![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Biotechnology, Faculty of Life Science and Biotechnology, Fukuyama University, 985 Sanzo, Higashimura, Fukuyama 729-0292, Japan,1 Department of Biofunctional Chemistry, Faculty of Agriculture, Kobe University, 1-1 Rokkodai, Nada, Kobe 657-8501, Japan2
Received 16 January 2007/ Accepted 26 April 2007
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our previous study showed that LmrA interacts with a cis sequence in the lmrAB promoter region. With the combination of DNA microarray analysis and genome-wide computational sequence analysis, another LmrA target, yxaGH, was identified, which was found in the putative yxaGH promoter region two tandem LmrA-binding cis sequences, each of which is similar to that in the lmrAB promoter region (30). However, the DNA binding of LmrA was not inhibited by lincomycin, with the inducer interacting with LmrA remaining to be identified (30). Besides these findings, it was reported that yxaG encodes quercetin 2,3-dioxygenase, which is involved in flavonol degradation (3), whereas yxaH encodes a membrane protein of unknown function.
Flavonoids including quercetin, which are considerably abundant in soil, are secreted from the roots of higher plants and released through the degradation of plant cells. Some flavonoids possess antibacterial activity; quercetin inhibits bacterial DNA gyrase, which induces DNA cleavage (16, 24). To avoid such harmful effects, some bacteria are equipped with a degradation system for flavonoids for detoxification (18). Hence, the gene product of yxaG, encoding quercetin 2,3-dioxygenase, appears to play a major role in flavonoid degradation in B. subtilis.
LmrA is paralogous to YxaF, the gene of which is located immediately upstream of the yxaGH operon and oriented in the same direction as yxaGH (Fig. 1). On the assumption that LmrA and YxaF might recognize and bind to almost the same cis sequences located in the promoter regions of the target genes, we performed in vitro and in vivo experiments to characterize the operon organization of the regulon regulated by LmrA and YxaF, the LmrA/YxaF regulon. We also carried out experiments to identify the cis sequences to which LmrA and YxaF bind, i.e., LmrA/YxaF boxes, and to determine if some flavonoids such as quercetin might be inducers for the LmrA/YxaF regulon.
|
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
B. subtilis cells were pregrown on plates with tryptose blood agar base (Difco) supplemented with 0.18% glucose (TBABG) containing chloramphenicol (5 µg/ml), erythromycin (0.3 µg/ml), tetracycline (10 µg/ml), and/or lincomycin (25 µg/ml), according to drug resistance, at 30°C overnight. The cells were inoculated into Luria-Bertani (LB) medium (19) and then incubated at 37°C with shaking.
Overexpression of the lmrA and yxaF genes in Escherichia coli and LmrA and YxaF preparation. E. coli strain JM109 bearing plasmid pLMRA, a pUC18-based plasmid carrying the lmrA gene, was used for lmrA expression, the construction of which was described previously (30). To construct plasmid pET-LMRA, a pET-22b(+)-based plasmid carrying the lmrA gene, a region containing lmrA was amplified by PCR with genomic DNA of B. subtilis strain 168 as a template and primer pair LmrAN/LmrAB (Table 2), digested with NdeI and BamHI, and then ligated with vector pET-22b(+), which had been treated with the same restriction enzymes. The ligated DNA was used for the transformation of E. coli strain JM109 to ampicillin resistance (50 µg/ml) to produce an expression plasmid, pET-LMRA.
To clone the B. subtilis yxaF gene into plasmids pUC18 and pET-22b(+), the two yxaF fragments were amplified by PCR with genomic DNA of strain 168 as a template and the primer pairs YxaFE/YxaFB and YxaFN/YxaFB, respectively (Table 2). The resultant PCR products were digested with EcoRI/BamHI and NdeI/BamHI and then ligated with the arms of plasmids pUC18 and pET-22b(+) digested with the same restriction enzymes, respectively. Also, the ligated mixtures were used for the transformation of strain JM109 to ampicillin resistance to yield two expression plasmids, pYXAF and pET-YXAF, respectively. The correct cloning of the lmrA and yxaF genes in all the resultant plasmids was confirmed by DNA sequencing.
E. coli strain JM109 bearing plasmid pLMRA or pYXAF and strain BL21(DE3) transformed with plasmid pET-LMRA or pET-YXAF were grown in 50 ml of LB medium supplemented with ampicillin (50 µg/ml) at 37°C to an optical density at 600 nm (OD600) of 0.6. After isopropyl-ß-D-thiogalactopyranoside (IPTG) was added to a final concentration of 1 mM, the culture was further cultivated for 3 h. The cells were harvested and then disrupted by sonication in 20 mM Tris-Cl buffer (pH 8.0) containing 10% (vol/vol) glycerol, 0.1 mM phenylmethylsulfonyl fluoride, and 1 mM dithiothreitol. After centrifugation (17,000 x g for 30 min), the supernatants were recovered as crude extracts for the in vitro binding analyses of LmrA and YxaF. The protein concentrations of the crude extracts were measured by a method described previously Bradford (4) using bovine serum albumin as a standard.
Primer extension analysis.
Primer extension analysis to determine the transcription start sites of the yxaF gene and the yxaGH operon was performed essentially as described previously (20, 29). Quercetin dissolved in dimethyl sulfoxide (DMSO) was added to the culture in the logarithmic growth phase (OD600 of 0.2) to give 200 µg/ml, which was followed by a further 1 h of cultivation; the same volume of DMSO as that of the quercetin solution was also added to another control culture. Total RNA was extracted from the harvested cells and then purified as described previously (29). For primer extension of the yxaF gene and yxaGH transcripts, total RNA (45 µg) was annealed to 1 pmol each of primers PyxaF2 and PyxaG2 (Table 2), which had been 5'-end labeled with a MEGALABEL kit (Takara-Bio) and [
-32P]ATP (GE Healthcare). The primer extension reaction was conducted with ThermoScript reverse transcriptase (Invitrogen). Templates for the dideoxy sequencing reactions for ladder preparation, starting from the same 5'-end-labeled primers as those for yxaF and yxaGH, were generated by PCR with genomic DNA of strain 168 as a template and primer pairs PyxaF1/PyxaF2 and PyxaG1/PyxaG2, respectively (Table 2). Autoradiograms were obtained and quantified using a Typhoon 9400 variable image analyzer (GE Healthcare).
DNase I footprinting analysis.
DNase I footprinting experiments were performed essentially as described previously (6, 30). The lmrAB, yxaGH, and yxaF probes for the footprinting analyses were prepared by PCR with genomic DNA of strain 168 as a template and primer pairs MK1/MK2, PyxaG1/PyxaG2, and PyxaF1/PyxaF2 (Table 2), respectively. Prior to the PCR amplification, only the 5' terminus of one of the primer pairs was labeled with [-32P]ATP using the MEGALABEL kit.
Gel retardation analysis.
The gel retardation analysis was performed essentially as described previously (27, 30). The LmrA and YxaF extracts prepared from BL21(DE3) cells bearing pET-LMRA and pET-YXAF, respectively, were dialyzed against the same buffer as that used for the extract preparation to eliminate contaminating small molecules from the cells, which might affect the DNA binding of the protein. The amounts of LmrA and YxaF compared to total protein in the dialyzed extracts were estimated by measuring the intensities of the bands on the polyacrylamide gel stained with Coomassie brilliant blue using NIH Image (http://rsb.info.nih.gov/nih-image/). The same lmrAB, yxaGH, and yxaF probes as those used for DNase I footprinting were labeled by PCR in the presence of [
-32P]dCTP (MP Biomedicals) with the same primer pairs as those used for the probe amplification.
DNA microarray analysis. DNA microarray analysis was performed as described previously (11, 15, 29). RNA samples were prepared from cells of strains 168 and FU877 that had been grown in LB medium (19) and harvested in the middle of logarithmic growth at an OD600 of 0.5.
lacZ fusion experiments to monitor lmrAB and yxaGH. Cells pregrown on TBABG plates containing an antibiotic(s) were inoculated into 50 ml of LB medium at an initial OD600 of 0.05 and then incubated at 37°C with shaking. When the OD600 reached 0.2, each of the flavonoids dissolved in DMSO was added to the medium to give a final concentration of 200 µg/ml, corresponding to 0.6, 0.7, 0.6, 0.7, 0.6, 0.7, 0.8, and 0.7 mM for quercetin, fisetin, tamrixetin, galangin, (+)-catechin, genistein, daidzein, and coumestrol, respectively. As a control, 200 µl of DMSO was added instead of the flavonoid solution. One-milliliter aliquots of the culture were then withdrawn at intervals of 1 h, and the ß-galactosidase (ß-Gal) activity in crude cell extracts was spectrophotometrically measured by using o-nitrophenyl-ß-D-galactopyranoside (Wako Pure Chemicals Industries, Japan) as a substrate, according to a procedure described previously (1). To reduce the chromatic disturbance of the ß-Gal assay by the flavonoid that adhered to the cells, the latter was washed with 100 mM phosphate buffer (pH 7.5) before lysozyme treatment.
Flavonoid screening. Cells of strain YXAGd were pregrown in LB medium containing 0.3 µg/ml erythromycin overnight at 25°C. The cells were spread well onto TBABG plates containing 50 µg/ml 5-bromo-4-chloro-3-indolyl-ß-D-galactoside, onto which paper disks soaked in 10 mM each of the various flavonoids were placed, and the plates were then incubated at 37°C. The blue coloration around the disk(s), resulting from the synthesis of ß-Gal regulated by LmrA and YxaF exposed to the flavonoids, was monitored.
Assay of lincomycin resistance. The lincomycin resistance of B. subtilis cells was assayed by a method described previously (14). Cells pregrown on TBABG plates were inoculated to an OD590 of 0.05 in 200 µl of LB medium containing serially diluted lincomycin solutions in the wells of a 96-well titer plate. The growth of cells incubated with shaking was monitored by measuring the OD590 4 h after inoculation, when the cells had reached the late logarithmic phase in medium containing no lincomycin. The concentration of lincomycin that resulted in 50% growth inhibition (IC50) was determined. Experiments were also conducted in the presence of quercetin (0.3 mM) to examine its effect on lincomycin resistance.
Assay of quercetin 2,3-dioxygenase. B. subtilis cells pregrown on TBABG plates were inoculated into 50 ml of LB medium to an OD600 of 0.05 and then cultivated at 37°C with shaking. When the OD600 reached 0.8, the cells were harvested and washed with 50 mM Tris-Cl buffer (pH 7.5) containing 100 mM NaCl, 20 mM NH4Cl, and 1% (vol/vol) glycerol. The cells were then suspended in the same buffer containing 50 µg/ml lysozyme and 5 µg/ml DNase I and incubated at 37°C for 20 min. After centrifugation (17,000 x g for 30 min), the supernatant was recovered as a crude extract. The protein concentrations of the crude extracts prepared were measured as described above.
Quercetin 2,3-dioxygenase activity was determined by measuring the decrease of the absorbance at the characteristic spectroscopic peak of quercetin (
367 = 22,700 M–1cm–1 at pH 7.5) due to quercetin decomposition, as reported previously (2). For each assay, 870 µl of 50 mM Tris-Cl buffer (pH 7.5) was added to a quartz cell. Quercetin dissolved in DMSO (30 µl) was then mixed to give 45 µM, followed by preincubation at 25°C for 5 min. The enzymatic reaction was started by adding 100 µl of a crude extract (10 to 100 µg total protein) to the mixture, and the absorbance at 367 nm was monitored at 25°C.
Flavonoids. Quercetin, fisetin, (+)-catechin, genistein, daidzein, 3-hydroxyflavone, myricetin, morin, and kaempferol were products from Sigma. Tamarixetin and galangin were purchased from Nacalai Tesque (Japan), coumestrol was obtained from Fluka, and robinetin was obtained from Extrasynthese SA.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
A RNA polymerase (7).
|
|
Quercetin was found to trigger yxaF and yxaGH transcription (Fig. 3), implying that it might be an inducer for the LmrA/YxaF regulon. When the DNase I footprinting experiments were carried out in the presence of 1.5 mM quercetin, the specifically protected regions of lmrAB, yxaGH, and yxaF disappeared (Fig. 4, lanes 6). These results clearly indicated that quercetin prevents the binding of LmrA and YxaF to the lmrAB, yxaGH, and yxaF LmrA/YxaF boxes, implying that quercetin might function as an inducer for the LmrA/YxaF regulon.
Quantitative analysis of quercetin inhibition of the LmrA and YxaF binding to the LmrA/YxaF boxes. To analyze quercetin inhibition of LmrA and YxaF binding to the LmrA/YxaF boxes quantitatively, we performed gel retardation analysis involving the LmrA or YxaF protein and each of the same lmrAB and yxaGH probes as those used for DNase I footprinting (Fig. 4). We attempted to purify the LmrA and YxaF proteins to near homogeneity but failed to obtain pure active LmrA due to its instability. Instead, we used the LmrA and YxaF preparations of 38 and 35% purity, respectively, as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (PAGE) (data not shown), for this gel retardation analysis, which had been dialyzed to remove small molecules derived from the cell extracts.
As shown in Fig. 5A, both LmrA and YxaF bound to each of the lmrAB and yxaGH probes containing the lmrAB and yxaGH promoter regions, which produced the retarded bands detected by PAGE. The binding of LmrA and YxaF to the lmrAB probe was weaker than that to yxaGH, and LmrA binding to the two probes was weaker than that of YxaF. (The binding and association constants for LmrA and YxaF binding to the lmrAB and yxaGH probes were very roughly estimated to be 15 and 5 nM, and 10 and 1 nM, respectively, as dimers [Fig. 5A]; LmrA and YxaF presumably each exist as homodimers, as deduced from the crystal structure of YxaF [21].) When YxaF was mixed with the yxaGH probe, two steps of retardation were observed, implying that each YxaF dimer might bind stepwise to the two LmrA/YxaF boxes in the yxaGH promoter. However, the binding of LmrA to the yxaGH probe did not exhibit such a stepwise retardation pattern clearly.
|
Relief from repression of the lmrAB and yxaGH promoters in vivo by LmrA and YxaF upon quercetin addition.
Since quercetin was found to inhibit the binding of LmrA and YxaF to the LmrA/YxaF boxes in vitro, we examined, through lacZ fusion experiments, if it actually functions as an inducer for the LmrA/YxaF regulon, i.e., if it releases the repression of the lmrAB and yxaGH promoters through the LmrA and YxaF repressors. Plasmid pMUTIN2 integration not only disrupts the target gene and replaces it with lacZ as a reporter of its expression, it also places its downstream genes under the control of the spac promoter, which is repressed by E. coli LacI (25). To monitor lacZ expression under the control of either the lmrAB or yxaGH promoter in the presence of LmrA or YxaF, we constructed a series of plasmid pMUTIN2 integrants of the lmrA, lmrB, yxaG, or yxaF gene, which carry either a yxaF or an lmrA disruption (Table 1). Strains FU878 (lmrB::pMUTIN2 
yxaF::cat) and FU879 (yxaG::pMUTIN2 yxaF::cat) were used to monitor the lmrAB and yxaGH promoters, respectively, controlled by only LmrA. When quercetin was added to the culture in the logarithmic growth phase, a clear induction of ß-Gal synthesis was observed in these strain cells, suggesting that quercetin was incorporated into the cells and functioned as an inducer to release the repression of the two promoters mediated only by LmrA (Fig. 6).
|
yxaF::cat) (3.2 nmol/min per OD), suggesting that the lmrAB promoter is repressed mainly by LmrA and not YxaF, and the promoter was further derepressed by quercetin (Fig. 6). (The basal ß-Gal activity of strain LMRBd [lmrB::pMUTIN] possessing both LmrA and YxaF was 1.48 nmol/min per OD [data not shown].) The yxaGH promoter controlled only by YxaF was monitored by using strain FU880 (yxaG::pMUTIN2 lmrA::tet), which indicated that quercetin derepressed it in a manner similar to that in the case of the lmrAB promoter. The basal activity of strain FU880 without the addition of quercetin (0.21 nmol/min per OD) was lower than that of strain FU879 (yxaG::pMUTIN2 yxaF::cat) (4.3 nmol/min per OD), suggesting that YxaF might bind to the two LmrA/YxaF boxes in the yxaGH promoter region more tightly than LmrA, causing strict repression of the promoter. (The basal ß-Gal activity of strain YXAGd [yxaG::pMUTIN] possessing LmrA and YxaF was 0.12 nmol/min per OD [data not shown].)
Although the promoter activities enhanced by quercetin gradually declined upon the further cultivation of strains FU878 (yxaF::cat lmrB::pMUTIN2) and LMRAd (lmrA::pMUTIN2), the activities were sustained during the cultivation of strains FU879 (yxaF::cat yxaG::pMUTIN2) and FU880 (lmrA::tet yxaG::pMUTIN2) (Fig. 6). This observation suggests that the yxaG disruption in the latter strains likely prohibits quercetin degradation, leading to the continuous derepression of LmrA and YxaF by quercetin.
Strains FU877 (lmrA::pMUTIN2 yxaF::cat) and FU881 (yxaG::pMUTIN2 lmrA:tet yxaF::cat) were constructed to monitor the activities of the lmrAB and yxaGH promoters in the absence of both LmrA and YaxF, respectively. Figure 6 shows that the ß-Gal activities of strains FU877 (41.5 nmol/min per OD) and FU881 (17.3 nmol/min per OD) were not further enhanced upon the addition of quercetin. (The constitutive expression of yxaGH in strain FU881 was three times lower than its quercetin-triggered maximal expression in strain FU880, which is not explained properly at present.) Interestingly, the addition of quercetin provoked dramatic growth inhibition of only strain FU877 (Fig. 6); the growth of strain FU881 was not seriously affected by the addition of quercetin. From these results, we considered that the addition of quercetin might be harmful to the cells in which either or both of the yxaG and yxaH genes, encoding quercetin 2,3-dioxygenase and a putative transporter, respectively, are completely derepressed. To determine if yxaH is involved in this growth inhibition, yxaH expression under the control of the IPTG-inducible spac promoter in strain YXAGd (yxaG::pMUTIN2) was induced by the addition of 5 mM IPTG. It was found that the induced expression of the yxaH gene did not affect cell growth even when quercetin was added to the medium (data not shown). Moreover, the addition of quercetin also provoked dramatic growth inhibition of strain FU967 (lmrA::tet yxaF::cat yxaH::pMUTIN2), as observed with strain FU877 (lmrA::pMUTIN2 yxaF::cat) (data not shown). These results suggest that the derepression of the synthesis of quercetin 2,3-dioxygenase encoded by yxaG synthesis is likely the reason for the growth defect in the presence of quercetin. This might be due to the accumulation of an unidentified intermediate(s) derived through rapid quercetin degradation by excess 2,3-dioxygenase, which is harmful to B. subtilis.
Screening of candidate flavonoids besides quercetin that derepress the control of LmrA and YxaF repressors. The fact that quercetin can interact with the LmrA and YxaF repressors to derepress the lmrAB and yxaGH promoters prompted us to investigate if other flavonoids could interact with them similarly. We devised a screening method for finding compounds such as flavonoids that inhibit the LmrA and YxaF repressors. We used strain YXAGd (yxaG::pMUTIN2) to monitor ß-Gal synthesis reflecting yxaGH expression under the control of both LmrA and YxaF. If a tested flavonoid interacts with these regulatory proteins to dissociate them from the promoter, blue coloration would be produced around paper disks containing the flavonoid on a cell lawn on a TBABG plate containing 5-bromo-4-chloro-3-indolyl-ß-D-galactoside. When 13 flavonoid compounds including quercetin were tested, quercetin, fisetin, tamarixetin, genistein, and coumestrol gave clear positive results, indicated by blue coloration around the disks, but 3-hydroxyflavone, galangin, robinetin, myricetin, morin, kaempferol, (+)-catechin, and daidzein did not (data not shown). We selected the newly found four positive compounds as candidate inducers and galangin, (+)-catechin, and daidzein as negative compounds, which were found with this screening system involving strain YXAGd, for further detailed in vitro and in vivo analyses to examine their interactions with LmrA and YxaF repressors.
Inhibitory effects of various flavonoids on the in vitro DNA binding of LmrA and YxaF. To determine whether the above-mentioned seven flavonoids besides quercetin can interact with LmrA and YxaF to specifically detach them from DNA, we carried out gel retardation analyses involving each of the flavonoids similarly to methods described above, the results of which are summarized in Table 3. When the lmrAB probe was incubated with LmrA in the presence of each of the flavonoids, clear inhibition of the DNA binding was observed with quercetin, fisetin, and (+)-catechin, whereas significant inhibition was observed with galangin, genistein, and coumestrol but not with tamarixetin and daidzein (Table 3). When binding analysis involving the lmrAB probe and the YxaF protein was conducted, clear inhibition was observed with flavonols (quercetin, fisetin, tamarixetin, and galangin), whereas significant inhibition was observed with genistein and coumestrol but not with (+)-catechin and daidzein.
|
Suppressive effects of various flavonoids on the repression of the lmrAB and yxaGH promoters by LmrA and YxaF in vivo. Since each of the seven flavonoids besides quercetin was found to inhibit the binding of LmrA and YxaF to the LmrA/YxaF boxes in vitro in different manners, we examined, through lacZ fusion experiments, if they actually release the repression of the lmrAB and yxaGH promoters through the LmrA and YxaF repressors, depending on the repressor or LmrA/YxaF box involved, as observed by in vitro gel retardation analysis.
The results are summarized in Table 3, together with the those obtained from the in vitro DNA binding assay. The lmrAB promoter activity controlled by only LmrA was monitored using strain FU878 (yxaF::cat lmrB::pMUTIN2). When quercetin, fisetin, genistein, or coumestrol was added to the medium, the ß-Gal activity increased significantly, with fisetin being the most effective (Table 3). The addition of (+)-catechin caused slight but continuous ß-Gal synthesis (data not shown), but tamarixetin, galangin, and daidzein hardly induced it (Table 3). The results indicate that the five flavonoids that enhance ß-Gal synthesis are able to be incorporated into the cells and to function as inducers to release LmrA from the promoter. Furthermore, the derepression of the lmrAB promoter repressed by YxaF was examined using strain LMRAd (lmrA::pMUTIN2) (Table 3). The ß-Gal activity was well induced not only by quercetin, fisetin, genistein, and coumestrol but also by tamarixetin and daidzein; the latter did not suppress the lmrAB repression by LmrA. However, galangin and (+)-catechin did not release this repression. These in vivo results essentially agreed with those obtained using the in vitro gel retardation analysis, except that only galangin did not suppress the repression of lmrAB through LmrA or YxaF.
The yxaGH promoter activity controlled by LmrA only was monitored by measuring ß-Gal synthesis in strain FU879 (yxaF::cat yxaG::pMUTIN2). When quercetin, fisetin, genistein, or coumestrol was added to the culture, the ß-Gal activity increased substantially, with fisetin being the most effective. The addition of (+)-catechin enhanced it slightly, and the elevated activity was sustained during cultivation (data not shown). These results coincided well with the results of the corresponding in vitro experiments, except that galangin did not release the repression in vivo. On the other hand, the activity of the yxaGH promoter repressed only by YxaF was monitored by using strain FU880 (lmrA::tet yxaG::pMUTIN2); the addition of quercetin, fisetin, or genistein induced ß-Gal synthesis significantly, which is consistent with the results of the in vitro analysis. However, in conflict with the in vitro results, tamarixetin and coumestrol also clearly induced ß-Gal synthesis (Table 3) The probable reasons for this contradiction are discussed below. Galangin did not induce ß-Gal synthesis in this case, either. The fact that galangin did not induce ß-Gal synthesis in any case, contrary to the in vitro results, suggests that galangin is unable to be incorporated into B. subtilis cells.
The ß-Gal syntheses reflecting the respective lmrAB and yxaGH promoter activities in strains FU877 (yxaF::cat lmrA::pMUTIN2) and FU881 (lmrA::tet yxaF::cat yxaG::pMUTIN2),defective in both LmrA and YxaF, were constitutive, regardless of the addition of each of the flavonoids. However, the severe growth defect of strain FU877 observed with quercetin (Fig. 6) did not occur with any of the other flavonoids (data not shown).
Dual regulation of MDR and quercetin 2,3-dioxygenase synthesis by LmrA and YxaF.
Previous studies showed that the inactivation of the LmrA protein caused the constitutive expression of lmrB to elevate resistance to drugs such as lincomycin and puromycin (14, 30). The results obtained in this work revealed that in addition to LmrA, YxaF also binds to the lmrAB promoter region to repress lmrAB expression. Thus, to determine whether YxaF also regulates this MDR reflecting lmrB expression, we measured the growth rate in LB medium containing various concentrations of lincomycin and determined IC50s of B. subtilis strains in which either or both LmrA and YxaF were inactivated (Table 4). Strain PLR2, in which mutant LmrA with two point mutations is unable to repress LmrB synthesis, showed higher lincomycin resistance than strain 168 (Table 4), as reported previously (30). On the other hand, strain FU875 (yxaF::cat) exhibited almost the same sensitivity to lincomycin as strain 168. The introduction of the yxaF disruption in a PLR2 background to the constructed strain FU899 caused a slight increase in lincomycin resistance in comparison with that of strain PLR2 (Table 4). These results indicate that LmrA regulates lmrAB expression predominantly, whereas YxaF contributes to its regulation accessorily.
|
yxaF::cat) exhibited considerably high enzymatic activity in comparison with that of strain 168, whereas the extract of the lmrA mutant strain PLR2 exhibited higher activity than that of strain FU875. The enzyme activity in the extract of the lmrA and yxaF doubly mutated strain FU899 was the highest among those of the four strains. The results indicate that both LmrA and YxaF repress yxaGH expression, although LmrA contributes more to its regulation than YxaF does. Moreover, YxaF likely represses the yxaGH promoter more preferentially than the lmrAB promoter. When the cells of the above-described four strains were exposed to quercetin for 2 h under the cultivation conditions adopted for the lacZ fusion experiments (Fig. 6), only the quercetin 2,3-dioxygenase activity in the lysate obtained from strain 168 was accurately determined due to the harmful effect of quercetin on the mutant cells, which was 0.860 ± 0.182 µmol/min per mg protein. This value is higher than the derepressed activity of strain FU899 (0.568 ± 0.136 µmol/min per mg), indicating that quercetin is able to completely inhibit LmrA and YxaF in vivo. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
yxaF::cat lmrA::pMUTIN2) deficient in LmrA and YxaF, which were grown to the mid-logarithmic phase in LB medium (data not shown). However, we were able to find no other candidate target genes for LmrA/YxaF regulon members, which are presumably repressed by both LmrA and YxaF, upon comparisons of the two transcriptome patterns, suggesting that the members of the LmrA/YxaF regulon might be only the above-described three genes/operons revealed in this study. The TetR family of bacterial transcriptional regulatory proteins is known to typically possess two functional domains, a highly conserved N-terminal DNA binding domain and a less conserved C-terminal domain involved in both dimerization and effector binding (17). Recently, a crystal structure of the YxaF protein was determined (21). As predicted from the crystal structure, the YxaF protein possesses the HTH motif constituting the core of the putative DNA binding domain in the N-terminal region and forms the dimer structure in which the C-terminal domains are associated with each other. In this study, we found that the binding of LmrA and YxaF to the above-described four LmrA/YxaF boxes is inhibited by several flavonoids, such as quercetin and fisetin, resulting in the suppression of the lmrAB, yxaF, and yxaGH promoters repressed by them (Fig. 3 to 6 and Table 3). Thus, the effectors interacting with the C-terminal domains of LmrA and YxaF, i.e., inducers of the LmrA/YxaF regulon, are concluded to be the effective flavonoids shown in Table 3. Although LmrA and YxaF are quite similar to each other in function, they are distinct from each other in the following aspects. First, LmrA is less stable than YxaF; LmrA loses the DNA binding activity easily during the purification steps (data not shown). Second, the affinity of binding of YxaF to the LmrA/YxaF boxes is higher than that of LmrA; YxaF exhibits extremely high affinity, especially to the tandem boxes in the yxaGH promoter (Fig. 5A). The lower binding affinity of LmrA than that of YxaF (Fig. 5A) might be partially attributed to the instability of LmrA. Finally, the flavonoids that specifically inhibit LmrA binding to the LmrA/YxaF boxes are substantially different from those that inhibit YxaF binding to them (Table 3).
According to the results of the in vitro experiments (Fig. 5B and Table 3), LmrA binding to the lmrAB and yxaGH LmrA/YxaF boxes was well inhibited by quercetin, fisetin, and (+)-catechin, whereas it was less inhibited by galangin, genistein, and coumestrol and was inhibited hardly at all by tamarixetin and daidzein. When the inhibitory effects of the flavonols (quercetin, fisetin, tamarixetin, and galangin) on the DNA binding of LmrA are compared (Table 3), those of quercetin and fisetin can be seen to be higher than that of galangin, and tamarixatin has no inhibitory effect on it. These results suggest that the 4'-hydroxyl group at the B ring of flavonols (refer to the flavone structure for the ring and carbon assignments in Table 3) is essential for the DNA binding of LmrA but that the 5-hydroxyl group at the A ring is dispensable. Moreover, (+)-catechin is very inhibitory to the DNA binding of LmrA, suggesting that the plane structure of the center ring of the flavonoid might not be a determinant for its inhibitory effect. Isoflavones (genistein, daidzein, and coumestrol) exhibit less inhibitory effects than quercetin and fisetin, implying that the positioning of the B ring at the 3 carbon of the A ring is improper for the inhibitory effect. Coumestrol was also less recognized by LmrA than quercetin and fisetin.
On the other hand, the in vitro experiments involving YxaF revealed that YxaF binding to the lmrAB probe is well inhibited by all the flavonols tested (quercetin, fisetin, tamarixetin, and galangin) but was less inhibited by the isoflavones tested (genistein, daidzein, and coumestrol) (Table 3). YxaF binding to the yxaGH probe carrying the tandem LmrA/YxaF boxes was tighter than that to the lmrAB probe; the binding and association constants were roughly 5 and 1 nM, respectively (Fig. 5A). Moreover, this tight binding of YxaF to the yxaGH probe was less inhibited by quercetin (Fig. 5B) and fisetin (data not shown) than that to the lmrAB probe; the Ki values of quercetin were approximately 0.05 mM and 0.5 mM for YxaF binding to the lmrAB and yxaGH probes, respectively. Smaller Ki values for YxaF binding to the lmrAB probe were also obtained with the other flavonoids (fisetin, tamarixetin, galangin, genistein, and coumestrol) (Table 3). In contrast to LmrA binding to the LmrA/YxaF boxes, YxaF binding was not inhibited by (+)-catechin, suggesting that the plane structure of the flavonol central ring might be indispensable for its inhibition effect.
The in vivo experiments involving the lacZ reporter gene showed that several flavonoids were able to induce the expression of the reporter gene placed downstream of the lmrAB or yxaGH promoter in the presence of either LmrA or YxaF, which well supports the in vitro results described above, except that galangin did not induce the ß-Gal activity in any case, probably due to insufficient incorporation of galangin into the cells, and except that in contrast to the in vitro results, the repression of yxaGH by YxaF is well released by effective flavonoids in vivo, such as tamarixetin and coumestrol, as observed with that of lmrAB (Table 3). The latter contradiction might be possibly attributed to the possibility that the concentration of YxaF used in the in vitro experiments (125 nM) is likely higher than that in vivo because yxaF expression was very low as observed with the lacZ fusion experiments (data not shown); a lower concentration of YxaF might be sufficient to cause the repression of the lmrAB and yxaGH promoters under the physiological in vivo conditions, which could be easily relieved by the effective flavonoids.
As shown in Fig. 4, the LmrA and YxaF repressors recognize and bind to the LmrA/YxaF boxes of the LmrA/YxaF regulon. Thus, the physiological significance of the dual repression system through LmrA and YxaF for the lmrAB and yxaGH operons should be addressed, focusing on the predominance of either LmrA or YxaF for the repression of the lmrAB and yxaGH promoters. As shown in Fig. 5A, YxaF exhibited the highest affinity to the yxaGH probe, although it is not clear to which probe LmrA exhibited higher affinity. The in vivo results show that the lmrAB promoter was relatively more repressed by LmrA than by YxaF, whereas the yxaGH promoter was more tightly repressed by YxaF than by LmrA (refer to the basal ß-Gal activities in Fig. 6). On the other hand, the lincomycin resistance reflecting lmrB expression was predominantly repressed by LmrA and was repressed by YxaF accessorily (Table 4). Furthermore, quercetin 2,3-dioxygenase synthesis reflecting yxaG expression was repressed well by YxaF and more by LmrA. From these in vivo results, it is concluded that LmrA represses both the lmrAB and yxaGH promoters but that YxaF represses the yxaGH promoter more preferentially.
The dual regulation of yxaGH expression through LmrA and YxaF might be related to the degradation of toxic flavonoids as follows. Some flavonoids are known to possess antibacterial activity (24). Out of the eight flavonoids tested for the inhibitory effect on the DNA binding of LmrA and YxaF, quercetin exhibited the most significant inhibitory effect on the growth of B. subtilis cells when each flavonoid was added at a concentration of 200 µg/ml (Fig. 6). This inhibitory effect of quercetin was most pronounced when it was added to the culture of strain FU877 (yxaF::cat lmrA::pMUTIN2), in which yxaGH expression was completely derepressed in the absence of LmrA and YxaF. This severe growth-inhibitory effect of quercetin was most likely due to the derepression of yxaG, encoding quercetin 2,3-dioxygenase, because the induction and disruption of yxaH did not enhance it, suggesting that the rapid accumulation of a harmful degradation intermediate(s) of quercetin might damage the mutant cells, resulting in growth inhibition. This fact might be related to the complicated dual regulation of yxaGH expression through LmrA and YxaF, as described above, because cells must rigorously regulate 2,3-dioxygenase activity toward a wide range of concentrations of toxic flavonoids such as quercetin.
The LmrA/YxaF regulon characterized in this work is the first system in the Bacillus genus in which flavonoids induce a regulon through the inactivation of LmrA and YxaF. Some other bacterial transcriptional regulators that recognize and respond to flavonoid compounds were reported previously. NodD regulators, which belong to the LysR family and control the transcription of the nod operons involved in the nodulation of Rhizobiales in response to flavonoid signals released by their leguminous hosts (9), have been characterized in detail. Moreover, in Pseudomonas putida DOT-T1E, the resistance-nodulation-cell division family transporter TtgABC and its cognate TetR family repressor TtgR both possess multidrug recognition properties, and several flavonoids are substrates of TtgABC and activate pump expression by binding to the TtgR-operator complex to dissociate it (23). Hence, it might not be rare that flavonoids function as signaling molecules for communication among soil bacteria and plants. However, our attempt to find a conserved motif for flavonoid recognition by aligning the amino acid sequences of these regulators was unsuccessful.
An interesting question arises regarding why some unique flavonoids participate in the resistance to structurally unrelated drugs such as lincomycin via the LmrA/YxaF regulation system in B. subtilis. All of the lmrAB-related MDR mutants isolated so far possess mutations in the LmrA/YxaF binding site of the lmrAB promoter and/or the lmrA gene, which causes lmrB overexpression (12, 14), and lincomycin itself does not act as an inducer molecule for LmrA (30) and YxaF (data not shown). Nonetheless, this puzzle remains to be elucidated. It is also unclear whether the degradation of quercetin to 2-protocatechuoyl-phloroglucinol carboxylic acid by the YxaG protein is aimed at detoxification or at dissimilation to utilize it as a carbon source. The yxaH gene encodes a putative membrane protein with nine transmembrane segments. Although its physiological function remains unknown, the YxaH membrane protein might contribute to either the degradation or efflux of related flavonoids. It would be intriguing to reveal the overall metabolic pathway for flavonoids in B. subtilis.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by a grant-in-aid for Young Scientists (B) to K. Hirooka, grants-in-aid for Scientific Research on Priority Areas, Scientific Research (B), and the High-Tech Research Center Project for Private Universities to Y. Fujita from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.
|
FOOTNOTES |
|---|
Published ahead of print on 4 May 2007.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||
-independent transcription terminators are illustrated. The –35 and –10 regions for each operon and gene are underlined, and their transcription start sites (+1) and Shine-Dalgarno (SD) sequences are boxed. The LmrA/YxaF boxes comprising 18 bp are shown in boldface type; the box sequence in the lmrAB promoter region is reversed.
