Department of Biotechnology, Faculty of Life Science and Biotechnology, Fukuyama University, 985 Sanzo, Higashimura,Fukuyama 729-0292, Japan
Received 16 February 2009/Accepted 18 March 2009
DNA microarray analysis revealed that transcription of the Bacillus subtilis yetM gene encoding a putativeflavin adenine dinucleotide-dependent monooxygenase was triggered by certain flavonoids during culture andwas derepressed by disruption of the yetL gene in the opposite orientation situated immediately upstream ofyetM, which encodes a putative MarR family transcriptional regulator. In vitro analyses, including DNase Ifootprinting and gel retardation analysis, indicated that YetL binds specifically to corresponding single sitesin the divergent yetL and yetM promoter regions, with higher affinity to the yetM region; the former regionoverlaps the Shine-Dalgarno sequence of yetL, and the latter region contains a perfect 18-bp palindromicsequence (TAGTTAGGCGCCTAACTA). In vitro gel retardation and in vivo lacZ fusion analyses indicated thatsome flavonoids (kaempferol, apigenin, and luteolin) effectively inhibit YetL binding to the yetM cis sequence,but quercetin, galangin, and chrysin do not inhibit this binding, implying that the 4-hydroxyl group on theB-ring of the flavone structure is indispensable for this inhibition and that the coexistence of the 3-hydroxylgroups on the B- and C-rings does not allow antagonism of YetL.
The rhizosphere is the surface region of soil that is directlyinfluenced by root secretions and associated soil microorgan-isms. A large population of bacteria is present in the rhizo-sphere, where the bacteria are able to feed on nutrients re-leased from plant cells, such as sugars, amino acids, and lipids,and they survive coordinately or hostilely with each other ac-cording to the environment in which they live (3).
Similar to nutrient material, flavonoids are exuded by plantcells, and therefore they are abundant in the soil, especially inthe rhizosphere. Certain flavonoids possess antibacterial activ-ity; quercetin inhibits bacterial DNA gyrase, which inducesDNA cleavage (20, 33). To avoid such harmful effects, somebacteria have a system for degradation of flavonoids that de-toxifies them (22). A gram-positive soil bacterium, Bacillussubtilis, possesses a quercetin 2,3-dioxygenase that convertsquercetin to 2-protocatechuoyl-phloroglucinol carboxylic acidand carbon monoxide (4). So far, quercetin 2,3-dioxygenasehas been isolated from several bacteria and fungi (12, 17);hence, this enzyme appears to be widely distributed and to playa major role in flavonoid degradation in soil microorganisms.
In B. subtilis, the yxaG gene encoding quercetin 2,3-dioxy-genase is a member of an operon containing the yxaH geneencoding a membrane protein with an unknown function (4,38). Our previous study demonstrated that the yxaGH operonis regulated by two paralogous transcriptional regulators,LmrA and YxaF, in response to certain flavonoids (9). LmrAand YxaF, both of which belong to the TetR family, similarlyrecognize and bind to the two cis sequences (LmrA/YxaF
boxes) located tandemly in the yxaGH promoter region, andthe binding of these two regulators is inhibited efficiently anddistinctly by flavonoids, such as quercetin and fisetin; in thisway transcription is induced. The lmrA gene is the first gene inthe lmrAB operon, and the product of the second gene, lmrB,is a member of the major facilitator superfamily involved inresistance to several drugs, such as lincomycin and puromycin(19). The yxaF gene is located immediately upstream of theyxaGH operon and is oriented in the same direction as yxaGH(38). LmrA and YxaF also regulate the lmrAB operon and theyxaF gene, binding to and becoming detached from the corre-sponding single LmrA/YxaF boxes in their promoter regions,as is the case for yxaGH (9).
It is intriguing that B. subtilis utilizes flavonoids as signalingmolecules to induce resistance to structurally unrelated anti-biotics, such as lincomycin and puromycin, through the LmrA/YxaF regulation system. We assume that this might be one ofthe strategies that B. subtilis uses in its struggle against othermicroorganisms in the mixed microbiological flora in the rhizo-sphere, the environmental conditions of which B. subtilis per-ceives through the abundant flavonoids (26). A similar situa-tion was observed for the habitat of Staphylococcus aureus, inwhich gene expression for the QacA major facilitator super-family pump controlled by QacR, a member of the TetR fam-ily, is induced in response to the plant alkaloid berberine (5).
LmrA and YxaF were the first characterized flavonoid-responsive regulators in the genus Bacillus. On the other hand,NodD regulators, which belong to the LysR family and controltranscription of the nod operons involved in nodulation ofRhizobiales in response to flavonoid signals released by theleguminous hosts, have been characterized in detail (13). Also,in Pseudomonas putida DOT-T1E, the resistance-nodulation-cell division family transporter TtgABC and the cognate TetRfamily repressor TtgR constitute a multidrug recognition sys-
* Corresponding author. Mailing address: Department of Biotech-nology, Faculty of Life Science and Biotechnology, Fukuyama Univer-sity, 985 Sanzo, Higashimura, Fukuyama 729-0292, Japan. Phone: 81-84-936-2111. Fax: 81-84-936-2023. E-mail: [email protected].
tem, and several flavonoids are substrates of TtgABC andtrigger pump expression through binding to the TtgR-operatorcomplex to dissociate it (30). Since it is not rare for flavonoidsto function as signaling molecules for communication amongsoil bacteria and plants, it was expected that, in addition to theLmrA/YxaF regulon, B. subtilis possesses genes involved inflavonoid degradation or another physiological function forintercellular communication via flavonoids, which are underthe control of unknown transcriptional regulators in responseto flavonoids.
In this study, in order to elucidate the comprehensive regu-latory system for the expression of the genes responsive toflavonoids in B. subtilis, we tried to identify additional genesthat are significantly induced by flavonoid addition by means ofDNA microarray analysis. Among the new candidate fla-vonoid-inducible genes found, we focused on the yetM geneencoding a putative flavin adenine dinucleotide (FAD)-depen-dent monooxygenase and on its transcriptional regulatorymechanism. DNA microarray analysis involving the wild-typestrain and a yetL disruptant, performed in the framework ofthe Japan Functional Analysis Network for B. subtilis
(JAFAN) (http://bacillus.genome.jp/), suggested that the prod-uct of the yetL gene, which encodes a putative transcriptionalregulator of the MarR family and is located immediately up-stream of the yetM gene in the opposite direction, negativelyregulates yetM transcription, which is induced by certain fla-vonoids. DNA binding experiments involving recombinantYetL showed that YetL binds to the corresponding single sitesin the yetL and yetM promoter regions, with particularly higheraffinity for the latter region. The DNA binding of YetL wasinhibited effectively by flavonoids such as kaempferol, apige-nin, and luteolin, and its weaker interaction with flavonoidssuch as quercetin and fisetin appears to be different from theinteraction of LmrA/YxaF. To date, the flavonoid-responsivetranscriptional regulators of several microorganisms have beenreported. However, to our knowledge, this is the first demon-stration that a MarR family member specifically responds toflavonoids, which provides a clue for elucidation of the entireregulatory mechanism for flavonoid-induced gene expression.
MATERIALS AND METHODS
B. subtilis strains and their construction and cultivation. The B. subtilis strainsused in this study are listed in Table 1. B. subtilis strain 168 was used as thestandard strain (wild type). Strain YETLd was constructed by integration ofplasmid pMUTIN2 (34) into the yetL gene of strain 168 (14; http://bacillus.genome.jp/).
Strain FU1033 carrying a yetL deletion was constructed as follows. Long-flanking homology PCR (35) was performed to create a DNA fragment in whichthe chloramphenicol acetyltransferase gene (cat) (11) was in the same orienta-tion as the original yetL gene and sandwiched by the upstream and downstreamregions of the yetL gene. The regions upstream and downstream of yetL wereboth amplified by PCR with the genomic DNA of strain 168 as the template andtwo primer pairs (yetLupF1/yetLupR_catup and yetLdownF_catdown/yetL-downR1 [Table 2], respectively). The cat cassette was amplified by PCR withprimer pair catF/catR (Table 2) and plasmid pCBB31 bearing the cat gene (24)as the template. The three PCR products described above were added to a
TABLE 1. B. subtilis strains used in this study
Strain Genotype Reference
168 trpC2 (standard strain, wild type)YETLd yetL::pMUTIN2 trpC2 14FU1033 �yetL::cat trpC2 This studyFU1034 �yetL::tet trpC2 This studyFU1035 amyE::�cat PyetL(�188 to 28)-lacZ� trpC2 This studyFU1036 amyE::�cat PyetL(�334 to 228)-lacZ� trpC2 This studyFU1037 amyE::�cat PyetM(�313 to 249)-lacZ� trpC2 This studyFU1038 �yetL::tet amyE::�cat PyetL(�188 to 28)-lacZ� trpC2 This studyFU1039 �yetL::tet amyE::�cat PyetL(�334 to 228)-lacZ� trpC2 This studyFU1040 �yetL::tet amyE::�cat PyetM(�313 to 249)-lacZ� trpC2 This study
TABLE 2. Oligonucleotide primers used in this study
reaction mixture containing an ExTaq DNA polymerase (Takara-bio, Japan) anddeoxynucleoside triphosphates without any primer oligonucleotide, and thendenaturation, annealing, and extension reactions were carried out to combine thethree fragments. Nested PCR with the resultant fragment as the template andprimer pair yetLupF2/yetLdownR2 (Table 2) was performed to amplify thecombined DNA fragment, which was then used to transform strain 168 tochloramphenicol resistance (5 �g/ml), which yielded strain FU1033 (�yetL::cat)(Table 1). Correct replacement of the yetL gene with cat was confirmed by PCRand DNA sequencing. Strain FU1033 was transformed with plasmid pCm::Tc(28) to change the chloramphenicol resistance to tetracycline resistance (10�g/ml), which yielded strain FU1034 (�yetL::tet) (Table 1).
To construct strain FU1035 carrying the yetL promoter region (bases �118 to28) fused to the lacZ reporter gene and strains FU1036 and FU1037, both ofwhich carried a fragment covering 200 bp of the open reading frame (ORF) ofyetL, the entire intergenic region between yetL and yetM, and 200 bp of the yetMORF fused to the lacZ gene in the opposite orientation (bases �334 to 228 forthe yetL promoter and bases �313 to 249 for the yetM promoter; base 1 was thetranscription start site for each of the sequences identified in this work), thecorresponding regions were amplified by PCR with genomic DNA of strain 168as the template and primer pairs PyetL_PEF/PyetL_PER, PyetL_200F/PyetL_200R, and PyetM_200F/PyetM_200R, respectively (Table 2). Each of thePCR products, trimmed by XbaI and BamHI digestion, was cloned into thepCRE-test2 vector (18), which had been treated with the same restriction en-zymes. Correct construction was confirmed by DNA sequencing. The resultantplasmids were linearized by PstI digestion and then integrated into the amyElocus of strain 168 through double-crossover transformation to obtain chloram-phenicol resistance, which resulted in strains FU1035, FU1036, and FU1037(Table 1), respectively.
Strains FU1035, FU1036, and FU1037 were transformed with the genomicDNA of strain FU1034 (�yetL::tet) to obtain tetracycline resistance, which re-sulted in strains FU1038, FU1039, and FU1040 (Table 1), respectively.
B. subtilis cells were pregrown on tryptose blood agar base (Difco) platessupplemented with 0.18% glucose containing chloramphenicol (5 �g/ml), eryth-romycin (0.3 �g/ml), and/or tetracycline (10 �g/ml) according to the drug resis-tance of the cells at 30°C overnight. The cells were inoculated into Luria-Bertani(LB) medium (23) or minimal medium containing 0.4% glucose, 0.2% glutamine,and 50 �g/ml tryptophan (MM medium) (38) supplemented with a mixture of 16amino acids (glutamine, histidine, tyrosine, and aspargine were omitted) (2) toobtain an optical density at 600 nm (OD600) of 0.05 and then incubated at 37°Cwith shaking.
DNA microarray analysis. DNA microarray analysis was performed as de-scribed previously (39). Strain 168 cells were cultivated at 37°C in 200 ml of MMmedium supplemented with 16 amino acids as described above until the OD600
reached 0.2, and either quercetin or fisetin dissolved in dimethyl sulfoxide(DMSO) was added to the medium at a final concentration of 200 �g/ml. Thesame volume of DMSO that was added to the flavonoid solution was added to acontrol culture. After further cultivation until the OD600 reached 0.8, the cellswere harvested by centrifugation, and then total RNA was extracted and purifiedfor synthesis of cDNA labeled with a fluorescent dye (Cy5 for the flavonoidtreatment sample and Cy3 for the control).
Primer extension analysis. Two sets of strains, strains FU1035 and FU1038and strains 168 and YETLd, were used for primer extension analysis to deter-mine the transcription start sites of the yetL and yetM genes, respectively. Cellsof each strain were grown in LB medium until the OD600 reached 1.0 andharvested, and then total RNA was extracted and purified as described previ-ously (39). For the primer extension reaction for the yetL and yetM transcripts,total RNA (45 �g) was annealed to 1 pmol each of primers PEpR and PyetMR(Table 2), respectively, which had been 5� end labeled with a MEGALABEL kit(Takara-bio) and [�-32P]ATP (MP Biomedicals), and then the primer extensionreaction was conducted with ThermoScript reverse transcriptase (Invitrogen) asdescribed previously (24). Templates for the dideoxy sequencing reactions forladder preparation, starting with the same 5�-end-labeled primers that were usedfor yetL and yetM reverse transcription, were generated by PCR with genomicDNA of strains FU1035 and 168 as the templates and primer pairs PEpF/PEpRand PyetMF/PyetMR, respectively (Table 2). Autoradiograms were obtained andquantified using a Typhoon 9400 variable image analyzer (GE Healthcare).
Production and purification of the YetL protein. The yetL ORF was amplifiedby PCR with genomic DNA of B. subtilis strain 168 as the template and primerpair yetLORF_NF/yetLORF_BR (Table 2), digested with NdeI and BamHI, andthen cloned into the pET-22b(�) vector (Novagen) which had been treated withthe same restriction enzymes, which yielded an expression plasmid, pET-YetL.Correct cloning of the yetL gene was confirmed by DNA sequencing.
Escherichia coli strain BL21(DE3) transformed with pET-YetL was grown in
LB medium supplemented with ampicillin (50 �g/ml) at 37°C to an OD600 of 0.4.After isopropyl--D-thiogalactopyranoside (IPTG) was added to a final concen-tration of 1 mM, the cells were cultivated for another 3 h. The cells harvestedfrom 4 liters of the culture were disrupted by sonication in 20 mM Tris-Cl buffer(pH 8.0) containing 10% (vol/vol) glycerol, 0.1 mM phenylmethylsulfonyl fluo-ride, and 1 mM dithiothreitol. After centrifugation (17,000 g, 4°C, 20 min) andfiltration (0.45 �m), the supernatant was recovered and subjected to (NH4)2SO4
precipitation. The supernatant fraction at 70% saturation was dialyzed againstthe same buffer that was used for sonication and then applied to a DEAE-Toyo-Pearl 650 M column (Tosoh, Japan) equilibrated with 20 mM Tris-Cl buffer (pH8.0) containing 10% (vol/vol) glycerol. The column was washed with the samebuffer that was in the column and was eluted with a linear 0 to 1 M NaCl gradientin the same buffer. The YetL fraction was collected and concentrated by ultra-filtration. The homogeneity of the YetL protein was confirmed by sodium do-decyl sulfate-polyacrylamide gel electrophoresis (PAGE) and staining with Coo-massie brilliant blue. The purified YetL protein was subjected to gel filtration(TSK gel G3000 SWXL column; 7.8 mm by 30 cm; Tosoh) with 0.1 M potassiumphosphate buffer (pH 7.0) containing 0.1 M Na2SO4 and 0.05% (wt/vol) NaN3 ata flow rate of 0.2 ml/min to determine the molecular mass of the native form ofYetL.
DNase I footprinting analysis. DNase I footprinting analysis was performed asdescribed previously (9). The PyetL and PyetM probes used for footprinting wereprepared by PCR with genomic DNA of strain 168 and primer pairs PyetLF/PyetLR and PyetMF/PyetMR (Table 2), respectively. Prior to PCR amplifica-tion, only the 5� terminus of one of the primers was labeled with [�-32P]ATPusing a MEGALABEL kit. The DNA probe (0.04 pmol) labeled at the 5� endwas mixed with the YetL protein prepared as described above to obtain aDNA-protein complex, which was then partially digested with DNase I (Takara-bio) in 50 �l of the reaction mixture, and this was followed by urea-PAGE withsequencing ladders prepared by using the same primer set and genomic DNA ofstrain 168. Incubation of the DNA probe with YetL followed by DNase Idigestion was also performed in the presence of 10 mM quercetin or apigenin.
Gel retardation analysis. Gel retardation analysis was performed essentially asdescribed previously (37). The PyetL and PyetM probes, which were the probesthat were used for DNase I footprinting, were labeled by PCR in the presence of[�-32P]dCTP (MP Biomedicals) with the same primer pairs. To create a PyetLprobe derivative from which the internal region was deleted, recombinant PCR(8) was performed with the internal overlapping primer pair PyetL_delEF/PyetL_delER (Table 2) along with the flanking primer pair PyetLF/PyetLR. Thelabeled probe (0.02 pmol) was mixed and incubated at 30°C for 10 min withvarious amounts of the YetL protein in a 25-�l reaction mixture, and then themixture was subjected to PAGE. To evaluate the inhibitory effects of flavonoidson DNA binding of the YetL protein, 1-�l portions of various concentrations ofeach flavonoid dissolved in DMSO were added to the reaction mixture, whichwas followed by similar incubation and then electrophoresis.
lacZ fusion analysis to monitor yetL and yetM promoters. B. subtilis cells weregrown in 50 ml of LB medium at 37°C with shaking. When the OD600 reached0.2, each of the flavonoids dissolved in DMSO was added to the medium toobtain a final concentration of 200 �g/ml, corresponding to concentrations of 0.6,0.7, 0.7, 0.7, 0.7, 0.7, 0.7, 0.8, 0.7, 0.7, 0.8, and 0.7 mM for quercetin, fisetin,galangin, kaempferol, morin, apigenin, luteolin, chrysin, (�)-catechin, genistein,daidzein, and coumestrol, respectively. As a control, 200 �l of DMSO was addedinstead of a flavonoid solution. Then 1-ml aliquots of the culture were withdrawnat 1-h intervals, and the -galactosidase (-Gal) activity in crude cell extracts wasmeasured spectrophotometrically using o-nitrophenyl--D-galactopyranoside(Wako Pure Chemicals Industries, Japan) as a substrate and the proceduredescribed previously (2). To reduce the chromatic disturbance of the -Gal assayby the flavonoid adhering to the cells, the collected cells were washed with 100mM phosphate buffer (pH 7.5) before lysozyme treatment.
Flavonoids. Quercetin, fisetin, kaempferol, morin, apigenin, chrysin, (�)-cat-echin, genistein, and daidzein were products of Sigma. Galangin was purchasedfrom Extrasynthese S.A., luteolin was purchased from Wako Pure ChemicalsIndustries, and coumestrol was purchased from Fluka.
RESULTS
DNA microarray analysis to find additional candidate genesinduced by flavonoids. In order to find candidate genes whoseexpression could be induced by quercetin or fisetin other thanthe members of the LmrA/YxaF regulon (yxaF, yxaGH, andlmrAB), we performed a DNA microarray analysis to compare
VOL. 191, 2009 B. SUBTILIS YetL REGULON RESPONSIVE TO FLAVONOIDS 3687
the transcriptomes of B. subtilis strain 168 cells grown in thepresence and absence of a flavonoid (the DNA microarraydata have been deposited in the KEGG Expression Database[http://www.genome.jp/kegg/expression]). As a result, we se-lected the yetM gene as a candidate, which had not been char-acterized previously but was predicted to encode an FAD-dependent monooxygenase based on a BLASTP sequencesimilarity search (http://www.genome.ad.jp/). (A study indi-cated that quercetin and fisetin are rather weak inducers ofyetM expression compared with some other flavonoids, as de-scribed below.) Immediately upstream of yetM, the yetL geneencoding a transcriptional regulator belonging to the MarRfamily (http://bacillus.genome.jp) is in the opposite orientation(Fig. 1). In the framework of the JAFAN (http://bacillus.genome.jp/), a comprehensive DNA microarray analysis ofhundreds of putative transcriptional regulators has been con-ducted, and a DNA microarray analysis involving strains 168(wild type) and YETLd (a yetL disruptant) indicated that theyetL disruption resulted in a significant increase in yetM tran-scription (data not shown). Based on all the information, wehypothesize that YetL represses the yetM gene by binding to itscis sequence in the promoter region and that some flavonoidscan inhibit DNA binding of YetL to derepress yetM transcrip-tion.
Determination of the transcription start sites of the yetL andyetM genes. To determine the transcription start site of theyetM gene by primer extension analysis, RNA samples wereprepared from cells of strains 168 (wild type) and YETLd (yetLdisruptant). As shown in Fig. 2 (left panel), the specific bandcontaining runoff cDNA representing yetM was detected onlywith the strain YETLd RNA sample, indicating that transcrip-tion of yetM is repressed by YetL. This allowed us to identifythe transcription initiation site of yetM, and we predicted thatthe “�35” and “�10” sequences of the yetM promoter areTTGACA and TAAGGT, respectively, with an 18-bp spacer(Fig. 1) and are similar to promoter sequences recognized by�A-RNA polymerase (7).
To determine the start site of the yetL transcript, we firstperformed primer extension using RNA samples from strains168 and YETLd as the templates and the radiolabeled primerspecific for the upper part of the yetL ORF. But both theprimer extension and DNA sequencing reactions were blockedinside the ORF (data not shown), probably due to blockage ofelongation by formation of specific RNA and DNA secondarystructures. Then we constructed strains FU1035 and FU1038without and with the yetL disruption, respectively, in which theyetL promoter fused to the lacZ gene was integrated into theamyE locus (Table 1). Also, we conducted primer extension
FIG. 1. Organization of the divergent yetL and yetM genes and their promoter regions. (A) Organization of the yetL and yetM genes. yetL, yetM,and neighboring genes are indicated by large open arrows, and the two hairpin structures that likely function as -independent transcriptionterminators are indicated by stem-loops structures. aa, amino acids. (B) Promoter regions of the yetL and yetM genes. The “�35” and “�10”sequences of the two genes are underlined, and the transcription start sites (�1) and the SD sequences are enclosed in boxes. The partial codingregions of yetL and yetM are indicated by lines, and the deleted region in PyetL-�E is also indicated by a line. The protected regions in the codingand noncoding strands detected by DNase I footprinting are indicated by bars. The 18 bp of a perfect palindromic sequence in the yetM promoteris indicated by a pair of facing arrows, and the sequences conserved between the YetL binding sites of yetL and yetM are indicated by bold type.
with a primer specific for lacZ. As shown in Fig. 2 (right panel),the specific band of runoff cDNA was detected with the RNAsamples from both strain FU1035 and strain FU1038, but theband derived from the RNA of strain FU1038 seemed to besubstantially more intense than the band derived from theRNA of strain FU1035, suggesting that the yetL gene is par-tially autorepressed. Thus, we determined the transcriptionstart site of yetL and predicted that the “�35” and “�10”sequences of the yetL promoter are TTGCGT and TATAATwith a 17-bp spacer (Fig. 1), which also seems to be recognizedby �A-RNA polymerase (7).
Preparation of the YetL protein. To prepare the YetL pro-tein for in vitro experiments, the yetL gene was cloned in thevector pET-22b(�), and recombinant YetL was overproducedin E. coli BL21(DE3) cells by means of IPTG addition. Puri-fication of YetL almost to homogeneity was achieved by(NH4)2SO4 precipitation followed by anion-exchange columnchromatography as described in Materials and Methods. On asodium dodecyl sulfate-PAGE gel, a single 19.2-kDa proteinspecies was visualized (data not shown). As determined by gelfiltration, the YetL protein had a molecular mass of 40.6 kDa(data not shown), indicating that it forms a dimer.
Identification of the binding sites of YetL in the yetL andyetM promoter regions. DNase I footprinting analysis was per-formed to identify each of the YetL binding sites in the yetLand yetM promoter regions. When the YetL protein was mixed
with the PyetL probe (bases �68 to 183; base 1 was the tran-scription start base), YetL protected a region in the yetL pro-moter against DNase I (bases 11 to 30 of the coding strand andbases 4 to 26 of the noncoding strand) (Fig. 3, upper panels,lanes 2, 3, 5, and 8). The protected sequence overlapped theShine-Dalgarno (SD) sequence for ribosome binding (Fig. 1).Next, we carried out DNase I footprinting experiments usingthe PyetM probe (bases �86 to 138; base 1 was the transcrip-tion start base). In this analysis, YetL was found to specificallyprotect its binding site in the yetM promoter region againstDNase I (bases �2 to 21 of the coding strand and bases �5 to21 of the noncoding strand) (Fig. 3, lower panels, lanes 2, 3, 5,and 8), and 18 bp of the complete palindrome sequence wasobserved (TAGTTAGGCGCCTAACTA; bases �1 to 17)(Fig. 1). These results suggest that YetL binds to the corre-sponding sites in the yetL and yetM promoter regions to represstheir transcription.
Quantitative evaluation of the DNA binding affinity of YetLand its inhibition by various flavonoids by in vitro analysis. Toquantitatively evaluate the YetL binding to the yetL and yetMsites and its inhibition by various flavonoids, we performed gelretardation analysis using the YetL protein and the PyetL andPyetM probes that were used for DNase I footprinting (Fig. 3and 4A).
As shown in Fig. 4, YetL bound to each of the PyetL andPyetM probes containing its binding site, which resulted in
FIG. 2. Determination of the transcription start sites of the yetM and yetL genes by primer extension analysis. Total RNAs of strains 168 (wildtype) (lane 1) and YETLd (yetL disruptant) (lane 2) used for determination of the transcription start site of yetM (left panel) and total RNAs ofFU1035 (yetL�) (lane 1) and FU1038 (�yetL::tet) (lane 2) used for determination of the transcription start site of yetL (right panel) were preparedand used for the reverse transcription reaction to generate the runoff cDNAs. Lanes G, A, T, and C contained the products of the dideoxysequencing reactions obtained with the same primer that was used for reverse transcription. The runoff cDNAs are each indicated by an arrow.The partial nucleotide sequences of the coding strands corresponding to the ladders are shown; the “�35” and “�10” sequences are underlined,and the transcription start sites (�1) and the SD sequence are enclosed in boxes.
VOL. 191, 2009 B. SUBTILIS YetL REGULON RESPONSIVE TO FLAVONOIDS 3689
retarded bands on a PAGE gel depending on the YetL con-centration. The binding affinity of YetL for the PyetL probewas weaker than that for the PyetM probe, and the apparentdissociation constants (Kd) of YetL for the PyetL and PyetMprobes were estimated to be 24 nM and 6 nM for a dimer,respectively (Fig. 4B). As mentioned above, the YetL bindingsite for yetM contains a complete 18-bp palindrome sequence,whereas the binding site for yetL contains only a portion of thispalindrome but overlaps the SD sequence of yetL. Sequencecomparison of the two binding sites revealed only 9 bp thatmatched in the 18-bp sequences (Fig. 4A). To confirm that thesite found to be protected by YetL by DNase I footprinting isactually essential for YetL binding to the yetL promoter, aPyetL probe derivative lacking this site (PyetL-�E) was exam-
ined for YetL binding affinity. The YetL binding affinity for thePyetL-�E probe was found to be remarkably lower (apparentKd, 1.2 �M) than that for the PyetL probe (24 nM), clearlyindicating that this site is indispensable for YetL binding (Fig.4). There was a second fragment shift from 800 nM to 6.2 �MYetL not only for the PyetL and PyetL-�E probes (Fig. 4B) butalso for the PyetM probe (data not shown), so the shift ap-peared to result from nonspecific binding of YetL to the DNAfragment.
We attempted to quantify the inhibitory effects of variousflavonoids on YetL binding to the PyetL and PyetM probes byperforming a gel retardation analysis with the YetL concen-tration fixed at 780 nM and 49 nM for the PyetL and PyetMprobes, respectively (Fig. 5); these concentrations were suffi-
FIG. 3. DNase I footprinting analysis of YetL in the yetL and yetM promoter regions. DNA probes corresponding to the yetL and yetM promoterregions (PyetL and PyetM), 5� end labeled on either the coding or noncoding strand, were prepared. The 5�-labeled probe (0.8 nM) was incubatedin the reaction mixture with the recombinant YetL protein (lanes 2, 5, 6, and 7 for the PyetL probe, 144 nM [a dimer]; lanes 3, 8, 9, and 10 forthe PyetL probe, 72 nM; lanes 2, 5, 6, and 7 for the PyetM probe, 36 nM; lanes 3, 8, 9, 10 for the PyetM probe, 18 nM) and without the YetL protein(lanes 1 and 4). A flavonoid solution in DMSO (2 �l) was added to the mixture to obtain a flavonoid concentration of 10 mM (lanes 6 and 9,quercetin; lanes 7 and 10, apigenin), and the same volume of DMSO was added to the mixtures in lanes 5 and 8 before incubation. After partialdigestion with DNase I, the resulting mixtures were subjected to urea-PAGE. Lanes G, A, T, and C contained the products of the dideoxysequencing reactions with the corresponding 5�-labeled primers. Nucleotide sequences protected by YetL are indicated on the right in each panel;the SD sequence and the transcription start site (�1) are enclosed in boxes, the perfect palindrome sequence is indicated by a pair of facing arrows,and the sequences conserved in the protected areas of the PyetL and PyetM probes are indicated by bold type.
cient to cause complete retardation of the probes (Fig. 4B).Five flavonols (quercetin, fisetin, kaempferol, morin, andgalangin), three flavones (apigenin, chrysin, and luteolin),three isoflavones (genistein, daidzein, and coumestrol), and(�)-catechin were tested at appropriate concentrations (4.9�M to 10 mM), and the results are summarized in Table 3; theresults for quercetin, fisetin, kaempferol, apigenin, and luteolinare shown in Fig. 5. All of the flavonoids tested except daidzeinand (�)-catechin inhibited YetL binding to the PyetL probe,and the inhibitory effects of fisetin, kaempferol, apigenin, lu-teolin, and coumestrol were prominent (Ki, �5 �M). On theother hand, clear inhibition of YetL binding to the PyetMprobe was observed with kaempferol, morin, apigenin, andluteolin (Ki, �2 mM), and there was slight inhibition by quer-cetin and galangin (Ki, �2 mM). But fisetin, chrysin, genistein,daidzein, coumestrol, and (�)-catechin did not inhibit YetLbinding to the PyetM probe even at a concentration of 10 mM.We also tested the inhibitory effects of quercetin and apigeninon YetL binding to the PyetL and PyetM probes using DNase
I footprinting. When DNase I digestion was carried out with 10mM quercetin or apigenin, the specifically protected regions ofPyetL and PyetM disappeared (Fig. 3, lanes 6, 7, 9, and 10). Theinhibitory effect of quercetin on binding to the PyetM probewas likely so weak that it was detected only by DNase I foot-printing.
Repression of the yetL and yetM promoters by YetL in vivoand release of this repression by certain flavonoids. BothDNase I footprinting and gel retardation analyses revealed theYetL binding sites of yetL and yetM, which are likely involvedin repression of the promoter activities of these genes. Toconfirm this in vivo, we constructed two sets of B. subtilisstrains without and with the yetL disruption, in which the yetLand yetM promoters fused to the lacZ gene in different orien-tations were integrated into the amyE locus, respectively.Strains FU1036 (yetL�) and FU1039 (�yetL::tet) were used toassess the yetL promoter activity in the presence and absenceof YetL, the yetL promoter region (bases �334 to 228), whichcovers 200 bp of the partial yetL ORF, the entire intergenic
FIG. 4. In vitro binding affinity of YetL to the yetL and yetM cis sequences. (A) Each DNA probe used for gel retardation is indicated by a bar.The YetL-binding cis sequences in the yetL and yetM promoter regions are indicated by cross-hatched boxes. The results of a comparison of thesesequences are shown as motif logos created with the B. subtilis Motif Location Search software (http://dbtbs.hgc.jp/motiflocationsearch.html).(B) The 32P-labeled DNA probes for the yetL and yetM promoter regions (PyetL and PyetM) and the 32P-labeled PyetL derivative lacking the cissequence (PyetL-�E) (0.8 nM) were incubated with purified YetL protein, which was followed by PAGE. The YetL protein solution was dilutedstepwise twofold, and an aliquot of each dilution was added to the mixture to obtain the concentrations used. The no protein lanes contained noYetL. The arrows indicate lanes for which the Kd was determined. The experiments were repeated at least two times, and the results ofrepresentative experiments are shown.
VOL. 191, 2009 B. SUBTILIS YetL REGULON RESPONSIVE TO FLAVONOIDS 3691
region between yetL and yetM, and 200 bp of the partial yetMORF, being fused to the lacZ gene (Table 1 and Fig. 6A).When the -Gal activity of each strain was monitored, theactivity of strain FU1039 was found to be fairly low but higherthan that of strain FU1036, suggesting that YetL represses theyetL promoter activity (Fig. 6B). Then we assessed the yetMpromoter activity using strains FU1037 (yetL�) and FU1040(�yetL::tet), the same region that was used for FU1036 andFU1039 (bases �313 to 249 for yetM) being inversely fused sothat lacZ was under control of the yetM promoter (Table 1 andFig. 6A). The -Gal activity of each strain was monitored, andit was found that the activity of strain FU1040 was always much
higher than that of strain FU1037, clearly indicating that YetLrepresses the yetM promoter activity (Fig. 6B). The dere-pressed promoter activities of both yetL and yetM graduallydecreased as the cultures reached the stationary growth phase,suggesting that these promoters were inactivated during thestationary phase, possibly due to a decrease in RNA polymer-ase activity associated with �A and/or an unknown regulatoryfactor(s) other than YetL.
Since each flavonoid had different inhibitory effects on thebinding of YetL to the cis sequences of yetL and yetM in vitro,we examined if a flavonoid releases repression of the yetMpromoter through the YetL repressor, i.e., if it actually inducesthe -Gal activity observed in the lacZ fusion experimentsinvolving strain FU1037. The inducing effects of flavonoids onthe yetL promoter were not examined because of the low ac-tivity of the intrinsic yetL promoter, as judged in the lacZfusion experiment involving strain FU1039 (�yetL::tet).
The 12 flavonoids examined in the gel retardation analysiswere also examined in lacZ fusion experiments, the results ofwhich are summarized in Table 3 together with those obtainedin the in vitro analysis. The induction profiles for the -Galactivity in the presence of quercetin, fisetin, kaempferol, api-genin, and luteolin are shown in Fig. 6C. The -Gal activity ofstrain FU1037 increased significantly in the presence ofkaempferol, apigenin, and luteolin, and kaempferol was themost effective flavonoid (Table 3 and Fig. 6C). Addition offisetin, morin, and coumestrol resulted in moderate inductionof the -Gal activity, while addition of quercetin induced -Galactivity only very slightly (Table 3 and Fig. 6C) and addition ofgalangin, crysin, genistein, daidzein, and (�)-catechin did notinduce -Gal activity at all (Table 3). These in vivo resultsessentially agreed with the results of the in vitro gel retardationanalysis and indicate that 3 of the 12 flavonoids (kaempferol,apigenin, and luteolin) have significant effects and 3 (fisetin,morin, and coumestrol) have moderate effects as inducers forYetL, the repressor of the yetL and yetM genes, and that theyappear to be incorporated (or permeable) in B. subtilis cells.
DISCUSSION
The B. subtilis yetL and yetM genes, which are diverselyoriented with respect to each other (Fig. 1), encode a tran-scriptional regulator belonging to the MarR family and a pu-tative FAD-dependent monooxygenase, respectively. The ori-entations of the yetL and yetM genes and neighboring genesstrongly suggest that yetL and yetM are monocistronic. Thetranscription initiation bases of the yetL and yetM genes wereidentified by primer extension analysis (Fig. 2), and the twopromoters were likely recognized by RNA polymerase possess-ing �A.
The DNase I footprinting analysis revealed that YetL bindsto the cis sequence in each of the yetL and yetM promoterregions, implying that YetL regulates the expression of thesegenes separately (Fig. 3). A 18-bp perfect palindrome se-quence (TAGTTAGGCGCCTAACTA) was found in thebinding site in the yetM promoter region, whereas a perfectpalindromic sequence was not found in the binding site in theyetL promoter region (Fig. 1). The YetL protein exhibitedmuch higher affinity for the cis sequence of the yetM promoterregion (PyetM probe) than for that of the yetL promoter region
FIG. 5. Evaluation of the inhibitory effect of each flavonoid onYetL binding to the yetL and yetM cis sequences by gel retardation.The 32P-labeled PyetL and PyetM probes (0.8 nM) were incubated with780 nM and 49 nM YetL, respectively, in the presence of each fla-vonoid indicated at concentrations of 4.9 �M to 10 mM (12 lanes;twofold dilution with DMSO 11 times before incubation). The DMSOlanes contained no flavonoid, and the no protein lanes contained noYetL. The experiments were repeated at least two times, and theresults of representative experiments are shown.
TABLE 3. Structures of flavonoids, their inhibitory effects on YetL binding to the PyetL and PyetM probes, and derepression of the yetMpromoter repressed by YetL
Flavonoid
Structure
Effect on yetLpromoter
in vitro (Ki��M�)a
Effect on yetMpromoter in vitro (Ki�mM�/effect on yetMpromoter in vivo)a
Quercetinb �� (5) � (�10)/�
Fisetinb ��� (�5) �/��
Kaempferolb ��� (�5) �� (1.9)/���
Morin �� (10) ��� (0.16)/��
Galangin � (39) � (5.0)/�
Apigeninb ��� (�5) ��� (0.24)/���
Luteolinb ��� (�5) ��� (0.16)/���
Chrysin �� (7) �/�
Genistein � (59) �/�
Daidzein � �/�
Continued on following page
VOL. 191, 2009 B. SUBTILIS YetL REGULON RESPONSIVE TO FLAVONOIDS 3693
(PyetL probe) as determined by gel retardation analysis (Fig.4B), which might be attributed to the complete palindromestructure in the binding site of yetM. The implication that yetMexpression is repressed well by YetL, whereas yetL repressionis moderately autoregulated, was confirmed by primer exten-sion analysis (Fig. 2) and the lacZ fusion assay involving thestrains without and with the yetL disruption (Fig. 6B).
We tried to find additional members of the YetL regulon byperforming DNA microarray analysis involving the wild-typeand yetL-deficient strains, as well as a motif search involvingthe B. subtilis Motif Location Search software (http://dbtbs.hgc.jp/motiflocationsearch.html) and the two YetL binding se-quences identified here. However, these approaches were notsuccessful for finding further candidate members of the YetLregulon (data not shown).
To evaluate the inhibitory effects of various flavonoids onthe binding of YetL to its cis sequences, a gel retardationanalysis with various concentrations of each flavonoid was con-ducted (Fig. 5 and Table 3). Twelve flavonoids were tested, andall of them except daidzein and (�)-catechin were found toreadily release YetL binding to the cis sequence of yetL; theinhibitory effects of fisetin, kaempferol, apigenin, luteolin, andcoumestrol were prominent. The inhibitory effects of thisbroad range of flavonoids were due to the lower affinity ofYetL for the yetL cis sequence. On the other hand, the high-affinity binding of YetL to the yetM cis sequence was clearlyinhibited by kaempferol, morin, apigenin, and luteolin andslightly inhibited by quercetin and galangin, but no inhibitionwas observed with the other flavonoids.
The in vivo lacZ fusion experiments showed that severalflavonoids were able to induce expression of the lacZ gene
placed downstream of the yetM promoter, which supports thein vitro results of the gel retardation analysis described above(i.e., clear induction by kaempferol, apigenin, and luteolin,moderate induction by fisetin, morin, and coumestrol, slightinduction by quercetin, and no induction by the other fla-vonoids) (Fig. 6C and Table 3). Based on these in vitro and invivo results, we concluded that kaempferol, apigenin, and lu-teolin certainly act as inducers that release YetL binding to thecis sequence of yetM for derepression of this gene.
To elucidate the structural requirements for flavonoids asinducers of YetL, the inhibitory effects of flavonols (quercetin,fisetin, kaempferol, morin, and galangin) and flavones (apige-nin, luteolin, and chrysin) on YetL binding to the yetM cissequence were compared in vitro and in vivo (Table 3). Theflavonol kaempferol and the flavone apigenin with a 4-hydroxylgroup on their B-rings were much more effective than thecorresponding compounds galangin and crysin without thisgroup (the flavone structure with the ring and carbon assign-ments is shown in Table 3), suggesting that this group is es-sential for YetL inhibition. In addition, kaempferol is moreeffective than quercetin, suggesting that the 3-hydroxyl groupon the B-ring of flavonols prevents the interaction with YetL asan inducer. However, when apigenin and luteolin were com-pared, they were found to be equally effective, which meansthat the 3-hydroxyl group on the B-ring of flavones does not actadversely. Hence, we suppose that a hydroxyl group at eitherposition 3 of the B-ring or position 3 of the C-ring is permissivebut that hydroxyl groups at both positions are nonpermissive.Because the effect of morin is similar to that of kaempferol, itseems that the 2-hydroxyl group on the B-ring does not se-verely impair the inducer function when a 3-hydroxyl group is
TABLE 3—Continued
Flavonoid
Structure
Effect on yetLpromoter
in vitro (Ki��M�)a
Effect on yetMpromoter in vitro (Ki�mM�/effect on yetMpromoter in vivo)a
Coumestrol ��� (�5) �/��
(�)-Catechin � �/�
a The data are the results of an evaluation of the in vitro results for the yetL promoter and a comparison of the in vitro and in vivo results for the yetM promoter.The in vitro results were obtained in experiments involving gel retardation, which were performed as described in the legend to Fig. 5, in which the results for additionof quercetin, fisetin, kaempferol, apigenin, or luteolin are shown. Based on the Ki values for the 12 flavonoids for binding of YetL to the PyetL probe, the inhibitoryeffects are expressed as follows: ���, �5 �M; ��, 5 to 10 �M; �, �10 �M; �, no inhibition. For YetL binding to the PyetM probe, the inhibitory effects are expressedas follows: ���, �0.5 mM; ��, 0.5 to 2 mM; �, �2 mM; �, no inhibition. The actual Ki values are indicated in parentheses. The in vivo results were obtained byperforming experiments with lacZ fusions as described in the legend to Fig. 6C, in which the results for addition of quercetin, fisetin, kaempferol, apigenin, or luteolinare shown. Based on the peak -Gal activity after addition of each flavonoid, the degrees of derepression of the yetM promoter repressed by YetL are expressed asfollows; ���, �5 nmol/min per OD600 unit; ��, 1 to 5 nmol/min per OD600 unit; �, �1 nmol/min per OD600 unit; �, no derepression.
b Data for the effects of quercetin, fisetin, kaempferol, apigenin, and luteolin are shown in Fig. 5 and 6C.
on the C-ring. Isoflavones (genistein, daizein, and coumestrol)and (�)-catechin are unlikely to have significant inhibitoryeffects, implying that the flavone structure might be an essen-tial feature for activity as a YetL inducer.
The specificity of YetL for its inducer flavonoids appears tobe distinct from the specificities of the LmrA and YxaF tran-
scriptional regulators described previously (9). While YetLbinding to the yetM cis sequence is not as affected by quercetinand fisetin, these flavonols substantially inhibit the binding ofLmrA and YxaF to their cis sequences (LmrA/YxaF boxes)(9). Moreover, the inducer specificities of LmrA and YxaFseem somewhat broader than that of YetL. Genistein and
FIG. 6. Derepression of the yetL and yetM promoter activities repressed by YetL in response to flavonoids. (A) The yetL and yetM promoterregions were fused to the lacZ reporter gene in opposite orientations and then integrated into the amyE loci of strains 168 and FU1034 (�yetL::tet)to monitor the yetL and yetM promoters, respectively. In the resulting strains, FU1036 (yetL�) and FU1039 (�yetL::tet), the lacZ reporter is underthe control of the yetL promoter, whereas in strains FU1037 (yetL�) and FU1040 (�yetL::tet), the lacZ reporter is controlled by the yetM promoter.(B) lacZ expression during cultivation of strains FU1036 and FU1037 (left panel) and strains FU1039 and FU1040 (right panel). Squares, OD600values for strains without the yetL disruption (strains FU1036 and FU1039); diamonds, OD600 values for strains with the yetL disruption (strainsFU1039 and FU1040); triangles, -Gal activities for strains without the yetL disruption (strains FU1036 and FU1037); circles, -Gal activities forstrains with the yetL disruption (strains FU1039 and FU1040). (C) Strain FU1037 was used to monitor the yetM promoter activity in response toaddition of each flavonoid. When the OD600 reached 0.2, each flavonoid was added to the culture (indicated by an arrow) at a concentration of200 �g/ml. The squares and triangles indicate the OD600 values and -Gal activities, respectively, and the open and filled symbols indicate cultureswithout and with flavonoids, respectively. The lacZ fusion experiments were repeated at least two times, and the results of representativeexperiments are shown.
VOL. 191, 2009 B. SUBTILIS YetL REGULON RESPONSIVE TO FLAVONOIDS 3695
coumestrol also affect the binding of both LmrA and YxaF totheir boxes, and (�)-catechin exhibits inhibitory activity onlyfor LmrA binding, whereas tamarixetin exhibits inhibitory ac-tivity only for YxaF binding. YetL is also distinct from LmrAand YxaF in domain structure. LmrA and YxaF belong to theTetR family of bacterial transcriptional regulatory proteins,which are known to typically possess two functional domains, ahighly conserved N-terminal DNA binding domain and a lessconserved C-terminal domain involved in both dimerizationand effecter binding (21). The crystal structure of the YxaFprotein showed that this protein actually has this structuralproperty of this family (25). On the other hand, YetL belongsto the MarR family of bacterial transcriptional regulators. Thecrystal structures of several MarR family members revealedthat they form a dimer structure with a common triangularshape, at the two corners of which winged helix-turn-helixDNA binding motifs are located (1, 5, 10). These DNA bindingmotifs consist of the internal region of each subunit, and theirN and C termini are intertwined with each other to form a coredomain. So far, several bacterial transcriptional regulators thatrecognize and respond to flavonoids have been reported (9, 13,30). However, to our knowledge, YetL is the first reportedmember of the MarR family which specifically responds toflavonoids.
The mechanisms underlying signal recognition by membersof the MarR family have not been well defined, and whether acommon recognition mechanism triggers their derepressionremains unclear. It has been reported that two members of theMarR family, B. subtilis OhrR and YodB, sense oxidative thiolstress through oxidative modification of their cysteine residues,which are located at the N terminus of OhrR and the N and Ctermini of YodB. This modification results in prevention ofDNA binding, which is followed by induction of the targetgenes involved in resistance to oxidizing compounds (10, 16).E. coli MarR, the prototype of the MarR family, can be dis-sociated from the operator DNA of the marRAB operon,which is involved in multidrug resistance through interactionwith a broad range of drugs, including salicylate (5). A highconcentration of salicylate was required to obtain the crystalstructure of MarR, in which a salicylate molecule was bound tothe surface of each of its DNA binding domains (1), suggestingthat inducer drugs are able to interfere with the MarR-DNAinteraction. The YetL inducers are unlikely to be so reactivethat covalent modification within the YetL readily occurs, as isthe case for B. subtilis OhrR and YodB, and are also unlikelyto directly interrupt the protein-DNA interaction, as is the casefor E. coli MarR, because YetL appears to recognize certainflavonoids strictly. We suppose that the flavonoid recognitionmechanism of YetL is distinct from those of OhrR, YodB, andMarR. Indeed, we found that YetL binding to the yetM cissequence is not inhibited by aromatic compounds, such ascatechol and salicylate, by means of gel retardation and lacZfusion analyses (data not shown).
In B. subtilis, the yxaG gene, one of the members of theLmrA/YxaF regulon, encodes quercetin 2,3-dioxygenase,which catalyzes the C-ring cleavage of quercetin, yielding2-protocatechuoyl-phloroglucinol carboxylic acid and carbonmonoxide (4). YxaG exhibits similar dioxygenase activity withseveral other flavonols (K. Hirooka and Y. Fujita, unpublishedresults). Thus, it is assumed that flavonols are converted by
YxaG to phenolic esters of aromatic carboxylic acids, whichcould be hydrolyzed to the corresponding aromatic carboxyliccompounds by an endogenous esterase with broad substratespecificity in B. subtilis. It has been reported that B. subtilispossesses the yfiE gene, whose expression is induced by cate-chol and which encodes catechol 2,3-dioxygenase, which con-verts catechol into 2-hydroxymuconic semialdehyde (29), andthe ywhB gene, which encodes 4-oxalocrotonate tautomerase,which catalyzes the interconversion of 2-hydroxymuconate and4-oxalocrotonate (36). This implies that catechol and its deriv-atives are degraded through the meta-cleavage pathway via thedehydrogenation route (27). Alternatively, highly electrophilicaromatic compounds, such as catechol and 2-methylhydroqui-none, can form S adducts with cellular thiols. These S adductsare assumed to be subjected to thiol-dependent ring cleavagefor detoxification by multiple dioxygenase/glyoxalase family en-zymes encoded by mhqA, mhqO, and mhqE, which respond tothiol stress and are regulated by MhqR, a MarR-type repressorwith an unknown derepression mechanism (32). The yetMgene, identified as the YetL target, is predicted to encode anFAD-dependent monooxygenase containing FAD and NADHbinding motifs. The superfamily that contains YetM also in-cludes salicylate monooxygenase that converts salicylate to cat-echol (15). In addition, it has been reported that an enzymefrom an Asteraceae plant species, Chrysanthemum segetum,catalyzes hydroxylation at position 8 of flavonols and flavonesusing FAD and NADPH as cofactors (6). Thus, we speculatethat YetM might catalyze the conversion of the salicylate de-rivatives derived from the flavonols through YxaG degradationor the direct hydroxylation of flavonols, followed by YxaGdegradation, yielding the catechol derivatives, which wouldflow into the catechol metabolic pathways described above.Our attempts to prepare the recombinant YetM protein in asoluble form were unsuccessful, and there was no difference ingrowth in the presence of any of the 12 flavonoids testedbetween the wild-type strain and the strain with yetM disrupted(data not shown); thus, the enzymatic function of YetM re-mains to be elucidated. Other approaches for characterizationof YetM are currently being examined.
ACKNOWLEDGMENTS
We are grateful to Y. Moritake, M. Watanabe, R. Yamamoto, K.Kumamoto, and T. Satomura for their help with the experiments. Wealso thank Kazuo Kobayashi (Nara Institute of Science and Technol-ogy, Japan) for providing strain YETLd and for providing the resultsof a DNA microarray analysis involving this strain, which involved asearch for candidate target genes for YetL.
This work was supported by a Grant-in-Aid for Scientific Researchon Priority Areas and by a grant from the High-Tech Research Projectfor Private Universities from the Ministry of Education, Culture,Sports, Science, and Technology of Japan.
REFERENCES
1. Alekshun, M. N., S. B. Levy, T. R. Mealy, B. A. Seaton, and J. F. Head. 2001.The crystal structure of MarR, a regulator of multiple antibiotic resistance,at 2.3 A resolution. Nat. Struct. Biol. 8:710–714.
2. Atkinson, M. R., L. V. Wray, Jr., and S. H. Fisher. 1990. Regulation ofhistidine and proline degradation enzymes by amino acid availability inBacillus subtilis. J. Bacteriol. 172:4758–4765.
3. Bais, H. P., T. L. Weir, L. G. Perry, S. Gilroy, and J. M. Vivanco. 2006. Therole of root exudates in rhizosphere interactions with plants and otherorganisms. Annu. Rev. Plant Biol. 57:233–266.
4. Bowater, L., S. A. Fairhurst, V. J. Just, and S. Bornemann. 2004. Bacillussubtilis YxaG is a novel Fe-containing quercetin 2,3-dioxygenase. FEBS Lett.557:45–48.
5. Grkovic, S., M. H. Brown, and R. A. Skurray. 2002. Regulation of bacterialdrug export systems. Microbiol. Mol. Biol. Rev. 66:671–701.
6. Halbwirth, H., and K. Stich. 2006. An NADPH and FAD dependent enzymecatalyzes hydroxylation of flavonoids in position 8. Phytochemistry 67:1080–1087.
7. Haldenwang, W. G. 1995. The sigma factors of Bacillus subtilis. Microbiol.Rev. 59:1–30.
8. Higuchi, R. 1990. Recombinant PCR, p. 177–183. In M. A. Innis, D. H.Gelfand, J. J. Sninsky, and T. J. White (ed.), PCR protocols: a guide tomethods and applications. Academic Press, San Diego, CA.
9. Hirooka, K., S. Kunikane, H. Matsuoka, K. Yoshida, K. Kumamoto, S. Tojo,and Y. Fujita. 2007. Dual regulation of the Bacillus subtilis regulon compris-ing the lmrAB and yxaGH operons and yxaF gene by two transcriptionalrepressors, LmrA and YxaF, in response to flavonoids. J. Bacteriol. 189:5170–5182.
10. Hong, M., M. Fuangthong, J. D. Helmann, and R. G. Brennan. 2005. Struc-ture of an OhrR-ohrA operator complex reveals the DNA binding mecha-nism of the MarR family. Mol. Cell 20:131–141.
11. Horinouchi, S., and B. Weisblum. 1982. Nucleotide sequence and functionalmap of pC194, a plasmid that specifies inducible chloramphenicol resistance.J. Bacteriol. 150:815–825.
12. Hund, H. K., J. Breuer, F. Lingens, J. Huttermann, R. Kappl, and S.Fetzner. 1999. Flavonol 2,4-dioxygenase from Aspergillus niger DSM 821, atype 2 CuII-containing glycoprotein. Eur. J. Biochem. 263:871–878.
13. Kobayashi, H., Y. Naciri-Graven, W. J. Broughton, and X. Perret. 2004.Flavonoids induce temporal shifts in gene-expression of nod-box controlledloci in Rhizobium sp. NGR234. Mol. Microbiol. 51:335–347.
14. Kobayashi, K., S. D. Ehrlich, A. Albertini, et al. 2003. Essential Bacillussubtilis genes. Proc. Natl. Acad. Sci. USA 100:4678–4683.
15. Lee, J. R., K. R. Min, Y. C. Kim, C. K. Kim, J. Y. Lim, H. J. Yoon, K. H. Min,K. S. Lee, and Y. Kim. 1995. Cloning of the salicylate hydroxylase gene andcatechol 2,3-dioxygenase gene, and sequencing of an intergenic sequencebetween the two genes of Pseudomonas putida KF715. Biochem. Biophys.Res. Commun. 211:382–388.
16. Leelakriangsak, M., N. T. T. Huyen, S. Towe, N. van Duy, D. Becher, M.Hecker, H. Antelmann, and P. Zuber. 2008. Regulation of quinone detoxi-fication by the thiol stress sensing DUF24/MarR-like repressor YodB inBacillus subtilis. Mol. Microbiol. 67:1108–1124.
17. Merkens, H., S. Sielker, K. Rose, and S. Fetzner. 2007. A new monocupinquercetinase of Streptomyces sp. FLA: identification and heterologous ex-pression of the queD gene and activity of the recombinant enzyme towardsdifferent flavonols. Arch. Microbiol. 187:475–487.
18. Miwa, Y., and Y. Fujita. 2001. Involvement of two distinct catabolite-respon-sive elements in catabolite repression of the Bacillus subtilis myo-inositol (iol)operon. J. Bacteriol. 183:5877–5884.
19. Murata, M., S. Ohno, M. Kumano, K. Yamane, and R. Ohki. 2003. Multi-drug resistant phenotype of Bacillus subtilis spontaneous mutants isolated inthe presence of puromycin and lincomycin. Can. J. Microbiol. 49:71–77.
20. Plaper, A., M. Golob, I. Hafner, M. Oblak, T. Solmajer, and R. Jerala. 2003.Characterization of quercetin binding site on DNA gyrase. Biochem. Bio-phys. Res. Commun. 306:530–536.
21. Ramos, J. L., M. Martínez-Bueno, A. J. Molina-Henares, W. Teran, K.Watanabe, X. Zhang, M. T. Gallegos, R. Brennan, and R. Tobes. 2005. TheTetR family of transcriptional repressors. Microbiol. Mol. Biol. Rev. 69:326–356.
22. Rao, J. R., and J. E. Cooper. 1994. Rhizobia catabolize nod gene-inducingflavonoids via C-ring fission mechanisms. J. Bacteriol. 176:5409–5413.
23. Sambrook, J., and D. W. Russell. 2001. Molecular cloning: a laboratory
manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Har-bor, NY.
24. Satomura, T., D. Shimura, K. Asai, Y. Sadaie, K. Hirooka, and Y. Fujita.2005. Enhancement of glutamine utilization in Bacillus subtilis through theGlnK-GlnL two-component regulatory system. J. Bacteriol. 187:4813–4821.
25. Seetharaman, J., D. Kumaran, J. B. Bonanno, S. K. Burley, and S. Swami-nathan. 2006. Crystal structure of a putative HTH-type transcriptional reg-ulator yxaF from Bacillus subtilis. Proteins 63:1087–1091.
26. Shaw, L. J., P. Morris, and J. E. Hooker. 2006. Perception and modificationof plant flavonoid signals by rhizosphere microorganisms. Environ. Micro-biol. 8:1867–1880.
27. Shingler, V., J. Powlowski, and U. Marklund. 1992. Nucleotide sequence andfunctional analysis of the complete phenol/3,4-dimethylphenol catabolicpathway of Pseudomonas sp. strain CF600. J. Bacteriol. 174:711–724.
28. Steinmetz, M., and R. Richter. 1994. Plasmids designed to alter the antibioticresistance expressed by insertion mutations in Bacillus subtilis, through invivo recombination. Gene 142:79–83.
29. Tam, L. T., C. Eymann, D. Albrecht, R. Sietmann, F. Schauer, M. Hecker,and H. Antelmann. 2006. Differential gene expression in response to phenoland catechol reveals different metabolic activities for the degradation ofaromatic compounds in Bacillus subtilis. Environ. Microbiol. 8:1408–1427.
30. Teran, W., T. Krell, J. L. Ramos, and M. T. Gallegos. 2006. Effector-repressor interactions, binding of a single effector molecule to the operator-bound TtgR homodimer mediates derepression. J. Biol. Chem. 281:7102–7109.
31. Tojo, S., T. Satomura, K. Kumamoto, K. Hirooka, and Y. Fujita. 2008.Molecular mechanisms underlying the positive stringent response of theBacillus subtilis ilv-leu operon, involved in the biosynthesis of branched-chainamino acids. J. Bacteriol. 190:6134–6147.
32. Towe, S., M. Leelakriangsak, K. Kobayashi, N. van Duy, M. Hecker, P.Zuber, and H. Antelmann. 2007. The MarR-type repressor MhqR (YkvE)regulates multiple dioxygenases/glyoxalases and an azoreductase which con-fer resistance to 2-methylhydroquinone and catechol in Bacillus subtilis. Mol.Microbiol. 66:40–54.
33. Ulanowska, K., A. Tkaczyk, G. Konopa, and G. Wegrzyn. 2006. Differentialantibacterial activity of genistein arising from global inhibition of DNA,RNA and protein synthesis in some bacterial strains. Arch. Microbiol. 184:271–278.
34. Vagner, V., E. Dervyn, and S. D. Ehrlich. 1998. A vector for systematic geneinactivation in Bacillus subtilis. Microbiology 144:3097–3104.
35. Wach, A. 1996. PCR-synthesis of marker cassettes with long flanking homol-ogy regions for gene disruptions in S. cerevisiae. Yeast 12:259–265.
36. Wang, S. C., W. H. Johnson, Jr., R. M. Czerwinski, S. L. Stamps, and C. P.Whitman. 2007. Kinetic and stereochemical analysis of YwhB, a 4-oxalocro-tonate tautomerase homologue in Bacillus subtilis: mechanistic implicationsfor the YwhB- and 4-oxalocrotonate tautomerase-catalyzed reactions. Bio-chemistry 46:11919–11929.
37. Yoshida, K., Y. Fujita, and A. Sarai. 1993. Missense mutations in the Bacillussubtilis gnt repressor that diminish operator binding ability. J. Mol. Biol.231:167–174.
38. Yoshida, K., I. Ishio, E. Nagakawa, Y. Yamamoto, M. Yamamoto, and Y.Fujita. 2000. Systematic study of gene expression and transcription organi-zation in the gntZ-ywaA region of the Bacillus subtilis genome. Microbiology146:573–579.
39. Yoshida, K., K. Kobayashi, Y. Miwa, C. M. Kang, M. Matsunaga, H.Yamaguchi, S. Tojo, M. Yamamoto, R. Nishi, N. Ogasawara, T. Nakayama,and Y. Fujita. 2001. Combined transcriptome and proteome analysis as apowerful approach to study genes under glucose repression in Bacillus sub-tilis. Nucleic Acids Res. 29:683–692.
VOL. 191, 2009 B. SUBTILIS YetL REGULON RESPONSIVE TO FLAVONOIDS 3697
