|Home | About | Journals | Submit | Contact Us | Français|
Broad-host-range catabolic plasmids play an important role in bacterial degradation of man-made compounds. To gain insight into the role of these plasmids in chloroaniline degradation, we determined the first complete nucleotide sequences of an IncP-1 chloroaniline degradation plasmid, pWDL7::rfp and its close relative pNB8c, as well as the expression pattern, function, and bioaugmentation potential of the putative 3-chloroaniline (3-CA) oxidation genes. Based on phylogenetic analysis of backbone proteins, both plasmids are members of a distinct clade within the IncP-1β subgroup. The plasmids are almost identical, but whereas pWDL7::rfp carries a duplicate inverted catabolic transposon, Tn6063, containing a putative 3-CA oxidation gene cluster, dcaQTA1A2BR, pNB8c contains only a single copy of the transposon. No genes for an aromatic ring cleavage pathway were detected on either plasmid, suggesting that only the upper 3-CA degradation pathway was present. The dcaA1A2B gene products expressed from a high-copy-number vector were shown to convert 3-CA to 4-chlorocatechol in Escherichia coli. Slight differences in the dca promoter region between the plasmids and lack of induction of transcription of the pNB8c dca genes by 3-CA may explain previous findings that pNB8C does not confer 3-CA transformation. Bioaugmentation of activated sludge with pWDL7::rfp accelerated removal of 3-CA, but only in the presence of an additional carbon source. Successful bioaugmentation requires complementation of the upper pathway genes with chlorocatechol cleavage genes in indigenous bacteria. The genome sequences of these plasmids thus help explain the molecular basis of their catabolic activities.
Aniline and its derivatives are used industrially in the production of pesticides, varnishes, photographic chemicals, rubber, azo dyes, and polyurethanes, and they also accumulate in the environment as a result of the microbial degradation of herbicides (11). Anilines, especially chloroanilines, are toxic and carcinogenic. There are reports of a few bacterial strains that are capable of degrading mono- and dichloroanilines (35, 64); however, the compounds are recalcitrant to degradation by the vast majority of bacteria (57). The general pathway for chloroaniline degradation is thought to follow the typical aerobic biodegradation pathway for chlorinated aromatic compounds. First, a peripheral (upper) pathway generates chlorocatechol by oxidative deamination. Then, modified ortho- or meta-cleavage (lower) pathways convert chlorocatechol to tricarboxylic acid (TCA) cycle intermediates (22, 25, 48). The accumulation of upper-pathway intermediates has been observed in some bacterial strains (38). Gene clusters involved in chloroaniline degradation have been found on plasmids as well as on bacterial chromosomes (7, 8, 14, 26).
Plasmid-mediated horizontal transfer of catabolic genes contributes to the ability of bacterial communities to degrade toxic man-made compounds, and this natural process can be exploited in bioaugmentation approaches (60, 56). While we have gained insight in recent decades into the genetic diversity and function of catabolic plasmids and their role in biodegradation of xenobiotic compounds, relatively few such plasmids have been rigorously analyzed and compared. In order to understand the evolution of these mobile elements and their contribution to the removal of toxic compounds in our environment, complete plasmid sequences are needed, combined with information on the biological significance of their gene products.
Several bacterial strains isolated for their ability to degrade man-made, chlorinated organic compounds carry genes encoding the relevant degradation pathways on plasmids of the incompatibility group IncP-1, which transfer to and replicate in a broad range of hosts (55). A few of these plasmids have recently been completely sequenced: pJP4, pEST4011 [2,4-dichlorophenoxyacetic acid degradation (58, 61)], pUO1 [haloacetate degradation (44)], pADP-1 [atrazine degradation (33)], pA81 [chlorobenzoate degradation (23)], and pCNB1 [4-chloronitrobenzene degradation (32)]; details regarding most of these plasmid sequences are also summarized in a recent review (60). While several catabolic plasmids that encode (partial) (chloro)aniline degradation pathways have been described, such as pCIT1 (2, 34), pTDN1 (18, 40), pYA1 (17), pNB1, pNB2, pNB8c, pC1 (7), and pWDL7 and pTB30 (14), their complete genome sequences have not yet been reported. Some of these plasmids were shown to transfer the ability to use 3-chloroaniline (3-CA) as the sole nitrogen or occasionally the sole carbon source to other hosts (3, 6, 7, 14). However, it is not clear if all of the genes necessary for complete degradation are located on the plasmids.
In this study, we describe the complete genome sequences of plasmids involved in 3-CA and aniline degradation: pWDL7::rfp, an 86.6-kb derivative of plasmid pWDL7 from Comamonas testosteroni strain WDL7, and pNB8c, a 60.4-kb plasmid from Delftia acidovorans strain B8c. These strains were obtained from an orchard soil with a 10-year history of treatment with a mixture of linuron, diuron, and simazine, and from activated sludge of a wastewater treatment plant, respectively (7, 13). Strains that carry plasmid pWDL7, C. testosteroni WDL7 and Cupriavidus pinatubonensis JMP228n (formerly Cupriavidus necator), were shown to completely degrade 3-CA, suggesting that the plasmid encoded pathways for at least partial and possibly complete degradation of this recalcitrant and toxic compound (13, 14). In contrast, plasmid pNB8c conferred aniline but not chloroaniline transformation in C. pinatubonensis JMP228gfp (7). Based on sequence and hybridization results, both plasmids had been assigned to the β subgroup of the IncP-1 plasmids (7, 14).
Our specific objectives were to (i) determine, analyze, and compare the complete sequences of a derivative of plasmid pWDL7 labeled with the red fluorescent protein, designated pWDL7::rfp, and its close relative, plasmid pNB8c, and infer their phylogenetic relationship to other plasmids; (ii) identify the chloroaniline catabolic genes and determine their transcriptional activity and function; and (iii) determine the utility of pWDL7::rfp as a tool to augment the catabolic capabilities of activated sludge exposed to 3-CA. The motivation for sequencing the rfp-marked version of pWDL7 and not the wild type is its utility in bioaugmentation studies due to the presence of fluorescent and antibiotic resistance markers.
The bacterial strains and plasmids used are listed in Table 1. Activated sludge used in batch mating studies was obtained from the University of California, Davis, wastewater treatment plant (Davis, CA).
Strains were cultured in Luria-Bertani medium (LB; Miller or Lennox medium ) with appropriate antibiotics at 30°C or 37°C. In experiments to test 3-chloroaniline (3-CA) utilization as the sole carbon source, minimal medium without additional carbon (MMO) was used (47). In MMO, 3-CA was the sole carbon source and ammonium (2.5 mM (NH4)2SO4) was available as a nitrogen source. In experiments to determine if 3-CA was used as a nitrogen source, minimal medium without nitrogen (MMN) was used, in the presence or absence of an additional carbon source such as pyruvate (7). The media were prepared as previously described (14).
The plasmid pWDL7 was tagged with the mini-Tn5 transposon from plasmid pTTN151 (52) as previously described (54), and the resulting plasmid was designated pWDL7::rfp. Tagged plasmids were transferred to C. pinatubonensis JMP228n. Red-fluorescing JMP228n(pWDL7::rfp) colonies were selected on MMN agar plates containing pyruvate (2 g/liter), 3-CA (100 mg/liter), kanamycin (Km; 50 mg/liter), and nalidixic acid (Nal; 100 mg/liter). Plasmid pWDL7::rfp was then transferred in a plate mating (see below) from a purified clone of JMP228n(pWDL7::rfp) to Pseudomonas putida UWC3, followed by selection on LB medium with rifampin (Rif; 50 mg/liter) and Km (50 mg/liter).
To obtain sufficient plasmid DNA of high quality for sequence determination by shotgun cloning, plasmids were first transferred by conjugation from JMP228n(pWDL7::rfp) and B8c to E. coli K12Nal, a nalidixic acid-resistant mutant of MG1655. In the case of pNB8c, transconjugants were not selected for a specific phenotype but were found by randomly choosing recipient colonies after the mating, followed by plasmid DNA isolation. E. coli transconjugants were minimally processed to avoid mutations in the plasmids. Plasmid DNA to be used for sequence determination and additional analyses was isolated from E. coli using a plasmid midi-kit (Qiagen, Valencia, CA) according to the manufacturer's instructions for low-copy-number plasmids. Standard methods were used for restriction analysis and pulse field gel electrophoresis (PFGE) (41). Restriction endonucleases and DNA modification enzymes were purchased from New England BioLabs (Ipswich, MA) and Fermentas Inc. (Glen Burnie, MD). DNA fragments were purified with a QIAquick gel extraction kit (Qiagen).
The genome sequences of plasmids pWDL7::rfp and pNB8c were determined by shotgun sequencing at the DOE Joint Genome Institute (Walnut Creek, CA), with 10× coverage for each plasmid. Regions of poor sequence quality were resequenced at the University of Idaho by PCR amplification followed by sequencing using a BigDye Terminator v.3.1 cycle sequencing kit and a 3730 DNA analyzer (Applied Biosystems, Carlsbad, CA). The sequence was automatically annotated by the J. Craig Venter Institute Annotation Service (http://www.jcvi.org/annotation/service/) and further annotated manually using Manatee (manatee.sourceforge.net).
Similarity searches were performed using BLAST (1). Plasmid map and alignment figures included in this work were generated using GENtle software (Magnus Manske, University of Cologne, Germany). Plasmid alignments were performed using Mauve (12) and then refined and rendered graphically using TRAPPIST, a Python-based sequence map and alignment drawing tool for plasmids (G. Van der Auwera, unpublished data), with identity scoring by ClustalW. The pdca promoter region was analyzed using BPROM (Softberry Inc., Mount Kisco, NY).
To infer the phylogenetic relationships of various IncP-1 plasmids, the deduced amino acid sequences of 24 genes from the 26 plasmids were aligned individually using MUSCLE (15) and then concatenated into a single alignment. RAxML-VI-HPC was used to infer a maximum likelihood phylogeny for the aligned sequences (46). The phylogeny with the largest likelihood was chosen from 100 iterations that altered the starting tree. The phylogeny inferred by this method was not substantially different from those inferred using maximum parsimony and neighbor joining. Proteins used were KfrA, KfrB, KlcA, KleE, KorA, KorB, KorC, TraD, TraE, TraF, TraG, TraI, TraJ, TraK, TraL, TrbA, TrbB, TrbC, TrbD, TrbF, TrbG, TrbI, TrbJ, and TrbK. The plasmids used in the analysis (and their accession numbers) were pR751 (NC_001735), pADP1 (NC_004956), pB4 (NC_003430), pA81 (NC_006830), pUO1 (NC_005088), pB10 (NC_004840), RK2 (NC_001621), pB3 (NC_006388), pA1 (NC_007353), pCNB1 (NC_010935), pJP4 (NC_005912), pAKD4 (GQ983559), pSP21 (CP002153), pBS228 (NC_008357), pBP136 (NC_008459), pTP6 (NC_007680), pAOVO02 (NC_008766), pTB11 (NC_006352), pB8 (NC_007502), pB11 (CP002152), pB5 (CP002151), pKJK5 (NC_008272), QKH54 (NC_008055), pAMMD1 (NC_008385), pWDL7::rfp (GQ495894), and pNB8c (JF274990).
To amplify fragments carrying dcaQT and dcaA1A2B, possibly encoding chloroaniline oxidation, total genomic DNAs from P. putida UWC3(pWDL7::rfp) and E. coli(pNB8c) were used as the templates in PCRs. The primers cloneADOF (5′-GCGGCGCTCGAGGTGACCGAATTGCGAGAAAACGATGAAAG-3′) and cloneADOR (5′-GCATGCTTTCATCGATGGCTTCAGGC-3′) were used to amplify the fragment carrying dcaA1A2B. The resulting PCR product from pWDL7::rfp was ligated to pBBR1MCS-2 (27) after digestion with XhoI and ClaI, forming pJTP500 (Table 1). Since the sequences of the dcaA1A2B genes from pWDL7::rfp and pNB8c were identical (see Results), a dcaA1A2B clone from pNB8c was not needed. The HindIII-KpnI fragment containing the dcaA1A2B genes from pJTP500 was ligated with HindIII-KpnI-digested pUC18 (41), resulting in pJTP510. The primers cloneAA2-F (5′-GCGGCGAATTCGGTACCGTGGCAGATGGTTGGGTAATTCG-3′) and cloneAA-R (5′-GGCGCCAAGCTTCTCGAGAAATCGCGGCTTCAGTTCATGTTGTTTCGG-3′) were used to amplify a fragment carrying dcaQT. After restriction digestion with EcoRI and HindIII, the resulting product was ligated with pJPC13 (37), forming pJTP800 and pJTP820 for pWDL7::rfp and pNB8c, respectively. All plasmid inserts were verified by DNA sequencing using a BigDye Terminator v.3.1 cycle sequencing kit and a 3730 DNA analyzer (Applied Biosystems, Carlsbad, CA). The ability of E. coli DH5α to convert 3-CA to chlorocatechol when containing these cloned dca genes was then tested using whole-cell biotransformation reactions with mixtures containing 3-CA (0.2 g/liter). The culture supernatant was extracted with NaOH-washed ethyl acetate, and the extract was analyzed by gas chromatography-mass spectrometry (GC-MS) as described previously (31, 39). Products were identified by comparison to chemical standards.
To analyze the promoter activity of the region upstream of gene dcaQ, two approaches were used. First, for a GFP reporter gene fusion assay the putative promoter region, i.e., the 307-bp and 289-bp tnpA-dcaQ intergenic fragments from plasmid pWDL7::rfp and pNB8c, respectively, were amplified with the primers pDcaQF (5′-aaagaattcTTGAACGCTCACTCATGTGC-3′) and pDcaQR (5′-aaaggtaccGCATAGATGCCGGTTGTTTAG-3′), which have EcoRI and KpnI sites at their 5′ ends (lowercase). Amplicons were cleaved with EcoRI-KpnI and cloned in front of the promoterless gfp cassette in the pG-GFP promoter probe reporter plasmid (28) (Table 1). E. coli DH5α cells with the respective cloned fragments or vector only, were grown overnight in LB medium supplemented with 100 mg/liter ampicillin (Amp), transferred into fresh medium and grown to an optical density at 600 nm (OD600) of 0.1 to 0.7. The promoter activity of these cultures was measured using a Modulus single-tube multimode reader (Promega Corp., Sunnyvale, CA), and the data are presented as fluorescence units/OD600 unit (FU). Data are mean values from 2 to 4 different clones (3 independent measurements each).
To determine transcriptional activity of the dca genes in C. pinatubonensis JMP228 upon induction with 3-CA, both plasmids pWDL7::rfp and pNB8c were transferred into C. pinatubonensis JMP228. For total RNA isolation, cultures grown overnight in LB were diluted 1:20 into fresh LB medium and grown for 3 h at 30°C with or without 50 mg/liter 3-CA. Culture volumes containing similar cell numbers based on OD600 readings were stabilized with RNAprotect bacterial reagent (Qiagen, Valencia, CA), and RNA was isolated using an RNeasy minikit (Qiagen, Valencia, CA) according to the manufacturer's instructions. The purified RNA was then treated with Turbo DNA-free kit (Applied Biosystems, Carlsbad, CA) and reverse-transcribed using a high-capacity cDNA reverse transcription kit with RNase inhibitor (Applied Biosystems, Carlsbad, CA) both according to the manufacturers' instructions. PCR amplification of a segment of the dcaA1 gene was carried out on cDNA, using Phusion Taq DNA polymerase (New England BioLabs, Ipswich, MA). The volumes of cDNA template solutions were normalized based on the levels of RNA transcribed, as determined with a Nanodrop ND-1000 spectrophotometer (Thermo Fisher, Wilmington, DE). The primers used were NB8c24880 (5′-ATGAATGGGACGACCTGGTG-3′) and NB8c25793r (5′-AAAGGATCCCAGATTGGGAAAGATGTTGAGG-3′), and the annealing temperature was 65°C. Amplicons were separated on 0.75% agarose gels and visualized with a GelDoc imaging system, and the intensity of bands was quantified using Quantity One software (Bio-Rad Laboratories, Hercules, CA).
P. putida UWC3(pWDL7::rfp) was used as the donor in all plasmid transfer experiments because it is Ilv− and does not grow in MMO. Plate matings were performed essentially as described previously (29). Briefly, mixtures of an overnight-grown donor culture of P. putida UWC3(pWDL7::rfp) and either activated sludge or an overnight grown recipient culture were placed on LB agar, allowing conjugation to occur for 12 h. Subsequently, the biomass of these mating mixtures and of negative controls (separate donor and recipient cultures) was scraped from the agar surface, and the cells were washed three times by centrifugation and resuspension in phosphate-buffered saline (PBS) to remove agar and nutrients. The mating mixtures were used to study the effect of bioaugmentation because it was not possible to isolate and stably maintain transconjugants capable of growing on 3-CA as the sole carbon source. The washed cell pellets were then resuspended in 1 ml PBS, and added to 125 ml screw-cap Erlenmeyer flasks with 25 ml of MMO, containing either 0.16 mM (20 mg/liter) or 0.39 mM (50 mg/liter) of 3-CA as the sole carbon source. Cell densities were adjusted to equal optical densities, and flasks were incubated on a rotary shaker (100 rpm) at 30°C. One-milliliter samples were removed at 24-h intervals and were prepared for high-pressure liquid chromatography (HPLC) analysis of 3-CA concentrations by centrifugation at 10,000 × g for 10 min. A mass balance of 3-CA, tested using autoclaved cells, revealed that 3-CA was in the supernatant (data not shown). Samples were analyzed using reverse-phase HPLC (Agilent Technologies, Santa Clara, CA) with a Phenomenex Prodigy 5 μm ODS-2 octyldecyl silane column (100 mm by 2 mm) using isocratic 50:50 water-methanol as the solvent at a flow rate of 0.25 ml/min and with detection of peaks at 210, 254, and 290 nm, essentially as previously described (16).
The sequences of pWDL7::rfp and pNB8c have been deposited in the GenBank database under accession numbers GQ495894 and JF274990, respectively.
Determination of the complete nucleotide sequence of pWDL7::rfp required a combination of approaches due to its unique and unexpected genetic structure. The initially assembled sequence obtained by shotgun sequencing was not consistent with the plasmid restriction fragment profiles and the observed size of uncut plasmid DNA separated by pulsed-field gel electrophoresis (PFGE) (data not shown). Results obtained by PCR also confirmed the presence of a repeated region (data not shown). By using these data, the complete genome sequence of plasmid pWDL7::rfp was reassembled (Fig. 1A). The plasmid is 86,647 bp long and contains two 22.4-kb inverted repeats, named transposon Tn6063, which are flanked by inverted IS1071 sequences. The artificial mini-Tn5 transposon used to mark the plasmid with red fluorescent protein and kanamycin resistance was found inserted into Tn6063 (Fig. 1A). The plasmid genome contains 88 assigned open reading frames (ORFs). The mean G+C content is 63 mol%, but that of the backbone, which consists of plasmid replication, maintenance/control, and transfer genes, is slightly higher (66%).
Plasmid pNB8c is very similar to pWDL7::rfp but much smaller (60,421 bp). The automatic sequence assembly in this case was consistent with the predicted size and restriction profiles. It has an overall G+C content of 65%, which is very similar to the estimated G+C content of the host strain D. acidovorans B8c (66.6%) (7). The typical IncP-1 plasmid backbone of pNB8c is interrupted by an 18,915-bp segment, which was named Tn6063′ because of its high similarity to Tn6063 on pWDL7::rfp (Fig. 1A). Sixty-four ORFs were assigned (Fig. 1A).
Plasmids pWDL7::rfp and pNB8c have backbone regions that are typical of plasmids of the incompatibility group IncP-1. The region contains 45 genes in clusters responsible for conjugative transfer, initiation of vegetative replication, stable inheritance, and regulatory networks for these functions (Fig. 1A). Three of these, upf30.5, encoding a putative outer membrane protein, upf31.7, encoding a site-specific methylase, and relE (stbE), a likely part of the plasmid stability system, are unique to backbones of the plasmid subgroup IncP-1β (42, 51, 58). The backbone regions are most similar to those of the IncP-1β plasmids pA81 from Achromobacter xylosoxidans A8 (accession no. AJ515144), and pCNB1 from Comamonas sp. CNB-1 (accession no. EF079106) (24, 32), both of which also encode proteins involved in catabolic functions (Fig. 2).
Phylogenetic analysis using 24 concatenated protein sequences encoded by IncP-1 plasmids confirmed that pWDL7::rfp and pNB8c belong to the β subgroup but form a distinct clade (Fig. 3). Other members of this clade are the catabolic plasmids pA81 and pCNB1 mentioned above, the cryptic plasmid pA1 of Sphingomonas sp. A1 (19), the Acidovorax sp. JS42 plasmid pAOVO02 (GenBank accession no. NC_008766), and the exogenously isolated multiresistance plasmid pB4 from an unknown host (50). This new clade has been named IncP-1β-2 by Norberg et al. (36). Very recently, plasmid pAKD26, which encodes mercury resistance and carries putative 2,4-dichlorophenoyacetic acid and catechol degradation (mocp) genes, was also shown to be a member (not shown in Fig. 3) (43). Thus, plasmids pWDL7::rfp and pNB8c have a backbone that is divergent from that of the archetypal IncP-1β plasmid R751 but very similar to that of other catabolic plasmids in the IncP-1β-2 clade.
The unusual structure of plasmid pWDL7::rfp raised questions about the orientation of the two backbone regions. Because the two Tn6063 transposons are identical, it was not possible to establish the final orientation of the backbone regions simply based on the assembled DNA sequence. Therefore, pWDL7::rfp DNA was digested with the restriction enzymes PsiI and SnaBI, and the resulting fragments were separated by PFGE. In the pWDL7::rfp backbone sequence, a single SnaBI site is located between upf30.5 and upf31.7 (at 15,699 bp), and a unique PsiI site is located at 63,156 bp and precedes relE (stbE) (Fig. 4B). The resulting restriction fragments in this case should be 39,189 and 47,458 bp long (form A). If the backbone regions were inverted (form B), different DNA fragments should be observed (20,932 and 65,715 bp). Surprisingly, PFGE yielded four DNA fragments in identical molar ratio, and their sizes corresponded to the fragments expected when both plasmid forms, A and B, are present (Fig. 4C). This means that the E. coli strain from which the plasmid DNA was purified before sequence determination contains both forms A and B of pWDL7::rfp. Thus, the sequence assembled and deposited in the GenBank database represents only one of these two orientations (form A).
Most IncP-1 plasmids described to date carry catabolic genes, mercury resistance determinants, or multidrug resistance determinants, which are usually located between the tra and trb regions (site 1) and/or between oriV and trfA (site 2). Surprisingly, in pWDL7::rfp, both sites 1 and 2 were found to be occupied by identical copies of transposon Tn6063, inserted in opposite orientations (Fig. 1A). The two Tn6063 insertion sites did not share any sequence similarity, but the copy located at site 2 is flanked by a pair of 5-bp direct repeats in form A of the plasmid (5′-TCGAT-3′), while the copy at site 1 is not. In pNB8c, transposon Tn6063′ is inserted in site 1. It is almost identical to Tn6063 on pWDL7::rfp except for slight differences in the dca operon sequence (see below). This suggests that the first transposition of Tn6063 into the pWDL7::rfp plasmid occurred at site 1, followed by recent insertion into site 2 (Fig. 4A).
Tn6063 has a complex structure composed of two insertion sequences (IS1071 and IS1071a) flanking a region that contains two other insertion sequences, IS21 and IS66, as well as an operon of catabolic genes, named dca by homology to recently identified dca genes from Variovorax sp. (5, 8). Moreover, the mini-Tn5 element that was used to mark pWDL7::rfp was found within the 3′ end of the second IS1071-like transposase gene. The transposases of the flanking IS1071 and IS1071a elements share 78% identity (88% similarity) without gaps at the protein level. The IS1071 element has two 108-bp long inverted repeat sequences (IR3L and IR3R), while IS1071a seems to contain only the left-side long inverted repeat (IRL; 110 bp) but not the corresponding right inverted repeat (IRR) (Fig. 1A). All three IRs show high sequence similarity to each other (Fig. 1B).
The operon that is likely involved in the degradation of chloroaniline (dca) by encoding a peripheral (upper) degradation pathway, contains six genes. Due to the extremely high DNA sequence similarity with the recently annotated dca genes of the linuron degrader Variovorax sp. SRS16, shown to be induced by 3,4-dichloroaniline (3,4-DCA) (5), we also named the operon on pWDL7:rfp and NB8c dca. It was previously shown that aniline dioxygenases consist of multiple components: proteins homologous to glutamine synthetase, glutamine amidotransferase, large and small subunits of the terminal dioxygenase, and a reductase (18, 30). In P. putida UCC22 and Delftia tsuruhatensis AD9, where these systems were first described, these components are encoded by, respectively, tdnQ-tadQ, tdnT-tadT, tdnA1-tadA1, tdnA2-tadA2, and tdnB-tadB (18, 30). The genes dcaQ, -T, -A1, -A2, -B, and -R on pWDL7::rfp and pNB8c show high sequence similarity to genes from all previously described systems, which together are thought to code for conversion of (chloro)aniline to (chloro)catechols (5, 17, 18, 35, 59). No homologs of aromatic ring cleavage pathway genes were identified on either pWDL7::rfp or pNB8c.
In spite of the presence of almost identical copies of dca genes on both plasmid pWDL7::rfp and pNB8c, clear differences in catabolic activity had been previously observed. While pWDL7 was shown to confer 3-CA and 3,4-DCA transformation after conjugative transfer in C. pinatubonensis JMP228n (14), pNB8c conferred only the ability to transform nonchlorinated aniline in an isogenic strain, even though its original host, NB8c, degraded 3-CA (7). To determine the molecular basis of these differences in catabolic activity between the two plasmids, the dca genes and their promoter regions, as well as their functions, were compared in detail. Discrete differences were revealed in two areas of the dca operons: the dcaQT genes and the dca promoter region (Fig. 5A and B).
First, on pNB8c, three single-nucleotide polymorphisms (SNPs) in the dcaQ gene caused amino acid substitutions, and a single G-to-A transition in the dcaT start codon (ATG) moved the start codon 75 bp downstream compared to the start site in pWDL7::rfp. These changes may affect the activity of the DcaQ and DcaT proteins, even though the type I glutamine amidotransferase (GATase 1) catalytic triad of the latter remained unchanged. Therefore, our objectives were to demonstrate (i) that the proteins encoded by the dca operons on pWDL7::rfp and pNB8c were functional and sufficient to allow transformation of 3-CA into 4-chlorocatechol (4-CC) and (ii) that the effect of the SNPs in dcaQT did not affect that function. The dcaQT genes from both plasmids and the dcaA1A2B from pWDL7::rfp were cloned separately into two vectors downstream of the plac promoter and their activity was measured in E. coli (Fig. 5B; Table 1). Since the dcaA1A2B sequences from both plasmids are identical, the dca genes on pJTP510 from pWDL7::rfp are representative of both. The E. coli DH5α strains carrying either plasmids pJTP800 and pJTP510 (expressing dcaQT and dcaA1A2B from pWDL7::rfp) or plasmids pJTP820 and pJTP510 (expressing dcaQT from pNB8c and dcaA1A2B from pWDL7::rfp/pNB8c) were both able to convert 3-CA to 4-CC, as judged by GC-MS analysis, with no observable differences. In contrast, E. coli carrying only the dcaQT genes from either pWDL7::rfp or pNB8c did not produce any 4-CC from 3-CA. However, the cloned dcaA1A2B genes alone [in E. coli DH5α(pJTP510)] conferred chloroaniline dioxygenase activity, as judged by the formation of 4-CC (Fig. 5B). These results indicate that in E. coli, the dcaA1A2B genes are sufficient for conversion of 3-CA into 4-CC and that the slight differences in the dcaQT coding regions of the two plasmids do not explain the drastic differences in the ability of pWDL7::rfp and pNB8c to confer this activity.
Second, compared to the sequence of pWDL7::rfp, deletions were observed in a 20-bp segment of pNB8c 78 bp upstream of the dcaQ start codon (Fig. 5A). In silico analysis revealed the presence of a hypothetical E. coli σ70-like promoter, and the mutations in pNB8c completely destroyed its −10 box (Fig. 5A). To analyze the effect of these deletions, we used two approaches. In a first approach the dca promoters from both plasmids were amplified and cloned in front of a promoterless gfp gene (Table 1). While the pWDL7::rfp fragment showed strong promoter activity (dcaQ, 30,557 ± 13,276 FU), the activity of the pNB8c fragment was as low as that of the control E. coli strain with the insert-free vector (~700 FU) (Fig. 5A). As these results were obtained in an E. coli host, they might not reflect promoter activity in C. pinatubonensis JMP228, in which catabolic activity differences were previously observed. Therefore, we compared the transcriptional activity of the dca operons from both plasmids in that host in the presence and absence of 3-CA. Repeated assays consistently showed that low levels of dcaA1 transcript were formed by both JMP228(pWDL7::rfp) and JMP228(pNB8c) grown in the absence of 3-CA. Yet when the cultures were grown with 3-CA for several hours before RNA was harvested, JMP228(pWDL7::rfp) showed a consistent increase in transcript levels compared to uninduced cultures, but JMP228(pNB8c) did not (data not shown). These results suggest that the modified segment of the pNB8c pdca promoter region was insufficient for regulation of the dca operon by 3-CA and thus likely explains the previously shown inability of pNB8c to confer 3-CA conversion in JMP228gfp (7).
Since plasmid pWDL7::rfp carries two copies of a fully functional dca gene cluster, it could be useful in bioaugmentation projects to accelerate removal of 3-CA from polluted waters or soils by transferring the genetic information to indigenous bacteria (56). The apparent lack of expression of the dca genes on pNB8c in some strains other than its native host B8c (7) makes this plasmid less useful for such applications, and it was therefore not analyzed further. We tested the ability of pWDL7::rfp to transfer 3-CA degradation capacity from P. putida UWC3 to a mixed bacterial community from the activated sludge of a wastewater treatment plant in plate matings. In parallel to the activated sludge communities, pure cultures of Delftia acidovorans ATCC 15668 and Comamonas testosteroni ATCC 11996, which are unable to degrade 3-CA, were used as control recipients for the plasmid, because other strains of these species have been reported to be involved in 3-CA degradation (6, 7, 9, 13, 22, 49, 63).
Mating mixtures were used to study the effect of bioaugmentation because it was not possible to isolate and stably maintain transconjugants capable of growing on 3-CA as the sole carbon source. When activated sludge samples bioaugmented with P. putida UWC3(pWDL7::rfp) were suspended in MMO with 0.16 mM (Fig. 6) or 0.39 mM 3-CA (data not shown) as the sole carbon source, little degradation occurred over the first 5 days. However, as soon as pyruvate (0.227 mM) was added as an additional carbon source on day 7, 3-CA was rapidly removed, with no 3-CA remaining after 9 days (Fig. 6). Bioaugmented cultures of D. acidovorans ATCC 15668 removed 100% of 3-CA after 9 days (Fig. 6). In contrast, mixtures with C. testosteroni ATCC 11996 as the recipient did not show any 3-CA degradation (data not shown). The uninoculated controls showed little 3-CA degradation until day 11, after which the 3-CA concentration rapidly decreased in all except the abiotic control. The donor strain P. putida UWC3 is an auxotrophic variant of KT2442 and is therefore incapable of growing with 3-CA as a sole carbon and energy source in mineral salts medium. Therefore, accelerated 3-CA degradation activity in mixtures of this donor with activated sludge communities and the other strains almost certainly must be attributed to recipients that acquired and expressed the plasmid-encoded degradation genes. Plasmid transfer was not monitored in these experiments, nor was the possible accumulation of degradation products. Overall, the addition of a donor carrying pWDL7::rfp to activated sludge decreased the lag time for 3-CA removal by 5 to 7 days. Such a time “saving” could be critical in bioaugmentation applications. Thus, providing the genes encoding the upper pathway for 3-CA degradation to a wastewater treatment reactor may greatly accelerate partial or complete chloroaniline removal, depending on the composition of the activated sludge community and the presence of a readily available carbon source.
We present the complete genome sequences of two plasmids, pWDL7::rfp and pNB8c, which contain genes involved in the degradation of 3-CA. These plasmids belong to the same clade within the IncP-1β plasmid group and have a broad host range, as they are able to transfer between beta- and gammaproteobacteria. Small genetic differences in the accessory regions of both plasmids can explain their different abilities to confer chloroaniline degradation in strains other than their native hosts (7). We showed that the catabolic gene products of both plasmids were able to convert 3-CA to 4-CC but that only the dca genes on pWDL7::rfp and not those on pNB8c were induced by 3-CA, likely due to changes in the pNB8c dca promoter region. Plasmid pWDL7::rfp was also shown to have potential as a bioaugmentation agent to augment microbial communities of wastewater treatment plants with the catabolic functions needed to convert chloroanilines.
An intriguing aspect of pWDL7::rfp is the presence of two exact copies of Tn6063, one in each of the two expected accessory gene regions of the plasmid, while only one copy of a very similar transposon is present on pNB8c. Duplication of Tn6063 may have been caused by intramolecular replicative transposition (Fig. 4). This mechanism can create two inversely oriented copies of a replicative transposon in one circular DNA molecule. It is accompanied by the inversion of the DNA segment between the target site of the transposon and the site where the transposon was originally located (10). In plasmid pWDL7, such an event could have been mediated by the IS1071-type transposase acting on the inverted repeats IRL and IR3L. Therefore, it is plausible that plasmid form B (Fig. 4B) was generated first, when Tn6063 transposed, and that form A was subsequently generated by homologous recombination between the duplicated transposon regions. Sota et al. (45) showed that the transposition of IS1071 produces a 5-bp target site duplication. The presence of 5-bp direct repeats flanking one of the Tn6063 copies strongly suggests the occurrence of replicative transposition. Based on the characteristics of IS1071, which can actively transpose only in specific bacterial species, it is highly unlikely that the Tn6063 duplication occurred in E. coli (45). To test this hypothesis, we confirmed the presence of two copies of Tn6063 in the original plasmid pWDL7 using a set of PCR primers specific for both insertion sites. The numbers and sizes of obtained amplicons were identical in the cases of the original and marked plasmids (data not shown). Moreover, data obtained from an unfinished C. testosteroni WDL7 genome sequencing project also showed the presence of two copies of Tn6063 in pWDL7 (P. Albers and D. Springael, personal communication). The presence of two identical catabolic transposons on one plasmid seems rare but illustrates the potential involvement of transposable elements in gene duplication.
We showed that the dcaA1A2B genes alone are sufficient for chloroaniline dioxygenase activity. These results are different from those of studies of the aniline dioxygenase from P. putida UCC22(pTDN1) and Frateuria sp. ANA18, which showed that strains of E. coli required all of the cloned tdnQ, -T, -A1, -A2, and -B genes for full aniline dioxygenase activity (18, 35). However, we used a very sensitive assay for the detection of 4-CC and do not know if, in the natural host, sufficient conversion of 3-CA would occur in the absence of dcaQT to allow growth. Further research is also necessary to determine if the presence of a second copy of Tn6063 on pWDL7::rfp significantly increases the enzyme concentration in the cell and therefore the rate of conversion of 3-CA to 4-CC, compared to a plasmid with a single Tn6063 copy like pNB8c.
The sequence difference in the dca promoter region on plasmid pNB8c compared to pWDL7::rfp was associated with lack of induction of dca transcription in the host strain C. pinatubonensis JMP228 compared to the strain carrying pWDL7::rfp. These data can explain the previously observed inability of plasmid pNB8c to confer metabolism of 3-CA (used as the sole nitrogen source) in C. pinatubonensis JMP228gfp, while its native host was able to degrade 3-CA (7). The mechanism by which the native host of pNB8c effectively metabolized 3-CA is not known. This finding shows how subtle genetic differences in catabolic operons can drastically affect the degradation phenotypes conferred by two very similar plasmids.
Previous results suggested that the native plasmid pWDL7 conferred complete mineralization of 3-CA to strain JMP228n (14), yet the marked derivative pWDL7::rfp does not contain any chlorocatechol degradation genes. Due to size differences of plasmid pWDL7 in different archived clones of JP228n(pWDL7), it is currently unclear whether that plasmid contained the complete degradation pathway before it was marked and transferred to other strains. Ongoing studies like the complete genome sequence determination of the original pWDL7 host, C. testosteroni WDL7 (Albers and Springael, personal communication), will be able to address this question.
Our results suggest that plasmid pWDL7::rfp might successfully bioaugment wastewater treatment plants by accelerating the removal of recalcitrant chloroanilines. Even though 3-CA was eventually removed by the activated sludge itself, the acceleration of degradation by almost a week can be of great benefit in a real treatment reactor. As more water can be treated in a given period of time, the plant footprint can be reduced, which translates into great cost savings. This accelerated removal of a chlorinated aromatic compounds is in line with previous reports of successful bioaugmentation of soils and wastewater treatment systems by conjugal transfer of plasmids that enable degradation of chlorinated aromatics (56). One study in particular showed that bioaugmentation by conjugative transfer of the IncP-1 plasmid pNB2 to activated sludge bacteria might enhance chloroaniline transformation (4). Like pWDL7::rfp, this plasmid is thought to contain only a partial degradative pathway specific for chloroanilines (3, 6, 7). Therefore, both in that study and ours, the genomes of strains that can use 3-CA as the sole carbon source upon acquisition of these plasmids likely encode a modified ortho- or meta-ring cleavage pathway to allow chlorocatechol metabolism. The efficient removal of 3-CA by the mating mixture of UWC3(pWDL7::rfp) and D. acidovorans ATCC 15668 (Fig. 6) was expected, since other D. acidovorans strains have been shown to possess the necessary lower-pathway genes needed for mineralization of 3-CA (22).
In summary, based on their complete genome sequences, the IncP-1β plasmids pWDL7::rfp and pNB8c were shown to contain, respectively, two copies and one copy of a chloroaniline dioxygenase gene cluster that encodes only the upper pathway of chloroaniline degradation. Moreover, comparison of the two plasmids' catabolic gene sequences and their transcription and function suggests that promoter differences can explain the lack of catabolic activity conferred by pNB8c outside its native host. Finally, bioaugmentation of an activated sludge community with pWDL7::rfp resulted in accelerated 3-CA removal, suggesting the plasmid's potential in aiding 3-CA cleanup by wastewater treatment plants.
This work was supported by grants EF-0627988 from the NSF Microbial Genome Sequencing Program, and NIH grants R01 GM073821 from the National Institute for General Medical Sciences, and P20RR16448 and P20RR016454 (COBRE and INBRE programs) from the National Center for Research Resources. Support was also provided by the University of California Toxic Substances Research and Teaching Program and Ecotoxicology Lead Campus Program to S.W. and by NSF grant MCB-1022362 to R.E.P.
We are grateful to the U.S. Department of Energy Joint Genome Institute for providing the draft plasmid genome sequences (with special thanks to Brian Foster, Alla Lapidus and Kerry Barrie). Their work is supported by the Office of Science of the U.S. Department of Energy under contract no. DE-AC02-05CH11231. We also thank P. Green for his assistance with HPLC method development, G. Van der Auwera for generating Fig. 2 and for useful discussions and suggestions, N. Boon for providing plasmid pNB8c, and JCVI for offering the JCVI Annotation Service, which provided us with automatic annotation data and the manual annotation tool Manatee.
Published ahead of print 18 November 2011