Bacillus cereus Fnr binds a [4Fe-4S] cluster and forms a ternary complex with ResD and PlcR
© Esbelin et al.; licensee BioMed Central Ltd. 2012
Received: 6 April 2012
Accepted: 11 June 2012
Published: 25 June 2012
Skip to main content
© Esbelin et al.; licensee BioMed Central Ltd. 2012
Received: 6 April 2012
Accepted: 11 June 2012
Published: 25 June 2012
Bacillus cereus is a facultative anaerobe that causes diarrheal disease in humans. Diarrheal syndrome may result from the secretion of various virulence factors including hemolysin BL and nonhemolytic enterotoxin Nhe. Expression of genes encoding Hbl and Nhe is regulated by the two redox systems, ResDE and Fnr, and the virulence regulator PlcR. B. cereus Fnr is a member of the Crp/Fnr family of iron-sulfur (Fe-S) proteins. Only its apo-form has so far been studied. A major goal in deciphering the Fnr-dependent regulation of enterotoxin genes is thus to obtain and characterize holoFnr.
Fnr has been subjected to in vitro Fe-S cluster reconstitution under anoxic conditions. UV-visible and EPR spectroscopic analyses together with the chemical estimation of the iron content indicated that Fnr binds one [4Fe-4S]2+ cluster per monomer. Atmospheric O2 causes disassembly of the Fe-S cluster, which exhibited a half-life of 15 min in air. Holo- and apoFnr have similar affinities for the nhe and hbl promoter regions, while holoFnr has a higher affinity for fnr promoter region than apoFnr. Both the apo- and holo-form of Fnr interact with ResD and PlcR to form a ternary complex.
Overall, this work shows that incorporation of the [4Fe-4S]2+ cluster is not required for DNA binding of Fnr to promoter regions of hbl and nhe enterotoxin genes or for the formation of a ternary complex with ResD and PlcR. This points to some new unusual properties of Fnr that may have physiological relevance in the redox regulation of enterotoxin gene regulation.
Bacillus cereus is a facultative anaerobic bacterium that can cause two types of food-borne illness in humans. Among these, the diarrheal syndrome may result from the production in the human host’s small intestine of various extracellular factors including hemolysin BL (Hbl) and nonhemolytic enterotoxin Nhe [1, 2]. The genes encoding Hbl and Nhe belong to the PlcR regulon . The ability of B. cereus to produce enterotoxins and grow well in an O2-limited environment such as that prevailing in the human small intestine is controlled by both the two-component system ResDE and the redox regulator Fnr. Unlike ResDE, Fnr is essential for B. cereus growth under anaerobic fermentative conditions and for hbl and nhe expression, irrespective of the oxygenation conditions [4, 5]. B. cereus Fnr is a member of the large Fnr/Crp superfamily of transcription factors that bind as homodimers to palindromic sequences of DNA, each subunit binding to one half-site . Like its homologue from Bacillus subtilis B. cereus Fnr contains a C-terminal extension with four cysteine residues, C(x4)C(x 2)C(x3)C. The last three cysteine residues were identified as [4Fe-4S]2+ cluster ligands in B. subtilis Fnr, the fourth ligand being an aspartate residue . The integrity of this oxygen-labile Fe-S cluster was found to be essential for the DNA binding activity of B. subtilis Fnr [7, 8]. By contrast, B. cereus Fnr appeared active in DNA-binding protein in its apo-form (cluster-free form). This has led to the conclusion that unlike its B. subtilis homologue, B. cereus Fnr is active in both its apo-form and its holo-form (bearing a Fe-S cluster) . However, data evidencing that B. cereus Fnr could coordinate a Fe-S cluster under anaerobiosis were lacking.
Here, we show that purified B. cereus apoFnr can bind one [4Fe-4S]2+ cluster per monomer upon incubation with iron, cysteine and cysteine desulfurase. Reconstituted Fnr (also referred to as holoFnr) showed enhanced DNA binding activity within the fnr promoter, but no activity difference with regard to the hbl and nhe promoters. Both the apo- and holo-form of Fnr interact with ResD and PlcR to form a ternary complex. Our results lend novel insight into the molecular control of enterotoxin gene expression in anaerobically-grown B. cereus.
This work has shown that the Fnr protein of B. cereus is homodimeric and can bind one [4Fe-4 S] iron-sulfur cluster per monomer. Our first challenge was to accurately assemble the Fe-S cluster via an enzymatic system since all our attempts to purify holoFnr under anaerobiosis failed. We demonstrated that CsdA from E. coli was capable of assembling the B. cereus Fnr Fe-S cluster. Interestingly, B. cereus synthesizes one pyridoxal 5-phosphate-containing enzyme (NP_834652)  that might be involved in Fe-S cluster biogenesis. When anaerobically reconstituted B. cereus Fnr was exposed to O2, we observed a rapid loss of the Fe-S cluster, demonstrating that Fnr functions as an oxygen sensor via its Fe-S cluster. Importantly, the cluster of the reconstituted B. cereus Fnr appeared extremely unstable, judging from its fast destruction on exposure to air. In this respect, the B. subtilis holoFnr, which is the closest homolog of B. cereus Fnr  displayed greater stability . Sequence comparison of the B. cereus and B. subtilis Fnr revealed a significant variation in the amino acid residues around the three C-terminal cysteine residues (C219-X 2-C222-X4-C227) that serve as ligands for the cluster (Additional file 3) . These observations imply that the occurrence of certain amino acid residues close to the cluster ligands may affect the stability of the B. cereus holoFnr, thus providing a possible explanation for its high susceptibility to oxygen damage . As a result, B. cereus Fnr might sense subtle changes in the redox status of the cells, a property that would reflect an adaptation of the pathogenic strain to the environment of its human host.
We proposed previously that B. cereus apoFnr binds promoter regions of enterotoxins only through the monomer pathway. In other words, we proposed that apoFnr was active as a DNA-binding protein only under its monomeric form . Here we showed that, when produced in a tag-less form, apoFnr is active as a DNA binding protein under its dimeric form. In addition, we showed that dimeric apoFnr-DNA complexes were stable in contrast to what we observed previously . We conclude that (i) in our previous studies, tags fused at the N-terminus and C-terminus of Fnr introduced steric hindrance that affected its oligomeric structure and/or DNA binding activity and (ii) B. cereus apoFnr may bind DNA both through the dimer and the monomer pathway under aerobiosis unlike its homologues of B. subtilis and E. coli. There are probably many variables affecting the choice for a monomer or dimer recognition pathway in vivo. Among them, there is the redox state of the cell that may impact directly the ratio of monomeric to dimeric apoFnr since we observed that the addition of reductant (DTT) affected the dimerization state of apoFnr in solution. Finally, the mechanism of apoFnr dependent regulation of enterotoxin is undoubtedly complex, and further extensive experiments are required to examine the role of the monomer and dimer pathways.
In conclusion, this work brings further evidence that B. cereus Fnr, unlike its counterpart from B. subtilis, is an active transcriptional regulator in both its apo- and holo- forms. This property may enable B. cereus to ensure optimal enterotoxin gene expression in response to changes in oxygen tension such as those encountered during infection of the human host.
Escherichia coli strain TOP10 (Invitrogen) was used as the general cloning host, and strain BL21 CodonPlus(DE3)-RIL (Stratagene) was used to overexpress fnr and resD. E. coli strain BL21λDE3, containing the pRep4 plasmid  was used to overexpress plcR. E. coli strains were routinely grown in Luria broth at 37°C.
The coding sequence for B. cereus fnr was PCR amplified from F4430/73 genomic DNA using primers PET101F (5'-CACCATGACATTATCTCAAG-3') and PET101R (5'-CTAATCAATGCTACAAACAGAAGC-3'). The amplicon was cloned as a blunt-end PCR product into pET101/D-TOPO (Invitrogen), yielding pET101fnr. B. cereus Fnr was produced as a recombinant protein in aerobically grown E. coli BL21(pET101fnr). The culture was grown in a 6-litre fermentor until OD600 reached ~1.0, and protein overproduction was then triggered by 0.2 mM isopropyl-ß-d-thiogalactopyranoside (IPTG). After incubation for 16 h at 20°C, cells were harvested by centrifuging at 10,000 × g for 15 min. Fnr was then purified as follows: the bacterial pellet was resuspended in 120 ml of buffer C (25 mM Tris–HCl [pH 8], 1 mM dithiothreitol (DTT)) and incubated with 0.2 mg.ml-1 of lysozyme and 5 mM EDTA for 10 min at 30°C. Cells were lysed by ultrasonication for 3 min using a Vibra cell ultrasonifier (Fisher Bioblock Scientific). Cell debris and membrane particles were removed by centrifuging at 43,000 × g for 1 h, and the resulting supernatant was loaded on a 30 ml DEAE-cellulose column (DE52; Whatman) equilibrated with buffer C. The non-retained fraction was adjusted to pH 7 with 1 M KH2PO4 and then loaded onto a 20 ml hydroxyapatite column (HA Ultrogel; Pall Corporation) equilibrated with buffer D (50 mM KH2PO4 [pH 7], 1 mM DTT). The column was developed with a linear gradient from 50 to 200 mM KH2PO4 at a flow rate of 2 ml/min. Fractions containing recombinant Fnr were pooled and concentrated by ultrafiltration through an Omega disc membrane (30 kDa cutoff, Ø 43 mm, Pall Corporation). A polishing step was then carried out by gel filtration on a column of Superdex SD200 (1.5 × 60 cm, Amersham Biosciences) equilibrated with buffer F (25 mM Tris–HCl [pH 7.5]1, 50 mM NaCl, 1 mM DTT). The purified protein, >90% pure by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; Additional file 1), was stored as pellets in liquid nitrogen.
The following procedure was carried out under anoxic conditions (O2 < 1 ppm) in a glove box maintained at 18°C (Jacomex, France). All buffers were degassed under argon and equilibrated for at least 16 h in the glove box before use. ApoFnr (2 g/L) was incubated with 1 μM cysteine desulfurase CsdA from E. coli, 1 mM l-cysteine, and 1 mM Fe(NH4)2(SO4)2 (Sigma-Aldrich) in the presence of 4 mM DTT in buffer F. Formation of the cluster was monitored by UV-visible spectroscopy using a Uvikon spectrophometer (Kontron) connected through optic fibers to the cuvette holder inside the glove box. The reaction was initiated by adding CsdA, and reached completion after 2 h (no further increase in the absorption at 416 nm). The protein was run through a 10 ml Sephadex G25 column (Amersham Biosciences) equilibrated in buffer F to remove excess reagents, and then concentrated by ultrafiltration using a Nanosep device with a molecular cutoff of 30 kDa (Pall Corporation).
Protein concentrations were determined by either a bicinchoninic acid (BCA) assay (Pierce) or a Biuret method insensitive to thiols . Bovine serum albumin (BSA) was used as a standard. The presence of the Fnr protein in purification fractions was monitored by SDS-PAGE , followed by Coomassie blue staining. The iron content of holoFnr was determined spectrophotometrically using a method adapted from Blair and Diehl . Briefly, 50 μl samples of holoFnr (2.8 g/L) were incubated at 100°C for 15 min with 30 μL of 6 N HCl. After dilution to 0.5 ml with H2O, samples were centrifuged at 12,000 × g for 5 min, and 100 μl aliquots of the supernatant fractions were mixed with 0.65 ml of 0.5 M Tris–HCl pH 8.5, 50 μl of 5% ascorbate and 0.2 ml of 0.1% bathophenanthroline (Sigma-Aldrich). Mixtures were incubated at room temperature for 1 h, and the absorbance was measured at 536 nm (ϵ 536 = 22.14 mM-1 cm-1) and compared with a blank lacking holoFnr.
Samples were prepared in an anaerobic glove box at 18°C. HoloFnr (0.1 mM) was tentatively reduced with 10 μM 5-deazaflavin (a gift from Prof J. Knappe, Heidelberg University, Germany) in the presence of 2.5 mM glycine as electron donor. Photoreduction was carried out in a 0.2 cm light path cuvette by exposing the protein sample to the light of a slide projector for 1 min time periods. Chemical reduction was also applied with an excess of sodium dithionite (2 mM) at pH 8.5. Progression of the reaction was monitored by recording UV-visible absorption spectra in the 300–700 nm range.
Samples were transferred into EPR tubes and immediately frozen in liquid nitrogen. EPR spectra were recorded at 10 K using a Bruker EMX spectrometer equipped with an Oxford Instruments ESR900 liquid helium cryostat.
To assess the sensitivity of holoFnr to oxygen, a fraction of the reconstituted protein was removed from the glove box and exposed to air. Absorbance spectra were recorded at time intervals with an HP8452 diode-array spectrophotometer (Agilent).
Far-Western assays and cross-linking reactions were carried out in an anaerobic glove box as described previously []. Revelation in Far-Western assays used biotinylated PlcR or biotinylated ResD. The cross-linked products were analyzed by 12% SDS-PAGE and detected by Western blotting using anti-Fnr and anti-ResD antibodies.
EMSAs were performed in an anaerobic glove box. Fragments containing the promoter regions of fnr hbl, and nhe were PCR-amplified and end-labeled with the following biotinylated primer pairs: FnrFbiot (5'-CGAACACTTCAGCAGGCATA-3') and FnrR (5'-AATGTCATACTGTTTGCCAC-3'), Hbl1Fbiot (5'-GGTAAGCAAGTGGGTGAAGC-3') and Hbl1R (5'-AATCGCAAATGCAGAGCACAA-3'), Hbl2Fbiot (5'-TTAACTTAATTCATATAACTT-3') and Hbl2R (5'-TACGCATTAAAAATTTAAT-3'), NheFbiot (5'-TGTTATTACGACAGTTCCAT-3') and NheR (5'-CTGTAACCAATAACCCTGTG-3'), respectively. DNA fragment used as negative control was part of sequence BC0007 (NC_004722) and was amplified with the biotinylated primer pairs: F16biot (5’-GGTAGTCCACGCCGTAAACG-3’) and R16 (5’-GAAAACCATGCACCACCTG-3’). The 5’-labeled amplicons were purified using the High Pure PCR Product Purification Kit (Roche). Binding reactions were performed for 30 min at 37°C by incubating biotin-labeled DNA fragments (2 nM per reaction) with the indicated amount of purified apo- or holoFnr (0.2, 0.4, 0.6 and 0.8 μM) in 10 mM Tris–HCl [pH 7.5] buffer containing 50 mM KCl, 1 mM DTT, 2.5% glycerol, 5 mM MgCl2 and 5 mg/L of poly(dI-dC). The samples were resolved by electrophoresis on a 6% non-denaturing polyacrylamide gel  and electrotransferred onto Nylon membranes (Amersham Hybond N+). Biotin-labeled DNAs were detected using the LightShift Chemiluminescent EMSA Kit (Pierce).
B. cereus F4430/73 protein lysates were prepared as follows: anaerobically-grown cells were harvested by centrifuging, washed twice with phosphate-buffered saline (PBS; 0.14 M NaCl, 2.68 mM KCl, 10.14 mM Na2HPO4, 1.76 mM KH2PO4 [pH 7.4]), resuspended in lysis buffer (10 mM Tris, 1 mM EDTA, [pH 8]), and mechanically disrupted using a FastPrep instrument (FP120; Bio101, Thermo Electron Corporation). Cell debris were removed by centrifuging (3500 × g, 10 min, 4°C). The protein lysate was then filtered through a 0.22 μm membrane; 100 μl of cleared lysate was incubated with 50 μl of anti-Fnr protein A-coated Dynabeads prepared by mixing 50 μl of polyclonal anti-Fnr  with 50 μl of protein A Dynabeads (Dynal). The beads were pelleted by centrifuging, washed three times with PBS buffer, and suspended in 20 μl of loading buffer. Samples were either directly analyzed by non-denaturing PAGE, or boiled and subjected to 12% SDS-PAGE. Resolved proteins were transferred to a nitrocellulose membrane (Amersham Bioscience) according to standard procedures (Bio-Rad). Membranes were probed with 1:2,000, 1:1,000 and 1:2,000 dilution of polyclonal rabbit sera raised against Fnr, ResD and PlcR, respectively [9, 11, 24]. The blotted membranes were developed with 1:2,000 dilution of goat anti-rabbit IgG peroxidase-conjugate (Sigma-Aldrich) and an enhanced chemiluminescence substrate (Immobilon Western, Millipore).
Electrophoretic Mobility Gel Shift Assay.
We thank D. Lereclus for kindly providing plasmids for recombinant expression of plcR and Stephen H. Leppla for sending us anti-PlcR antibodies. We thank E. Mulliez for the gift of purified CsdA, and S. Ollagnier and E. Mulliez for their help in cluster reconstitution experiments. We also thank N. Duraffourg for recording and comments on the EPR spectra.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.