INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

Polychlorinated biphenyls (PCBs) are synthesized by direct chlorination of biphenyl, resulting in 209 possible congeners. The physical and chemical nature of PCBs, including lipophilicity, heat resistance, and relative inertness, has resulted in their widespread industrial application. The nature of PCBs has also led to their persistence and bioaccumulation and has caused health problems in contaminated organisms (reviewed in references 8 and 15). As a result, the production of most PCBs was banned in the United States in 1976 and in Canada in 1977. However, several locations, including areas in the Canadian Arctic, remain polluted with PCBs. Canadian environmental legislation and aboriginal land settlement agreements require remediation of many of these Arctic sites. Conventional clean-up strategies at such remote locations are very expensive because excavation and transportation of polluted soil to treatment facilities are generally required. Bioremediation of PCBs, then, may be the most cost-effective strategy, since it allows on-site treatment.

Several mesophilic bacteria capable of cometabolizing PCBs via the biphenyl catabolic pathway have been described (1, 8, 13). Most of these organisms are able to degrade biphenyls with up to four chlorine substituents, although bacteria capable of degrading biphenyls with up to seven chlorine substituents have been isolated (5, 9). Bioremediation strategies in Arctic and temperate areas that use cold-adapted PCB-degrading bacteria may be more efficient than strategies using mesophilic PCB-degrading bacteria, since heating requirements for growth and PCB degradation activity may be reduced. In addition, PCB-degrading bacteria indigenous to Arctic soil have presumably adapted to various soil characteristics that limit the survival or activity of foreign PCB-degrading microorganisms. Others have shown removal of biphenyls with as many as three chlorine substituents at 4°C by PCB-contaminated river sediment (28). However, PCB-degrading bacterial isolates from these river sediments have not been reported. Psychrophilic and psychrotolerant PCB-degrading bacteria indigenous to Arctic soils have been reported (21), but such organisms have not been characterized.

In this study we isolated and characterized, for the first time, PCB-degrading psychrotolerant bacteria from PCB-contaminated Arctic soil. We also characterized Sag-50G, a psychrotolerant PCB-degrading bacterium previously isolated (21). The extent and rate of removal of a PCB mixture by each Arctic soil isolate at a range of temperatures were compared with those of the mesophile Burkholderia cepacia LB400 (17) (previously known as Pseudomonas sp. strain LB400 [7]). LB400, compared to most other PCB-degrading bacteria that have been studied, degrades a wide range of PCB congeners (4). Our results suggest that PCB degradation enzymes of the Arctic soil bacterial isolates are cold adapted. Since the psychrotolerant PCB-degrading organisms grow at lower temperatures than mesophiles, such as LB400, and appear to express genes encoding cold-adapted PCB-degrading enzymes, these psychrotolerant isolates are potentially useful for PCB bioremediation at low temperatures.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Isolation of PCB-degrading bacteria from Arctic soil samples. Soil samples were obtained from a Vancouver location or from PCB-contaminated (5 to 100 µg/g of dry soil) or pristine Arctic locations: Saglek, Labrador (57°N, 63°W), Cambridge Bay, Northwest Territories (69°N, 105°W), and Iqaluit, Northwest Territories (63°N, 69°W). Enrichment cultures were prepared with 1.0 g of soil suspended in 100 ml of phosphate-buffered mineral medium (PAS) (6) containing 200 mg of biphenyl (Aldrich)/liter as the sole organic carbon source. The enrichment cultures were incubated on a shaker for 3 weeks at 7°C. Subsequent enrichment cultures had 1% inocula. If growth occurred, 0.1% of the culture was transferred to homologous medium. Enriched cultures were diluted in PAS medium, then spread on PAS with 1.5% purified agar and 200 mg of biphenyl/liter on the agar surface. Distinct colonies were picked, suspended in 2.5 ml of PAS medium with 200 mg of biphenyl/liter, and incubated at 7°C. Isolates which grew were stored at 70°C with 8% dimethyl sulfoxide.

PCB degradation by cell suspensions. Arctic soil isolates were grown at 7°C, and B. cepacia LB400 was grown at 15°C in 1-liter cultures to late-logarithmic phase. Remaining biphenyl crystals were allowed to precipitate; then cultures were decanted and centrifuged at 8,000 × g for 10 min and washed with PAS buffer. Final cell suspensions contained 600 to 800 µg of cell protein per ml of buffer. Cell samples of 2.5 ml were transferred to solvent-washed screw-cap tubes with Teflon-lined caps (n = 3). Control cell suspensions were boiled. PCBs (100 mg of Aroclor 1242/liter and 10 mg of 2,2',4,4',6,6'-hexachlorobiphenyl/liter) were added to each tube, and then the reaction tubes were incubated at 7, 37, or 50°C. Reaction tubes were transferred to 20°C at regular time intervals over 24 h to stop PCB removal activity. Control tubes were incubated for 0, 4, and 24 h. Initial rates of removal of selected PCB congeners were calculated from the slopes of initially linear curves, generated by plotting the percentage of the PCB congener remaining versus time, and were standardized for protein concentration. The protein concentrations of cultures were determined by a bicinchoninic acid protein assay (2).

Analysis of PCBs. Remaining PCBs from batch cultures and cell suspensions were extracted twice with an equal volume of hexane. The extracts were pooled, then mixed with Na₂SO₄ to dry the organic phase. The extracts were analyzed by using a gas chromatograph (GC) equipped with an ion trap mass spectrophotometer (21). The extraction efficiency was 98% ± 0.5%.

Characterization of isolates. Gram staining, cell size, and motility tests were performed on liquid cultures in late-logarithmic phase. Colony morphology determination, oxidase tests, and catalase tests were performed on colonies grown on PAS agar with biphenyl at 7°C (25). Anaerobic respiration with nitrate by cultures using glucose as a carbon source was tested as previously described (29). The abilities of each isolate to use a range of substrates for growth at 7°C (see Table 3) and to oxidize substrates contained in GN microplates (Biolog, Hayward, Calif.) were also determined. The percent removal of 100 mg of Aroclor 1242/liter at 7°C by Arctic soil isolates growing on primary substrates other than biphenyl was determined by extracting and analyzing remaining PCBs from batch cultures after 5 weeks of incubation (n = 2).

16S rDNA PCR and DNA sequencing. Cell pellets from 500-µl aliquots of late-logarithmic-phase cultures grown at 7°C with 200 mg of biphenyl/liter were washed with 0.8% NaCl, suspended in 90 µl of sterile, deionized water, and boiled for 10 min. Cell debris was pelleted by brief centrifugation. The 16S ribosomal DNA (rDNA) was amplified from each isolate by PCR using the reagents and procedures of Gibco BRL Life Technologies, Inc. (Gaithersburg, Md.) and primers hybridizing to positions 8 to 27 and 1541 to 1525 (18). The Oligonucleotide Synthesis Laboratory, University of British Columbia, synthesized these primers. Thermal cycling was performed in a PowerBlock II System (ERICOMP) according to the following program: initial denaturation at 94°C for 3 min; 30 cycles of 94°C for 30 s, 43°C for 1 min, and 72°C for 2 min; and a final extension at 72°C for 10 min.

Phylogenetic analysis. Each 16S rDNA sequence that was determined was compared to other prokaryotic 16S rDNA sequences by using the Similarity_ Rank analysis service of the Ribosomal Database Project (19). The 16S rDNA sequences of the closest relatives to the Arctic soil isolates env. JAP501 (GenBank accession no. U09827), env. JAP412 (GenBank accession no. U09773), and Pseudomonas syringae A501 (GenBank accession no. L24787), as well as representative -proteobacteria Acetobacter aceti (ATCC 15973), Agrobacterium tumefaciens (ATCC 4720), Sphingomonas capsulata (GIFV 11526), Sphingomonas paucimobilis (ATCC 29837), and Sphingomonas terrae (IFO 15098), -proteobacteria Alcaligenes denitrificans (ATCC 15173), B. cepacia (ATCC 25416), Comamonas testosteroni (ATCC 11996), Stenotrophomonas maltophilia (ATCC 13637), and Zoogloea ramigera (ATCC 25935), and -proteobacteria Pseudomonas aeruginosa (ATCC 10145), Pseudomonas fluorescens (ATCC 13525), Pseudomonas putida (ATCC 12633), Pseudomonas stutzeri (ATCC 17588), and Escherichia coli (GenBank accession no. J01695), were retrieved from the Ribosomal Database Project and aligned with the 16S rDNA sequences of the Arctic soil isolates by using Clustal X. One hundred bootstrapped data sets of the aligned sequences were obtained by using SEQBOOT. Phylogeny estimates for each of these data sets were obtained by using the default parameters of DNADIST. A phylogenetic tree was obtained by analyzing the resulting distance matrices with the default parameters of NEIGHBOR and CONSENSE.

Nucleotide sequence accession numbers. The 16S rDNA sequences determined for Cam-1, Sag-1, Iqa-1, and Sag-50G have been deposited in the EMBL Nucleotide Database and have accession numbers AF098464, AF098467, AF098465, and AF098466, respectively.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

Isolation of PCB degraders. Of approximately 50 biphenyl-degrading bacteria isolated from PCB-contaminated or pristine Arctic soil or pristine Vancouver soil, three (Cam-1, Sag-1, and Iqa-1) could remove PCBs. Interestingly, all of the PCB-degrading isolates were obtained from PCB-contaminated Arctic soil samples; no PCB-degrading bacteria were isolated from pristine soil samples. Most of the PCB congeners present in Aroclor 1221 were removed at 7°C by batch cultures of the three Arctic soil isolates identified above, albeit to different extents (Fig. 1; Table 2). The congener ranges of these isolates were similar to the congener range of Sag-50G, which was previously reported (21). None of the isolates could remove 2,2'-chlorobiphenyl, although Cam-1 and Sag-1 did remove 4,4'-chlorobiphenyl. This result is consistent with results for other PCB-degrading bacteria that express 2,3-biphenyl dioxygenase activity and not 3,4-biphenyl dioxygenase activity (4). We also tested removal of Aroclor 1242 at 7°C and at 15°C by batch cultures of each isolate. The total removal of Aroclor 1242 by Cam-1 and Sag-1 was approximately 30 and 15% higher at 15°C than at 7°C, respectively (Table 2). However, the total removal of Aroclor 1242 by Iqa-1 was approximately 30% higher at 7°C than at 15°C. Differences in relative removal (per unit of final biomass) were consistent with differences in absolute removal. The range of Aroclor 1242 congeners removed by each of the isolates at 7 and 15°C was similar to the range of congeners removed by Cam-1 at 7°C (Fig. 2) and to those of most mesophilic PCB-degrading bacteria that have been reported (reviewed in reference 1). Cam-1 also appeared to remove some tetrachlorobiphenyls (peaks 18, 21, and 26) at 15°C. The removal of 2,3,4,3',4'-pentachlorobiphenyl by Cam-1 was considered unlikely, since removal of 2,5,3',4'- and 2,4,3',4'-tetrachlorobiphenyl was not observed. However, the position of chlorine substituents, and not only the number of chlorine substituents, can determine whether a PCB congener is degraded (4). Experiments showing removal of pure tetrachlorobiphenyls and 2,3,4,3',4'-pentachlorobiphenyl would be required to confirm that Cam-1 can remove such PCB congeners.

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Percent removal of major congeners in Aroclor 1221 by batch cultures of Arctic soil isolates incubated at 7°C for 5 weeks (n = 2; error bars indicate ranges). Peaks are identified in Table 1.

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Percent removal of major congeners in Aroclor 1242 by a batch culture of Cam-1 incubated at 7°C for 5 weeks (n = 3; error bars indicate standard deviations). Peaks are identified in Table 1.

Morphology, physiology, and phylogeny. All of the PCB-degrading bacteria isolated were most closely related to the genus Pseudomonas. Both physiological analyses and 16S rDNA sequence analyses support this conclusion. All isolates were gram-negative, catalase-positive, motile rods. Cam-1 and Iqa-1 were oxidase positive; Sag-1 was oxidase negative. The isolates were obligately aerobic and failed to grow anaerobically on glucose, fermentatively or with nitrate. When grown on biphenyl at 7°C, Cam-1 cells were 1.5 to 3.0 by 1.0 µm, Sag-1 cells were 1.5 to 4.0 by 1.0 µm, and Iqa-1 cells were 1.2 to 4.0 by 1.0 µm. On PAS agar with biphenyl, colonies of each isolate were circular with a slightly undulate edge, convex, butyrous, smooth, grey, and opaque, except for Sag-1, which was translucent. The use of different primary growth substrates and oxidation of test substrates in GN Biolog plates (Table 3) confirmed that these isolates are distinct from each other and from Sag-50G. All of the Arctic soil isolates used glucose and galactose as primary growth substrates. None of the Arctic soil isolates used 2-chlorobiphenyl, 3-chlorobiphenyl, camphor, citronellol, cymene, dehydroabietic acid, limonene, methanol, n-hexadecane, pentachlorophenol, phenol, or pinene as primary growth substrates.

View larger version (20K): [in this window] [in a new window]   FIG. 3.   Unrooted tree showing phylogenetic relationships of Arctic soil isolates (in bold) and representative members of -, -, and -proteobacteria. The phylogenetic tree was generated with nearly complete 16S rDNA sequences. The scale shows evolutionary distance. Numbers indicate bootstrap values.

Effect of primary substrate on PCB degradation. Arctic soil isolates growing on primary substrates other than biphenyl also partially removed Aroclor 1242. Compared to relative removal of Aroclor 1242 in cultures growing on biphenyl, relative removal was lower in cultures growing on all other primary substrates (Table 4). Exceptions were linoleic acid-grown and, to a lesser extent, pyruvate-grown cultures of Sag-1, whose relative removals were higher than those of biphenyl-grown cultures. Partial removal of monochlorobiphenyls and of some di- and tri-chlorobiphenyls was detected in cultures of each isolate growing on linoleic acid, glucose, glycerol, acetate, or pyruvate. Cultures of Cam-1 and Sag-1 growing on benzoate failed to remove PCBs. Cultures of Sag-50G growing on benzoate removed comparatively small amounts of the PCB mixture. It is possible that benzoate inhibits induction of the biphenyl operon in these organisms, since benzoate is an end product of the biphenyl degradative pathway. In some cases, substrates supported similar absolute removals, but different relative removals, of Aroclor 1242 by a particular isolate. Therefore, primary substrates affect PCB removal by affecting both cell density and the efficiency of PCB removal by a constant amount of biomass.

Temperature optimum for growth. With biphenyl as their organic substrate, Cam-1 and Iqa-1 grew optimally at 15°C, and Sag-1 grew optimally at 7°C (Fig. 4). Up to 2 weeks of incubation at 7°C were required before growth was detected, and the exponential phase continued up to 3 days. The mesophile B. cepacia LB400 did not grow at 7°C but did grow at 15 and 37°C. Temperatures higher than 37°C were not tested. Psychrophilic and psychrotolerant bacteria are defined by optimal growth at temperatures below 20°C and growth at temperatures as low as 0°C. Psychrotolerant bacteria are distinguished by growth at temperatures above 20°C (22). Since Cam-1, Iqa-1, and Sag-1 grew at 22°C, these organisms are psychrotolerant rather than psychrophilic bacteria.

View larger version (16K): [in this window] [in a new window]   FIG. 4.   Growth rates of Arctic soil isolates grown on biphenyl at various temperatures (n = 3; error bars indicate standard deviations). , Cam-1; , Sag-1; , Iqa-1.

PCB degradation by cell suspensions. High rates of PCB removal at low temperatures and the temperature sensitivity of PCB removal activity suggest that PCB degradation enzymes expressed by the Arctic soil bacteria are cold adapted. The ranges of PCBs removed by cell suspensions of Arctic soil isolates at 7°C (data not shown) did not differ from those of batch cultures growing at 7°C (Fig. 2). Although no significant removal of peak 10 by Cam-1 was detected, an initial rate of removal was evident (Table 5). Cam-1 probably removed 2,6,4'-trichlorobiphenyl, which is a minor component of peak 10, rather than 2,3,2'-trichlorobiphenyl, which is the major component of peak 10. The inability of Cam-1 to remove 2,3,2'-trichlorobiphenyl seems consistent with its inability to remove other 2,2'-substituted biphenyls. The range of PCBs removed by B. cepacia LB400 differed from those of the Arctic soil isolates, since LB400 could remove 2,2'-dichlorobiphenyl. No significant differences in relative areas of peaks corresponding to PCBs on gas chromatographs were detected between psychrotolerant killed cells and LB400 killed cells or between killed cell controls incubated for different times.

View larger version (35K): [in this window] [in a new window]   FIG. 5.   Rates of degradation of individual PCB congeners in Aroclor 1242 by cell suspensions at various temperatures (n = 2). , peak 4; , peak 5; , peak 12; , peak 14. Peaks are identified in Table 1.