Regulation of Iron Uptake by Fine-Tuning the Iron Responsiveness of the Iron Sensor Fur

Iron is a fundamental metal ion for living organisms as it facilitates various biological processes. The ferric uptake regulator (Fur) protein controls iron homeostasis in various bacterial species. It is believed that Fur’s iron-dependent regulatory action is sufficient for it to function as an iron sensor. However, we now establish that the bacterial pathogen Salmonella enables Fur to properly reflect changes in surrounding iron availability by fine-tuning its responsiveness to iron. This process requires a protein that hampers Fur DNA binding at low iron concentrations. In this way, Salmonella broadens the range of iron concentrations that Fur responds to. Our findings reveal a potentially widespread control mechanism of bacterial iron homeostasis.

ABSTRACT

Iron is one of most abundant environmental metal ions but is highly limited in organisms. It is an important metal ion as it facilitates various biological processes, including catalysis of metabolic enzymes and DNA biogenesis. In bacteria, the ferric uptake regulator (Fur) protein controls iron uptake by regulating genes coding for iron transporters in response to iron concentration. This iron response is ascribed to Fur’s intrinsic affinity for iron because its binding to iron dictates its regulatory function. However, we now report that the pathogen Salmonella achieves a proper response of Fur to changes in environmental iron concentrations via EIIA^Ntr (a nitrogen metabolic phosphotransferase system component). We establish that EIIA^Ntr increases expression of iron transporter-coding genes under low-iron conditions (i.e., nanomolar ranges) in a Fur-dependent manner, which promotes Salmonella growth under such conditions. EIIA^Ntr directly hampers Fur binding to DNA, thereby inducing expression of those genes. This regulation allows Salmonella to express Fur-regulated genes under low-iron conditions. Our findings reveal a potentially widespread control mechanism of bacterial iron uptake systems operating in response to iron availability.

IMPORTANCE Iron is a fundamental metal ion for living organisms as it facilitates various biological processes. The ferric uptake regulator (Fur) protein controls iron homeostasis in various bacterial species. It is believed that Fur’s iron-dependent regulatory action is sufficient for it to function as an iron sensor. However, we now establish that the bacterial pathogen Salmonella enables Fur to properly reflect changes in surrounding iron availability by fine-tuning its responsiveness to iron. This process requires a protein that hampers Fur DNA binding at low iron concentrations. In this way, Salmonella broadens the range of iron concentrations that Fur responds to. Our findings reveal a potentially widespread control mechanism of bacterial iron homeostasis.

INTRODUCTION

Iron is an abundant metal on Earth but is very limited in organisms due to its poor solubility (1). Because this metal is an essential cofactor for many biological processes, including reactions catalyzed by metabolic enzymes and DNA biogenesis (2), iron homeostasis is important for living organisms (1, 2). Not surprisingly, mammalian hosts and bacterial pathogens compete for this limiting metal ion during infection (3, 4). Therefore, it is important for bacteria to properly modulate their iron acquisition system in response to changes in iron availability, especially when the available iron concentration is low. Here, we report how a microorganism tunes the responsiveness of an iron sensor to properly control the iron uptake system.

The transcription factor ferric uptake regulator (Fur) functions as an iron sensor and plays a leading role in maintaining iron homeostasis in various bacterial species by controlling iron transporter-coding genes (5, 6). Fur represses expression of iron transporter genes under high-iron conditions where iron-bound Fur binds to its target promoters, thereby repressing gene transcription (5–7). It is reported that external iron concentrations below 5 to 10 µM cause dissociation of Fur from its targets, allowing expression of iron transporter-coding genes (5). Fur influences the expression of a variety of genes, including those that participate in iron acquisition and virulence (5, 8).

The ptsN gene encodes EIIA^Ntr, a component of the nitrogen-metabolic phosphotransferase system (PTS) (9). This system lacks the membrane-bound complex that would normally control the activities of sugar PTSs in response to particular sugar availabilities (9). EIIA^Ntr plays regulatory functions in various bacterial species by interacting with proteins involved in a variety of cellular processes. EIIA^Ntr controls potassium uptake via TrkA and KdpD (10, 11), phosphate uptake via PhoR (12), virulence via SsrB (13), the stringent response via SpoT (14, 15), and amino sugar homeostasis via GlmS (16). The phosphorylation status of EIIA^Ntr contributes some of those EIIA^Ntr-mediated regulatory functions (10, 11, 15, 16). The phosphorylation of EIIA^Ntr is known to be controlled by the extracellular abundance of nitrogen sources or by the cellular concentration of glutamine (15, 17).

The intracellular pathogen Salmonella enterica serovar Typhimurium is the etiologic agent of human gastroenteritis and murine typhoid fever. Salmonella resides inside macrophage phagosomes (18) where iron availability is limited due to the iron transporters being recruited by the mammalian host pumping out iron from those vesicles (19). However, the phagosome does not represent an iron-depleted condition given that iron-responding systems are activated in that compartment (20–22). Lack of the iron sensor Fur attenuates Salmonella virulence (23), suggesting that Salmonella must manage the low iron availability inside the phagosome and that the iron-sensing ability of Fur is critical for its virulence. Fur’s iron-sensing function is ascribed to its intrinsic affinity to iron (6, 24) given that iron binding dictates its DNA binding ability (6). The affinity of Fur for iron is in the low micromolar range (6, 25), which is believed sufficient for Fur to operate as an iron sensor.

Here, we now report that Fur’s response to iron requires EIIA^Ntr in addition to its own ability to sense iron. EIIA^Ntr tunes the responsiveness of Fur to iron by hampering Fur binding to DNA. This allows expression of the iron uptake system when surrounding iron concentrations drop to the nanomolar range (hereafter, low-iron conditions), which enables the intracellular pathogen Salmonella enterica to properly control its iron uptake system in response to iron availability.

RESULTS

Expression of iron uptake genes requires EIIA^Ntr under low-iron conditions.

A recent proteomic study reported that the Escherichia coli ptsN mutant reduces the abundance of proteins involved in iron uptake system (26), which leads us to question whether EIIA^Ntr has a role in iron response. To test this idea, we first examined expression of genes involved in iron importation (including fhuA, fepA, and iroB) in the wild-type strain and the ptsN mutant grown in acidified defined media to mimic the experience of Salmonella in acidic phagosomes, where iron uptake is important (18, 19). The wild-type strain displayed higher mRNA abundance of those genes than the ptsN mutant (Fig. 1A), indicating that EIIA^Ntr is involved in controlling iron response genes. We further investigated the expression of the fhuA gene by measuring β-galactosidase activity produced by Salmonella strains with a p_fhuA-lacZ fusion from its normal chromosomal location; the wild-type strain had ∼9-fold-higher activities than the ptsN mutant (Fig. 1B). Moreover, the defective fhuA expression of the ptsN mutant was due to a lack of EIIA^Ntr protein because a plasmid expressing EIIA^Ntr from a heterologous promoter restored fhuA expression to wild-type levels, but the empty vector did not (Fig. 1B).

FIG 1.

EIIA^Ntr G expression of iron uptake genes under low-iron conditions. (A) mRNA abundance of fhuA, fepA, and iroB genes was determined in the wild-type (WT) and the ptsN mutant strains grown in acidified M9 medium (pH 5.8) to mid-log phase. (B and C) β-Galactosidase activities of Salmonella with a p_fhuA-lacZ fusion in the normal chromosomal location and isogenic ptsN mutants with denoted plasmids (empty vector [pVec] or plasmid expressing EIIA^Ntr from a heterologous promoter [pPtsN]) were determined. Bacteria were grown to mid-log phase in acidified M9 medium (pH 5.8) with 8 nM FeSO₄ and IPTG at the denoted concentrations (B) or in acidified M9 medium (pH 5.8) with FeSO₄ at the denoted concentrations (C). The means and standard deviations (SD) of results from at least three independent experiments are shown as follows: symbols or bars, mean values; error bars, SD. Two-tailed t tests were performed for comparisons between the wild-type strain and the ptsN mutant or between indicated strains, and statistical significance is indicated as follows: ns, not significant; *, P < 0.05; **, P < 0.01; ****, P < 0.0001.

We next investigated fhuA expression under conditions of various iron concentrations. Wild-type Salmonella induced an ∼18-fold increase in fhuA expression when iron concentrations dropped from 100 µM to 8 nM (Fig. 1C). Surprisingly, however, the ptsN mutant failed to do so under conditions of the same changes in iron concentrations even though it has the iron sensor Fur (Fig. 1C); the ptsN mutant displayed only an approximately 2-fold increase in fhuA expression (Fig. 1C). This EIIA^Ntr-dependent regulation of fhuA expression was evident when iron concentrations were lower than 1 µM (Fig. 1C). These results suggest that Salmonella EIIA^Ntr controls the iron uptake genes under low-iron conditions.

EIIA^Ntr controls iron uptake gene expression in a Fur-dependent manner.

Given that expression of those iron uptake genes (including fhuA) is repressed by Fur (24, 27), we hypothesized that EIIA^Ntr controls the iron uptake gene expression via Fur. Consistent with the previous notion that Fur represses its target genes under iron-replete conditions (5–7), a deletion of the fur gene did not alter fhuA expression under low-iron conditions. Lack of Fur abrogated the regulatory effects of EIIA^Ntr on fhuA expression (Fig. 2A), indicating that Fur is necessary for EIIA^Ntr-mediated regulation of fhuA expression. If EIIA^Ntr regulates fhuA expression by altering Fur expression (i.e., if changes in Fur expression could in turn control expression of fhuA gene), heterologous EIIA^Ntr-independent expression of Fur should abolish regulatory effects of EIIA^Ntr on fhuA expression. However, lack of EIIA^Ntr reduced fhuA transcription even when Fur was expressed from a heterologous promoter (Fig. 2A), indicating that EIIA^Ntr likely controls fhuA expression independently of Fur expression and that Fur represses fhuA expression when EIIA^Ntr is absent under low-iron conditions. Moreover, fur transcription occurred independently of EIIA^Ntr (see Fig. S1A in the supplemental material), and Fur protein amounts were also comparable in the wild-type and the ptsN mutant strains (Fig. S1B). These findings suggest that the regulatory effect of EIIA^Ntr occurs at posttranscriptional levels.

FIG 2.

EIIA^Ntr-mediated regulation of fhuA requires Fur. β-Galactosidase activities of Salmonella with a p_fhuA-lacZ fusion in the normal chromosomal location and isogenic strains deleted for ptsN, fur, and fur ptsN genes with or without a plasmid expressing Fur from a heterologous promoter (pFur) were determined. Bacteria were grown to mid-log phase in acidified M9 medium (pH 5.8) (A) or the same medium with (+) or without (−) 100 µM FeSO₄ (B). The means and standard deviations (SD) of results from at least three independent experiments are shown as follows: bars, mean values; error bars, SD. Two-tailed t tests were performed for comparisons between indicated strains, and statistical significance is indicated as follows: ns, not significant; **, P < 0.01; ****, P < 0.0001.

Given that iron renders Fur’s regulatory function (6) and that Fur repressed fhuA expression in the ptsN mutant under low-iron conditions (Fig. 2A), we next investigated the effects of exogenous iron on EIIA^Ntr- and Fur-mediated regulation of fhuA. The addition of iron (100 μM of FeSO₄) greatly reduced fhuA expression levels in the wild-type strain (Fig. 2B). Under this condition, the ptsN mutant showed fhuA expression comparable to that exhibited by the wild-type strain (Fig. 2B). Moreover, lack of Fur resulted in high levels of fhuA expression independently of both EIIA^Ntr and iron levels (Fig. 2B). Taken together, these findings indicate that EIIA^Ntr promotes expression of iron uptake genes (including fhuA) via Fur under low-iron conditions.

EIIA^Ntr interacts with Fur.

Since EIIA^Ntr controls its targets via protein-protein interaction (10, 11, 13–16), we wondered whether it interacts with Fur. To examine interactions of EIIA^Ntr and Fur, a bacterial two-hybrid assay was used in which β-galactosidase levels are dependent on the proximity of fused proteins to fragments (i.e., T25 and T18) of the Bordetella pertussis adenylate cyclase in an E. coli strain lacking its own adenylate cyclase (28). Coexpression of T25-EIIA^Ntr and T-18-Fur produced ∼33-fold-higher levels of β-galactosidase activity than the strain expressing T25-EIIA^Ntr and the T-18 fragment or empty vectors (Fig. 3A). However, β-galactosidase activities from the strain expressing T25-EIIA^Ntr and T-18-Fur were lower than seen with expression of positive-control plasmids. Given that some EIIA^Ntr-mediated regulation of biological functions is dependent on its phosphorylation status (10, 11, 15, 16), we wondered if phosphorylation is necessary for the interaction of EIIA^Ntr with Fur. The unphosphorylatable EIIA^Ntr variant (H73A) gave expression levels of β-galactosidase activities similar to those seen with wild-type EIIA^Ntr (Fig. 3A), indicating that the phosphorylation status of EIIA^Ntr is not critical for the interaction with Fur. This is similar to results revealing other roles of EIIA^Ntr occurring independently of its phosphorylation status (12–14, 16, 29). Taken together, these results indicate that EIIA^Ntr interacts with Fur.

FIG 3.

EIIA^Ntr inhibits Fur binding to DNA. (A) β-Galactosidase activities were determined from cya E. coli mutant strains harboring the denoted plasmid combinations grown in acidified M9 medium containing 0.5 mM IPTG. The means and standard deviations (SD) of results from at least three independent experiments are shown as follows: bars, mean values; error bars, SD. Two-tailed t tests were performed for comparisons between indicated strains, and statistical significance is indicated as follows: ns, not significant; **, P < 0.01. (B) In vitro binding of Fur to the fhuA promoter with or without EIIA^Ntr. The fhuA promoter DNA (80 fmol) was incubated with Fur (2 µM) and EIIA^Ntr (10 µM) proteins. Data are representative of results from at least three independent experiments.

EIIA^Ntr controls iron uptake gene expression by hampering Fur binding to DNA.

Given that EIIA^Ntr interacts with Fur (Fig. 3A), it is possible that EIIA^Ntr controls fhuA expression by interfering with Fur binding to DNA. To test this, a gel shift assay was conducted using purified Fur and EIIA^Ntr proteins with the fhuA promoter DNA. Purified Fur bound to the fhuA promoter DNA, which formed a complex in vitro (Fig. 3B). Addition of EIIA^Ntr to this reaction hampered Fur binding to the fhuA promoter (Fig. 3B). However, EIIA^Ntr alone did not bind to the fhuA promoter (Fig. 3B), indicating that prevention of Fur binding to DNA by EIIA^Ntr was not due to competition between EIIA^Ntr and Fur for binding the promoter. These results suggest that EIIA^Ntr promotes iron uptake gene expression by relieving Fur-mediated repression via inhibition of Fur binding to its target promoter regions under low-iron conditions.

Control of iron uptake gene expression by EIIA^Ntr under low-iron conditions ensures Salmonella growth.

Given that EIIA^Ntr is required for expression of iron uptake genes under low-iron conditions (Fig. 1C; see also Fig. 2B) and that EIIA^Ntr hampers Fur binding to DNA by interacting with Fur (Fig. 3), we wondered if excess amounts of iron might alter interaction between EIIA^Ntr and Fur. To test this idea, the bacterial two-hybrid assay was done with a supply of 100 µM iron. Although coexpression of T25-EIIA^Ntr and T-18-Fur produced ∼33-fold-higher levels of β-galactosidase activity than were seen with the control strains (Fig. 3A), the same strain showed basal levels of β-galactosidase activity in the presence of excess iron (Fig. S2A). Furthermore, the addition of iron to the mixture of purified Fur with or without EIIA^Ntr and the target promoter DNA abolished EIIA^Ntr effects on Fur binding to target DNA (Fig. S2B). This is consistent with the finding that regulatory effects of EIIA^Ntr were not observed under conditions of high iron concentrations (Fig. 1C; see also Fig. 2B). These results further support the notion that EIIA^Ntr plays a critical role in controlling iron uptake gene expression under low-iron conditions, such as inside the host phagosome.

We then wondered whether EIIA^Ntr-mediated regulation of iron uptake genes could provide any advantage with respect to Salmonella physiology. Given that iron supports bacterial growth (30, 31), we examined iron-dependent growth phenotypes of the Salmonella wild-type strain and the ptsN mutant using iron concentrations differentially activating iron uptake genes in those two strains (Fig. 1C). As expected, supply of low concentrations of iron (250 nM FeSO₄) increased growth of wild-type Salmonella by >80% (Fig. 4). However, it resulted in an increase of only <30% for the Salmonella ptsN mutant (Fig. 4). These results suggest that the EIIA^Ntr-mediated activation of iron uptake genes promotes Salmonella growth under low-iron conditions.

FIG 4.

EIIA^Ntr promotes Salmonella growth under low-iron conditions. Bacterial growth was determined for the Salmonella wild-type strain and the ptsN mutant in acidified M9 medium with (+) or without (−) 0.25 µM FeSO₄. The optical density at 600 nm (OD₆₀₀) of bacteria was measured. The means and standard deviations (SD) from three independent experiments are shown as follows: bars, mean values; error bars, SD. Two-tailed t tests were performed for comparisons between indicated strains, and statistical significance is indicated as follows: *, P < 0.05; ***, P < 0.001.

DISCUSSION

Signal-sensing regulatory systems respond to environmental changes. The output of such systems is dependent on the sensitivity of the system to environmental signals. How does an organism tune the sensitivity of certain sensory systems? Here, we have established that EIIA^Ntr allows the iron sensor Fur to appropriately respond to environmental iron concentrations. Our findings suggest that Salmonella fine-tunes the responsiveness of Fur to iron via EIIA^Ntr (Fig. 5), thereby properly controlling the iron uptake system and potentially other biological functions (5, 8, 27). This provides a new insight into the mechanism of Fur’s iron sensing. Given that EIIA^Ntr and Fur proteins coexist in >600 different bacterial organisms (see Table S1 in the supplemental material), the EIIA^Ntr-mediated tuning of Fur’s iron response is likely a widespread mechanism in other bacteria.

FIG 5.

EIIA^Ntr enables Salmonella to properly control iron uptake gene expression in response to iron availability by modulating Fur. A model depicting iron-dependent activation of iron uptake gene expression by Fur and EIIA^Ntr is depicted. Transcription of iron uptake genes (including fhuA) is repressed by Fur at high iron concentrations. When iron concentrations decrease, iron-unbound Fur levels increase, but only a low level of expression of those genes is seen in the ptsN mutant. In the wild-type strain, EIIA^Ntr facilitates expression of iron uptake genes by hampering binding of Fur to its target promoters. This enables Salmonella to properly control the iron uptake system under low-iron conditions.

As described above, it has been believed that Fur is sufficient to operate as an iron sensor by itself because Fur binding to iron dictates its DNA binding ability (5, 6). We have now established that Fur actually requires an additional protein factor, EIIA^Ntr, to properly respond to changes of iron concentrations (Fig. 1; see also Fig. 5) in addition to its own ability to sense iron. In the absence of EIIA^Ntr, the high affinity of Fur for iron (6, 25) favors formation of the iron-bound form and repression of iron uptake genes even under conditions of low iron concentrations (Fig. 1A). When iron concentrations drop to below certain levels, iron-unbound Fur levels increase due to less availability of iron. Under this condition, EIIA^Ntr promotes iron uptake gene expression (Fig. 1A), possibly because Fur has a chance of interacting with EIIA^Ntr than iron. Alternatively or in addition, EIIA^Ntr binding to Fur may block Fur’s binding to iron.

Bacteria have perhaps chosen to evolve by tuning the activity of Fur through the use of auxiliary protein instead of altering the affinity of Fur for iron, which might be feasible with respect to adjusting Fur-mediated iron metabolism in various bacterial species, depending on their habitats. As a result of tuning of Fur’s activity via EIIA^Ntr, Fur not only attains better resolution in reflecting iron availability but also could potentially assimilate signals sensed and/or processed via EIIA^Ntr. The human pathogen Acinetobacter baumannii controls iron metabolism by altering the activity of Fur via BlsA (32), which is a photoreceptor that responds to light and temperature signals (33). Thus, BlsA allows A. baumannii to integrate light and temperature signals into iron uptake systems. Our findings suggest that EIIA^Ntr may enable Salmonella to control iron metabolism in response to environmental changes potentially impacting EIIA^Ntr, such as changes in nitrogen or amino sugar sources (16, 17) or other potential signals changing cellular amounts of EIIA^Ntr.

In addition to Fur, there is a membrane-bound iron sensor, PmrB, that is a constituent of the PmrA/PmrB two-component regulatory system (34). Interestingly, the Fur and PmrA iron-activated regulators negatively control Salmonella pathogenicity island 2 (SPI-2) gene expression (20, 21). As PmrB and Fur sense extracytoplasmic iron and cytoplasmic iron, respectively (5, 35), Salmonella controls SPI-2 gene expression in response to both extracytoplasmic and cytoplasmic iron levels via those iron sensors. As EIIA^Ntr hampers Fur binding to DNA (Fig. 3B), it probably favors induction of SPI-2 gene expression by relieving Fur-mediated negative regulation. Paradoxically, EIIA^Ntr reduces the transcription of SPI-2 genes by interfering with the binding of the major regulator of SPI-2, SsrB, to SPI-2 gene promoters (13). By impeding both Fur and SsrB, Salmonella probably achieves appropriate expression of virulence genes under low-iron conditions inside the host (19, 36).

We have shown that EIIA^Ntr plays an important role in controlling iron uptake systems under low-iron conditions. This ability to cope with low iron availability is probably critical for Salmonella during infection given that mammalian hosts utilize a strategy to withhold iron (37), because an increase of iron availability results in enhanced growth and/or virulence of many bacterial pathogens in vitro and in vivo (30, 31) and iron depletion reduces intracellular bacterial growth (38). As NRAMP1 protein removes iron from macrophage phagosomes (19), an Nramp1^−/− mouse lacking this iron transporter is highly sensitive to Salmonella infection (39). Moreover, lack of Fur highly attenuates Salmonella virulence in Nramp1^+/+ mice whereas this defect in virulence is reduced in Nramp1^−/− mice (23). We previously reported that EIIA^Ntr promotes virulence by preventing hyperactivation of Salmonella pathogenicity island 2 (SPI-2) gene expression (13). Our findings reported here suggest that the attenuated virulence of the ptsN mutant Salmonella in mice (13) might be due to both ectopic expression of SPI-2 genes and failure to import sufficient iron inside the host.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The S. enterica serovar Typhimurium strains used in this study were derived from strain SL1344. The strains and plasmids used in this study are listed in Table 1. Phage P22-mediated transduction was performed as described previously (40). All Salmonella strains were grown aerobically at 30 or 37°C in LB or M9 minimal medium at pH 5.8 supplemented with 0.5% Casamino Acids. Antibiotics were used at the following concentrations: ampicillin, 50 μg/ml; chloramphenicol, 25 μg/ml; kanamycin (Km), 50 μg/ml; and streptomycin, 50 μg/ml. Primers used for the construction of bacterial strains and plasmids, reverse transcription-quantitative PCR (qRT-PCR), and electrophoretic mobility shift assay (EMSA) are listed in Table 2.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Descriptiona	Reference or source
Strains
Salmonella enterica SL1344	Wild type, Strr	46
Salmonella enterica SR3203	ΔptsN	13
Salmonella enterica SR4101	pfur-lacZ (Kmr)	This work
Salmonella enterica SR4102	pfur-lacZ (Kmr) ΔptsN	This work
Salmonella enterica SR4103	fur-FLAG	This work
Salmonella enterica SR4104	fur-FLAG ΔptsN	This work
Salmonella enterica SR4125	pfhuA-lacZ (Kmr)	This work
Salmonella enterica SR4125	pfhuA-lacZ (Kmr) ΔptsN	This work
Salmonella enterica SR4131	pfhuA-lacZ (Kmr) Δfur	This work
Salmonella enterica SR4132	pfhuA-lacZ (Kmr) Δfur ΔptsN	This work
Escherichia coli DH5α	F– supE44 ΔlacU169 (φ80 lacZΔM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1	47
Escherichia coli BTH101	F– cya-99 araD139 galE15 galK16 rpsL1 (Strr) hsdR2 mcrA1 mcrB1	28
Plasmids
pCP20	reppSC101ts Apr Cmr FLP+ cI857+	41
pCE70	repR6K Kmr Cmr FRT tnpR lacZY	43
pKD3	repR6K Apr FRT Cmr FRT	41
pKD13	repR6K Apr FRT Kmr FRT	41
pKD46	reppSC101ts Apr paraBAD γ β exo	41
pUHE21-2lacIq	reppMB1 Apr lacIq	48
pFur	reppMB1 Apr lacIq fur	21
pJJ37	reppMB1 Apr lacIq ptsN-His6	13
pJJ48	reppMB1 Apr lacIq fur-His6	This work
pKT25	Kmr repp15A	44
pUT18	Apr reppMB1	44
pT25-ptsN	Kmr repp15A ptsN	13
pT25-ptsN(H73A)	Kmr repp15A ptsN(H73A)	12
pT25-zip	Kmr repp15A zip	44
pT18-fur	Apr reppMB1 fur	This work
pT18-zip	Apr reppMB1 zip	44

^aAp^r, ampicillin resistance; Cm^r, chloramphenicol resistance; Km^r, kanamycin resistance; Str^r, streptomycin resistance.

TABLE 2.

Primers used in this study

Primer	Sequence (5′–3′)
Strain construction
fur-Red-F	CGC TTC CTC GTT TAA AAA TTC TGG AAG TTC TTC AGG AAC CTG TAG GCT GGA GCT GCT TCG
fur-Red-F	TCG TGA TGA TGC TGT TGC GTC AGT TCA AAA ACG GAT TTA CAT TCC GGG GAT CCG TCG ACC
fur-lacZ-F	GCG ACT GCC GCG AAG ACG AGC ACG CGC ACG ATG ACG CGA CTA AAT AAG TGC CCG TCG TTT TAC AAC GTC G
fur-lacZ-R	CAA CAT CAA GCG GCA GGA AAG AGG AGG ATA TAA AAA AGC CAA CCG GGC GGC GTG TAG GCT GGA GCT GCT TC
fhuA-Red-F	ATC GTT TAC GTT ATC ATT CAC TTT CAT CAG AGA TAT ACC ATG TAG GCT GGA GCT GCT TCG
fhuA-Red-R	CCT GCG CTA ATG GGT TGG TTG GAT CGG CGG TCA GGT TAT TAT TCC GGG GAT CCG TCG ACC
Plasmid construction
pHis6x-Fur-F	CAT GTT CTG AAT TCA AAT TAT GCA TCA CCA TCA CCA TCA CGC AAT GAC TGA CAA CAA TAC CGC ATT AAA
pHis6x-Fur-R	ACC GGG CGG TTG GAT CCT CGA AAG ATT T
pUT18C-Fur-F	TTA GCA ACA GGA GGA TCC CCG CAT GAC T
qRT-PCR
fhuA-qRT-F	GTT CAA CCG AAA GAA GAA ACC ATT A
fhuA-qRT-R	GTT TTT TCG ATA GGT GTA TCA GTT TTG
fepA-qRT-F	AGA AGA TTC ATT CCC TGA CCT TAC TG
fepA-qRT-R	TAT CGG TTT TGT CTT CCG CCA TCA
iroB-qRT-F	ATG CGT ATT CTG TTT GTC GGT CCA
iroB-qRT-R	CAG TAC TTC ATG GCC ATT AAC ACG A
rrs-qRT-F	CCA CAA AAC TTA TGG ATT TAT GCG T
rrs-qRT-R	TTT ACG CCC AGT AAT TCC GAT T
EMSA
Fur-EMSA-F1	CTC GAC GAC ATC CTC AAC GCC TAA TCT
Fur-EMSA-R1	AAA CGA GGA AGC GTT ACT TTC AGG CCA G

Construction of bacterial strains.

The method of Datsenko and Wanner (41) was used for chromosomal gene deletion and epitope tagging.

For construction of the fur deletion strain, the kanamycin resistance (Km^r) cassette from plasmid pKD13 was amplified using primers fur-RED-F and fur-RED-R. The resulting PCR products were introduced into the SL1344 strain containing plasmid pKD46, followed by selection for Δfur::kan transformants. The Km^r cassette was removed using plasmid pCP20 (41).

A Salmonella strain expressing the Fur protein with a FLAG tag at the C terminus in the normal fur chromosomal location was constructed. The Km^r cassette from plasmid pKD13 was amplified using primers Fur-FLAG-F and Fur-FLAG-R, and the PCR products were introduced into the SL1344 strain harboring plasmid pKD46. The Km^r cassette was removed using plasmid pCP20.

A strain carrying a lacZ fusion to the fhuA gene was constructed as described previously (42). The Km^r cassette from plasmid pKD13 was amplified using primers fhuA-RED-F and fhuA-RED-R. The resulting PCR products were introduced into the SL1344 strain harboring plasmid pKD46, and the Km^r cassette was removed using plasmid pCP20. Finally, the lacZY genes were introduced into the flippase recognition target (FRT) site using plasmid pCE70 (43).

A strain carrying a lacZ fusion to the fur gene was constructed as described previously (42). The Km^r cassette from plasmid pKD13 was amplified using primers fur-lacZ-F/fur-lacZ-R. The resulting PCR products were introduced into the SL1344 strain harboring plasmid pKD46, and the Km^r cassette was removed using plasmid pCP20. Finally, the lacZY genes were introduced into the FRT site using plasmid pCE70 (43).

Plasmid construction.

A plasmid expressing Fur protein with a His₆ tag at N terminus from the lac promoter was constructed as follows: the fur gene was amplified from wild-type Salmonella (SL1344) using primers pHis6x-Fur-F/pHis6x-Fur-R, and the PCR fragments were introduced between the EcoRI and BamHI sites of the pUHE21-2lacI^q plasmid vector.

A plasmid expressing T18-Fur fusion protein was constructed as follows: the fur gene was amplified from wild-type Salmonella (SL1344) using primers pUT18C-Fur-F/pHis6x-Fur-R and then introduced between the BamHI and EcoRI sites of pUT18C (44).

β-Galactosidase assay.

β-Galactosidase assays were carried out with at least three biological replicates with technical duplicates, and the activity was determined as described previously (45).

RNA isolation and reverse transcriptase-quantitative PCR (qRT-PCR).

Salmonella strains were grown as described above, and total RNA was isolated using an RNeasy minikit (Qiagen). After DNase treatment of the isolated RNA, cDNA was synthesized using Omni Transcript reverse transcription reagents (Qiagen) and random hexamers (Invitrogen). Quantification of the cDNA was carried out using 2× iQ SYBR green Supermix (Bio-Rad), and real-time amplification of the PCR products was performed using an iCycler iQ real-time detection system (Bio-Rad). The primers used for detection of the gene transcripts are listed in Table 2. Data were normalized to 16S rRNA expression levels.

Western blotting.

Salmonella strains encoding the Fur-FLAG protein from the normal chromosomal location were grown under the indicated conditions. Bacteria were collected by centrifugation, and cell lysates were prepared using B-PER solution (Pierce). Proteins from cell lysates were resolved by 12% SDS-PAGE, and the Fur and DnaK proteins were detected using anti-FLAG (Sigma; 1:2,000) and anti-DnaK (Abcam; 1:5,000) antibodies. The blots were developed using anti-mouse IgG horseradish peroxidase-linked antibody (GE Healthcare; 1:5,000) with an ECL detection system (Amersham Biosciences).

Bacterial two-hybrid assay.

E. coli BTH101 (which lacks the cya gene) was used for this assay. Derivatives of plasmids pUT18 and pKT25 were introduced into BTH101 cells as indicated. These cells were grown overnight in M9 minimal media containing ampicillin (100 µg/ml) and kanamycin (50 µg/ml). They were transferred to 1 ml of the same fresh medium containing 0.5 mM IPTG (isopropyl-β-d-thiogalactopyranoside) at a dilution of 1:100 and grown for 8 h with shaking at 30°C. β-Galactosidase activities were determined as described above.

Protein purification.

His6-tagged Fur protein was expressed in E. coli strain BL21(DE3) as follows: bacterial cells were grown in LB at 37°C until the optical density at 600 nm (OD₆₀₀) reached 0.5, and expression of those proteins was induced by addition of IPTG (0.5 M) followed by growth at 30°C for 5 h. Cells were harvested, washed, and suspended in buffer A (20 mM Tris [pH 8.0], 150 mM NaCl, 20 mM imidazole). The cells were then disrupted by sonication, and cell debris was removed by centrifugation at 20,000 × g at 4°C for 30 min. The supernatant was applied to a 1.5-ml nickel-nitrilotriacetic acid (Ni-NTA) agarose column equilibrated in buffer A, washed with 25 column volumes of the same buffer, and eluted using a gradient of buffer A and buffer B (20 mM Tris [pH 8.0], 150 mM NaCl, 250 mM imidazole). The fractions were then collected and analyzed by SDS/PAGE, and selected fractions were dialyzed against buffer C (20 mM Tris [pH 8.0], 150 mM NaCl, 10% glycerol).

Electrophoretic mobility shift assay (EMSA).

EMSAs were performed to determine binding of Fur to DNA in vitro. DNA fragments corresponding to the fhuA promoter was amplified by PCR using ³²P-labeled primers EMSA-fhuA-F1/EMSA-fhuA-R1 with wild-type Salmonella chromosomal DNA as a template. The promoter DNA was purified from agarose gels using a gel extraction kit (Qiagen). The labeled DNA probe (16 fmol) was incubated with the His₆-Fur protein in the presence or absence of EIIA^Ntr-His₆ at room temperature for 20 min in 20 μl of binding buffer (10 mM Tris [pH 7.5], 50 mM KCl, 5 mM MgCl₂, 0.5 mM dithiothreitol [DTT], 2.5% glycerol) containing 50 ng/μl poly(dI-dC). The reaction mixtures were resolved by 6% PAGE, and the radiolabeled DNA fragments were visualized using a BAS2500 system (Fuji Film).

Protein co-occurrence.

Co-occurrence of EIIA^Ntr and Fur proteins across sequenced organisms was analyzed using STRING software version 11.0.