WhiB7, an Fe-S-dependent Transcription Factor That Activates Species-specific Repertoires of Drug Resistance Determinants in Actinobacteria*

Background: WhiB7 is essential for antibiotic resistance in M. tuberculosis. Results: WhiB7 requires conserved residues, including a redox-sensitive center and DNA-binding motif, to coordinate transcription of species-specific drug resistance genes in diverse Actinobacteria. Conclusion: WhiB7 activates species-specific drug resistance genes in Actinobacteria. Significance: Understanding WhiB7 activity may allow the development of drugs that sensitize bacteria to antibiotics. WhiB-like (Wbl) proteins are well known for their diverse roles in actinobacterial morphogenesis, cell division, virulence, primary and secondary metabolism, and intrinsic antibiotic resistance. Gene disruption experiments showed that three different Actinobacteria (Mycobacterium smegmatis, Streptomyces lividans, and Rhodococcus jostii) each exhibited a different whiB7-dependent resistance profile. Heterologous expression of whiB7 genes showed these resistance profiles reflected the host's repertoire of endogenous whiB7-dependent genes. Transcriptional activation of two resistance genes in the whiB7 regulon, tap (a multidrug transporter) and erm(37) (a ribosomal methyltransferase), required interaction of WhiB7 with their promoters. Furthermore, heterologous expression of tap genes isolated from Mycobacterium species demonstrated that divergencies in drug specificity of homologous structural proteins contribute to the variation of WhiB7-dependent drug resistance. WhiB7 has a specific tryptophan/glycine-rich region and four conserved cysteine residues; it also has a peptide sequence (AT-hook) at its C terminus that binds AT-rich DNA sequence motifs upstream of the promoters it activates. Targeted mutagenesis showed that these motifs were required to provide antibiotic resistance in vivo. Anaerobically purified WhiB7 from S. lividans was dimeric and contained 2.1 ± 0.3 and 2.2 ± 0.3 mol of iron and sulfur, respectively, per protomer (consistent with the presence of a 2Fe-2S cluster). However, the properties of the dimer's absorption spectrum were most consistent with the presence of an oxygen-labile 4Fe-4S cluster, suggesting 50% occupancy. These data provide the first insights into WhiB7 iron-sulfur clusters as they exist in vivo, a major unresolved issue in studies of Wbl proteins.

WhiB-like proteins (Wbl) are found as multiple paralogs in Actinobacteria (1), where they play diverse roles in essential functions, as well as antibiotic resistance (2). The best known Actinobacteria include Streptomyces, which produce the majority of known antibiotics (3), and Mycobacterium tuberculosis, a leading cause of human mortality due to bacterial infection. Streptomyces sp. display diverse species-specific drug resistance patterns (4), which may be related to the evolution of thousands of antibiotic biosynthetic pathways in their genomes (3). Other saprophytic Actinobacteria such as Mycobacterium smegmatis (5) and Rhodococcus jostii may require intrinsic drug resistance determinants to survive in various soil ecosystems that may contain antibiotics and toxins produced by competing organisms.
Chemotherapeutic options for tuberculosis are limited by the efficient intrinsic antibiotic resistance system of M. tuberculosis (6,7). Many of these resistance phenotypes require WhiB7 (7) and are linked to redox metabolism (8). Disruption of whiB7 sensitizes M. tuberculosis to several antibiotics with different chemical structures and mechanisms of action, whereas overexpression promotes resistance (7). More extensive screens using M. smegmatis as a model system identified other whiB7 activators, including compounds that perturb respiration, redox balance, transmembrane ion flux, as well as heat shock and iron starvation (9,10). Antibiotic exposure caused autoinduction of the whiB7 promoter and a whiB7-dependent increase of cellular thiol reducing power (9). Some of these inducers may alter cellular redox conditions comparable with those encountered by M. tuberculosis during host macrophage invasion or chemotherapeutic treatments. Indeed, whiB7 is highly up-regulated in macrophages (11) and in the lungs of infected animals (12). WhiB7 activates expression of its own regulon that includes resistance genes involved in antibiotic efflux (tap, Rv1258c) (13), ribosome modification (erm, Rv1988) (14,15), modulation of the host immune responses (eis, Rv2416c, and erm) (16,17), and mycobacterial survival within macrophages (eis) (7).
In addition to drug resistance, Wbl genes have roles in morphogenesis, cell division, virulence, and both primary and secondary metabolism. Two members of the family, whiB and whiD, were first identified in Streptomyces coelicolor where they play essential roles in sporulation. Wbl proteins contain around 100 amino acids. They are characterized by four conserved cysteine residues (18) that can act as ligands for iron (19), a tryptophan/glycine-rich motif, predicted to form a ␤-turn, and positively charged amino acids at their C termini (18) that match a sequence (RKRPRGRPRK) of a peptide that binds to AT-rich DNA sequences (AT-hook) (20). The AT-hook domain is found in eukaryotic non-histone chromosomal proteins (HMGAs) that have roles in chromatin architecture and transcriptional regulation (21). The cysteine motif resembles those known to coordinate redox-active Fe-S clusters. Indeed, many laboratories have demonstrated subpopulations of Fe-S clusters in Wbl proteins purified aerobically, often under denaturing conditions (19,(22)(23)(24)(25). These preparations can be used to reconstitute iron-sulfur clusters; however, proteins assembled in this way contain substoichiometric concentrations of iron (19,22,23,26). To our knowledge, there is only one report describing anaerobic purification of a soluble Wbl protein, presumably in its native form (27). Anaerobically purified WhiD contains a 4Fe-4S cluster having about 4 mol of iron per monomer (Ͼ70% occupancy). Mutagenesis experiments have shown that the conserved cysteines are needed for the biological functions of WhiB3 (25), WhiB4 (26), WhiD (19), WhiB2 (28), and the WhiB homolog of mycobacteriophage TM4 (22). However, the possibility that these mutant proteins are unstable has not been systematically addressed (except in one case (28)).
Although genetic and microarray studies suggest that Wbl proteins function as transcription regulators (7,29), some members of this family may provide protein-disulfide reductase (23) or chaperone (30) activities. Direct biochemical evidence (transcriptional run off experiments) has assigned transcriptional regulatory functions to WhiB proteins, including whiB1 (31,32), a repressor, and whiB7, an activator (33). In vitro transcriptional run-off studies have shown that the M. smegmatis WhiB7 protein is a redox-sensitive transcriptional activator of its own promoter (33). The whiB7 promoter, conserved across mycobacteria and other Actinobacteria, includes an AT-rich motif directly upstream of its Ϫ35 hexamer that is targeted by WhiB7 to promote transcription (9,33). AT-rich motifs are also found upstream of other promoters in the whiB7 regulon (9) and are necessary for in vitro transcriptional activation (33). WhiB3 or WhiB7 function depends on their direct interactions with the vegetative factor RpoV (33,34), and at least one Streptomyces factor gene encodes an N-terminal extension homologous to Wbl proteins (35).
Here, we show that WhiB7-mediated multidrug resistance spectra in Streptomyces lividans (closely related to S. coelicolor having an identical whiB7 sequence), M. smegmatis, and R. jostii are dependent on genome-specific resistance determinants and that WhiB7 activity requires the consensus sequence motifs of the Wbl protein family as well as its distinctive AT-hook DNA binding domain. In addition, the Streptomyces WhiB7 protein was, for the first time, anaerobically purified as dimers that coordinated a fully reduced, oxygen-sensitive Fe-S cluster.

Bacterial Strains, Plasmids, Growth Conditions, and Chemicals
Strains and plasmids used in this study are described in Table  1. The sequences of the oligonucleotides used are available upon request. M. smegmatis was routinely grown at 37°C in Middlebrook 7H9 broth (Difco) supplemented with 10% Middlebrook albumin/dextrose/catalase (Difco), 0.5% glycerol, and 0.05% (v/v) tyloxapol or on Middlebrook 7H10 agar plates (Difco) supplemented with 10% (v/v) oleic acid albumin/dextrose/catalase (Difco). R. jostii RHA1 and S. lividans were grown at 30°C in NE medium (36). S. lividans spores were prepared as described elsewhere (37) and stored at Ϫ20°C. Escherichia coli was grown at 37°C in LB broth or on LB agar plates. Plasmids were maintained in E. coli with appropriate antibiotics for selection (100 g/ml ampicillin and 20 g/ml kanamycin). For the selection of resistance markers in M. smegmatis and R. jostii, hygromycin (50 g/ml) or kanamycin (30 g/ml) and apramycin (50 g/ml), respectively, were added to the cultures. Antibiotics and chemicals used in sensitivity tests were obtained from Sigma.

Plasmid Construction
DNA manipulations were carried out using standard techniques (38). E. coli, M. smegmatis, and R. jostii were transformed by electroporation with a Gene Pulser Xcell TM (Bio-Rad).
N-terminal Truncations of the whiB7 ST Gene-Truncated forms of the whiB7 gene from S. coelicolor were constructed from pMV361:B7 ST . Two truncated forms were generated at nucleotide position 28 (valine; Val-10) and at nucleotide posi-tion 85 (proline; Pro-29). The DNA fragments contained engineered EcoRI and HindIII restriction sites that allowed cloning into pMV361 vector. The generated plasmids were transformed into M. smegmatis mc 2 6 ⌬whiB7.
Mutagenesis of the whiB7 ST Gene-Five cysteine-motif mutated alleles of whiB7 ST (C49S, C72S, C75S, C81S, and C72S/C75S) were constructed by a two-stage PCR mutagenesis, cloned into pGEM-T Easy, and their sequences verified. The inserts were removed by EcoRI/HindIII digestion and cloned into pMV361 (pLN6, pLN7, pLN8, pLN9, and pLN10). An inframe deletion of whiB7 ST was created in which codons encoding eight amino acids of the glycine/tryptophan-rich motif (WGVWGGEL) were removed. In addition, three mutated whiB7 ST alleles were generated with mutations in the AT-hook motif and cloned into pMV361 (pLN1, pLN2, pLN3, pLN4, and pLN5). The sequences of all mutant alleles were confirmed by sequencing.
Mutagenesis of C-terminal His-tagged whiB7 ST -The whiB7 ST gene was PCR-amplified from pB00-1 (pET19b: whiB7 ST ; see below) eliminating the whiB7 ST stop codon. The DNA fragment was cloned into the NdeI and BglII sites of plasmid pET30b (Novagen). The entire whiB7 ST gene fused to a C-terminal hexahistidine tag was amplified and subcloned into the PstI and ClaI sites of pMV261. Five cysteine-motif mutated alleles of whiB7 ST (C49S, C72S, C75S, C81S, and C72S/C75S) were constructed by site-directed mutagenesis of the C-terminal His-tagged whiB7 ST gene cloned in pMV261 using QuikChange Lightning site-directed mutagenesis kit (Stratagene). The presence of all mutant alleles was confirmed by DNA sequencing. The resulting plasmids, both wild type and cysteine mutant constructs, were electroporated into M. smegmatis mc 2 155 ⌬whiB7.
Cloning of His-whiB7 ST in E. coli-The whiB7 ST gene was amplified from S. coelicolor A3(2) genomic DNA. The DNA

Strain Constructions
The S. lividans whiB7 mutant has been reported previously (7). Replacement of whiB7 (MSMEG_1953) by a hygromycin resistance gene in M. smegmatis mc 2 6 was performed by specialized transduction (40). The whiB7-5Ј (660 bp) and -3Ј (735 bp) flanking regions were PCR-amplified and inserted on either side of the resistance marker gene in plasmid pYUB854. Hygromycin-resistant colonies were analyzed, and the replacement was confirmed by recovering the mutant locus by PCR using primers flanking the allelic exchange substrate. The mutant DNA generated a larger PCR fragment corresponding to the inserted hygromycin cassette (data not shown). The hygromycin resistance chromosomal marker was then transferred by generalized transduction from M. smegmatis mc 2 6 to M. smegmatis mc 2 155 using bacteriophage Bxz1 (41). M. smegmatis mc 2 155 is a high frequency transformation mutant of M. smegmatis mc 2 6 (42) that was used to confirm the genetic background of the mutation. Briefly, 250 ng of a mycobacterial replicative plasmid with a kanamycin resistance marker was electroporated into competent cells of M. smegmatis mc 2 6 and mc 2 155 and whiB7 mutant derivatives. Transformants were selected on 7H10 agar medium containing 30 g/ml kanamycin. In the mc 2 155 background, an average of 2 ϫ 10 5 transformants/g of DNA could be recovered from both wild type and whiB7 mutant strains, although only 300 transformants were recovered using the mc 2 6 background in both wild type and whiB7 mutant strains. These results confirmed the whiB7 mutation in the mc 2 155 host background. Replacement of the whiB7 gene was confirmed by PCR as described above (data not shown).
Replacement of the whiB7 gene in R. jostii RHA1 (ro06383) by an apramycin resistance cassette was done as described previously (43). Briefly, in a first step the chloramphenicol resistance gene of fosmid RF00120I22 was replaced by a hygromycin resistance gene. This fosmid harbors a 34.2-kb insert of genomic RHA1 DNA containing whiB7 (44). In a second step, the whiB7 gene was replaced by an apramycin resistance cassette, yielding the RFMD16 fosmid. RFMD16 was conjugated into RHA1 cells. Allelic exchange between the fosmid and the chromosome resulted in replacement of whiB7 with an apramycin resistance cassette. To confirm replacement of the whiB7 gene in R. jostii, apramycin-resistant and hygromycinsensitive double crossover ex-conjugant colonies were selected and analyzed by PCR using combinations of primers homologous to the whiB7 gene, flanking the replacement locus and within the apramycin resistance gene cassette. In the mutant, both 5Ј and 3Ј junctions of the replacement locus were verified using primers for flanking regions and those within the apramycin resistance gene cassette. Substitution of whiB7 with the apramycin resistance gene cassette was further confirmed by Southern blot analysis. Genomic DNA (500 ng) of the whiB7 mutant and the wild type R. jostii RHA1 was digested with EcoRI and probed with digoxigenin-labeled 1.45-kb PCR fragment amplified from the fosmid RF0012OI22. The mutant and the wild type strains generated the expected hybridization signals (data not shown).
For homologous and heterologous complementation experiments in M. smegmatis, wild type mc 2 155 and whiB7 mutant strains were transformed with the integrative plasmid pMV361 harboring the whiB7 genes from M. tuberculosis, M. smegmatis, R. jostii, or S. lividans under the control of the constitutive mycobacterial hsp promoter. For whiB7-dependent drug resistance experiments, the erm (37) and tap genes from M. tuberculosis and the tap gene from Mycobacterium fortuitum were cloned under the control of their native promoters (including 500 -700 bp upstream of the annotated start codon of the gene) into a promoter-less site of the pSUM36 vector (45), or under the control of the mycobacterial hsp constitutive promoter of the pMV361 vector (cloning the functional gene from the annotated start codon). Constructs were introduced into M. smegmatis mc 2 155 and corresponding whiB7 mutant. For complementation experiments in R. jostii RHA1, whiB7 mutant strains were transformed with the replicative plasmid pTIP-QC1 (46) harboring the whiB7 genes from M. tuberculosis, M. smegmatis, R. jostii, or S. lividans under the control of the thiostreptoninducible tipA promoter.

Western Blot Analyses of C-terminal His-tagged WhiB7 Mutants
Plasmids pMV261 harboring wild type and cysteine mutant constructs of C-terminal His-tagged whiB7 ST were transformed into the mc 2 155 whiB7 mutant and grown in 7H9 until the cell density reached an A 600 of ϳ1. Cultures (1 ml) were then subjected to a 45°C heat shock treatment for 30 min in a water bath. Cells were chilled on ice, harvested by centrifugation at 4,000 rpm for 15 min, resuspended in 100 l of SDS loading buffer, boiled at 95°C for 10 min, and placed for 1 min on ice. The mixture was centrifuged to pellet the insoluble materials, and 3 l of the sample supernatant was loaded onto a 15% SDS-polyacrylamide gel. Proteins were transferred to nitrocellulose membranes using a semi-dry transfer apparatus (Bio-Rad) at 12 V for 75 min. Transfer Buffer was composed of 25 mM Tris, 190 mM glycine, and 20% methanol at pH 8.5. Following transfer to a nitrocellulose membrane, the membrane was blocked overnight at 4°C with Blocking Buffer for fluorescent Western blotting (MB-070, Rockland Immunochemicals). After three washing steps of 20 min each in PBS containing 0.05% Tween 80, the blot was incubated for 3 h in the primary antibody solution composed of a mouse anti-His 6 tag antibody (catalog no. 200-301-382, Rockland Immunochemicals) at 0.5 g/ml in Blocking Buffer containing 0.05% Tween 80. The blot was washed three times for 20 min with PBS plus 0.05% Tween 80, incubated for 3 h with the secondary antibody (goat antimouse IgG F(c) Antibody DyLight TM 680-conjugated, catalog no. 610-144-003, Rockland Immunochemicals), diluted to 0.2 g/ml in Blocking Buffer with 0.05% Tween 80, and then washed three times for 20 min each with PBS plus 0.05% Tween 80. The presence of the hexahistidine epitope on the fusion WhiB7 ST was visualized using a Li-Cor Odyssey Fluorescent Imager. NOVEMBER 29, 2013 • VOLUME 288 • NUMBER 48

WhiB7 ST Expression, Purification, and Characterization
For expression of WhiB7 ST , electrocompetent E. coli BL21(DE3) cells were co-transformed with plasmids pB00-1 and pRKISC, harboring isc genes to promote iron-sulfur cluster assembly (39). Transformants were grown on LB agar containing ampicillin (100 g/ml) and tetracycline (20 g/ml). Single colonies of transformants were inoculated into LB broth containing antibiotics for plasmid maintenance and grown overnight at 37°C and 200 rpm. The starter culture was diluted 200-fold into 1 liter of LB media containing 100 g/ml ammonia iron(III) citrate and selective antibiotics and grown at 37°C with shaking until they reached an A 600 of ϳ0.5. Cultures were then transferred to a water bath, and the temperature increased to 42°C to induce expression of heat shock chaperone proteins. After 30 min, culture temperature was quickly lowered to 21°C, and expression of WhiB7 ST was induced using 0.1 mM isopropyl 1-thio-␤-D-galactopyranoside. After overnight expression (16 -18 h), cells were harvested by centrifugation at 4,000 rpm for 15 min and frozen at Ϫ80°C. For purification, frozen cells were gently thawed at 4°C overnight and then resuspended in Lysis Buffer (50 mM Tris/HCl, pH 7.6, plus 1 mg of DNase I (Roche Applied Science), 10 l of 1 M CaCl 2 , and 10 l of 1 M MgCl 2 per 10 ml of buffer) at a ratio of 10 ml of Lysis Buffer per 1 liter of the original culture. Resuspended cells were lysed by passage through a pressure homogenizer at 10,000 p.s.i. until a clear homogenate was observed, transferred to an ultracentrifugation tube, and immediately bubbled with nitrogen for 30 min. Oxygen-degassed homogenates were ultracentrifuged using an Optima L-90k (Beckman) centrifuge with a type 70 Ti rotor at 45,000 rpm for 45 min at 4°C. From this point, all procedures were performed inside a LabMaster model 100 glove box (M. Braun, Inc., Peabody, MA) operated under anaerobic conditions (O 2 level Ͻ5 ppm). All buffer solutions and nickel-agarose beads were oxygen-degassed overnight inside the globe box before use. A nickel affinity chromatography strategy was used to purify WhiB7 ST . Nickel-agarose beads were equilibrated with Buffer A (50 mM Tris/HCl, pH 7.6, 20 mM imidazole). Supernatants were passed through a 0.45-m filter before loading onto the nickel column and incubated for 30 min. Three washing steps were performed with 4 column volumes of Buffer B (Buffer A plus 300 mM NaCl), Buffer C (50 mM Tris/HCl, pH 7.6, 50 mM imidazole), and Buffer D (50 mM Tris/HCl, pH 7.6, 300 mM imidazole). WhiB7 ST was eluted in the 300 mM imidazole fraction. To remove imidazole, the protein-containing fraction was dialyzed three times against a buffer containing 20 mM Tris/HCl, pH 7.6, and 2 mM DTT using an Amicon filter (Millipore; 10-kDa pore size) and concentrated to 10 mg/ml. The purified protein was flash-frozen by adding solution drops into liquid nitrogen and stored at Ϫ80°C until further use. Prior to analysis, two or three beads of purified protein were desalted into the corresponding assay buffer using a G-25 gel filtration matrix, previously incubated overnight inside the glove box to eliminate any trace of oxygen. Purification of N-terminal histidine-tagged WhiB7 ST under anaerobic conditions yielded 15 mg of protein from 1 liter of culture.
Expression and purification of WhiB7 ST was monitored by electrophoresis using a Bio-Rad Power Pac 3000 apparatus with 16% Tricine 4 /SDS-polyacrylamide resolving gels (47). Protein bands were visualized using Coomassie Blue staining (GelCode Blue Stain Reagent).
For molecular weight determinations, size exclusion chromatography was performed using an ÄKTA Explorer (GE Healthcare). Anaerobically purified WhiB7 ST was loaded onto a HiLoad 26/60 Superdex 75 column (GE Healthcare) operated at an average flow rate of 3 ml/min. Oxygen within the column was purged by applying 1 column volume of the following: (i) anaerobic water, (ii) anaerobic water containing 2 mM of sodium dithionite, and (iii) anaerobic Running Buffer (20 mM Tris/HCl, pH 7.8, 200 mM NaCl, 1 mM DTT). The column was calibrated using low molecular mass standards as follows: cytochrome c (13.6 kDa), carbonic anhydrase (32 kDa), ␤-lactoglobulin (36 kDa), and bovine serum albumin (68 kDa). The calibration curve was constructed by plotting K av versus log M r , where where V e is the measured elution volume of each standard; V 0 is the void volume of the column determined by V e of blue dextran, and V t is the total volume as specified by the manufacturer. Protein elution was followed by monitoring absorbance (nm) at 280 and 413.
For spectroscopic analysis of anaerobically purified WhiB7 ST , UVvisible absorption spectra were recorded using a Cary 5000 spectrophotometer equipped with a thermojacketed cuvette holder (Varian, Walnut Creek, CA). Samples (protein concentration ϳ15 M) were assayed in 20 mM MOPS, 80 mM NaCl, pH 7.6 buffer. A molecular mass of 16 kDa for the N-terminal His-tagged WhiB7 ST protein was used to calculate molar concentrations. Protein concentrations were determined using the Bradford assay with bovine serum albumin as a standard. Acidlabile sulfur content of samples was determined colorimetrically using the N,N-dimethylparaphenylenediamine assay (48). Iron content was determined using the Ferene S assay (49).

Drug Sensitivity Assays
The MICs of the compounds were determined using either resazurin or 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assays (50). Briefly, serial 2-fold dilutions of the compounds were assayed in NE medium. The initial inoculum for M. smegmatis and R. jostii was 10 5 cells/ml; 10 6 spores/ml were used for S. lividans. Sensitivity dilution end points for M. smegmatis, R. jostii, and S. lividans were recorded at 96, 48 (or 72), or 24 h, respectively. The MIC value was defined as the lowest drug concentration that prevented a change in dye color. For some compounds, values were also confirmed in NE agar by a conventional antibiotic-containing disc assay (Pasteur Diagnostics) or using E-test strips (AB Biodisks) and scored after 48 -72 h of growth at 37°C. For complementation assays, fold changes in the sensitivity of a given engineered strain are expressed with respect to the appropriate control strain. In the case of R. jostii, genes cloned under the control of the tipA promoter were induced with thiostrepton (0.5 g/ml) for 1 h at 30°C. Cultures were then added to 2-fold serial dilutions of erythromycin and tetracycline at a 1:1 ratio. Experiments were done in triplicate, each from three independent samples.

RESULTS
Inactivation of whiB7 Generates Species-specific Drug Sensitivity Profiles in Actinobacteria-A representative collection of 38 drugs and chemicals was used to analyze the resistance profiles and possible physiological roles of the whiB7 gene using mutants constructed in different Actinobacteria. Orthologous whiB7 genes were disrupted in M. smegmatis, R. jostii, and S. lividans. Chemicals were selected to specifically inhibit different cellular processes such as cell wall assembly, DNA replication, transcription, translation, or redox balance. Quantitative analysis showed that the whiB7 disruptions generated a unique sensitivity pattern in each of the three species (Fig. 1). All three whiB7 mutants exhibited increased sensitivities to various antibiotics targeting protein biosynthesis. Interestingly, the increase in macrolide (azithromycin, clarithromycin, erythromycin, roxithromycin, and spiramycin) sensitivity caused by the mutation varied dramatically from 4-to 512-fold, depending on the species. Similarly, but to a lesser extent, mutants became more sensitive to other protein synthesis-targeting compounds such as aminoglycosides, tetracycline, lincomycin, fusidic acid, and phenicols (Fig. 1).
R. jostii and S. lividans whiB7 mutants also displayed sensitivity to drugs targeting other essential cellular processes (Fig.  1), including compounds that disrupt fatty acid biosynthesis, DNA replication, and redox balance. Cerulenin, a fatty-acid synthase inhibitor, was 256 times more active against the R. jostii mutant. The mutation also affected sensitivity to compounds altering the intracellular oxidative state of these bacteria, such as the thiol oxidant diamide, the thiol reductant DTT, the thiol-specific alkylating agent mBBr, the reactive oxygen species generators menadione, or hydrogen peroxide. The whiB7 mutant in S. lividans was slightly more sensitive to type 2 NADH dehydrogenase inhibitors (phenothiazines: chlorpromazine and thioridazine), suggesting a potential link between WhiB7 and respiration/redox stress. Finally, whiB7 mutations in R. jostii and S. lividans caused 4-fold increased sensitivity to rifampin, an inhibitor of the RNA polymerase, whereas the mutation in M. smegmatis did not alter its rifampin sensitivity.
In summary, whiB7 mutations led to multidrug sensitivity in Actinobacteria. Compounds targeting ribosome function showed the most dramatic effects, but increased sensitivities to inhibitors of other cellular processes were also observed. The spectrum and levels of sensitivity varied in different species of the actinobacterial phylum. This suggested that actinobacterial species each carry specific repertoires of whiB7-controlled drug resistance genes.
WhiB7 Proteins Are Functional in Heterologous Actinobacteria-WhiB7 proteins displayed a conserved core sequence ( Fig. 2; their variable N termini are annotated) with characteristic residues, including cysteines that may bind ironsulfur clusters, a glycine-tryptophan-rich region that predicts a unique turn in the protein conformation, and an AT-hook DNA-binding motif. These sequence motifs suggested a universal function and regulatory features conserved in actinobacterial WhiB7 proteins. In principle, the fact that inactivation of the whiB7 gene conferred distinct species-specific changes in resistance patterns could be due to divergent functional speci- WhiB7 Activates Species-specific Drug Resistance NOVEMBER 29, 2013 • VOLUME 288 • NUMBER 48 ficities (drug recognition, promoter recognition, regulatory partners, or thiol redox properties) of the corresponding WhiB7 proteins. To rule out the possibility that the speciesspecific drug resistance spectra were determined directly by recognition specificities of the different WhiB7 proteins or by genomic drug resistance proteins, four whiB7 orthologs were expressed from the mycobacterial hsp promoter (expression in M. smegmatis) or a thiostrepton-inducible promoter (expression in R. jostii). These plasmids were introduced into whiB7 mutant strains and assayed for their abilities to restore the species-specific wild type resistance spectra. The sensitivities to erythromycin, spectinomycin, and tetracycline in M. smegmatis and erythromycin and tetracycline in R. jostii were used as representative indicators of WhiB7 activities. The whiB7 genes from either M. smegmatis (whiB7 SM ), M. tuberculosis (whiB7 TB ), S. coelicolor (whiB7 ST ), or R. jostii (whiB7 RH ) were able to fully or partially suppress the sensitivity phenotypes of the mutant strains of either M. smegmatis (Table 2) or R. jostii (Table 3). In addition, when constitutively expressed by the hsp promoter in the parental M. smegmatis strain, whiB7 genes also provided slightly increased (2-fold) resistance to the specific spectrum of drugs affected by the whiB7 mutation (data not shown). These results demonstrated that the abilities of heterologous whiB7 genes to contribute to drug resistance were defined by specificities of genes in their hosts' whiB7 drug resistance regulons and not by divergent functional specificities of the corresponding WhiB7 proteins. Expression of Heterologous whiB7-dependent Genes in M. smegmatis Defines Their Resistance Functions-In M. tuberculosis, overexpression and inactivation of whiB7 led to the identification of genes in the whiB7 regulon (7). The whiB7 regulon includes several genes associated with antibiotic resistance, such as tap (Rv1258c), an efflux pump that confers resistance to aminoglycosides and tetracyclines and is also involved in maintaining cellular homeostasis in the nonreplicative state (13), and erm (37) (Rv1988) (14), a ribosomal methyltransferase that confers macrolide resistance by modification of the 23 S rRNA. Genetic approaches and drug sensitivity assays were used to demonstrate a direct relation between their expression and whiB7 mutant phenotypes (Fig. 3). The erm (37) and tap genes from M. tuberculosis (tap TB ) or M. fortuitum (tap FR ), cloned downstream of hsp promoter (to provide for WhiB7-independent gene transcription), or their native promoters (WhiB7-dependent gene transcription) were analyzed in M. smegmatis wild type or whiB7 mutant backgrounds. Sensitivities of these strains to representative antibiotics affected by the whiB7 mutation in M. smegmatis, i.e. spectinomycin, erythromycin, and tetracycline, are shown in Table 2.   155) or its ⌬whiB7 derivative, carried pMV361 (hsp promoter to provide expression of cloned genes) or pSUM36 (without vectorencoded promoter). b MICs were assayed over a range of 2-fold serial drug dilutions. ERY, erythromycin; SPT, spectinomycin; TET, tetracycline. c Fold change in sensitivity compared with the control strain. The ⌬whiB7 strain had increased sensitivity (1) compared with the parental control. All other decreases in sensitivity generated by cloned genes are expressed relative to the corresponding host with empty vector.

WhiB7 Activates Species-specific Drug Resistance
The erm (37) gene under the control of the hsp promoter strongly increased erythromycin resistance in the whiB7 mutant (64-fold), reaching levels similar to those of the wild type strain, but did not alter resistance to spectinomycin or tetracycline. Under the control of its native M. tuberculosis promoter, erm (37) increased resistance (2-4-fold) to erythromycin in the wild type strain but had no effect on resistance in the whiB7 mutant (Table 2).
Analogous studies were done using the tap genes (tap TB and tap FR ). Expression of tap TB from the constitutive hsp60 promoter provided full phenotypic suppression of the whiB7 mutation in M. smegmatis for spectinomycin and tetracycline resistance but not for erythromycin. When the tap genes from M. tuberculosis or M. fortuitum were cloned under the control of their native promoters, resistance was whiB7-dependent. The tap FR gene expressed from its native promoter generated a different resistance spectrum, providing high levels of resistance (32-64-fold) to tetracycline but minimal or no resistance to the other two drugs. Similar to the erm (37) and tap TB constructs, this phenotype was impaired in the whiB7 mutant (Table 2). In summary, these data demonstrated that whiB7 was essential for the function of drug-resistant determinants encoded by erm (37), tap TB , and tap FR genes. In addition, expression of orthologous tap genes conferred different patterns of drug resistance in the same genetic background.

Conserved Sequence Motifs Are Required for WhiB7
Activity in Vivo-The WhiB7 protein from S. lividans (WhiB7 ST ) was mutagenized to define requirements for the in vivo function of the predicted N-terminal amino acids, the conserved cysteines, the glycine/tryptophan-rich region, and the AT-hook DNAbinding motifs (Fig. 2). Two shorter versions of the annotated WhiB7 ST protein, deleting 9 and 28 amino acids from the annotated N terminus (Val-10 and Pro-29, respectively) were cloned under the control of the hsp60 constitutive promoter. Both forms restored wild type resistance levels to erythromycin and spectinomycin when introduced to a whiB7 mutant in M. smegmatis (Table 4). However, complementation was more efficient in cultures expressing the shorter protein, Pro-29; resistance levels were consistently restored 24 h earlier compared with Val-10 or protein corresponding to the annotated start site. The annotated full-length WhiB7 ST protein has a calculated molecular mass of 17 kDa, and the Pro-29 form is 14 kDa. Pro-29 was closely linked to an in-frame TTG potential start codon (annotated Leu-27) and a potential upstream ribosome-binding site.
To test whether the AT-hook motif was required for WhiB7 ST activity, clones carrying four mutated alleles were assayed for their abilities to restore resistance to three antibiotics: erythromycin, spectinomycin, or chloramphenicol (Table  4). LN2, encoding a protein lacking the five C-terminal amino acids adjacent to the AT-hook motif, restored wild type resistance levels, suggesting that this region of the C terminus was not required for WhiB7 ST activity. In contrast, the LN1 form, lacking both the AT-hook and the five C-terminal amino acids, and LN4, lacking the AT-hook but retaining the five C-terminal amino acids, did not complement the whiB7 mutation. The double point mutations in LN3, in which the two prolines of the AT-hook were mutated to alanine (P106A/P118A; residues needed to maintain a trans-configuration in the AT-hook for proper functionality (51,52)), provided decreased complementing activity. These data suggested that the AT-hook DNA binding domain was required for full WhiB7 ST activity.
Another Wbl family sequence signature is a tryptophan/glycine-rich motif (Fig. 2). A mutant form of WhiB7 ST (LN5) in which the motif was mutated (deletion of eight amino acids; WGVWGGEL) was unable to restore antibiotic resistance ( Table 4), suggesting that it is required for WhiB7 function. As this motif is conserved in all WhiB-like proteins, it probably provides a similar, essential role throughout the Wbl family.  Tap and Erm(37) functions were studied in the parental and whiB7 mutant backgrounds measuring their abilities to increase the MIC to tetracycline, spectinomycin, and erythromycin in the parental strain or to restore the sensitivity of the whiB7 mutant to that of the parental strain. Quantitative data represented in this schematic are presented in Table 2.

JOURNAL OF BIOLOGICAL CHEMISTRY 34521
To evaluate the requirement of the four conserved cysteine residues for WhiB7 ST function, five alleles of whiB7 ST were constructed in which triplets encoding each individual or the two most closely linked cysteines were mutated to serine (C49S, C72S, C75S, C81S, and C72S/C75S) ( Table 4). Mutant strains had obvious defects in activating antibiotic resistance genes (erythromycin, chloramphenicol, and spectinomycin) in an M. smegmatis whiB7 mutant background. Interestingly, the C49S allele partially complemented the spectinomycin sensitivity profile (MICs: wild type, Ͼ30 g/ml; whiB7 mutant, Ͻ5 g/ml; and C49S, 10 g/ml), suggesting that this particular WhiB7 ST mutation retained partial activity. These data suggested that at least three of the four conserved cysteines were essential for WhiB7 activity.
The conserved cysteines of WhiB-like proteins are generally needed to coordinate their iron-sulfur clusters (19, 22, 23, 26 -28, 53-55) that can provide altered structures and functions to transcriptional regulatory proteins (19, 22, 23, 26 -28, 53-56). Western blot analyses were performed using C-terminal His-tagged WhiB7 ST proteins to determine whether the lack of complementation by cysteine-mutated forms of the WhiB7 ST protein could be attributed to reduced expression or increased degradation of the mutant proteins (Fig. 4). The addition of a His 6 tag at the C terminus of the WhiB7 ST protein did not alter its biological activity; His-tagged WhiB7 ST cysteine mutants had similar activities as corresponding untagged proteins, i.e. the only mutant protein with detectable activity was C49S (data not shown). In Western blot analyses of the parental strain, WhiB7 ST migrated roughly as predicted by its sequence (ϳ17 kDa). WhiB7 ST C49S co-migrated with wild type protein, although the C72S, C75S, C81S and C72S/C75S forms migrated faster (ϳ14 kDa) (Fig. 4). Interestingly, migration patterns correlated with in vivo activity, suggesting unstable iron-sulfur clusters were present in the holoprotein when cysteines at positions 72, 75, and 81 were replaced by serines, leading to cleavage of the protein. In contrast, the C49S mutation might still allow weaker stabilization of the holoprotein providing some in vivo activity. Although most iron-sulfur clusters are bound by four cysteines, in some cases the protein activity can be retained when other amino acids replace one of the cysteines (57,58).
Anaerobically Purified WhiB7 ST Coordinated an Oxygenlabile Fe-S Cluster-WhiB7 ST was heterologously produced in E. coli and purified anaerobically for biochemical studies. When cells containing WhiB7 ST were disrupted and centrifuged to remove insoluble material, the protein was present in both the supernatant and insoluble pellet fractions. The soluble fraction had an intense dark brown color that could be monitored throughout the purification process. Over 95% of the purified protein was WhiB7 ST , as determined by SDS-PAGE analysis (data not shown). The UV-visible spectrum of anaerobically purified WhiB7 ST was similar to anaerobically purified WhiD (27) and reconstituted WhiB proteins containing 4Fe-4S clusters (23). Removal of the N-terminal His tag by enterokinase did not alter the absorbance spectra (data not shown). The preparation contained 2.1 Ϯ 0.3 and 2.2 Ϯ 0.3 mol of iron and sulfur, respectively, per mol of WhiB7 ST protein monomer. The molar concentration of iron per protein was also estimated by spectroscopy. Molar absorption coefficients (⑀) of iron in Fe-S clusters at 400 nm are ϳ4,000 M Ϫ1 cm Ϫ1 (19); this predicted an iron/protein molar ratio of 1.76.
The electronic absorption spectrum of purified WhiB7 ST was used to monitor redox sensitivity. The spectrum was stable for weeks even when incubated at room temperature, provided oxygen was excluded. Immediately following the introduction of air, however, the Fe-S cluster broke down rapidly (Fig. 5A) as evident from the monotonic decay (t1 ⁄ 2 ϭ 28 min) of the absorbance band at 413 nm (Fig. 5A, inset). Similarly, reaction of WhiB7 ST with a 10-fold molar excess of potassium ferricyanide resulted in the eradication of this absorbance band indicating degradation of the Fe-S cluster (Fig. 5B). The absorption spectrum of the WhiB7 ST protein did not change after exposure to sodium dithionite. The fact that sodium dithionite was unable to alter the spectrum indicated that the anaerobically purified WhiB7 ST protein had a reduced Fe-S cluster and that the breakdown of this cluster is triggered by its oxidation (19,59).

TABLE 4 Genetic elucidation of the in vivo role of the N-terminal region, conserved cysteines, the glycine-tryptophan-rich region, and the DNAbinding AT-hook motif of the WhiB7 ST protein
Refer to Fig. 2 for locations of mutations.

Parental
None a MICs were determined using E-test (AB Biodisk). The parental strain, mc 2 6, carried pMV361-derived vectors, some containing whiB7 mutant genes from S. coelicolor (whiB7 ST ): V10, nine amino acids deleted from the N terminus; P29, 28 amino acids deleted from the N terminus; LN1, lacking both the AT-hook and the last five C-terminal amino acids; LN2, lacking the five C-terminal amino acids; LN3, two prolines mutated to alanine (P106A and P118A); LN4, without the AT-hook but retaining the last five C-terminal amino acids; LN5, the tryptophan/glycine-rich motif (WGVWGGEL) deleted; and a series of mutants with the conserved cysteines mutated to serine: C49S (LN6), C72S (LN7), C75S (LN8), C81S (LN9), and C72S/C75S (LN10). ERY, erythromycin; CHL, chloramphenicol; SPT, spectinomycin. The oligomeric state of the anaerobically purified WhiB7 ST was also determined under anaerobic conditions using size exclusion chromatography. The elution volume of WhiB7 ST corresponded to a molecular mass of 34 kDa (Fig. 5C). Because the theoretical molecular mass of the N-terminal His-tagged WhiB7 ST is 16 kDa, and provided that the protein is globular in character, these results indicated that N-terminal His-tagged WhiB7 ST protein was a dimer when purified anaerobically.

DISCUSSION
Multiple Wbl genes are a unique feature of actinobacterial genomes, one of the largest and most diverse bacterial phyla (29). These genes may have evolved independently to provide specific functions in Actinobacteria having different shapes, developmental systems, habitats, and metabolisms and have also been adopted by actinobacterial plasmids and phages (29). Most whiB paralogs are not found in all genomes, reflecting evolutionary divergence of genes having specialized functions. In contrast, whiB1, whiB2, whiB3, whiB4, and whiB7 are ubiquitous (29), implying that they are the foundations for common core phenotypes that are retained throughout this bacterial taxon. Retention of whiB7 paralogs suggests that genes in their regulons provide selective advantage under stress conditions commonly experienced by Actinobacteria. Our results demonstrate that native whiB7 genes are required for expression of resistance determinants encoded by their genomes (represented by erm (37) and tap) and that whiB7 genes are also able to activate drug resistance genes present in heterologous Actinobacteria. This supports the concept that many antibiotics induce a common physiological response that activates WhiB7 in all Actinobacteria (9). Here, we used genetic and biochemical tools to investigate the iron-sulfur cluster, a redox sensor, as a probable antibiotic-induced effector of changes in WhiB7 structure and activity and then characterized resistance genes under whiB7 control.
Experiments showing that whiB7 genes function effectively in heterologous actinobacterial hosts suggested the presence of conserved sensor and effector regulatory partners. Extending what has been reported for WhiB3 TB (34), our studies of WhiB7 SM (33) have shown that it binds to a conserved amino acid sequence in the primary factor found in all Actinobacteria. It is noteworthy that the promoter region of whiB7, tap, erm, and eis genes from M. tuberculosis includes a conserved WhiB7 AT-rich binding site upstream of the Ϫ35 promoter hexamer (9). WhiB7 functions as a transcriptional activator, and the requirement for the AT-hook module suggests conserved target promoter sequence recognition motifs serving to control genes that provide multidrug resistance. Our genetic approach provided in vivo evidence supporting this hypothesis (Fig. 3). Iron-sulfur clusters in WhiB7 may respond to changes in redox balance in a way that localizes RNA polymerase to promoters containing AT-rich motifs and thereby activates transcription (33,34). Although WhiB7 recognizes an AT-rich motif, recognition sequences for other Wbl proteins are not defined. In the case of WhiB7, there is evidence that other regulators are involved (9,60).
To characterize more homogeneous, biologically relevant forms of WhiB7, we studied anaerobically purified WhiB7. Previous studies have typically utilized Wbl proteins that were isolated as inclusion bodies, solubilized in urea, purified under aerobic conditions, and reconstituted in vitro. The absorption curve of anaerobically purified WhiB7 from S. lividans suggested that it could coordinate a 4Fe-4S cluster (Fig. 5A). This spectrum was indistinguishable from anaerobically purified WhiD from S. coelicolor, reported to contain a 4Fe-4S cluster (27). It also lacked the characteristic broad shoulders (424, 460, and 560 -580 nm) of aerobically purified WhiB7 from M. tuberculosis and WhiD from S. coelicolor proteins reported to have 2Fe-2S clusters (19,23). However, chemical determination of the iron content in our anaerobically purified WhiB7 ST demonstrated the presence of only two atoms each of iron and sulfur atoms per protein monomer. The apparent inconsistencies between these data might reflect a mixed population of apoand holo-forms of the protein, containing only about 50% in the 4Fe-4S form. It is not uncommon to observe low occupancy rates in Fe-S cluster coordinating proteins, including Wbl proteins. This has also been reported for other Fe-S cluster coordinating proteins; Layer et al. (61) reported that anaerobically purified HemN protein from E. coli coordinated a 4Fe-4S cluster with incomplete iron incorporation (2 mol of iron/mol of HemN). Alternatively, it is conceivable that dimeric WhiB7 harbors a single 4Fe-4S cluster. Additional biochemical studies are required to test this hypothesis. Finally, the strong reductant sodium dithionite was unable to further reduce WhiB7 ST (Fig. 5B), suggesting that the cluster is fully reduced as anaerobically purified.
Replacement of cysteine with serine destabilizes iron-sulfur clusters, and therefore cysteine to serine mutations have been employed to probe whether particular cysteine residues participate as iron-sulfur ligands (62). Intuitively, the three most adjacent cysteines in WhiB7 ST (Cys-72, Cys-75, and Cys-81) should form closer contacts with iron atoms, whereas the distal cysteine (Cys-49) might form a loop to stabilize the iron-sulfur pocket. Sequence analyses showed that these three conserved cysteines are invariant among Wbl protein sequences, whereas the first cysteine is occasionally replaced by aspartate (18,19,26,28). Our mutagenesis data suggested that these more vicinal cysteines (Cys-72, Cys-75, and Cys-81) were essential for normal WhiB7 ST activity, whereas the first cysteine (Cys-49) was FIGURE 5. Absorbance spectra and oligomeric state of anaerobically purified WhiB7 ST proteins. A, samples were air-bubbled for 30 s and remained exposed to the air for 15 h. The blue line represents anaerobic absorbance spectra (t ϭ 0 h), red lines aerobic spectra (from 0 to 1 h), and black lines absorbance spectra from 1 to 15 h after oxygen exposure. Inset, a one-phase exponential decay curve was fitted and calculated a T1 ⁄2 (413 nm) ϭ 23.19 to 34.65 min (95% confident interval. R 2 ϭ 0.9827) and T1 ⁄2 (280/413) ϭ 32.78 to 41.41 min (95% confident interval. R 2 ϭ 0.9940). B, effect of the strong oxidant ferricyanide and reductant sodium dithionite on the absorption spectrum of anaerobically purified WhiB7 ST proteins. Samples were exposed for 1 h to the redox agent at a molar protein/redox agent ratio of 1:10 and then desalted before recording the UV-visible spectra. C, anaerobic determination of the WhiB7 ST oligomeric state. Size exclusion chromatography was performed, and protein elution was followed at absorbance values at 280 and 413 nm. Inset, theoretical calculation of WhiB7 oligomeric size. The molecular mass for the N-terminal His-tagged WhiB7 ST protomer was ϳ16 kDa. partially dispensable. Mutagenesis of corresponding cysteines in other Wbl proteins has suggested that they are essential for in vitro assembly of the iron-sulfur cluster (53,55). In vivo, all four residues are essential for the function of WhiD (19); however, in other Wbl proteins, WhmD (28) and WhiBTM4 (22), their N-terminal conserved cysteines were not essential. Cysteine mutations at these essential sites probably led to inefficient assembly or increased release of Fe-S clusters that can cause instability of WhiB7 (33) and WhiBTM4 proteins (22). This could generate inactive forms of the intact apoprotein in vivo. Unfolded proteins without disulfide bonds typically migrate more slowly on SDS-PAGE than their oxidized forms, an effect that can also be achieved by the inclusion of reducing agents such as mercaptoethanol (22,28,(63)(64)(65) or cysteine mutations (22,28,63,64). Because secondary structure cannot explain the faster migration of mutant proteins on our SDS-polyacrylamide gels, we conclude that the faster migrating peptides detected by antibodies against the C-terminal hexahistidine tag were likely to have been generated by cleavage near their N termini. Therefore, these data, as well as mutational studies of other Fe-Scontaining proteins, suggest that the cysteine to serine (62) or cysteine to alanine mutations (33) might affect protein folding pathways or structural features that determine stability and function. Clarification of how these cysteine residues form the iron-sulfur cluster of Wbl proteins under reducing conditions and unfold in the presence of oxygen awaits three-dimensional structural determinations.
Iron-sulfur clusters can serve as sensors of iron and oxygen (56,66,67) that modulate the activities of transcriptional and translational regulatory proteins (68 -70). In Mycobacterium, expression of whiB7 responds to diverse antibiotics, an effect amplified by the presence of a reducing agent in the medium (9). It is also induced by iron starvation (10), a stimulus that may explain activation early after macrophage infection (11). It is noteworthy here that the redox-responsive transcription regulator of antibiotic resistance and oxidative stress response in E. coli, SoxR, also contains an essential iron-sulfur cluster (71). SoxR is constitutively expressed at low levels in its reduced state. Reactive oxygen radicals or nitric oxide oxidize its ironsulfur cluster, which generates a form of the protein that promotes expression of the transcription factor SoxS. SoxS activates genes providing for repair of DNA damage, removal of superoxide, and resistance to antibiotics. By analogy, WhiB7 or other WhiB-like proteins are likely to be modulated by altered oxidative states. The oxygen sensitivity of a 4Fe-4S cluster of the Streptomyces WhiD sporulation regulatory protein is dependent on four conserved cysteine residues; mutations of these cysteines resulted in sporulation defects (19). Oxygen limitation may be also a factor in triggering the Streptomyces sporulation program (72). M. tuberculosis entering the dormant state in host macrophages must adapt to low oxygen concentrations, which trigger metabolic shifts required for survival. These processes may result in many spore-like characteristics of inactive persistent mycobacterial cells, including enhanced antibiotic resistance (73). The N-terminal module containing the iron-sulfur cluster is linked to a conserved motif predicted to encode a ␤-turn and the AT-hook motif. Site-di-rected mutagenesis showed that both are necessary for WhiB7 ST activity in vivo.
The role of AT-hooks in transcription regulation and chromosome architecture has been a focus of research in eukaryotes; prokaryotic regulatory proteins containing AThooks have also been identified (74 -76). Often found as an auxiliary module within DNA-binding proteins or protein complexes, the AT-hook binds to the minor groove of AT-rich sequences and alters DNA conformation (20,77). Although the AT-hook motif has no detectable secondary structure in solution, NMR studies have demonstrated that the two prolines are in the trans configuration (51,52). This restricts the flexibility of the protein and creates a bend in solution (52) that may facilitate initial DNA binding. When these residues are mutated to alanine or when their position in the peptide motif is altered, the mutant peptide no longer binds to AT-rich DNA sequences in vitro (78); these mutant alleles lose their biological function in vivo (79). Exchanging corresponding prolines to alanines in WhiB7 ST was therefore expected to weaken binding of the AThook to its DNA targets, and indeed our result showed that these mutant alleles were functionally impaired and could only provide low levels of antibiotic resistance in the whiB7 mutant. Taken together, these results suggested that the AT-hook DNA-binding motif facilitated binding of WhiB7 ST to DNA targets and may have enhanced the activities of other transcription regulators or factors by altering protein-protein interactions or DNA conformation. Although other Wbl proteins may not contain a motif that has been demonstrated to bind DNA, the positively charged C terminus of other Wbl proteins is believed to provide a similar function (18,22,53).
The iron-sulfur cluster in SoxR is essential for transcription activation but is not required for the initial folding or to maintain its structure and DNA-binding affinity (71). It was suggested that the iron-sulfur activated form of SoxR functions in remodeling the soxR promoters such that they can form an "active" complex with RNA polymerase preceding transcriptional initiation (71). Similarly, most AT-hook proteins are also unstructured or marginally structured in solution and undergo a specific type of folding when bound to DNA targets or protein partners (77). Thus, WhiB7 proteins display common regulatory partners and control target genes that provide comparable multidrug resistance functions. Interactions with potential sensory and regulatory partners, such as RNA polymerase factors, promoter-binding sites, reductases, etc., must also be conserved. In addition, these results provide important information for structural elucidation of WhiB7 and other WhiB-like proteins, which might help develop novel antimycobacterial compounds. Such inhibitors might render pathogenic mycobacteria more sensitive to available repertoires of clinically approved antibiotics, which may widen chemotherapeutic options to treat mycobacterial diseases.