271; and lane 10, H-NS plus F4. (C) Lane 1, free DNA F3 fragment (113 bp); lane 2, H-NSplus F3; lane 3, Ler plus F3. �, free DNA; �, DNA plus protein.

1625


.org/D

ownloaded from

http://jb.asm.org/

for 30 min. To ensure the specificity of the PCR products, melting curve analysiswas performed by heating products from 65°C to 95°C in increments of 0.5°Cevery 5 s while monitoring the fluorescence. These assays were performed at leastin triplicate for each strain to obtain statistical data.

DNA manipulations. Standard methods were used to perform plasmid purifi-cation, PCR, and gel electrophoresis (25). All oligonucleotides used are listed inTable 2. Each PCR fragment was purified and concentrated using Microconcolumns (Millipore), and the DNA concentration was calculated by observingthe absorbance of samples at 260 nm.

Expression and purification of His-tagged H-NS and Ler proteins. The H-NS–His6 or Ler-His6 protein was expressed and purified as described previously (29).Briefly, E. coli BL21/pLys21 harboring the pT6HNS or pT6Ler plasmid wasgrown to mid-logarithmic phase at 37°C. L-(�)-Arabinose (Sigma-Aldrich, St.Louis, MO) was added to a final concentration of 0.1%, and the bacteria werefurther incubated for 4 h at 30°C. Cells were pelleted by centrifugation at 8,000rpm for 10 min at 4°C, resuspended in 8 M urea buffer (pH 8.0), and disruptedby sonication. The cleared supernatant was applied to a HiTrap Ni2�-chelatingcolumn (Pro-Bond; Invitrogen). Proteins were eluted with a pH gradient (frompH 8.0 to 4.5) of 8 M urea buffer, and fractions containing purified H-NS–His6

or Ler-His6 were selected based on SDS-PAGE separation and Coomassiestaining. The fractions were loaded onto membrane dialysis cassettes (Slide-A-Lyzer cassette with a molecular mass cutoff of 6,000 to 8,000 Da; Fisher, Pitts-burgh, PA) and dialyzed at 4°C in buffer containing decreasing amounts of urea(4 M, 1 M, 0.2 M, and no urea). Aliquots of the purified proteins were stored at�80°C, and the protein concentration was determined using the Bradford pro-tein assay.

EMSAs. Approximately 50 ng of PCR-generated DNA fragments correspond-ing to the different segments of the lpf1 promoter region and lpfA1 gene (frag-ments F1, F2, F3, F4, F5e, F6e, and F7e) was mixed with 150 nM (50 ng) orincreasing concentrations of purified Ler-His6 or H-NS–His6 protein (solubilizedin a buffer containing 0.4 M HEPES, 0.08 M MgCl2, 0.5 M KCl, 0.01 M dithio-threitol [DTT], 0.5% NP-40, and 1 mg/ml bovine serum albumin [BSA]). Thereaction mixtures were incubated for 20 min at room temperature and thenseparated by electrophoresis in a 6% polyacrylamide gel with Tris-borate-EDTA

buffer at room temperature. The DNA bands were stained with ethidium bro-mide and visualized in a UV transilluminator. A fragment containing the lerpromoter region from the LEE was used as a positive control, and 16S rRNA wasused as a negative control.

DNase I footprinting assay. Oligonucleotides PLPFA258F (32P labeled) andPLPFA129R and oligonucleotides 120F (32P labeled) and 9R were used toamplify by PCR the lpf1 F1 and F4 regions, respectively. Labeled DNA frag-ments were incubated with increasing concentrations (150, 750, 1,000, and 1,500nM) of purified Ler-His6 and H-NS–His6 in binding buffer (40 mM HEPES [pH7.9], 50 mM KCl, 8 mM MgCl2, 1 mM dithiothreitol [DTT], and 1 mg/ml BSA)and 1 �g of poly(dI-dC) and incubated at room temperature for 20 min. Thereaction mixtures were then treated with 0.1 U of DNase I. Samples wereresolved by electrophoresis in 8% polyacrylamide-8 M urea gels along with thecorresponding sequencing ladder, which was generated with primersPLPFA258F and PLPFA120F, using pPLPFA as the template and the Sequenaseversion 2.0 DNA sequencing kit (USB, Cleveland, OH), according to the man-ufacturer’s instructions.

RESULTS

H-NS and Ler are key regulators controlling lpf1 transcrip-tion. We have previously shown, using transcriptional genefusions, that Ler and H-NS regulate the expression of the lpf1operon (29). To fully demonstrate that the regulatory effect ofboth H-NS and Ler occurs at the chromosomally carried lpf1operon, real-time RT-PCR experiments were performed usingtotal RNA obtained from the EHEC wild type (wt) and itsisogenic ler and hns mutants grown under LEE-inducing con-ditions. Transcription of the lpfA1 gene was measured andnormalized against a reference gene (e.g., rrsB), as described inMaterials and Methods. We observed that expression of the

FIG. 3. H-NS exhibited affinity for the region between �248 and �520 bp in the structural region of the lpfA1 gene. A total of 50 ng of threedifferent-sized PCR products from the structural region of lpfA1 gene were combined with increasing amounts of H-NS. Lane 1, free DNA F5efragment (141 bp); lane 2, 37 nM (12.5 ng) of H-NS plus F5e; lane 3, 75 nM (25 ng) of H-NS plus F5e; lane 4, 150 nM (50 ng) of H-NS plus F5e;lane 5, free DNA F6e fragment (257 bp); lane 6, 37 nM H-NS plus F6e; lane 7, 75 nM H-NS plus F6e; lane 8, 150 nM H-NS plus F6e; lane 9, freeDNA F7e fragment (520 bp); lane 10, 37 nM H-NS plus F7e; lane 11, 75 nM ng of H-NS plus F7e; and lane 12, 150 nM H-NS plus F7e. �, freeDNA; �, DNA with protein.



.org/D

ownloaded from

http://jb.asm.org/

lpfA1 gene in the hns mutant strain (AC425) increased by3.9-fold with respect to that of the wt strain (P � 0.03). Com-plementation of the AC425 mutant with plasmid pK3-HNSrestored H-NS repression of lpfA to wild-type levels (Fig. 1A)(P � 0.03). In contrast, expression of the lpfA1 gene in the lermutant (strain CB49) resulted in approximately a 3-fold reduc-tion with respect to that of the wild-type strain (Fig. 1B) (P �0.001). Complementation of the EHEC ler mutant with plas-mid pT6Ler restored lpfA1 expression to wild-type levels (Fig.1B). Our data confirmed the role of H-NS and Ler as regula-tors of the chromosomally carried lpf1 operon.

Ler and H-NS bind to defined DNA regions over the lpf1promoter (lpfA1p). We have previously reported that the DNAsequence required for H-NS and Ler binding to the regulatoryregion of the lpf1 operon (29) consists of 271 bp, locatedbetween positions �262 and �9, with respect to the startcodon of the lpfA gene (lpfA1p271 includes 262 bp of lpfA1p and9 bp within the structural region of the lpfA1 gene). ThisAT-rich region (79%) contains multiple AT runs of up to 19bp, a characteristic associated to H-NS DNA binding. To furthermap the regions that were required for H-NS- and Ler-specificbinding, four fragments spanning the lpfA1p271 sequence andthree fragments located within the lpfA1 gene were used forelectrophoretic mobility shift assays (EMSAs).

The EMSAs with different DNA fragments and purified Lerand H-NS proteins were performed using increasing concen-trations from 0.04 to 150 nM (12.5 to 50 ng) of each proteinand a fixed amount of lpfA1p fragments (150 nM [50 ng]) andamplified from the promoter region, as follows: F1 (129 bp),F2 (138 bp), and F4 (129 bp). We determined that the optimalprotein concentration to achieve a shift in the DNA-proteincomplex was 150 nM (50 ng) (data not shown). Ler binds to thethree promoter fragments, and more striking, Ler appeared toshift completely the free DNA of F2 and F4 fragments (Fig.2A, lanes 8 and 10). Our results indicate that Ler preferentiallybinds to a region extending from �197 bp upstream of theATG start codon to �9 bp, including the two lpf promoters,which we have previously identified (29). As a positive control,we used a 460-bp fragment amplified from the regulatory re-gion of the EHEC ler gene (previously shown to bind Ler), andas a negative control, we used a 780-bp fragment from EHEC16S rRNA (29).

Similarly, EMSAs were also performed with 150 nM pu-rified H-NS and the F1, F2, and F4 PCR fragments. Theresults indicate that H-NS binds to the 3 fragments from thelpfA1p sequence and shifts the entire free DNA from the F2and F4 fragments, confirming that the promoter region is a

FIG. 4. Scheme of localization of putative SRSs and LBSs in lpfA1p271 and lpfA1. EMSAs and footprinting assays were used to define the DNAregions for Ler and H-NS binding. (A) Representation of the DNA fragments F1, F2, F3, and F4 in lpfA1p and F5e, F6e, and F7e within the lpfA1gene. The name and size of each fragment are indicated at the left. At the end of each fragment, the position with respect to the lpfA1 ATG startcodon is indicated. P1 and P2 represent the lpfA1 transcriptional promoters and are depicted in full-length lpfAp271 and in the F2, F3, and F4fragments. At the right of each fragment, we indicated whether the proteins were binding as follows: by complete shift of DNA in EMSAs (�),by no shift of the DNA-protein complex (�), by partial shift of the DNA-protein complex in EMSAs (�/�). (B) We identified three Ler bindingsites upstream of the ATG, LBS1 (�220 to �193 bp), LBS2 (�177 to �170 bp), and LBS3 (�94 bp to �34 bp). Two SRSs were also identifiedupstream of the ATG start codon, SRS1 (�220 to �202 bp) and SRS2 (�105 to �61 bp). A third putative H-NS binding site was identifieddownstream of the ATG start codon (�283 to �545 bp), within the structural sequence of the lpfA1 gene.



.org/D

ownloaded from

http://jb.asm.org/

critical DNA target sequence for binding of this protein(Fig. 2B).

To further delineate the specific binding regions for theseproteins, we amplified by PCR the F3 fragment (113 bp), whichis identical to the F2 sequence, except for a 30-bp truncation atthe 5� end (Fig. 4A). When we examined the F3 fragment inthe EMSA, Ler was no longer able to bind to this fragment,suggesting that a putative Ler binding site is located at the first30 bp of F2 (Fig. 2C, lane 3). In contrast, H-NS retains theability to bind this fragment (Fig. 2C, lane 2).

H-NS, but not Ler, is binding to a region within the lpfA1 gene.The in silico analysis of the lpfA1 sequence using MegAlign(DNASTAR, Madison, WI) and the Cn3D program (NCBI,Bethesda, MD) revealed that putative areas of curvature existin some areas of the lpfA1 gene sequence due to the presenceof high AT content segments (data not shown). Therefore,three DNA fragments (F5e, F6e, and F7e) were PCR amplified

and used for EMSA analysis. We combined each DNA frag-ment with increasing amounts of the purified proteins, 37, 75,and 150 nM (12.5, 25, and 50 ng, respectively), and found thatH-NS binds to the longest fragment (F7e) at a concentration of25 ng, because the fragment changes mobility as the proteinconcentration increases (Fig. 3, lanes 11 and 12). The F7efragment extends from �24 bp to �545 in the lpfA1 gene (Fig.4A). In contrast, we did not observe any binding of the purifiedLer protein to any of the three different DNA fragments lo-cated within the lpfA1 gene, which suggests to us that theregulatory effect exerted by this protein occurs only at thepromoter region (data not shown).

Based on the information collected with the EMSA experi-ments, we began delimiting two putative Ler binding sites(LBSs). The first LBS spans from �197 bp to �167 bp up-stream of the ATG start codon. The Ler protein also bound tothe second LBS, which was mapped from �54 bp upstream of

FIG. 5. DNase I footprinting of Ler and H-NS on lpfA1p. (A) The 32P-labeled fragments (F1 and F4) of the lpfA1 promoter were incubatedin the absence or presence of increasing concentrations of Ler-His6 protein (0, 150, 750, 1,000, and 1,500 nM). (B) Footprinting analysis withH-NS–His6 was performed in the same manner as with the Ler protein and using the same protein concentrations. Lanes G, A, T, and C are thesequence ladders, and the protected areas are depicted as black boxes. To the right of both of the footprints, the DNA sequence of the lpfAp isdepicted, and the protected DNA areas are represented by light gray (Ler-protected) and dark gray (H-NS-protected) bars. The �10 and �35boxes, ribosome-binding sites (RBS), and transcriptional promoters are also illustrated.



.org/D

ownloaded from

http://jb.asm.org/

the start codon to �9 bp downstream within the lpfA1 gene, aregion which was located in the F4 fragment. Similarly, we alsoidentified the putative silencer regulatory sequence (SRS) inthe promoter region. The SRS contains an H-NS binding siteand is located between �197 bp upstream of the ATG to �9bp downstream within the lpfA1 gene. Because the H-NS pro-tein bound to F7e, but not to the F5e or F6e fragments, anH-NS binding site, outside the regulatory region, was delimitedfrom �283 to �545 bp in the lpfA1 gene (Fig. 4B).

The H-NS and Ler binding sites overlap on the lpf1 pro-moter region. Our previous published data (29) showed thatthe promoter region is sufficient to control the expression oflpfA1 by Ler and H-NS, and the minimal region to interact withthose proteins is approximately 271 bp. The EMSA resultsindicated that the LBS and SRS existed over the promoterregion. To further characterize the LBSs and SRSs within thepromoter region, DNase footprinting assays were performedas described in Material and Methods. DNA fragments F1 andF4, which cover the lpfAp271 (Fig. 4A), and increasing concen-trations of Ler (0.15, 0.75, 1.0, and 1.5 �M) were used, and theDNA sequences protected by Ler-His6 were identified. TheLBSs were remapped, and flanking regions were defined as

follows: Ler binding site 1 (LBS1) is located between �220 and�193 bp, LBS2 from �177 to �170 bp, and LBS3 from �94 to�34 bp, upstream from the ATG codon. These LBSs delimitedthe specific interactions sites of Ler and corroborated the spec-ificity of binding of this protein on lpfA1p (Fig. 4B and 5A). Asecond set of DNase footprinting studies using different con-centrations of H-NS–His6 (0.15, 0.75, 1.0, and 1.5 �M) wereperformed, and we delimited the following SRS sites onlpfA1p: SRS1 from �220 to �202 bp and SRS2 between �105and �61 bp (Fig. 4B and 5B).

DISCUSSION

In the current study, we have shown that H-NS and Ler bindto define segments within the promoter region of lpfA1(lpfA1p) and, apparently, within the lpfA1 gene to silence/antisilence the expression of lpf1. In our previous studies, wefound that these two regulatory proteins played an importantrole in the transcriptional control of lpf1 (29), and now, wefully demonstrated that Ler and H-NS regulate the expressionof the chromosomally carried lpf1 operon. Recent data fromour laboratory have suggested that in the hns ler double mu-

FIG. 5—Continued.



.org/D

ownloaded from

http://jb.asm.org/

tant, an additional factor (most likely a positive regulator) isneeded for lpf1 control, because lpfA1 expression increasedsignificantly and the fold difference is more than the one ob-served in the hns or ler single mutants (our unpublished re-sults). Interestingly, it has been reported that SlyA, a memberof the MarR family of transcriptional regulators, works inSalmonella as an antirepressor only when H-NS is present (22),but this does not seems to be the case in lpf1, because when wecompared the hns mutant to the wild-type strain, maximal lpf1expression is observed in the mutant strain. Further, the pres-ence of another antirepressor protein counteracting H-NSfunction is unlikely, because in the absence of Ler, lpfA1 ex-

pression decreased significantly. Therefore, we are currentlyexploring the phenotype observed in the hns ler mutant toidentify the additional regulatory factor controlling lpf1 expres-sion in EHEC.

Analysis of the minimal regulatory sequence required forlpf1 expression demonstrated characteristics found in thoseDNA sequences targeted by H-NS, for example, a high contentof AT (79%) and evident DNA curvature. In bacteria such asEPEC and other enteric pathogens, H-NS is implicated inregulating the expression of horizontally acquired virulencegenes with low percentages of GC (reviewed in references 5and 27), and its main function is to participate as a negative

FIG. 6. Model of lpf1 expression mediated by the regulators H-NS and Ler. The silencing mechanism mediated by H-NS occurs by binding onthe SRS1 and SRS2 sites, limiting the access of the RNA polymerase and preventing transcription initiation (steps I, II, and III). The antisilencingmechanism mediated by Ler counteracts the effect of H-NS and might be occurring in two ways, either by modifying the local DNA topology uponbinding to specific sites and/or by competition with H-NS for target specific sequences and/or the promoters (step IV). These mechanisms are notmutually exclusive.



.org/D

ownloaded from

http://jb.asm.org/

regulator, repressing transcription in response to changes inenvironmental signals, such as temperature and osmolarity (9,32). For example, H-NS represses transcription of the eltABoperon, encoding the heat-labile enterotoxin (LT) of entero-toxigenic E. coli (ETEC) by binding to two silencing regionslocated downstream of the eltAB promoter. Further in silicoanalysis also suggested that these sequences exhibit a highdegree of DNA curvature (33).

In EPEC, the divergently transcribed operons LEE2 andLEE3 have been shown to be regulated by H-NS/DNA inter-actions (2, 26). It was shown that the promoter region pos-sesses two silencer regulatory sequences (SRS1 and SRS2),and they are used for H-NS binding and repression of geneexpression. It was also demonstrated that Ler can bind to anddestabilize the H-NS/DNA complex, competing for the DNAand leading to an increase in expression from these operons. InEPEC and EHEC, H-NS can also regulate the expression ofLEE5 and the escD-LEE4 operons (12, 20), the bundle-form-ing pilus (bfp) genes and the perABC locus (only in EPEC)(24), and the stcE gene (only in EHEC) (8), and all of thesegenes are considered horizontally acquired genetic elements.However, the mechanism of regulation by H-NS in all of thesemodels varies because this protein binds to minimal regions of10 bp located downstream, upstream, or within the promoterregions (1, 19). It is believed that these sequences act in cis,interacting with H-NS and forming a nucleoprotein complex inthe promoter region (6, 7, 10).

Our efforts to map lpfA1p271 by footprinting analysis re-vealed that the H-NS protein has two specific silencing regu-latory sequences in the promoter region, SRS1 and SRS2 (18and 45 bp in length, respectively). SRS1 and SRS2 are 78% ATrich, and SRS2 overlaps the promoter at �83 bp from ATG.According to our previous data on lacZ gene reporter tran-scriptional fusions, we established that the promoter region issufficient for the regulation by Ler and H-NS (29). We havetested a transcriptional fusion, lacking SRS1, and observedthat its expression is increased by 3-fold in the hns mutantcompared to that of the wild-type strain, which suggests thatSRS2 is still a functional sequence target for regulation byH-NS (our unpublished data). Interestingly, the presence ofSRSs on lpfA1p suggests that the presumptive mechanism ofregulation used is similar to those proposed in Vibrio choleraeToxT and Salmonella enterica OmpS1 models, where H-NSbinds to the promoter region, forming a transcriptional nucleo-repressor filament (4, 34).

The repression mediated by H-NS is released by the antag-onistic properties of antisilencers or positive transcription fac-tors that either are acquired by horizontal gene transfer or areendogenous to the bacterial cell. The Ler protein, a horizon-tally acquired antisilencer, counteracts the effect of H-NS onthe EPEC promoters of the grlRA, LEE2-LEE3, and LEE5operons and in the EHEC lpf1 regulatory region (1, 2, 12, 29).In the current study, we found three Ler binding sites onlpfA1p, LBS1 (23 bp in length, overlapping SRS1), LBS2 (7bp), and finally, LBS3 ( 60 bp, overlapping SRS2). In LBS1and LBS2 (separated from each other by 16 bp) and LBS3,Ler binds, and it is likely that it modifies the local DNAtopology and/or competes with H-NS to release lpfA1p, allow-ing the RNA polymerase to initiate transcription (Fig. 6). Ithas been reported that as part of the positive regulation of

individual genes and operons of the LEE pathogenicity island,the Ler protein increases the transcription of divergent LEEoperons possessing both overlapping and nonoverlapping pro-moter regions, as in the case of the LEE2-LEE3 divergentoperons and the grlRA operon, respectively (1, 26). Further-more, it has also been observed that Ler binds to AT-rich DNAregions, but no consensus sequence has been defined as theLer binding site in the grlRA, LEE2-3, and LEE5 operons.These data suggest that Ler recognizes structural DNA motifsand not specific nucleotide sequences (2, 12).

In support of our data, we are now proposing a model toexplain the regulatory role of H-NS and Ler on lpf1. Thesilencer H-NS protein binds the lpfA1p in two defined regions(SRS1 and SRS2), and this protein shares these binding siteswith Ler (Fig. 6). A third putative SRS site in the lpfA1 genehas been located (Fig. 4B); however, the contribution of thisregion to the regulatory control of the lpf1 needs further study.Our model of regulation indicates that the target sequencesfacilitate H-NS binding on SRS1 and SRS2 (Fig. 6, steps II andIII), hiding or limiting the access of the RNA polymerase tothe transcriptional start site. In our prior study (29), we dem-onstrated that increasing concentrations of Ler are able todisrupt the H-NS/DNA complex, counteracting the silencingeffect of H-NS. Binding of Ler to LBSs promotes changes inthe DNA topology and/or competes through the shared DNAsequences with H-NS (Fig. 6, step IV). This antisilencing Lereffect releases lpfA1p and permits transcription by the RNApolymerase. We are currently further improving this workingmodel and are actively defining whether a relationship existsbetween SRS1, SRS2, and the third putative H-NS binding sitelocated in the structural region of lpfA1 and whether theyinfluence the regulatory control of the lpf1 operon.

ACKNOWLEDGMENTS

We thank Mardelle Susman and Douglas Botkin for critical readingof the manuscript.

This work was supported in part by a John Sealy Memorial Endow-ment Fund Bridging Grant and the NIH AI079154-01A2 grant toA.G.T. The laboratory of Y.M.-L was supported in part by institutionalfunds from the VIEP-BUAP MALI-NAT09-I. The work performed inJ.L.P.’s laboratory was supported by grants from Consejo Nacional deCiencia y Tecnología (CONACyT) (60796) and Direccion General deAsuntos del Personal Academico (IN227410). Fellowships fromCONACYT, Mexico, and SURP, UTMB (funding was awarded byARRA NIH grant AI079154-01A2S1), supported the work of M.R.-L.

This work and its contents are solely the responsibility of the coau-thors and do not necessarily represent the official views of the NIH.

REFERENCES

1. Barba, J., et al. 2005. A positive regulatory loop controls expression of thelocus of enterocyte effacement-encoded regulators Ler and GrlA. J. Bacte-riol. 187:7918–7930.

2. Bustamante, V. H., F. J. Santana, E. Calva, and J. L. Puente. 2001. Tran-scriptional regulation of type III secretion genes in enteropathogenic Esch-erichia coli: Ler antagonizes H-NS-dependent repression. Mol. Microbiol.39:664–678.

3. Casadaban, M. J. 1976. Transposition and fusion of the lac genes to selectedpromoters in Escherichia coli using bacteriophage lambda and Mu. J. Mol.Biol. 104:541–555.

4. De la Cruz, M. A., et al. 2007. LeuO antagonizes H-NS and StpA-dependentrepression in Salmonella enterica ompS1. Mol. Microbiol. 66:727–743.

5. Dorman, C. J. 2007. H-NS, the genome sentinel. Nat. Rev. Microbiol. 5:157–161.

6. Dorman, C. J. 2004. H-NS: a universal regulator for a dynamic genome. Nat.Rev. Microbiol. 2:391–400.

7. Dorman, C. J., and P. Deighan. 2003. Regulation of gene expression byhistone-like proteins in bacteria. Curr. Opin. Genet. Dev. 123:179–184.



.org/D

ownloaded from

http://jb.asm.org/

8. Elliott, S. J., et al. 2000. The locus of enterocyte effacement (LEE)-encodedregulator controls expression of both LEE- and non-LEE-encoded virulencefactors in enteropathogenic and enterohemorrhagic Escherichia coli. Infect.Immun. 68:6115–6126.

9. Erol, I., et al. 2006. H-NS controls metabolism and stress tolerance inEscherichia coli O157:H7 that influence mouse passage. BMC Microbiol.6:72.

10. Fang, F. C., and S. Rimsky. 2008. New insights into transcriptional regulationby H-NS. Curr. Opin. Microbiol. 11:113–120.

11. Fitzhenry, R., et al. 2006. Long polar fimbriae and tissue tropism in Esche-richia coli O157:H7. Microbes Infect. 8:1741–1749.

12. Haack, K. R., C. L. Robinson, K. J. Miller, J. W. Fowlkes, and J. L. Mellies.2003. Interaction of Ler at the LEE5 (tir) operon of enteropathogenic Esch-erichia coli. Infect. Immun. 71:384–392.

13. Hayashi, T., et al. 2001. Complete genome sequence of enterohemorrhagicEscherichia coli O157:H7 and genomic comparison with a laboratory strainK-12. DNA Res. 8:11–22.

14. Jordan, D. M., et al. 2004. Long polar fimbriae contribute to colonization byEscherichia coli O157:H7 in vivo. Infect. Immun. 72:6168–6171.

15. Kaper, J. B., J. P. Nataro, and H. L. T. Mobley. 2004. Pathogenic Escherichiacoli. Nat. Rev. Microbiol. 2:123–140.

16. Leverton, L. Q., and J. B. Kaper. 2005. Temporal expression of enteropatho-genic Escherichia coli virulence genes in an in vitro model of infection. Infect.Immun. 73:1034–1043.

17. Mayer, M. P. 1995. A new set of useful cloning and expression vectorsderived from pBlueScript. Gene 163:41–46.

18. Mellies, J. L., A. M. Barron, and A. M. Carmona. 2007. Enteropathogenicand enterohemorrhagic Escherichia coli virulence gene regulation. Infect.Immun. 75:4199–4210.

19. Mellies, J. L., S. J. Elliott, V. Sperandio, M. S. Donnenberg, and J. B. Kaper.1999. The Per regulon of enteropathogenic Escherichia coli: identification ofa regulatory cascade and a novel transcriptional activator, the locus of en-terocyte effacement (LEE)-encoded regulator (Ler). Mol. Microbiol. 33:296–306.

20. Mellies, J. L., et al. 2008. Ler interdomain linker is essential for anti-silencingactivity in enteropathogenic Escherichia coli. Microbiology 154:3624–3638.

21. Nataro, J. P., and J. B. Kaper. 1998. Diarrheogenic Escherichia coli. Clin.Microbiol. Rev. 11:142–210.

22. Perez, C. J., T. Latiff, and E. A. Groisman. 2008. Overcoming H-NS-medi-ated transcriptional silencing of horizontally acquired genes by the PhoP andSlyA proteins in Salmonella enterica. J. Biol. Chem. 283:10773–10783.

23. Perna, N. T., et al. 2001. Genome sequence of enterohaemorrhagic Esche-richia coli O157:H7. Nature 409:529–532.

24. Porter, M. E., et al. 2004. Direct and indirect transcriptional activation ofvirulence genes by an AraC-like protein, PerA from enteropathogenic Esch-erichia coli. Mol. Microbiol. 54:1117–1133.

25. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, NY.

26. Sperandio, V., et al. 2000. Activation of enteropathogenic Escherichia coli(EPEC) LEE2 and LEE3 operons by Ler. Mol. Microbiol. 38:781–793.

27. Tendeng, C., and P. Bertin. 2003. H-NS in Gram-negative bacteria: a familyof multifaceted proteins. Trends Microbiol. 11:511–518.

28. Torres, A. G., et al. 2002. Identification and characterization of lpfABCC�DE,a fimbrial operon of enterohemorrhagic Escherichia coli O157:H7. Infect.Immun. 70:5416–5427.

29. Torres, A. G., et al. 2007. Ler and H-NS, regulators controlling expression ofthe long polar fimbriae of Escherichia coli O157:H7. J. Bacteriol. 189:5916–5928.

30. Torres, A. G., et al. 2007. Environmental regulation and colonization attri-butes of the long polar fimbriae of Escherichia coli O157:H7. Int. J. Med.Microbiol. 297:177–185.

31. Torres, A. G., X. Zhou, and J. B. Kaper. 2005. Adherence of diarrheagenicEscherichia coli strains to epithelial cells. Infect. Immun. 73:18–29.

32. White-Ziegler, C., A. Villapakkam, K. Ronaszeki, and S. Young. 2000. H-NScontrols pap and daa fimbrial transcription in Escherichia coli in response tomultiple environmental cues. J. Bacteriol. 182:6391–6400.

33. Yang, J., M. S. R. Tauschek, and R. M. Robins-Browne. 2005. The H-NSprotein represses transcription of the eltAB operon, which encodes heat-labile enterotoxin in enterotoxigenic Escherichia coli, by binding to regionsdownstream of the promoter. Microbiology 151:1199–1208.

34. Yu, R. R., and V. J. DiRita. 2002. Regulation of gene expression in Vibriocholerae by ToxT involves both antirepression and RNA polymerase stimu-lation. Mol. Microbiol. 43:119–134.



.org/D

ownloaded from

http://jb.asm.org/

regulatory control of the escherichia coli o157:h7 lpf1 by...

Documents