Versatile Transcription of Biphenyl Catabolic bphOperon in Pseudomonas pseudoalcaligenes KF707*

Pseudomonas pseudoalcaligenes KF707 possesses a chromosomally encoded bph gene cluster responsible for the catabolism of biphenyl/polychlorinated biphenyls. The gene cluster consists of(orf0)bphA1A2(orf3)bphA3A4BCX0X1X2X3D. We studied the role of orf0 and transcription in the KF707bph operon. Primer extension analyses revealed that at least as many as six transcriptional initiation sites exist upstream oforf0, bphA1, bphX0,bphX1, and bphD, including two upstream ofbphD. The orf0-disruptant failed to grow on biphenyl but accumulated large amounts of the biphenyl ringmeta-cleavage yellow compound (2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoate). Western blot analysis revealed that ORF0 protein is inducibly expressed in KF707 in the presence of biphenyl. Gel shift assay revealed that ORF0 directly binds to the orf0 operator region. This binding was greatly enhanced by addition of the biphenyl ring meta-cleavage yellow compound. These results indicated that orf0,bphA1A2(orf3)bphA3A4BC and bphX0X1X2X3Dare independently transcribed, and that ORF0 protein belonging to the GntR family is involved in the regulation of the bphoperon in KF707 and is absolutely required for the expression oforf0 and bphX0X1X2X3D.

Despite the detailed biochemical and genetic analyses of bph genes of various soil bacteria, the knowledge concerning regulation have remained unclear. Recently, it was suggested that the bph gene cluster in Tn4371 found from Ralstonia eutropha A5, bphEFGA1A2A3BCD, forms an operon transcribed from a 70 promoter and that the bphS gene product negatively regulates the transcription of the bph gene cluster as a repressor (18). On the other hand, the bph operon in Gram-positive Rhodococcus sp. M5, bpdC1C2BADEF, is regulated by the twocomponent signal transduction system of bpdS and bpdT. These bpd genes are inducibly transcribed by biphenyl. In this system, BpdS and BpdT function as a sensor histidine kinase and a response regulator, respectively (19). Here we report the versatile transcription of the KF707 bph gene cluster, indicating that the ORF0 protein is an activator and is absolutely involved in the expressions of its own orf0 and bphX0X1X2X3D and that at least six transcriptional initiation sites are present in the KF707 bph gene cluster.

EXPERIMENTAL PROCEDURES
Bacterial Strains and Growth Conditions-The strains and plasmids used in this study are listed in Table I. The biphenyl-utilizing strain P. pseudoalcaligenes KF707 was grown in basal salt medium (BSM) supplemented with 0.2% (w/v) biphenyl as a sole source of carbon as described previously (1). Biphenyl-utilizing defective derivatives of KF707 such as KF730 (bphA1::Tn5-B21), KF748 (bphB::Tn5-B21), and KF744 (bphC::Tn5-B21) were previously constructed (20), and KF7095 (orf0::Tc R ) was constructed in this study. They were grown in BSM supplemented with 0.1% (w/v) sodium succinate and tetracycline (Tc) * This work was supported in part by CREST (Core Research for Evolutional Science and Technology) of the Japan Science and Technology Corporation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) D85852, M83673, D85853, and D85851.
(30 g/ml). For agar plates (1.5% w/v), biphenyl was supplied as vapor in the inverted lid of a Petri dish.
Enzyme Assay-Strain KF707 and the derivatives were pregrown in BSM supplemented with biphenyl or succinate for 24 h, and the one hundredth culture was transferred to the same fresh medium to grow to the stationary phase. Cells were disrupted with a French pressure cell (Ohtake Co., Tokyo) and centrifuged at 14,700 ϫ g for 30 min; the supernatant was used as the crude extract.
BphC (23DHBP dioxygenase) and BphD (HOPD hydrolase) activities were assayed as described previously (1). One unit of BphC activity was defined as the amount of enzyme to produce 1 mol of biphenyl ring meta-cleavage compound (HOPD)/min. One unit of BphD activity was defined as the amount of enzyme to degrade 1 mol of HOPD/min. BphX0 (glutathione S-transferase) and BphX2 (acetaldehyde dehydrogenase) activities were measured as described by Habig and Jakoby (21) and Shingler et al. (22), respectively.
DNA Manipulations, Sequencing, and RNA Preparation-DNA manipulations were performed essentially as described by Sambrook et al. (23). Plasmids were prepared by the rapid alkaline procedure. The DNA fragments to be sequenced were cloned into pUC19 (Takara Shuzo) or pBluescriptII® SK(Ϫ) (Stratagene). By using appropriate primers (Sawady Technology Co., Tokyo), transcriptional initiation sites were determined by the primer extension method. RNA preparation was performed for the cells grown to an A 600 of 0.7 as described by Ausubel et al. (24). The concentration and purity of the RNA were estimated from the A 260 /A 280 ratio and from the RNA bands on formaldehydedenatured agarose gel electrophoresis.
Primer Extension Analyses-Primer extension was performed with the primer extension kit according to the manufacturer's instruction (Promega). The primers used in this study were labeled with [␥-32 P]ATP (Amersham Pharmacia Biotech) by T4 polynucleotide kinase (Takara Shuzo). A 20 -50-g aliquot of total RNA was used in each primer extension reaction. The end-labeled primers were complementary to 23-24-bp sequences located 22-23 nucleotides downstream of the first nucleotides of the orf0, bphX1, and bphD start codon, respectively. In the analysis of bphX0, we used the following primer: 5Ј-CGTAAGGGC-GACGGTTTTGCGTAAGTA-3Ј (in bphC coding sequence). Signal intensities were quantified with a densitometer using Bio-PROFIL, version 6.0 (Vilber Lourmat).
RT-PCR-RT-PCR was performed using the RT-PCR kit according to the manufacturer's instruction (TOYOBO, Osaka). For the detection of the orf0-mRNA, the following primers were used for 8 g of total RNA: 5Ј-ATGAATACGAGAACTCCAAGC-3Ј and 5Ј-CCGCTGACCTCATAG-CGC-3Ј within orf0 coding sequence. To examine whether orf0 and bphA1 could be co-transcribed, the following primers encompassing the both genes were used: 5Ј-CGAGATGCCGAGCGCTGT-3Ј within orf0 coding sequence and 5Ј-CTTCTTTGATTGATGAGCTCAT-3Ј within bphA1 coding sequence. The PCR products were subjected to electrophoresis through 2.0% agarose gel.
Construction of orf0 Expression Plasmid-The orf0 of KF707 was amplified from chromosomal DNA using the following primers: 5Ј-AGCTCCATGGGAATGAATACG-AGAACTCCAAGCGGC-3Ј for the forward sequence, where the NcoI site is underlined and the start codon ATG is in bold type. For the reverse primer sequence, 5Ј-AAGGCTGTC-GACCCTCAACAAACCGAATTCCCGAAC-3Ј was used, where the SalI site is underlined. Amplification of orf0 was carried out for 25 cycles under the following conditions: denaturation, 94°C for 1 min; primer annealing, 52°C for 1.5 min; and primer extension, 72°C for 1.5 min. The PCR products were digested by NcoI and SalI and inserted at the NcoI and XhoI sites of pET32b(ϩ) (Novagen), which contains an isopropyl-␤-D-thiogalactopyranoside-inducible T7 promoter. The orf0 was designed to be fused to Trx (thioredoxin protein)⅐Tag™, S⅐Tag™, and His⅐Tag® at 5Ј end and His⅐Tag® at the 3Ј end. The resulting plasmid, pET32b(ϩ)-orf0 (pTWF3), was transformed into Escherichia coli BL21 (DE3) (Novagen).
subjected to SDS-polyacrylamide gel electrophoresis and transferred to Trans-Blot ® Transfer Medium (Bio-Rad). Anti-ORF0 antibody was prepared by the method described by Imajoh-Ohmi et al. (25). Western blot analysis using anti-ORF0 antibody was carried out to detect the expression of ORF0 protein in KF707 by the method described by Imajoh-Ohmi et al. (25). Signal intensities were quantified with a densitometer.
Construction of pSUPB101-The Tc R gene (4.3 kb) was removed from pSUP102::Tn5-B30 (20, 26) by XhoI digestion. A 2.5-kb SalI fragment containing the orf0 gene was inserted into the XhoI site of pSUP102::Tn5-B30⌬Tc R . Because the resultant plasmid has the unique XhoI site in the middle of the inserted fragment, a purified Tc R gene from pSUP102::Tn5-B30 was then inserted into the XhoI site within orf0 to generate pSUPB101 (Table I). pSUPB101 containing orf0 disrupted by the Tc R gene was introduced into E. coli S17-1 (27) by transformation.
Southern Blot Analysis-Southern blot analysis was performed using the DIG DNA Labeling and Detection Kit according to the manufacturer's instruction (Roche Diagnostics). The chromosomal DNA of the Pseudomonas strains listed in Table I were digested with XhoI and subjected to electrophoresis through 1.0% agarose gels. The digested DNA fragments were transferred onto Biodyne® B (PALL). Hybridization was performed with the DIG-11-dUTP-labeled BamHI-EcoRI fragment (3.9 kb) from pSUP102::Tn5-B30 vector as a probe.
RNase Protection Assay-RNase protection assay was performed using the RNase protection kit according to the manufacturer's instructions (Roche Diagnostics). The RNA probe used in this assay was labeled with DIG-11-UTP by the methods of in vitro transcription using the DIG RNA Labeling Kit (Roche Diagnostics). The probe for the bphX0-mRNA was prepared by cloning a ClaI-PstI fragment containing bphC and bphX0 into pBluescriptII® KS(ϩ) to create pTWF7, which was then linearized with HincII to use as a probe.
E. coli JM109 carrying pTWF6 (pTV118N-orf0) or pTV118N was grown in LB medium containing ampicillin (50 g/ml) and 0.1 mM isopropyl-␤-D-thiogalactopyranoside overnight. The cells were then suspended in EDTA-and dithiothreitol-free binding buffer (20 mM HEPES, pH 7.9, 50 mM KCl, 10% glycerol) and disrupted by a French pressure cell (Ohtake). Cell debris was removed by centrifugation, and the supernatant was mixed with nickel-nitrilotriacetic acid agarose and incubated on ice for 1 h. The matrix was washed twice with buffer (20 mM HEPES, pH 7.9, 50 mM KCl, 10% glycerol, 20 mM imidazole). The proteins were eluted with the same buffer, containing 250 mM imidazole. Protein (purified ORF0) was analyzed by SDS-polyacrylamide gel electrophoresis. The protein concentration was determined using the Bio-Rad protein assay kit (Bio-Rad). Bovine serum albumin was used as a standard.
Preparation of Biphenyl Ring meta-Cleavage Yellow Compound (HOPD)-KF7095 (orf0-disruptant, Table I) was grown in BSM supplemented with 0.1% (w/v) sodium succinate and Tc (30 g/ml) overnight. After centrifugation, the cells were suspended in binding buffer (20 mM HEPES, pH 7.9, 50 mM KCl, 0.5 mM EDTA, 0.5 mM dithiothreitol, 10% glycerol), and 0.2% (w/v) of 23DHBP was added to the cell suspension. After 2 h of incubation, cells were removed by centrifugation at 14,700 ϫ g for 30 min, and the supernatant was filter-sterilized and stocked at Ϫ80°C until use.

RESULTS
Disruption of orf0 -It was previously shown that orf0 is not directly responsible for the enzymatic activity of biphenyl dioxygenase (encoded by bphA1A2A3A4) (28). The orf0 encoded a putative protein of 245 amino acids, and the data base search revealed that the ORF0 protein shares 28% identity to GntR, a regulatory protein known as the transcriptional repressor of the gluconate operon in Bacillus subtilis (29). These two proteins exhibited a high similarity of amino acid sequence in the N-terminal region ( Fig. 2A). Furthermore, a helix-turn-helix motif involved in DNA-binding was predicted in the N-terminal region of ORF0 (Fig. 2B). Therefore, ORF0 was considered to be a possible regulatory protein belonging to the GntR family (30).
To investigate the function, we disrupted orf0 in strain KF707. For this purpose, the plasmid pSUPB101 in which the Tc R gene was inserted in orf0 was constructed as described under "Experimental Procedures." E. coli S17-1 cells carrying pSUPB101 were filter-mated with KF707 overnight. pSUPB101 cannot replicate in Pseudomonas strains; therefore, single cross-over recombinants were first screened on BSM agar plates supplemented with 0.1% (w/v) succinate, Tc (30 g/ml), and gentamicin (20 g/ml). Such single cross-over recombinants were repeatedly subcultured in LB broth to obtain double cross-over recombinants. These were checked for growth on BSM plates supplemented with succinate, Tc, and gentamicin. The loss of the vector-borne gentamicin resistance gene was confirmed by Southern blot analysis (data not shown).
Strain KF7095 (orf0-disruptant) failed to grow on biphenyl and accumulated large amounts of the biphenyl ring metacleavage yellow compound (HOPD) (Fig. 3). In the case of strain KF707 (wild type), HOPD accumulated temporarily, but reduced gradually because of the conversion to benzoic acid by the BphD enzyme (HOPD hydrolase). These results strongly suggested that the bphA1A2A3A4BC genes are expressed in KF7095, but the bphD gene is not.
Primer Extension Analysis of the orf0 and Detection of RNA Transcript-Based on the enzyme assays using KF7095, it was indicated that there are at least two different expression systems in the bph gene clusters in KF707. One is the orf0independent system for the expression of the bphA1A2-(orf3)bphA3A4BC genes, and the other is the orf0-dependent system for the expression of the bphX region and bphD genes.
To examine how the orf0 is transcribed, we performed primer extension analysis. One transcriptional initiation site was found at 106 nucleotides upstream of the start codon of orf0 (Fig. 5A). The possible promoter sequence was assigned to TTGACA (Ϫ35 region) and TAAGTT (Ϫ10 region) (Fig. 5B). This promoter sequence is similar to the consensus sequence of the E. coli 70 -dependent promoter. The extension signal from the biphenyl-grown cells (Fig. 5A, lane B) was much (2.5 times) stronger than that of the succinate-grown cells (Fig. 5A, lane S), indicating that the transcription of orf0 is induced by biphenyl. We also performed Northern blot analysis using antisense orf0 RNA probe but failed to obtain a clear signal corresponding to the orf0 (data not shown).
In addition, we performed RT-PCR against the total RNAs of KF707 and KF7095 using the primers designed to anneal the orf0-coding region from the start codon to the upstream of the unique XhoI site in which the Tc R gene is inserted in KF7095. The results clearly show that 200-bp RT-PCR products were detected when KF707 was grown on either biphenyl or succinate (Fig. 5C, lanes 1 and 2). The amount of RT-PCR product was much higher in the biphenyl-grown cells than that of the succinate-grown cells. However, no amplified DNA was detected in KF7095 (Fig. 5C, lane 3). The result clearly indicated that orf0 is not transcribed in KF7095. We also performed RT-PCR to examine whether orf0 and bphA1 are co-transcribed using the primers encompassing the both genes, but no amplified DNA was detected (Fig. 5C, lane 4). This result indicated that orf0 and bphA1 are independently transcribed.
Western Analysis of ORF0 Protein-We were successful in expressing the ORF0 protein as His-tagged form in E. coli. After purification, we prepared antibody against the ORF0 protein. Using the anti-ORF0 antibody, we detected the ORF0 protein of 27.7 kDa in size from KF707 cell extracts by Western blot analysis (Fig. 6). These data allowed us to conclude that orf0 is absolutely expressed in KF707. The signal from the biphenyl-grown cells (Fig. 6, lane 1) was much stronger (3.1 times) than that of succinate-grown cells, indicating that the expression of ORF0 is induced in the presence of biphenyl. This is consistent with the result obtained in the primer extension analysis and the RT-PCR analysis (Fig. 5, A and C).
Binding of ORF0 Protein to the Operator Region-The results obtained in primer extension analysis and Western analysis strongly indicated that orf0 is inducibly expressed in the presence of biphenyl (Figs. 5A and 6). In addition, the ORF0 protein could be considered to be a regulatory protein in the KF707 bph operon because ORF0 exhibits a helix-turn-helix DNA binding motif in the N-terminal half (Fig. 2B). Furthermore, the sequence boxed in Fig. 5B could be a possible ORF0binding region (termed an operator region). This sequence is located between the transcriptional initiation site and the start codon of orf0 and is similar to the binding sites of some GntRtype regulators such as AphS in the aph operon of a phenoldegrader Comamonas testosteroni TA441 (Fig. 7 and Ref. 31).
To confirm whether ORF0 directly binds to the operator region, we performed gel shift assay. As described under "Experimental Procedures," a 38-bp fragment located between the transcriptional initiation site and start codon of orf0 was labeled with DIG-11-ddUTP. The ORF0 protein fused to His⅐Tag® at the C terminus was expressed in E. coli carrying pTV118N-orf0 (pTWF6). The purified ORF0 protein (28 kDa) was used in gel shift assay. When the ORF0 protein was mixed with the labeled DNA fragment, its mobility was retarded (Fig.  8, shifted band I in lane 2). The labeled DNA fragment was not shifted when an excess amount of cold target DNA was added (Fig. 8, lane 3). The addition of an unrelated DNA did not affect the binding of ORF0 to the target DNA (Fig. 8, lane 6). The shifted band was not detected when the cell extract of E. coli carrying pTV118N was used as a control (Fig. 8, lane 7). These results strongly indicated that the ORF0 protein specifically FIG. 3. Accumulation of HOPD (yellow compound) in KF7095  (orf0-disruptant). These strains were grown in BSM supplemented with biphenyl and succinate, and their cell extracts were used as the crude enzyme in this assay. Although KF707 accumulated HOPD temporarily, HOPD was simultaneously hydrolyzed by the BphD enzyme activity. The data are the means from two independent experiments.
FIG. 4. Enzyme activities of BphC, BphX0, BphX2, and BphD in KF707 and KF707 mutants. Transposon Tn5-B21 shown as a stemloop structure and the tetracycline-resistant (Tc R ) gene shown as a filled box are schematically drawn. The shaded boxes show the inactive genes. ϩ, enzyme activity identical to that of wild type KF707; Ϫ, no enzyme activity. BphC activity of KF707 could not be measured quantitatively as described in Fig. 3 legend. In addition, we failed to perform quantitative assay of the BphX0 and BphX2 activities because of the instability of their substrates (1-chloro-2,4-dinitrochlorobenzene and glutathione for BphX0; NAD ϩ and acetaldehyde for BphX2). The data of the BphD enzyme activity represent the means of triplicate assays. binds to the operator region that lies between its transcriptional initiation site and the start codon.
More interestingly, in the presence of biphenyl-metabolites such as 23DHBP and HOPD, the retardation of shifted bands were much more notable, and the signal intensities were stronger, particularly in the presence of HOPD (Fig. 8, shifted band  II in lanes 4 and 5). These observations indicated that the ORF0 protein strongly binds to the operator region in the presence of biphenyl-metabolites.
Determination of Transcriptional Initiation Sites of bphX0, bphX1, and bphD-Because KF7095 did not have any enzyme activities of BphX0, BphX2, and BphD (Fig. 4), it is conceivable that there are at least two independent transcripts of bphA1A2(orf3)A3A4BC and bphX0X1X2X3D. The transcriptional initiation site was found far upstream of 452 nucleotides from the start codon of bphX0 (Fig. 9A, lane B). This site lies within the center of the coding sequence of bphC. of biphenyl (Fig. 9A, lane S), indicating that the expression of bphX0 is induced by biphenyl or its metabolites. In the case of KF7095, a faint band was detected (Fig. 9A, lane M), but this was concluded to be an artificial band from the following experiments. Using the antisense RNA probe of bphX0 (Fig. 9C), we performed RNase protection assay to detect the mRNA of bphX0. As shown in Fig. 9B, the mRNA of bphX0 was detected in KF707 but not in KF7095. Furthermore, we did primer extension analysis using the primer complementary to the sequence located 24 nucleotides downstream of the bphX0 start codon. No extension signal was detected in KF7095 (data not shown). These results were consistent with the fact that no BphX0 enzyme activity is detected and allowed us to conclude that bphX0 is not expressed in KF7095 and that orf0 is involved in the transcription of bphX0.
Further primer extension analyses revealed one transcriptional initiation site at 48 nucleotides upstream of the start codon of bphX1 (Fig. 10A, lane B). However, no signal was detected from succinate-grown cells (Fig. 10A, lane S) nor from KF7095 (Fig. 10A, lane M). These results suggest that BphX1 is induced by biphenyl or its metabolites and that the orf0 product is absolutely required in the expression.
In addition, two transcriptional initiation sites were found at 30 and 29 nucleotides upstream of the start codon of bphD (Fig.  10B, lane B). The other faint bands detected in lane B were ignored because they were not detected in several primer extension analyses. These two extension signals were detected, but weakly, even in the absence of biphenyl (Fig. 10B, lane S). The intensities of the extension signals detected in the presence of biphenyl (Fig. 10B, lane B) were much stronger (2.7 times) than those in the absence of biphenyl (Fig. 10B, lane S), indicating that bphD is inducibly transcribed in the presence of biphenyl and its metabolites. On the other hand, KF7095 exhibited no extension signal of the bphD gene (Fig. 10B, lane M), indicating that orf0 is absolutely required for the expression of bphD as in the case of bphX0 and bphX1. DISCUSSION In this study, it was found that the ORF0 protein belonging to the GntR family is involved in the regulation of the bph operon in P. pseudoalcaligenes KF707. The orf0 was inducibly expressed in the presence of biphenyl (Figs. 5A and 6). It was further demonstrated that ORF0 itself directly binds to the operator region (Figs. 5B, 7, and 8) by gel shift assays. The retardation was greatly enhanced in the presence of biphenyl ring meta-cleavage compound (HOPD) (Fig. 8). These results indicate that the ORF0 protein may change the conformation in the presence of HOPD and strongly binds to the operator region and promotes the transcription of its own gene.
It was first reported that the GntR protein negatively regulates the gnt operon in B. subtilis (29). The GntR-type regulatory proteins involved in the degradation of aromatic compounds are also reported to be repressors (18,31). However, ORF0 in KF707 is not likely to be a repressor (silencer) but more likely to be an activator being required for the expression of the lower pathway genes of bphX0X1X2X3D (Figs. 4, 9B and  10 (A and B)).
We have preliminary data demonstrating that ORF0 protein also binds to the upstream region (about 200 bp) of bphX0 and bphX1, respectively (data not shown). Thus, it seems that ORF0 controls its own expression and acts as an activator for the expression of bphX0X1X2X3D. However, the same protein is not involved in the expression of the bphA1A2(orf3)bphA3A4BC genes, because the orf0-disruptant (KF7095) converts biphenyl to the HOPD (Fig.  3), indicating that these genes are expressed without the ORF0 protein. Along with this finding, we previously reported that one transcriptional initiation site is located 104 nucleotides upstream of the start codon of bphA1 (3). Furthermore, the result obtained in RT-PCR (Fig. 5, lane 4) revealed that no mRNA encompassing both orf0 and bphA1 is present. These results allowed us to conclude that the orf0 is transcribed alone but not co-transcribed with other degradative genes. The bphS and aphS regulatory genes are independently transcribed with respect to their degradative genes (18,31). On the other hand, the gntR gene in the gnt operon of B. subtilis is inducibly co-transcribed with the other genes. In this case the expression of the GntR repressor is posttranscriptionally regulated (32).
In this and the previous studies (3), primer extension analyses revealed that transcriptional initiation sites exist upstream of orf0, bphA1, bphX0, bphX1, and bphD, respectively. These results revealed that the bph gene cluster has at least six transcriptional initiation sites, including two for the bphD gene. It was reported previously that three transcriptional initiation sites are found upstream of bphA in Pseudomonas sp. strain LB400 whose bph gene cluster is nearly identical to that of KF707 (6). However, the transcription and regulation of LB400 bph genes have not yet been studied in detail since then.
As shown in Figs. 3 and 4, KF7095 expressed BphC but not BphX0, BphX2, and BphD. Primer extension analysis identified the transcriptional initiation site upstream of bphX0 in wild type KF707 (Fig. 9A). These results strongly indicate that the bphA1A2(orf3)A3A4BC and the bphX0X1X2X3D are independently transcribed. However, no typical transcriptional termination signal was found downstream of bphC. At the present, we could determine six transcriptional initiation sites in the KF707 bph operon by primer extension analyses but failed to obtain information about the promoters except for orf0 and the transcriptional terminator(s) in this study. No typical terminator sequences were found in the operon. To investigate the sizes of mRNA transcripts, we performed Northern blot analyses using the antisense probes of bph genes. However, we failed to detect the clear transcript signals because they were smeared (data not shown).
A number of studies in the past revealed that many of the gene clusters involved in the degradations of aromatic compounds including biphenyl are present on movable elements such as plasmids and transposons. Such elements permit microorganisms to degrade aromatic compounds by horizontal gene transfer. It was reported that the bph gene cluster involved in the metabolism of biphenyl/4-chlorobiphenyl is located in the 55-kb transposon Tn4371 (17,18,33,34). Our recent study also revealed that the 90-kb bph-sal element coding for biphenyl and salicylate metabolism behaves like a conjugative transposon in Pseudomonas putida KF715 whose bph gene cluster is nearly identical to that of KF707 except that the bphX0X1X2X3 genes were deleted (35). If this is the case in KF707, the bph genes could be foreign genes derived from other strains. Such genes might be regulated and expressed in different fashions because the sigma factor(s) in KF707 may recognize certain regions as promoters. This could be the reason why so many transcriptional initiation sites were detected in the KF707 bph gene clusters.