Unraveling the Specific Regulation of the Central Pathway for Anaerobic Degradation of 3-Methylbenzoate

Background: The specific transcriptional regulation of the mbd pathway for anaerobic 3-methylbenzoate degradation is unknown. Results: The MbdR/3-methylbenzoyl-CoA couple controls the induction of the mbd genes. Conclusion: MbdR is the regulator of the mbd pathway in Azoarcus sp. CIB. Significance: This work highlights the importance of the regulatory systems in the evolution and adaptation of bacteria to the anaerobic degradation of aromatic compounds.


tion and adaptation of bacteria to the anaerobic degradation of aromatic compounds.
Aromatic compounds are included among the most widespread organic compounds in nature, and some of them are man-made environmental pollutants (1)(2)(3)(4). Microorganisms play a fundamental role in the degradation of these aromatic compounds in diverse ecological niches (3,(5)(6)(7)(8). Many habitats containing large amounts of aromatic compounds are often anoxic. In the last decades, biochemical studies concerning the anaerobic degradation of aromatic compounds have been steadily accumulating, with benzoyl-CoA representing the intermediate to which most monocyclic aromatic compounds are converted (3-5, 9 -12). On the contrary, the study on the specific regulatory systems controlling the expression of the gene clusters involved in the anaerobic degradation of aromatic compounds has been mainly restricted to the characterization of a few transcriptional regulators.
Anaerobic benzoate degradation via benzoyl-CoA was shown to be controlled by the two-component BamVW regulatory system (13) or the BgeR regulator (14) in the obligate anaerobes Geobacter strains, and by the BadR/BadM (15,16) and BzdR/BoxR (17)(18)(19)(20) regulators in the facultative anaerobes Rhodopseudomonas palustris and Azoarcus strains, respectively. Moreover, a few global regulators, e.g. AadR, AcpR, and AccR, that influence the anaerobic expression of the benzoyl-CoA central pathway have been reported (15,21,22). A TdiSR (TutC1B1) two-component regulatory system was described for the regulation of the bss/bbs genes encoding the peripheral pathway that converts toluene into benzoyl-CoA in denitrifying bacteria (4,(23)(24)(25). It was also reported that the regulation of the peripheral routes that funnel 4-hydroxybenzoate and p-coumarate into the benzoyl-CoA central pathway in the phototrophic R. palustris strain is accomplished by the HbaR and CouR proteins, respectively (26,27). However, no specific-tran-scriptional regulators that control anaerobic degradation pathways, other than that of benzoyl-CoA and some peripheral routes that converge to the latter, have been described so far.
Azoarcus sp. CIB is a denitrifying ␤-proteobacterium able to anaerobically degrade different aromatic compounds, including some hydrocarbons such as toluene, via benzoyl-CoA, and m-xylene, via 3-methylbenzoyl-CoA (28). The Azoarcus sp. CIB bzd genes responsible for the anaerobic degradation of benzoate are clustered and consist of the P N promoter-driven bzd-NOPQMSTUVWXYZA catabolic operon and the bzdR regulatory gene (29). BzdR-mediated repression of P N is alleviated by the inducer molecule benzoyl-CoA, the first intermediate of the catabolic pathway (17,18). In addition, the P N promoter is also subject to control by the benzoyl-CoA-dependent BoxR repressor, a BzdR paralog that regulates the expression of the box genes responsible for the aerobic degradation of benzoate in Azoarcus sp. CIB (20). The mbd cluster of Azoarcus sp. CIB encodes the central pathway responsible for the degradation of the 3-methylbenzoyl-CoA formed during the anaerobic degradation of m-xylene and 3-methylbenzoate ( Fig. 1) (28). The mbd cluster is organized in at least three operons, i.e. the mbdO-orf9, mbdB1-mbdA, and mbdR operons (Fig. 1A). The mbdB1-mbdA operon is driven by the P B1 promoter and encodes a putative 3-methylbenzoate ABC transporter (MbdB1B2B3B4B5) and the 3-methylbenzoate-CoA ligase (MbdA) that activates 3-methylbenzoate to 3-methylbenzoyl-CoA (peripheral pathway) (Fig. 1B). The mbdO-orf9 operon is regulated by the P O promoter and encodes the enzymes for the anaerobic conversion of 3-methylbenzoyl-CoA to a hydroxymethylpimelyl-CoA (MbdMNOPQW-XYZ) (upper central pathway) and the further degradation of the latter to the central metabolism (Orf1-9) (lower central pathway) (Fig. 1) (28). The mbdR gene was proposed to encode a transcriptional regulator of the TetR family that might regulate the inducible expression of the catabolic mbd genes (28). The efficient expression of the bzd and mbd genes required the oxygen-dependent AcpR activator, and it was under the control of AccR-mediated carbon catabolite repression by some organic acids and amino acids (22,28).
In this work we have characterized the promoters of the mbd clusteranddemonstratedthe3-methylbenzoyl-CoA/MbdR-dependent transcriptional control of the mbd genes in Azoarcus sp. CIB. The studies on the structural-functional relationships of the MbdR protein expand our current view on the transcriptional regulation of anaerobic pathways, and highlight the importance of the regulatory systems in the evolution and adaptation of bacteria to the anaerobic degradation of aromatic compounds.

EXPERIMENTAL PROCEDURES
Bacterial Strains, Plasmids, and Growth Conditions-Bacterial strains and plasmids used are listed in Table 1. Escherichia coli strains were grown in lysogeny broth (LB) medium (31) at 37°C. When required, E. coli cells were grown anaerobically in M63 minimal medium (40) at 30°C using the corresponding necessary nutritional supplements, 20 mM glycerol, as carbon source, and 10 mM nitrate, as terminal electron acceptor. Azo-arcus sp. CIB strains were grown anaerobically in MC medium at 30°C, using the indicated carbon source(s) and 10 mM nitrate as the terminal electron acceptor, as described previously (29). For aerobic cultivation of Azoarcus strains, the same MC medium was used but without nitrate. When appropriate, antibiotics were added at the following concentrations: ampicillin (100 g ml Ϫ1 ), gentamicin (7.5 g ml Ϫ1 ), and kanamycin (50 g ml Ϫ1 ).
Molecular Biology Techniques-Standard molecular biology techniques were performed as described previously (31). Plasmid DNA was prepared with a High Pure plasmid isolation kit (Roche Applied Science). DNA fragments were purified with Gene-Clean Turbo (Q-biogene). Oligonucleotides were supplied by Sigma. The oligonucleotides employed for PCR amplification of the cloned fragments and other molecular biology techniques are summarized in Table 2. All cloned inserts and DNA fragments were confirmed by DNA sequencing with fluorescently labeled dideoxynucleotide terminators (41) and AmpliTaq FS DNA polymerase (Applied Biosystems) in an ABI Prism 377 automated DNA sequencer (Applied Biosystems). Transformation of E. coli cells was carried out by using the RbCl method or by electroporation (Gene Pulser; Bio-Rad) (31). The proteins were analyzed by SDS-PAGE and Coomassie-stained as described previously (31). The protein concentration was determined by the method of Bradford (42) using bovine serum albumin as the standard. Nucleotide sequence analyses were done at the National Center for Biotechnology Information (NCBI) server (www.ncbi.nlm.nih.gov). Pairwise and multiple protein sequence alignments were made with the ClustalW program (43) at the EMBL-EBI server.
Synthesis and Purification of 3-Methylbenzoyl-CoA-The 3-methylbenzoyl-CoA was synthesized from the corresponding carboxylic acid via its succinimide ester as described (44). The CoA ester formed was purified by preparative reversed phase HPLC on a 1525 Binary HPLC Pump system (Waters) equipped with a NUCLEOSIL100 -7 C 18 column (Macherey-Nagel, 50 ml total volume) using acetonitrile in 50 mM potassium phosphate buffer, pH 6.8, at a flow rate of 8 ml min Ϫ1 . The column was equilibrated with 5% acetonitrile; elution was at 25% acetonitrile in buffer. For removal of phosphate, the freezedried CoA ester was suspended in 2% aqueous acetonitrile; elution was with 25% aqueous acetonitrile. The purity was checked by reversed phase HPLC as described above and by the UVvisible spectrum. 3-Methylbenzoyl-CoA was stored at Ϫ20°C as freeze-dried powder.
Construction of Azoarcus sp. CIBdmbdR and Azoarcus sp. CIBdmbdB1 Mutant Strains-For insertional disruption of mbdR and mbdB1 through single homologous recombination, an internal region of each gene was PCR-amplified with the primer pairs 5ЈmbdRmut2/3ЈmbdRmut2 and mbdB1mutEcoRI5Ј/ mbdB1mutXbaI3Ј ( Table 2). The obtained fragments were double-digested with the appropriate restriction enzymes and cloned into double-digested pK18mob vector, generating the pK18mbdRnew and pK18mbdB1 recombinant plasmids (Table  1). These plasmids were transferred from E. coli S17-1pir (donor strain) to Azoarcus sp. CIB (recipient strain) by biparental filter mating (32), and exconjugant strains Azoarcus sp. CIBdmbdR and Azoarcus sp. CIBdmbdB1 were isolated aerobi-cally on kanamycin-containing MC medium harboring 10 mM glutarate as the sole carbon source for counterselection of donor cells. The mutant strains were analyzed by PCR to confirm the disruption of the target genes.
Construction of Azoarcus sp. CIB⌬P A Mutant Strain-The P A promoter was deleted by allelic exchange through homologous recombination using the mobilizable plasmid pK18mobsacB, which allows positive selections of double-site recombinants using the sacB gene of Bacillus subtilis (34). In summary, two primer pairs (Table 2) were used to PCR-amplify the 1191-bp (Z1 fragment) and 1451-bp (Z2 fragment) flanking regions of the P A promoter. Both fragments were digested with restriction endonuclease KpnI and ligated, and the chimeric DNA harboring a deleted P A promoter was PCR-amplified, double-digested, and cloned into the pK18mobsacB plasmid. The resulting pK18mobsacB⌬P A plasmid was transformed into the E. coli S17-1pir strain (donor strain) and then transferred to Azoarcus sp. CIB (recipient strain) by biparental filter mating (32). Exconjugants containing first site recombination were selected on kanamycin-containing MC medium harboring 10 mM glutarate as the sole carbon source for counterselection of donor cells. Second site recombination was selected by growth on the same medium supplemented with 5% sucrose and by plating on glutarate-containing MC plates supplemented with 5% sucrose. Correct allelic exchange in sucrose-resistant and kanamycin-sensitive Azoarcus sp. CIB⌬P A was verified by PCR with the appropriate primers ( Table 2).
Construction of a P A ::lacZ Translational Fusion-The intergenic region between mbdB5 and mbdA genes that includes the P A promoter was PCR-amplified using the primers Inter.mbdB5-A5Ј and Inter.mbdB5-A3Ј.2 ( Table 2). The resulting 238-bp fragment was KpnI/XbaI double-digested and cloned upstream of the lacZ gene into the double-digested pSJ3 promoter probe vector, generating plasmid pSJ3P A ( Table 1). The recombinant pSJ3P A plasmid was KpnI/HindIII doubledigested, and the 3.3-kb fragment containing the P A ::lacZ translational fusion was then cloned into the broad host-range pIZ1016 cloning vector (Table 1). To this end, pIZ1016 was KpnI/HindIII double-digested and its Ptac promoter and polylinker region were replaced by the P A ::lacZ translational fusion, generating plasmid pIZP A ( Table 1).
Construction of a P 3R ::lacZ Translational Fusion-The intergenic region between tdiR and mbdR genes that includes the P 3R promoter was PCR-amplified using the primers PmbdRKpnI5Ј and PmbdRXbaI3Ј ( Table 2). The resulting 451-bp fragment was KpnI/XbaI double-digested and cloned upstream of a lacZ gene into the double-digested pSJ3 promoter probe vector, generating plasmid pSJ3P 3R (Table 1). The recombinant pSJ3P 3R plasmid was KpnI/HindIII double-digested, and the 3.5-kb fragment containing the P 3R ::lacZ translational fusion  was then cloned into the broad host range pIZ1016 cloning vector (Table 1). To this end, pIZ1016 was KpnI/HindIII double-digested and its Ptac promoter and polylinker region were replaced by the P 3R ::lacZ translational fusion, generating plasmid pIZP 3R ( Table 1). Construction of the pIZmbdA and pCKmbdR Plasmids-The pIZmbdA plasmid is a broad host range plasmid that expresses the mbdA gene under the control of the P tac promoter (Table 1). For the construction of pIZmbdA, the 1.7-kb HindIII/XbaI fragment containing the mbdA gene from pUCmbdA (28) was cloned into HindIII/XbaI double-digested pIZ1016 plasmid. The pCKmbdR plasmid (Table 1) expresses the mbdR gene under control of the Plac promoter in the pCK01 cloning vector. To this end, the mbdR gene was PCR-amplified as a 676-bp fragment using mbdRSalI5Ј and mbdRPstI3Ј oligonucleotides ( Table 2). The SalI/PstI double-digested PCR fragment was then cloned into double-digested pCK01 plasmid to generate pCKmbdR.
Overproduction and Purification of MbdR-The recombinant pETmbdR plasmid (Table 1) carries the mbdR gene, which was PCR-amplified (651-bp) with primers mbdRNdeI5Ј and mbdRXhoI3Ј (Table 2), with a His 6 tag coding sequence at its 3Ј-end, under control of the P T7 promoter that is recognized by the T7 phage RNA polymerase. The gene encoding T7 phage RNA polymerase is present in monocopy in E. coli BL21(DE3), and its transcription is controlled by the Plac promoter and the LacI repressor, making the system inducible by the addition of isopropyl 1-thio-␤-D-galactopyranoside (IPTG). 5 E. coli BL21 (DE3) (pETmbdR) cells were grown at 37°C in 100 ml of kanamycin-containing LB medium until the culture reached an A 600 of 0.5. Overexpression of the His-tagged protein was then induced during 5 h by the addition of 0.5 mM IPTG. Cells were harvested at 4°C, resuspended in 10 ml of 20 mM imidazolecontaining working buffer (50 mM NaH 2 PO 4 , pH 8, 300 mM KCl), and disrupted by passage through a French press operated at a pressure of 20,000 p.s.i. Cell debris was removed by centrifugation at 16,000 ϫ g for 20 min at 4°C, and the resulting supernatant was used as crude cell extract. The MbdR-His 6 protein was purified from the crude cell extract by a single-step nickel-chelating chromatography (nickel-nitrilotriacetic acid spin columns, Qiagen). The column was equilibrated with resuspension buffer, loaded with the crude extract, and washed four times with working buffer plus increasing concentrations of imidazole (20, 75, and 100 mM). The MbdR-His 6 protein was eluted in three steps adding to the column working buffer plus increasing concentrations of imidazole (250 and 500 mM and 1 M). The purity of MbdR-His 6 protein was analyzed by SDS-12.5% PAGE. When necessary, the protein solutions were dialyzed against working buffer plus 20 mM imidazole, concentrated using Vivaspin 500 columns (Sartorius, 10,000 molecular weight cutoff membrane), and stored at 4°C where they maintained their activity for at least 6 months.
Analytical Ultracentrifugation Methods-Sedimentation velocity and equilibrium were performed to determine the state of association of MbdR-His 6 . The analytical ultracentrifugation analysis was performed using several protein concentrations (from 11 to 46 M). All samples were equilibrated in buffer containing 50 mM NaH 2 PO 4 , 300 mM KCl, 20 mM imidazole, pH 8. The sedimentation velocity experiments were carried out at 48,000 rpm and 20°C in an Optima XL-A analytical ultracentrifuge (Beckman-Coulter Inc.) equipped with UV-visible optic detection system, using an An50Ti rotor and 12-mm double sector centerpieces. Sedimentation profiles were registered every 1-5 min at 260 and 275 nm. The sedimentation coefficient distributions were calculated by least squares boundary modeling of sedimentation velocity data using the c(s) method (45), as implemented in the SEDFIT program. These s values were corrected to standard conditions (water at 20°C and infinite dilution) using the SEDNTERP program (46) to get the corresponding standard s values (s 20 , w ). Sedimentation equilibrium assays were carried out at speeds ranging from 5000 to 15,000 rpm (depending upon the samples analyzed) and at several wavelengths (260, 280, and 290 nm) with short columns (85-95 l), using the same experimental conditions and instrument as in the sedimentation velocity experiments. After the equilibrium scans, a high speed centrifugation run (40,000 rpm) was done to estimate the corresponding baseline offsets. The measured low speed equilibrium concentration (signal) gradients of MbdR-His 6 were fitted using an equation that characterizes the equilibrium gradient of an ideally sedimenting solute (using a MATLAB program, kindly provided by Dr. Allen Minton, National Institutes of Health) to obtain the corresponding buoyant signal average molecular weight.
Crystallization and X-ray Crystal Structure Determination of MbdR-To determine the three-dimensional structure of MbdR, the mbdR gene from Azoarcus sp. CIB was cloned into pEHISTEV vector (37). To this end, the mbdR gene was PCRamplified with primers mbdRBspHI5Ј and mbdRBamHI3Ј ( Table 2) by using genomic DNA of Azoarcus sp. CIB as template, digested with BspHI and BamHI, and then ligated into the NcoI/BamHI double-digested pEHISTEV vector, giving rise to plasmid pEHISTEVMbdR. Protein expression of the selenomethionine (SeMet)-substituted recombinant MbdR protein was carried out in E. coli B834(DE3) strain (Table 1) transformed with pEHISTEVMbdR, and purification was carried out essentially as described previously (47). The purified SeMet MbdR protein has an extra glycine and alanine at the N terminus resulting from cleavage of the engineered hexa-histidine tag. Crystallization of SeMet MbdR was carried out as described previously (47), and the MbdR crystals were finally grown in the optimized condition of 0.1 M MOPS, pH 7.0, 28% PEG3550, and 0.08% (NH 4 ) 2 PO 4 . Structure was determined using SeMet MAD data and refined using CCP4 package (48). The atomic coordinates and structure factors have been deposited in the Protein Data Bank (PDB) under accession number 4uds. Crystallization of MbdR⅐inducer complex was tried out using the purified MbdR protein with 3-methylbenzoyl-CoA either by co-crystallization or crystal socking, but in both cases the production of crystals failed.
RNA Extraction and RT-PCR Assays-Azoarcus cells grown in MC medium harboring the appropriate carbon source were harvested at the mid-exponential phase of growth and stored at Ϫ80°C. Pellets were thawed, and cells were lysed in TE buffer (10 Tris-HCl, pH 7.5, 1 mM EDTA) containing 50 mg ml Ϫ1 lysozyme. Total RNA was extracted using the RNeasy mini kit (Qiagen), including a DNase treatment according to the manufacturer's instructions (Ambion), precipitated with ethanol, washed, and resuspended in RNase-free water. The concentration and purity of the RNA samples were measured by using a, ND1000 spectrophotometer (Nanodrop Technologies) according to the manufacturer's protocols. Synthesis of total cDNA was carried out with 20 l of reverse transcription reactions containing 400 ng of RNA, 0.5 mM concentrations of each dNTP, 200 units of SuperScript II reverse transcriptase (Invitrogen), and 5 M concentrations of random hexamers as primers in the buffer recommended by the manufacturer. Samples were initially heated at 65°C for 5 min then incubated at 42°C for 2 h, and the reactions were terminated by incubation at 70°C for 15 min. In standard RT-PCRs, the cDNA was amplified with 1 unit of AmpliTaq DNA polymerase (Biotools) and 0.5 M concentrations of the corresponding primer pairs ( Table  2). Control reactions in which reverse transcriptase was omitted from the reaction mixture ensured that DNA products resulted from the amplification of cDNA rather than from DNA contamination. The dnaE gene encoding the ␣-subunit of DNA polymerase III was used to provide an internal control cDNA that was amplified with oligonucleotides 5ЈPOLIIIHK/ 3ЈPOLIIIHK ( Table 2). The expression of the internal control was shown to be constant across all samples analyzed. For real time RT-PCR assays, the cDNA was purified using the GENECLEAN Turbo kit (MP Biomedicals), and the concentration was measured using an ND1000 spectrophotometer (Nanodrop Technologies). The IQ5 Multicolor Real Time PCR Detection System (Bio-Rad) was used for real time PCR in a 25-l reaction containing 10 l of diluted cDNA (5 ng in each reaction), 0.2 M primer 5Ј, 0.2 M primer 3Ј, and 12.5 l of SYBR Green Mix (Applied Biosystems). The oligonucleotides used to amplify a fragment of mbdA were mbdAQ-RT-PCRF3 and mbdAQ-RT-PCRR5 (Table 2). PCR amplifications were carried out as follows: 1 initial cycle of denaturation (95°C for 4 min) followed by 30 cycles of amplification (95°C, 1 min; test annealing temperature, 60°C, 1 min; elongation and signal acquisition, 72°C, 30 s). Each reaction was performed in triplicate. After the PCR, a melting curve was generated to confirm the amplification of a single product. For relative quantification of the fluorescence values, a calibration curve was constructed by 5-fold serial dilutions of an Azoarcus sp. CIB genomic DNA sample ranging from 0.5 to 0.5 ϫ 10 Ϫ4 ng. This curve was then used as a reference standard for extrapolating the relative abundance of the cDNA target within the linear range of the curve. Results were normalized relative to those obtained for the dnaE internal control.
Gel Retardation Assays-DNA probes containing P O , P B1 , P A , and P 3R promoters were PCR-amplified with the corresponding primers indicated in Table 2. The amplified DNA was then digested with ScaI and EcoRI restriction enzymes and single end-labeled by filling in the overhanging EcoRI-digested end with [␣-32 ]dATP (6000 Ci/mmol; PerkinElmer Life Sciences) and the Klenow fragment of E. coli DNA polymerase I as described previously (31). The labeled fragments (P O , P B1 , P A , and P 3R probes) were purified using GENECLEAN Turbo (Qbiogen). The retardation reaction mixtures contained 20 mM Tris-HCl, pH 7.5, 10% glycerol, 50 mM KCl, 0.05 nM DNA probe, 250 g/ml bovine serum albumin, 50 g/ml unspecific salmon sperm DNA, and purified MbdR-His 6 protein in a 9-l final volume. After incubation of the retardation mixtures for 20 min at 30°C, mixtures were fractionated by electrophoresis in 5% polyacrylamide gels buffered with 0.5ϫ TBE (45 mM Tris borate, 1 mM EDTA). The gels were dried onto Whatman 3MM paper and exposed to Hyperfilm MP (Amersham Biosciences) accompanied by amplifier screens (Cronex Lightning Plus, DuPont). The radioactivity present in the retardation complexes and free probes was quantified by using a densitometer with the Quantity One software (Bio-Rad).
DNase I Footprinting Assays-The DNA 32 P-probes used for these experiments were labeled as indicated for the gel retardation assays. The reaction mixture contained 2 nM DNA probe (P O , P B1 , or P A ), 500 g/ml bovine serum albumin, and purified MbdR-His 6 protein in 15 l of buffer (20 mM Tris-HCl, pH 7.5, 10% glycerol, 50 mM KCl). This mixture was incubated for 20 min at 30°C, after which 3 l (0.05 units) of DNase I (Roche Applied Science) (prepared in 10 mM CaCl 2 , 10 mM MgCl 2 , 125 mM KCl, and 10 mM Tris-HCl, pH 7.5) was added, and the incubation was continued at 37°C for 20 s. The reaction was stopped by the addition of 180 l of a solution containing 0.4 M sodium acetate, 2.5 mM EDTA, 50 g/ml salmon sperm DNA, and 0.3 l/ml glycogen. After phenol extraction, DNA fragments were precipitated with absolute ethanol, washed with 70% ethanol, dried, and directly resuspended in 90% (v/v) formamide-loading gel buffer (10 mM Tris-HCl, pH 8, 20 mM EDTA, pH 8, 0.05% w/v bromphenol blue, 0.05% w/v xylene cyanol). Samples were then denatured at 95°C for 3 min and fractionated in a 6% polyacrylamide-urea gel. AϩG Maxam and Gilbert reactions (49) were carried out with the same fragments and loaded in the gels along with the footprinting samples. The gels were dried onto Whatman 3MM paper and visualized by autoradiography as described previously.
Primer Extension Analyses-Azoarcus sp. CIB cells were grown anaerobically on MC medium plus 3-methylbenzoate (inducing conditions) or benzoate (control condition) until mid-exponential phase. For the primer extension analysis of P O and P B1 promoters, total RNA was isolated by using RNeasy mini kit (Qiagen) according to the instructions of the supplier. In the case of P A and P 3R promoters, the procedure was the same but Azoarcus sp. CIB strains harboring pIZP A or pIZP 3R plasmids were used instead of the parental strain due to the weaker nature of these promoters. Primer extension reactions were carried out with the avian myeloblastosis virus reverse transcriptase (Promega) and 15 g of total RNA as described previously (17), using oligonucleotides CIBϩ 1P mbdO 3Ј, CIBϩ1P mbdB1 3Ј, PmbdREcoRI3Ј, and PmbdAEcoRI3Ј (Table 2), which hybridize with the coding strand of the mbdO, mbdB1, mbdR, and mbdA genes, respectively. These oligonucleotides were labeled at their 5Ј-end with phage T4 polynucleotide kinase and [␥-32 P]ATP (3000 Ci/mmol; PerkinElmer Life Sciences). To determine the length of the primer extension products, sequencing reactions of plasmids pSJ3P O , pSJ3P B1 , pIZP A , and pIZP 3R (Table 1) were carried out with oligonucle-otides CIBϩ1P mbdO 3Ј, CIBϩ1P mbdB1 3Ј, PmbdAEcoRI3Ј, and PmbdREcoRI3Ј, respectively, using the T7 sequencing kit and [␣ 32 P]dATP (PerkinElmer Life Sciences) as indicated by the supplier. Products were analyzed on 6% polyacrylamide-urea gels. The gels were dried on Whatman 3MM paper and exposed to Hyperfilm MP (Amersham Biosciences).
In Vitro Transcription Experiments-Multiple-round in vitro transcription assays were performed as published previously (50). Plasmids pJCDP O and pJCDP B1 (Table 1)

mbdR Gene Encodes a Specific Repressor of the P O and P B1
Promoters in Azoarcus sp. CIB-In silico analysis at the 3Ј-end of the mbd cluster revealed a gene, mbdR, that encodes a putative specific transcriptional regulator ( Fig. 1) (28). To analyze the role of the mbdR gene in the expression of the catabolic and transport mbd genes, an mbdR disruptional insertion mutant (Azoarcus sp. CIBdmbdR strain; Table 1) was constructed. Because Azoarcus sp. CIBdmbdR mutant strain grew normally on minimal medium containing 3-methylbenzoate as the only carbon source, the mbdR gene does not seem to function as a transcriptional activator of the mbd genes. Wild-type Azoarcus sp. CIB strain and Azoarcus sp. CIBdmbdR mutant strain were grown anaerobically on minimal medium containing benzoate (control condition) or 3-methylbenzoate (inducing condition) as the only carbon sources, and the expression from P O and P B1 promoters was analyzed by RT-PCR experiments. Whereas the wild-type strain showed a clear induction of the P O and P B1 promoters when grown in 3-methylbenzoate, the MbdR mutant exhibited expression from the P O and P B1 promoters when growing both in benzoate or 3-methylbenzoate (Fig. 2, A  and B). Hence, these results support the idea that MbdR acts as a specific transcriptional repressor of the P O and P B1 promoters.
MbdR Is a New Member of the TetR Family of Transcriptional Regulators-Analysis of the primary structure of MbdR shows an overall low amino acid sequence similarity to members of the TetR family of transcriptional regulators (Fig. 3) (51, 52). To determine the structure of the MbdR repressor, we cloned and expressed in the pETmbdR plasmid (Table 1) a C-terminally His-tagged version of the MbdR protein. The MbdR protein (24.9 kDa) was overproduced in E. coli BL21 (DE3) cells harboring plasmid pETmbdR and purified from the soluble protein fraction by a single-step affinity chromatography (data not shown). The oligomeric state of MbdR protein in solution was determined by analytical ultracentrifugation experiments carried out at different concentrations (11-46 M) of MbdR. Sedimentation velocity analysis of 11 M MbdR revealed a single species with a sedimentation (s) value of 2.9 Ϯ 0.1 (data not shown). The molecular mass of the 2.9 S species, as measured by sedimentation equilibrium, is compatible with the mass of the MbdR dimer (data not shown). Because the frictional ratio f/f 0 was 1.46, the shape of the MbdR dimer deviates from that expected for a globular protein and suggests a slightly elongated dimer.
The crystal structure of MbdR was determined using multiple wavelength anomalous diffraction data, and it was refined to 1.76 Å resolution. A summary of the crystallographic statistics is shown in Table 3. The crystal structure reveals that the crystallographic asymmetric unit contains a monomer of the protein (Fig. 4A). The N-terminal 14 amino acids, residues Thr-46 and Lys-47, and the C-terminal 10 residues in the structure are disordered. Helices ␣1 to ␣3 (Ala-13 to Phe-54) make up the N-terminal DNA binding domain and contain the helix-turnhelix motif (Fig. 3). The larger C-terminal ligand binding domain of MbdR (Fig. 3) consists of helices ␣4 to ␣9 (Lys-57 to Val-204) (Fig. 4A). The long axis of helices ␣4, ␣5, ␣7, ␣8, and ␣9 are approximately parallel and at right angles to ␣1. The short helix ␣6 lies approximately parallel to ␣1 and bisects the C-terminal domain with ␣4 and ␣7 on the one side and ␣5, ␣8, and ␣9 on the other side (Fig. 4A). A 2-fold crystallographic symmetry operator (arises in space group I222) sits parallel to ␣4 and generates a dimeric arrangement. The dimer interface is formed mainly by helices ␣8 and ␣9 with small contributions from helices ␣6 and ␣7. In total, the dimer buries 1759 Å 2 / monomer of surface area with mostly hydrophobic residues (Fig. 4B).
Taken together, all these results indicate that the MbdR homodimer shows the characteristic structure of the TetR family regulators. The members of the TetR family are mostly repressors (51,52), and MbdR behaves also as a transcriptional repressor of the mbd genes responsible for the anaerobic catabolism of 3-methylbenzoate.
MbdR Binds to Palindrome Operator Sites within P O and P B1 Promoters-To confirm in vitro that the MbdR regulator directly interacts with the P O and P B1 promoters, gel retardation experiments were carried out with purified MbdR and a 271-bp DNA harboring P O or a 251-bp DNA containing P B1 as probes. The MbdR protein was able to retard the migration of both DNA probes in a protein concentration-dependent manner (Fig. 2, C and D). The affinity of MbdR for both P O and P B1 probes was very similar, showing a relative K d of 1.71 Ϯ 0.18 and 3.72 Ϯ 0.03 nM, respectively. To further study the interaction of the MbdR protein with the P O and P B1 promoters, we mapped the transcription start sites of both promoters. Primer extension analyses were performed with total RNA isolated from Azoarcus sp. CIB cells grown exponentially in benzoate (control condition) or 3-methylbenzoate (inducing condition).  Table 2. Lane M, molecular size markers (HaeIII-digested ⌽X174 DNA). Numbers on the left represent the sizes of the markers (in base pairs). C and D, the MbdR protein binds to the P O and P B1 promoters. Gel retardation assays were performed as indicated under "Experimental Procedures." C shows the interaction between increasing concentrations of purified MbdR-His 6 protein and a DNA probe (271-bp) containing the P O promoter. D shows the interaction between increasing concentrations of purified MbdR-His 6 protein and a DNA probe (251-bp) containing the P B1 promoter. Lane numbers refer to the MbdR-His 6 protein concentration (nanomolar) used for each reaction. P O and P B1 probes as well as the major P O ⅐MbdR and P B1 ⅐MbdR complexes are marked with arrows.
Whereas no transcript was observed from cells growing in benzoate, a transcript band was visible from cells growing in 3-methylbenzoate (Fig. 5, A and B), confirming a 3-methyl-benzoate-dependent activation of the P O and P B1 promoters. The transcription start site at the P O and P B1 promoters was mapped at a guanine located 137 and 138 bp upstream of the ATG translation initiation codon of the mbdO and mbdB1 genes, respectively.
To characterize the DNA-binding sites of MbdR within the P O and P B1 promoters, we performed DNase I footprinting assays. As shown in Fig. 5, C and D, MbdR protected DNA regions spanning from positions ϩ18 to Ϫ16 and from Ϫ4 to Ϫ34 with respect to the transcription start sites of the P O and P B1 promoters, respectively. The protected regions contained a conserved palindromic sequence (ATACN 10 GTAT) that is suggested to be the operator sequence recognized by MbdR. The MbdR operator in P O and P B1 promoters spans the transcription initiation sites as well as the Ϫ10 and Ϫ35 (only in P B1 ) sequences for recognition of the 70 -dependent RNA polymerase (Fig. 5, C and D). Therefore, the characterization of the MbdR operator supports the observed repressor role of MbdR at the P O and P B1 promoters (Fig. 2, A and B).
3-Methylbenzoyl-CoA Is the Inducer That Alleviates the MbdR-dependent Repression of the mbd Genes-To identify the inducer molecule that alleviates the specific repression exerted by MbdR on the expression of the mbd genes, we first accomplished an in vivo approach. Thus, the activity of a P B1 ::lacZ translational fusion in plasmid pIZP B1 (Table 1)    in E. coli cells harboring also the pCKmbdR plasmid that expresses the mbdR gene under the IPTG-controlled Plac promoter ( Table 1). As shown in Fig. 6A, the ␤-galactosidase activity levels of recombinant E. coli cells expressing the mbdR gene and grown anaerobically in minimal medium with glycerol as sole carbon source were significantly lower than those obtained in E. coli control cells lacking the mbdR gene. This result confirms in a heterologous host the role of MbdR as a transcriptional repressor of the mbd genes. Interestingly, the addition of 3-methylbenzoate to the culture medium of recombinant E. coli cells unable to metabolize this aromatic acid did not alleviate the repression exerted by MbdR (Fig. 6A), suggesting that 3-methylbenzoate, the substrate of the mbd pathway, is not the specific inducer of the P B1 promoter. It has been described previously that the transcriptional activation of benzoate degradation operons in Azoarcus sp. CIB requires benzoyl-CoA, the first intermediate of the anaerobic/aerobic degradation pathways, as inducer molecule (17,20). Thus, we checked whether 3-methylbenzoyl-CoA, the first CoA-derived intermediate of the mbd pathway, could be the specific inducer molecule of the mbd genes. To this end, we expressed the mbdA gene encoding the 3-methylbenzoate-CoA ligase (MbdA) that catalyzes the transformation of 3-methylbenzoate to 3-methylbenzoyl-CoA (28), in the reporter E. coli strain containing plasmids pIZP B1 and pCKmbdR. As shown in Fig. 6A, the activity of the P B1 promoter increased after the addition of 3-methylbenzoate to the culture medium, suggesting that 3-methylbenzoyl-CoA is the specific inducer of the MbdR repressor.
In vitro experiments were then performed to confirm the direct role of 3-methylbenzoyl-CoA as the inducer molecule of the mbd cluster. First, gel retardation experiments showed that the presence of 3-methylbenzoyl-CoA inhibited the interaction of MbdR with the P O and P B1 probes (Fig. 6B). On the contrary, 3-methylbenzoate or some 3-methylbenzoyl-CoA analogs, such as benzoyl-CoA or phenylacetyl-CoA, did not avoid the interaction of MbdR with its target promoters (Fig. 6C), suggesting that MbdR recognizes 3-methylbenzoyl-CoA specifically. The inducing effect of 3-methylbenzoyl-CoA was also observed in footprinting assays where the addition of 3-methylbenzoyl-CoA reverted the protection of MbdR against the DNase I digestion on the P O and P B1 promoters (Fig. 5, C and D).
The role of MbdR as a specific transcriptional repressor of the P O and P B1 promoters and 3-methylbenzoyl-CoA as the cognate inducer was also demonstrated by in vitro transcription assays using supercoiled DNA templates bearing each of the two promoters. Thus, Fig. 6D shows the MbdR-dependent repression of the P O and P B1 promoters, and it also reveals how the addition of increasing amounts of 3-methylbenzoyl-CoA leads to formation of the expected transcripts from both promoters.
Identification of Additional MbdR-dependent Promoters in the mbd Cluster, the P 3R and P A Promoters-Nucleotide sequence analysis of the intergenic regions of the mbd cluster revealed putative MbdR binding regions that contain the conserved (ATACN 10 GTAT) palindromic sequence in the P 3R promoter that drives the expression of mbdR (Fig. 1) (28) and upstream of the mbdA gene encoding the 3-methylbenzoate-CoA ligase (putative P A promoter). To experimentally validate that P 3R and P A are functional promoters of the mbd cluster, the upstream region of mbdR and the mbdB5-mbdA intergenic region were cloned into the promoter probe vector pSJ3, rendering plasmids pSJ3P 3R and pSJ3P A that contain the P 3R ::lacZ and P A ::lacZ translational fusions, respectively (Table 1). Both  Table 1) that produce 227-and 224-nucleotide mRNAs (arrows) from P O and P B1 promoters, respectively, and 30 nM E. coli RNA polymerase. The transcription reactions were carried in the absence of repressor (lanes Ϫ) or in the presence of 100 nM MbdR-His 6 with increasing concentrations of 3-methylbenzoyl-CoA (3MBzCoA). Lane numbers refer to the 3-methylbenzoyl-CoA concentration (M) used for each assay.
translational fusions were then subcloned into the broad host range vector pIZ1016 giving rise to plasmids pIZP 3R (P 3R ::lacZ) and pIZP A (P A ::lacZ) ( Table 1). E. coli cells containing plasmids pIZP 3R or pIZP A were grown in M63 minimal medium, and they showed 75 and 50 Miller units of ␤-galactosidase activity, respectively, suggesting that P 3R and P A are functional but weak promoters. Primer extension experiments revealed that the transcription initiation sites (ϩ1) of P 3R and P A promoters are located 120 bp (data not shown) and 117 bp (Fig. 7A) upstream of the mbdR and mbdA start codons, respectively.
To demonstrate the direct interaction of MbdR with the P 3R and P A promoters, gel retardation assays were performed. To this end, purified MbdR was incubated either with a 352-bp DNA probe carrying the P 3R promoter or with a 225-bp DNA fragment containing the P A promoter. Fig. 8, A and C, shows that MbdR was able to retard the migration of both DNA probes in a protein concentration-dependent manner. The binding was specific, because the addition of unlabeled heterologous DNA did not affect the protein-DNA binding, but the addition of unlabeled specific DNA inhibited the retardation of the probes (data not shown). Several P 3R -MbdR retardation bands were observed (Fig. 8C), which agrees with the fact that several MbdR operator regions were suggested in P 3R (Fig. 8E). As observed previously with the P O and P B1 promoters, 3-methylbenzoyl-CoA behaved as the inducer of MbdR because binding of this protein to the P A and P 3R promoters was significantly diminished in the presence of this aromatic CoA ester (Fig. 8, B  and D). FIGURE 7. MbdR protein interacts with the P A promoter region. A, determination of the transcription start site at the P A promoter. Total RNA was isolated from Azoarcus sp. CIB cells growing on 3-methylbenzoate (lane 3M) as sole carbon source as described under "Experimental Procedures." The size of the extended product was determined by comparison with the DNA sequencing ladder (lanes A, T, C, and G) of the P A promoter region. Primer extension and sequencing reactions of the P A promoter were performed with primer PmbdAEcoRI3Ј (Table 2), as described under "Experimental Procedures." An expanded view of the nucleotides surrounding the transcription initiation site (circled) in the noncoding strand is shown. The longest extension product is pointed by an arrow. B, DNase I footprinting analyses of the interaction of purified MbdR protein and the P A promoter region. The DNase I footprinting experiments were carried out using the P A probe labeled as indicated under "Experimental Procedures." Lane AϩG shows the AϩG Maxam and Gilbert sequencing reaction. Lanes A-D show footprinting assays containing increasing concentrations of MbdR-His 6 . Lane E shows a footprinting assay containing MbdR-His 6 (25 nM) in the presence of 250 M 3-methylbenzoyl-CoA. Left side, an expanded view of the P A promoter region is shown. The protected region is shaded in gray over the promoter sequence. The Ϫ10/Ϫ35 regions are boxed, and the transcription initiation site (ϩ1) is underlined. The predicted MbdR operator is flanked by palindrome sequences indicated by convergent arrows.
Although the role of P 3R driving the expression of the mbdR regulatory gene is obvious, the role of the P A promoter located within the P B1 -driven operon (Fig. 1) is puzzling, and therefore, it was further investigated. P A and P B1 Promoters Are Essential for Growth of Azoarcus sp. CIB on 3-Methylbenzoate-As described previously, the P B1 promoter drives the expression of the mbdB1B2B3B4B5mbdA operon ( Fig. 1) (28). We have shown above (Fig. 8A) that a new MbdR-dependent promoter, the P A promoter, is located upstream of mbdA within the P B1 -driven operon (Fig. 1). To explore whether both promoters share a similar MbdR-dependent regulation, the sequence of the P A promoter recognized by MbdR was experimentally determined by DNase I footprinting assays. Fig. 7B shows that the region of P A protected by MbdR against the DNase I digestion includes the predicted (ATACN 10 GTAC) operator region (Fig. 8E), and it spans the Ϫ35 sequence for recognition of the 70 -dependent RNA polymerase. Moreover, the addition of 3-methylbenzoyl-CoA released the MbdR-dependent protection (Fig. 7B), confirming the role of this molecule as inducer. All these data support the hypothesis that MbdR behaves also as a transcriptional repressor for the P A promoter. To confirm in vivo the repressor role of MbdR on the P A promoter, the activity of a P A ::lacZ translational fusion in plasmid pIZP A (Table 1) was measured in E. coli MC4100 cells harboring also the pCKmbdR and pUCmbdA plasmids that express the mbdR and mbdA genes under the IPTG-controlled Plac promoter, respectively ( Table 1). The ␤-galactosidase activity levels (5 Miller units) of recombinant E. coli cells expressing the mbdR/mbdA genes and grown anaerobically were significantly lower than those obtained in E. coli control cells expressing the P A ::lacZ translational fusion but lacking the mbdR/mbdA genes (50 Miller units). However, the addition of 3-methylbenzoate to the culture medium, which is transformed to 3-methylbenzoyl-CoA by the MbdA activity, alleviated the repression exerted by MbdR, and values of ␤-galactosidase activity of about 40 Miller units were obtained. Therefore, these results show that MbdR behaves as a functional repressor of the P A promoter, and 3-methylbenzoyl-CoA acts as the inducer molecule.
As suggested above by comparing the ␤-galactosidase values in E. coli cells expressing P A ::lacZ (50 Miller units) and P B1 ::lacZ (4000 Miller units) fusions, the P A promoter appears to be significantly weaker than P B1 . To confirm the major role of P B1 in the expression of the mbdA gene in the homologous system, we checked by real time RT-PCR the expression of mbdA in the wild-type Azoarcus sp. CIB strain and in Azoarcus sp. CIBdmbdB1, a mutant strain that contains an insertion within the mbdB1 gene that should block transcription from the P B1 promoter but maintains a functional P A promoter ( Table 1). The expression levels of the mbdA gene in Azoarcus sp. CIBdmbdB1 grown in the presence of 3-methylbenzoate were similar to the basal levels observed with the wild-type CIB strain grown in the absence of 3-methylbenzoate, and they were more than 47 times lower than those observed in the wild-type CIB strain grown in 3-methylbenzoate (data not shown). These data suggested that P B1 , but not P A , has indeed a major contribution to the mbdA expression in Azoarcus sp. CIB. In agreement with this observation, the Azoarcus sp. CIBdmbdB1 mutant strain was unable to use 3-methylbenzoate as sole carbon source (Fig. 9A), and growth was restored when the mbdA FIGURE 8. MbdR protein binds to the P A and P 3R promoters and 3-methylbenzoyl-CoA acts as inducer. Gel retardation assays were performed as indicated under "Experimental Procedures." A shows the interaction between increasing concentrations of purified MbdR-His 6 protein and a DNA probe (225-bp) containing the P A promoter. B shows the interaction between MbdR-His 6 protein (30 nM), the P A DNA probe, and increasing concentrations of 3-methylbenzoyl-CoA (3MBzCoA). Lane Ϫ, free P A probe. Lanes 0 to 100 (A) and 0 to 50 (B) refer to the MbdR-His 6 protein concentration (nM) and the 3-methylbenzoyl-CoA concentration (M) used for each assay, respectively. P A probe as well as the major P A ⅐MbdR complex are marked with arrows. C shows the interaction between increasing concentrations of purified MbdR-His 6 protein and a DNA probe (352-bp) containing the P 3R promoter. Lanes 0 to 1 refer to the MbdR-His 6 protein concentration (M) used for each reaction. P 3R probe as well as the P 3R ⅐MbdR complexes are marked with an arrow and a bracket, respectively. D shows the interaction between MbdR-His 6 protein (0.5 M), the P 3R DNA probe, and 0 or 250 M of 3-methylbenzoyl-CoA (3MBzCoA). Lane Ϫ, free P 3R DNA probe. E, nucleotide sequence of the predicted MbdR operator regions in promoters P A (O A ) and P 3R (O1 3R , O2 3R , and O3 3R ). The flanking ATAC and GTAT palindrome regions are indicated by convergent arrows, and the nonconserved nucleotides are boxed. Nucleotides that extend the palindromic regions are indicated by triangles.
gene was provided in trans in plasmid pIZmbdA (Fig. 9A). In contrast, Azoarcus sp. CIBdmbdB1 mutant strain was still able to use m-xylene as a sole carbon source (data not shown), which is in agreement with the fact that the Bss-Bbs peripheral pathway for the anaerobic degradation of m-xylene generates 3-methylbenzoyl-CoA without the need of a specific CoA ligase activity ( Fig. 1) (53)(54)(55)(56). Taken together, all of these results indicated that P B1 is essential for growth of Azoarcus sp. CIB in 3-methylbenzoate by providing an efficient expression of the mbdA gene rather than by transcribing the mbdB1-B5 genes encoding a putative 3-methylbenzoate ABC transporter.
Nevertheless, the presence of the P A promoter within the P B1 -driven operon raised a question about the role of this weak promoter in 3-methylbenzoate degradation. To confirm whether P A is essential for the anaerobic degradation of 3-methylbenzoate, an Azoarcus sp. CIB⌬P A mutant strain harboring a deletion of the P A promoter but maintaining a complete mbdA gene and the native P B1 promoter was constructed (Table 1). Interestingly, Azoarcus sp. CIB⌬P A was not able to grow anaerobically in 3-methylbenzoate (Fig. 9A), suggesting that P A is also necessary for an efficient expression of the mbdA gene, which in turn supports the presence of P A within the P B1 -driven operon.
Because P B1 accounts for most of the mbdA expression, the role of the weak P A promoter might be related to the initial induction of the mbdA expression when the cells start to grow in 3-methylbenzoate. To check this hypothesis, the activity of the P B1 and P A promoters was analyzed by ␤-galactosidase assays along the growth curve of Azoarcus sp. CIB harboring pIZP B1 (P B1 ::lacZ) and Azoarcus sp. CIB harboring pIZP A (P A ::lacZ) grown in the presence of 3-methylbenzoate. The activity of the weak P A promoter was always higher than that of P B1 up to 6 h after the addition of 3-methylbenzoate, and then P B1 showed a significant induction and reached values about 20-fold higher than those of P A (Fig. 9B). Therefore, these results suggest that the fast and modest induction of the P A promoter will be critical to provide the required amount of the inducer molecule 3-methylbenzoyl-CoA for triggering the induction of the P B1 promoter and to allow growth on 3-methylbenzoate.

DISCUSSION
Bacterial metabolism of some compounds that usually are nonpreferred carbon sources, e.g. aromatic compounds, is generally strictly regulated at the transcriptional level (8). In this work, we have characterized the specific regulation of the mbd central cluster, which is responsible for anaerobic 3-methylbenzoate degradation in Azoarcus sp. CIB, by the MbdR transcriptional repressor. MbdR is an efficient repressor of the mbd genes whose expression can only be switched on when the Azoarcus sp. CIB cells grow anaerobically on 3-methylbenzoate (28) but not on benzoate (Fig. 2, A and B). This finding provides an explanation to the fact that Azoarcus sp. CIBdbzdN, a strain lacking a functional benzoate degradation (bzd) pathway, cannot use benzoate anaerobically despite the Mbd enzymes that can activate benzoate to benzoyl-CoA and further metabolize this CoA-derived compound (28). On the other hand, it is worth noting that the bzd genes are not induced when Azoarcus sp. CIB grows anaerobically in 3-methylbenzoate (data not shown). Therefore, these results reveal that there is no crossinduction between the bzd and mbd pathways, supporting the existence of devoted BzdR-and MbdR-dependent regulatory systems that control, respectively, each of these two central catabolic pathways in Azoarcus sp. CIB.
Analytical ultracentrifugation and crystallographic data indicate that MbdR is a homodimer in solution, a common feature shared by most TetR-like regulators (Fig. 4D) (51,52). Like other members of the TetR family, e.g. TetR (57), QacR (58), ActR (59), FadR (60,61), PfmR (62), and the MbdR monomeric structure includes two domains with nine helices (␣1 to ␣9) linked by loops (Fig. 4A). The N-terminal DNA binding domain (helices ␣1 to ␣3) contains the helix-turn-helix motif whose amino acid sequence is rather conserved in other TetR-like transcriptional regulators (Fig. 3). An electrostatic surface representation of the MbdR dimer structure shows a positively charged patch at the N-terminal domain of both monomers (Fig. 4C), which might contact the phosphate backbone of the target operator region as in the cases of other TetR family proteins (52). An 18-bp conserved palindromic sequence (ATACN 10 GTAT) was suggested to be the operator region recognized by MbdR in P O and P B1 promoters (Fig. 5). The MbdR binding regions in P O and P B1 promoters span the transcription initiation sites as well as the Ϫ10 and Ϫ35 (only in P B1 ) sequences for recognition of the 70 -dependent RNA polymerase (Fig. 5, C and D), which is in agreement with the observed repressor role of MbdR at both promoters (Fig. 2, A and B), and it supports MbdR as a transcriptional repressor of the mbd cluster. Although the length of the MbdR operator region is similar to that of other TetR operators, their different consensus sequences agree with the fact that the DNA-binding mechanisms differ among the TetR family proteins (52).
The C-terminal domain of TetR-like regulators is highly variable, with its specific surfaces required for the dimerization of the protein and for the interaction with the inducer (51,52,57). Based on the previously published studies of other TetR-like regulators, ligand binding usually induces a conformational change in the protein that leads to changes in DNA recognition and interaction, causing the dissociation of the repressor from the cognate promoter (52). To date, all ligands bind in the same general location at or near the dimer interface. However, it has been shown that in some members of the TetR family, for example AcrR (68), the ligand binds in a large internal cavity in the C-terminal region, surrounded by helices ␣4 through ␣8 of each monomer. In contrast, MbdR and other members of TetR family, such as QacR (58), do not have such a cavity (Fig. 4, A and C). By superimposing the apo-MbdR structure with the structure of the QacR⅐diamidine hexamidine complex (69), we could suggest the binding site of 3-methylbenzoyl-CoA in MbdR and a model of the MbdR-3-methylbenzoyl-CoA interaction (Fig. 4E). Binding of 3-methylbenzoyl-CoA would require the movements of helices ␣5, ␣6, ␣8, and ␣9 in MbdR, similar to that described as the "induced fit" mechanism of QacR bound to its ligand (69,70). Similar to what has been observed in the QacR⅐ligand complex structure, the movement of ␣6 after 3-methylbenzoyl-CoA binding to MbdR would induce a rotation of the helix-turn-helix domain (Fig. 4E), and as a consequence, this DNA binding domain would lose its DNA binding ability. Sequence comparison of MbdR and PaaR (Fig. 3), another member of the TetR family which uses phenylacetyl-CoA as inducer (63), shows two MbdR-specific hydrophobic clusters, Gln-107 to Gly-123 within ␣6 and the ␣6/␣7 linkage loop, and Ser-165 to Ile-176 within ␣8. Some residues within these two clusters could be involved in discriminating between the 3-methylbenzoyl group of 3-methylbenzoyl-CoA and the phenylacetyl group of phenylacetyl-CoA (Fig. 4F). Nevertheless, further experiments are needed to determine the structure of the MbdR⅐3-methylbenzoyl-CoA complex for understanding the inducer specificity determinants and the molecular mechanism of transcriptional de-repression at the target promoters. P A and P 3R are two additional promoters within the mbd cluster whose activity levels are lower than those of P O and P B1 but that share with the latter the 3-methylbenzoyl-CoA/MbdRdependent control (Fig. 8). The P 3R promoter drives the expression of the regulatory mbdR gene (Fig. 1). Interestingly, the amount of MbdR needed for the retardation of 50% of the P 3R probe was at least 1 order of magnitude higher than that needed for the retardation of the P A (Fig. 8A), P O (Fig. 2C), and P B1 (Fig.  2D) promoters. The fact that the activity from the P 3R promoter is under auto-repression by MbdR at high protein concentrations underlines the importance of a negative feedback loop that would restrict the intracellular concentration of the transcriptional repressor when it reaches a given concentration. The P A promoter is located within the P B1 -driven operon (Fig.  1). The predicted MbdR operator region (ATACN 10 GTAT) (Fig. 8E) spans the Ϫ35 sequence for recognition of the 70 -dependent RNA polymerase in the P A promoter (Fig. 7B), thus supporting the observed repressor role of MbdR on this promoter. Whereas the role of P 3R driving the expression of the mbdR regulatory gene is obvious, the role of the P A promoter was puzzling, and therefore, it was further investigated.
Inactivation of either the strong (P B1 ) or the weak (P A ) promoters in Azoarcus sp. CIBdmbdB1 and Azoarcus sp. CIB⌬P A mutant strains, respectively, revealed that both promoters are essential for the anaerobic growth of strain CIB in 3-methylbenzoate (Fig. 9A). However, whereas P B1 accounts for most of the mbdA expression when the cells are actively growing in 3-methylbenzoate, the P A promoter allows the initial induction of the mbdA expression when the cells start to grow in this aromatic compound (Fig. 9B). Therefore, these results suggest that the fast and modest induction of the P A promoter in the presence of 3-methylbenzoate leads to an increase of mbdA expression that, in turn, would enhance the amount of the inducer molecule 3-methylbenzoyl-CoA triggering the induction of the P B1 promoter. The expression of the mbdA gene driven by the induced P B1 promoter will provide the required amount of MbdA for the efficient degradation of 3-methylbenzoate and thus will allow growth on this aromatic compound. In summary, these studies highlight the main role of some minor regulatory loops that control the expression of CoA ligases for triggering the efficient expression of aromatic catabolic pathways that use aryl-CoA compounds as central intermediates.
Mbd enzymes are able to activate benzoate and further convert benzoyl-CoA in vitro (28). We have shown here that MbdR is an efficient repressor of the mbd genes, and it recognizes 3-methylbenzoyl-CoA, but not benzoyl-CoA, as inducer. These results suggest that the broad substrate range mbd catabolic genes have recruited a regulatory system based on the MbdR regulator and its target promoters to evolve to a distinct central aromatic catabolic pathway that is only expressed for the anaerobic degradation of aromatic compounds that generate 3-methylbenzoyl-CoA as central metabolite. Thus, the existence in Azoarcus sp. CIB of two different central pathways, i.e. the bzd pathway, for the anaerobic degradation of aromatic compounds that generate benzoyl-CoA as central intermediate, and the mbd pathway, for the anaerobic degradation of aromatic compounds that generate 3-methylbenzoyl-CoA as central intermediate, could be mainly driven by the high specificity of the corresponding repressors, i.e. BzdR and MbdR, for their cognate inducers, i.e. benzoyl-CoA and 3-methylbenzoyl-CoA, respectively. If correct, this highlights the importance of the regulatory systems in the evolution and adaptation of bacteria to the anaerobic degradation of aromatic compounds.
The studies presented in this work expand our knowledge on the specific regulation of anaerobic pathways for the catabolism of aromatic compounds (4,9,14,17,20,27,28). Moreover, it worth noting that 3-methylbenzoyl-CoA is an uncommon metabolite in living cells, and MbdR-responsive promoters are likely to be also very infrequent in nature. Therefore, the P B1 promoter, mbdR regulator, and mbdA genes become potential BioBricks for creating new conditional expression systems that respond to 3-methylbenzoate in a fashion minimally influenced by the host and that has no impact on the host physiology (biological orthogonality), two major desirable traits in current synthetic biology approaches (71).