Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Iida, T. Articles by Kudo, T. Search for Related Content PubMed PubMed Citation Articles by Iida, T. Articles by Kudo, T.

 Previous Article  |  Next Article 

Journal of Bacteriology, January 2009, p. 123-134, Vol. 191, No. 1
0021-9193/09/$08.00+0     doi:10.1128/JB.01112-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.

The GAF-Like-Domain-Containing Transcriptional Regulator DfdR Is a Sensor Protein for Dibenzofuran and Several Hydrophobic Aromatic Compounds ^,

Toshiya Iida,^1* Taro Waki,^1,2 Kaoru Nakamura,¹ Yuki Mukouzaka,¹ and Toshiaki Kudo^1,2

Environmental Molecular Biology Laboratory, RIKEN (The Institute of Physical and Chemical Research), Wako, Saitama 351-0198, Japan,¹ Supramolecular Biology, International Graduate School of Arts and Science, Yokohama City University, Suehiro, Tsurumi-ku, Yokohama 230-0045, Japan²

Received 8 August 2008/ Accepted 15 October 2008

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Dibenzofuran (DF) is one of the dioxin carbon skeletal compounds used as a model to study the microbial degradation of dioxins. This study analyzed the transcriptional regulation of the DF dioxygenase genes dfdA1 to dfdA4 in the DF-utilizing actinomycetes Rhodococcus sp. strain YK2 and Terrabacter sp. strain YK3. An open reading frame designated dfdR was detected downstream of the dfdC genes. The C-terminal part of the DfdR amino acid sequence has high levels of similarity to several LuxR-type DNA binding helix-turn-helix domains, and a GAF domain sequence in the central part was detected by a domain search analysis. A derivative of YK2 with dfdR disrupted was not able to utilize DF and did not exhibit DF-dependent dfdA1 transcriptional induction ability, and these dysfunctions were compensated for by introduction of dfdR. Promoter analysis of dfdA1 in Rhodococcus strains indicated that activation of the dfdA1 promoter (P_dfdA1) was dependent on dfdR and DF and not on a metabolite of the DF pathway. The cell extract of a Rhodococcus strain that heterologously expressed DfdR showed electrophoretic mobility shift (EMS) activity for the P_dfdA1 DNA fragment in a DF-dependent manner. In addition, P_dfdA1 activation and EMS activity were observed with hydrophobic aromatic compounds comprising two or more aromatic rings, suggesting that DfdR has broad effector molecule specificity for several hydrophobic aromatic compounds.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Concern about environmental pollution caused by chlorinated dioxins is increasing as these compounds are unintentionally produced as by-products of several industrial processes. We studied the microbial degradation of dioxins by using dibenzofuran (DF) and dibenzo-p-dioxin as model compounds. To date, several DF-utilizing bacterial strains have been isolated, and the routes of DF and dibenzo-p-dioxin catabolism have been analyzed (35, 55). The gram-negative DF-utilizing bacterium Sphingomonas wittichii RW1 has been isolated (56) and analyzed in detail. High-G+C-content gram-positive bacterial strains capable of utilizing DF as a carbon and energy source, such as Terrabacter sp. strain DBF63 (34), Terrabacter sp. strain YK3, and Rhodococcus sp. strain YK2 (21), have also been isolated. In addition, we recently reported isolation of a low-G+C-content spore-forming DF-utilizing bacterium, Paenibacillus sp. strain YK5 (22).

The initial pathway for bacterial degradation of DF and the enzymes catalyzing the reactions in DF-utilizing actinomycetes are shown in Fig. 1A. The first step of DF degradation is catalyzed by the Rieske nonheme iron oxygenases DfdA1 to DfdA4; these enzymes have regiospecificities for the oxygenation of DF at position 4,4a and are thus termed angular dioxygenases. The 4,4a-dihydroxylated reaction product (compound II) is unstable and spontaneously converted to 2,2',3-trihydroxybiphenyl (compound III). Subsequent reactions are catalyzed by extradiol dioxygenase (DfdB) and hydrolase (DfdC) and produce salicylate (compound V) and 2-hydroxypenta-2,4-dienoate (compound VI). Previously, we reported cloning of two gene clusters involved in DF degradation in two DF-utilizing actinomycetes, Terrabacter sp. strain YK3 and Rhodococcus sp. strain YK2. A 5.2-kb PstI fragment of the large circular plasmid pYK3 in strain YK3 encodes an angular dioxygenase for DF (dfdA1 to dfdA4) (Fig. 1B) (20), and a 3.4-kb Sau3AI fragment of YK2 containing the dfdBC genes catalyzes the biotransformation of 2,2',3-trihydroxybiphenyl to salicylate (21) (clone pDMR5) (Fig. 1C, right panel). Southern hybridization analysis indicated that the dfd genes are well conserved among several DF-utilizing actinomycetes belonging to different genera (20, 21). In addition, genes that closely resemble dfdA have recently been identified in two DF-utilizing actinomycetes, Nocardioides sp. strain DF412 (33) and Rhodococcus sp. strain HA01 (3). These results suggest that DF-utilizing actinomycetes recently acquired dfd gene clusters by lateral gene transfer.

View larger version (23K): [in this window] [in a new window]

FIG. 1. Initial catabolic pathway for DF and organization of the dfd gene cluster. (A) The enzymes responsible for utilization of DF in actinomycetes are indicated above the conversion steps. The following compounds are involved in this process: DF (compound I), 4,4a-dihydroxy-4-hydrodibenzofuran (compound II), 2,2',3-trihydroxybiphenyl (compound III), 2-hydroxy-6-oxo-6-(2-hydroxyphenyl)hexa-2,4-dienoate (compound IV), salicylate (compound V), and 2-hydroxypenta-2,4-dienoate (compound VI). 4,4a-Dihydroxy-4-hydrodibenzofuran is unstable and spontaneously transforms into 2,2',3-trihydroxybiphenyl. (B and C) dfd gene cluster organization in Terrabacter sp. strain YK3 (B) and Rhodococcus sp. strain YK2 (C). Dashed lines indicate the DNA regions that were cloned previously (gray dashed lines) and in this study (black dashed lines). The sizes and directions of the ORFs are indicated by arrows. Genes that encode proteins whose functions are unknown, genes involved in DF catabolism, and the regulator genes that were analyzed in this study are indicated by open, gray, and black arrows, respectively. The positions of ClaI (C), EcoRI (E), PstI (P), and HpaI (H) restriction enzyme recognition sites are indicated. The position of a portion of the Sau3AI (S) site is indicated in panel C in parentheses. The DNA regions whose nucleotide sequences were not determined in this study (downstream of the HpaI site) are indicated by gray lines.

Genes involved in the angular dioxygenation of DF have been found in strains RW1 (dxnA1A2, fdx1, and redA2) (5-8, 11), DBF63 (dbfA1A2) (25), YK3 (dfdA1 to dfdA4) (20), and YK5 (dbfA1A2_YK5) (22). The presence of a transcriptional regulation system for genes derived from DBF63, YK3, and YK5 has been demonstrated. Transcription of the dfdA1 and dbfA1_YK5 genes was induced in strains YK3 and YK5, respectively, by cultivation with DF (20, 22). In contrast, transcription of the dbfA1_DBF63 gene was induced not only by DF but also by fluorene (36), and it was subsequently demonstrated that the dbfA1A2 genes are linked to the fluorene-phthalate catabolic gene cluster (16). The transcriptional regulation mechanism for genes involved in the degradation of DF has not been determined yet; so far, we have no molecular information on the transcriptional regulation system for DF catabolic pathways in bacteria.

Several regulators of the catabolic pathway for aromatic compounds with chemical structures similar to that of DF have been studied. In gram-negative bacteria, two regulators of the carbazole degradation pathway have been described. Transcription of the carbazole and anthranilate degradation gene clusters in Pseudomonas resinovorans CA10 was controlled by the AraC/XylS family protein CarR_CA10 and anthranilate, a metabolite of carbazole (51). A different transcriptional regulator, CarR_J3, which is involved in transcriptional control of the car gene cluster, has been identified in Janthinobacterium sp. strain J3 (32). This regulator is a member of the GntR family, and a metabolite of carbazole, 2-hydroxy-6-oxo-6-(2'-aminophenyl) hexa-2,4-dienoate, acts as the effector. Several metabolic pathway genes responsible for biphenyl degradation have been described. In gram-negative bacteria, transcriptional regulators belonging to the GntR and LysR families were identified as transcriptional control factors for the bph gene cluster in Pseudomonas sp. strain KKS102 (37) and Pseudomonas pseudoalcaligenes KF707 (53, 54), and a metabolic intermediate of biphenyl, 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoic acid, acted as an effector. The transcriptional regulation system in high-G+C-content gram-positive Rhodococcus strains has a different regulatory mechanism for bph gene expression. Biphenyl-utilizing Rhodococcus strains use the two-component signal transduction (TCST) system (4, 26, 48), and the primary substrate biphenyl acts as an effector for transcriptional activation of the bph gene cluster in Rhodococcus sp. strain RHA1 (48).

Molecular information concerning the pathway regulators involved in the degradation of hydrophobic xenobiotic compounds should increase our understanding of the mechanisms that recognize molecules that exhibit relatively high degrees of hydrophobicity and that are occasionally cytotoxic to bacteria. In the present paper, we describe cloning and functional analysis of a transcriptional regulator termed DfdR, which is encoded by a gene in the dfd gene cluster in the DF-utilizing high-G+C-content gram-positive actinomycetes Rhodococcus sp. strain YK2 and Terrabacter sp. strain YK3. The dfdR gene product affects the promoter activity of the dfdA genes, which is involved in the initial hydroxylation of DF. We describe the peculiarities of effector-compound specificities and the protein domain structure of DfdR as a transcriptional regulator for aromatic compound metabolic pathway genes.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bacterial strains and culture conditions. Rhodococcus sp. strain YK2 and Terrabacter sp. strain YK3 are capable of using DF as a sole source of carbon and energy (21). Rhodococcus sp. strain RD2 is a mutant of Rhodococcus sp. strain YK2 that spontaneously lost the ability to utilize DF during continuous cultivation in rich media (21). This strain lost the dfd gene cluster, including dfdA1, dfdB, and dfdR. R. erythropolis JCM 2892 was used as a host strain for measurement of promoter activity. Rhodococcus and Terrabacter strains were cultured at 30°C on Luria-Bertani (LB) medium supplemented with 10 mM glucose (LBG). Escherichia coli strains were cultured at 37°C on LB medium. MM3Y supplemented with an appropriate carbon source was used as the minimal medium (21). Solid media contained 1.5% to 2.5% agar powder or purified agar powder. The following carbon sources were used in this study: 10 mM glucose and succinate; 1% glycerol; 3 mM sodium salicylate, sodium gentisate, 2,2',3-trihydroxybiphenyl, and 2-hydroxydibenzofuran; and 0.1% DF, dibenzo-p-dioxin, xanthene, carbazole, dibenzothiophene, fluorene, biphenyl, naphthalene, and phenanthrene. Antibiotics were added to media for selection of transformants at the following concentrations: ampicillin, 200 µg·ml^–1; kanamycin, 25 µg·ml^–1; and chloramphenicol, 25 µg·ml^–1. All the chemicals used in this study were the highest purity and were obtained from Sigma-Aldrich Chemicals (St. Louis, MO), Nacalai Tesque (Kyoto, Japan), Kanto Chemical (Tokyo, Japan), and Wako Pure Chemicals (Osaka, Japan).

Manipulation of DNA and RNA. The procedures used for isolation of genomic DNA, plasmid DNA, and total RNA of bacterial strains and the procedures used for Southern and Northern hybridization analyses have been described previously (20-22). PCR amplification was carried out using ExTaq DNA polymerase (TAKARA BIO, Shiga, Japan) according to the manufacturer's protocol. A PCR-amplified DNA fragment was initially cloned into pGEM-T (Promega, Madison, WI) or other suitable vector plasmids and used after nucleotide sequence analysis to select clones with no PCR errors. For high-fidelity PCR cloning, we used the KOD plus DNA polymerase (TOYOBO, Osaka, Japan) and confirmed that the nucleotide sequences of at least two individual PCR clones matched completely. The PCR primers used in this study are shown in Table S1 in the supplemental material.

Induction cultivation and RT-PCR analysis. Cells grown in LBG were collected at an optical density of 600 nm (OD₆₀₀) of 1.0, washed once with saline-phosphate buffer (pH 7.0), and resuspended in MM3Y to obtain an OD₆₀₀ of 5.0. The cell suspension was distributed in test tubes, and carbon sources (10 mM glucose and 0.1% DF) were added. After 6 h of incubation at 30°C and rotary shaking at 180 rpm, the total RNA was extracted as described above.

Reverse transcription (RT) was performed using 1 µg of total RNA treated with RNase-free DNase (TOYOBO), random nonamer primers, and ReverTra Ace (TOYOBO) at 50°C for 30 min according to the manufacturer's instructions. PCR amplification was performed with primers specific for dfdA1, dfdR, and 16S rRNA (see Table S1 in the supplemental material), a diluted RT product (50x dilution for YK2 samples, 50x dilution for YK3 samples for dfdA1 and 16S rRNA, and 20x dilution for dfdR), and ExTaq DNA polymerase by using the following parameters: 94°C for 5 min; 25 cycles of 94°C for 20 s, 58°C for 30 s, and 72°C for 60 s; and a final incubation at 72°C for 5 min. For 16S rRNA sequence amplification, the annealing temperature and number of cycles were 48°C and 18, respectively. DNase-treated RNA samples that were not subjected to RT-PCR were used as PCR templates to verify the absence of contaminating genomic DNA. We analyzed 10-µl PCR samples on 1.7% agarose gels.

Gene disruption of Rhodococcus sp. strain YK2. For dfdR (orf4_YK2) gene disruption of Rhodococcus sp. strain YK2, a chloramphenicol resistance gene was amplified by PCR from pRF30 (13) (a kind gift from Genetello, Ghent University, Ghent, Belgium) to introduce NcoI sites at both ends of the gene and then cloned into the pGEM-T vector. After nucleotide sequence confirmation, the NcoI fragment was introduced into the AflIII site of pBluescript II KS(+) (Stratagene, La Jolla, CA), and the resultant plasmid was designated pBS-RfCm. An internal region of dfdR (680 bp) that was amplified by PCR was introduced into the EcoRV site of pBS-RfCm and then used for transformation of Rhodococcus sp. strain YK2. The transformation protocol used for Rhodococcus species has been described previously (21). Total DNA of the YK2 transformants selected on LBG with 25 µg·ml^–1 chloramphenicol in agarose plates was used to confirm gene disruption.

Construction of a plasmid for the promoter activity assay. The promoter probe shuttle vector pRKlux was constructed by modifying pKLA1 (48). Bacterial luciferase genes (luxAB) were amplified by PCR from pKLA1, which was introduced into StuI, KpnI, and PmaCI sites at the 5' end and into XbaI and StuI sites at the 3' end, and then cloned into the pGEM-T vector. After confirmation of the nucleotide sequence, the 2.1-kb StuI fragment was recovered and ligated with a 4.0-kb PvuII fragment of the Rhodococcus-E. coli shuttle vector pRK401 without the multiple cloning sites and the lacZ- gene of the original shuttle vector to obtain the pRKlux vector plasmid. The PCR-amplified promoter region of the dfdA1 gene, which was introduced into KpnI and SfoI sites at the 5' and 3' ends, respectively, was cloned into pGEM-T and confirmed by nucleotide sequencing. Promoter DNA fragments that were recovered by KpnI and SfoI treatment were introduced into the KpnI and PmaCI sites of the pRKlux vector.

Bacterial luciferase assay. Rhodococcus strains introducing pRKlux derivatives were cultured in LBG with kanamycin until the OD₆₀₀ was approximately 0.6. Cells were collected, washed, and resuspended in 0.2x LB medium (0.2% Bacto tryptone, 0.1% Bacto yeast extract, 1% NaCl) with carbon sources and then incubated for 6 h at 30°C. The washed induced cells were resuspended in Lux buffer (50 mM sodium phosphate buffer, pH 7.0), and 0.1 ml of the suspension was placed into a White Microwell SI plate (Nalge Nunc International, Rochester, NY). Luciferase activity was measured with a Wallac 1420 ARVO SX (Wallac-Perkin Elmer, Wellesley, MA) at 25°C by adding 5 µl of 0.02% (vol/vol) decanal (Sigma-Aldrich) in ethanol as a substrate. The light output in 1 s from the reaction was collected, and the number of light units was calculated by determining the light output per OD₆₀₀ unit.

Primer extension analysis. Strains YK2 and YK3 were cultured in MM3Y with glucose or with DF plus glucose, and 20 µg of the total RNA was isolated and treated with DNase I (Stratagene). After phenol-chloroform treatment, the DNA-free total RNA was recovered by ethanol precipitation. RT was performed using ReverTra Ace (TOYOBO), RNase inhibitor (TOYOBO), a deoxynucleoside triphosphate mixture (dATP, dTTP, dCTP, and 7-deaza-dGTP), and Texas Red-labeled RTdfdA1 primer (see Table S1 in the supplemental material) according to the manufacturer's protocol. Nucleotide sequencing of P_dfdA1 was performed with a Thermo-Sequenase cycle sequencing kit (GE Healthcare Bio-Sciences, Piscataway, NJ) using the RTdfdA1 primer. The reaction products were analyzed using a DNA sequencer (SQ-5500; Hitachi, Tokyo, Japan).

Heterologous expression of DfdR and EMSA. The 129-bp constitutive promoter P_dfdB derived from the sequence upstream of dfdB in Rhodococcus sp. strain YK2 (21) was amplified using PCR with addition of the SfoI site downstream of the start codon of dfdB. A PCR-amplified dfdR gene fragment that included the PmaCI site at the 5' end was ligated into the SfoI site downstream of P_dfdB on the shuttle vector pRK401 and then used for transformation of Rhodococcus species. Transformants of the dfdR expression constructs cultured in LBG with kanamycin were disrupted by sonication in phosphate-buffered saline containing 10% glycerol and 0.5 mM dithiothreitol with or without 0.1% aromatic substrates by using an ultrasonic processor (UP50H; Hielscher, Teltow, Germany). Supernatants obtained after centrifugation at 15,000 rpm for 10 min at 4°C were used for an electrophoretic mobility shift assay (EMSA). To activate DfdR in cell extracts, 50x concentrations of aromatic compound stock solutions (DF, dibenzo-p-dioxin, biphenyl, and naphthalene) dissolved in dimethylformamide were added to the 10-µg·µl^–1 cell extracts, incubated at 4°C for 30 min, and then used for EMSA. The same amount of dimethylformamide was added to the cell extract for the negative control.

The EMSA was performed by using a DIG-Gel shift kit (Roche, Mannheim, Germany) with digoxigenin (DIG)-labeled P_dfdA1 and the cell extract, using the method described by the manufacturer. The binding buffer for DfdR contained 100 mM NaCl, 4 mM Tris-HCl (pH 7.5), 0.2 mM dithiothreitol, 10% glycerol, 0.05% NP-40, 0.05 µg·µl^–1 poly(dI-C), and 0.02 ng·µl^–1 DIG-labeled probe. The concentration of cold probe DNA used for the competition assay was 1 ng·µl^–1 (50 times the amount of the probe). Shift band intensities were quantified by using the Image Gauge software (version 3.1; Fujifilm, Tokyo, Japan).

Nucleotide sequence accession numbers. The nucleotide sequences of the dfd gene clusters have been deposited in the DNA Data Bank of Japan (DDBJ). The accession numbers for the dfdA1A2A3A4 genes of YK2, the dfdBCR genes of YK2, and the dfdA1A2A3A4BCR genes of YK3 are AB302135, AB070453, and AB075242, respectively.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
New DNA-binding protein gene detected in the nucleotide sequence of the dfd gene cluster of Terrabacter sp. strain YK3. In order to identify the regulator gene for dfdA in Terrabacter sp. strain YK3, we determined the nucleotide sequence of the dfd gene cluster present in plasmid pYK3. In this study, we performed a Southern hybridization analysis of pYK3 by using dfdA1 from YK3 and dfdB from YK2 as probes, and then we cloned two large fragments of pYK3, a 14.9-kb EcoRI fragment containing the dfdA1 and dfdB genes and a 11.6-kb ClaI fragment containing the dfdB gene. We analyzed the entire nucleotide sequence of the 14,869-bp EcoRI fragment and the 4,621-bp dfdBC coding region of the 11.6-kb ClaI fragment (5' end of the ClaI-to-HpaI site) (Fig. 1B). The results revealed an approximately 2.5-kb overlap between the two sequences. The two gene clusters containing dfdA1 to dfdA4 or dfdBC are located in opposite directions and are separated by a 2.9-kb DNA sequence that contains transposase-like genes (orf1 and orf10). The organization of the 16,983-bp dfd gene cluster of YK3 is shown in Fig. 1B; the characteristics of the predicted open reading frame (ORF) products and the homologous proteins as determined by BLASTP analysis are shown in Table S2 in the supplemental material.

The gene organization downstream of the dfdA4 gene suggests that the gene cluster involved in DF utilization ends at dfdA4. This region contains four ORFs, orf6 to orf9, and the products of these genes exhibit homology to a transposase, a putative acyltransferase, a regulatory protein, and a putative hydroxylase, respectively.

We identified the dfdBC genes in the nucleotide sequence opposite the dfdA genes, which completely matched the corresponding region of YK2. In the region downstream of the dfdBC genes, we identified an ORF, orf13. orf13 encodes a 402-amino-acid protein with an estimated molecular mass of 44 kDa, which exhibited a homology to several kinds of LuxR family transcriptional regulators (Fig. 2). However, the homology was restricted to an approximately 60-amino-acid region in the C-terminal part of Orf13. An amino acid sequence analysis using Pfam (14) demonstrated that the C-terminal region (amino acids 343 to 400) of Orf13 is similar to a DNA-binding domain present in transcription regulators of the LuxR/FixJ family of response regulators (GerE; PF00196) with a high level of confidence (E value, 8.1e-20) (Fig. 2). In addition, the Pfam analysis predicted a GAF domain in the central part of Orf13 (PF01590; 135- to 277-amino-acid region of DfdR) (Fig. 2) with a significant E value (4.2e-06). These results suggest that the orf13 product is a novel DNA-binding protein that has a LuxR-type helix-turn-helix DNA-binding motif and a GAF-like domain that is probably involved in small-molecule binding and regulation. We designated this gene dfdR on the basis of the results of this study.

View larger version (96K): [in this window] [in a new window]

FIG. 2. Amino acid sequence alignment of DfdR (Orf13/Orf4) and homologous gene products. The amino acid sequence alignment was generated by using ClustalX, version 1.81. Conserved and homologous amino acids in the alignment are indicated by asterisks and colons or periods, respectively, below the alignment. The amino acid regions indicated by a black background and by bold type are a putative GAF domain and a DNA-binding domain of DfdR, respectively. Conserved amino acid regions in the alignment are indicated by shading. The nucleotide sequence accession numbers and strains used are as follows: DfdR/Orf13YK3, accession number AB075242 and Terrabacter sp. strain YK3; DfdR/Orf4YK2, accession number AB070453 and Rhodococcus sp. strain YK2; Mvan_0348, Mvan_0349, Mvan_3973, and Mvan_3991, accession number NC_008726 and M. vanbaalenii PYR-1; MSMEG_0330 and MSMEG_0331, accession number NC_008596 and M. smegmatis MC2155; MUL_2876, accession number NC_008611.1 and M. ulcerans Agy99; FRAAL1658 and FRAAL1659, accession number NC_008278 and F. alni ACN14a; and orf25, accession number DQ306260 and Rhodococcus sp. strain T104.

During the course of this study, we discovered genes with overall similarity to dfdR in the sequence databases, while the levels of amino acid sequence identity were 34% or lower (Fig. 2). Of the 10 sequences examined, 9 were obtained in genome sequence analyses of actinomycetes, the pyrene-utilizing strain Mycobacterium vanbaalenii PYR-1 (accession number NC_008726; Mvan_0348, Mvan_0349, Mvan_3973, and Mvan_3991), the pathogenic strain Mycobacterium ulcerans Agy99 (accession number NC_008611; MUL_2876), the nonpathogenic strain Mycobacterium smegmatis MC² 155 (accession number NC_008596; MSMEG_0330 and MSMEG_0331), and the nitrogen-fixing bacterium Frankia alni ACN14a (accession number NC_008278; FRAAL1658 and FRAAL1659). The remaining sequence was orf25 of the terpene-degrading organism Rhodococcus sp. strain T104 (accession number DQ306260) (10). We found a GAF-like domain in the Mvan_0348, Mvan_3973, Mvan_3991, MSMEG_0331, and Orf25 sequences by performing a Pfam analysis, and the levels of confidence were significant (E values, 9.6e-06, 3.6e-04, 2.9e-09, 3.6e-06, and 1.7e-04, respectively). The following conserved amino acid sequence in actiomycetous proteins homologous to DfdR was found in the C-terminal part of the GAF domain region: Y-V-(A/V)-A-P-(V/I/L)-X₅-(V/M)-(I/V/L/M)-G-(F/L)-(L/I)-H-(A/T)-(D/S) (Fig. 2).

The sequence that was most similar to the DfdR sequence (34% identity) was that of Mvan_3991 of M. vanbaalenii PYR-1, and the gene localized with genes whose functions have not been identified yet. M. vanbaalenii PYR-1 harbors four copies of genes that are homologous to dfdR. Mvan_3973 (whose product shows 27% identity to DfdR) is located 16.8 kb upstream of the Mvan_3991 gene in the reverse direction along with the putative cytochrome P450 gene (Mvan_3974). A protein gene that is highly similar to Mvan_3973 (82% identity) and whose product showed 27% identity to DfdR was detected in M. ulcerans Agy99 (MUL_2876) along with a pseudo-P450 gene (MUL_2875). The two remaining genes in PYR-1 that are homologous to dfdR (Mvan_0348 and Mvan_0349) are tandemly arranged along with a gene cluster whose functions have not been determined yet (Mvan_0347 to Mvan_0339). Mvan_0347 in this gene cluster encodes a protein that is similar to 2-nitropropane dioxygenase. Mvan_0348 and Mvan_0349 show 27 and 22% identity to DfdR, respectively. Genes homologous to these tandem regulator genes were detected in the databases for M. smegmatis (MSMEG_0330 and MSMEG_0331) and F. alni (FRAAL1658 and FRAAL1659); the products of these genes show 21 to 27% identity to DfdR. As observed in the case of M. vanbaalenii, the tandemly arranged regulator genes in M. smegmatis and F. alni that are homologous to dfdR are localized with genes whose functions have not been elucidated yet. The gene product of orf25 shows 20% identity to DfdR, and the gene is located near the cmt gene cluster, which is involved in p-cumate catabolism in the terpene-degrading organism Rhodococcus sp. strain T104 (10). The orf25 gene is located immediately downstream of cmtC, which encodes 2,3-dihydroxy-p-cumate dioxygenase (nucleotide sequence accession number ABC33901). In the cmt cluster, three copies of IclR-type regulator genes (cmtR1 to cmtR3) have also been discovered (10), and the functional relationship between orf25 and the regulation of the upstream cmt cluster has not been determined. Other than the cmt cluster in Rhodococcus sp. strain T104, we could not find any genes that are involved in the catabolism of aromatic compounds in the region surrounding the genes homologous to dfdR.

Nucleotide sequence of Rhodococcus sp. strain YK2 dfd gene clusters. To analyze the molecular information for dfdA transcriptional regulation, the availability of a transformation system is important. The DF-utilizing organism Rhodococcus sp. strain YK2 could be transformed by the vectors for Rhodococcus species. Thus, we decided to analyze the nucleotide sequence of the Rhodococcus sp. strain YK2 dfd gene cluster (Fig. 1C; see Table S2 in the supplemental material).

A 5.2-kb PstI fragment of YK2 containing the dfdA gene cluster was cloned from a genomic DNA library of YK2 based on homology to dfdA1 of YK3 (Fig. 1C, left panel). The length of the PstI fragment is identical to the length of the YK3 sequence (5,211 bp), and the level of nucleotide sequence identity is 99.8% (5,202/5,211 nucleotides). For the dfdA genes, eight of the nine nucleotide substitutions in the PstI fragment were present in the DfdA1 coding region. Consequently, seven amino acids were different in the DfdA1 proteins (98.5% identity) of the YK2 and YK3 strains. The remaining nucleotide substitution was in the noncoding region between dfdA3 and dfdA4. Inducible transcription of dfdA1 in YK2 occurred in cultures in the presence of DF (Fig. 3B, lanes YK2). No nucleotide substitutions have been found in the 5' end 778 bp upstream of dfdA1, which might be involved in the transcriptional regulation of dfdA. These results indicate that the regulation machinery for dfdA might be similar or identical in the two strains.

View larger version (33K): [in this window] [in a new window]

FIG. 3. Characterization of orf4 disruptants in Rhodococcus sp. strain YK2. (A) Growth of YK2 and derivatives of this strain in MM3Y supplemented with DF. The cell densities in cultures of YK2, RD2 (non-DF-utilizing mutant of YK2), YK2-dR transformed with vector plasmid pRK401 (dfdR disruptant), and YK2-dR transformed with pRK401-dfdR (dfdR transformant of YK2-dR) were measured by determining the OD600. (B) Northern hybridization analysis using the dfdA1 coding region as a probe. We isolated the total RNA of YK2, YK2-dR, and YK2-dR transformed with pRK401-dfdR cultured in a medium with glucose alone (lanes G) or with DF plus glucose (lanes GD) for 6 h, and we loaded 4 µg in each lane. The results of ethidium bromide-stained denatured agarose gel electrophoresis are shown below the hybridization results. The arrowheads indicate the positions of 16S and 23S rRNAs.

Previously, a DNA fragment of YK2 containing the dfdBC genes was cloned (Fig. 1C, right panel) (clone pDMR5 [21]). In this study a DNA fragment downstream of dfdC was isolated from YK2 genomic DNA by high-fidelity PCR. The nucleotide sequence of the resultant 2.3-kb amplified fragment corresponding to the HpaI site was analyzed (Fig. 1C, right panel), and an ORF (orf4_YK2) identical to the dfdR ORF in YK3 was detected; hence, we designated this ORF dfdR. The nucleotide sequence of the dfdBC-dfdR region (4,622 bp) exhibited 100% identity to the nucleotide sequence of YK3.

Transcription of the putative regulator gene. Transcription of dfdR (orf4_YK2 and orf13_YK3) was determined by semiquantitative RT-PCR analysis. Figure 4 shows that apparent transcription of dfdA1 was detected in DF-induced RNA samples from both strains; this is consistent with the results of the Northern hybridization analysis (20). In the PCR samples, using primers specific for dfdR, we detected almost the same level of weak transcription signals in culture samples grown with and without DF. The amounts of amplified 16S rRNA products for the culture samples grown with and without DF could not be distinguished. These results suggest that the transcription of putative regulator genes is constitutive and probably weaker than the transcription of dfdA, although the analysis was not an accurate quantitative assay.

View larger version (36K): [in this window] [in a new window]

FIG. 4. Transcriptional expression analysis of dfdA1 and dfdR genes: agarose gel electrophoresis of RT-PCR products amplified from Rhodococcus sp. strain YK2 (upper panel) and Terrabacter sp. strain YK3 (lower panel). The total RNAs isolated from cells induced with glucose (lanes G) or with DF plus glucose (lanes GD) were used as templates for RT (lanes +). RNA samples without an RT reaction were used as negative controls (lanes –). The positions of specific PCR products of dfdA1 (489 bp), dfdR (689 bp), and 16S rRNA (408 bp) are indicated by arrowheads. The PCR conditions used for amplification of each gene are described in Materials and Methods.

DfdR is involved in transcriptional activation of dfdA1. Disruption of the dfdR gene in YK2 was performed to analyze the function of the dfdR gene product. Southern hybridization analysis of a dfdR disruptant was performed to confirm the accuracy of the disruption (data not shown). The disruptant was designated strain YK2-dR. No differences in growth were observed for the disruptant YK2-dR and the parent strain YK2 in MM3Y with glucose or glycerol (data not shown). The ability of YK2-dR to utilize DF was determined (Fig. 3A). Rhodococcus sp. strain RD2 was used as a negative control in this analysis. The growth curve for vector plasmid pRK401-transformed YK2-dR in MM3Y containing DF was almost the same as that for RD2, indicating that the YK2-dR strain could not utilize DF. A pRK401-derived plasmid carrying dfdR and the 5' upstream 107-bp region (that is, the internal region of dfdC and dfdR) was constructed. We identified a promoter-like element in the 107-bp region (TTCGGA-N₁₇-TGGCCT) upstream of the putative Shine-Delgarno sequence (AAGGAG) of dfdR. The optical density of the YK2-dR transformant with the plasmid containing dfdR increased during culture and was comparable to that of the wild-type YK2 strain.

The transcriptional induction of dfdA1 in YK2-dR was investigated. Figure 3B shows the results of a Northern hybridization analysis performed by using the dfdA1 coding region as a probe. These results demonstrate that the YK2-dR strain did not have dfdA1 transcriptional induction ability in the presence of glucose and DF (Fig. 3B, lanes YK2-dR) and that the strain with dfdR restored exhibited transcriptional induction ability (lanes YK2-dR+dfdR). These findings indicate that the dfdR gene product involved in DF utilization is probably important for the transcriptional activation of dfdA.

DfdR is the transcriptional activator for the dfdA1 promoter, while DF is an inducer. To analyze the activity of the dfdA1 promoter in Rhodococcus strains, we prepared a fusion construct comprising the region upstream of the dfdA1 start codon and the luxAB gene of Vibrio harveyi introduced into the pRKlux vector (designated pRKluxP_A) and used this construct to transform Rhodococcus sp. strain YK2. In preliminary experiments, we obtained similar results for two fusion constructs that differed in promoter length (data not shown); one construct contained the region upstream of the start codon up to 778 bp (to the 5' end of the PstI fragment of YK2) (Fig. 1C, left panel), and the other contained the region upstream of the start codon up to 481 bp. We decided to use the 481-bp region as the promoter for dfdA1 (P_dfdA1).

Figure 5A shows the luciferase activity of the pRKluxP_A transformant after cultivation in the presence and in the absence of DF. The cells induced with DF exhibited strong promoter activity that was approximately 86 times higher than that of the cultures grown in the absence of DF (Fig. 5A). Since host strain YK2 can utilize DF, the substrate was catabolized via the dfd catabolic pathway during cultivation in the presence of an inducer. The three metabolic intermediates in the DF utilization pathway (2,2',3-trihydroxybiphenyl, salicylate, and gentisate) were also tested, and no P_dfdA1 activation was observed. These results indicate that the P_dfdA1 regulator responds to DF and not to its catabolites.

View larger version (28K): [in this window] [in a new window]

FIG. 5. Promoter activity of PdfdA1 in Rhodococcus strains. (A) Promoter activity of PdfdA1 in Rhodococcus sp. strain YK2 determined by using the PdfdA1::luxAB transcriptional reporter fusion construct. Transformants of YK2 with the the pRKlux promoter probe vector inserted into the 481-bp dfdA1 promoter region (pRKluxPA) were induced by culturing them for 6 h in 0.2x LB medium with several carbon sources prior to use in the luciferase activity assay. The carbon sources added to the assay media are indicated on the left. The bars and error bars indicate the means and standard deviations obtained from three determinations. The numbers of relative light units (LU) were calculated by determining the number of light counts per second per OD600 unit. (B) Promoter activity of PdfdA1 in (panel I) Rhodococcus sp. strain RD2 and (panel II) R. erythropolis JCM 2892 determined by using PdfdA1::luxAB transcriptional reporter fusion constructs. Transformants with pRKluxPA, pRKluxPA+dfdR containing dfdR with 5' upstream 107 bp, or pRKluxPA+dfdRo/p containing dfdR under controls with a constitutive overproducing promoter (PdfdB) were cultured in 0.2x LB medium for 6 h in the absence (striped bars) or presence (gray bars) of 0.1% DF, which was subsequently used for measurement of luciferase activity. The numbers of relative light units were calculated by determining the number of light counts per second per OD600 unit. The bars and error bars indicate the means and standard deviations obtained from three determinations.

Rhodococcus sp. strain YK2 utilizes glucose, glycerol, and succinate as preferred carbon and energy sources. This study investigated whether these carbon sources affect the DF-dependent promoter activation of P_dfdA1 (Fig. 5A). The promoter activities of the strains cultured with these compounds and with DF were almost comparable to those of DF cultures. This suggests that the carbon sources used in this study do not exert carbon catabolite repression on P_dfdA1 activation via DF.

Next, we analyzed the relationships between DF-dependent P_dfdA1 activation and the dfdR gene product by using Rhodococcus sp. strain RD2 and R. erythropolis JCM 2892 as host strains; these strains are unable to transform DF. We used two kinds of dfdR gene cassettes for expression of DfdR: (i) dfdR with its own promoter (107-bp region present in dfdC and dfdR) and (ii) the coding region of dfdR inserted downstream of the heterologous promoter P_dfdB, which is the 5' upstream 129-bp region of dfdB that is responsible for constitutive transcription in the genus Rhodococcus. Previously, we reported that the transcription of dfdB is constitutive in strain YK2 (21), and the P_dfdB promoter activity measured using the luciferase reporter assay system in LBG-grown strain RD2 was 2.1 x 10⁶ counts·s^–1·OD₆₀₀ unit^–1 (standard deviation, 1.4 x 10⁵ counts·s^–1·OD₆₀₀ unit^–1). The cassettes were introduced into the XbaI site of pRKluxP_A, and the resulting constructs were designated pRKluxP_A+dfdR and pRKluxP_A+dfdRo/p, respectively, and then were used for transformation of strains RD2 and JCM 2892. The promoter activities of the recombinants with dfdR introduced are shown in Fig. 5B. A comparison of the results for the two strains with pRKluxP_A and pRKluxP_A+dfdR revealed that P_dfdA1 promoter activities were detected only in the presence of dfdR and DF, indicating that P_dfdA1 activation via DF involves the gene product of dfdR. The promoter activity levels of pRKluxP_A+dfdRo/p-containing strains induced with DF were comparable to the promoter activity levels of the pRKluxP_A+dfdR-containing strains. However, DfdR overproduction occurred after P_dfdA1 activation independent of DF and was approximately 37- and 49-fold higher than the overproduction by the non-DfdR-overproducing transformants (pRKluxP_A-dfdR) of RD2 (Fig. 5B, panel I) and JCM 2892 (Fig. 5B, panel II), respectively, under induction conditions without DF. These results suggest that DfdR overproduction affects the DF dependence of DfdR during transcriptional activation of P_dfdA1.

DfdR binds to P_dfdA1 in a DF-dependent fashion. Prior to assaying the interaction between DfdR and P_dfdA1, we determined the transcription start site of dfdA1 by primer extension analysis. A single RT product appeared only in RNA samples of both the YK2 and YK3 strains grown on DF-supplemented media (Fig. 6A, lanes 1 and 3). The signal corresponded to the adenine residue 172 bp upstream of the start codon of dfdA1 (Fig. 6B). A deduced promoter sequence was found upstream of the transcription start (Fig. 6A).

View larger version (67K): [in this window] [in a new window]

FIG. 6. DfdR-PdfdA1 binding analysis. (A) Determination of the transcriptional start site of PdfdA1 in YK2 (lanes 1 and 2) and YK3 (lanes 3 and 4). Fluorescent primer extension was performed using Texas Red-labeled RTdfdA1 primer and DNase I-treated RNA (10 µg) isolated from cells induced with glucose (lanes 2 and 4) or with DF plus glucose (lanes 1 and 3). The RTdfdA1 primer is complementary to a sequence 36 bp downstream from the putative dfdA1 initiation codon. The sequencing products of PdfdA1 obtained with the RTdfdA1 primer (lanes A, C, G, and T) were simultaneously electrophoresed with the RT products. The arrows indicate the signals of RT products and the transcriptional start site. The nucleotide sequence around the transcriptional start (indicated by bold type) and a deduced promoter (indicated by underlining) are shown in the box below the gel. (B) DIG-labeled probes for PdfdA1 upstream (probe-1; positions –180 to 1) and downstream (probe-2; positions –34 to 172) sequences. The results of an EMSA using cell extract of DfdR-expressing Rhodococcus sp. strain RD2 are also shown. Lanes 1 to 7 and lanes 8 to 10 contained probe-1 and probe-2, respectively. Lanes 1 and 8 contained control samples without cell extract, lanes 2 and 9 contained cell extracts prepared without DF, and lanes 3 to 7 and 10 contained cell extracts prepared with DF. The amounts of cell extract were 2.5 and 5.0 µg in lanes 3 and 4, respectively, and 10 µg in the other lanes. Lanes 6 and 7 were supplemented with 10 ng of nonlabeled cold probe-1 and probe-2, respectively. The open and filled arrowheads indicate the positions of free and shift probes, respectively. The arrow indicates the positions of host-derived, DfdR-independent faint shift signals. +1 indicates the start codon of dfdA1. (C) EMSA using DIG-labeled probe-1 and 10 µg of cell extract of DfdR-expressing RD2 prepared without aromatic substrates. The control lanes contained control samples lacking cell extract. The other lanes contained cell extracts treated with different concentrations (0 to 2,000 µM) of aromatic compounds (DF, dibenzo-p-dioxin, biphenyl, and naphthalene) prior to EMSA, as indicated at the top. The arrowheads indicate the positions of the shift probes. (D). Comparison of signal intensities with different aromatic compounds by using EMSA. Band intensity values (LAU) obtained by using Image Gauge software (Fujifilm) minus the background values, which were the values for lanes without aromatic substrates in the blots, are plotted on the y axis. The values for DF, dibenzo-p-dioxin, biphenyl, and naphthalene for each concentration are plotted.

DfdR binding to P_dfdA1 was analyzed by the EMSA. We attempted to demonstrate heterologous expression of DfdR in the E. coli protein expression system using several systems, including the pET system (Novagene), the pCold system (TAKARA BIO), and coexpression of several chaperones with the chaperone plasmid set (TAKARA BIO); however, we could not obtain proteins suitable for analysis. Therefore, we used Rhodococcus sp. strain RD2 and P_dfdB for expression of DfdR in this study. We prepared the two P_dfdA1 probes shown in Fig. 6B. We detected faint bands with probe-2 (Fig. 6B, lanes 9 and 10); however, these weak signals also appeared with the cell extract of RD2 that did not carry dfdR (data not shown), suggesting that the probe shift was due to an unknown host-specific factor. The apparent mobility shift was observed with probe-1 (Fig. 6B, lanes 3 to 7). The shift band was never observed with cell extract of RD2 that did not carry dfdR (data not shown). The shift band intensities with probe-1 increased with the amount of protein (2.5, 5.0, and 10 µg) (Fig. 6B, lanes 3 to 5). No shift band appeared with cell extracts prepared without DF (lane 2). Excess amounts (50x) of nonlabeled probe-1 DNA resulted in a decrease in the intensity of the shift band (lane 6), but no effects were observed when the same amount of cold probe-2 DNA was used (lane 7), indicating that the interaction is specific. Figure 6C shows EMSA results obtained using probe-1 and DfdR-expressing cell extracts treated with different concentrations of DF. The shift band intensities obtained in EMSA are shown in Fig. 6D. The results revealed that the shift band intensities increased with the DF concentration and were maximal with 50 µM DF. These results indicate that DfdR directly binds to P_dfdA1 upstream of the transcriptional start site and that DF activates the binding.

DfdR activates P_dfdA1 with several aromatic compounds. The effector molecule specificities of DfdR for aromatic compounds were determined by using the reporter assay system. The R. erythropolis JCM 2892 transformant with pRKluxP_A-dfdR was cultured with several aromatic compounds that were not utilized as sole carbon and energy sources by strains JCM 2892, YK2, and YK3. The relative induction abilities of P_dfdA1 in the presence of DF and several aromatic compounds are shown in Table 1. All the aromatic substrates except 2-hydroxydibenzofuran exhibited P_dfdA1 activation that was broadly comparable to that of DF, and the activity values ranged from 60% with phenanthrene to 119% with anthracene.

View this table: [in this window] [in a new window]

TABLE 1. Promoter activities of PdfdA1 in the presence of different aromatic compounds

We investigated the activation of DNA binding of DfdR using dibenzo-p-dioxin, biphenyl, and naphthalene with EMSA (Fig. 6C and D). DNA-binding activation was observed with these three aromatic compounds, and the activities were significantly lower than the activities with DF under the in vitro binding conditions used in this study. Dibenzo-p-dioxin activated binding as well as DF, but the quantity of substrate required for activation was 20 times greater. The activity with biphenyl was approximately one-half the activity with dibenzo-p-dioxin. Very weak signals were observed with samples containing 1 and 2 mM naphthalene, while this compound activated P_dfdA1 substantially (87% of the DF activity) under the in vivo reporter assay conditions with redundant substrates (0.1%) during induction cultivation. Note that no band shift was observed with the cell extract of RD2 without the DfdR expression plasmid when the three aromatic compounds were used (data not shown), suggesting that the band shift was caused by DfdR.

On the basis of these in vivo and in vitro assay results, we suggest that DfdR has broad effector molecule specificities for P_dfdA1 activation and has a preference for DF under in vitro binding conditions.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this study, we identified a novel transcriptional regulator, DfdR, which is involved in the regulation of DF degradation genes in DF-utilizing actinomycetes. DfdR has a distinctive feature; it is a DNA-binding positive regulatory protein that responds to aromatic compounds composed of two or more aromatic rings with relatively low water solubility, such as DF, dibenzo-p-dioxin, and biphenyl. To our knowledge, this is the first report to identify the transcriptional regulator involved in the DF catabolic pathway, and the bacterial DNA-binding transcriptional regulator can directly identify the aromatic compounds as effectors.

Several kinds of transcriptional regulators for aromatic compound catabolic pathway genes have been reported previously (for a review, see reference 49). These regulators have been divided into two groups on the basis of the effector substrate that they recognize, either (i) a catabolite of the pathway substrates or (ii) the pathway substrate itself. The former group is represented by several regulators, particularly regulators involved in pathways for compounds with two or more aromatic rings. In the case of regulators for the biphenyl, naphthalene, and carbazole catabolic pathways in gram-negative bacteria, one or more catabolites of the pathway substrates function as the effectors; these effectors are 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoate (37), salicylate (18), and anthranilate (51) and 2-hydroxy-6-oxo-6-(2'-aminophenyl)hexa-2,4-dienoate (32), respectively. The regulation system for the first group is better than that for the second group with regard to the water solubilities of the signaling molecules. Hydrophobic chemicals are highly insoluble in water once they are captured by hydrophobic cell materials such as cell membranes; therefore, they are unsuitable as signal messengers for cytosolically localized signal accepters. Catabolic intermediates formed by oxidative degradation pathways generally increase the water solubility; they may behave efficiently as signal messengers by increasing both the intracellular and extracellular concentrations. Therefore, it is suggested that hydrophilic metabolites might function better as signal transduction molecules than as hydrophobic pathway substrates.

The second group includes the XylR/DmpR subclass of the NtrC family regulators found in gram-negative bacteria, and XylR (39) and DmpR (46, 47) have been extensively studied. Generally, regulators of this type identify pathway substrates as effectors and require ATP as a cofactor for transcriptional activation, as well as substrates involved in catabolic pathways for monocyclic aromatic compounds such as toluene, xylene, and phenol. The only exception is HbpR, which was identified by using a 2-hydroxybiphenyl catabolic pathway regulator in Pseudomonas azelaica HBP1 (23). The effector is the pathway substrate, and it also recognizes 2,2'-dihydroxybiphenyl and 2-aminobiphenyl but not nonhydroxylated biphenyl. We emphasize that the aromatic effector substrates of this family of regulators are hydrophobic but dissolve relatively well in water compared to compounds possessing two or more aromatic rings without a hydroxyl group. The solubility values for chemicals obtained from the ChemIDplus database (National Library of Medicine) at a water temperature of 25°C (unless indicated otherwise) are as follows: toluene, 5.71 mM; m-xylene, 1.52 mM; styrene, 2.98 mM; phenol, 87.9 mM; salicylic acid, 12.4 mM; anthranilic acid, 25.5 mM (solubility at 20°C); and 2-hydroxybiphenyl, 4.11 mM. These values are significantly higher than those for the pathway substrates composed of two or more aromatic rings, such as DF (18.4 µM), dibenzo-p-dioxin (4.89 µM), biphenyl (45.0 µM), naphthalene (242 µM), and carbazole (10.8 µM).

Another example of the second group is a member of the TCST system family. Transcriptional regulation of metabolic pathways for toluene (tod) and styrene (sty) is involved in the TodST and StySR TCST systems, respectively (27, 38), and the ligand for the TCST system is the primary substrate of the target pathway. The biphenyl catabolic gene regulator in high-G+C-content gram-positive Rhodococcus species is a TCST system, and biphenyl, which contains two aromatic rings, is an effector (4, 26, 48). Genes for a member of the TCST system were detected downstream of the biphenyl catabolic pathway genes of three Rhodococcus strains and encoded a particularly large histidine kinase component (BphS or BpdS) and a response regulator (RR) component (BphT or BpdT). Although the domain architecture and topology of the sensor kinase have not been determined, the large kinase might be anchored to the cell membrane and detect biphenyl molecules in the vicinity to activate the cognate RR via phosphotransfer.

Similar to the regulators of the second group, the DfdR protein also directly identifies the primary pathway substrate DF. However, this suggests that (i) DfdR is not an integral membrane protein like BphS since no hydrophobic transmembrane segment-like sequence could be detected in the DfdR sequence and (ii) DfdR does not contain kinase or ATP-binding domains; therefore, DfdR probably activates transcription without energy consumption, unlike the transcriptional regulators belonging to the second group described above. Therefore, the DfdR protein is a different kind of transcriptional regulator for aromatic compound catabolic pathways than the regulators described previously. We strongly suggest that the DfdR protein identifies hydrophobic aromatic compounds within the cell via the GAF-like domain in the central part of DfdR and then activates transcription by directly binding to the promoter region of dfdA1. The DF-dependent in vitro DNA-binding ability of DfdR supports this suggestion.

Recent bacterial genome sequence analysis revealed diverse genes that may be involved in environmental signal sensing and response. Some of the protein families exhibiting such a function are one-component systems, and they have been compared with well-known two-component systems (31, 50). The protein architecture of a one-component system is simpler than that of a TCST system; a single protein has an input signal-sensing domain such as GAF or Per/Arnt/Sim (PAS) and an output domain such as a DNA-binding domain, and it lacks histidine kinase and receiver domains that are present in the TCST system proteins. Based on its domain architecture, the transcriptional regulator DfdR is categorized as a one-component system. We detected only 10 sequences that permitted detection of the overall similarity to DfdR when a protein-protein BLAST search was used. However, several sequences with domain architecture similar to that of DfdR have been discovered, including sequences having domains for GAF and helix-turn-helix DNA binding in a single protein. We detected at least 51 sequences with similar domain architectures by using SMART analysis (28). All the structurally similar sequences were sequences from bacteria belonging to the classes Actinobacteria, Cyanobacteria, Firmicutes, and Proteobacteria. Almost all the genes were detected by genome sequence analysis (the exceptions were orf25 of Rhodococcus sp. strain T104 and AAA50358 of Anabaena sp. strain 7120), and all of the proteins are functionally unidentified proteins.

In addition to the proteins described above, thousands of diverse GAF-containing proteins have been discovered thus far; however, there is less information regarding their ligand molecules. Cyclic nucleotide monophosphates bind GAF domain-containing eukaryotic cyclic nucleotide phosphodiesterases and cyanobacterial adenylyl cyclases, and they have been well studied, including structural analysis (for reviews, see references 19, 30, and 57). Other examples include the transcriptional activator FhlA in E. coli, which contains two GAF domains that bind formate to activate the formate-hydrogen lyase system (17); the NifA protein in Azotobacter vinelandii, which is a regulator for nitrogen fixation and modulates the activity in response to a second regulatory protein, NifL, by binding of 2-oxoglutarate to NifA-GAF (29); the histidine kinase DosS in Mycobacterium tuberculosis, which binds heme in the N-terminal GAF domain (43); and NorR in E. coli, which responds to nitric oxide via a mononuclear nonheme iron center in the N-terminal GAF domain (12). To our knowledge, this is the first report of a GAF-containing protein in which hydrophobic aromatic compounds function as ligands.

We identified 10 proteins that exhibit considerable homology with the complete DfdR sequence. The functions of these proteins have not been identified, but some of them have been annotated as RRs of the TCST system in the database, probably because they exhibit considerable amino acid sequence similarity to several RRs in their C-terminal DNA-binding domains. It is noteworthy that Pfam analysis did not reveal any conserved receiver domains for RRs among the proteins homologous to DfdR. Although we cannot exclude the possibility that these proteins behave like RRs in the TCST system (i.e., they activate the transcription of target genes by receiving signals from a partner sensor kinase), we believe that they act as one-component systems for executing regulatory functions.

We noted that all the mycobacterial genes similar to dfdR are located near a gene cluster that is homologous to mce3 of M. tuberculosis (1, 2) at distances ranging from 1.98 kb (M. vanbaalenii Mvan_3973) to 14.0 kb (M. smegmatis MSMEG_0330). There are four copies of mce clusters in M. tuberculosis H37Rv (mce1 to mce4), which are known to be related to M. tuberculosis virulence (9, 15, 24, 40, 45, 52). mce3 expression in M. tuberculosis is negatively regulated by the TetR family regulator Mce3R, but the signals that inactivate the Mce3R repressor have not been identified (41, 42). In this study, genes homologous to mce3R were identified farther upstream of the mce3 cluster than the genes homologous to dfdR in loci of two mycobacterial strains (Mvan_3994 and Mul_2880). Although we did not detect genes that were similar to dfdR in the M. tuberculosis H37Rv genome, the gene organization of mycobacterial DfdR-type regulators and mce3 suggests that there is a regulatory relationship between the mce3 cluster and the DfdR-type regulator in Mycobacterium strains.

The organization of the dfd cluster appears to be unusual, and it suggests that the dfdBCR genes constitute an operon that is regulated by DfdR. However, although we have not excluded this possibility, we found that the dfdR product is not involved in regulation of the dfdBC genes. We previously described high-level constitutive transcription of dfdB in YK2 (21), while regulatory transcription of dfdB has been observed in YK3 (unpublished results), even though the nucleotide sequences of the dfdBCR genes and the 5' upstream 1,589-bp regions are identical in the two strains. These results strongly suggest that the product of dfdR is not directly involved in transcription of dfdB. Additionally, we demonstrated that the DNA fragment that includes dfdR and the 5' upstream 107-bp region, which contains the predicted promoter elements, fully complements the DF utilization function and transcriptional activation of dfdA in the dfdR disruptant of YK2. This suggests that the 5' upstream 107-bp region of dfdR contains a functional promoter for dfdR that is responsible for the weak constitutive transcription of this gene, as shown in Fig. 4. In addition, when the gene cassette for dfdR expression under the control of P_dfdB on a multicopy plasmid was introduced into Rhodococcus strains harboring pRKluxP_A, a DF-independent response of DfdR was observed due to the high-level expression of dfdR (Fig. 5B). This implies that if the dfdR gene was also regulated polycistronically by P_dfdB, it would have caused DF-independent transcriptional activation of dfdA in YK2.

The dfd catabolic pathway genes have been identified in at least four genera of actinomycetes (3, 20, 21, 33, 44), and the levels of nucleotide sequence conservation are high. By using Southern hybridization analysis, we determined that the transcription regulator gene dfdR was highly conserved in all six strains of DF-utilizing actinomycetes that we isolated and analyzed (data not shown), suggesting that dfdR was laterally transferred as a part of the dfd gene cluster in several DF-utilizing actinomycetes. Significant relationships between the effector molecule specificities of DfdR and genes controlled by DfdR, a DfdA aromatic hydrocarbon dioxygenase were found. Previously, we reported that DfdA exhibits broad substrate specificity (20). Resting cells of DfdA-expressing Rhodococcus strains were capable of transforming all of the aromatic compounds listed in Table 1 except 2-hydroxydibenzofuran, which was not activated by DfdR. This suggests that DfdR and DfdA function as a hydroxylation system for diverse hydrophobic aromatic compounds in actinomycetes. We have reported that a remnant of the dihydrodiol dehydrogenase-like gene was detected between dfdA3 and dfdA4, suggesting that the dfdA gene cluster lost the dehydrogenase gene during adaptation and evolution for DF utilization (20) since dehydrogenase activity is not required in this process (Fig. 1A). In addition, the products of the dfdB and dfdC genes also exhibit broad substrate specificity, and the biphenyl metabolite 2,3-dihydroxybiphenyl can be converted to benzoate by resting cells of pDMR5-transformed E. coli JM109 (data not shown). These results suggest that the ancestor of the dfdA-dfdBCR cluster might have been involved in biphenyl catabolism. The evolutionary scheme for the dfd gene cluster would be interesting, and more sequence information on homologous gene clusters in different DF-utilizing actinomycetes strains should be discussed.

 
We are grateful to Masao Fukuda and Keisuke Miyauchi (Nagaoka University of Technology, Nagaoka, Japan) for providing the shuttle reporter plasmid pKLA1. We thank Christiane Genetello (Ghent University, Ghent, Belgium) for providing plasmid pRF30. We also thank Yoshiyuki Ohtsubo for helpful discussions.

This work was supported in part by a grant from the Ecomolecular Science Research Program of RIKEN and by grant-in-aid for young scientists 17780069 from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.

 
* Corresponding author. Mailing address: Environmental Molecular Biology Laboratory, RIKEN, Wako, Saitama 351-0198, Japan. Phone: 81 (48) 462-1111, ext. 5674. Fax: 81 (48) 462-4672. E-mail: tiida{at}riken.jp

Published ahead of print on 24 October 2008.

Supplemental material for this article may be found at http://jp.asm.org/.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Ahmad, S., S. El-Shazly, A. S. Mustafa, and R. Al-Attiyah. 2004. Mammalian cell-entry proteins encoded by the mce3 operon of Mycobacterium tuberculosis are expressed during natural infection in humans. Scand. J. Immunol. 60:382-391.[CrossRef][Medline]
2
2. Ahmad, S., S. El-Shazly, A. S. Mustafa, and R. Al-Attiyah. 2005. The six mammalian cell entry proteins (Mce3A-F) encoded by the mce3 operon are expressed during in vitro growth of Mycobacterium tuberculosis. Scand. J. Immunol. 62:16-24.[CrossRef][Medline]
3
3. Aly, H. A., N. B. Huu, V. Wray, H. Junca, and D. H. Pieper. 2008. Two angular dioxygenases contribute to the metabolic versatility of dibenzofuran-degrading Rhodococcus sp. strain HA01. Appl. Environ. Microbiol. 74:3812-3822.[Abstract/Free Full Text]
4
4. Arai, H., S. Kosono, K. Taguchi, M. Maeda, E. Song, F. Fuji, S.-Y. Chung, and T. Kudo. 1998. Two sets of biphenyl and PCB degradation genes on a linear plasmid in Rhodococcus erythropolis TA421. J. Ferment. Bioeng. 86:595-599.[CrossRef]
5
5. Armengaud, J., J. Gaillard, and K. N. Timmis. 2000. A second [2Fe-2S] ferredoxin from Sphingomonas sp. strain RW1 can function as an electron donor for the dioxin dioxygenase. J. Bacteriol. 182:2238-2244.[Abstract/Free Full Text]
6
6. Armengaud, J., B. Happe, and K. N. Timmis. 1998. Genetic analysis of dioxin dioxygenase of Sphingomonas sp. strain RW1: catabolic genes dispersed on the genome. J. Bacteriol. 180:3954-3966.[Abstract/Free Full Text]
7
7. Armengaud, J., and K. N. Timmis. 1997. Molecular characterization of Fdx1, a putidaredoxin-type [2Fe-2S] ferredoxin able to transfer electrons to the dioxin dioxygenase of Sphingomonas sp. RW1. Eur. J. Biochem. 247:833-842.[Medline]
8
8. Armengaud, J., and K. N. Timmis. 1998. The reductase RedA2 of the multi-component dioxin dioxygenase system of Sphingomonas sp. RW1 is related to class-I cytochrome P450-type reductases. Eur. J. Biochem. 253:437-444.[Medline]
9
9. Arruda, S., G. Bomfim, R. Knights, T. Huima-Byron, and L. W. Riley. 1993. Cloning of an M. tuberculosis DNA fragment associated with entry and survival inside cells. Science 261:1454-1457.[Abstract/Free Full Text]
10
10. Bae, H. W., D. Kim, K. Y. Choi, N. R. Kwon, J. C. Chae, G. J. Zylstra, S. C. Koh, C. H. Lee, and E. Kim. 2007. Functional identification of p-cumate operons in the terpene-degrading Rhodococcus sp. strain T104. FEMS Microbiol. Lett. 266:54-59.[CrossRef][Medline]
11
11. Bunz, P. V., and A. M. Cook. 1993. Dibenzofuran 4,4a-dioxygenase from Sphingomonas sp. strain RW1: angular dioxygenation by a three-component enzyme system. J. Bacteriol. 175:6467-6475.[Abstract/Free Full Text]
12
12. D'Autreaux, B., N. P. Tucker, R. Dixon, and S. Spiro. 2005. A non-haem iron centre in the transcription factor NorR senses nitric oxide. Nature 437:769-772.[CrossRef][Medline]
13
13. Desomer, J., P. Dhaese, and M. V. Montagu. 1990. Transformation of Rhodococcus fascians by high-voltage electroporation and development of R. fascians cloning vectors. Appl. Environ. Microbiol. 56:2818-2825.[Abstract/Free Full Text]
14
14. Finn, R. D., J. Mistry, B. Schuster-Bockler, S. Griffiths-Jones, V. Hollich, T. Lassmann, S. Moxon, M. Marshall, A. Khanna, R. Durbin, S. R. Eddy, E. L. Sonnhammer, and A. Bateman. 2006. Pfam: clans, web tools and services. Nucleic Acids Res. 34:D247-D251.[Abstract/Free Full Text]
15
15. Gioffre, A., E. Infante, D. Aguilar, M. P. Santangelo, L. Klepp, A. Amadio, V. Meikle, I. Etchechoury, M. I. Romano, A. Cataldi, R. P. Hernandez, and F. Bigi. 2005. Mutation in mce operons attenuates Mycobacterium tuberculosis virulence. Microbes Infect. 7:325-734.[CrossRef][Medline]
16
16. Habe, H., M. Miyakoshi, J. Chung, K. Kasuga, T. Yoshida, H. Nojiri, and T. Omori. 2003. Phthalate catabolic gene cluster is linked to the angular dioxygenase gene in Terrabacter sp. strain DBF63. Appl. Microbiol. Biotechnol. 61:44-54.[Medline]
17
17. Hopper, S., and A. Bock. 1995. Effector-mediated stimulation of ATPase activity by the sigma 54-dependent transcriptional activator FHLA from Escherichia coli. J. Bacteriol. 177:2798-2803.[Abstract/Free Full Text]
18
18. Huang, J. Z., and M. A. Schell. 1991. In vivo interactions of the NahR transcriptional activator with its target sequences. Inducer-mediated changes resulting in transcription activation. J. Biol. Chem. 266:10830-10838.[Abstract/Free Full Text]
19
19. Hurley, J. H. 2003. GAF domains: cyclic nucleotides come full circle. Sci. STKE 2003:PE1.[Medline]
20
20. Iida, T., Y. Mukouzaka, K. Nakamura, and T. Kudo. 2002. Plasmid-borne genes code for an angular dioxygenase involved in dibenzofuran degradation by Terrabacter sp. strain YK3. Appl. Environ. Microbiol. 68:3716-3723.[Abstract/Free Full Text]
21
21. Iida, T., Y. Mukouzaka, K. Nakamura, I. Yamaguchi, and T. Kudo. 2002. Isolation and characterization of dibenzofuran-degrading actinomycetes: analysis of multiple extradiol dioxygenase genes in dibenzofuran-degrading Rhodococcus species. Biosci. Biotechnol. Biochem. 66:1462-1472.[CrossRef][Medline]
22
22. Iida, T., K. Nakamura, A. Izumi, Y. Mukouzaka, and T. Kudo. 2006. Isolation and characterization of a gene cluster for dibenzofuran degradation in a new dibenzofuran-utilizing bacterium, Paenibacillus sp. strain YK5. Arch. Microbiol. 184:305-315.[CrossRef][Medline]
23
23. Jaspers, M. C., W. A. Suske, A. Schmid, D. A. Goslings, H. P. Kohler, and J. R. van der Meer. 2000. HbpR, a new member of the XylR/DmpR subclass within the NtrC family of bacterial transcriptional activators, regulates expression of 2-hydroxybiphenyl metabolism in Pseudomonas azelaica HBP1. J. Bacteriol. 182:405-417.[Abstract/Free Full Text]
24
24. Joshi, S. M., A. K. Pandey, N. Capite, S. M. Fortune, E. J. Rubin, and C. M. Sassetti. 2006. Characterization of mycobacterial virulence genes through genetic interaction mapping. Proc. Natl. Acad. Sci. USA 103:11760-11765.[Abstract/Free Full Text]
25
25. Kasuga, K., H. Habe, J. S. Chung, T. Yoshida, H. Nojiri, H. Yamane, and T. Omori. 2001. Isolation and characterization of the genes encoding a novel oxygenase component of angular dioxygenase from the gram-positive dibenzofuran-degrader Terrabacter sp. strain DBF63. Biochem. Biophys. Res. Commun. 283:195-204.[CrossRef][Medline]
26
26. Labbe, D., J. Garnon, and P. C. Lau. 1997. Characterization of the genes encoding a receptor-like histidine kinase and a cognate response regulator from a biphenyl/polychlorobiphenyl-degrading bacterium, Rhodococcus sp. strain M5. J. Bacteriol. 179:2772-2776.[Abstract/Free Full Text]
27
27. Lau, P. C., Y. Wang, A. Patel, D. Labbe, H. Bergeron, R. Brousseau, Y. Konishi, and M. Rawlings. 1997. A bacterial basic region leucine zipper histidine kinase regulating toluene degradation. Proc. Natl. Acad. Sci. USA 94:1453-1458.[Abstract/Free Full Text]
28
28. Letunic, I., R. R. Copley, B. Pils, S. Pinkert, J. Schultz, and P. Bork. 2006. SMART 5: domains in the context of genomes and networks. Nucleic Acids Res. 34:D257-D260.[Abstract/Free Full Text]
29
29. Little, R., and R. Dixon. 2003. The amino-terminal GAF domain of Azotobacter vinelandii NifA binds 2-oxoglutarate to resist inhibition by NifL under nitrogen-limiting conditions. J. Biol. Chem. 278:28711-28718.[Abstract/Free Full Text]
30
30. Martinez, S. E., J. A. Beavo, and W. G. Hol. 2002. GAF domains: two-billion-year-old molecular switches that bind cyclic nucleotides. Mol. Interv. 2:317-323.[Abstract/Free Full Text]
31
31. Mascher, T., J. D. Helmann, and G. Unden. 2006. Stimulus perception in bacterial signal-transducing histidine kinases. Microbiol. Mol. Biol. Rev. 70:910-938.[Abstract/Free Full Text]
32
32. Miyakoshi, M., M. Urata, H. Habe, T. Omori, H. Yamane, and H. Nojiri. 2006. Differentiation of carbazole catabolic operons by replacement of the regulated promoter via transposition of an insertion sequence. J. Biol. Chem. 281:8450-8457.[Abstract/Free Full Text]
33
33. Miyauchi, K., P. Sukda, T. Nishida, E. Ito, Y. Matsumoto, E. Masai, and M. Fukuda. 2008. Isolation of dibenzofuran-degrading bacterium, Nocardioides sp. DF412, and characterization of its dibenzofuran degradation genes. J. Biosci. Bioeng. 105:628-635.[CrossRef][Medline]
34
34. Monna, L., T. Omori, and T. Kodama. 1993. Microbial degradation of dibenzofuran, fluorene, and dibenzo-p-dioxin by Staphylococcus auriculans DBF63. Appl. Environ. Microbiol. 59:285-289.[Abstract/Free Full Text]
35
35. Nojiri, H., H. Habe, and T. Omori. 2001. Bacterial degradation of aromatic compounds via angular dioxygenation. J. Gen. Appl. Microbiol. 47:279-305.[CrossRef][Medline]
36
36. Nojiri, H., M. Kamakura, M. Urata, T. Tanaka, J. S. Chung, T. Takemura, T. Yoshida, H. Habe, and T. Omori. 2002. Dioxin catabolic genes are dispersed on the Terrabacter sp. DBF63 genome. Biochem. Biophys. Res. Commun. 296:233-240.[CrossRef][Medline]
37
37. Ohtsubo, Y., M. Delawary, K. Kimbara, M. Takagi, A. Ohta, and Y. Nagata. 2001. BphS, a key transcriptional regulator of bph genes involved in polychlorinated biphenyl/biphenyl degradation in Pseudomonas sp. KKS102. J. Biol. Chem. 276:36146-36154.[Abstract/Free Full Text]
38
38. O'Leary, N. D., W. A. Duetz, A. D. Dobson, and K. E. O'Connor. 2002. Induction and repression of the sty operon in Pseudomonas putida CA-3 during growth on phenylacetic acid under organic and inorganic nutrient-limiting continuous culture conditions. FEMS Microbiol. Lett. 208:263-268.[CrossRef][Medline]
39
39. Ramos, J. L., S. Marques, and K. N. Timmis. 1997. Transcriptional control of the Pseudomonas TOL plasmid catabolic operons is achieved through an interplay of host factors and plasmid-encoded regulators. Annu. Rev. Microbiol. 51:341-373.[CrossRef][Medline]
40
40. Rengarajan, J., B. R. Bloom, and E. J. Rubin. 2005. Genome-wide requirements for Mycobacterium tuberculosis adaptation and survival in macrophages. Proc. Natl. Acad. Sci. USA 102:8327-8332.[Abstract/Free Full Text]
41
41. Santangelo, M. P., F. C. Blanco, M. V. Bianco, L. I. Klepp, O. Zabal, A. A. Cataldi, and F. Bigi. 2008. Study of the role of Mce3R on the transcription of mce genes of Mycobacterium tuberculosis. BMC Microbiol. 8:38.[CrossRef][Medline]
42
42. Santangelo, M. P., J. Goldstein, A. Alito, A. Gioffre, K. Caimi, O. Zabal, M. Zumarraga, M. I. Romano, A. A. Cataldi, and F. Bigi. 2002. Negative transcriptional regulation of the mce3 operon in Mycobacterium tuberculosis. Microbiology 148:2997-3006.[Abstract/Free Full Text]
43
43. Sardiwal, S., S. L. Kendall, F. Movahedzadeh, S. C. Rison, N. G. Stoker, and S. Djordjevic. 2005. A GAF domain in the hypoxia/NO-inducible Mycobacterium tuberculosis DosS protein binds haem. J. Mol. Biol. 353:929-936.[Medline]
44
44. Schmid, A., B. Rothe, J. Altenbuchner, W. Ludwig, and K. H. Engesser. 1997. Characterization of three distinct extradiol dioxygenases involved in mineralization of dibenzofuran by Terrabacter sp. strain DPO360. J. Bacteriol. 179:53-62.[Abstract/Free Full Text]
45
45. Senaratne, R. H., B. Sidders, P. Sequeira, G. Saunders, K. Dunphy, O. Marjanovic, J. R. Reader, P. Lima, S. Chan, S. Kendall, J. McFadden, and L. W. Riley. 2008. Mycobacterium tuberculosis strains disrupted in mce3 and mce4 operons are attenuated in mice. J. Med. Microbiol. 57:164-170.[Abstract/Free Full Text]
46
46. Shingler, V., M. Bartilson, and T. Moore. 1993. Cloning and nucleotide sequence of the gene encoding the positive regulator (DmpR) of the phenol catabolic pathway encoded by pVI150 and identification of DmpR as a member of the NtrC family of transcriptional activators. J. Bacteriol. 175:1596-1604.[Abstract/Free Full Text]
47
47. Sze, C. C., L. M. Bernardo, and V. Shingler. 2002. Integration of global regulation of two aromatic-responsive ⁵⁴-dependent systems: a common phenotype by different mechanisms. J. Bacteriol. 184:760-770.[Abstract/Free Full Text]
48
48. Takeda, H., A. Yamada, K. Miyauchi, E. Masai, and M. Fukuda. 2004. Characterization of transcriptional regulatory genes for biphenyl degradation in Rhodococcus sp. strain RHA1. J. Bacteriol. 186:2134-2146.[Abstract/Free Full Text]
49
49. Tropel, D., and J. R. van der Meer. 2004. Bacterial transcriptional regulators for degradation pathways of aromatic compounds. Microbiol. Mol. Biol. Rev. 68:474-500.[Abstract/Free Full Text]
50
50. Ulrich, L. E., E. V. Koonin, and I. B. Zhulin. 2005. One-component systems dominate signal transduction in prokaryotes. Trends Microbiol. 13:52-56.[CrossRef][Medline]
51
51. Urata, M., M. Miyakoshi, S. Kai, K. Maeda, H. Habe, T. Omori, H. Yamane, and H. Nojiri. 2004. Transcriptional regulation of the ant operon, encoding two-component anthranilate 1,2-dioxygenase, on the carbazole-degradative plasmid pCAR1 of Pseudomonas resinovorans strain CA10. J. Bacteriol. 186:6815-6823.[Abstract/Free Full Text]
52
52. Van der Geize, R., K. Yam, T. Heuser, M. H. Wilbrink, H. Hara, M. C. Anderton, E. Sim, L. Dijkhuizen, J. E. Davies, W. W. Mohn, and L. D. Eltis. 2007. A gene cluster encoding cholesterol catabolism in a soil actinomycete provides insight into Mycobacterium tuberculosis survival in macrophages. Proc. Natl. Acad. Sci. USA 104:1947-1952.[Abstract/Free Full Text]
53
53. Watanabe, T., H. Fujihara, and K. Furukawa. 2003. Characterization of the second LysR-type regulator in the biphenyl-catabolic gene cluster of Pseudomonas pseudoalcaligenes KF707. J. Bacteriol. 185:3575-3582.[Abstract/Free Full Text]
54
54. Watanabe, T., R. Inoue, N. Kimura, and K. Furukawa. 2000. Versatile transcription of biphenyl catabolic bph operon in Pseudomonas pseudoalcaligenes KF707. J. Biol. Chem. 275:31016-31023.[Abstract/Free Full Text]
55
55. Wittich, R.-M. 1998. Degradation of dioxin-like compounds by microorganisms. Appl. Microbiol. Biotechnol. 49:489-499.[CrossRef][Medline]
56
56. Wittich, R. M., H. Wilkes, V. Sinnwell, W. Francke, and P. Fortnagel. 1992. Metabolism of dibenzo-p-dioxin by Sphingomonas sp. strain RW1. Appl. Environ. Microbiol. 58:1005-1010.[Abstract/Free Full Text]
57
57. Zoraghi, R., J. D. Corbin, and S. H. Francis. 2004. Properties and functions of GAF domains in cyclic nucleotide phosphodiesterases and other proteins. Mol. Pharmacol. 65:267-278.[Free Full Text]

---

Journal of Bacteriology, January 2009, p. 123-134, Vol. 191, No. 1
0021-9193/09/$08.00+0     doi:10.1128/JB.01112-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Iida, T. Articles by Kudo, T. Search for Related Content PubMed PubMed Citation Articles by Iida, T. Articles by Kudo, T.