• mmi_4920.fm

Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382X© 2005 The Authors; Journal compilation © 2005 Blackwell Publishing Ltd 200558513541367Original ArticleRegulation of acid resistance by FurA. G. Oglesby, E. R. Mur-
?
phy, V. R. Iyer and S. M. Payne
Molecular Microbiology (2005) 58(5), 1354–1367 doi:10.1111/j.1365-2958.2005.04920.x
First published online 24 October 2005
Fur regulates acid resistance in Shigella flexneri via
RyhB and ydeP
Amanda G. Oglesby, Erin R. Murphy, Introduction
Vishwanath R. Iyer and Shelley M. Payne*
Shigella flexneri, a causative agent of bacillary dysentery,
Section of Molecular Genetics and Microbiology, The
initiates disease by invading colonic epithelial cells, mul-
University of Texas at Austin, Austin, TX 78712, USA.
tiplying intracellularly, and spreading to adjacent epithelial
cells (Jennison and Verma, 2004). During infection, S.
Summary flexneri encounters different environmental conditions and
must be able to respond efficiently to each of these in
Shigella flexneri requires iron for survival, and the
order to survive and cause disease. In particular, S. flex-
genes for iron uptake and homeostasis are regulated
neri can survive the acidic conditions it encounters in the
by the Fur protein. Microarrays were used to identify
stomach, an ability that may contribute to its low infectious
genes regulated by Fur and to study the physiological
dose (10–500 organisms) (DuPont et al., 1989). While S.
effects of iron availability in S. flexneri. These assays
flexneri will not grow at a pH below 4.8 (Lin et al., 1995),
showed that the expression of genes involved in iron
this pathogen can survive for several hours at pH 2.5, a
acquisition and acid response was induced by low-
property termed acid resistance (Gorden and Small,
iron availability and by inactivation of fur. A fur null
1993).
mutant was acid sensitive in media at pH 2.5, and acid
S. flexneri possesses two acid resistance systems. The
sensitivity was also observed in the wild-type strain
first is dependent upon the presence of glutamate in the
grown under iron-limiting conditions. Acid resistance
medium during acid challenge (Lin et al., 1995). This sys-
of the fur mutant in minimal medium was restored by
tem requires the expression of glutamate decarboxylase,
addition of glutamate during acid challenge, indicat-
as well as an antiporter which imports glutamate and
ing that the glutamate-dependent acid resistance sys-
exports its decarboxylated product, gamma-aminobu-
tem was not defective. Inactivation of ryhB, a small
tyrate (Hersh et al., 1996; Waterman and Small, 1996; De
regulatory RNA whose expression is repressed by
Biase et al., 1999). The decarboxylation of glutamate is
Fur, restored acid resistance in the fur mutant, while
thought to protect bacteria by increasing the internal pH
overexpressing ryhB increased acid sensitivity in the
through the consumption of intracellular protons. In S.
wild-type strain. RyhB-regulated genes were identi-
flexneri, gadA and gadB encode glutamate decarboxylase
fied by microarray analysis. The expression of one of
isozymes, and the glutamate/GABA antiporter is encoded
the RyhB-repressed genes, ydeP, which encodes a
by gadC (Waterman and Small, 1996). Two similar sys-
putative oxidoreductase, suppressed acid sensitivity
tems that use arginine or lysine instead of glutamate are
in the fur mutant. Furthermore, an S. flexneri ydeP
also present in Escherichia coli (Park et al., 1996), but
mutant was defective for both glutamate-independent
homologous genes are not present in the S. flexneri
and glutamate-dependent acid resistance. The repres-
genome (Jin et al., 2002; Wei et al., 2003).
sion of ydeP by RyhB may be indirect, as real time
The regulation of glutamate-dependent acid resistance
polymerase chain reaction (PCR) experiments indi-
has been described in E. coli. This system is induced in
cated that RyhB negatively regulates evgA, which
logarithmically growing cells exposed to low pH, or in cells
encodes an activator of ydeP. These results demon-
grown to stationary phase. The EvgA response regulator
strate that the acid sensitivity defect of the S. flexneri
activates glutamate-dependent acid resistance during
fur mutant is due to repression of ydeP by RyhB, most
exponential growth at low pH (Masuda and Church, 2002;
likely via repression of evgA.
2003; Nishino et al., 2003; Ma et al., 2004), whereas
RpoS induces expression of this system during stationary
phase (Castanie-Cornet et al., 1999; De Biase et al.,
1999).
The second acid resistance system is glutamate-
Accepted 16 September, 2005. *For correspondence. E-mail
payne@mail.utexas.edu; Tel. (+1) 512 471 9258; Fax (+1) 512 independent and is less well characterized than the
471 7088. glutamate-dependent acid resistance system. This
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd
 Regulation of acid resistance by Fur 1355
glutamate-independent acid-resistance system protects be regulated by iron were also identified in this analysis,
cells in the absence of glutamate at pH 2.5 and is induced several of which are members of the acid regulon in
by RpoS in complex medium during stationary phase (Lin E. coli (Tucker et al., 2002). This report describes the role
et al., 1995). Overexpression of evgA also induces this of Fur and RyhB in the regulation of acid resistance in
system in exponential phase cells (Masuda and Church, S. flexneri.
2002). This system has also been called the oxidative acid
resistance system, as its function is repressed by fermen-
tative growth in media containing glucose and its expres- Results
sion requires the cAMP receptor protein (CRP) (Lin et al.,
Acid response genes are induced in response to iron
1995; Castanie-Cornet et al., 1999).
depletion and fur inactivation in S. flexneri
S. flexneri encounters iron-restricted, as well as acidic,
environments in the host. Most iron in the host is seques- To study the physiological effects of decreased iron avail-
tered in proteins such as lactoferrin, transferrin, and ability in S. flexneri, microarrays were used to identify iron-
haemoglobin and is not readily available to bacterial and Fur-regulated genes. We compared global gene
pathogens. S. flexneri expresses several iron acquisition expression of wild-type S. flexneri grown in high- and low-
systems in iron-restricted environments in vivo and in vitro iron conditions, as well as the gene expression profiles of
(Payne and Mey, 2004). Iron restriction limits bacterial the wild type and fur mutant grown in high iron. Selected
growth, but high intracellular concentrations of iron can be genes whose expression was induced in low iron and in
harmful due to its ability to accelerate the formation of the fur mutant are shown in Table 1.
oxygen radicals in aerobically respiring bacteria. The As expected, known iron acquisition genes were
acquisition of iron is therefore tightly regulated in response induced in cells grown in the presence of the iron chelator
to the intracellular iron concentration. Iron homeostasis is EDDA compared with those grown in iron-replete condi-
largely achieved through the action of Fur (ferric uptake tions, and these genes were expressed at a higher level
repressor), a 17 kDa iron-binding repressor protein in the fur mutant compared with the wild-type strain.
(Hantke, 2001). Under iron-replete conditions, the Fur pro- These iron acquisition genes have previously been shown
tein becomes ferrated and binds to a 19 bp sequence to be repressed by iron via the Fur protein. Among these
called the Fur box in the promoters of iron acquisition genes are the iucA-D and iutA genes required for aero-
genes, thereby repressing their expression (Bagg and bactin synthesis and uptake (Payne et al., 1983; de
Neilands, 1987; Escolar et al., 1999; Baichoo and Hel- Lorenzo et al., 1987), the sit and feo genes involved in
mann, 2002). In iron-depleted conditions, the apo-Fur pro- ferrous iron uptake (McHugh et al., 2003; Runyen-
tein is inactive, and transcription of iron acquisition genes Janecky et al., 2003) and fhuE, fhuF and exbB (McHugh
proceeds. et al., 2003).
In E. coli, Fur also represses the expression of ryhB, Several genes that are induced in E. coli by growth in
which encodes a small regulatory RNA that blocks gene acidic conditions (Tucker et al., 2002) also responded to
expression by binding to and destabilizing target mRNAs iron limitation and fur inactivation in S. flexneri (Table 1).
(Masse and Gottesman, 2002; Masse et al., 2003). gadB and gadC, encoding glutamate decarboxylase and
Degradation of the RyhB-mRNA complex is RNaseE- the glutamate/GABA antiporter respectively, were both
dependent (Masse et al., 2003). RyhB stability and mRNA induced upon iron depletion and fur inactivation, as was
binding are promoted by the small RNA chaperone protein hdeA, encoding a putative periplasmic acid resistance
Hfq (Masse et al., 2003). In E. coli, RyhB contributes to chaperone. A gene encoding a central activator of acid
iron homeostasis by repressing the production of iron- resistance, gadE (yhiE) (Ma et al., 2003) was also more
containing proteins during growth in iron-depleted condi- highly expressed during growth in the presence of the iron
tions. These RyhB-regulated genes include sdhCDAB chelator (low iron) than in high iron and in the fur mutant
(succinate dehydrogenase), sodB (Fe-superoxide dismu- as compared with wild type. Although GadE induces
tase), acnA (aconitase), fumA (fumarase) and ftn (ferritin) expression gadA, as well as gadB and C in E. coli
(Masse and Gottesman, 2002). The sequence of the S. (Castanie-Cornet and Foster, 2001), gadA induction in
flexneri ryhB gene is identical to that of E. coli K-12 (Wei response to low iron or loss of Fur was not observed in
et al., 2003), and the encoded small RNA likely serves these experiments.
similar regulatory functions. There were differences in the extent of induction of
In this study, we analysed the physiological effects of the acid response genes when comparing individual
iron regulation by identifying targets of Fur regulation in microarray experiments, and this variation was greater
S. flexneri using microarrays. As expected, the iron acqui- than that observed for the iron acquisition genes. In
sition genes were induced during growth in iron-limiting some experiments, the acid resistance genes were
conditions and in a fur mutant. Other genes not known to induced 10- to 20-fold while in other experiments, the
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 58, 1354–1367
1356 A. G. Oglesby, E. R. Murphy, V. R. Iyer and S. M. Payne
Table 1. Iron acquisition and stress response genes induced by iron depletion or fur inactivation during exponential growth.a
Fold inductionb in:
Gene name UID Description of gene product low iron (+ EDDA) fur mutant
Iron metabolism genes
iutA IUTA Aerobactin outer membrane receptord 11.6 ± 3.0 20.6 ± 11.5
iucA IUCA Aerobactin synthesisd 3.3 ± 0.9 4.6 ± 2.1
iucB IUCB Aerobactin synthesisd 9.4 ± 1.9 24.3 ± 8.2
iucC IUCC Aerobactin synthesisd 8.3 ± 2.4 28.6 ± 22.4
fhuE B1102 Coprogen outer membrane receptore,f 5.9 ± 2.9 7.2 ± 5.4
fhuF B4367 Hydroxamate reductasef,g 12.2 ± 3.3 13.0 ± 7.0
exbB B3006 TonB accessory proteinf,h 6.0 ± 1.1 6.5 ± 3.9
feoA B3408 Ferrous iron uptakef,i 2.4 ± 0.6 5.1 ± 3.8
sitA SITA Ferrous iron uptakej 10.9 ± 2.3 4.0 ± 3.3
sitB SITB Ferrous iron uptakej 10.0 ± 3.3 3.7 ± 3.8
sitC SITC Ferrous iron uptakej 5.8 ± 1.8 2.3 ± 1.6
sitD SITD Ferrous iron uptakej 8.7 ± 1.7 2.3 ± 0.6
ryhB N/A Fur-repressed small RNA ND 68c
Acid response genes
gadC B1492 Glutamate/GABA antiporterk 9.7 ± 10.0 7.0 ± 8.8
gadB B1493 Glutamate decarboxylasek 10.3 ± 10.8 6.1 ± 7.3
hdeA B3510 Putative acid resistance chaperonel 8.3 ± 8.8 6.3 ± 8.5
gadE B3512 Regulator of acid response genesm 6.2 ± 6.1 3.4 ± 3.6
Oxidative stress genes
dps B0812 DNA protection during starvationn 4.4 ± 4.2 3.4 ± 3.9
sufB B1683 Fe-S cluster assemblyo 2.7 ± 1.9 1.9 ± 0.8
sufC B1682 Fe-S cluster assemblyo 2.2 ± 1.6 1.4 ± 0.8
sufD B1681 Fe-S cluster assemblyo 2.4 ± 1.7 1.7 ± 0.8
sufE B1679 Fe-S cluster assemblyo 2.9 ± 1.5 1.9 ± 0.9
sufS B1680 Fe-S cluster assemblyo 3.0 ± 2.3 1.7 ± 0.8
a. Strains were grown to mid-logarithmic phase at 37°C in L broth. Experiments were performed at least three times using PCR-amplified
microarrays as described in Experimental procedures.
b. Fold induction indicates the ratio of hybridized Cy5-labelled cDNA (fur or wild type, + EDDA) to Cy3-labelled cDNA (wild type or wild type, no
EDDA) to the indicated spot. ND, not determined.
c. Determined by real time PCR analysis.
d. (Payne et al., 1983; Bagg and Neilands, 1987).
e. (Hantke, 1983).
f. (McHugh et al., 2003).
g. (Matzanke et al., 2004).
h. (Fischer et al., 1989).
i. (Kammler et al., 1993).
j. (Runyen-Janecky et al., 2003).
k. (Hersh et al., 1996).
l. (Gajiwala and Burley, 2000).
m. (Hommais et al., 2004).
n. (Martinez and Kolter, 1997).
o. (Outten et al., 2004).
induction of these genes was less than twofold (data not The fur mutant does not acidify the culture medium more
shown). In contrast, the genes encoding iron transport than wild type
systems were consistently induced more than twofold in
the low-iron cultures and in the fur mutant. These differ- A possible explanation for the observed induction of acid
ences in induction of the acid resistance genes are response genes in the fur mutant is acidification of the
therefore likely due to subtle variations in culture condi- medium caused by altered expression of metabolic path-
tions unrelated to changes in iron availability. This same ways. To test this, the pH of the growth medium was
pattern of variation was seen with another group of iron determined for both the wild type and the fur mutant
and Fur-regulated genes, the suf and dps genes (Table 1 grown under the same conditions as used for the
and Fig. 1). The suf genes are induced both by low iron microarray experiments (Fig. 2). The pH of the culture
and by oxidative stress (Outten et al., 2004). Thus the medium for both strains remained above 6.5 throughout
extent of induction of the acid resistance genes by low the growth period. As the induction of acid response
iron or in the fur mutant may depend on oxidative stress genes during logarithmic growth requires an external pH
or other environmental conditions in addition to loss of of 5.5 or lower (Castanie-Cornet and Foster, 2001), it
iron or loss of Fur. appears that induction of the acid response genes in the
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 58, 1354–1367
 Regulation of acid resistance by Fur 1357
concentration of glutamate in the complex medium (Luria
-/+ Iron
-/+ Iron
-/+ Iron
-/+ Iron
-/+ Fur
-/+ Fur
broth [L broth]) was unknown, the assays were repeated
in minimal medium with or without glutamate supplemen-
tation to determine if the acid resistance defect was spe-
iucA cific to either the glutamate-dependent or glutamate-
fhuE independent mechanism. As was observed in complex
iutA medium, only 1% of fur mutant cells survived after 2 h
fhuF incubation in M9 minimal medium at pH 2.5 (Fig. 3A, open
exbB bars). This phenotype was suppressed by the addition of
iucB glutamate to the medium, indicating that the fur mutant is
iucC defective for glutamate-independent acid resistance
feoA (Fig. 3A, closed bars). As the induction of acid response
sitB genes was also observed in wild type S. flexneri grown in
sitA low iron, the effect of iron limitation on acid resistance in
sitC the wild type was tested. Wild-type S. flexneri grown prior
sitD to acid shock in medium containing the iron chelator
gadC EDDA to reduce iron availability displayed the same acid
gadB sensitivity phenotype as the fur mutant in both complex
hdeA (data not shown) and minimal media (Fig. 3A). The acid-
gadE sensitive phenotype of the fur mutant was complemented
dps by expression of fur from a plasmid (Fig. 3B). Overall,
sufC these results indicate that both iron and Fur are required
sufB for glutamate-independent acid resistance in S. flexneri.
sufD
sufS
sufE Acid sensitivity in the fur mutant is due to induction of ryhB
ftn Although the expression of the acid resistance genes was
sodB regulated by iron and Fur, we did not observe an apparent
tdcA Fur box in the promoters of the acid-response genes or
ydeP their identified regulators. This suggested that the effect
of Fur on the acid-response genes was indirect. One
Fig. 1. Expression of Shigella genes in response to iron limitation
and loss of Fur. Expression data from the microarray experiments mechanism for indirect regulation by Fur is via RyhB, a
described in Table 1 were used for hierarchical clustering as Fur-repressed small RNA that regulates a number of
described in Experimental procedures and displayed in an expression
map. Each column represents a single experiment and each row a
single gene. The brightness of each square indicates the ratio of
expression in either low-iron compared with high-iron conditions 7.5
(–/+Fe) or in the fur mutant compared with wild type (–/+Fur). Red Wild type
indicates induction while green indicates repression of that gene. fur
Grey indicates that data for that gene were unavailable for analysis.
7.0
fur mutant was not a result of acidification of the culture
pH
medium.
6.5
The fur mutant is defective for glutamate-independent acid
resistance
6.0
The induction of the acid response genes in cells grown
0 0.2 0.4 0.6 0.8 1
in low iron and in the fur mutant suggested that the cells
were experiencing acid stress and that acid resistance OD650
could be affected in these cells. Therefore, we determined Fig. 2. Change in culture pH of wild type and fur cultures. Wild type
the acid resistance phenotype of the fur mutant. Following (SA101-diamonds) and fur (SA211-squares) were diluted 1:100 into
incubation in complex medium at pH 2.5, less than 1% of L broth from overnight cultures in L broth, and were grown in batch
culture at 37°C. pH and growth readings were taken every 30 min until
the fur mutant cells survived, while greater than 10% of strains reached late-logarithmic growth phase. Displayed are results
wild-type cells were viable (data not shown). Because the of a representative experiment.
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 58, 1354–1367
1358 A. G. Oglesby, E. R. Murphy, V. R. Iyer and S. M. Payne
The data suggest that RyhB represses glutamate-inde-
A 1000 M9 pendent acid resistance in S. flexneri. To test this directly,
M9 + Glutamate wild-type S. flexneri was transformed with a plasmid car-
rying ryhB under the control of an IPTG-inducible pro-
100
moter. This strain was resistant to acid challenge in the
% Survival
presence of glutamate, but exhibited acid sensitivity in the
10 absence of glutamate, proportional to the concentration
of IPTG in the growth medium (Fig. 4B). As expected,
ryhB expression was proportional to the concentrations
1 of IPTG, as measured by real time PCR (Fig. 4B).
Thus, overexpression of ryhB suppressed glutamate-
0.1
Wild Type fur Wild Type fur
no EDDA + EDDA A 1000
M9
M9 + Glutamate
B
100
M9 100
% Survival
M9 + Glutamate
% Survival
10
10
1
Wild type fur fur ryhB ryhB
Relative
1 ryhB 1 4 0 0
WT / fur / fur / expression
Vector Vector Fur+
B 1000
Fig. 3. Iron and Fur are required for glutamate-independent acid
M9
resistance.
M9 + Glutamate
A. Wild type (SA100) and fur (SA1301), grown with and without EDDA 100
as indicated, were incubated in M9, pH 2.5 without (open bars) or with
(closed bars) 0.012% glutamate for 2 h.
% Survival
B. Wild type (SM100) or the fur mutant (SM1301) carrying pACYC184
10
(Vector) or pFS2 (Fur+), as indicated, were incubated for 2 h in M9,
pH 2.5 without (open bars) or with (closed bars) glutamate 1
supplementation.
Per cent survival was determined as described in the Experimental
procedures. Data shown are averages of three independent experi- 0.1
ments. Error bars indicate one standard deviation.
0.01
- IPTG + 100 µM - IPTG + 25 µM + 50 µM + 100 µM
IPTG IPTG IPTG IPTG
Vector Inducible RyhB
genes in E. coli (Masse and Gottesman, 2002). Real time
polymerase chain reaction (PCR) confirmed that ryhB was
Relative ryhB expression: 1 3 4 5
induced under the same conditions that caused induction
of the acid-resistance genes (Table 1) and in the growth Fig. 4. RyhB represses glutamate-independent acid resistance.
conditions used for acid sensitivity assays (Fig. 4). To A. Wild type (SM100), fur (SM1301), fur ryhB (SM1302) and ryhB
(SM100; ryhB::cam) were incubated in M9, pH 2.5 without (open bars)
determine whether RyhB was influencing acid resistance, or with (closed bars) glutamate supplementation for 2 h.
a ryhB deletion mutant and a fur ryhB double mutant were B. Wild-type S. flexneri (SA100) carrying pQE-2 (Vector) or pERM111
constructed. While the fur mutant was sensitive to acid (RyhB+) were incubated in M9, pH 2.5 without (open bars) or with
(closed bars) glutamate supplementation for 2 h in the presence of
challenge, the ryhB mutation suppressed acid sensitivity the indicated concentration of IPTG.
in the fur mutant, with nearly 100% of the fur ryhB cells Per cent survival was determined as described in the Experimental
surviving acid challenge in the absence of glutamate procedures. Data shown are averages of three independent experi-
ments. Error bars indicate one standard deviation. Numbers below
(Fig. 4A). The ryhB mutation alone had no effect on acid each graph represent the relative amount of ryhB expression as
resistance (Fig. 4A). determined by real time PCR.
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 58, 1354–1367
 Regulation of acid resistance by Fur 1359
independent acid resistance in S. flexneri. Taken together, A
these results suggest that RyhB represses one or more