Involvement of the Cra Global Regulatory Protein in the Expression of the iscRSUA Operon, Revealed during Studies of Tricarballylate Catabolism in Salmonella enterica

Abstract

In Salmonella enterica, tricarballylate (Tcb) catabolism requires function of TcuB, a membrane-bound protein that contains [4Fe-4S] clusters and heme. TcuB transfers electrons from reduced flavin adenine dinucleotide in the Tcb dehydrogenase (TcuA) to electron acceptors in the membrane. We recently showed that functions needed to assemble [Fe-S] clusters (i.e., the iscRSUA-hscBA-fdx operon) compensate for the lack of ApbC during growth of an apbC strain on Tcb. ApbC had been linked to [Fe-S] cluster metabolism, and we showed that an apbC strain had decreased TcuB activity. Here we report findings that expand our understanding of the regulation of expression of the iscRSUA genes in Salmonella enterica. We investigated why low levels of glucose or other saccharides restored growth of an apbC strain on Tcb. Here we report the following findings. (i) A ≤1 mM concentration of glucose, fructose, ribose, or glycerol restores growth of an apbC strain on Tcb. (ii) The saccharide effect results in increased levels of TcuB activity. (iii) The saccharide effect depends on the global regulatory protein Cra. (iv) Putative Cra binding sites are present in the regulatory region of the iscRSUA operon. (v) Cra protein binds to all three sites in the iscRSUA promoter region in a concentration-dependent fashion. To our knowledge, this is the first report of the involvement of Cra in [Fe-S] cluster assembly.

Tricarballylate (Tcb) is a citrate analog that causes grass tetany in ruminants, a disease characterized by acute hypomagnesia caused by the excretion of the Mg Tcb chelate (31, 37). Unlike the normal rumen flora, Salmonella enterica uses Tcb as a source of carbon and energy (17). Previous work identified the genes encoding tricarballylate utilization (tcuRABC) functions in this bacterium (24) and established their biochemical activities (Fig. 1). TcuA is a flavin adenine dinucleotide (FAD⁺)-dependent Tcb dehydrogenase (22), TcuB is an electron transfer protein (23), TcuC is a citrate and Tcb transporter (24, 38), and TcuR is the transcription activator of the tcuABC operon (24). Although the TcuABC proteins are sufficient for S. enterica to grow on Tcb, [Fe-S] assembly systems and heme and FAD biosynthetic functions are required for the synthesis of functional TcuA and TcuB proteins.

FIG. 1.

The Tcb utilization pathway of S. enterica. Of relevance to this paper is the function of the TcuB protein, which requires two [4Fe-4S] clusters for function (23). TcuB is required to shuttle electrons from the reduced FAD cofactor of TcuA to the quinone pool in the cell membrane. Both ApbC and IscU play a functional role in [4Fe-4S] cluster maturation in TcuB.

The apbC locus of S. enterica was initially described within the context of thiamine biosynthesis (27), and subsequent studies linked ApbC activity to the functionality of enzymes containing [Fe-S] clusters (40). Boyd et al. recently reported that a mutation in the apbC gene also blocks growth of S. enterica on Tcb (5). Growth of the apbC strain on Tcb was restored by increasing the expression of the iscRSUA operon and, in particular, the iscU gene, which encodes an [Fe-S] cluster scaffold protein (5). It has been hypothesized that ApbC specifically loads [Fe-S] clusters into certain target proteins and that IscU can compensate for the absence of ApbC only when it is present at high levels. Precedent in support of this idea is found in the studies by Dos Santos et al. with Azotobacter vinelandii. Dos Santos et al. found that IscU could compensate for NifU in nitrogenase [Fe-S] cluster maturation only at elevated expression levels (12).

Here we show that low concentrations of glucose and other saccharides (<1 mM) restore growth of an apbC strain on Tcb, that the saccharide effect is mediated via the global regulatory protein Cra, that Cra directly affects the expression of the iscRSUA operon, and that the saccharide effect results in increased levels of TcuB activity.

MATERIALS AND METHODS

Bacterial strains, culture media, and growth conditions.

A list of strains and plasmids used and their genotypes is provided in Table 1. All chemicals were purchased from Sigma unless otherwise stated. Escherichia coli cultures were maintained in lysogeny broth (LB; Difco) (3, 4). Nutrient broth (Difco) was used as rich medium for S. enterica. Antibiotic concentrations were as follows: ampicillin (Ap), 100 μg/ml; chloramphenicol (Cm), 20 μg/ml; kanamycin (Km), 50 μg/ml; and tetracycline (Tc), 20 μg/ml. No-carbon essential (NCE) medium (2) was used as minimal medium and was supplemented with MgSO₄ (1 mM) and trace minerals (1, 11). When added, oxidized l-glutathione (GSSG) was used at 2.5 mM. Unless otherwise noted, whenever used as the sole carbon and energy sources, Tcb, citrate, pyruvate, and succinate were all at 20 mM. The carbon sources tested for rescue of growth of the apbC strain on Tcb were glucose, fructose, ribose, glycerol, pyruvate, citrate, and succinate, all of which were present at 1 mM, unless otherwise noted. The abbreviation “Tcb + Gluc” indicates Tcb was the main carbon and energy source and glucose was present at 0.5 to 1 mM.

TABLE 1.

Strains and plasmids used in this study^a

Strain or plasmid	Genotype	Source or reference
Strains
E. coli
DH5α/F′	F′ endA1 hsdR17(rK− mK+) supE44 thi-1 recA1 gyrA (Na1r) relA1 Δ(lacZYA-argF)U169 deoR [φ80dlacΔ(lacZ)M15]	New England Biolabs
BL21-AI	F−ompT hsdSB(rB− mB−) gal dcm araB::T7 RNAP-tetA	Invitrogen
S. enterica
TR6583	metE205 ara-9	K. Sanderson via J. Roth
Derivatives of TR6583
JE6692	pKD46
JE10432	iscR101::kan+
JE7212	tcuA33::MudJ(lacZ+kan+)b
JE10614	tcuA33::MudJ(lacZ+kan+)/pTCU5
JE10621	tcuA33::MudJ(lacZ+kan+) apbC55::Tn10d(tet+)c
JE10622	tcuA33:: MudJ(lacZ+kan+) apbC55::Tn10d(tet+)/pTCU5
DM10310	ara-9	D. Downs
Derivatives of DM10310
DM10325	Δisc5::cat+ (deletes iscS→fdx)	42
JE10452	gshA102::MudJ(lacZ+kan+)
JE10440	cra426::MudJ(lacZ+kan+); the lacZ+ gene is not under the control of the cra promoter and thus is not a reporter of cra expression
JE10281	apbC55::Tn10d(tet+) Δisc5::cat+ (deletes iscS→fdx)
JE10444	apbC55::Tn10d(tet+) Δisc5::cat+ (deletes iscS→fdx) /pDB1282(bla+)
JE10453	apbC55::Tn10d(tet+) gshA102::MudJ(lacZ+kan+)
JE10438	apbC55::Tn10d(tet+) sufS::cat+
JE10457	apbC55::Tn10d(tet+) ybfM106::MudJ(lacZ+kan+) tcuB17
JE10435	apbC55::Tn10d(tet+) ΔiscR102
JE10434	apbC55::Tn10d(tet+) iscR101::kan+
JE10441	apbC55::Tn10d(tet+) cra426::MudJ(kan+)
JE10442	apbC55::Tn10d(tet+) cra426::MudJ(kan+)/pBAD30
JE10443	apbC55::Tn10d(tet+) cra426::MudJ(kan+)/pCRA1
JE10433	iscR101::kan+
JE10836	ΔiscR102
JE10837	cra426::MudJ(lacZ+kan+) ΔiscR102
JE10838	cra426::MudJ(lacZ+kan+) ΔiscR102 apbC55::Tn10d(tet+)
DM10300	apbC55::Tn10d(tet+)	D. Downs
Plasmids
pARA13	PBAD expression vector; bla+ (Apr)	6
pDB1282	A. vinelandii iscSUA hscBA fdx orf3 ndx in pARA13; bla+ (Apr)	D. Dean
pTCU5	tcuC+ in pBAD30; bla+ (Apr)	24
pJMB106	cra+ in pET20b; bla+ (Apr)
pCRA1	cra+ in pBAD30; bla+ (Apr)
pBAD30	PBAD expression vector; bla+ (Apr)	18
pISC2	PiscR in pFZY1; bla+ (Apr)
pFZY1	Promoterless lacZ fusion vector; bla+ (Apr)	20
pKD46	araBp-gam-bet-exo repA101(Ts) oriR101	10
pG-Tf2	groES groEL tig cat+ (Cmr)	Takara Bio, Inc.

^aUnless otherwise stated, all strains and plasmids were constructed during the course of these studies.

^bMudJ(kan⁺) is an abbreviation of MudI1734 (7).

^cTn10d(tet⁺) is an abbreviation of Tn10Δ16Δ17 (44).

Skovran et al. previously reported that an isc mutant required nicotinic acid and thiamine for growth (41). The requirements for these nutrients were bypassed by using overnight cultures grown in LB, without washing the cells, to inoculate fresh medium.

For growth curves, S. enterica strains were grown overnight in LB, and 2 μl of the overnight cultures (∼4 × 10⁶ CFU) was used to inoculate 198 μl of fresh NCE medium in 96-well microtiter dishes. Growth was monitored using an ELx808 high-throughput spectrophotometer (BioTek Instruments). Absorbance readings at either 630 or 650 nm were taken every 30 min for 36 h. Cultures were shaken for 240 s between measurements. The incubation chamber was maintained at 37°C. Each growth experiment was performed in triplicate.

Genetic crosses.

Transductions involving phage P22 HT105 int-201 were performed as described previously (8, 11, 34, 35).

Construction of an in-frame deletion of the iscR gene.

An in-frame deletion of iscR was constructed using a modification of the Datsenko protocol (10). Briefly, the kan⁺ cassette of plasmid pKD4 was amplified using a 5′ primer containing 51 bp that matched the 5′ end of iscR (5′-CATTTTACAATAAAAAACCCCGGGCAGGGGCGAGTTTGAGGTTAAGTAAGACATGGTG TAGGCTGGAGCTGCTTC-3′) and a 3′ primer containing 51 bp that matched the 5′ end of iscS (5′-CAACACGCGGGTCCACCGGCGTGGTTGCGGAGTAGTCGAGATAAATCGGTAATTTCATATGAATATCCTCC TTAG-3′). Manipulations were performed in strain JE6692 (Table 1). Insertion of the kan⁺ gene into iscR was verified by sequencing. P22 phage grown on a strain carrying the iscR101::kan⁺ insertion was used as a donor to transduce strain DM10300 [apbC::Tn10d(tet⁺)] to kanamycin resistance. Resolution of the kan⁺ gene was achieved as described previously (10).

Recombinant DNA techniques.

Restriction and modification enzymes were purchased from MBI Fermentas unless otherwise stated and were used according to the manufacturer's instructions. All DNA manipulations were performed in E. coli DH5α. Plasmids were transformed into E. coli cells by CaCl₂ heat shock as described previously (19). Plasmids isolated from E. coli were transformed into S. enterica via electroporation (25). Plasmid DNA was isolated using the Wizard Plus SV plasmid miniprep kit from Promega. DNA fragments were isolated from 1% (wt/vol) agarose gels and purified using the Qiaquick gel extraction kit (Qiagen). PCRs were purified using the Qiaquick PCR purification kit (Qiagen). Nonradioactive BigDye ABI-PRISM DNA sequencing was performed as per the manufacturer's instructions, and reaction mixtures were resolved and analyzed at the Biotechnology Center at the University of Wisconsin—Madison.

Plasmid constructions.

Plasmids were propagated in E. coli strain DH5α, except where noted. Genomic DNA for PCR was prepared from S. enterica strain TR6583 using the Wizard SV genomic purification system from Promega. All primers used for PCR amplifications were purchased from Integrated DNA Technologies.

Plasmid pCRA1.

The cra⁺ gene from S. enterica was PCR amplified using primers with a 5′ KpnI site and 3′ XbaI site and cloned into plasmid pBAD30 (18) using the same restriction sites. Plasmid pCRA1 was 5.9 kb and encoded Ap^r.

Plasmid pISC2.

The first 500 bp immediately 5′ to the start codon for iscR were cloned into pFZY1 (promoterless lacZ fusion vector) (20) using 5′ EcoRI and 3′ HindIII sites. Plasmid pISC2 was 11.8 kb and encoded Ap^r.

Plasmid pJMB106.

The cra⁺ gene from S. enterica was amplified using primers with a 5′ NdeI site and a 3′ XhoI site, and the resulting amplicon was cloned into plasmid pET20b cut with the same restriction enzymes. The resulting plasmid (pJMB106) was 4.7 kb and encoded Ap^r.

Overproduction and purification of the TcuA protein.

TcuA protein was purified and reconstituted in vitro with FAD⁺ as described previously (22).

Overproduction and purification of Cra. (i) Overproduction.

Escherichia coli strain BL21-AI containing the protein expression plasmids pJMB106 and pG-Tf-2 (Takara) was grown at 37°C in LB containing Ap (100 μg/ml) and Cm (20 μg/ml). The pGTf-2 plasmid was used to overproduce GroEL, GroES, and Tig, which had previously been shown to increase Cra solubility (30). When the optical density at 650 nm (OD₆₅₀) of the culture was 0.6, the medium was cooled to 15°C and l-(+)-arabinose (1 mM), isopropyl-β-d-thiogalactopyranoside (IPTG; 0.1 mM), and Tc (10 ng/ml) were added. Cells were incubated for 18 h before harvesting by centrifugation. Cell paste was washed with buffer A (50 mM Tris-HCl [pH 8.0]), flash frozen with liquid nitrogen, and stored at −80°C.

Protein purification.

Frozen cell paste was suspended in an equal volume of buffer A containing DNase (0.03 mg/ml). Cell suspensions were passed three times through a chilled French pressure cell at 4°C and subsequently clarified by centrifugation (39,000 × g for 40 min at 4°C). The clarified cell extract was loaded onto a 1.6- by 10-cm pre-equilibrated Ni²⁺-loaded chelating Sepharose fast flow column (GE Healthcare) and washed with 20 column volumes of 50 mM Tris (pH 8.0) plus 1 M NaCl. The column was re-equilibrated with buffer A, and recombinant protein was eluted during a 30-column-volume linear gradient from 0 to 100% elution buffer (50 mM Tris [pH 8.0], 250 mM imidazole). Fractions were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (21) and subsequent gel staining with Coomassie G-250 (33). Fractions that contained Cra protein at >95% purity were pooled and concentrated over a 30,000-Da molecular-mass-cutoff membrane (Amicon YM30). After concentration, Cra was dialyzed overnight in a mixture of 50 mM Tris-HCl (pH 8.0), 10% (vol/vol) glycerol, and 150 mM NaCl. Finally, 100-μl samples of a Cra protein solution were flash frozen in liquid nitrogen and were stored at −80°C until needed. All steps were performed at 4°C, and buffers used for dialysis had the pH adjusted at 4°C.

Protein concentration determination.

The protein concentration was determined using a copper-based colorimetric assay and used a reagent containing bicinchoninic acid to detect the cupreous ion (Pierce). Bovine serum albumin (2 mg/ml) was used as a standard.

Nonradioactive EMSAs.

Nonradioactive electrophoresis mobility shift assays (EMSAs) using Cra-H₆ were performed using the LightShift chemiluminescent EMSA kit (Pierce). A 239-bp PCR product containing the three putative Cra binding sites in the iscR promoter region was amplified using the following primers: (i) upstream primer 5′-TATCCGCTGGTGGACGATCTT-3′ and (ii) downstream primer 5′-TCAGGTATTTAGCGTTCCGTG-3′. For a positive control, the icd promoter region (331 bp) was amplified using the primers described by Ramseier et al. (30). Both PCR fragments were biotinylated using the biotin 3′-end DNA labeling kit (Pierce) as per the manufacturer's instructions. The labeling efficiency was measured at >90%. Cra binding assays were performed as described previously (30). Binding assays were carried out in a reaction volume of 20 μl containing various amounts of Cra protein (69 to 1,104 nM of monomer), 20 fmol labeled DNA, 1 μg nonspecific poly(dI-dC)·poly(dI-dC) DNA, Tris-HCl (4 mM; pH 7.0), MgCl₂ (2 mM), sodium chloride (5 mM), glycerol (6.5% [vol/vol]), and dithiothreitol (DTT; 1 mM). Binding reaction mixtures were incubated at room temperature for 20 min before being loaded onto a 4% polyacrylamide gel (29:1 acrylamide-bisacrylamide) that had been prerun at 4°C for 1 h in 45 mM Tris-borate (pH 8.3) containing 1 mM EDTA at 200 V. Prior to loading, 5 μl of 5× sample loading buffer (Pierce) was added to each reaction mixture, and 20 μl of each reaction mixture was loaded onto the gel. Electrophoresis was allowed to proceed for 1.5 h. Samples were transferred to an Immobilon-Ny+ membrane for 30 min in 45 mM Tris-borate (pH 8.3) containing EDTA (1 mM) at 200 V. DNA was cross-linked to the membrane using a transilluminator for 10 min. Detection of biotin-labeled DNA was performed using the LightShift chemiluminescent EMSA kit (Pierce) as per the manufacturer's instructions. Chemiluminescent detection was performed using a Typhoon imager (Molecular Devices).

In vitro activity assays for TcuB.

The amount of active TcuB protein in cell extracts was measured as a function of the TcuB-dependent Tcb dehydrogenase (TcuA) enzyme as described previously (23). cis-Aconitate, the product of the TcuA reaction, was quantified by high-performance liquid chromatography (22). The following strains were grown on minimal medium supplemented with glucose (10 mM) and Tcb (10 mM): DM10310 (apbC⁺), DM10300 (apbC), and JE10457 (tcuB). Strain DM10310 (apbC⁺) was used as a control and was grown on minimal medium supplemented with only glucose (10 mM). Cell extracts were prepared from each culture as described previously (23). Reaction mixtures (200 μl) contained TcuA (5 μg), TcuB-enriched extract (100 μg), 2-morpholinoethanesulfonic acid (MES; 100 mM [pH 6.5 at 30°C]), and DTT (1 mM). Assays were performed at 30°C, and product formation was measured after a 20-min incubation period.

β-Galactosidase activity assays.

β-Galactosidase activity was measured as described previously (13). One unit of enzyme activity was defined as the amount of enzyme required to hydrolyze 1 nmol of o-nitrophenyl-β-d-galactopyranoside (ONPG) per min. Specific activity is reported as the number of units per OD₆₅₀ unit. Enzyme activity was measured in mid-log cultures (i.e., OD₆₅₀ of ∼0.4 to 0.6). Cell density was monitored with a Spectronic 20D spectrometer (Manostat). Unless otherwise noted, 5-ml cultures were used.

RESULTS AND DISCUSSION

Growth of an apbC strain on Tcb is restored by low concentrations of saccharides.

While growth of S. enterica on Tcb alone required apbC function (Fig. 2), we noticed that growth of the apbC strain on Tcb was restored by the addition of low concentrations of glucose (0.5 to 1 mM) to the medium (Fig. 2). Although the growth rate of the apbC strain on Tcb + Gluc was slower than that of the apbC⁺ strain (8 and 6 h for the apbC strain with 0.5 and 1 mM glucose, respectively, versus 2 h for the apbC⁺ strain), the cultures of the apbC strain growing in the presence of low glucose levels reached full density. Results from control experiments showed that the contribution of glucose to the final cell density was minimal (Fig. 2). This stimulatory effect was not specific to glucose. Low concentrations (i.e., <1 mM) of fructose, ribose, and glycerol also restored growth of the apbC strain on Tcb. However, carboxylic acids such as pyruvate, citrate, and succinate did not (see Fig. S8 in the supplemental material).

FIG. 2.

Glucose (Gluc) restores growth of an apbC strain on Tcb. apbC⁺ identifies strain DM10310, and apbC identifies strain DM10300 [apbC55::Tn10d(tet⁺)]. The experiment was performed in triplicate; error bars denote standard deviations.

During growth on Tcb + Gluc, the level of TcuB activity doubles in the apbC strain.

We investigated whether glucose had an effect on the level or activity of the TcuB protein, a protein that contains two [4Fe-4S] clusters. To this end, we measured the level of TcuB activity in apbC and apbC⁺ strains grown on glucose with or without Tcb in the medium. Inclusion of Tcb in the glucose medium (thus inducing expression of the tcuABC operon) resulted in a 15-fold increase in TcuA activity in the apbC⁺ strain (Fig. 3), and the activity depended on the presence of TcuB (Fig. 3). Surprisingly, the level of TcuB-dependent TcuA activity in the apbC strain grown on Tcb + Gluc was ∼2-fold higher than that in the apbC⁺ strain (Fig. 3), a result that suggested that, under the conditions tested, ApbC had a negative effect on the assembly of [Fe-S] clusters in the TcuB protein in the wild-type strain. One possible explanation for this observation was that tcuB expression increased in the apbC strain. To test whether the increase in TcuB activity in the apbC strain grown on Tcb + Gluc was due to increased transcription of the tcuABC operon, we quantified the expression of a lacZ transcriptional reporter under the control of the tcu promoter [i.e., tcuA::MudJ(lacZ⁺)] in the apbC⁺ (JE7212) and apbC (JE10621) strains. Strains JE7212 and JE10621 were transformed with plasmid pTCU5, which contained the wild-type allele of tcuC, the gene that encodes the Tcb transporter; β-galactosidase activity was quantified in mid-log-phase cultures. Levels of expression of the tcuA::MudJ(lacZ⁺) reporter were not substantially different in the apbC⁺ strain (JE10614) and the apbC strain (JE10622) grown on either LB plus Tcb (304 ± 4 versus 336 ± 15 U/OD₆₅₀, respectively) or NCE-Gluc + Tcb (870 ± 30 versus 760 ± 70 U/OD₆₅₀, respectively). These results suggested that the high TcuB activity measured in the apbC strain during growth on Tcb + Gluc (Fig. 3) was not due to increased tcuABC expression. However, these results did not exclude the possibility that TcuB was regulated posttranscriptionally. Unfortunately, antibodies for TcuB were not available, and we were unable to test whether TcuB protein levels were different between the apbC strain and the wild-type strain. Regardless, these experiments showed that when glucose is present, ApbC protein is not necessary for loading of [Fe-S] clusters into TcuB. In the absence of glucose, however, ApbC is required for TcuB activity (5), suggesting that ApbC is not the only means of loading [Fe-S] clusters into TcuB, but the expression of the redundant ApbC-like activity requires the presence of glucose.

FIG. 3.

TcuB activity in cell extracts of apbC and apbC⁺ strains. Shown are the amounts of cis-aconitate produced in 20 min in a reaction mixture that contained Tcb dehydrogenase (TcuA; 5 μg = 98 pmol) and 100 μg of protein from TcuB-enriched cell extracts. apbC⁺ identifies strain DM10310, apbC identifies strain DM10300 [apbC55::Tn10d(tet⁺)], and tcuB identifies strain JE10457 (tcuB17). Tcb and Gluc denote 20 mM Tcb and 10 mM glucose, respectively. The tcuB strain was a negative control for lack of TcuB protein, while the glucose-alone condition was a negative control for lack of induction of the tcuABC operon. The experiment was performed in triplicate; error bars denote standard deviations.

The effect of glucose on the growth of an apbC strain on Tcb requires iscSUA functions.

The documented redundancy of ApbC with IscU (5) led us to investigate a possible connection between the saccharide effect and the expression of the iscRSUA operon. If this connection existed, mutations that inactivated iscRSUA genes in an apbC strain would eliminate the positive effects of saccharides on growth of the apbC strain on Tcb. To test this idea, we introduced isc mutations into an apbC strain and assessed the effect of glucose on the growth of the resulting strain on Tcb. Strain JE10281 (apbC ΔiscSUA) did not grow on Tcb + Gluc. Growth of the apbC ΔiscSUA strain on Tcb + Gluc was restored when the wild-type iscSUA⁺ functions from Azotobacter vinelandii were provided on a plasmid (see Table 1 for a complete description of the genes on plasmid pDB1282). In contrast, the growth behavior of a strain lacking the iscSUA functions (DM10325) was similar to that of the apbC isc/pDB1282 strain, indicating that as long as the cell had functional ApbC, the Isc functions were dispensable during growth on Tcb (see Fig. S9 in the supplemental material). Together, these results supported the idea that the effect of glucose on the ability of an apbC strain to grow on Tcb depended on isc functions.

Glutathione is required for the glucose effect.

A role for glutathione in [Fe-S] cluster assembly/repair was previously reported (16, 39, 43). Hence we predicted that inactivation of the gshA gene (which encodes l-glutamate:l-cysteine γ-ligase [ADP forming]; EC 6.3.2.2) in the apbC strain would prevent growth on Tcb + Gluc. Indeed, that was the case (see Fig. S10 in the supplemental information). Addition of GSSG (2.5 mM) to the medium restored growth of the apbC gshA strain on Tcb + Gluc, confirming that glutathione was required for the glucose effect, a finding consistent with the reported involvement of glutathione in [Fe-S] cluster assembly/repair.

suf functions are not involved in the glucose effect.

The genome of S. enterica contains suf genes, whose functions contribute to the assembly of [Fe-S] clusters under conditions of iron limitation and oxidative stress (26, 45). Strain JE10438 (apbC sufS) grew on Tcb + Gluc (data not shown), suggesting that suf functions were dispensable during growth of an apbC strain on Tcb + Gluc.

Saccharide-dependent growth of an apbC mutant strain on Tcb requires cra function.

One hypothesis was that glucose stimulated expression of the iscRSUA operon, thereby allowing the cell to bypass the requirement for ApbC to grow on Tcb. Because the effect was not limited to glucose, we tested whether cra (formerly fruR) function was involved. The catabolite repressor/activator (Cra) protein was initially characterized as the fructose repressor, FruR (9, 14). Cra acts as an activator or repressor, and its DNA-binding activity is abrogated by fructose-1-phosphate and fructose-1,6-bisphosphate (30). Cra regulates numerous genes involved in carbon and energy metabolism and is believed to regulate the direction of carbon flow (28, 32), but to our knowledge, there were no reports in the literature linking Cra to the regulation of iscRSUA operon expression.

For the sake of simplicity, all of the experiments were done with glucose as the representative saccharide that stimulated growth of the apbC strain on Tcb. Strain JE10441 (apbC cra) did not grow on Tcb + Gluc (Fig. 4) but grew on glucose (10 mM) alone (Fig. 4), suggesting that the inability of the apbC cra strain to catabolize Tcb was not due to a pleiotropic effect of Cra. Furthermore, strain JE10440 (apbC⁺ cra) grew on Tcb + Gluc (Fig. 4), indicating that in an ApbC-proficient strain, cra function was not required for Tcb catabolism. Strain 10441 (apbC cra/pcra⁺) grew on Tcb + Gluc (Fig. 4), although at a somewhat slower growth rate than the apbC strain alone (doubling time of 8 versus 6 h); the reasons for the observed slower growth are unclear.

FIG. 4.

The effect of glucose (Gluc) on growth of an apbC strain on Tcb requires cra function. cra represents strain JE10440 [cra426::MudJ(kan⁺)] and apbC cra represents strain JE10441 [apbC55::Tn10d(tet⁺) cra426::MudJ(kan⁺)]. pBAD30 was the cloning vector control. pcra⁺ represents plasmid pCRA1. The experiment was performed in triplicate; error bars denote standard deviations.

Cra directly regulates the expression of the iscRSUA operon of S. enterica.

One explanation for these results is that Cra can activate iscRSUA expression. To test this possibility, we constructed an apbC cra ΔiscR triple-mutant strain and assessed its growth on Tcb + Gluc. Data shown in Fig. 4 indicate that the lack of growth of the apbC cra strain on Tcb + Gluc was due to the repressive effect of IscR on iscRSUA expression. Although it is possible that Cra activates iscRSUA expression, the data in Fig. 4 show that expression of the operon in the absence of Cra is sufficient to support growth on Tcb. It is possible that further activation of iscRSUA by Cra is needed under some as-yet-unidentified growth conditions.

Cra protein modulates iscRSUA operon expression.

To determine whether Cra affected iscRSUA transcription, the 500 bp immediately 5′ to the iscR start codon were cloned into a promoter-less lacZ⁺ fusion in the low-copy-number vector pFZY1. The involvement of Cra in iscRSUA expression was clear. In an apbC⁺ cra strain (JE10440), the level of β-galactosidase activity was fivefold lower than in an apbC⁺ cra⁺ strain (DM10310) when cells were grown in glucose or Tcb + Gluc medium (Table 2). In spite of this sharp decrease in P_iscR-lacZ expression, growth of the cra strain on Tcb was not affected (Fig. 4), consistent with the idea that Cra was not required for Tcb catabolism. In contrast, P_iscR-lacZ expression in the apbC cra⁺ strain on Tcb + Gluc medium was nearly twofold higher than the level measured in the apbC⁺ cra⁺ strain (Table 2). This twofold increase in iscRSUA expression correlated with the ability of the apbC cra⁺ strain to grow on Tcb only when glucose was present in the medium (Fig. 2).

TABLE 2.

Cra affects iscRSUA expression^a

Expt	Relevant genotypeb	β-Galactosidase sp act (U/OD650) in cells grown on carbon and energy sourcec:
		10 mM Glc	10 mM Tcb + 1 mM Glc (growth)
1	apbC+cra+isc+ PiscR-lacZ+	70 ± 5	75 ± 1 (+)
2	apbC+cra isc+ PiscR-lacZ+	15 ± 1	15 ± 1 (+)
3	apbC cra+isc+ PiscR-lacZ+	110 ± 5	140 ± 15 (+)
4	apbC cra isc+ PiscR-lacZ+	125 ± 5	80 ± 5 (−)
5	apbC+cra+ ΔiscR PiscR-lacZ+	220 ± 14	200 ± 10 (+)
6	apbC+cra ΔiscR PiscR-lacZ+	710 ± 30	640 ± 70 (+)

^aThese experiments were performed three times with very similar results. The data in the table are those from one representative set of experiments.

^bAll strains carry an ara-9 mutation in their chromosome. In the strains used in experiments 1 to 4, isc⁺ represents iscRSUA⁺ hscBA⁺ fdx⁺ orf3. The strains used in experiments 5 and 6 had an in-frame deletion of iscR (ΔiscR), with the remaining genes of the isc locus (see above) remaining wild type. P_iscR-lacZ⁺ is encoded by the pISC2 plasmid.

^cA unit of β-galactosidase activity is defined in Materials and Methods. Cultures were grown to the mid-log phase (OD₆₅₀ of ∼0.4 to 0.5). Values in parentheses indicate increase (+) or decrease (−) versus growth on glucose alone. Each experiment was performed in triplicate. Means ± standard deviations are shown.

Inactivation of the cra gene in the apbC strain reduced P_iscR-lacZ expression to the levels measured in the apbC⁺ cra⁺ strain (Table 2), consistent with the lack of growth on Tcb + Gluc medium (Fig. 4). The abovementioned glucose effect on P_iscR-lacZ expression depended on a functional Cra protein (Table 2).

In summary, increased expression of the iscRSUA operon was required for growth on Tcb only in strains lacking ApbC, a result that was in agreement with our previous data (5).

One caveat of the P_iscR-lacZ expression experiments, in contrast to the growth experiments, was that the glucose concentration used in the P_iscR-lacZ expression experiments was held at 10 mM versus 1 mM for the growth experiments. The higher concentration of glucose in the expression experiments was required because the apbC cra strain would not grow on Tcb if the concentration of glucose was kept at 1 mM. We note that at 10 mM glucose, fructose-1,6-bisphosphate would accumulate to levels that would negatively affect Cra activity (32). Thus, these experiments may, in fact, underestimate the magnitude of the Cra effect on iscRSUA expression.

The combined absence of Cra and IscR derepresses P_iscR-lacZ expression.

When we assessed the expression of the P_iscR-lacZ reporter in a cra ΔiscR strain, the level of β-galactosidase activity was threefold higher than that in the cra⁺ ΔiscR strain (Table 2) and approximately ninefold higher than that in the isc⁺ cra⁺ strain (Table 2). These results were unexpected since they cast Cra as a repressor of iscRSUA expression, a role that would be inconsistent with the results of the growth experiments. At this point, we cannot rule out the involvement of other regulators, which in the absence of Cra and IscR may trigger iscRSUA expression. Further analysis is needed to dissect the regulatory network controlling the expression of the iscRSUA operon.

Putative Cra binding sites are present within the isc promoter.

Shown in Fig. 5 is an alignment of the iscR region with the consensus Cra DNA-binding motif (32). If the effect of Cra were direct, Cra would bind to one or more nonconsensus sites within the iscR region. Cra is known to bind to at least one nonconsensus binding site within the cydAB promoter (29).

FIG. 5.

Putative Cra binding sites within the promoter region of iscR. The promoter of iscR (bottom sequence) was aligned with the consensus Cra box (top sequence). Potential Cra binding sites are underlined. The ATG start codon of iscR is boxed. The transcription initiation site is labeled as +1 (36). IscR binding motifs are shaded (15). Arrows denote primers used for amplification of the fragment used for EMSA.

Cra binds in vitro to one or more of the putative Cra binding sites in the iscRSUA promoter.

To determine whether Cra could bind to the iscR promoter region, EMSAs were performed with purified Cra protein. A DNA fragment containing the three putative Cra binding sites upstream of the iscR transcriptional start site was used to assess Cra binding in vitro (Fig. 5). Indeed, Cra bound to the iscR promoter region, albeit with relatively low affinity (Fig. 6 and 7). A relatively large amount of Cra (278 pmol) was added to the reaction mixture before we observed binding of Cra to one site on the DNA (0.02 pmol); occupancy of the second and third sites occurred at higher Cra concentrations. The weak binding of Cra to the iscR promoter was not surprising given the lack of a consensus sequence. Furthermore, the relatively weak binding was consistent with our in vivo data, which suggested that the effect of Cra on iscRSUA expression is more subtle than the effect of IscR. Additionally, there is precedent for Cra interacting with other transcriptional regulators, thereby modulating its affinity for a site. For example, Cra binds to a nonconsensus site in the cydAB promoter region only when high (278 nM) concentrations of Cra and Fnr are both present (29). Detailed in vitro analyses are needed to dissect the role of Cra in iscRSUA expression and to determine whether other transcription factors modulate the Cra effect. This is, however, the first evidence of a regulator, other than IscR, involved in the regulation of the iscRSUA operon and the first evidence that the assembly of [Fe-S] clusters is activated in response to sources of carbon and energy whose catabolism is under the control of Cra.

FIG. 6.

Cra binds to the iscR promoter region in vitro. Cra DNA binding assays were performed on a 239-bp DNA fragment containing three putative Cra binding sites or a 330-bp positive control fragment containing the icd promoter region. Cra protein concentrations are given in terms of monomeric protein, although Cra is tetrameric.

FIG. 7.

Model for the roles of IscR and Cra in iscRSUA expression. In the model, IscR represses iscRSUA transcription when bound to a [2Fe-2S] cluster, while Cra can activate iscRSUA expression. Cra binding sites are numbered. Cra binding was only shown to occur at sites 3, 4, and 5. Whether Cra binds to site 1 and/or 2 remains an open question.