 Springer-Verlag 1996

Appl Microbiol Biotechnol (1996) 46: 673 – 677

ORIGINAL PAPER

M. R. Natarajan · W.-M. Wu · J. Nye · H. Wang L. Bhatnagar · M. K. Jain

Dechlorination of polychlorinated biphenyl congeners by an anaerobic microbial consortium

Received: 30 April 1996 / Received revision: 26 July 1996 / Accepted: 5 August 1996

Abstract An anaerobic methanogenic microbial consortium, developed in a granular form, exhibited extensive dechlorination of defined polychlorinated biphenyl (PCB) congeners. A 2,3,4,5,6-pentachlorobiphenyl was dechlorinated to biphenyl via 2,3,4,6-tetrachlorobiphenyl, 2,4,6-trichlorobiphenyl, 2,4-dichlorobiphenyl and 2-chlorobiphenyl (CB). Removal of chlorine atoms from all three positions of the biphenyl ring, i.e., ortho, meta and para, was observed during this reductive dechlorination process. Biphenyl was identified as one of the end-products of the reductive dechlorination by GCMS. After 20 weeks, the concentrations of the dechlorination products 2,4,6-CB, 2,4-CB, 2-CB and biphenyl were 8.1, 41.2, 3.0 and 47.8 lM respectively, from an initial 105 lM 2,3,4,5,6-CB. The extent and pattern of the dechlorination were further confirmed by the dechlorination of lightly chlorinated congeners including 2-CB, 3-CB, 4-CB, 2,4-CB and 2,6-CB individually. This study indicates that the dechlorination of 2,3,4,5,6-CB to biphenyl is due to ortho, meta and para dechlorination by this anaerobic microbial consortium.

Introduction Polychlorinated biphenyls (PCBs) are ubiquitous anthropogenic pollutants of major environmental concern (Abramowicz 1990; Loganathan and Kannan 1991). Problems of PCBs in the global ecosystem and their impact on public health and wildlife through bioaccumulation in biota have been extensively addressed by several researchers (Loganathan et al. 1993; Mes et al. 1993). Biological approaches have been explored as an effective way to decontaminate PCBs from river, lake and estuarine sediments (Brown et al. 1987a,b; Quensen

et al. 1988; Rhee et al. 1993; Sokol et al. 1994). Several reports are now available on reductive dechlorination of highly chlorinated PCBs by indigenous anaerobic microflora from river and lake sediments and the resultant accumulation of residual ortho-chlorinated congeners, such as 2-chlorobiphenyl (2-CB) and 2,6-CB as endproducts (Abramowicz 1990; Alder et al. 1993). These lower ortho-chlorinated products are not amenable to further anoxic dechlorination and degradation. However, recent reports have indicated that dechlorination of the ortho chlorines of 2,3,5,6-CB and 2,4,6-CB congeners occurs under anaerobic conditions in the Woods Pond sediment (Van Dort and Bedard 1991; Williams 1994). Neither pure nor mixed cultures of anaerobic microorganisms capable of dechlorinating PCBs have been isolated and the mechanism of reductive dechlorination is still not fully known (Assaf-Anid et al. 1992). In our laboratory, we have developed an anaerobic microbial consortium in the form of granules in an upflow anaerobic sludge-blanket reactor (Bhatnagar et al. 1991). These granules contain self-immobilized anaerobic microorganisms such as syntrophic, acetogenic and methanogenic species and are grown in mineral salts medium containing volatile fatty acids (VFA) and glucose. The granules did not lose their metabolic activity even after exposure to oxygen and storage for a brief period (Natarajan et al. 1995). In this paper, we report the dechlorinating ability of our anaerobic methanogenic microbial granules using a defined PCB congeners. Production of lightly-chlorinated congeners and biphenyl is demonstrated by stepwise dechlorination of 2,3,4,5,6-CB and further validated by dechlorination of mono- and dichlorobiphenyls.

Materials and methods M. R. Natarajan · W. -M. Wu · J. Nye H. Wang · L. Bhatnagar · M. K. Jain (&) MBI International, 3900 Collins Road, P. O. Box 27609, Lansing, MI 48909–0609, USA. Fax: (517) 337–2122 e-mail: [email protected]

Chemicals and gases Defined polychlorinated biphenyl congeners and octochloronapthalene were purchased from AccuStandard (New Haven, Conn.). Ultrapure GC-grade solvents, acetone, hexane and iso-oc-

674 tane, were purchased from J. T. Baker (Phillipsburg, N.J.). Florisil, copper powder and all other chemicals were purchased from Sigma Chemical Co. (St. Louis, Mo.). Nitrogen gas and gas mixtures were obtained from Linde Division, Union Carbide Corp. (Warren, Mich.), and passed over heated copper filings to remove traces of O2. All PCB stock solutions were prepared in acetone before use. Cultivation of Anerobic Microbial Granules Anaerobic methanogenic microbial granules that dechlorinated PCBs were grown continuously in an upflow anaerobic sludgeblanket reactor as described previously (Bhatnagar et al. 1991; Wu et al. 1993; Natarajan et al. 1995). The granules were grown and maintained at 20–22 °C with a continuous feed of medium (Wu et al. 1993) containing glucose, methanol, butyric acid, and propionic acid as carbon and electron sources. Recycling of treated effluent from the top of the reactor to the reactor inlet (a recycle ratio of approximately 40 : 1 of the influent flow rate) was used to maintain the required hydraulic flux rate. The reactor was continuously monitored for production of methane and degradation of volatile fatty acids. In addition, the granules were examined periodically for retention of the PCB-dechlorinating activity using defined PCB congeners. Actively growing granules were withdrawn from the reactor when required as inoculum for the experiments.

(20 ml/min) were used as carrier and make-up gas respectively. The temperature program for PCB analysis was as follows: 0.5 min hold at 140 °C, 5 min at 140–280 °C, and a 1-min hold at 280 °C with a total run time of 29.5 min. The injector and detector temperatures were at 280 °C and 320 °C respectively. Peak identification and quantification were based on analysis of the PCB standard mixture performed in our laboratory. Quantification of PCB (5– 40 mg/l) on a micromolar basis (lm) and peak integration were performed by using a PE Nelson Turbochrome 3.1 software system and PCB standard mixture. Generally, the variation between samples was observed to be within a 10% range. On the basis of the GC-ECD response factor for defined congeners, a PCB calibration table was developed in our laboratory with a standard mixture containing 2-CB, 3-CB, 4-CB, 2,4-CB, 2,6-CB, 2,3,6-CB, 2,4,6-CB, 2,3,4,6-CB, 2,4,5,6-CB and 2,3,4,5,6-CB. The dechlorination product, biphenyl in the organic phase, was analyzed by a gas chromatography/mass spectrometer (Hewlett-Packard 5890) equipped with a DB5 MS column (30 m × 0.32 mm × 0.25 mm; J & W Scientific, Folsom, Calif.). Electron-impact ionization (70 eV ionization energy) was used. Samples were injected in a splitless mode. The flow rate of the carrier gas, helium, was 1 ml/min. The temperature program was run from 40 °C to 100 °C at a rate of 10 °C/ min and from 100 °C to 200 °C at 5 °C/min.

Results Dechlorination experiment Dechlorination studies were performed under anaerobic conditions in 158-ml serum vials (Wheaton Scientific, Millville, N.J.) containing 45 ml phosphate-buffered basal medium (pH 7.2) that were sealed with Teflon-coated rubber stoppers and aluminum seals. The basal medium contained (g/l) NaCl 0.5, NH4Cl 0.5, MgCl2.6H2O 0.2, CaCl2·2H2O 0.1, resazurin 0.02, and 10 ml trace element solution (Jain et al. 1991). The medium was reduced with Na2S·9H2O (stock 2.5%, w/v) and the pH was adjusted with phosphate buffer (15% KH2PO4 and 29% K2HPO4) before inoculation. Defined PCB congeners were spiked with a glass Hamilton syringe into the experimental bottles at a final concentration of 40 mg/l before inoculation. The defined PCB congeners, 2-CB, 3-CB, 4-CB, 2,4-CB, 2,6-CB and 2,3,4,5,6-CB were dissolved (approximately 20 mg/ml stock) in acetone. The anaerobic serum vials were inoculated in an anaerobic glove box (Coy Manufacturing Co., Ann Arbor, Mich.). Two replications were kept for each PCB congener. In addition to the acetone (carrier for PCBs), glucose (0.2 g/l, w/v) and/or methanol (2 g/l, w/v) was added into each bottle as growth substrate and energy source to support the reductive dechlorination. Each vial also received 5 ml microbial granules (approximately 0.1 g dry weight) as inoculum. The headspace of each vial, sealed with a Teflon-coated stopper and aluminum cap, was purged with nitrogen gas and the vials were incubated at 30 °C without shaking. Periodically, vials were transported to the anaerobic glove box and 2- to 3-ml aliquots containing granules and broth were sampled in tubes with wide-mouth pipettes. The vials were resealed, removed from the glove box, and incubated at the same conditions.

To determine the pattern and extent of dechlorination, a defined pentachlorobiphenyl congener, 2,3,4,56-CB, containing chlorines at the ortho, meta and para positions on one biphenyl ring was chosen. This congener was selected so that the dechlorination products might be easily identified and the kind of chlorines removed and pattern of dechlorination determined. Anaerobic granules in PBB medium supplemented with glucose and methanol as the source of energy and electron donors showed extensive dechlorination of the 2,3,4,5,6-CB (Fig. 1). Glucose and methanol were supplemented in the experimental vials after every sampling period, since reductive dechlorination is an electron-dependent process in which chlorine is replaced by a hydrogen atom (Nies and Vogel 1991).

PCB extraction and analysis PCBs from the samples were extracted with acetone containing octachloronapthalene (1.6 mg/l) as an internal standard followed by extraction with a mixture of acetone and hexane (1:9 v/v) according to the method described by Quensen et al. (1990). During the extraction procedure, the granules completely disintegrated. Glass columns packed with Florisild copper powder were used to remove interferences such as sulfur and fine granular debris from the extracted samples. The PCBs in the organic fraction were analyzed using a gas chromatograph (GC, Varian 3400) equipped with an electron-capture detector (ECD) and a DB-5 capillary column (30 m × 0.53 mm inner diameter, 1.5 lm film thickness; J & W Sci-entific, Folsom, Calif.). Helium (1.0 ml/min) and N2

Fig. 1 Reductive dechlorination of 2,3,4,5,6-pentachlorobiphenyl (PeCB) and its dechlorination products by the anaerobic microbial granules. TeCB tetrachlorobiphenyl, TCB trichlorobiphenyl, DCB dichlorobiphenyl, CB chlorobiphenyl

675

Fig. 2 Gas chromatography/mass spectroscopy analysis of biphenyl, a dechlorination product of 2,3,4,5,6-CB, after 20 weeks of incubation

A time course of 2,3,4,5,6-CB dechlorination is shown in Fig. 1. In 4 weeks, from an initial 105 lM, approximately 33 lM 2,3,4,5,6-CB was dechlorinated and 27 lM 2,3,4,6-CB and 12 lM of 2,4,6-CB were produced via preferential removal of two meta-chlorines. The 2,3,4,6-CB was concurrently produced from the dechlorination of 2,3,4,5,6-CB and was further dechlorinated. At 8 and 12 weeks, the total amount of this congener increased relative to the amount at 4 weeks. In 16 weeks, the 2,3,4,5,6-CB was completely dechlorinated and the concentrations of its dechlorination products 2,3,4,6-CB, 2,4,6-CB, 2,4-CB, 2-CB and biphenyl were 1, 60, 4, 16 and 33 lM respectively. The concentration of 2,4,6-CB remained at the same level up to 12 weeks; however, at 16 weeks, an approximately fivefold increase was observed in its level. Interestingly, by the 20th week, about 85% of the 2,4,6-CB was dechlorinated to 2,4-CB. The expected dechlorination products of 2,4,6-CB, such as 2,4-CB and 2-CB, were observed only after 12 weeks of incubation. No dechlorination was observed in the presence of autoclaved granules or when the granules were not added. The appearance of biphenyl, as a dechlorination product, was observed and confirmed by the GC-MS analysis. The mass spectra of the biphenyl detected from the reductive dechlorination of 2,3,4,5,6-CB is shown in Fig. 2. Biphenyl formation was detected in an increasing concentration at 16 and 20 weeks indicating that all chlorines including ortho chlorines from 2,3,4,5,6-CB were removed. The mass balance during this study was Fig. 3 Proposed pathway for anaerobic dechlorination of 2,3,4,5,6-CB by anaerobic microbial granules. Chlorine removal was observed in the order meta, ortho and para followed by ortho dechlorination

estimated to be between 80% and 109%. A sequential dechlorination pathway was proposed on the basis of these results: 2,3,4,5,6-CB→2,3,4,6-CB (a product of one meta chlorine removed)→2,4,6-CB (a product of two meta chlorines removed) →2,4-CB (a product of two meta and one ortho chlorines removed) →2 CB (a product of two meta, one ortho and one para chlorines removed)→biphenyl (Fig. 3). To validate these results further, we examined the fate of defined monochlorobiphenyl congeners such as 2-CB (ortho chlorine), 3-CB (meta chlorine) and 4-CB (para chlorine) and dichlorobiphenyl congeners such as 2,4-CB (ortho + para chlorines), and 2,6-CB (two ortho chlorines) individually. These compounds were chosen on the basis of the kind of chlorine present and the type of congeners that are not considered to be dechlorinated anaerobically in natural environments. Anaerobic reductive dechlorination has been mostly shown to dechlorinate highly chlorinated PCBs to lightly orthochlorinated end-products such as 2-CB and 2,6-CB (Brown et al. 1987b; Quensen et al. 1990). Figure 4 shows the disappearance of lightly-chlorinated defined PCB congeners in the presence of anaerobic microbial granules. PCB congeners were respiked after a 12-week incubation period. At this time, the sampling interval was reduced from the initial 6 weeks to every 3 weeks. In this study, the 2-CB was dechlorinated at a higher rate than the 2,6-CB, 3-CB, 4-CB and 2,4-CB. The dechlorination of 2,4-CB was slower than that of the other congeners during the first 12 weeks of incubation. In the second stage of incubation, i.e. after 12 weeks, its dechlorination rate was comparable to that of other congeners. Production of biphenyl was examined in the vials of 2-CB, 3CB and 4-CB at 21 weeks of incubation and was confirmed by GC-MS (data not shown). Biphenyl was not quantified. Relative to other chlorinated congeners such as 2,3,4,5,6-CB, 2,3,4,6-CB, 2,4,6-CB and 2,4-CB, dechlorination of 2-CB was observed at a higher rate.

Discussion In this study, we demonstrate a complete sequential dechlorination of 2,3,4,5,6-CB by methanogenic microbial granules. Previously, only partial reductive dechlorination of defined PCB congeners such as 2,3,4,5,6CB (Nies and Vogel 1991) and 2,3,5,6-CB (Van Dort and Bedard 1991) had been observed in the con-

676

Fig. 4 Dechlorination of lightly chlorinated polychlorinated-biphenyl (PCB)- defined congeners such as 2-CB, 3-CB, 4-CB, 2,4-CB and 2,6CB by anaerobic microbial granules. PCBs were spiked twice at time zero and after 12 weeks of incubation

taminated anaerobic sediments. In the methanogenic pond sediment, removal of only one ortho chlorine (from position 6) of 2,3,5,6-CB (Van Dort and Bedard 1991) was observed. We show here the removal of both ortho chlorines (from the 2 and 6 positions) and the production of biphenyl from 2,3,4,5,6-CB by the sediment-free methanogenic microbial consortium under anaerobic conditions. Intermediary products of dechlorination such as 2,3,4,6-CB, 2,4,6-CB, 2,4-CB and 2-CB were observed and analyzed but intermediates such as 2,3,6CB, 2,6-CB and 4-CB were not detected during this experiment. This suggests that 2,3,4,5,6-CB was dechlorinated via only one stepwise dechlorinating pathway by this consortium. These results indicate that the reductive dechlorination began with the removal of two meta chlorines (first the chlorine from the 5th and then from the 3rd position) followed by one ortho chlorine (from the 6th position), and one para chlorine (from the 4th position) and was completed with the removal of another ortho chlorine (from the 2nd position). A similar initial meta dechlorination, i.e. preferential removal of chlorines from the 5th and 3rd positions, was also observed in a previous study with anaerobic sediment (Van Dort and Bedard 1991). It was interesting that ortho and not para chlorine was preferentially removed from the 2,4,6-CB. This is in contrast to the earlier studies with anaerobic river sediments, in which para dechlorination was frequently reported but ortho dechlorination was not detected (Abramowicz 1990; Fish and Principe 1994). This is the first report of the dechlorination of all chlorines (ortho, meta and para) from defined PCB congeners with biphenyl as the end-product by our culturable anaerobic consortium. These results are supported by a good mass balance. Production of biphenyl unequivocally demonstrated the complete dechlorination of 2,3,4,5,6-CB under

anaerobic conditions. The product, biphenyl, can be easily degraded and mineralized by aerobic microorganisms (Bedard et al. 1987; Ahmed and Focht 1973; Hickey et al. 1993). The reductive dechlorination pattern of this study is distinct from the patterns demonstrated with sediment microflora in several laboratories. Dechlorination of highly chlorinated PCBs to the less toxic, non-chlorinated biphenyl is essentially needed for the development of an anaerobic bioremediation process for PCB. In this study, the 2,4-CB persisted longer than other congeners, indicating that the microbial consortium probably required an adaptation for congeners containing one ortho plus one para chlorine on a single biphenyl ring, such as 2,4-CB, in comparison to a twoortho-chlorine congener such as 2,6-CB (Fig. 4). In contrast, the 2,6-CB that appeared as a major dechlorination end-product of PCBs in the river sediments (Quensen et al. 1990) has been observed to persist in anaerobic sediments. The congeners that contained only one para chlorine, such as 4-CB, were dechlorinated similarly to the dechlorination of ortho- (2-CB) and metasubstituted (3-CB) congeners in this study (Fig. 4). The dechlorination of 2-CB and 2,6-CB by the anaerobic microbial consortium was observed for the first time under anoxic conditions. The GC-MS detection of biphenyl after the treatment of 2-CB, 3-CB and 4-CB with microbial granules strengthens the conclusion that these congeners were also reductively dechlorinated by the anaerobic consortium. Furthermore, these results clearly indicate the ability of these anaerobic microbial granules to remove ortho chlorine from more than one PCB congener (2,3,4,5,6-CB, 2-CB and 2,6-CB). In summary, this anaerobic methanogenic microbial consortium removed chlorines from all three positions, i.e. ortho, meta and para, resulting in the complete dechlorination of 2,3,4,5,6-CB to biphenyl. Acknowledgements This study was supported by funds from the Great Lakes Protection Fund (GLPF FG2911054) and the Kellogg Foundation. We thank Dr. John Quensen, Michigan State University, East Lansing, Michigan, for assistance in setting up the PCB analytical system at MBI and Dr. Raj Rajan of EFX Systems Inc., Lansing, Michigan, for assistance and help in data acquisition and data analysis.

References Abramowicz DA (1990) Aerobic and anaerobic biodegradation of PCBs. A review. Crit Rev Biotechnol 10: 241–249 Ahmed M, Focht DD (1973) Degradation of polychlorinated biphenyls by two species of Achromobacter. Can J Microbiol 19: 47–52 Alder AC, Haggblom MM, Oppenbeimes SR, Young LY (1993) Reductive dechlorination of polychlorinated biphenyls in anaerobic sediments. Environ Sci Technol 27: 530–536 Assaf-Anid N, Nies L, Vogel T (1992) Reductive dechlorination of a polychlorinated biphenyl congener and hexachlorobenzene by vitamin B12. Appl Environ Microbiol 58: 1057–1060 Bedard DL, Wagner RE, Brennan MJ, Haberl ML, Brown JF Jr (1987) Extensive degradation of Aroclors and environmentally transformed polychlorinated biphenyls by Alcaligenes eutrophus H850. Appl Environ Microbiol 53: 1094–1102

677 Bhatnagar L, Wu W-M, Jain MK, Zeikus JG (1991) Design and function of biomethanation granules for hazardous waste treatment. In: Proceedings of the International Symposium on Environmental Biotechnology, Royal Flemish Society of Engineering, Ostend, Belgium. pp 1–10 Brown JF, Wagner RE, Feng H, Bedard DL, Brennan MJ, Carnahan JC, May RJ (1987a) Environmental dechlorination of PCBs. Environ Toxicol Chem 6: 579–593 Brown JF, Bedard DL, Brennan MJ, Carnahan JC, Feng H, Wagner RE (1987b) PCB dechlorination in aquatic sediments. Science 236: 709–713 Fish K, Principe J (1994) Biotransformation of Aroclor 1242 in Hudson River test tube microcosms. Appl Environ Microbiol 60: 4289–4296 Hickey WJ, Searles DB, Focht DD (1993) Enhanced mineralization of polychlorinated biphenyls in soil inoculated with chlorobenzoate-degrading bacteria. Appl Environ Microbiol 59: 1194–1200 Jain MK, Zeikus JG, Bhatnagar L (1991) Methanogens. In: Levell PN (ed) Anaerobic microbiology, a practical approach. Oxford University Press, Oxford, pp 223–245 Loganathan BG, Kannan K (1991) Time perspectives of organochlorine contamination in the global environment. Mar Pollut Bull 22: 582–584 Loganathan BG, Tanabe S, Hidaka Y, Kawano M, Hidaka H, Tatsukawa R (1993) Temporal trends of persistent organochlorine residues in human adipose tissue from Japan, 1928– 1985. Environ Pollut 81: 31–39 Mes J, Davies DJ, Doucet J, Weber D, Mullen ED (1993) Specific polychlorinated biphenyl congener distribution in breast milk of Canadian women. Environ Technol 14: 555–565 Natarajan MR, Wang H, Hickey R, Bhatnagar L (1995) Effect of oxygen and storage conditions on the metabolic activities of

polychlorinated biphenyl (PCBs) dechlorinating granules. Appl Microbiol Biotechnol 43: 733–738 Nies L, Vogel T (1991) Identification of proton sources for the microbial reductive dechlorination of 2,3,4,5,6-pentachlorobiphenyl. Appl Environ Microbiol 57: 2771–2774 Quensen JF, Boyd SA, Tiedje JM (1990) Dechlorination of four common polychlorinated biphenyl mixtures (Aroclors) by anaerobic microorganisms from sediments. Appl Environ Microbiol 56: 2360–2369 Quensen JF, Tiedje JM, Boyd SA (1991) Reductive dechlorination of polychlorinated biphenyls by anaerobic microorganisms from sediments. Science 242: 752–754 Rhee G, Roger CS, Bethoney CM, Bush B (1993) Dechlorination of polychlorinated biphenyl by Hudson River sediment microorganisms: specificity to the chlorination pattern of congeners. Environ Sci Technol 27: 1190–1192 Sokol RC, Kwon OS, Bethoney CM, Rhee GY (1994) Reductive dechlorination of polychlorinated biphenyls in St. Lawrence River sediments and variations in dechlorination characteristics. Environ Sci Technol 28: 2054–2064 Van Dort HM, Bedard DL (1991) Reductive ortho and meta dechlorination of polychlorinated biphenyl congener by anaerobic microorganisms. Appl Environ Microbiol 57: 1576–1578 Williams WA (1994) Microbial reductive dechlorination of trichlorobiphenyls in anaerobic sediment slurries. Environ Sci Technol 28: 630–635 Wu WM, Nye J, Hickey, RF, Bhatnager L (1993) Anaerobic granules developed for reductive dechlorination of chlorophenols and chlorinated ethylene. In: Proceedings of the 48th Industrial Waste Conference. Lewis, Boca Raton, Fla, pp 483– 493