Download Identification and characterization of mutants with increased

FEMS Immunology and Medical Microbiology 28 (2000) 25^35
www.fems-microbiology.org
Identi¢cation and characterization of mutants with increased
expression of hilA, the invasion gene transcriptional activator of
Salmonella typhimurium
Thomas F. Fahlen, Nitin Mathur, Bradley D. Jones *
Department of Microbiology, University of Iowa School of Medicine, 3-330 Bowen Science Bldg., 51 Newton Rd., Iowa City, IA 52242-1109, USA
Received 8 October 1999; accepted 8 December 1999
Abstract
Induction of invasion gene transcription and expression of the invasive phenotype of Salmonella strains are regulated by environmental
conditions. Experimental evidence indicates that oxygen, pH, and osmotic conditions need to closely resemble those of the host intestinal
lumen for invasion gene activation. The hilA gene, encoded on Salmonella pathogenicity island 1 (SPI-1), is a transcriptional activator which
is required for invasion and whose expression is modulated by oxygen, pH, and osmolarity. Additionally, hilA is regulated by genetic
elements encoded on SPI-1 (hilC/sirC/sprA and hilD), as well as by elements which reside outside of SPI-1 (phoP/phoQ and sirA), although
how environmental signals modulate hilA is unknown. In an effort to further characterize the Salmonella invasion gene regulon, we have
created and preliminarily characterized 18 Tn5 insertions which result in upregulation of a hilA: :lacZY fusion. We have classified the
mutations based on location and phenotype into three classes. Six class 1 and six class 2 mutants have insertions in SPI-1 near the invasion
gene orgA or the invasion gene regulator hilD, respectively. Six class 3 mutants reside outside of SPI-1 in four different loci. The class 2 and 3
mutations induce overexpression of an episomal hilA: :lacZY fusion and significantly increase S. typhimurium invasion of HEp-2 cells in a
standard invasion assay. These data implicate new regions of SPI-1 as being involved in the regulation of invasion by S. typhimurium and
identify new invasion gene regulators located outside of SPI-1. ß 2000 Federation of European Microbiological Societies. Published by
Elsevier Science B.V. All rights reserved.
Keywords : Salmonella typhimurium ; Invasion gene regulation ; hilA
1. Introduction
Pathogenic Salmonella typhimurium strains cause two
types of infectious disease. In healthy humans the bacteria
cause a self-limiting gastroenteritis of the small intestine
while infection of mice leads to a systemic enteric fever [1].
Since S. typhimurium infections in mice closely resemble
human enteric fever, murine infection models are commonly used as models of typhoid fever. The pathogenic
strategy of enteroinvasive Salmonella species requires that
the bacteria invade host intestinal cells to initiate disease.
Infection begins when organisms invade the apical surface
of enterocytes in the small intestine [2^4]. The entry process occurs via actin rearrangement and ru¥ing of the cell
membrane that results in cellular projections which engulf
the bacteria [5]. Invasion causes destruction of the cell
which allows the bacteria to proceed to submucosal sites
* Corresponding author. E-mail : [email protected]
where they are taken up by macrophages and, in an appropriate host, persist within the phagolysosomal compartment. Survival within circulating macrophages allows
dissemination via the lymphatic system to the liver and
spleen of the host causing systemic infection (enteric fever).
Genetic elements responsible for the invasive phenotype
of Salmonella have been identi¢ed and localized to a 40-kb
chromosomal region at centisome 63 termed Salmonella
pathogenicity island 1 (SPI-1) (reviewed in [6]). The SPI1 genes are predicted to encode products with similarity to
the structural components of type III secretion systems
found in both plant and animal pathogens including Pseudomonas, Rhizobium, Erwinia, Yersinia, Shigella, and enteropathogenic Escherichia coli [7]. Type III secretion systems are composed of 20 or more proteins which allow
e¡ector proteins to be secreted across the inner and outer
membranes of Gram-negative bacteria and are typically
involved in the translocation of virulence proteins to eukaryotic cells [8]. In addition to virulence loci within
0928-8244 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 8 - 8 2 4 4 ( 9 9 ) 0 0 1 9 9 - 6
FEMSIM 1177 5-4-00
26
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
SPI-1, loci outside of SPI-1, which encode proteins secreted via the type III secretion apparatus of SPI-1, have
been found to induce cellular changes in tissue culture cells
and/or to a¡ect the ability of Salmonella to invade tissue
culture cells [9^13].
The expression of the invasion-associated type III secretion apparatus and the secreted e¡ector proteins are
tightly regulated in Salmonella in response to environmental conditions including osmolarity, oxygen, and pH such
that all conditions must be appropriate to allow invasion
gene expression [14^18]. The SPI-1-encoded hilA gene encodes a transcriptional activator which modulates expression of the type III secretion apparatus and/or the secreted
e¡ector proteins [19]. The modulation of hilA expression
appears to be the primary system for regulation of the
invasive phenotype of Salmonella since overexpression of
hilA confers a hyperinvasive phenotype and counteracts
the e¡ects of invasion-repressing signals [14,19,20]. Results
in our laboratory and others [21,22] show that null mutations in hilA result in dramatic attenuation of invasion
demonstrating the importance of hilA in the invasion phenotype. It appears that the signi¢cance of hilA in the invasion phenotype stems from its role in the activation of
expression of components of the type III secretion apparatus as well as the secreted proteins encoded on SPI-1
[14,23].
As invasion gene regulation is studied more extensively
in Salmonella, it is evident that a complex regulatory
scheme exists to allow expression of invasion genes only
under optimal conditions. The elements involved in this
scheme and their roles in regulation are not well understood. In an e¡ort to elucidate pathways involved in invasion gene regulation, we conducted a search for genetic
elements involved in the modulation of hilA expression.
We describe here the identi¢cation of 18 mutations created
through random Tn5 mutagenesis which result in increased expression of a chromosomal hilA: :lacZ transcriptional fusion.
2. Materials and methods
2.1. Bacterial strains and growth conditions
Bacterial strains and plasmids used in this study are
listed in Table 1. Liquid cultures were grown in Lennox
broth (LB) (Gibco BRL) with appropriate antibiotics
added to the following ¢nal concentrations (Wg ml31 ):
ampicillin, 100; kanamycin, 50; streptomycin, 100; and
tetracycline, 20. MacConkey lactose agar (Difco) was
used to assay the L-galactosidase phenotype of S. typhimurium reporter strains. High and low oxygen bacterial
growth conditions were used to repress or induce invasion
gene transcription. High oxygen cultures were obtained by
inoculating 3 ml of LB broth with 10 Wl of a stationaryphase culture and growing vigorously at 200 rpm at 37³C
to early exponential phase (V1U108 CFU ml31 ). Low
oxygen cultures were initiated by inoculating 3 ml of LB
broth with 10 Wl of a stationary-phase culture and growing
the bacteria statically overnight at 37³C to a density of
V4^5U108 CFU ml31 [21,24].
2.2. Plasmid construction
The plasmid pJK001 encodes the prgH, -I, -J, -K and
orgA genes with the native prg promoter. The plasmid was
constructed by combining portions of pBDJ116 and
pBDJ134 [25]. Brie£y, a HindIII-BamHI fragment from
pBDJ116 that contains prgH, prgI and the ¢rst 185 bp
of prgJ was cloned into pACYC177 cut with HindIII
and BamHI and designated pBDJ233. A second plasmid,
pBDJ234, was constructed with the terminal portion of
prgJ, and intact prgK and orgA genes. In addition,
pBDJ234 carries a kanamycin cassette just downstream
of the orgA gene. Digestion of pBDJ234 with BamHI
and BclI released a fragment containing the last 118 bp
of prgJ, prgK, orgA and the kanamycin gene which was
cloned into pBDJ233 digested with BamHI to create
pJK001. Creation of the desired plasmid was veri¢ed by
restriction enzyme digestion and agarose gel electrophoresis.
2.3. Mutagenesis of S. typhimurium with Tn5
Transposon mutagenesis of the S. typhimurium hilA: :
lacZY reporter strain BJ57 was performed using the plasmid pRTP1: :Tn5. This plasmid harbors the Tn5 transposon (kanR ), the rpsL gene, which confers sensitivity to
streptomycin, and an ampicillin resistance marker. In
BJ57, which is streptomycin-resistant due to a mutation
in the chromosomal rpsL gene, pRTP1: :Tn5 imparts resistance to kanamycin and ampicillin yet the strain is
streptomycin-sensitive because the rpsL allele on the
pRTP1 : :Tn5 plasmid is dominant to the chromosomal
rpsL gene. To create random chromosomal Tn5 insertions,
BJ57 was transformed with pRTP1 : :Tn5, several hundred
colonies were plated onto each of 10 plates, and the colonies were replica-plated to LB agar containing streptomycin and kanamycin. Growth on this medium results
from transposition of Tn5 to the chromosome and subsequent loss of pRTP1: :Tn5. The bacteria that grew on each
of the 10 selective plates were collected to form 10 separate
pools. At least 3000 colonies from each of the 10 pools of
streptomycin- and kanamycin-resistant bacteria were
plated onto MacConkey agar to identify mutants with
deviations in the `¢sheye' phenotype of BJ57.
2.4. Identi¢cation of Tn5 insertion sites
The site of the Tn5 insertion in each mutant, except
BJ749 and TF43, was determined by cloning the kanamycin resistance marker and £anking DNA from the chro-
FEMSIM 1177 5-4-00
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
27
Table 1
Bacterial strains, plasmids, and phage used in this study
Strain/plasmid
Strain
E. coli
DH12S
S. typhimurium
SL1344
EE251
BJ35
BJ57
BJ606
BJ749
TF25
TF26
TF27
TF28
TF29
TF30
TF37
TF38
TF39
TF40
TF41
TF42
TF43
TF44
TF45
TF46
TF47
TF48
TF49
TF50
TF51
TF52
TF53
TF54
TF55
TF56
TF57
TF58
TF59
TF60
TF61
TF62
TF63
TF64
TF65
Plasmid
pJK001
pLS31
pRTP1: :Tn5
Relevant genotype
Reference/source
mcrA v(mrr-hsdRMS-mcrBC) FP lacIq lacZvM15
Gibco BRL
wild-type virulent strain
invasive LT2 derivative, rpsL
EE251 pepT: :Mud1(x) vzda88Tn10
EE251 hilA: :Tn5lacZY
EE251 hilA: :Tn5lacZY prgH3 : :Tn5
EE251 prgH3: :Tn5
EE251 hilA: :Tn5lacZY prgK1 : :Tn5
EE251 hilA: :Tn5lacZY orgA1 : :Tn5
EE251 hilA: :Tn5lacZY orgA2 : :Tn5
EE251 hilA: :Tn5lacZY orgA3 : :Tn5
EE251 hilA: :Tn5lacZY orgA4 : :Tn5
EE251 hilA: :Tn5lacZY orgA5 : :Tn5
EE251 hilA: :Tn5lacZY prgH4 : :Tn5
EE251 hilA: :Tn5lacZY prgH5 : :Tn5
EE251 hilA: :Tn5lacZY prgH6 : :Tn5
EE251 hilA: :Tn5lacZY prgH7 : :Tn5
EE251 hilA: :Tn5lacZY prgH8 : :Tn5
EE251 prgH4: :Tn5
EE251 prgH5: :Tn5
EE251 prgH6: :Tn5
EE251 prgH7: :Tn5
EE251 prgH8: :Tn5
EE251 hilA: :Tn5lacZY ams1: :Tn5
EE251 hilA: :Tn5lacZY hupB1: :Tn5
EE251 hilA: :Tn5lacZY pag1 : :Tn5
EE251 hilA: :Tn5lacZY hilE3: :Tn5
EE251 hilA: :Tn5lacZY hilE2: :Tn5
EE251 hilA: :Tn5lacZY hilE1: :Tn5
EE251 ams1: :Tn5
EE251 hupB1: :Tn5
EE251 pag1: :Tn5
EE251 hilE3: :Tn5
EE251 hilE2: :Tn5
EE251 hilE1: :Tn5
EE251 pepT: :Mud1(x)
EE251 pepT: :Mud1(x) ams1: :Tn5
EE251 pepT: :Mud1(x) hupB1 : :Tn5
EE251 pepT: :Mud1(x) pag1: :Tn5
EE251 pepT: :Mud1(x) hilE3: :Tn5
EE251 pepT: :Mud1(x) hilE2: :Tn5
EE251 pepT: :Mud1(x) hilE1: :Tn5
[49]
[25]
This
[21]
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
This
pACYC derivative encoding prgH, -I, -J, -K and orgA, ampR , kanR
low copy plasmid with a hilA-lacZY fusion, tetR
Tn5 transposon inserted onto plasmid pRTP1 which carries the rpsL gene (strS ) ampR , kanR
Jones Laboratory
C. Lee
[50,25]
mosome into the plasmid vector pBluescript (Stratagene).
Approximately 0.5 Wg of chromosomal DNA was isolated
using a Dneasy tissue kit (Qiagen), digested with BamHI, and ligated into the unique BamHI site of pBluescript. Ligated DNA was then used to transform E. coli
DH12S (Gibco BRL) to kanamycin resistance via electroporation. Because of a unique BamHI site in Tn5, kanamycin-resistant clones carry the kanamycin resistance
marker and chromosomal DNA £anking the Tn5 insertion
site. These plasmids were isolated from DH12S and used
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
work
as template with the primer oxregA (5P-TTCAGGACGCTACTTGTG-3P) in a sequencing reaction to determine the
chromosomal sequence £anking the Tn5 insertion site.
Due to di¤culty in cloning the TF43 insertion site, PCR
was used to determine the location of Tn5 in this mutant.
The general area of insertion was ¢rst identi¢ed using
PCR with the primers oxregA (5P-TTCAGGACGCTACTTGTG-3P) and prgH3P (5P-AGATCTAAAGTGGGCTTGGGAAATAC-3P) which ampli¢ed a 1.9-kb band
using TF43 chromosome as template. This band was ex-
FEMSIM 1177 5-4-00
28
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
cised from an agarose gel and the DNA was eluted using a
gel extraction kit (Qiagen). The puri¢ed PCR product was
then sequenced as above using the oxregA primer to direct
sequencing o¡ the end of the Tn5 segment into the ampli¢ed £anking DNA.
Inverse PCR [26] was used to identify the location of the
Tn5 insertion in S. typhimurium BJ749. Chromosomal
DNA from BJ749 was isolated using the technique of
Marmur [27] and V1 Wg was digested to completion
with BamHI, which cuts within the Tn5 transposon. The
BamHI-digested chromosomal DNA was diluted 1:100 in
dH2 0 and ligated overnight at 16³C to form closed circular
DNA. The ligated DNA was used as template in a 50 Wl
Expand Long Template PCR reaction (Promega). The
primers Tn5B21L (5P-GTTCAGGACGCTACTTGTG-3P)
and Mathur (5P-AGGACGACGAGGCTTGCA-3P) anneal to sequences adjacent to the BamHI site of Tn5 and
direct DNA synthesis away from each other. PCR ampli¢cation with these primers resulted in ampli¢cation of a
4.5-kb product which was ligated into pGEM-T (Promega) and electroporated into E. coli DH12S. Plasmid DNA
was isolated from white colonies on X-gal plates and the
presence of an insert in the vector was veri¢ed by restriction enzyme digestion and agarose gel electrophoresis.
Plasmid DNA was used as template with Tn5B21L primer
(5P-GTTCAGGACGCTACTTGTG-3P) to direct sequencing into DNA £anking the Tn5 insertion site.
DNA sequencing was conducted by the DNA core facility at the University of Iowa. Sequences obtained were
used to search DNA databases available at the National
Center for Biotechnology Information (www.ncbi.nlm.
nih.gov), the Sanger Centre (www.sanger.ac.uk), and the
Genome Sequencing Center at the Washington University
School of Medicine (genome.wustl.edu/gsc/bacterial/salmonella.shtml).
2.5. P22-mediated transductions
Mutations marked by antibiotic resistance genes were
moved by transduction with P22 HT int3 as previously
described [28]. Transductants were selected on LB plates
with appropriate antibiotics and EGTA to prevent secondary infections by P22.
2.6. L-Galactosidase assays
L-Galactosidase production from lacZ reporter constructs was quantitated using the method of Miller [29]
after growth of the bacteria under high or low oxygen
conditions.
2.7. Tissue culture conditions and invasion assays
HEp-2 tissue culture cells [30] were maintained in RPMI
1640 (Gibco BRL) containing 10% (v/v) fetal bovine serum (FBS) and passaged every 2^3 days. Invasion assays
were done as previously described [24]. Brie£y, wells of a
24-well dish containing 105 HEp-2 cells per well seeded the
previous day were inoculated with 107 bacteria. Bacteria
were allowed to invade for 1 h before washing each well
with PBS three times. One ml of RPMI 1640 medium
containing 50 Wg ml31 of gentamicin was added to each
well and the plates were incubated at 37³C for 90 min.
Wells were then washed with PBS again before lysing
the cells by the addition of 100 Wl of 1% Triton X-100
to each well. After incubation at 37³C for 10 min, 400
Wl of LB was added to each well and dilutions from
each well were plated to determine the number of intracellular bacteria. The percent invasion is de¢ned as the
number of gentamicin-resistant bacteria recovered from
the HEp-2 cells divided by the number of bacteria in the
inoculum multiplied by 100.
3. Results
3.1. Identi¢cation of S. typhimurium mutants with altered
hilA expression
To identify mutants with aberrant hilA expression, we
took advantage of the MacConkey agar phenotype of S.
typhimurium strain BJ57 which carries an oxygen-regulated hilA: :Tn5lacZY transcriptional fusion. Colonies of
this strain have a `¢sheye' phenotype on MacConkey agar
(red center and a white periphery) due to the expression of
the lacZ fusion in the low oxygen environment at the
center of the colony [25,31]. We were interested in isolating solid red mutants on MacConkey agar since these may
carry disruptions in negative regulators of hilA. S. typhimurium BJ57 was mutagenized with a plasmid-mediated
transposon delivery system that allowed us to create Tn5
insertions randomly throughout the S. typhimurium chromosome. Ten separate pools of random Tn5 chromosomal
insertions were created and screened by plating for isolated colonies on MacConkey agar. More than 15 000 colonies were screened which yielded 30 mutants (three from
each pool) that exhibited a darker red phenotype on MacConkey agar compared to wild-type. To con¢rm a single
Tn5 insertion as the cause of the altered phenotype, the
Tn5 was moved from each mutant to S. typhimurium
EE251 (wild-type) and then back into BJ57 via P22 phage
transduction. Strains isolated in this manner contained
single Tn5 insertions which a¡ected hilA: :lacZ expression
and were characterized further.
3.2. The class 1 mutants cluster near the orgA invasion gene
The insertion site of Tn5 in each of the mutants was
determined by cloning kanamycin resistance and £anking
DNA (see Section 2). Sequence analysis of the mutants
revealed that several of the transposons had inserted
near orgA, a gene encoded on SPI-1 which is required
FEMSIM 1177 5-4-00
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
29
for invasion and the secretion of virulence proteins [21,25].
This group of mutants has been designated class 1.
Although these mutants were isolated as colonies which
appeared darker red than wild-type on MacConkey agar,
the hilA: :lacZ expression of these strains as measured in a
standard L-galactosidase assay was not signi¢cantly di¡erent from wild-type (data not shown). Therefore, these mutations were not characterized further.
3.3. The class 2 mutants cluster upstream of the hilA
derepressor hilD and cause increased expression of
chromosomal and episomal hilA : :lacZ
Six Tn5 insertions termed class 2 were found to cluster
in the region of SPI-1 encoding the prgH operon and the
newly described hilD gene [32,33] (Fig. 1B). Five of these
insertions lie within the prgH, -I, -J, -K and orgA operon
([32] and J. Klein, manuscript in preparation) while the
sixth insertion in this region lies between the prgH and
hilD genes and as such does not disrupt either of these
protein coding regions. The class 2 mutations alter the
normal expression of the chromosomal hilA: :lacZ fusion
which is typically induced approximately 2^3-fold by
growth in low versus high oxygen conditions. Increases
caused by these mutations range from 0- to 3-fold high
oxygen and 0.5- to 14-fold in low oxygen (Fig. 1A).
To further characterize the class 2 mutants, strains containing the mutations in an otherwise wild-type back-
Fig. 1. Location of the class 2 mutations and their e¡ect on chromosomal hilA: :lacZ expression. A: Chromosomal hilA : :lacZ expression in
class 2 mutant backgrounds. Strains were grown under high oxygen
(white bars) or low oxygen (dark bars) and assays were conducted to
quantitate L-galactosidase activity (see Section 2). BJ57 is the hilA: :lacZ
parent strain. Data presented represent the averages of at least three independent experiments. Error bars depict the standard deviation from
the mean. B: Partial map of SPI-1 showing the location of the class 2
mutant Tn5 insertions in the S. typhimurium chromosome. The distances
in nucleotides of the insertions to the ¢rst nucleotide of the HilD coding
sequence are as follows : TF37, 936; TF38, 811 ; TF39, 664; BJ606, 652,
TF40, 349 ; TF41, 99.
Fig. 2. E¡ect of the class 2 mutations on expression of an episomal
hilA : :lacZ fusion. Strains were grown under high oxygen (white bars)
or low oxygen (dark bars) and assays were conducted to quantitate
L-galactosidase activity (see Section 2). EE251 is the wild-type parent
strain. Plasmid pLS31 encodes a hilA: :lacZ fusion in the low copy vector pRW50 [33]. Data presented represent the averages of at least three
independent experiments. Error bars depict the standard deviation from
the mean.
ground were transformed with the low copy hilA: :lacZ
reporter plasmid, pLS31 [33]. As shown in Fig. 2, these
mutations cause a signi¢cant increase in hilA: :lacZ expression from pLS31 demonstrating that the mutations can
a¡ect hilA expression in trans.
3.4. Invasion by class 2 mutants is not repressed by high
oxygen conditions
Overexpression of hilA has been shown to result in a
hyperinvasive phenotype for tissue culture cells [20]. Since
the class 2 mutations cause a signi¢cant increase in hilA
expression, we were interested in examining the e¡ect of
these mutations on the invasion phenotype. As ¢ve of the
class 2 mutations disrupt the prgH, -I, -J, -K and orgA
genes, which are required for invasion, strains containing
these mutations are attenuated for invasion in a standard
invasion assay of HEp-2 cell monolayers (data not
shown). However, when these strains are transformed
with plasmid pJK001, which encodes the prgH, -I, -J,
-K, and orgA, genes, low oxygen invasion is restored to
levels near wild-type and high oxygen invasion is 10^100fold above wild-type, with the exception of TF43 and
TF45 (Fig. 3). The increase in invasion (high or low oxygen) by strains TF42, TF44, BJ749, and TF46 is not due
to the presence of pJK001 since invasion by the wild-type
strain containing this plasmid is still repressed by growth
in high oxygen (Fig. 3). One of the class 2 mutants, TF46,
has a Tn5 insertion between prgH and hilD that does not
disrupt genes required for invasion (Fig. 1B). Similar to
the other class 2 mutants with pJK001 present, TF46 is
nearly 100-fold more invasive under high oxygen conditions than the wild-type strain (Fig. 3). Taken together,
these data indicate that four of the class 2 mutations result
in a decreased ability to respond to the invasion-repressing
growth condition of high oxygen and that the decreased
ability is not due to the disruption of prgH -I, -J, -K or
orgA.
FEMSIM 1177 5-4-00
30
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
Fig. 3. E¡ect of the class 2 mutations on invasion of HEp-2 cell monolayers. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and invasion assays were conducted as described in Section 2. EE251 is the wild-type parent strain. Data presented represent
the average of at least three independent experiments with duplicate
wells. Error bars depict the standard deviation from the mean.
Fig. 4. E¡ect of the class 3 mutations on expression of chromosomal
hilA : :lacZ. Strains were grown under high oxygen (white bars) or low
oxygen (dark bars) and assays were conducted to quantitate L-galactosidase activity (see Section 2). BJ57 is the hilA : :lacZ parent strain. Data
presented represent the average of at least three independent experiments. Error bars depict the standard deviation from the mean.
3.5. The class 3 mutations map outside of SPI-1
ducted on strains harboring chromosomal or episomal
hilA: :lacZ fusions. The class 3 mutations cause small effects on chromosomal hilA: :lacZ expression under noninducing conditions but up to 3-fold increases of hilA : :
lacZ under inducing conditions as measured by L-galactosidase assay (Fig. 4). The ams mutation and one of the
mutations in hilE appear to express hilA: :lacZ only
slightly higher than wild-type. It should be noted that
the hilE3: :Tn5 mutation in TF50 is at the extreme 3P
end of the putative ORF and thus the mutation may
only partially disrupt activity of the protein. The other
strains with Tn5 insertions in hilE, TF51 and TF52, cause
disruptions in the 5P end and result in a 1.5-fold increase in
chromosomal hilA: :lacZ expression under our low oxygen
assay conditions (Fig. 4).
To further characterize the e¡ect of the class 3 mutations on hilA expression, we transformed mutant strains
with the hilA: :lacZ reporter plasmid pLS31 and measured
L-galactosidase production. Interestingly, the patterns of
hilA: :lacZ expression in high and low oxygen di¡er in
Six Tn5 insertions a¡ecting hilA: :lacZ expression were
determined to reside outside of SPI-1 and were designated
class 3 mutations. The S. typhimurium genes disrupted by
these insertions were identi¢ed by using the DNA sequence £anking the transposon to search databases at
the National Center for Biotechnology Information
(NCBI) or the Sanger Centre S. typhi sequencing project
(www.sanger.ac.uk). Sequences identi¢ed in this manner
were examined for open reading frames (ORFs) disrupted
by the Tn5 using a DNA sequence analysis program. Homologs of each of the genes identi¢ed in S. typhi were
found in S. typhimurium by using the S. typhi sequence
to search the S. typhimurium genome database available
from the Washington University School of Medicine
(genome.wustl.edu/gsc/bacterial/salmonella.shtml).
One of the strains (TF47) was found to have Tn5 inserted at amino acid 194 of the S. typhimurium ams gene,
which, in E. coli, encodes the endoribonuclease RNase E
[34]. Another mutant (TF48) was found to have a Tn5
insertion 188 bp upstream of the putative translational
start of the hupB gene which encodes the L subunit of
the heterodimeric histone-like protein HU [35]. A third
insertion site (in TF49) was determined to be in a small
ORF of unknown function between the pagK and pagM
genes which we refer to here as pag [36]. Interestingly, the
remaining three class 3 mutations (in TF50, TF51, and
TF52) reside in a previously uncharacterized ORF that
we propose be termed hilE. Computer analysis of the
hilE gene predicts that it encodes a soluble protein but a
secondary structure analysis of the putative protein
showed no helix turn helix or coiled coil motifs.
3.6. The class 3 mutations upregulate chromosomal and
episomal hilA: :lacZ fusions
To quantitate the e¡ects of the class 3 mutations on
hilA : :lacZ expression, L-galactosidase assays were con-
Fig. 5. E¡ect of the class 3 mutations on expression of an episomal hilA
: :lacZ fusion. Strains were grown under high oxygen (white bars) or
low oxygen (dark bars) and assays were conducted to quantitate L-galactosidase activity (see Section 2). EE251 is the wild-type parent strain.
Plasmid pLS31 encodes a hilA : :lacZ fusion in the low copy vector
pRW50 [33]. Data presented represent the average of at least three independent experiments. Error bars depict the standard deviation from the
mean.
FEMSIM 1177 5-4-00
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
these mutants when the hilA: :lacZ reporter fusion is encoded on the plasmid rather than on the chromosome
(Fig. 5). Among these di¡erences is the loss of oxygen
regulation of hilA: :lacZ on the plasmid in the ams1 : :Tn5
background of TF53. Additionally, plasmid encoded hilA
: :lacZ in the hupB1: :Tn5 background is expressed at
levels above wild-type in high oxygen growth conditions while chromosomal hilA: :lacZ expression in the
hupB1 : :Tn5 background is repressed to levels similar to
that of wild-type (Fig. 4). Although pLS31 is a low copy
plasmid, the di¡erences seen in plasmid and chromosomal
hilA : :lacZ expression may be due to di¡erences in copy
number. Alternatively, the 500 bp of native DNA upstream of hilA present on pLS31 may not be su¤cient to
mimic the chromosomal context of this gene thus resulting
in the observed di¡erences between chromosomal and episomal hilA: :lacZ expression. It is clear from these data,
however, that the class 3 mutations exert a trans e¡ect on
hilA : :lacZ expression.
3.7. The class 3 mutants are hyperinvasive for HEp-2 cell
monolayers
To determine the e¡ect of the class 3 mutations on the
invasion phenotype, standard invasion assays of HEp-2
monolayers were conducted with strains grown under
high oxygen (non-inducing) or low oxygen (inducing) conditions. As shown in Fig. 6, members of this class show
moderate to dramatic increases in invasion of HEp-2 cells.
Most notable is the increase in invasion of TF53
(ams1: :Tn5) which is 33-fold more invasive under high
oxygen and 4.3-fold more invasive under low oxygen conditions compared to wild-type. This strain appears to be
unable to repress invasion under high oxygen conditions.
TF57 (hilE2 : :Tn5) and TF58 (hilE1 : :Tn5) show a 2^3fold increase in invasion under low oxygen while TF56
(hilE3 : :Tn5) does not show increased invasion compared
to wild-type. This phenotype coincides with our hilA : :lacZ
fusion data which show that hilA: :lacZ expression in the
hilE3 : :Tn5 mutant background is not a¡ected to the
degree of the two other mutations in the same ORF
(Fig. 4). The mutations in TF54 (hupB1: :Tn5) and TF55
(pag1 : :Tn5) each cause moderate (1.5-fold) increases in
invasion.
31
Fig. 6. E¡ect of the class 3 mutations on invasion of HEp-2 cell monolayers. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and invasion assays were conducted as described in Section 2. EE251 is the wild-type parent strain. Data presented represent at
least three independent experiments with duplicate wells. Error bars depict the standard deviation from the mean.
3.8. The class 3 mutants speci¢cally upregulate hilA
To test the hypothesis that the e¡ect of the class 3 mutations on hilA: :lacZ transcription is speci¢c to hilA : :
lacZ, we transduced each of these mutations into TF59,
an EE251 derivative with a pepT: :lacZ transcriptional fusion, and measured L-galactosidase production under high
and low oxygen conditions. The pepT gene has been characterized in Salmonella and encodes an endopeptidase
which is regulated at the transcriptional level by oxygen
[31] and is not involved in the invasion phenotype of Salmonella [25]. Although we observed that the ams1: :Tn5
mutation slightly reduced pepT: :lacZ expression in
TF60, there is no signi¢cant e¡ect of the class 3 mutations
on pepT: :lacZ expression (Table 2) and certainly no increase in pepT: :lacZ expression as observed hilA: :lacZ
reporter fusions (Fig. 4). This control experiment suggests
that the upregulation of hilA by the class 3 mutations is
speci¢c to hilA.
4. Discussion
As the regulation of Salmonella invasion gene expression is studied more intensively, it is apparent that a complex regulatory network involving a multitude of regulators is present to ensure that invasion occurs under
Table 2
E¡ect of class 3 mutations on expression of pepT: :lacZ
Strain
TF59 (wild-type)
TF60
TF61
TF62
TF63
TF64
TF65
L-Galactosidase activity
Induction ratio
High oxygen
Low oxygen
388.3 þ 29.5
100.1 þ 70.4
426.4 þ 21.4
378.6 þ 28.9
282.4 þ 16.3
366.1 þ 33.4
268.4 þ 9.87
959.8 þ 87.4
641.9 þ 22.3
1057.5 þ 124
979.2 þ 79.4
790.8 þ 20.5
895.6 þ 140.1
740.5 þ 24
FEMSIM 1177 5-4-00
2.47
6.41
2.48
2.59
2.80
2.45
2.76
32
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
optimal conditions, those which mimic the intestinal lumen of the host. At the center of invasion gene regulation
in S. typhimurium is hilA which encodes a transcriptional
activator homologous to the ompR/toxR family of transcriptional activators [19]. In an e¡ort to identify factors
involved in the complex regulation of hilA and Salmonella
invasion, we conducted a search for negative regulators of
hilA. We have isolated and characterized 18 Tn5 insertions
in the S. typhimurium chromosome that result in increased
expression of hilA on MacConkey agar. Twelve of these
insertions, which we have divided into classes 1 and 2,
de¢ne two loci within SPI-1 ; one which has a minor e¡ect
on hilA (class 1) and one which has a major e¡ect on hilA
(class 2). A third class of mutants, comprised of six isolates, identi¢es four loci outside of SPI-1 involved in hilA
regulation: ams, hupB, pag, and hilE.
Six of these mutations cluster in the prgH-hilD region of
SPI-1 and have been termed class 2 mutations (Fig. 1).
Each of these mutations upregulates chromosomal hilA: :
lacZ under low oxygen 2^6-fold, with the prgH7: :Tn5
mutation causing the greatest increase. Interestingly, we
noticed that two of the class 2 mutants, BJ606 and
TF40, increase expression of hilA: :lacZ after growth in
high and low oxygen conditions. This observation suggests
the presence of two regulatory regions near hilD, one involved in low oxygen expression and one involved in both
high and low oxygen expression of hilA.
Most of the class 2 insertions interrupt the invasion
gene prgH, which results in polar disruption of the invasion genes prgI, -J, -K and orgA ([32], and J. Klein, manuscript in preparation). In Yersinia, virulence proteins
(Yops) are secreted via a type III secretion system similar
to that of Salmonella and blockage of secretion of the Yop
proteins causes a feedback repression of Yop synthesis
[37]. By comparison, it is possible that the trans e¡ect of
class 2 mutations on hilA transcription is due to a lack of
production of the PrgH, -I, -J, -K, and OrgA proteins
which are required for secretion of invasion e¡ector proteins [21,32]. We tested invasion in these strains by complementation of the disrupted genes with plasmid pJK001
and observed a hyperinvasive phenotype. The plasmid
pJK001 restored high oxygen invasion in the mutants to
levels above wild-type yet did not alter the oxygen regulation of the wild-type strain. These experiments, along
with the isolation of a mutation which does not disrupt
prgH (prgH8: :Tn5), indicate that disruption of the prgH,
-I, -J, -K and orgA genes by the class 2 mutations is not
responsible for the upregulation of hilA. The class 2 mutations are in close proximity to hilD, a recently identi¢ed
derepressor of hilA that is 300 nucleotides upstream and
transcribed in the opposite orientation as prgH (Fig. 1B)
[33]. Therefore, a potential e¡ect of these mutations is to
increase hilD transcription which in turn would increase
hilA transcription. Consistent with this are data showing
that class 2 mutations increase hilA transcription in trans
(Fig. 2).
With the exception of TF43 and TF45, the e¡ect of the
class 2 mutants on the invasion phenotype appears to be
to reduce the ability of high oxygen conditions to repress
invasion (Fig. 3). Although low oxygen invasion was increased in these strains, the mutants displayed 10^100-fold
higher invasion compared to wild-type after growth in
high oxygen conditions. This `oxygen blind' invasion phenotype suggests that the class 2 mutations are impaired in
their ability to sense oxygen. However, hilA expression in
these strains, as measured with the lacZ reporter, is oxygen-regulated (Fig. 1). This may be explained as a phenomenon of invasion saturation. That is, e¡ects causing
increases above 0.01% invasion are detectable in these assays whereas those causing increases above 1% invasion
are not. Such a phenomenon would make strains overexpressing hilA under both high and low oxygen conditions
appear oxygen-insensitive in invasion assays. It seems clear
from these assays, however, that the class 2 mutations
allow increased invasion after high oxygen growth.
During the course of this work we noticed that although
the class 2 mutants showed upregulation of chromosomal
hilA: :lacZ only under low oxygen conditions, the invasion
phenotype of these mutants showed increases under high
oxygen conditions as well, suggesting high oxygen overexpression of hilA (Figs. 1 and 3). Also, strains containing
the plasmid reporter hilA: :lacZ showed increases in
hilA: :lacZ expression under both high and low oxygen
conditions (Fig. 2), an observation consistent with the invasion phenotype and inconsistent with the chromosomal
hilA: :lacZ reporter data. Perhaps a critical di¡erence between these strains is the fact that the chromosomal
hilA: :lacZ creates a hilA null mutation. Since the class 2
mutations in the chromosomal hilA: :lacZ reporter strains
do not cause overexpression of hilA whereas the mutations
in hilA‡ background apparently do, it is possible that
overexpression of hilA under high oxygen conditions requires hilA. This suggests autoregulation by HilA. This is
in contrast to work done by Bajaj et al. [14] who examined
autoregulation by hilA and found that hilA does not regulate its own expression. Therefore, the reason for the
di¡erence between high oxygen expression of hilA in hilA3
and hilA‡ backgrounds remains unclear.
Two of the class 2 mutants did not appear hyperinvasive
in our assays. One of these, TF43, exhibited wild-type
levels of invasion under both high and low oxygen conditions (Fig. 3). Although dissimilar to our other class 2
mutant invasion phenotypes, this phenotype is in agreement with data in Fig. 1 which show that this mutation
only slightly increases (less than 2-fold) hilA: :lacZ expression. The other strain, TF45, is less invasive than wildtype cells. This is inconsistent with data in Fig. 1 which
show the mutation in this strain causes a large increase in
hilA expression which should result in a hyperinvasive
phenotype [20]. Observations in our laboratory, and
others, suggest that strains that express hilA at very high
levels have growth defects. This situation selects for muta-
FEMSIM 1177 5-4-00
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
tions that downregulate the expression of the invasion
genes.
Six of the Tn5 insertions map outside of SPI-1 and are
designated class 3 mutations. One of these mutations disrupts amino acid 194 of the S. typhimurium ams gene and
is the ¢rst reported ams mutation in Salmonella. Although
this mutation caused a slight increase in expression of
chromosomal hilA: :lacZ (Fig. 4), both plasmid-encoded
hilA : :lacZ expression and invasiveness were signi¢cantly
increased in high oxygen growth conditions (Figs. 5 and
6). This suggests that the ams gene is involved in the repression of hilA under high oxygen conditions or, as discussed above, a functional hilA gene is required for high
oxygen over expression of hilA in an ams null mutant
background. In E. coli, ams encodes RNase E, an endoribonuclease with an N-terminal catalytic domain [38] that
cleaves A/U-rich RNA species [39]. The mutation that we
have identi¢ed disrupts this catalytic domain in the ams
gene of S. typhimurium. Interestingly, Schechter et al. [33]
have recently determined the transcriptional start site of
hilA and found that the message for the HilA protein has
a long 5P untranslated region which is A/U-rich. One possible role of RNase E in invasion is to degrade hilA
mRNA to modulate translation of the transcript. Another
class 3 mutation disrupts the regulatory region of the hupB
gene which leads to increased expression of chromosomal
and plasmid hilA: :lacZ reporter fusions (Figs. 4 and 5)
but does not signi¢cantly increase invasion (Fig. 6). In
E. coli, hupB encodes the L subunit of the histone-like
protein HU which binds to and condenses DNA [40^42].
The promoter region of hupB has been characterized extensively in E. coli [43,44] and by comparison to these
studies, our hupB1: :Tn5 mutation disrupts the crpcAMP binding site within the hupB promoter region. Since
a Tn5 insertion in this region disrupts the positive regulation of hupB by crp-cAMP as well as the negative regulation by FIS we currently do not know if the transposon
insertion increases or decreases normal expression of
hupB. The K subunit of the HU protein is encoded by
the hupA gene. Since HU protein can exist as a homoor heterodimer of K and L subunits, alterations in hupA
or hupB expression a¡ect the ratio of KK, KL, or LL dimers
present in the bacterium. Thus, it is possible that our
hupB1 : :Tn5 mutation alters the composition of the HU
dimer and thus a¡ects hilA expression.
A third locus identi¢ed by our screen for negative regulators of hilA is a small ORF between the phoP/Q-activated genes pagK and pagM [45] and downstream of the
macrophage-induced (mig) gene mig-3 [46] which we refer
to as pag. This mutation increases hilA: :lacZ expression
(Figs. 4 and 5) but does not cause a signi¢cant increase in
invasion (Fig. 6). Since phoP/Q is integral in regulating
invasion genes [32,47] as well as genes important for interacting with macrophages [36,48], this mutant is of great
interest. Future work will be aimed at analyzing this mutation in a phoQ constitutive background, which represses
33
hilA, to determine whether this ORF is involved in phoP/
Q-mediated repression of hilA.
Three of the class 3 Tn5 insertions disrupt a previously
unidenti¢ed ORF which we named hilE. This putative
gene is predicted to encode a protein which lacks homology to any known protein. Two of the insertions,
hilE1: :Tn5 and hilE2 : :Tn5, disrupt the 5P end of hilE
and presumably abolish the production of a functional
protein from this locus. Strains containing these mutations
have signi¢cantly increased hilA: :lacZ expression (Figs. 4
and 5) and invasion 2^3-fold higher under low oxygen
conditions (Fig. 6). This suggests that this ORF is involved in the modulation of hilA expression and the
invasion phenotype. A third mutation in this locus,
hilE3: :Tn5, is at the 3P end of the ORF and has a smaller
e¡ect on hilA: : lacZ expression and invasion (Figs. 4^6).
This suggests that in this mutant a protein is made that
retains some ability to repress hilA. These data show that
we have identi¢ed a novel ORF involved in hilA regulation
and expression of the invasive phenotype.
By isolating the class 3 mutations, we have identi¢ed
four new genes which appear to be involved in the regulation of hilA. However, these mutations are due to the
insertional inactivation of the genes by Tn5 which can
exert polar e¡ects on downstream genes. Future work
aimed at characterizing these genes will include determining whether the e¡ect of the transposon insertions on hilA
transcription is due to a polar e¡ect on downstream genes
or to disruption of the gene at the insertion site.
In this study we have searched for repressors of hilA by
transposon mutagenesis, identi¢ed several loci involved in
hilA regulation both on SPI-1 and elsewhere in the chromosome, and reported a preliminary characterization of
the identi¢ed loci. Future work to characterize these loci
will help to elucidate the complex regulatory scheme that
serves to ensure expression of invasion genes only under
conditions optimal for invasion.
Acknowledgements
We thank Lisa Schechter and Cathy Lee for the gift of
plasmid pLS31 and helpful discussions in the preparation
of the manuscript. T.F.F. is supported by a predoctoral
fellowship on NIH Training Grant 5T32 AI 07511. This
work was supported by National Institutes of Health
Grant AI 38268 and a grant from the Roy J. Carver
Charitable Trust (B.D.J.).
References
[1] Jones, B.D. and Falkow, S. (1996) Typhoid fever: Host immune
response and Salmonella virulence determinants. Annu. Rev. Immunol. 14, 533^561.
[2] Jones, B.D., Ghori, N. and Falkow, S. (1994) Salmonella typhimuri-
FEMSIM 1177 5-4-00
34
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
um initiates murine infection by penetrating and destroying the specialized epithelial M cells of the Peyer's patches. J. Exp. Med. 180,
15^23.
Clark, M.A., Jepson, M.A., Simmons, N.L. and Hirst, B.H. (1994)
Preferential interaction of Salmonella typhimurium with mouse Peyer's patch M cells. Res. Microbiol. 145, 543^552.
Jones, B., Pascopella, L. and Falkow, S. (1995) Entry of microbes
into the host: Using M cells to break the mucosal barrier. Curr.
Opin. Immunol. 7, 474^478.
Finlay, B.B. and Falkow, S. (1997) Common themes in microbial
pathogenicity revisited. Microbiol. Mol. Biol. Rev. 61, 136^169.
Darwin, K.H. and Miller, V.L. (1999) Molecular basis of the interaction of Salmonella with the intestinal mucosa. Clin. Microbiol. Rev.
12, 405^428.
Groisman, E.A. and Ochman, H. (1993) Cognate gene clusters govern invasion of host epithelial cells by Salmonella typhimurium and
Shigella £exneri. EMBO J. 12, 3779^3787.
Lee, C.A. (1997) Type III secretion systems: machines to deliver
bacterial proteins into eukaryotic cells? Trends Microbiol. 5, 148^
156.
Wood, M.W., Rosqvist, R., Mullan, P.B., Edwards, M.H. and
Galyov, E.E. (1996) SopE, a secreted protein of Salmonella dublin,
is translocated into the target eukaryotic cell via a sip-dependent
mechanism and promotes bacterial entry. Mol. Microbiol. 22, 327^
338.
Kaniga, K., Uralil, J., Bliska, J.B. and Galan, J.E. (1996) A secreted
protein tyrosine phosphatase with modular e¡ector domains in the
bacterial pathogen Salmonella typhimurium. Mol. Microbiol. 21, 633^
641.
Fu, Y. and Galan, J.E. (1998) The Salmonella typhimurium tyrosine
phosphatase SptP is translocated into host cells and disrupts the actin
cytoskeleton. Mol. Microbiol. 27, 359^368.
Hardt, W.D., Chen, L.M., Schuebel, K.E., Bustelo, X.R. and Galan,
J.E. (1998) S. typhimurium encodes an activator of Rho GTPases that
induces membrane ru¥ing and nuclear responses in host cells. Cell
93, 815^826.
Hong, K.H. and Miller, V.L. (1998) Identi¢cation of a novel Salmonella invasion locus homologous to Shigella ipgDE. J. Bacteriol. 180,
1793^1802.
Bajaj, V., Lucas, R.L., Hwang, C. and Lee, C.A. (1996) Co-ordinate
regulation of Salmonella typhimurium invasion genes by environmental and regulatory factors is mediated by control of hilA expression.
Mol. Microbiol. 22, 703^714.
Ernst, R.K., Dombroski, D.M. and Merrick, J.M. (1990) Anaerobiosis, type 1 ¢mbriae, and growth phase are factors that a¡ect invasion
of HEp-2 cells by Salmonella typhimurium. Infect. Immun. 58, 2014^
2016.
Galän, J.E. and Curtiss III, R. (1990) Expression of Salmonella typhimurium genes required for invasion is regulated by changes in DNA
supercoiling. Infect. Immun. 58, 1879^1885.
Lee, C.A. and Falkow, S. (1990) The ability of Salmonella to enter
mammalian cells is a¡ected by bacterial growth state. Proc. Natl.
Acad. Sci. USA 87, 4304^4308.
Schiemann, D.A. and Shope, S.R. (1991) Anaerobic growth of Salmonella typhimurium results in increased uptake by Henle 407 epithelial and mouse peritoneal cells in vitro and repression of a major
outer membrane protein. Infect. Immun. 59, 437^440.
Bajaj, V., Hwang, C. and Lee, C.A. (1995) hilA is a novel ompR/toxR
family member that activates the expression of Salmonella typhimurium invasion genes. Mol. Microbiol. 18, 715^727.
Lee, C.A., Jones, B.D. and Falkow, S. (1992) Identi¢cation of a
Salmonella typhimurium invasion locus by selection for hyperinvasive
mutants. Proc. Natl. Acad. Sci. USA 89, 1847^1851.
Penheiter, K.L., Mathur, N., Giles, D., Fahlen, T. and Jones, B.D.
(1997) Non-invasive Salmonella typhimurium mutants are avirulent
because of an inability to enter and destroy M cells of ileal Peyer's
patches. Mol. Microbiol. 24, 697^709.
[22] Bajaj, V., Hwang, C. and Lee, C.A. (1995) hilA is a novel ompR/toxR
family member that activates the expression of Salmonella typhimurium invasion genes. Mol. Microbiol. 18, 715^727.
[23] Eichelberg, K., Hardt, W.D. and Galan, J.E. (1999) Characterization
of SprA, an AraC-like transcriptional regulator encoded within the
Salmonella typhimurium pathogenicity island 1. Mol. Microbiol. 33,
139^152.
[24] Jones, B.D., Lee, C.A. and Falkow, S. (1992) Invasion by Salmonella
typhimurium is a¡ected by the direction of £agellar rotation. Infect.
Immun. 60, 2475^2480.
[25] Jones, B.D. and Falkow, S. (1994) Identi¢cation and characterization
of a Salmonella typhimurium oxygen-regulated gene required for bacterial internalization. Infect. Immun. 62, 3745^3752.
[26] Ling, M.M. and Robinson, B.H. (1997) Approaches to DNA mutagenesis: an overview. Anal. Biochem. 254, 157^178.
[27] Marmur, J. (1961) A procedure for the isolation of deoxyribonucleic
acid from micro-organisms. J. Mol. Biol. 3, 208^218.
[28] Davis, R.W., Botstein, D. and Roth, J.R. (1980) Advanced Bacterial
Genetics : A Manual for Genetic Engineering. Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY.
[29] Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring
Harbor Laboratory, Cold Spring Harbor, NY.
[30] Moore, A.E., Sabachewsky, L. and Toolan, H.W. (1955) Culture
characteristics of four permanent lines of human cancer cells. Cancer
Res. 15, 598.
[31] Strauch, K.L., Lenk, J.B., Gamble, B.L. and Miller, C.G. (1985)
Oxygen regulation in Salmonella typhimurium. J. Bacteriol. 161,
673^680.
[32] Pegues, D.A., Hantman, M.J., Behlau, I. and Miller, S.I. (1995)
PhoP/PhoQ transcriptional repression of Salmonella typhimurium invasion genes: evidence for a role in protein secretion. Mol. Microbiol.
17, 169^181.
[33] Schechter, L.M., Damrauer, S.M. and Lee, C.A. (1999) Two AraC/
XylS family members can independently counteract the e¡ect of repressing sequences upstream of the hilA promoter. Mol. Microbiol.
32, 629^642.
[34] McDowall, K.J. and Cohen, S.N. (1996) The N-terminal domain of
the rne gene product has RNase E activity and is non-overlapping
with the arginine-rich RNA-binding site. J. Mol. Biol. 255, 349^355.
[35] Marsh, M. and Hillyard, D.R. (1988) Nucleotide sequence of the
HU-1 gene of Salmonella typhimurium. Nucleic Acids Res. 16, 7196.
[36] Belden, W.J. and Miller, S.I. (1994) Further characterization of the
PhoP regulon: identi¢cation of new PhoP-activated virulence loci.
Infect. Immun. 62, 5095^6101.
[37] Allaoui, A., Schulte, R. and Cornelis, G.R. (1995) Mutational analysis of the Yersinia enterocolitica virC operon: characterization of
yscE, F, G, I, J, K required for Yop secretion and yscH encoding
YopR. Mol. Microbiol. 18, 343^355.
[38] Taraseviciene, L., Bjork, G.R. and Uhlin, B.E. (1995) Evidence for
an RNA binding region in the Escherichia coli processing endoribonuclease RNase E. J. Biol. Chem. 270, 26391^26398.
[39] McDowall, K.J., Lin, C.S. and Cohen, S.N. (1994) A+U content
rather than a particular nucleotide order determines the speci¢city
of RNase E cleavage. J. Biol. Chem. 269, 10790^10796.
[40] Rouviere, Y.J. and Gros, F. (1975) Characterization of a novel, lowmolecular-weight DNA-binding protein from Escherichia coli. Proc.
Natl. Acad. Sci. USA 72, 3428^3432.
[41] Rouviere, Y.J., Yaniv, M. and Germond, J.E. (1979) E. coli DNA
binding protein HU forms nucleosomelike structure with circular
double-stranded DNA. Cell 17, 265^274.
[42] Kano, Y., Wada, M., Nagase, T. and Imamoto, F. (1986) Genetic
characterization of the gene hupB encoding the HU-1 protein of Escherichia coli. Gene 45, 37^44.
[43] Kohno, K., Wada, M., Kano, Y. and Imamoto, F. (1990) Promoters
and autogenous control of the Escherichia coli hupA and hupB genes.
J. Mol. Biol. 213, 27^36.
[44] Claret, L. and Rouviere, Y.J. (1996) Regulation of HU alpha and
FEMSIM 1177 5-4-00
T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35
HU beta by CRP and FIS in Escherichia coli. J. Mol. Biol. 263, 126^
139.
[45] Gunn, J.S., Belden, W.J. and Miller, S.I. (1998) Identi¢cation of
PhoP-PhoQ activated genes within a duplicated region of the Salmonella typhimurium chromosome. Microb. Pathog. 25, 77^90.
[46] Valdivia, R.H. and Falkow, S. (1997) Fluorescence-based isolation of
bacterial genes expressed within host cells. Science 277, 2007^2011.
[47] Behlau, I. and Miller, S.I. (1993) A PhoP-repressed gene promotes
Salmonella typhimurium invasion of epithelial cells. J. Bacteriol. 175,
4475^4484.
35
[48] Miller, S.I., Kukral, A.M. and Mekalanos, J.J. (1989) A two-component regulatory system (phoP/phoQ) controls Salmonella typhimurium
virulence. Proc. Natl. Acad. Sci. USA 86, 5054^5058.
[49] Wray, C. and Sojka, W.J. (1978) Experimental Salmonella typhimurium in calves. Res. Vet. Sci. 25, 139^143.
[50] Stibitz, S., Black, W. and Falkow, S. (1986) The construction of a
cloning vector designed for gene replacement in Bordetella pertussis.
Gene 50, 133^140.
FEMSIM 1177 5-4-00