Analysis of the 2,4-dichlorophenoxyacetic acid-degradative plasmid pEST4011 of Achromobacter xylosoxidans subsp. denitrificans strain EST4002

Gene 255 (2000) 281–288 www.elsevier.com/locate/gene

Analysis of the 2,4-dichlorophenoxyacetic acid-degradative plasmid pEST4011 of Achromobacter xylosoxidans subsp. denitrificans strain EST4002 Eve Vedler *, Viia Ko˜iv, Ain Heinaru Institute of Molecular and Cell Biology, Tartu University, 23 Riia Street, 51010 Tartu, Estonia Received 5 January 2000; received in revised form 7 May 2000; accepted 17 July 2000 Received by D.L. Court

Abstract The 2,4-dichlorophenoxyacetic acid (2,4-D)-degradative bacterium Achromobacter xylosoxidans subsp. denitrificans strain EST4002, isolated in Estonia more than 10 years ago, was found to contain the 70 kb plasmid pEST4011 that is responsible for the bacterium having had obtained a stable 2,4-D+ phenotype. The tfd-like genes for 2,4-D degradation of the strain EST4002 were located on a 10.5 kb region of pEST4011, but without functional genes coding for chloromuconate cycloisomerase and chlorodienelactone hydrolase. The latter two genes are probably encoded by homologous, tcb-like genes, located elsewhere on pEST4011. We also present evidence of two copies of insertion element IS1071-like sequences on pEST4011. IS1071 is a class II (Tn3 family) insertion element, associated with different catabolic genes and operons and globally distributed in the recent past. We speculate that this insertion element might have had a role in the formation of plasmid pEST4011. The 28 kb plasmid pEST4012 is generated by deletion from pEST4011 when cells of A. xylosoxidans EST4002 are grown in the absence of 2,4-D in growth medium. We propose that this is the result of homologous recombination between the two putative copies of IS1071-like sequences on pEST4011. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Chloroaromatic catabolism; Genes coding for enzymes of 2,4-dichlorophenoxyacetic acid degradation; Sequence analysis

1. Introduction Many microorganisms are able to degrade the widely used herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) (Loos et al., 1979; Fournier, 1980; Don and Pemberton, 1981; Ou, 1984; Ka and Tiedje, 1994; Fulthorpe et al., 1995; Maltseva et al., 1996). The 2,4-D degraders that belong to the b and c subdivisions of Proteobacteria contain tfd-like (pJP4-like) genes for 2,4-D degradation, often encoded on transmissible plasmids and horizontally spread among the representatives of these two Abbreviations: A, adenosine; bp, base pair(s); C, cytosine; 2,4-D, 2,4-dichlorophenoxyacetic acid; G, guanosine; IR, inverted repeat sequence; IS, insertion sequence; kb, kilobase(s) or 1000 bp; LB, Luria– Bertani (medium); MAR, maleylacetate reductase; ORF, open reading frame; PCR, polymerase chain reaction; RBS, ribosome-binding site(s); rRNA, ribosomal RNA; S, sedimentation constant; SDS, sodium dodecyl sulfate; SSC, 0.15 M NaCl/0.015 M Na · citrate 3 pH 7.6; T, thymidine. * Corresponding author. Tel.: +3727-375014; fax: +3727-420286. E-mail address: [email protected] (E. Vedler)

subdivisions ( Ka et al., 1994a,b,c; Top et al., 1995). Plasmid pJP4 from the well-characterized bacterium Ralstonia eutropha (formerly Alcaligenes eutrophus) JMP134 (Don and Pemberton, 1985; Streber et al., 1987; Kaphammer et al., 1990; Perkins et al., 1990; Matrubutham and Harker, 1994; Leveau and van der Meer, 1996; Leveau et al., 1998), originally isolated in Australia, carries a 22 kb DNA region encoding tfd genes responsible for converting 2,4-D to a common intermediate metabolite 3-oxoadipate ( Figs. 1 and 2). This DNA region has a mosaic-like structure and has apparently been composed of DNA fragments of different bacterial origins ( Fulthorpe et al., 1995). The catabolic genes of pJP4 — tfdA, tfdB and tfdCDEF, as well as the module 2 genes tfdD C E F B and II II II II II tfdK — are all equally expressed in the presence of 2,4-D (Leveau et al., 1999). There is strong homology between the tfdCDEF 2,4-dichlorocatechol degradative operon and other chlorocatechol degradative operons — tcbCDEF from 1,2,4-trichlorobenzene degradative plasmid pP51 of Pseudomonas sp. strain P51 (van der Meer

0378-1119/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 3 7 8 -1 1 1 9 ( 0 0 ) 0 0 32 9 - 2

282

E. Vedler et al. / Gene 255 (2000) 281–288

Fig. 1. (A) Degradation pathway of 2,4-D along with homologous genes coding for enzymes of different chlorocatechol degradation pathways. (B) Putative 2,4-D degradation pathway coded by the genes of plasmid pEST4011.

Fig. 2. Restriction map of plasmids pEST4011 and pEST4012 and putative positions of IS elements on pEST4011. Organization of BamHI-B and HindIII-E fragments of pEST4011 and the region of pJP4 encoding tfd genes (Leveau and van der Meer, 1996; GenBank accession number U16782; see Sections 1 and 3.2). Directions of transcription are shown for each gene and ORF. The truncated tfdD gene of pEST4011 is designated ORF . The gray regions represent significant homology between plasmids pEST4011 and pJP4. 243bp

E. Vedler et al. / Gene 255 (2000) 281–288

et al., 1991) and clcABDE from 3-chlorobenzoate degradative plasmid pAC27 of Pseudomonas putida (Ghosal and You, 1989) (Fig. 1). The stable 2,4-D degradative plasmid pEST4011 (70 kb) is a natural deletant plasmid generated from unstable pEST4002 (78 kb) of the bacterium (originally named Pseudomonas sp. EST4002) isolated in Estonia (Ausmees and Heinaru, 1990). In this study we compared the 2,4-D degradative strain EST4002, which now was found to contain the stable plasmid pEST4011, with databank strains by using Biolog identification and 16S rRNA gene sequencing. The pEST4011 genes tfdR, tfdC and tfdB have been sequenced and it has been shown that TfdR is a LysRtype transcriptional activator, its effector molecule is 2,4-dichloro-cis,cis-muconate and it activates the expression of the divergently transcribed tfdCB operon ( Ko˜iv et al., 1996; Vedler et al., 2000) (Fig. 2). In this study we finished determination of the nucleotide sequence of the pEST4011 region, which had been shown to hybridize to pJP4 tfd genes (Ma¨e et al., 1993). It was shown that a 10.5 kb region of pEST4011 contains almost the whole pathway of tfd-like genes for 2,4-D degradation (Figs. 1 and 2), but there are no functional tfd-like genes coding for chloromuconate cycloisomerase (tfdD) and chlorodienelactone hydrolase (tfdE ) either on plasmid pEST4011 or chromosome of the strain EST4002. Instead, these two enzymes are probably encoded by homologous, tcb-like (pP51-like) genes (tcbD and tcbE ), located on pEST4011. Also found were sequences on pEST4011 similar to the insertion element IS1071, associated with catabolic genes or operons from different geographically distinct isolates (Di Gioia et al., 1998).

2. Material and methods 2.1. Bacterial strains and plasmids used, media and culture conditions The bacterial strains used in this study were Escherichia coli DH5a (BLR), Achromobacter xylosoxidans subsp. denitrificans EST4002 [pEST4011] with 2,4-D+ phenotype and A. xylosoxidans subsp. denitrificans EST4003 [pEST4012] with 2,4-D− phenotype. Vector plasmid pBluescript(SK ) (Stratagene) was used to clone pEST4011 BamHI-B and HindIII-E fragments for sequence analysis. E. coli and A. xylosoxidans were grown in LB at 37°C and 30°C respectively. 2.2. DNA manipulations and sequence analysis Plasmid and chromosomal DNA preparations and other DNA manipulations were carried out by established procedures (Sanger et al., 1977; Sambrook et al., 1989). 16S rRNA gene of EST4002 was PCR-amplified

283

using the forward primer PCRI (5∞ AGAGTTTGATCATGGCTCAG 3∞), complementary to E. coli 16S rRNA gene positions 6 to 26, and the reverse primer PCRII (5∞ TACGGTTACCTTGTTACGACTT 3∞), complementary to E. coli 16S rRNA gene positions 1513 to 1492. The PCR product (about 1.5 kb) was sequenced with T7 Sequenase v2.0 PCR product sequencing kit (Amersham Life Science, Inc. 1997), using primers PCRI, PCRII and the internal reverse primers SEQ1 (5∞ GTATTACCGCGGCTGCTGG 3∞), SEQ2 (5∞ TTGCGCTCGTTGCGGGACT 3∞), SEQ3 (5∞ ACGGGCGGTGTGTACAAG 3∞) and S4 (5∞ CCAGGGTATCTAATCC 3∞), complementary to conserved regions of different eubacterial 16S rRNA genes. Fragments of putative chloromuconate cycloisomerase and chlorodienelactone hydrolase were respectively PCR-amplified using primer pairs TCBD1 (5∞ GGCACAATCGGTTCAAGGT 3∞)–TCBD2 (5∞ TACAACGTCTCGAACTCGC 3∞), complementary to regions in Variovorax paradoxus plasmid pTV1 putative tfdD gene conserved between different cycloisomerase genes, and TCBE1 (5∞ GTCGCGGTGATCGGCTATTG 3∞)–TCBE2 (5∞ CGAACGAGTGCCCGGCGTC 3∞), complementary to regions in Pseudomonas sp. strain P51 plasmid pP51 tcbE gene conserved between different hydrolase genes, and the plasmid pEST4011 DNA as a template. The fragment of putative chloromuconate cycloisomerase gene was sequenced with Taq DNA Sequencing Kit for Cycle Sequencing (Boehringer Mannheim GmbH ), using the same radioactively labeled primers that were used in PCR amplification, and TCBD PCR product as a template. T4 polynucleotide kinase ( Fermentas) and [c-32P]dATP were used to end-label primers TCBD1 and TCBD2 for sequence analysis and primers HINDE (5∞ CAATGAAATCAACGGTCTA 3∞), TCBD2 and TCBE2 for Southern hybridization analyses. Computer analysis of DNA and amino acid sequences was performed with the program PC/GENE (Genofit, Geneva, Switzerland ). Database searches were performed with the BLASTN and BLASTX programs at the National Center for Biotechnology Information (NCBI ). 2.3. Southern hybridization analyses Plasmids pEST4011, pEST4012 and the chromosome of strain EST4002 were digested with restrictional endonuclease BamHI ( Fermentas) overnight at 37°C, fractionated by 0.7% agarose gel electrophoresis and transferred to a Hybond-N+ nylon filter (Amersham). When the 191 bp NdeI–StuI fragment of the truncated tfdD sequence and the 305 bp PflMI–PvuI fragment of dysfunctional tfdE sequence of pEST4011 BamHI-B fragment were used as the probes, filters were hybridized overnight at +65°C in 6× SSC, 0.5% SDS, 5× Denhardt’s and 200 mg of denatured Salmon sperm

284

E. Vedler et al. / Gene 255 (2000) 281–288

DNA per milliliter. Probes were labeled with [a-32 P]dCTP by random priming (Amersham). After hybridization, membranes were washed in standard washes (6× SSC–0.1% SDS, 2× SSC–0.1% SDS, 1× SSC– 0.1% SDS at +65°C for 30 min). When the end-labeled primers HINDE, TCBD2 and TCBE2 were used as the probes, filters were hybridized for 2 h at +50°C in 6× SSC, 0.5% SDS, 5× Denhardt’s and 200 mg of denatured Salmon sperm DNA per milliliter. After hybridization, membranes were washed twice in 4× SSC–0.1% SDS at +37°C for 10 min. Images were made with a PhosphorImager and Image Quant software (Molecular Dynamics, Sunnywale, CA).

3. Results and discussion 3.1. Isolation of pEST4011 from strain EST4002 and identification of this 2,4-D degradative strain Ausmees and Heinaru (1990) reported on the isolation of a 2,4-D degradative bacterium, containing a 78 kb plasmid pEST4002, in Estonia. This strain had a very unstable 2,4-D+ phenotype: 80% of the cells had lost their ability to grow on 2,4-D after growth in a non-selective LB medium for 4 h. To overcome this instability and difficulties with DNA extraction from the wild-type strain, pEST4002 was transferred by conjugation into the plasmid-free recipient strain P. putida PaW340 (Fig. 3). The achieved 2,4-D+ transconjugants of PaW340 contained a 70 kb plasmid pEST4011. Subsequently, P. putida strain PaW85 was transformed with pEST4011 and a stable 2,4-D+ transformant (pEST4011) was obtained and named EST4021 (Ma¨e et al., 1993). The original 2,4-D degradative strain

EST4002 was maintained on minimal medium agar plates containing 0.1% 2,4-D. In 1999, we found that this strain contained a plasmid that had the same size and restrictional pattern as pEST4011 did, and this plasmid was also named pEST4011. We found that pEST4011 is very stable in EST4002: growth of EST4002 in LB medium for 24 h resulted only in 1–2% cells with irreversible 2,4-D− phenotype, which all contained the same 28 kb deletant plasmid. The latter strain was named EST4003. The corresponding deletant plasmid pEST4012 was also achieved by Ma¨e et al. (1993) from 2,4-D− P. putida strain EST4022 when they cultivated P. putida EST4021 [pEST4011] cells in LB for 30 generations. The correlated restrictional maps of pEST4011 and pEST4012 are given in Fig. 2. The 2,4-D degradative strain EST4002 was originally named Pseudomonas sp. (Ma¨e et al., 1993). In this study this strain was analyzed at a physiological level by microtiter plate-bound substrate utilization assay (Biolog; Biolog Inc., Hayward, CA) and at the phylogenetic level by sequencing and comparison of the nearly complete PCR-amplified 16S rRNA gene. The closest relative by Biolog identification (readings were taken after 24 h) appeared to be A. xylosoxidans subsp. denitrificans (formerly Alcaligenes xylosoxidans subsp. denitrificans; belongs to the b subdivision of Proteobacteria) with similarity coefficient 0.951 (1.0 indicates complete homology to a databank species). The comparison of the 16S rRNA gene of strain EST4002 with databank strains supported the result of Biolog identification, as the most relative strain appeared to be A. xylosoxidans subsp. denitrificans (GenBank accession number M22509) (only 12 mismatches between these two genes). In this alignment, 21 unspecified nucleotides of the GenBank sequence M22509 were not used. According

Fig. 3. Schematic presentation of different strains and plasmids used and discussed in this study.

E. Vedler et al. / Gene 255 (2000) 281–288

to the results given above, we named our 2,4-D degradative strain A. xylosoxidans subsp. denitrificans EST4002. 3.2. Sequencing and databank comparison of 2,4-D degradation genes of pEST4011 We have previously shown (Ma¨e et al., 1993) that different fragments of pJP4 containing at least parts of the genes tfdA, tfdS, tfdR, tfdD , tfdT, tfdCDEF, tfdB II and ISJP4 hybridized only with the 10.8 kb BamHI-B restriction fragment of pEST4011 (Fig. 2). Sequence analysis of the 5 kb XhoI–StuI fragment of this pEST4011 region and expression studies ( Ko˜iv et al., 1996; Vedler et al., 2000) revealed three functional genes necessary for 2,4-D degradation: tfdR, tfdC and tfdB. The latter two genes form one operon transcribed from tfdC to tfdB and they code for chlorocatechol 1,2-dioxygenase and 2,4-dichlorophenol hydroxylase respectively. The tfdR gene is located upstream and transcribed divergently from the tfdCB operon, and its product, a LysR-type transcriptional activator, activates the expression of the tfdCB operon. tfdR, tfdC and tfdB of pEST4011 were shown to be the most similar to the pJP4 module 2 genes tfdR, tfdC and tfdB . Although II II the pEST4011 nucleotide sequence between tfdC and tfdB is very similar to the pJP4 module 2 tfdE gene, II deletion of GC dinucleotide (corresponding to the positions 73 and 74 in the pJP4 tfdE gene) in the coding II region leads to frameshifting and no functional protein could be translated from this region (GenBank accession number U32188). We also found a 243 bp nucleotide sequence in the tfdR–tfdC intergenic region similar to the beginning of pJP4 tfdD gene, which encodes chloroII muconate cycloisomerase. This region was designated ORF .We have shown that ORF is the most 243bp 243bp upstream part of the tfdCB operon, but it is not a translated ORF since there is no RBS in front of it because the first nucleotide in its ATG translation initiation codon is the transcription start site of the tfdCB operon ( Vedler et al., 2000). In order to find all genes necessary for 2,4-D degradation we finished sequencing the 10.8 kb BamHI-B and the 5 kb HindIII-E restriction fragments of pEST4011 (Fig. 2). We found three additional ORFs downstream of the tfdB gene, all preceded by potential RBS (AGGAG in all cases). ORF1 was found 99 bp downstream of tfdB, it was 1392 bp long and its amino acid sequence was most similar to pJP4 TfdK (73% identity over the 456 amino-terminal amino acids). It has been shown that the pJP4 tfdK gene product facilitates uptake of 2,4-D by R. eutropha JMP134 (Leveau et al., 1998). The predicted ORF1 polypeptide bears all the putative transmembrane domains and motifs, which is characteristic for the members of the major facilitator superfamily of transporter proteins (including the pJP4 TfdK ) (Jessen-Marshall et al., 1995). Thus we speculate that

285

the pEST4011 TfdK may be a functional transporter of 2,4-D. ORF2 was found 149 bp downstream of ORF1; it was 864 bp long and its amino acid sequence was similar to pJP4 TfdA, the a-ketoglutarate-dependent dioxygenase (Fukumori and Hausinger, 1993) (81% identity). As we have shown that 2,4-D dioxygenase ( TfdA) activity could be detected in P. putida PaW85 cells carrying the 10.8 kb BamHI fragment of pEST4011 cloned into the vector plasmid pKT240 (Ma¨e et al., 1993), we assume that ORF2 codes for an active 2,4-D dioxygenase. ORF3 was found 1811 bp downstream of ORF2; it was 1074 bp long and its amino acid sequence was most similar to pJP4 TfdF, the maleylacetate reductase (MAR) ( Kasberg et al, 1995) (48% identity over the 352 amino-terminal amino acids). Alignment of the deduced ORF3 polypeptide with different MARs showed that ORF3 contains all patterns typical for this enzyme (Seibert et al., 1998). These results propose that ORF3 may be a functional MAR. Alignment results, obtained previously ( Ko˜iv et al., 1996; Vedler et al., 2000) and in this study, show that the tfd genes, present on the 10.5 kb region of pEST4011, obviously comprise genes of different origins. The genes tfdR to tfdK of pEST4011 (Fig. 2) are very similar to, and organized in the same way as, the pJP4 module 2 genes tfdR to tfdK (GenBank accession number U16782). Thus it is very likely that these regions of pEST4011 and pJP4 have evolved from the same origin. In pJP4, it has been suggested that the module 2 tfd genes have been incorporated into an early version of this plasmid as one unit by two copies of the insertion element ISJP4 (Leveau and van der Meer, 1997). In pEST4011, the regions corresponding to the pJP4 tfdF gene and the downstream part of the tfdD gene, II II as well as the GC dinucleotide in the middle of the tfdE gene, have been lost during evolution. II McGowan et al. (1998) have sequenced internal conserved regions of tfdA genes of phylogenetically different 2,4-D degraders. When we aligned the corresponding region of pEST4011 tfdA (bases 453–765) to these sequences, we found that this fragment was identical to class III tfdA sequences that have been found in phylogenetically distant 2,4-D degraders of Comamonas–Rhodoferax group (b-Proteobacteria) and Halomonas group (c-Proteobacteria). The pJP4 tfdA gene belongs to class I sequences that have been found in Burkholderia, Comamonas–Rhodoferax and Alcaligenes–Ralstonia groups (all b-Proteobacteria). Considering these results, we presume that the pEST4011 tfdA gene has a different origin than the pJP4 tfdA gene does, and this gene has been recruited into a precursor plasmid of pEST4011 independently from other tfd genes. While the (putative) products of the genes tfdA, tfdB, tfdC, ‘tfdE’ (frameshift eliminated for alignment) and tfdK of pEST4011 are about 75–90% similar to corre-

286

E. Vedler et al. / Gene 255 (2000) 281–288

sponding proteins coded by pJP4 module 2 genes, the deduced amino acid sequence of pEST4011 tfdF gene is only about 48% identical to pJP4 module 1 tfdF gene product and about 44% identical to module 2 tfdF II gene product. With each other, the known MARs share about 45–60% identity, thus being more heterogeneous than other enzymes for degradation of chlorocatechols. The pEST4011 tfdF gene has obviously evolved from an origin not shared with other known genes coding for MARs and has apparently been independently incorporated into a precursor plasmid of pEST4011. 3.3. Detection of two copies of the insertion element IS1071-like sequences on plasmid pEST4011 About 3.7 kb downstream of ORF3 we found a 280 bp sequence (we sequenced this region only to the HindIII restriction site) ( Fig. 2) that was 100% identical at the nucleotide level to IS1071 region containing the 110 bp inverted repeat sequence (IR) and the beginning of the transposase gene. IS1071 is a unique class II (Tn3 family) insertion sequence: it is 3.2 kb long and contains one ORF of 2.91 kb coding for a transposase protein ( TnpA) (Nakatsu et al., 1991). It has been shown that this insertion element has an important role in the mobilization of different catabolic genes and operons into composite transposon structures on plasmids and it has been distributed globally in a number of host genera in the recent past (Di Gioia et al., 1998). In order to find out whether there may be more copies of IS1071 on pEST4011, we performed Southern hybridization using the radioactively labeled oligonucleotide HINDE, complementary to a part of the 110 bp IR of IS1071, as the probe. The analysis showed that this oligonucleotide hybridized to three different BamHI restriction fragments of pEST4011, namely BamHI-A, BamHI-F and BamHI-G, and to the BamHI-A restriction fragment of pEST4012 (Fig. 2). The nucleotide sequence of IS1071 has one BamHI site and two HindIII sites (at a distance of 1.1 kb) (GenBank accession number M65135). The Southern hybridization result proposes that there may be two copies of IS1071 on pEST4011, the IRs of one element being on BamHI-A and BamHI-F, and the IRs of the other element being on BamHI-A and BamHI-G restriction fragments. This result is supported by the presence of two 1.1 kb HindIII restriction fragments on pEST4011 at the same positions where the two elements might be according to the Southern analysis. Although these results do not show whether these two putative IS elements of pEST4011 are complete and functional transposable elements, forming a composite transposon, we can speculate that they have had a role in the formation of the 2,4-D degradative plasmid pEST4011. As the Southern analysis showed that there (is) are (one) two IRs of IS1071 on the BamHI-A fragment of the deletant plasmid

pEST4012 and that pEST4012 actually consists only, or mainly, of this BamHI-A fragment (Fig. 2), there is a possibility that the formation of pEST4012 in the absence of 2,4-D in the growth medium may be the result of homologous recombination between the two putative direct copies of this IS element on pEST4011. If this is the case, the 2,4-D degradation genes of pEST4011 must be coded on the 39 kb region between the IS1071-like elements ( Fig. 2). 3.4. Determination of genes coding for chloromuconate cycloisomerase and chlorodienelactone hydrolase As described in Section 3.2, all pJP4 tfd genes hybridized only with pEST4011 BamHI-B fragment (Fig. 2). Sequence analysis of this region revealed that there are no functional genes coding for chloromuconate cycloisomerase (tfdD) and chlorodienelactone hydrolase (tfdE ) either on the BamHI-B or the HindIII-E fragment. In order to find out whether the other copies of tfdD and/or tfdE genes, similar to the ones on pEST4011 BamHI-B fragment, might be on pEST4011 or chromosome of EST4002, we undertook Southern hybridization using the 191 bp NdeI–StuI fragment of the truncated tfdD sequence, ORF , in front of the tfdC gene of 243bp pEST4011 and the 305 bp PflMI–PvuI fragment of the dysfunctional tfdE sequence of pEST4011 as the probes. These studies showed that the two probes did not hybridize to any BamHI fragment of pEST4011 or the chromosome of EST4002 other than to the BamHI-B fragment of pEST4011. As isolation of EST4002 chromosome also produced small amounts of plasmid DNA, in the case of hybridization with EST4002 chromosomal DNA only the fragment corresponding to the plasmid hybridized. Thus it seems very unlikely that there are functional genes for chloromuconate cycloisomerase and chlorodienelactone hydrolase in the strain EST4002, either on plasmid or chromosome, the nucleotide sequences of which are most similar to the corresponding genes of the tfd pathway. Database searches done with pEST4011 tfd genes revealed that the published nucleotide sequence of V. paradoxus plasmid pTV1 (GenBank accession number AB028643) is nearly 100% identical to the corresponding sequence of pEST4011. 2.6 kb downstream of pTV1 tfdR there is an ORF coding for putative chloromuconate cycloisomerase (tfdD), the nucleotide sequence of which is most similar to the tcbD gene of Pseudomonas sp. strain P51 plasmid pP51 (81% over 769 bp). This finding led us to an idea to design a pair of oligonucleotides for both putative cycloisomerase and hydrolase genes of pEST4011 — TCBD1 and TCBD2, complementary to regions in pTV1 putative tfdD gene conserved between different cycloisomerase genes, and TCBE1 and TCBE2, complementary to regions in pP51 tcbE gene conserved between different hydrolase genes. Southern

E. Vedler et al. / Gene 255 (2000) 281–288

hybridization, done with radioactively labeled TCBD2 and TCBE2 primers, showed that both primers hybridized to BamHI-C, BamHI-D and/or BamHI-E fragments of pEST4011 (BamHI-D and BamHI-E fragments are indistinguishable by gel electrophoresis as they are of the same size). TCBE2 primer hybridized very weakly to the pEST4011 BamHI-B fragment as well. We performed PCR amplification with these two pairs of primers using the plasmid pEST4011 as a template. The results showed a PCR product of about 300 bp (as expected) only with TCBD primers. Sequence analysis of the TCBD PCR product (293 bp of it was sequenced ) showed that it was 100% identical to the corresponding fragment of V. paradoxus pTV1 tfdD gene and 80% identical to the tcbD gene of Pseudomonas sp. strain P51 plasmid pP51 at the nucleotide level. Amino acid sequence (97 amino acids) of TCBD PCR product was 85% identical to TcbD of pP51 and CbnB of R. eutropha NH9 plasmid pENH91 (Ogawa and Miyashita, 1999), 75% identical to ClcB of P. putida plasmid pAC27, 59% identical to TfdD of pJP4 and only 43% identical (over 66 amino acids) to TfdD of pJP4. The TCBD PCR II product must have been amplified from putative tcblike gene(s) somewhere on the unsequenced region of pEST4011, because the region that could bind the TCBD primers is not present in ORF of the pEST4011 243bp BamHI-B fragment. As the sequenced TCBD PCR product was amplified from the whole plasmid pEST4011 as a template, it shows that this is the only variant of this type of gene on pEST4011, and according to the Southern analysis there is probably at least one copy of this gene on pEST4011 BamHI-C, BamHI-D (and/or BamHI-E) fragments. In the case of TCBE primers, we expected that the product would be amplified from putative tcb-like gene(s), and not from the dysfunctional tfdE sequence of the BamHI-B fragment. Nevertheless, TCBE primers were obviously not complementary enough to anneal with (any) putative chlorodienelactone hydrolase gene(s) on pEST4011. As the Southern analysis showed that TCBE2 primer hybridized much more strongly with pEST4011 BamHI-C and BamHI-D/E fragments than with BamHI-B fragment, we believe that there are at least two copies of a gene coding for chlorodienelactone hydrolase on pEST4011 which are more similar to tcbE than to tfdE gene. The 39 kb region of pEST4011 between the two putative IS1071-like elements contains genes of different origins which all serve for 2,4-D degradation. All these genes are deleted from pEST4011 when cells, harboring this plasmid, are grown in the absence of 2,4-D in growth medium. We have sequenced only about 16 kb of this region on which a part of the 2,4-D degradation pathway is coded by the tfd-like genes. In this study we have found evidence that the remaining region of about 23 kb does not contain any tfd-like genes and that genes, more similar to tcb-like sequences, have been recruited

287

into pEST4011 to code for at least two enzymes of the 2,4-D degradation pathway — chloromuconate cycloisomerase and chlorodienelactone hydrolase (Fig. 1). It will be interesting to finish analysis of the whole 39 kb region of pEST4011 in order to determine all genes present on that fragment. The nucleotide sequence of EST4002 16S rRNA gene has been assigned GenBank accession number AF232712. The nucleotide sequences of the plasmid pEST4011, discussed and partly sequenced in this study, have been assigned GenBank accession numbers U32188 (the genes tfdR, tfdC, ‘tfdE’, tfdB, tfdK, tfdA and tfdF ), AF233095 (a fragment of putative element IS1071) and AF233096 (a fragment of putative gene for chloromuconate cycloisomerase).

4. Conclusions (1) The 2,4-D degradative bacterium, isolated in Estonia more than 10 years ago, is A. xylosoxidans subsp. denitrificans strain EST4002. Maintaining this strain on minimal medium agar plates containing 2,4-D for several years has resulted in a much more stable 2,4-D+ phenotype due to the deletion of about 8 kb fragment of pEST4002 which gave rise to the plasmid pEST4011. (2) A 10.5 kb region of pEST4011 contains almost the whole pathway of tfd-like genes for 2,4-D degradation without functional genes for chloromuconate cycloisomerase and chlorodienelactone hydrolase. The latter two genes are obviously coded by tcb-like genes located elsewhere on pEST4011. (3) The two putative copies of insertion element IS1071-like sequences could have had a role in the formation of a precursor plasmid of the 2,4-D degradative plasmid pEST4011. There is also a possibility that the generation of the deletant plasmid pEST4012 in the absence of 2,4-D in the growth medium is the result of homologous recombination between the two IS1071-like sequences.

Acknowledgement This work was supported by the Institute of Molecular and Cell Biology, by grant 2334 from the Estonian Science Foundation.

References Ausmees, N.R., Heinaru, A.L., 1990. New plasmids of herbicide 2,4-D degradation. Genetika 26, 770–772. (in Russian). Di Gioia, D., Peel, M., Fava, F., Campbell, R.C., 1998. Structures of homologous composite transposons carrying cbaABC genes from

288

E. Vedler et al. / Gene 255 (2000) 281–288

Europe and North America. Appl. Environ. Microbiol. 64, 1940–1946. Don, R.H., Pemberton, J.M., 1981. Properties of six pesticide degradation plasmids isolated from Alcaligenes paradoxus and Alcaligenes eutrophus. J. Bacteriol. 145, 681–686. Don, R.H., Pemberton, J.M., 1985. Genetics and physical map of the 2,4-dichlorophenoxyacetic acid degratative plasmid pJP4. J. Bacteriol. 161, 466–468. Fournier, J.C., 1980. Enumeration of the soil microorganisms able to degrade 2,4-D by metabolism or cometabolism. Chemosphere 9, 169–174. Fukumori, F., Hausinger, R.P., 1993. Alcaligenes eutrophus JMP134 ‘‘2,4-dichlorophenoxyacetate monooxygenase’’ is an a-ketoglutarate-dependent dioxygenase. J. Bacteriol. 175, 2083–2086. Fulthorpe, R.R., McGowan, C., Maltseva, O.V., Holben, W.E., Tiedje, J.M., 1995. 2,4-Dichlorophenoxyacetic acid degrading bacteria contain mosaics of catabolic genes. Appl. Environ. Microbiol. 61, 3274–3281. Ghosal, D., You, I.-S., 1989. Operon structure and nucleotide homology of the chlorocatechol oxidation genes of plasmids pJP4 and pAC27. Gene 83, 225–232. Jessen-Marshall, A.E., Paul, N.J., Brooker, R.J., 1995. The conserved motif, GXXX(D/E) (R/K )XG[ X ](R/K ) (R/K ), in hydrophilic loop 2/3 of the lactose permease. J. Biol. Chem. 270, 16 251–16 257. Ka, J.O., Tiedje, J.M., 1994. Integration and excision of a 2,4-dichlorophenoxyacetic acid-degradative plasmid in Alcaligenes paradoxus and evidence of its natural intergenic transfer. J. Bacteriol. 176, 5284–5289. Ka, J.O., Holben, W.E., Tiedje, J.M., 1994a. Analysis of competition in soil among 2,4-dichlorophenoxyacetic acid-degrading bacteria. Microbiology 60, 1121–1128. Ka, J.O., Holben, W.E., Tiedje, J.M., 1994b. Use of gene probes to aid in recovery and identification of functionally dominant 2,4-dichlorophenoxyacetic acid-degrading populations in soil. Appl. Environ. Microbiol. 60, 1116–1120. Ka, J.O., Holben, W.E., Tiedje, J.M., 1994c. Genetic and phenotypic diversity of 2,4-dichlorophenoxyacetic acid (2,4-D)-degrading bacteria isolated from 2,4-D-treated soils. Appl. Environ. Microbiol. 60, 1106–1115. Kaphammer, B., Kukor, J.J., Olsen, R.H., 1990. Regulation of tfdCDEF by tfdR of the 2,4-dichlorophenoxyacetic acid degradation plasmid pJP4. J. Bacteriol. 172, 2280–2286. Kasberg, T., Daubaras, D.L., Chakrabarty, A.M., Kinzelt, D., Reineke, W., 1995. Evidence that operons tcb, tfd, and clc encode maleylacetate reductase, the fourth enzyme of the modified ortho pathway. J. Bacteriol. 177, 3885–3889. Ko˜iv, V., Marits, R., Heinaru, A., 1996. Sequence analysis of the 2,4-dichlorophenol hydroxylase gene tfdB and 3,5-dichlorocatechol 1,2-dioxygenase gene tfdC of 2,4-dichlorophenoxyacetic acid degrading plasmid pEST4011. Gene 174, 293–297. Leveau, J.H.J., van der Meer, J.R., 1996. The tfdR gene can successfully take over the role of the insertion element-inactivated TfdT protein as a transcriptional activator of the tfdCDEF gene cluster, which encodes chlorocatechol degradation in Ralstonia eutropha JMP134(pJP4). J. Bacteriol. 178, 6824–6832. Leveau, J.H., van der Meer, J.R., 1997. Characterization of insertion sequence ISJP4 on plasmid pJP4 from Ralstonia eutropha JMP134. Gene 202, 103–114. Leveau, J.H.J., Zehnder, A.J.B., van der Meer, J.R., 1998. The tfdK gene product facilitates uptake of 2,4-dichlorophenoxyacetate by Ralstonia eutropha JMP134(pJP4). J. Bacteriol. 180, 2237–2243.

Leveau, J.H.J., Ko¨nig, F., Fo¨chslin, H., Werlen, C., van der Meer, J.R., 1999. Dynamics of multigene expression during catabolic adaptation of Ralstonia eutropha JMP134 (pJP4) to the herbicide 2,4-dichlorophenoxyacetate. Mol. Microbiol. 33, 396–406. Loos, M.A., Schlosser, I.F., Mapham, W.R., 1979. Phenoxy herbicide degradation in soils: quantitive studies of 2,4-D- and MCBAdegrading microbial populations. Soil Biol. Biochem. 11, 377–385. Ma¨e, A., Marits, R., Ausmees, N., Ko˜iv, V., Heinaru, A., 1993. Characterization of a new 2,4-dichlorophenoxyacetic acid degrading plasmid pEST4011: physical map and localization of catabolic genes. J. Gen. Microbiol. 139, 3165–3170. Maltseva, O., McGowan, C., Fulthorpe, R., Oriel, P., 1996. Degradation of 2,4-dichlorophenoxyacetic acid by haloalcalophilic bacteria. Microbiology 142, 1115–1122. Matrubutham, U., Harker, A.R., 1994. Analysis of duplicated gene sequences associated with tfdR and tfdS in Alcaligenes eutrophus JMP134. J. Bacteriol. 176, 2348–2353. McGowan, C., Fulthorpe, R., Wright, A., Tiedje, J.M., 1998. Evidence for interspecies gene transfer in the evolution of 2,4-dichlorophenoxyacetic acid degraders. Appl. Environ. Microbiol 64, 4089–4092. Nakatsu, C., Ng, J., Singh, R., Straus, N., Wyndham, R.C., 1991. Chlorobenzoate catabolic transposon Tn5271 is a composite class I element with flanking class II insertion sequences. Proc. Natl. Acad. Sci. U. S. A. 88, 8312–8316. Ogawa, N., Miyashita, K., 1999. The chlorocatechol-catabolic transposon Tn5707 of Alcaligenes eutrophus NH9, carrying a gene cluster highly homologous to that in the 1,2,4-trichlorobenzene-degrading bacterium Pseudomonas sp. strain P51, confers the ability to grow on 3-chlorobenzoate. Appl. Environ. Microbiol. 65, 724–731. Ou, L.T., 1984. 2,4-D degradation and 2,4-D degradation microorganisms in soils. Soil Sci. 137, 100–107. Perkins, E.J., Gordon, M.P., Caceres, O., Lurquin, P.F., 1990. Organization and sequence analysis of the 2,4-dichlorophenol hydroxylase and dichlorocatechol oxidative operons of plasmid pJP4. J. Bacteriol. 172, 2351–2359. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: a Laboratory Manual. second edition, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Sanger, F., Nicklen, S., Coulson, A.R., 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. U. S. A. 74, 5463–5467. Seibert, V., Kourbatova, E.M., Golovleva, L.A., Schlo¨mann, M., 1998. Characterization of the maleylacetate reductase MacA of Rhodococcus opacus 1CP and evidence for the presence of an isofunctional enzyme. J. Bacteriol. 180, 3503–3508. Streber, W.R., Timmis, K.N., Zenk, M.H., 1987. Analysis, cloning, and high-level expression of 2,4-dichlorophenoxyacetate monooxygenase gene tfdA of Alcaligenes eutrophus JMP134. J. Bacteriol. 169, 2950–2955. Top, E.M., Holben, W.E., Forney, L.J., 1995. Characterization of diverse 2,4-dichlorophenoxyacetic acid-degradative plasmids isolated from soil by complementation. Appl. Environ. Microbiol. 61, 1691–1698. Van der Meer, J.R., Eggen, R.I.L., Zehnder, A.J.B., Devos, W.M., 1991. Sequence analysis of the Pseudomonas sp. strain P51 tcb gene cluster, which encodes metabolism of chlorinated catechols: evidence for specialization of catechol 1,2-dioxygenases for chlorinated catechols substrates. J. Bacteriol. 173, 2425–2434. Vedler, E., Ko˜iv, V., Heinaru, A., 2000. TfdR, the LysR-type transcriptional activator, is responsible for the activation of the tfdCB operon of Pseudomonas putida 2,4-dichlorophenoxyacetic acid degradative plasmid pEST4011. Gene 245, 161–168.