INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Bioremediation of soil contaminated with polychlorinated biphenyls (PCBs) is an attractive clean-up strategy due to its potential to mineralize pollutants and to be inexpensive. Many PCB-degrading bacteria have been isolated and characterized (2, 3, 6, 8, 10, 22, 39). Some of these bacteria can grown on monochlorinated and dichlorinated biphenyls, and most cometabolize more highly chlorinated biphenyls while using biphenyl as a growth substrate (1, 7, 11). In some cases, the presence of biphenyl as a potential growth substrate and inducer of PCB metabolism (14) is important for maintaining PCB biodegradation activity in soil (5, 17). However, adding biphenyl to soil to stimulate PCB degradation activity is problematic due to the low water solubility of biphenyl and its possible adverse health effects (1, 19). Biphenyl is rare in natural environments, and it is possible that other, more common compounds also induce genes encoding biphenyl-degrading enzymes, termed bph genes (23). Such inducers may be less toxic and more water soluble than biphenyl, so that they could be added to soil to stimulate PCB degradation activity in bioremediation projects.

Several studies have investigated induction of PCB removal in cell suspensions of PCB-degrading bacteria by compounds other than biphenyl. Notably, cell suspensions of Arthrobacter sp. strain B1B grown on fructose medium supplemented with L-carvone, limonene, p-cymene, or isoprene remove Aroclor 1242 (27). Alcaligenes eutrophus H850 and Corynebacterium sp. strain MB1 grown on plant phenolic compounds and Pseudomonas sp. strain LB400 (10) (now a member of the genus Burkholderia [47]) grown on plant phenolic compounds, glucose, or glycerol degrade certain PCB congeners (9, 15). Also, other workers have amplified mRNA transcripts of 2,3-dihydroxybiphenyl dioxygenase (bphC) in A. eutrophus H850 grown on fructose plus L-carvone; however, these transcripts were not quantified to determine if there is a significant difference between the levels of bphC mRNA in cells grown on fructose alone and the levels of bphC mRNA in cells grown with carvone (38). Finally, Cellulomonas sp. strain T109 and Rhodococcus rhodochrous T100 grown on cymene and limonene, respectively, remove over 80% more Aroclor 1242 than these organisms grown on glucose (28). These studies support the hypothesis that certain compounds other than biphenyl may be used to stimulate PCB biodegradation. However, investigations so far have not shown that bacteria grown on substrates other than biphenyl remove PCBs as a result of induction of bph genes at levels above constitutive levels. Moreover, it is possible that the compounds used to induce bacterial PCB degradation activity did not induce bph genes but instead induced genes that encode other enzymes that also degrade PCBs or stimulated PCB degradation via mechanisms other than genetic regulation.

To determine if compounds other than biphenyl induce bph genes (Fig. 1), we constructed a chromosomal bphA-lacZ reporter in the psychrotolerant PCB-degrading bacterium Pseudomonas sp. strain Cam-1 (34) to generate strain Cam-10. We also constructed a chromosomal bphA-lacZ reporter in the mesophilic PCB-degrading bacterium Burkholderia sp. strain LB400 to generate strain LB400-1. Construction of Cam-10 and LB400-1 allowed us to study the regulation of bph genes in a chromosomal context. We incubated Cam-10 and LB400-1 with compounds that previously have been shown to stimulate PCB degradation in other bacteria or that are structurally similar to biphenyl. Then we performed beta-galactosidase assays to determine if the lacZ reporter gene was induced. Induction of beta-galactosidase activity was correlated to induction of bphA. Our results suggest that regulation of bphA in Cam-1 is highly specific. In contrast, the beta-galactosidase activities were indistinguishable in LB400-1 cells incubated with biphenyl and LB400-1 cells incubated with pyruvate, suggesting that in LB400 bphA is expressed constitutively.

View larger version (16K): [in this window] [in a new window]   FIG. 1.   Organization and similarity of the bph gene clusters in Pseudomonas sp. strain Cam-1 (A) and Burkholderia sp. strain LB400 (B). bphA, gene encoding the terminal dioxygenase large subunit; bphE, gene encoding the terminal dioxygenase small subunit; bphF, gene encoding ferredoxin; bphG, gene encoding ferredoxin reductase; bphB, gene encoding dihydrodiol dehydrogenase; bphC, gene encoding 2,3-dihydroxybiphenyl dioxygenase. In the LB400 operon the locations of promoter regions are indicated by p1, p2, and p3.

Few studies thus far have compared how different PCB-degrading bacteria regulate genes that encode enzymes involved in the biphenyl degradative pathway. We investigated induction of bphA in two PCB-degrading bacteria and found that bph genes in these organisms are regulated differently. This result has implications for PCB bioremediation strategies, as it suggests that the optimal methods for stimulating PCB degradation activity may depend on which PCB-degrading bacteria are present.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Bacterial strains, plasmids, and culture conditions. The bacterial strains and plasmids used in this study are listed in Table 1. Unless otherwise specified, Escherichia coli was cultured at 37°C in Luria-Bertani (LB) medium, and Pseudomonas sp. strain Cam-1 and Burkholderia sp. strain LB400 were grown at 15 or 30°C in tryptic soy broth or mineral medium (6) containing 1.3 or 9 mM pyruvate as the growth substrate.

Chemicals. Biphenyl (99%), (±)-camphor (96%), (s)-(+)-carvone (96%), beta-citronellol (95%), cumene (99%), p-cymene (99%), anthracene (99%), benzoate (99%), fluorene (99%), naphthalene (99%), 2-methylnaphthalene (97%), 1,4-dimethylnaphthalene (95%), and phenanthrene (99.5%) were obtained from Aldrich Chemical Co. (+)-Limonene (97%), linoleic acid (60%), myricetin (85%), naringenin (95%), (+)-()-pinene (99%), salicylic acid (99%), and o-nitrophenyl-beta-D-galactopyranoside were obtained from Sigma. Benzene (99.9%) and toluene (99.8%) were obtained from Fisher Scientific. 2-Chlorobiphenyl (99%), 3-chlorobiphenyl (99%), 4-chlorobiphenyl (99%), 2,2'-dichlorobiphenyl (99%), 4,4'-dichlorobiphenyl (99%), and Aroclor 1242 (99%) were obtained from AccuStandard.

Cloning bph genes from strain Cam-1. Total genomic DNA was isolated from strain Cam-1 by using hexadecyltrimethylammonium bromide (4) and was partially digested with Sau3A. The partially digested DNA was size fractionated with a 10 to 40% linear sucrose gradient. DNA fragments approximately 20 kb long were cloned into SuperCos by following the instructions of the manufacturer (Stratagene). In vitro packaging of the recombinant molecules was performed with GigapackII Gold packaging extract (Stratagene), and packaging reactions were used to infect E. coli XL1-Blue MR. The resulting cosmid library was amplified and screened for production of 2-hydroxy-6-oxo-6-phenyl-hexa-2,4-dienoic acid (a yellow meta-cleavage product) from biphenyl (32). Removal of biphenyl by yellow colonies was confirmed by adding 25 mg of biphenyl per liter to cell suspensions of the clones and then extracting the remaining biphenyl with hexane after incubation and analyzing the extracts by gas chromatography-mass spectrometry (34). Cosmid pEM1 was obtained from constructs that transformed biphenyl. Restriction fragments of cosmid pEM1 were separated on an agarose gel, transferred to a maximum-strength Nytran Plus nylon membrane (S&S Nytran Plus), and then hybridized with ³²P-labeled bphA, bphF, and bphG from pT7-6a (Table 1). A nick translation system from Life Technologies was used to label bphA and bphG with [-³²P]dCTP. An Oligolabelling Kit from Pharmacia Biotech (Uppsala, Sweden) was used to label bphF with [-³²P]dCTP. A SacI restriction fragment that hybridized to all three probes was subcloned into pUC19, giving pEM10. The sequence of the cloned DNA from Cam-1 was obtained by generating successive unidirectional deletions of pEM10 with the double-stranded nested-deletion system from Pharmacia Biotech. Oligonucleotide primers synthesized at the Nucleic Acid and Protein Services Unit of the University of British Columbia were used to sequence DNA regions not covered by the deletions. DNA sequences were determined by the Nucleic Acid and Protein Services Unit by using AmpliTaq FS dyedeoxy terminator cycle sequencing chemistry (Applied Biosystems) and Centri-Sep columns (Princeton Separation, Adelphia, N.J.) to purify the extension products. ClustalX was used to align the cloned Cam-1 DNA sequence with the bph operon sequence from LB400 (Fig. 1). Vent polymerase (New England Biolabs) and PCR primers with 5' extensions containing BamHI recognition sites were used to subclone bphAEFG from pEM10 into pT7-7, which yielded pT7-7a.

Insertion of a lacZ-Gm^r cassette into the bph operon. The pEX100T gene replacement vector (42, 43) was used to insert a selectable lacZ reporter gene cassette between the bphA and bphE genes in Cam-1 and LB400 (Fig. 2). Plasmids pEM2, pEM20, and pEM21 were selected in E. coli DH5 grown on LB medium containing appropriate antibiotics and 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal). Plasmid pEM21 was transformed into the mobilizer strain E. coli S17-1 and conjugally transferred into Cam-1 and LB400 (24). Cam-1 and LB400 grow on pyruvate; however, E. coli S17-1 does not. Therefore, exconjugants were plated on minimal medium containing 9 mM pyruvate and 10 µg of gentamicin per ml, and colonies which appeared on this medium after 48 h of incubation at 30°C were streaked onto LB agar containing 10 µg of gentamicin per ml. Colonies of Cam-1 and LB400 in which double homologous recombination had occurred were selected on LB medium containing 10 µg of gentamicin per ml and 5% sucrose and were designated Cam-10 and LB400-1, respectively. Sucrose-resistant colonies were also sensitive to ampicillin, indicating that these colonies had lost the pEX100T vector-associated sequences. Gene insertions in Cam-10 and LB400-1 were also confirmed by performing colony PCR with 20-mer primers annealing to the 3' region of bphA (5'-GACCTGGCAGAACAGCGACT) and the 5' regions of bphE (5'-TCTGCACATGCACGTCCAGC-3') and the lacZ reporter gene (5'-GTATCGCTCGCCACTTCAAC-3') (50).

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Construction of a selectable lacZ reporter cassette and pEM21. Step A was construction of pEM20. The 877-bp XbaI fragment from pUCGm was gel purified and ligated into the XbaI site of pIND/lacZ. Restriction digests obtained with EcoRV were used to isolate plasmids containing lacZ and the gene encoding gentamicin acetyltransferase 3-1 in the same transcriptional orientation. Step B was construction of pEM2. The 4-kb BamHI fragment from pT7-7a was treated with the large fragment of DNA polymerase I (PolK) and then ligated into the SmaI site of pEX100T. Restriction digests obtained with SacI and SacII were used to isolate plasmids containing lacZ and bphAEFG in the same transcriptional orientation. Step C was construction of pEM21. The 4-kb PmeI fragment of pEM20 was gel purified and ligated into the PolK-treated EcoRI site of pEM2. Colonies containing bphAEFG and lacZ-Gmr in the same transcriptional orientation were detected by formation of a blue color when the organisms were grown on LB medium supplemented with gentamicin, ampicillin, and X-Gal. The locations of restriction sites and genes and their transcriptional orientations are shown. Ap, -lactamase-encoding gene; Gm, gentamicin acetyltransferase 3-1-encoding gene; Neo, neomycin resistance gene; oriT, origin of transfer.

Beta-galactosidase assays. Unless otherwise specified, Cam-10 and LB400-1 were grown at 30°C on 9 mM pyruvate in the presence of 10 µg of gentamicin per ml to the mid-log phase and then cooled on ice for 15 min. Cultures were centrifuged at 5,000 × g for 15 min at 4°C and washed with mineral buffer. Washed cells were suspended in mineral medium with 1 mM pyruvate and adjusted to a final optical density at 600 nm of 0.6. Cell suspensions (20 ml) were prepared in 125-ml Erlenmeyer flasks and then were inoculated with potential inducers of the bphA gene at concentrations of 0.001 to 1 mM. Unless otherwise specified, triplicate cell suspensions were incubated with potential inducers for 3 h at 30°C on a rotary shaker at 200 rpm. Beta-galactosidase assays were performed as described by Miller (35). Precise volumes of chloroform (20 µl) and 0.1% sodium dodecyl sulfate (10 µl) were used to permeabilize cells (25). Test samples without o-nitrophenyl--D-galactopyranoside were used as negative controls. Protein concentrations of cell suspensions were determined by a bicinchoninic acid protein assay (4).

Biphenyl removal by Cam-10 and LB400-1. Cell suspensions of Cam-10 and LB400-1 were prepared as described above, and then 2.5-ml aliquots were transferred to Teflon-lined screw-cap tubes. Duplicate cell preparations were inoculated with 0.1 mM biphenyl and then incubated on a tube roller at 30°C for 3 or 6 h. Boiled cells and mineral medium containing 0.1 mM biphenyl were used as negative controls. The remaining biphenyl was extracted from cell suspensions with hexane, and extracts were analyzed by gas chromatography as described previously (34).

Nucleotide sequence accession number. The Cam-1 nucleotide sequence determined in this study has been deposited in the GenBank database under accession no. AY027651.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Optimization of lacZ reporter gene expression in Cam-10 and LB400-1. Maximum expression of the lacZ reporter gene in Pseudomonas sp. strain Cam-10 was observed after 3 h of incubation with 1 mM biphenyl (Fig. 3A). Other compounds that were studied to determine their abilities to induce beta-galactosidase activity in Cam-10 were tested under these conditions. Expression of the lacZ reporter gene in Cam-10 did not increase as the amount of biphenyl was increased above 1 mM. At concentrations of biphenyl less than 0.1 mM, beta-galactosidase activity was consistently higher when 1 mM pyruvate was also supplied. Pyruvate may provide cells with energy that allows greater beta-galactosidase production. Thus, unless otherwise stated, 1 mM pyruvate was added to all subsequent preparations in which potential inducers of bphA were tested.

View larger version (15K): [in this window] [in a new window]   FIG. 3.   Induction of beta-galactosidase activity at 30°C in Cam-10 (A) and LB400-1 (B). The error bars indicate standard deviations (n = 3). The treatments consisted of 1 mM pyruvate alone (+), 1 mM pyruvate plus 0.01 mM biphenyl (), 1 mM pyruvate plus 0.33 mM biphenyl, (), and 1 mM pyruvate plus 1 mM biphenyl ().

Inducers of beta-galactosidase activity in Cam-10. Biphenyl induced beta-galactosidase activity in Cam-10 to a level approximately six times greater than the basal level of expression in cells grown with pyruvate (Fig. 4A). At a concentration of 1 mM, 2-chlorobiphenyl, 4-chlorobiphenyl, salicylate, and naphthalene induced beta-galactosidase activity to levels greater than the basal levels. Thus, these compounds appear to be inducers of bphA in Cam-1. However, none of them appears to be as strong an inducer as biphenyl, as none of them induced beta-galactosidase to the same level of activity in Cam-10 as biphenyl did. At noninhibitory concentrations (Table 2), none of the other potential inducers tested induced beta-galactosidase activity to levels greater than the basal levels in Cam-10. The levels of beta-galactosidase activity after exposure to benzene, carvone, 3-chlorobiphenyl, and pyruvate (Fig. 4A) were typical of those observed after exposure to other noninducing aromatic compounds, terpenoids, chlorobiphenyls, sugars, alcohols, and organic acids (data not shown) (the compounds tested are listed in Table 2).

View larger version (31K): [in this window] [in a new window]   FIG. 4.   Induction of beta-galactosidase activity for 3 h at 30°C in Cam-10 (A) and LB400-1 (B). The error bars indicate standard deviations (n = 3). The concentrations of potential inducers used to test induction of beta-galactosidase activity in Cam-10 are indicated in parentheses. All preparations were supplemented with 1 mM pyruvate.

Inducers of beta-galactosidase activity in LB400-1. In contrast to the results obtained with Cam-10, the beta-galactosidase activities in cell suspensions of LB400-1 containing 1 mM biphenyl, carvone, cumene, cymene, pinene, limonene, fluorene, 3-chlorobiphenyl, or toluene were similar to the beta-galactosidase activity observed in cell suspensions containing only pyruvate (Fig. 4B). Interestingly, 1 mM 2,3-dihydroxybiphenyl had a slight inhibitory effect on induction of beta-galactosidase activity in LB400-1 (Fig. 4B). These results are consistent with S1 nuclease mapping studies of bph genes in LB400 done by other workers which identified three transcriptional initiation sites (16). Activation from the promoter region furthest upstream from the biphenyl dioxygenase translation start site (p3) is dependent on biphenyl. However, activation from the two proximal promoter regions (p2 and p1) is constitutive (16).

Inhibition effects of potential inducers. At a concentration of 1 mM, most of the potential inducers tested actually inhibited induction by biphenyl of beta-galactosidase activity in Cam-10 (Table 2). The exceptions were compounds previously found to be inducers, naphthalene, salicylate, 2-chlorobiphenyl, and 4-chlorobiphenyl, as well as naringenin, fructose, glucose, and glycerol. With the exception of benzoate, acenaphthalene, fluorene, dioxin, anthracene, 3-chlorobiphenyl, 2-methylnaphthalene, and dimethylnaphthalene, compounds that inhibited induction also inhibited cell growth. At concentrations less than 0.1 mM, none of the potential inducers were inhibitory to cell growth, yet at such concentrations several of these compounds substantially inhibited induction of beta-galactosidase. Clearly, in complex environments, inhibitory effects such as those found here can be expected to modulate expression of genes essential for PCB biodegradation. The inhibitory effect of soil extracts (Table 2) is consistent with this expectation.

Temperature dependence of bphA induction in Cam-1 and LB400. Cam-1 was isolated from PCB-contaminated arctic soil and was studied to determine the feasibility of bioremediating PCB-contaminated arctic soil with indigenous soil bacteria. We found that at 7°C Cam-1 removed PCBs at higher rates than LB400 removed PCBs (34). To investigate the role of bphA induction in the efficiency of PCB removal at low temperatures, we compared the beta-galactosidase activities in cell suspensions of Cam-10 and LB400-1 incubated at 7°C with pyruvate or biphenyl plus pyruvate. Cell suspensions of Cam-10 were prepared by using cells grown on pyruvate at 7°C. Since LB400 does not grow at 7°C (34), cell suspensions of LB400-1 were prepared by using cells grown on pyruvate at 15°C. Samples of the cell suspensions were obtained at several time points over 24 h and transferred to 28°C to measure beta-galactosidase activity.