JOURNAL OF BACTERIOLOGY, July 2008, p. 4736–4748 Vol. 190, No.

13
0021-9193/08/$08.00⫹0 doi:10.1128/JB.00089-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Integration Host Factor Positively Regulates Virulence Gene

Expression in Vibrio cholerae䌤
Emily Stonehouse, Gabriela Kovacikova, Ronald K. Taylor, and Karen Skorupski*
Department of Microbiology and Immunology, Dartmouth Medical School, Hanover, New Hampshire 03755
Received 17 January 2008/Accepted 25 April 2008

Virulence gene expression in Vibrio cholerae is dependent upon a complex transcriptional cascade that is
influenced by both specific and global regulators in response to environmental stimuli. Here, we report that the
global regulator integration host factor (IHF) positively affects virulence gene expression in V. cholerae.
Inactivation of ihfA and ihfB, the genes encoding the IHF subunits, decreased the expression levels of the two
main virulence factors tcpA and ctx and prevented toxin-coregulated pilus and cholera toxin production. IHF
was found to directly bind to and bend the tcpA promoter region at an IHF consensus site centered at position
ⴚ162 by using gel mobility shift assays and DNase I footprinting experiments. Deletion or mutation of the tcpA
IHF consensus site resulted in the loss of IHF binding and additionally disrupted the binding of the repressor
H-NS. DNase I footprinting revealed that H-NS protection overlaps with both the IHF and the ToxT binding
sites at the tcpA promoter. In addition, disruption of ihfA in an hns or toxT mutant background had no effect
on tcpA expression. These results suggest that IHF may function at the tcpA promoter to alleviate H-NS
repression.

The ability of pathogenic bacteria to modulate the expres- H-NS dimers bind to nucleating high-affinity sites, correspond-
sion of genes in response to environmental stimuli is wide- ing with the identified 10-bp consensus motif TCGATAAATT,
spread and critical for bacterial survival within specific niches. located within AT-rich regions, and polymerize outward along
Nucleoid-associated proteins, such as H-NS, integration host the DNA, eventually resulting in the condensation of DNA
factor (IHF), Fis, and HU play important roles in this adaptive through the formation of protein bridges between H-NS bind-
process. These proteins comprise a group of abundant DNA- ing sites (4, 29). This DNA remodeling can directly interfere
binding proteins that condense the bacterial chromosome, with RNA polymerase action or block the binding of specific
modulating its structure and influencing many cellular pro- activators at a nearby promoter region.
cesses, such as DNA replication (51), site-specific recombina- IHF is a heterodimeric protein composed of the subunits
tion (5, 16), and transcription (35). The nucleoprotein com- IHF␣ (11.3 kDa) and IHF␤ (10.6 kDa), encoded by the ihfA
plexes formed in response to environmental cues have both (himA) and ihfB (hipB) genes, respectively. The two subunits
negative and positive influences on transcription, and these have approximately 35% identity to each other. Unlike H-NS,
proteins function with one another and with gene-specific reg- IHF recognizes a specific asymmetric site on DNA, the 13-bp
ulatory factors to fine tune gene expression. consensus target sequence WATCAANNNNTTR, where W
The nucleoid-associated protein H-NS is an abundant global represents A or T and R represents A or G (10). Upon binding,
regulator that influences various cellular processes (13, 14, 41). IHF is known to bend DNA by as much as 180° (48, 61). In E.
H-NS binds DNA with relatively low sequence specificity but coli and S. enterica serovar Typhimurium, IHF is involved in
instead has a preference for AT-rich segments with intrinsic controlling the transcriptions of over 100 genes of various
planar curvature (49, 68). Studies using chromatin immuno- functions (2, 33, 35). In addition, IHF has increasingly been
precipitation-on-chip technology have identified the H-NS identified as a regulator of virulence gene expression in various
binding regions over the Escherichia coli and Salmonella en- pathogens. IHF is required for appropriate expression of the
terica serovar Typhimurium genomes (18, 32, 42, 44). Their Brucella abortus virB operon, encoding the type IV secretion
results demonstrate that H-NS binds to regions of high AT system (52). In Shigella flexneri, IHF positively regulates viru-
content, including genes acquired by horizontal gene transfer. lence gene expression at multiple promoters, including the
Of interest was the finding that H-NS interacts with extensive virF, virB, and icsA promoters (47). Flagellum production in
DNA stretches on pathogenicity islands and represses the tran- enteropathogenic and enterohemorrhagic E. coli strains is re-
scription of the associated virulence genes. This discovery has pressed by IHF (69). IHF can act as a transcriptional regulator
raised speculation that H-NS may function to silence foreign by different mechanisms. In several cases, IHF produces a
DNA until dedicated transcription factors are able to relieve DNA bend that enables the direct interaction of a distant,
H-NS repression (14). Recent studies have proposed that DNA-bound transcriptional regulator with RNA polymerase
to enhance transcription (15). DNA bending by IHF can stim-
ulate the recruitment of RNA polymerase to activate transcrip-
* Corresponding author. Mailing address: Department of Microbi- tion by favoring the interaction of the ␣ subunit C-terminal
ology and Immunology, HB7550, Dartmouth Medical School, Hano-
ver, NH 03755. Phone: (603) 650-1634. Fax: (603) 650-1318. E-mail:
domain with upstream DNA UP elements (3). IHF binding can
[email protected]. also serve to lower the activation energy for open complex
䌤
Published ahead of print on 2 May 2008. formation (45). IHF typically acts as an architectural factor

4736
VOL. 190, 2008 IHF REGULATION OF tcpA EXPRESSION 4737

that leads to the correct promoter structure necessary for tran- MATERIALS AND METHODS
scriptional regulation. An alternative mechanism of IHF acti- Bacterial strains and growth. The V. cholerae and E. coli strains and plasmids
vation is antirepression of H-NS-repressed promoters (63). used in this study are listed in Table 1. Strains were maintained at ⫺70°C in
Virulence gene expression in the human enteric pathogen Luria-Bertani (LB) medium (36) containing 30% (vol/vol) glycerol. Antibiotics
were used at the following concentrations in LB medium: ampicillin (Amp), 100
Vibrio cholerae is dependent upon a complex transcriptional
␮g/ml; gentamicin, 30 ␮g/ml; kanamycin (Kan), 45 ␮g/ml; polymyxin B, 50 IU/ml;
cascade that is influenced by both specific regulators, such as and streptomycin, 1 mg/ml. 5-Bromo-4-chloro-3-indolyl-␤-D-galactopyranoside
ToxR/S, TcpP/H, and ToxT, and global regulators, such as (X-Gal) was used in LB agar at 40 ␮g/ml (Sigma).
cyclic AMP-cyclic AMP receptor protein and H-NS, in re- Construction of chromosomal ihf deletions. The ihfA subunit deletion was
sponse to environmental stimuli (for reviews, see references 9 constructed in pKAS32 (53) by PCR using primer HMEco (oligonucleotide
sequences are listed in Table 2) with HMNot1 and HMNot2 with HMXba. A
and 55). Production of the two main virulence factors, the Kanr fragment was inserted between the two fragments at the NotI site, and the
toxin-coregulated pilus (TCP) (60) and the cholera toxin (CT) resulting plasmid, pGKK158, was used for allelic exchange. The ihfB subunit
(23), is required for human colonization and causes the severe deletion construct was created in pKAS46 (53) by PCR using primer HBEco with
watery diarrhea that is characteristic of the disease. Both of HBNotI and primer HBNot2 with HBXba. A Genr fragment was inserted be-
these virulence factors are carried on genetic elements ac- tween the two fragments at the NotI site, and the resulting plasmid, pGKK162,
was used for allelic exchange. Deletions were confirmed by PCR.
quired by horizontal gene transfer that have a higher AT con- Expression plasmids. Plasmid pKAS161 was constructed by PCR amplifying a
tent (⬎60% AT rich) than the ancestral chromosome (52% AT 1-kb fragment containing the ihfA subunit with primers HMEco and HMXba.
rich). The tcp genes are carried on the Vibrio pathogenicity The resulting product was directly cloned into pBAD-TOPO (Invitrogen) and
island (VPI), and the ctx genes are carried on the CT phage. sequenced.
The V. cholerae hns expression plasmid pEAS10 was constructed by PCR
The expression of the tcp and ctx genes is transcriptionally
amplifying the 414-bp hns gene from O395 chromosomal DNA by using HNS1
controlled in response to environmental stimuli by a regulatory and HNS2. The resulting fragment was digested with NdeI and SapI and ligated
cascade, termed the ToxR virulence regulon, that involves pro- into similarly cut pTXB-1 to generate pEAS10. The nucleotide sequence was
teins encoded within both the ancestral Vibrio genome and the confirmed by DNA sequencing.
VPI. The cascade is initiated by two chromosomally encoded Construction of IHF binding site mutations. The 21-bp deletion of the puta-
tive IHF consensus site in the tcpA promoter was constructed using primer pairs
proteins, AphA and AphB, functioning synergistically to acti-
H-Eco/Sap1B and ABgl2/Ear1T to amplify chromosomal sequences from O395.
vate transcription of tcpPH (25, 56). TcpP and TcpH (6, 20) are The fragments were digested with SapI and EarI to produce seamless junctions,
transmembrane proteins encoded on the VPI that, along with followed by BglII and EcoRI digestion and ligation into pKAS32 to generate
a second pair of transmembrane proteins, ToxR and ToxS (37, pEAS9. The resulting plasmid was confirmed by sequencing. To construct four
38), activate transcription at the toxT promoter in a coopera- base pair mutations (shown in Fig. 4C) in the most highly conserved residues of
the IHF consensus site, the primer pairs H-Eco/Sap2B and Ear3T/ABgl2 were
tive manner (21, 28). ToxT directly activates the expression of used to amplify O395 chromosomal DNA. After digestion as described above,
many genes, including the tcp operon (7, 12, 22, 71), the ctx the fragments were ligated into pKAS32, generating p4PM-tcpA, and sequenced.
operon (12), and the accessory colonization factor genes acfA These mutations were introduced into the tcpA-lacZ fusion strain (MBN135) or
and acfD (46, 65). In addition, ToxT autoregulates its own O395 by allelic exchange and subsequently confirmed by sequencing.
expression through activation of the tcpA promoter and ␤-Galactosidase assays. ␤-Galactosidase assays (36) were carried out after
overnight growth in LB medium with a starting pH of 6.5 at 30°C and shaking.
readthrough into toxT (70). Due to TCP-mediated bacterial autoagglutination, the specific activities of
In addition to the specific regulatory proteins of the ToxR strains in the ctx-lacZ background were calculated using the protein concentra-
virulence regulon mentioned above, the nucleoid-associated tion determined by a bicinchoninic acid procedure (Pierce) rather than using the
protein H-NS has been shown to play an important role in V. more standard optical density normalization method.
cholerae virulence gene expression (17, 27, 43). Mutation of V. CT assays. GM1 ganglioside enzyme-linked immunosorbent CT assays (23)
were carried out on the supernatants of overnight cultures grown in LB medium
cholerae hns causes derepression of toxT, tcpA, and ctx expres- with a starting pH of 6.5 at 30°C and shaking.
sion to near-wild-type levels even in the absence of cognate Immunoblot analysis. Whole-cell protein extracts prepared from the overnight
activator proteins (43). A variety of data indicate that H-NS cultures used for CT assays were subjected to sodium dodecyl sulfate-12.5%
directly influences transcription at each of these promoters. polyacrylamide gel electrophoresis, transferred to a nitrocellulose membrane,
probed with an anti-TcpA antibody (59) or an anti-ToxT polyclonal antibody,
Thus, H-NS functions at multiple promoters within the cascade
and visualized using an enhanced chemiluminescence detection system (Amer-
to reduce virulence gene expression under particular environ- sham).
mental conditions. Purification of the V. cholerae IHF heterodimer. Plasmids pEAS2 and pEAS3
In the present study, we show that, in contrast to H-NS, the were generated to overexpress the genes encoding the IHF␣ and IHF␤ subunits,
nucleoid protein IHF positively influences virulence gene ex- respectively. The ihfA gene was amplified from O395 chromosomal DNA by
using primers ihfA1 and ihfA2, digested with EcoRI and SphI, and ligated into
pression in V. cholerae. Deletion of either the ihfA or the ihfB
similarly cut pBAD22, creating pEAS2. The ihfB gene was amplified from O395
gene decreased chromosomally encoded ctx-lacZ and tcpA- chromosomal DNA with the primer pair ihfB1/ihfB2, digested with EcoRI and
lacZ promoter fusion expression and prevented TCP and CT SphI, and ligated into pEAS1, creating pEAS3. pEAS1 was generated by ligating
production. The same mutations had no effect on tcpP-lacZ or the 50-bp ribosomal binding site from pBAD22 (NheI/SphI-digested fragment)
toxT-lacZ fusions. Gel mobility shift assays and DNase I foot- into pKAS178 (a pBAD33 Kmr derivative) to replace the pBAD33 ribosomal
binding site. This ensured that the translational efficiencies of ihfB from pEAS3
printing experiments with purified V. cholerae IHF demon- and ihfA from pEAS2 were comparable. The nucleotide sequences of both
strate that IHF recognizes a putative IHF consensus site within constructs were confirmed by sequencing. V. cholerae O395 harboring both
the tcpA promoter and introduces a bend upon binding. DNase pEAS2 and pEAS3 was grown overnight in 10 ml of LB at 37°C with shaking and
I footprinting at tcpA revealed that H-NS binding covers the used to inoculate 2 liters of LB containing Amp/Kan. The culture was grown for
identified IHF consensus site as well as the ToxT site. In 3 h, and arabinose was then added to give a final concentration of 0.1%. Fol-
lowing overnight incubation at 37°C with shaking, the V. cholerae IHF het-
addition, we determined that the dependence on IHF for op- erodimer was extracted as soluble protein as previously described (40). Briefly,
timal tcpA expression requires the presence of both H-NS and the culture was centrifuged, sonicated, and recentrifuged and the supernatant
ToxT. was precipitated with 50% ammonium sulfate. After centrifugation, the resulting
4738 STONEHOUSE ET AL. J. BACTERIOL.

TABLE 1. Bacterial stains and plasmids used in this study

Strain or plasmid Characteristic(s) Reference or source

Strains
V. cholerae
O395 Sm Classical Ogawa; Smr 60
CG842 O395; ⌬lacZ 8
KSK218 CG842; ctx-lacZ; Smr Cmr 54
GK628 KSK218; ⌬ihfA::kan This work
GK646 KSK218; ⌬ihfB::gen This work
KSK618 CG842; tcpP-lacZ 56
GK630 KSK618; ⌬ihfA::kan This work
GK652 KSK618; ⌬ihfB::gen This work
MBN032 CG842; toxT-lacZ1 43
GK632 MBN032; ⌬ihfA::kan This work
GK663 MBN032; ⌬ihfB::gen This work
MBN135 CG842; tcpA-lacZ 43
GK648 MBN135; ⌬ihfA::kan This work
GK650 MBN135; ⌬ihfB::gen This work
ES23 MBN135; ⌬consensus at tcpA This work
ES24 MBN135; 4PM mutation in tcpA IHF consensus site This work
MBN142 MBN135; ⌬toxT 43
ES41 MBN135; ⌬toxT ⌬ihfA::kan This work
MBN148 MBN135; ⌬hns 43
ES42 MBN135; ⌬hns ⌬ihfA::kan This work
MBN168 MBN135; ⌬toxT ⌬hns 43
ES43 MBN135; ⌬toxT ⌬hns ⌬ihfA::kan This work
ES21 O395; ⌬ihfA::kan This work

E. coli ER2566 Expression strain; T7 RNA polymerase New England Biolabs

Plasmids
pKAS32 pGP704; rpsL; Apr 53
pGKK158 pKAS32; ⌬ihfA::kan; Apr Kmr This work
pKAS46 pKAS32; Kmr 53
pGKK162 pKAS46; ⌬ihfB::gen; Kmr Gmr This work
pBAD-TOPO Expression vector; Kmr Invitrogen
pKAS161 pBAD-TOPO; ihfA; Kmr This work
pKAS178 pBAD33; Kmr Cmr 24
pBAD22 Expression vector; Apr 19
pEAS1 pKAS178; RBS from pBAD22; Kmr This work
pEAS2 pBAD22; ihfA; Apr This work
pEAS3 pEAS1, ihfB; Kmr This work
pBlueScript Cloning vector; Apr Stratagene
pEAS4 pBlueScript; 232-bp tcpA promoter fragment This work
pEAS5 pBlueScript; 360-bp tcpA promoter fragment This work
pBEND5 Vector for circular permutation assay; Apr 72
pEAS7 pBEND5; 66-bp tcpA IHF consensus site This work
pEAS9 pKAS32; 21-bp deletion including tcpA IHF consensus site (⌬consensus) This work
pTXB-1 Expression vector for intein/chitin binding domain fusion; Apr New England Biolabs
pEAS10 pTXB-1; hns Apr This work
p4PM-tcpA pKAS32; tcpA IHF consensus 4-point mutations (4PM) This work
pRH81 pBAD22; His6 toxT 22

supernatant was precipitated with 70% ammonium sulfate and the pellet was strain ER2566, containing the hns expression plasmid pEAS10, was grown in LB
resuspended in 1 ml of TG (50 mM Tris [pH 7.4], 10% glycerol) containing 10 containing Amp at 37°C for 3 h, induced with 1 mM IPTG (isopropyl-␤-D-
mM KCl. The protein was then dialyzed overnight against TG containing 10 mM thiogalactopyranoside), and grown overnight at 16°C. The cells were collected by
KCl, loaded onto a phosphocellulose column, and eluted with a KCl salt gradient centrifugation and resuspended in column buffer (20 mM Tris [pH 8.0], 500 mM
from 10 mM to 1.2 M in TG. The IHF-containing fractions were pooled, dialyzed NaCl, 1 mM EDTA). The extract was sonicated and clarified by centrifugation at
against a 25 mM NaH2PO4 (pH 5.5), 1 mM EDTA, 5% glycerol buffer, frac- 14,000 rpm for 30 min in an SS34 rotor. The supernatant was then loaded onto
tionated over a cation exchange Macro-Prep High S support (Bio-Rad) at pH a chitin column equilibrated with column buffer to bind the fusion protein and
5.5, and developed with a linear gradient (75 to 800 mM) of NaCl in 25 mM washed with 10 volumes of column buffer and high-salt column buffer (containing
NaH2PO4 buffer. The fractions were analyzed on an 18% urea polyacrylamide 1 M NaCl). Three volumes of cleavage buffer (column buffer containing 100 mM
gel, and the IHF-containing fractions were pooled and further dialyzed against dithiothreitol [DTT]) were added, and the column was incubated overnight at
25 mM NaH2PO4, 1 mM EDTA, 5% glycerol. The estimated purity of the IHF 4°C. H-NS was then eluted off the column with column buffer. The eluted protein
heterodimer was 95% as determined by densitometry. was dialyzed overnight in a solution of 20 mM Tris (pH 8.0), 1 mM EDTA, 5 mM
Purification of V. cholerae H-NS. H-NS was purified using the IMPACT-CN potassium glutamate, and 0.1 mM DTT. Glycerol was added to the protein to
system, which uses a cleavable chitin binding domain fused to the C terminus of 10%, and the protein was stored at ⫺70°C. The final purity of H-NS was
a protein for purification utilizing a chitin column (New England Biolabs). E. coli estimated to be greater than 90% as determined by densitometry.
VOL. 190, 2008 IHF REGULATION OF tcpA EXPRESSION 4739

TABLE 2. Oligonucleotides used in this study sequencing. Permutated DNA fragments were generated from pEAS7 by diges-
tion with the enzymes listed in Fig. 5, and the resulting 190-bp fragments were
Name Sequencea
purified from polyacrylamide gels and labeled with DIG. A standard gel mobility
HMEco.........GATCGGAATTCCAGCGGTAGAGGCGATTGTG shift assay with purified IHF was performed on each fragment as described
HMNot1.......GATCGGCGGCCGCAAAACTTCCCTCAAAGCTAAG above.
HMNot2.......GATCGGCGGCCGCGACCGAGCAATAGACCACGC
HMXba ........GATCGTCTAGATTTAGCCCGAGCCACCATAG
HBEco..........GATCGGAATTCCGTGACCGTATCGAAGATGC
HBNot1 ........GATCGGCGGCCGCTAGTTTCCCTCTTCGAGTTG
RESULTS
HBNot2 ........GATCGGCGGCCGCTCTGAGTAAAGTTTGTTCAC
HBXba .........GATCGTCTAGATCTTGCTGGGCGCTACGATG IHF positively regulates expression of the tcpA and ctx pro-
H-Eco ...........GATCGGAATTCCACGTTATCAAACATTGCCAG moters. To assess the influence of IHF on the regulation of
Sap1B ...........GATCGGCTCTTCGCTTGTAATTTAATGCAAATAAAATGC
ABgl2............GATCGAGATCTGCTGTCGCATCAGCTGTTGC virulence gene expression in V. cholerae, ⌬ihfA::kan and
Ear1T............GATCGCTCTTCGAAGTTGGCCTTTTTTTAGTGTC ⌬ihfB::gen mutants were constructed in the ctx-lacZ fusion
Sap2B ...........GATCGGCTCTTCGACCTTTAAAGTAATTTAATGCAA
Ear3T............GATCGCTCTTCGGGTTTTGAAGTGAATAAGTTGGC strain KSK218 (54). Under conditions known to induce viru-
ihfA1.............GATCGGAATTCACCATGGCGCTCACAAAGGCCGAATTG lence gene expression, LB with a starting pH of 6.5 at 30°C,
ihfA2.............GATCGGCATGCGTCTTATTTTTCGACTTTGATG
ihfB1 .............GATCGGAATTCACCATGACTAAGTCCGAACTGATAG both ihf mutants showed fivefold reductions in ctx-lacZ expres-
ihfB2 .............GATCGGCATGCGATTACAAGTTAACGCGTTCG sion (Fig. 1A). Since tcpA expression is coordinate with that of
HNS1 ............GATCGCATATGGTAATGTCGGAAATCACTAAG
HNS2 ............GATCGGCTCTTCAGCACAGAGCGAATTCTTCCAGAGA
ctx, the influence of the ihf mutations on a tcpA-lacZ fusion
TCP-SAL .....GACTCGTCGACAATTTCGATCTCCACTCCGG strain, MBN135 (43), was assessed and expression was also
TCP-XHO....GTCAACTCGAGCATATTTATGTAACTCCACC reduced approximately fourfold (Fig. 1A). These results show
ctxP2 .............GATCGGAATTCTACCGGATTTTAATCACTTTG
ctxP1 .............GATCGGGATCCCATCAGGAGGTCTAGAATCTG that IHF plays a role in the positive activations of the ctx and
tcp1 ...............GACTCTCTAGAAGTGTCTGATTTATTTTGTG tcpA promoters. Expression of the ␣ subunit of IHF in pBAD-
tcp2 ...............GACTCGGATCCTTGCTGTGTTTTTTTATTTT
tcp7 ...............GATCGGAATTCATCCACGTAGGTGGGTATAG TOPO (pKAS161) in the ihfA mutant strain restored normal
tcp9 ...............GATCGGGATCCCATATTTATGTAACTCCAC ctx-lacZ and tcpA-lacZ expression, confirming that the de-
tcp12 .............GCATTTTATTTGCATTAAATTAC
tcp13 .............GACACTAAAAAAAGGCCAAC creased expression levels observed at these promoters were
a
due to the lack of functional IHF in the ⌬ihfA::kan strain
All primer sequences are shown 5⬘ to 3⬘. Underlines indicate sequences
encoding relevant restriction sites, while substitutions are in bold. (Fig. 1B).
Introduction of the ⌬ihfA::kan mutation into wild-type O395
confirmed that the production of TCP and CT in V. cholerae is
Gel mobility shift assays. DNA fragments containing the putative IHF binding dependent upon IHF. As observed in Fig. 2A, the ⌬ihfA::kan
sites in the promoter regions of the tcpA and ctx operons were generated by PCR mutant ES21 is considerably reduced for TcpA and CT pro-
from V. cholerae O395 chromosomal DNA or from the pEAS9 or p4PM-tcpA duction when grown under inducing conditions. In addition,
plasmid with the TCP-SAL (⫺226)/TCP-XHO (⫹78), TCP-SAL/tcp2 (⫺50),
tcp1 (⫺100)/TCP-XHO, or ctxP2 (⫺194)/ctxP1 (⫹118) oligonucleotide pair. IHF
both of these products were restored with ihfA overexpression
gel mobility shift assays were carried out in a volume of 20 ␮l containing from pEAS2 (Fig. 2A).
DNA-binding buffer (50 mM Tris [pH 7.4], 70 mM KCl, 1 mM EDTA, 1 mM The influence of IHF on virulence gene expression could be
DTT, 100 ␮g/ml bovine serum albumin, 5% glycerol), 10 ng of digoxigenin the result of a direct effect at the ctx and tcpA promoters, or the
(DIG)-labeled DNA, 1.5 ␮g of calf thymus DNA, and purified IHF het-
effect could be indirect due to an influence of IHF at promot-
erodimers. After incubation at 30°C for 15 min, protein-DNA complexes were
separated from free probes by electrophoresis on 5% nondenaturing polyacryl- ers further upstream in the virulence cascade, such as the
amide gels and transferred to nylon. H-NS gel mobility shift assays were carried tcpPH, toxR, or toxT promoter. To address the IHF effect on
out in a volume of 20 ␮l containing DNA-binding buffer (40 mM HEPES, 100 tcpPH, the ⌬ihfA::kan and ⌬ihfB::gen mutations were intro-
mM potassium glutamate, 10 mM magnesium aspartate, 0.022% NP-40, 100 duced into the tcpPH-lacZ strain KSK618 (56). The expression
␮g/ml bovine serum albumin, 10% glycerol), 10 ng of DIG-labeled DNA, 1 ␮g of
calf thymus DNA, and purified H-NS. After incubation at room temperature for
of tcpPH was not influenced by either ihf mutation (data not
20 min, protein-DNA complexes were separated by 5% nondenaturing polyac- shown). Levels of ToxR, a member of the protein pair that
rylamide gel electrophoresis and transferred to nylon. DNA was visualized with cooperates with TcpP/H to activate toxT expression, were also
an anti-DIG detection kit (Roche), followed by enhanced chemiluminescence unaffected by the ⌬ihfA::kan mutation, as detected by immu-
detection.
noblot analysis (data not shown). The expression of toxT is
DNase I footprinting. DNA fragments of 232 and 360 bp were generated from
the O395 tcpA promoter by amplification with the tcp7 (⫺282)/tcp2 (⫺50) or influenced by two promoters, one that is TcpP/ToxR depen-
tcp7/tcp9 (⫹78) primer pair. These fragments were ligated into pBlueScript dent, located immediately upstream of toxT, and one that is
(Stratagene), generating pEAS4 and pEAS5, respectively. Fragments for labeling ToxT dependent and initiates upstream of tcpA (see Fig. 8). To
were digested from pEAS4 and pEAS5 with EcoRI and BamHI. Strand labeling determine the effect of IHF on toxT expression from the TcpP/
was carried out essentially as described by Kovacikova and Skorupski (26).
Binding reactions were performed in a 6.5-␮l volume containing 33P-end-labeled
ToxR-dependent promoter, the ⌬ihfA::kan and ⌬ihfB::gen mu-
DNA, purified IHF or H-NS, and the corresponding binding buffer as described tations were introduced into the toxT-lacZ strain MBN032,
above. After incubation at 30°C for 15 min, protein-DNA complexes were di- which is lacking ToxT (43). Expression from this promoter was
gested with 1 ␮l of various dilutions of DNase I (Promega) for 5 min at 30°C. The not influenced by either ihf mutation (data not shown). To
reactions were stopped by addition of 2 mM EGTA, followed by heat inactiva-
address whether IHF influences toxT expression from the
tion at 65°C for 10 min and then spot dialysis for 30 min against 10 mM Tris (pH
8.0), 0.1 mM EDTA. The reaction mixtures were then heated to 90°C for 5 min ToxT-dependent promoter upstream of tcpA, a Western blot
in formamide loading buffer (Epicenter) and separated on a 6% acrylamide analysis was performed to assess whether total ToxT levels
sequencing gel with 1⫻ Tris-borate-EDTA. The gels were dried and visualized by were altered in the ihfA mutant. As shown in Fig. 2B, ToxT
autoradiography. levels were lower in the ihfA mutant than in the wild type upon
Circular permutation assay. The ability of IHF to bend DNA at its binding site
was tested as described by Wu and Crothers (66). A 67-bp fragment containing
growth to late log phase under inducing conditions. Addition-
the IHF binding site was amplified with primers tcp12 and tcp13 and cloned into ally, when total toxT transcript from nonfusion backgrounds
the HpaI site of pBEND5 (72), generating pEAS7. The insert was confirmed by was measured by quantitative real-time PCR from cultures
4740 STONEHOUSE ET AL. J. BACTERIOL.

FIG. 2. TcpA, CT, and ToxT production in O395 and the ihfA
mutant. Strains were grown overnight in LB medium with a starting pH
of 6.5 at 30°C with shaking. Western blot analysis with the anti-TcpA
(␣-TcpA) antibody (A) or ␣-ToxT polyclonal antibody (B) was per-
formed with 16 ␮g of total protein extract. CT represents the amount
of CT detected in the culture supernatant as measured by a GM1-
ganglioside enzyme-linked immunosorbent CT assay (A) (expressed in
ng/mg of protein). WT, wild type.

present within the tcpA promoter. One site overlaps the ⫺35
region (5⬘-CATCTGTCAATTG), while the second is located
FIG. 1. Influence of IHF on ctx-lacZ and tcpA-lacZ expression. upstream and centered at position ⫺162 (5⬘-AATCATTTGA
(A) Wild-type (WT) (tcpA-lacZ or ctx-lacZ), ihfA (⌬ihfA::kan), or ihfB ATT), about 80 bp upstream of the ToxT binding site (posi-
(⌬ihfB::gen) mutants. (B) WT, ihfA, or ihfA mutants with the empty vector tions ⫺84 to ⫺41). IHF was capable of binding to the promoter
pBAD-TOPO or the IHF␣ expression vector pIHFA (pKAS161). Cul- fragment containing the ⫺162 site alone (⫺226 to ⫺50) (A2),
tures were grown overnight in LB medium with a starting pH of 6.5 at
30°C with shaking (those with plasmids contained 0.1% arabinose) and whereas it was not capable of binding to the ⫺35 site alone
assayed for ␤-galactosidase production. The values are the averages for at (⫺100 to ⫹78) (A3) (Fig. 3). Despite the presence of only one
least two independent experiments. shifted complex on the A1 fragment containing both putative
sites, it is formally possible that IHF binds the ⫺35 site only in
the presence of the ⫺162 site. This possibility was examined in
grown overnight under inducing conditions, toxT transcript DNase I footprinting experiments where both sites were
levels were decreased similarly to tcpA transcript levels in the present on the same fragment (see below). Notably, the DNA
ihfA mutant compared to wild-type levels (data not shown). surrounding the ⫺162 site from position ⫺200 to ⫺125 has
The observation that IHF does not appear to influence expres- 76% AT content. The high AT content, coupled with the
sion from the TcpP/ToxR-dependent promoter and yet total knowledge that the nucleoid-associated protein H-NS directly
transcript levels (TcpP/ToxR-dependent and ToxT-dependent represses tcpA expression (71) and itself preferentially binds
toxT transcripts) were reduced in the ihfA mutant suggests that AT-rich DNA (18, 32, 42, 44), suggests that the region around
IHF might have a direct effect on the ToxT-dependent pro- position ⫺162 may also be an H-NS binding site. Scanning of
moter upstream of tcpA. The decreases in ctx expression and this region for putative H-NS nucleation sites by use of the
CT production in this mutant are most likely due to reduced 10-bp consensus site identified by Lang et al. (29) revealed two
toxT expression resulting from the influence of IHF at the tcpA matches adjacent to the IHF consensus site (Fig. 4C, boxes),
promoter. This is further supported by studies of IHF binding further suggesting that IHF and H-NS could both bind this
at these promoters (see below). region of the tcpA promoter.
Purified IHF from V. cholerae binds the tcpA promoter re- IHF and H-NS binding sites overlap the ⴚ162 IHF consen-
gion between positions ⴚ226 and ⴚ50. Gel mobility shift as- sus site. DNase I footprinting was used to determine if IHF
says were used to assess whether IHF directly binds to the and H-NS bind the same ⫺162 region within the tcpA pro-
virulence gene promoter regions. Increasing amounts of puri- moter. Radiolabeled fragments that spanned the tcpA pro-
fied V. cholerae IHF (180 and 360 nM) were incubated with 10 moter region from position ⫺282 to ⫺50 (IHF footprint) and
ng of ctx (C) or tcpA (A1) promoter fragments. IHF did not from position ⫺282 to ⫹78 (H-NS footprint) were generated,
bind the ctx fragment (C) but did bind the tcpA fragment (A1) incubated with increasing amounts of IHF (1.2 ␮M and 2.4
(Fig. 3), suggesting a direct role for IHF at the tcpA promoter ␮M) and H-NS (1 ␮M, 2 ␮M, and 4 ␮M), respectively, and
and not the ctx promoter. tcpP and toxT promoter fragments digested with DNase I as described in Materials and Methods.
also failed to bind IHF in gel mobility shift assays (data not The reaction mixtures were then subjected to electrophoresis
shown). These results are consistent with expression data that on a denaturing polyacrylamide gel. IHF protected a region
suggest that IHF does not have a direct regulatory role at the from position ⫺185 to ⫺126 on the top strand and a region
tcpP or toxT promoters. from position ⫺198 to ⫺136 on the bottom strand (Fig. 4A).
To locate the IHF binding site within the tcpA promoter This protection overlapped the putative IHF consensus site at
region, smaller fragments (A2 and A3) were used to determine position ⫺162 and the two putative H-NS nucleation sites, as
where binding occurs. Two putative IHF consensus sites based shown in Fig. 4C. DNase I-hypersensitive sites, a characteristic
on the E. coli IHF consensus site WATCAANNNNTTR, of DNA bending, flank the IHF protection, suggesting that
where W represents A or T and R represents A or G, are IHF might be bending the tcpA promoter upon binding. IHF
VOL. 190, 2008 IHF REGULATION OF tcpA EXPRESSION 4741

more centrally located the bending site, the more retarded

the migration of the protein-DNA complex. Gel mobility shift
assays with each fragment in the presence of IHF were carried
out as described above. All free DNA migrated at the same
rate, suggesting that there was no intrinsic bend in the DNA.
Mobility upon incubation with IHF was slowest when the IHF
binding site was centrally located (EcoRV) and maximal when
the site was closest to the end of the fragment (BglII and
BamHI). These results are consistent with the center of the
bend being located at the IHF binding site. Migration of the
fragments bound by IHF was measured from the well, and
the bending angle was calculated to be 120° by the method of
Thompson and Landy (61).
The influence of IHF consensus site mutations on IHF bind-
ing and tcpA expression. The IHF consensus site was altered in
order to disrupt IHF binding and activity at the tcpA promoter.
The first binding site mutation, termed ⌬consensus, was a
21-bp deletion that included the 13-bp IHF consensus site (Fig.
4C). The second, termed 4PM, was the introduction of 4-base-
pair changes in the most highly conserved consensus site res-
idues, in which the TCA and the second-to-last T were
changed to GGT and G, respectively (Fig. 4C). Fragments of
FIG. 3. Gel mobility shift assay for IHF binding to ctx and tcpA pro- the tcpA promoter region from position ⫺226 to ⫹78 contain-
moter DNA. The DIG-labeled promoter fragments of ctx from position ing these mutations were assessed for their effects on IHF
⫺194 to ⫹118 (C) and tcpA from position ⫺226 to ⫹78 (A1), ⫺226 to
⫺50 (A2), and ⫺100 to ⫹78 (A3) were incubated with purified IHF and binding in gel mobility shift assays as described above. IHF did
analyzed by a gel mobility shift assay. The first lane in each set has no not shift fragments containing the ⌬consensus or the 4PM
protein added, the second lane contains 180 nM IHF, and the third land mutation (Fig. 6A). These results suggest that this site is es-
contains 360 nM IHF. The boxes labeled ⫺162 and ⫺35 represent puta- sential for IHF binding. These mutations were then assayed for
tive IHF consensus sites on the tcpA promoter fragments.
their effects on tcpA transcription by integration into the tcpA-
lacZ background. Contrary to what was expected, neither the
⌬consensus nor the 4PM mutation mimicked an ihf mutant in
did not protect the putative IHF consensus site at the ⫺35 site its effect on tcpA expression. Rather than the 4-fold decrease
on the longer (⫺282 to ⫹78) tcpA fragment that contained seen previously in the ihf mutants, the promoter IHF consen-
both IHF consensus sites (data not shown). This suggests that sus mutations caused only a 1.6-fold decrease in tcpA-lacZ
the putative IHF site at position ⫺35 is not bound by IHF even expression (data not shown). In addition, TcpA protein pro-
in the presence of the ⫺162 site and that the single shifted duction was similar to wild-type levels when these mutations
complex observed with the A1 fragment in Fig. 3 is IHF bind- were placed in the otherwise wild-type O395 background (data
ing to the ⫺162 site alone. not shown). This result suggests that the mechanism of IHF
H-NS binding was more widespread than IHF binding within activation at this site may be more complicated than simply
the tcpA promoter region. H-NS bound over the IHF binding binding to the IHF consensus site.
site at position ⫺162 and strongly protected at least part of the H-NS binding to the IHF consensus mutations in gel mo-
ToxT binding site that spans from position ⫺84 to ⫺41 in the bility shift assays. One of the mechanisms by which IHF can
promoter region (71) (Fig. 4B). This binding region has an AT regulate transcription is alleviation of H-NS-mediated repres-
content ⬎70% similar to that of the IHF binding region. These sion of transcription (63, 64). Based on the idea that IHF can
results suggest that H-NS binding overlaps the binding sites of act as an antirepressor and given the DNase I protection data
both activators, IHF and ToxT, at the tcpA promoter and could that show overlapping H-NS and IHF sites, we chose to inves-
act to block the binding of these activator proteins. tigate whether the IHF consensus site mutations affected H-NS
IHF binding induces bending of the DNA upstream of tcpA. binding to the tcpA promoter fragments. Gel mobility shift
IHF has been shown to bend DNA in a number of systems assays were performed with wild-type and mutant tcpA pro-
(61), and this property has been suggested to be central to its moter fragments from position ⫺226 to ⫹78 with increasing
activities (39, 50, 57). We used the circular permutation assay amounts of H-NS (Fig. 6B). At low concentrations of H-NS,
of Wu and Crothers (66) to determine the ability of IHF to two shifted bands occurred with the wild-type fragment, a
bend 190-bp restriction fragments containing a 67-bp fragment result consistent with our DNase I protection determination
of the tcpA promoter region encompassing the IHF binding that showed multiple H-NS binding sites within this promoter
site at different positions relative to the ends of the restriction region. The wild-type fragment was completely shifted in the
fragment, as shown in Fig. 5A. The bending assay is based on presence of 320 nM H-NS, whereas fragments containing the
the observation that differences exist in the migration of linear ⌬consensus and 4PM mutations were not fully shifted at 320
and bent DNA fragments and that the electrophoretic mobility nM H-NS. These results suggest that the IHF consensus site
of bent protein-bound DNA differs with the positioning of the within the tcpA promoter not only is critical for IHF binding
DNA-binding site relative to either end of the fragment. The but also contributes to optimal H-NS binding. The reduced
4742 STONEHOUSE ET AL. J. BACTERIOL.

FIG. 4. DNase I protection by IHF and H-NS at the tcpA promoter. The top and bottom strands of the tcpA promoter fragments were incubated
with increasing amounts of IHF (1.2 and 2.4 ␮M) (A) or H-NS (1, 2, and 4 ␮M) (B) and treated with DNase I. (A) IHF protection from DNase
I on a promoter fragment from position ⫺282 to ⫺50 is shown with hatched bars. Arrows designate sites that were hypersensitive to DNase I
treatment. NP, no protein. (B) H-NS protection on a promoter fragment from position ⫺282 to ⫹78 is shown with black bars to the right. On the
left, the hatched bars designate the IHF consensus site and the gray boxes designate ToxT binding sites. (C) tcpA promoter sequence from position
⫺200 to ⫺125. Black bars designate the IHF protection on the top and bottom strands, and the IHF consensus site within the protected region
is in bold and aligned with the E. coli consensus site below. W, A/T; R, A/G; N, any nucleotide. The 21 bp that were deleted in the ⌬consensus
mutation are marked above the sequence. The four base pair mutations that were incorporated into the consensus site are designated 4PM below
the consensus alignment. Boxes indicate putative H-NS nucleation sites with seven matches to the 10-bp H-NS consensus site (29).

H-NS binding in the ⌬consensus and 4PM mutants may Addition of either 120 or 240 nM of IHF to a prebound H-NS
thereby serve to decrease repression of the tcpA promoter and fragment did not influence H-NS binding. However, addition
partially alleviate the requirement for IHF. The ⌬consensus of 420 nM IHF resulted in displacement of H-NS from the
promoter mutation deletes 5 of the 10 bp that comprise the DNA and a shift to a predominately IHF-DNA complex. Ad-
second putative H-NS high-affinity nucleation site (Fig. 4C) dition of 160 to 320 nM H-NS to a prebound IHF-DNA frag-
and could function to disrupt H-NS nucleation and polymer- ment did not significantly influence the IHF-DNA complex.
ization from this site. These results suggest that IHF effectively displaces H-NS from
Competitive gel mobility shift assays were performed be- the tcpA promoter fragment and that, once bound, IHF can
tween IHF and H-NS on the tcpA promoter region to examine essentially prevent H-NS binding.
the effect of adding either protein to prebound DNA com- Interplay of IHF, H-NS, and ToxT in the regulation of tcpA
plexes of the other. The DNA was incubated with the lowest expression. It is known that H-NS represses tcpA transcription.
concentrations of H-NS or IHF (320 or 420 nM, respectively) It has been shown that IHF alleviates H-NS repression at the
that gave a complete shift of the free DNA. After 15 min at bacteriophage Mu early (Pe) promoter and at the virB pro-
30°C, increasing amounts of IHF or H-NS, respectively, were moter in Shigella flexneri (47, 63). In an attempt to establish
added and incubation was continued for an additional 15 min. whether IHF functions as an antirepressor at the tcpA pro-
VOL. 190, 2008 IHF REGULATION OF tcpA EXPRESSION 4743

previously described by Hulbert and Taylor (22), there is a loss

of tcpA-lacZ expression in the absence of toxT (MBN142)
compared to that in a toxT-containing strain (MBN135) (Table
3). There was no further loss of tcpA-lacZ expression in the
ihfA toxT double mutant (ES41) compared to that in the toxT
mutant alone, suggesting that the positive effect of IHF at tcpA
requires ToxT. Furthermore, there was no affect of the ihf
mutant in an hns toxT background (MBN168) (Table 3). Taken
together, these results suggest that the IHF effect on tcpA
expression requires the presence of the repressor H-NS and
the activator ToxT.
Given that toxT expression is higher in an hns mutant due
to derepression (43) and that the ihf effect requires the
presence of ToxT, it is possible that the level of ToxT within
the cell determines the necessity for IHF at tcpA. We there-
fore examined the effect of overexpressing toxT in the
ihfA::kan background to determine if higher levels of ToxT
could overcome the role of IHF in activation. Wild-type
O395 or the O395 ihfA mutant ES21, harboring an empty
vector or the toxT overexpression vector pRH81, was grown
with or without arabinose to regulate toxT expression. Total
proteins were extracted, separated, and subjected to West-
ern blotting. The levels of ToxT were slightly reduced in the
FIG. 5. Bending of the tcpA promoter DNA by IHF as deter- ihfA mutant strain compared to wild-type levels when toxT
mined by a circular permutation assay. (A) Diagram of 190-bp was not induced from the plasmid (Fig. 7), consistent with
linear probes produced by digestion of plasmid pEAS7 (pBEND5- Fig. 2B. In addition, as seen in Fig. 2A, there is a drastic
tcpA) with five restriction enzymes. The probes contain the IHF
consensus site (hatched box) within a 67-bp tcpA promoter frag- reduction in TcpA levels in the ihfA mutant compared to
ment (⫺193 to ⫺127) (black box) at various positions relative to the wild-type levels, despite this modest difference in ToxT lev-
ends of the fragments. (B) Circular permutation analysis was per- els. In the wild-type strain, overexpression of toxT did not
formed on the tcpA IHF binding site with purified IHF. Each frag- detectably increase TcpA production (Fig. 7). However,
ment was incubated without protein or with 180 nM IHF and
analyzed by a gel mobility shift assay. The bending angle, ␣, was
TcpA was restored to near-wild-type levels upon overex-
determined to be 120° using the equation (␮M/␮E) ⫽ cos(␣/2), pression of toxT in the ihf mutant. These results suggest that
where ␮M is the relative migration of the fragment with the bend if ToxT levels are sufficiently high, IHF is not essential for
center in the middle and ␮E is the relative migration of the fragment maximal TcpA production. ToxT levels accumulate during
with the bend center at the end (61). exponential growth but decline in stationary phase (data not
shown). IHF may be necessary when ToxT levels are low
(early exponential phase and stationary phase under viru-
moter, we investigated expression levels in strains harboring lence gene-inducing conditions) to facilitate ToxT binding
wild-type or mutant copies of hns and ihfA. Cultures of the and activation at tcpA by overcoming some H-NS repression
tcpA-lacZ strains MBN135 (wild type), GK648 (ihfA::kan), and/or altering promoter structure to facilitate ToxT bind-
MBN148 (⌬hns), and ES42 (ihfA::kan ⌬hns) were grown over- ing. Similar to E. coli IHF (1), V. cholerae IHF is produced
night in LB with a starting pH of 6.5 at 30°C, and ␤-galacto- throughout the growth curve (under both virulence-inducing
sidase activity was measured. As was previously described by and -noninducing conditions), but levels are maximal in
Nye et al. (43), the level of tcpA-lacZ expression in the absence stationary phase (data not shown). This is consistent with a
of hns (MBN148) was increased about 1.5-fold compared to role for IHF in facilitating ToxT-mediated activation of tcpA
the wild-type (MBN135) level (Table 3). In the absence of hns when ToxT levels are low.
(MBN148), there was no further effect of deleting ihf on tcpA-
lacZ expression, suggesting that the ihf mutation is epistatic to DISCUSSION
the hns mutation at the tcpA promoter. Furthermore, deleting
hns has a much greater derepressive effect on tcpA-lacZ ex- In the V. cholerae El Tor strain N16961, McLeod et al. have
pression in the absence of ihfA than when ihfA is present found that, as in E. coli and Salmonella strains (2, 33), over 100
(501% versus 145%) (Table 3). This suggests that IHF coun- genes are differentially regulated in an ihfA mutant (34). In the
teracts H-NS repression at tcpA. work presented here, we show that IHF positively influences
In an attempt to determine if IHF-mediated activation at the the expression of virulence genes in the V. cholerae classical
tcpA promoter is ToxT dependent, we investigated expression O395 strain. The introduction of the ihfA::kan and ihfB::gen
levels in strains harboring wild-type or mutant copies of toxT alleles into the tcpA-lacZ (MBN135) and ctx-lacZ (KSK218)
and ihfA. Cultures of the tcpA-lacZ strains MBN135 (wild fusion strains resulted in a four- to fivefold reduction in the
type), GK648 (ihfA::kan), MBN142 (⌬toxT), and ES41 (ihfA:: expression of lacZ and prevented TCP and CT production. A
kan ⌬toxT) were grown overnight in LB with a starting pH of reduction in expression was not observed at the TcpP/ToxR-
6.5 at 30°C, and ␤-galactosidase activity was measured. As dependent toxT promoter immediately upstream of the toxT
4744 STONEHOUSE ET AL. J. BACTERIOL.

FIG. 6. Mutations in the tcpA IHF consensus site alter IHF and H-NS binding. Mutations in the tcpA promoter were constructed to disrupt
IHF binding. The wild-type (WT) and mutant promoter fragments (⌬consensus and 4PM) from position ⫺226 to ⫹78 were incubated in the
absence of protein or with 180 nM IHF (A) or 80, 160, 250, or 320 nM H-NS (B) and analyzed by a gel mobility shift assay. (C) Competitive gel
mobility shift assay between H-NS and IHF on the wild-type promoter fragment. Lane 1, no protein added. Lane 2, 320 nM H-NS; lane 3, 420 nM
IHF. Lanes 4 to 6, 320 nM H-NS and 120 nM (lane 4), 240 nM (lane 5), and 420 nM (lane 6) IHF. Lanes 7 to 9, 420 nM IHF and 160 nM (lane
7), 250 nM (lane 8), and 320 nM (lane 9) H-NS.

gene (Fig. 8). However, the observation that the levels of ToxT site centered at position ⫺162 in gel mobility shift and DNase
protein in the ihfA mutant were lower than those in the wild I footprinting assays. In addition, the protein was found to
type suggests that IHF might exert its effects at the ToxT- bend the DNA 120°. Since IHF did not appear to bind to the
dependent promoter upstream of tcpA that also influences toxT ctx promoter, these results suggest that the IHF effect on ctx
expression. A direct role for IHF at the tcpA promoter was
demonstrated by the ability of the purified protein to bind to a

TABLE 3. Effect of IHF on expression of the tcpA-lacZ fusion in

wild-type, hns, toxT, and hns toxT strains
␤-Galactosidase activity (%)a Fold
Strain
With ihfA Without ihfA change

MBN135 (wild type) 18,591 ⫾ 520 (100) 4,890 ⫾ 270 (100) 3.8
MBN148 (hns) 26,956 ⫾ 750 (145) 24,528 ⫾ 526 (502) 1.1
MBN142 (toxT) 331 ⫾ 27 (1.8) 355 ⫾ 15 (7.3) 1.1 FIG. 7. Overexpression of toxT in the ihfA mutant background.
MBN168 (hns toxT) 1,475 ⫾ 84 (7.9) 1,770 ⫾ 93 (36.2) 1.2 Strains were grown overnight in LB medium with a starting pH of 6.5 at
a
␤-Galactosidase activities (in Miller units) are shown for each strain and 30°C with shaking and with the addition of 0.1% arabinose where indi-
expressed as percentages of the wild-type level. The values are the averages for cated. Western blot analysis with anti-ToxT (␣-ToxT) and ␣-TcpA anti-
at least two independent experiments. bodies was performed with 16 ␮g of total protein extract. WT, wild type.
VOL. 190, 2008 IHF REGULATION OF tcpA EXPRESSION 4745

FIG. 8. Schematic of the transcriptional regulation of V. cholerae tcpA, toxT, and ctx expression. (A) OFF/INITIAL: H-NS represses gene
expression at multiple promoters within the ToxR regulon, including tcpA to -F, toxT, and ctx. H-NS binding in the tcpA promoter region overlaps
the IHF consensus sequence at position ⫺162 as well as the ToxT binding site spanning from position ⫺84 to ⫺41. (B) ON: After a transition to
virulence-inducing conditions, ToxR/S and TcpP/H in the inner membrane cooperate at the toxT promoter to activate toxT expression (TcpP/ToxR
dependent). ToxT-mediated activation of tcpA-toxT is facilitated by IHF binding, bending of the DNA, and displacement of H-NS, allowing for
greater accessibility of ToxT to its binding site. As ToxT levels accumulate, full activation of tcpA to -F, toxT, and ctx expression occurs.

expression is indirect due to lowered toxT expression driven the transcription of tcpA to -F. H-NS also appears to directly
from the ToxT-dependent promoter upstream of tcpA. repress toxT and ctx expression (43, 71). Under inducing con-
A direct role for H-NS in the repression of tcpA expression ditions (Fig. 8B), ToxT levels from the TcpP/ToxR-dependent
has previously been described (43, 71). Here, we show direct promoter increase due to cooperation between ToxR/S and
binding of H-NS to the tcpA promoter by gel mobility shift TcpP/H in the inner membrane. Activation of the ToxT-de-
assays and DNase I footprinting. H-NS protection covered the pendent tcpA-toxT transcript is facilitated by binding of IHF to
IHF binding site (position ⫺162) as well as the ToxT binding the site centered at position ⫺162 that introduces a bend in the
site (which spans from ⫺84 to ⫺41) (Fig. 8A). Genetic studies promoter that partially displaces H-NS from the DNA. This
of tcpA-lacZ expression in strains harboring wild-type or mu- allows greater accessibility of ToxT to its binding site, resulting
tant copies of ihfA, hns, or toxT showed that activation by IHF in transcriptional activation of the tcpA to -F genes as well as
appears to occur only when both H-NS and ToxT are present; toxT. The subsequent increase in ToxT protein levels permits
the ihfA mutation is epistatic to the hns and toxT mutations. full activation of tcpA to -F and ctx expression. In the absence
The functional dependence of IHF on H-NS and ToxT at tcpA, of IHF, we propose that under inducing conditions, the in-
along with the observed overlapping binding sites of these creased levels of ToxT from the TcpP/ToxR-dependent pro-
proteins, suggests a possible model whereby IHF functions as moter are insufficient to fully overcome H-NS repression, and
an antirepressor at this promoter. This is further supported by thus, expression of tcpA to -F, toxT, and ctx is not induced to
the binding studies that show that IHF can displace H-NS from maximal levels.
the tcpA regulatory region and, once bound, occlude H-NS Consistent with this model is the finding that loss of the IHF
binding. protein results in a fourfold reduction in tcpA expression,
As proposed in Fig. 8A, under noninducing conditions, whereas disruption of the IHF binding site by introduction of
H-NS binds directly to the tcpA promoter covering the IHF the ⌬consensus or 4PM mutation, both of which alter IHF and
consensus sequence at position ⫺162 as well as the ToxT H-NS binding, causes only a small reduction in tcpA expres-
binding site (spanning from ⫺84 to ⫺41), thereby repressing sion. Because the IHF and H-NS binding sites overlap, these
4746 STONEHOUSE ET AL. J. BACTERIOL.

promoter mutations phenotypically negate each other by si- (17). This group saw that the in vivo expression levels of ctx and
multaneously decreasing tcpA expression due to a reduction in tcpA were increased in an hns mutant 8 h after infection of the
IHF-mediated activation and increasing tcpA expression due to rabbit ileal loop. In addition, they noted that CT production
reduced H-NS repression. The hypothesis that IHF functions was about 2.5-fold greater in the hns mutant than in the wild-
as an antirepressor at the tcpA promoter is also supported by type O395 strain. These data demonstrate that H-NS is highly
the finding that IHF is not required upon overexpression of the expressed and active as a repressor at the ctx and tcpA pro-
primary activator ToxT. moters in this in vivo model. Proper colonization most likely
There are a number of reports of site-specific DNA-binding requires a mechanism to counteract some of this H-NS repres-
proteins acting as H-NS antagonists by displacing H-NS from sion in order to achieve optimal virulence gene expression and
common or adjacent binding sites. For example, H-NS is dis- cause disease. Yu and DiRita (71) showed that ToxT-mediated
placed by VirB binding at the icsB promoter in S. flexneri (62), activations of the ctx and tcpA promoters are mechanistically
LeuO binding at the ompS1 promoter in S. enterica (11), and distinct. At ctx, ToxT has a higher affinity for its binding site
SlyA binding to two distal SlyA/H-NS binding sites at hlyE in E. than it does at tcpA (71). However, Lee et al. have shown that
coli K-12 (31). At the yjjQ-bglJ promoter in E. coli, LeuO- during infection, maximal tcpA expression occurs before max-
induced DNA looping is proposed to prevent H-NS oligomer- imal ctx expression (30). This is likely due, in part, to the
ization along the promoter (58). Similarly, IHF may function at greater H-NS repressive effect on ctx expression than on tcpA
tcpA to displace H-NS from their overlapping binding site, and expression (43, 71). The differential expression might also be
the bend introduced upon IHF binding may prevent oligomer- due to the action of IHF at tcpA that could serve to lower the
ization along the promoter. A role for IHF in overcoming threshold level of ToxT that is required to activate tcpA ex-
H-NS repression at the early promoter (Pe) of bacteriophage pression despite the lower affinity of ToxT for its binding site at
Mu has already been established (63). The Pe promoter of this promoter. Further analysis of the molecular interactions
bacteriophage Mu appears to be specifically inhibited by the between IHF, H-NS, and ToxT at the tcpA promoter will pro-
binding of H-NS to the promoter region. Binding of IHF to the vide additional insights into the mechanism by which these
upstream site within the Pe promoter results in direct stimu- proteins interact to influence the later steps in the cascade that
lation of the promoter and also interferes with the formation of regulates virulence gene expression in V. cholerae.
DNA-protein complexes between H-NS and Pe, thereby coun-
teracting H-NS repression of this promoter. IHF also alleviates ACKNOWLEDGMENTS
H-NS repression at the virB promoter in S. flexneri (47). In this This work was supported by grants AI039654 (to R.K.T.) and
example, IHF binding does not overlap the H-NS binding site AI041558 (to K.S.) from the National Institutes of Health. E.S. is the
as in the case of the Pe promoter of Mu but is believed to recipient of a Training Grant award from NIH (AI007519).
We thank Steven Goodman for the anti-IHF polyclonal antibody
generate a conformational change in the DNA from an up- and Stephanie Batchelet for critical reading and insight into the manu-
stream location that facilitates the binding of the AraC-like script.
activator VirF, which itself overcomes H-NS repression. Sim-
REFERENCES
ilarly, ToxT appears to function, at least in part, to alleviate
1. Ali Azam, T., A. Iwata, A. Nishimura, S. Ueda, and A. Ishihama. 1999.
H-NS repression at the tcpA promoter, in addition to its role as Growth phase-dependent variation in protein composition of the Escherichia
an activator, since there is less H-NS repression of tcpA in the coli nucleoid. J. Bacteriol. 181:6361–6370.
presence of ToxT than in its absence (1.4-fold [wild type versus 2. Arfin, S. M., A. D. Long, E. T. Ito, L. Tolleri, M. M. Riehle, E. S. Paegle,
and G. W. Hatfield. 2000. Global gene expression profiling in Escherichia
⌬hns] versus 4.5-fold [⌬toxT versus ⌬toxT ⌬hns]). coli K12. The effects of integration host factor. J. Biol. Chem. 275:29672–
Many bacterial pathogens have acquired virulence genes by 29684.
3. Bertoni, G., N. Fujita, A. Ishihama, and V. de Lorenzo. 1998. Active recruit-
horizontal gene transfer, and it has been suggested that gene ment of ␴54-RNA polymerase to the Pu promoter of Pseudomonas putida:
silencing by H-NS allows the newly acquired genes to be inte- role of IHF and ␣CTD. EMBO J. 17:5120–5128.
grated into regulatory networks, like the ToxR regulon in V. 4. Bouffartigues, E., M. Buckle, C. Badaut, A. Travers, and S. Rimsky. 2007.
H-NS cooperative binding to high-affinity sites in a regulatory element re-
cholerae, over time without initially negatively affecting the sults in transcriptional silencing. Nat. Struct. Mol. Biol. 14:441–448.
fitness of the recipient bacterium (14). Enteric bacteria, like V. 5. Bushman, W., J. F. Thompson, L. Vargas, and A. Landy. 1985. Control of
cholerae, encounter multiple environmental stimuli both inside directionality in lambda site specific recombination. Science 230:906–911.
6. Carroll, P. A., K. T. Tashima, M. B. Rogers, V. J. DiRita, and S. B. Calder-
and outside the human host. It is not surprising, therefore, that wood. 1997. Phase variation in tcpH modulates expression of the ToxR
bacteria have evolved complex regulatory systems to overcome regulon in Vibrio cholerae. Mol. Microbiol. 25:1099–1111.
7. Champion, G. A., M. N. Neely, M. A. Brennan, and V. J. DiRita. 1997. A
H-NS silencing and respond to a range of environmental in- branch in the ToxR regulatory cascade of Vibrio cholerae revealed by char-
puts. As far as we know, the primary role of TCP and CT is in acterization of toxT mutant strains. Mol. Microbiol. 23:323–331.
human infection. The role of the global regulators H-NS and 8. Chiang, S. L., R. K. Taylor, M. Koomey, and J. J. Mekalanos. 1995. Single
amino acid substitutions in the N-terminus of Vibrio cholerae TcpA affect
IHF in infection has been hard to define due to their pleiotro- colonization, autoagglutination, and serum resistance. Mol. Microbiol. 17:
pic effects on transcription and the growth defects associated 1133–1142.
with their loss. It is likely that H-NS and IHF function to 9. Cotter, P. A., and V. J. DiRita. 2000. Bacterial virulence gene regulation: an
evolutionary perspective. Annu. Rev. Microbiol. 54:519–565.
modulate the topology of the promoters on which the more 10. Craig, N. L., and H. A. Nash. 1984. E. coli integration host factor binds to
specific regulators, like ToxT, can act to optimize virulence specific sites in DNA. Cell 39:707–716.
11. De la Cruz, M. A., M. Fernandez-Mora, C. Guadarrama, M. A. Flores-
gene expression in V. cholerae. Using microarray analysis, Xu Valdez, V. H. Bustamante, A. Vazquez, and E. Calva. 2007. LeuO antago-
et al. observed that hns is one of the most highly expressed V. nizes H-NS and StpA-dependent repression in Salmonella enterica ompS1.
cholerae genes in vitro and in the rabbit ileal loop model of V. Mol. Microbiol. 66:727–743.
12. DiRita, V. J., C. Parsot, G. Jander, and J. J. Mekalanos. 1991. Regulatory
cholerae infection (67). Additionally, Ghosh et al. (17) re- cascade controls virulence in Vibrio cholerae. Proc. Natl. Acad. Sci. USA
ported that H-NS represses ctx and tcpA expression in vivo 88:5403–5407.
VOL. 190, 2008 IHF REGULATION OF tcpA EXPRESSION 4747

13. Dorman, C. J. 2004. H-NS: a universal regulator for a dynamic genome. Nat. 40. Nash, H. A., C. A. Robertson, E. Flamm, R. A. Weisberg, and H. I. Miller.
Rev. Microbiol. 2:391–400. 1987. Overproduction of Escherichia coli integration host factor, a protein
14. Dorman, C. J. 2007. H-NS, the genome sentinel. Nat. Rev. Microbiol. 5:157– with nonidentical subunits. J. Bacteriol. 169:4124–4127.
161. 41. Navarre, W. W., M. McClelland, S. J. Libby, and F. C. Fang. 2007.
15. Engelhorn, M., and J. Geiselmann. 1998. Maximal transcriptional activation Silencing of xenogeneic DNA by H-NS-facilitation of lateral gene transfer
by the IHF protein of Escherichia coli depends on optimal DNA bending by in bacteria by a defense system that recognizes foreign DNA. Genes Dev.
the activator. Mol. Microbiol. 30:431–441. 21:1456–1471.
16. Esposito, D., J. S. Thrower, and J. J. Scocca. 2001. Protein and DNA 42. Navarre, W. W., S. Porwollik, Y. Wang, M. McClelland, H. Rosen, S. J.
requirements of the bacteriophage HP1 recombination system: a model for Libby, and F. C. Fang. 2006. Selective silencing of foreign DNA with low GC
intasome formation. Nucleic Acids Res. 29:3955–3964. content by the H-NS protein in Salmonella. Science 313:236–238.
17. Ghosh, A., K. Paul, and R. Chowdhury. 2006. Role of the histone-like 43. Nye, M. B., J. D. Pfau, K. Skorupski, and R. K. Taylor. 2000. Vibrio cholerae
nucleoid structuring protein in colonization, motility, and bile-dependent H-NS silences virulence gene expression at multiple steps in the ToxR
repression of virulence gene expression in Vibrio cholerae. Infect. Immun. regulatory cascade. J. Bacteriol. 182:4295–4303.
74:3060–3064. 44. Oshima, T., S. Ishikawa, K. Kurokawa, H. Aiba, and N. Ogasawara. 2006.
18. Grainger, D. C., D. Hurd, M. D. Goldberg, and S. J. Busby. 2006. Association Escherichia coli histone-like protein H-NS preferentially binds to horizon-
of nucleoid proteins with coding and non-coding segments of the Escherichia tally acquired DNA in association with RNA polymerase. DNA Res. 13:141–
coli genome. Nucleic Acids Res. 34:4642–4652. 153.
19. Guzman, L. M., D. Belin, M. J. Carson, and J. Beckwith. 1995. Tight regu- 45. Parekh, B. S., and G. W. Hatfield. 1996. Transcriptional activation by pro-
lation, modulation, and high-level expression by vectors containing the tein-induced DNA bending: evidence for a DNA structural transmission
arabinose PBAD promoter. J. Bacteriol. 177:4121–4130. model. Proc. Natl. Acad. Sci. USA 93:1173–1177.
20. Hase, C. C., and J. J. Mekalanos. 1998. TcpP protein is a positive regulator 46. Peterson, K. M., and J. J. Mekalanos. 1988. Characterization of the Vibrio
of virulence gene expression in Vibrio cholerae. Proc. Natl. Acad. Sci. USA cholerae ToxR regulon: identification of novel genes involved in intestinal
95:730–734. colonization. Infect. Immun. 56:2822–2829.
21. Higgins, D. E., and V. J. DiRita. 1994. Transcriptional control of toxT, a 47. Porter, M. E., and C. J. Dorman. 1997. Positive regulation of Shigella flexneri
regulatory gene in the ToxR regulon of Vibrio cholerae. Mol. Microbiol. virulence genes by integration host factor. J. Bacteriol. 179:6537–6550.
14:17–29. 48. Rice, P. A., S. Yang, K. Mizuuchi, and H. A. Nash. 1996. Crystal structure of
22. Hulbert, R. R., and R. K. Taylor. 2002. Mechanism of ToxT-dependent an IHF-DNA complex: a protein-induced DNA U-turn. Cell 87:1295–1306.
transcriptional activation at the Vibrio cholerae tcpA promoter. J. Bacteriol. 49. Rimsky, S., and A. Spassky. 1990. Sequence determinants for H1 binding on
184:5533–5544. Escherichia coli lac and gal promoters. Biochemistry 29:3765–3771.
23. Kaper, J. B., J. G. J. Morris, and M. M. Levine. 1995. Cholera. Clin. 50. Robertson, C. A., and H. A. Nash. 1988. Bending of the bacteriophage
Microbiol. Rev. 8:48–86. lambda attachment site by Escherichia coli integration host factor. J. Biol.
24. Kovacikova, G., W. Lin, and K. Skorupski. 2005. Dual regulation of genes Chem. 263:3554–3557.
involved in acetoin biosynthesis and motility/biofilm formation by the viru- 51. Ryan, V. T., J. E. Grimwade, C. J. Nievera, and A. C. Leonard. 2002. IHF and
lence activator AphA and the acetate-responsive LysR-type regulator AlsR HU stimulate assembly of pre-replication complexes at Escherichia coli oriC
in Vibrio cholerae. Mol. Microbiol. 57:420–433. by two different mechanisms. Mol. Microbiol. 46:113–124.
25. Kovacikova, G., and K. Skorupski. 1999. A Vibrio cholerae LysR homolog,
52. Sieira, R., D. J. Comerci, L. I. Pietrasanta, and R. A. Ugalde. 2004. Integra-
AphB, cooperates with AphA at the tcpPH promoter to activate expression
tion host factor is involved in transcriptional regulation of the Brucella
of the ToxR virulence cascade. J. Bacteriol. 181:4250–4256.
abortus virB operon. Mol. Microbiol. 54:808–822.
26. Kovacikova, G., and K. Skorupski. 2001. Overlapping binding sites for the
53. Skorupski, K., and R. K. Taylor. 1996. Positive selection vectors for allelic
virulence gene regulators AphA, AphB and cAMP-CRP at the Vibrio chol-
exchange. Gene 169:47–52.
erae tcpPH promoter. Mol. Microbiol. 41:393–407.
54. Skorupski, K., and R. K. Taylor. 1997. Cyclic AMP and its receptor protein
27. Krishnan, H. H., A. Ghosh, K. Paul, and R. Chowdhury. 2004. Effect of
negatively regulate the coordinate expression of cholera toxin and toxin-
anaerobiosis on expression of virulence factors in Vibrio cholerae. Infect.
coregulated pilus in Vibrio cholerae. Proc. Natl. Acad. Sci. USA 94:265–270.
Immun. 72:3961–3967.
55. Skorupski, K., and R. K. Taylor. 1997. Control of the ToxR virulence
28. Krukonis, E. S., R. R. Yu, and V. J. Dirita. 2000. The Vibrio cholerae
regulon in Vibrio cholerae by environmental stimuli. Mol. Microbiol. 25:
ToxR/TcpP/ToxT virulence cascade: distinct roles for two membrane-
1003–1009.
localized transcriptional activators on a single promoter. Mol. Microbiol.
38:67–84. 56. Skorupski, K., and R. K. Taylor. 1999. A new level in the Vibrio cholerae
29. Lang, B., N. Blot, E. Bouffartigues, M. Buckle, M. Geertz, C. O. Gualerzi, R. ToxR virulence cascade: AphA is required for transcriptional activation of
Mavathur, G. Muskhelishvili, C. L. Pon, S. Rimsky, S. Stella, M. M. Babu, the tcpPH operon. Mol. Microbiol. 31:763–771.
and A. Travers. 2007. High-affinity DNA binding sites for H-NS provide a 57. Stenzel, T. T., P. Patel, and D. Bastia. 1987. The integration host factor of
molecular basis for selective silencing within proteobacterial genomes. Nu- Escherichia coli binds to bent DNA at the origin of replication of the plasmid
cleic Acids Res. 35:6330–6337. pSC101. Cell 49:709–717.
30. Lee, S. H., D. L. Hava, M. K. Waldor, and A. Camilli. 1999. Regulation and 58. Stratmann, T., S. Madhusudan, and K. Schnetz. 2008. Regulation of the
temporal expression patterns of Vibrio cholerae virulence genes during in- yjjQ-bglJ operon, encoding LuxR-type transcription factors, and the diver-
fection. Cell 99:625–634. gent yjjP gene by H-NS and LeuO. J. Bacteriol. 190:926–935.
31. Lithgow, J. K., F. Haider, I. S. Roberts, and J. Green. 2007. Alternate SlyA 59. Sun, D. X., J. M. Seyer, I. Kovari, R. A. Sumrada, and R. K. Taylor. 1991.
and H-NS nucleoprotein complexes control hlyE expression in Escherichia Localization of protective epitopes within the pilin subunit of the Vibrio
coli K-12. Mol. Microbiol. 66:685–698. cholerae toxin-coregulated pilus. Infect. Immun. 59:114–118.
32. Lucchini, S., G. Rowley, M. D. Goldberg, D. Hurd, M. Harrison, and J. C. 60. Taylor, R. K., V. L. Miller, D. B. Furlong, and J. J. Mekalanos. 1987. Use of
Hinton. 2006. H-NS mediates the silencing of laterally acquired genes in phoA gene fusions to identify a pilus colonization factor coordinately regu-
bacteria. PLoS Pathog. 2:e81. lated with cholera toxin. Proc. Natl. Acad. Sci. USA 84:2833–2837.
33. Mangan, M. W., S. Lucchini, V. Danino, T. O. Croinin, J. C. Hinton, and 61. Thompson, J. F., and A. Landy. 1988. Empirical estimation of protein-
C. J. Dorman. 2006. The integration host factor (IHF) integrates stationary- induced DNA bending angles: applications to lambda site-specific recombi-
phase and virulence gene expression in Salmonella enterica serovar Typhi- nation complexes. Nucleic Acids Res. 16:9687–9705.
murium. Mol. Microbiol. 59:1831–1847. 62. Turner, E. C., and C. J. Dorman. 2007. H-NS antagonism in Shigella flexneri
34. McLeod, S. M., V. Burrus, and M. K. Waldor. 2006. Requirement for Vibrio by VirB, a virulence gene transcription regulator that is closely related to
cholerae integration host factor in conjugative DNA transfer. J. Bacteriol. plasmid partition factors. J. Bacteriol. 189:3403–3413.
188:5704–5711. 63. van Ulsen, P., M. Hillebrand, L. Zulianello, P. van de Putte, and N. Goosen.
35. McLeod, S. M., and R. C. Johnson. 2001. Control of transcription by nucle- 1996. Integration host factor alleviates the H-NS-mediated repression of the
oid proteins. Curr. Opin. Microbiol. 4:152–159. early promoter of bacteriophage Mu. Mol. Microbiol. 21:567–578.
36. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor 64. van Ulsen, P., M. Hillebrand, L. Zulianello, P. van de Putte, and N. Goosen.
Laboratory Press, Cold Spring Harbor, NY. 1997. The integration host factor-DNA complex upstream of the early pro-
37. Miller, V. L., V. J. DiRita, and J. J. Mekalanos. 1989. Identification of toxS, moter of bacteriophage Mu is functionally symmetric. J. Bacteriol. 179:3073–
a regulatory gene whose product enhances ToxR-mediated activation of the 3075.
cholera toxin promoter. J. Bacteriol. 171:1288–1293. 65. Withey, J. H., and V. J. DiRita. 2005. Activation of both acfA and acfD
38. Miller, V. L., R. K. Taylor, and J. J. Mekalanos. 1987. Cholera toxin tran- transcription by Vibrio cholerae ToxT requires binding to two centrally lo-
scriptional activator ToxR is a transmembrane DNA binding protein. Cell cated DNA sites in an inverted repeat conformation. Mol. Microbiol. 56:
48:271–279. 1062–1077.
39. Moitoso de Vargas, L., S. Kim, and A. Landy. 1989. DNA looping generated 66. Wu, H. M., and D. M. Crothers. 1984. The locus of sequence-directed and
by DNA bending protein IHF and the two domains of lambda integrase. protein-induced DNA bending. Nature 308:509–513.
Science 244:1457–1461. 67. Xu, Q., M. Dziejman, and J. J. Mekalanos. 2003. Determination of the
4748 STONEHOUSE ET AL. J. BACTERIOL.

transcriptome of Vibrio cholerae during intraintestinal growth and midexpo- 70. Yu, R. R., and V. J. DiRita. 1999. Analysis of an autoregulatory loop con-
nential phase in vitro. Proc. Natl. Acad. Sci. USA 100:1286–1291. trolling ToxT, cholera toxin, and toxin-coregulated pilus production in Vibrio
68. Yamada, H., S. Muramatsu, and T. Mizuno. 1990. An Escherichia coli pro- cholerae. J. Bacteriol. 181:2584–2592.
tein that preferentially binds to sharply curved DNA. J. Biochem. (Tokyo) 71. Yu, R. R., and V. J. DiRita. 2002. Regulation of gene expression in Vibrio
108:420–425. cholerae by ToxT involves both antirepression and RNA polymerase stimu-
69. Yona-Nadler, C., T. Umanski, S. Aizawa, D. Friedberg, and I. Rosenshine. 2003. lation. Mol. Microbiol. 43:119–134.
Integration host factor (IHF) mediates repression of flagella in enteropathogenic 72. Zwieb, C., and S. Adhya. 1994. Improved plasmid vectors for the analysis of
and enterohaemorrhagic Escherichia coli. Microbiology 149:877–884. protein-induced DNA bending. Methods Mol. Biol. 30:281–294.