Establishment of a Polychlorinated Biphenyl-Dechlorinating Microbial Consortium, Specific for Doubly Flanked Chlorines, in a Defined, Sediment-Free Medium

Abstract

Estuarine sediment from Charleston Harbor, South Carolina, was used as inoculum for the development of an anaerobic enrichment culture that specifically dechlorinates doubly flanked chlorines (i.e., chlorines bound to carbon that are flanked on both sides by other chlorine-carbon bonds) of polychlorinated biphenyls (PCBs). Dechlorination was restricted to the para chlorine in cultures enriched with 10 mM fumarate, 50 ppm (173 μM) 2,3,4,5-tetrachlorobiphenyl, and no sediment. Initially the rate of dechlorination decreased upon the removal of sediment from the medium. However, the dechlorinating activity was sustainable, and following sequential transfer in a defined, sediment-free estuarine medium, the activity increased to levels near that observed with sediment. The culture was nonmethanogenic, and molybdate, ampicillin, chloramphenicol, neomycin, and streptomycin inhibited dechlorination activity; bromoethanesulfonate and vancomycin did not. Addition of 17 PCB congeners indicated that the culture specifically removes double flanked chlorines, preferably in the para position, and does not attack ortho chlorines. This is the first microbial consortium shown to para or meta dechlorinate a PCB congener in a defined sediment-free medium. It is the second PCB-dechlorinating enrichment culture to be sustained in the absence of sediment, but its dechlorinating capabilities are entirely different from those of the other sediment-free PCB-dechlorinating culture, an ortho-dechlorinating consortium, and do not match any previously published Aroclor-dechlorinating patterns.

Microbial reductive dechlorination of polychlorinated biphenyls (PCBs) under anaerobic conditions has been shown to occur in freshwater, brackish, and marine sediments (reviewed in reference 2). The isolation of PCB-dechlorinating microorganisms is necessary to better understand the physiological mechanisms and catalysis of microbial PCB dechlorination. However, all reported attempts to isolate microorganisms that catalyze these reactions have been unsuccessful. Most of the PCB-dechlorinating enrichment cultures reported are derived from PCB-contaminated freshwater sediment (2, 4, 8, 17, 20, 22). The dechlorinating activities of these cultures are directed at meta and/or para chlorines and have not been demonstrated in the absence of soil or sediment. The transformation of biogenic halogenated organic compounds in estuarine and marine environments suggests that haloaromatic dechlorinators, including PCB dechlorinators, are present in these habitats. Recently it has been reported that microorganisms derived from estuarine sediment from Baltimore Harbor (Baltimore, Md.) are able to dechlorinate single PCB congeners (5) and Aroclor 1260 (19) by removal of meta, para, and ortho chlorines. A 2,3,5,6-tetrachlorobiphenyl (2,3,5,6-CB), ortho-dechlorinating culture has been enriched from these sediments, and the culture has since been transferred and sustained in a sediment-free medium (8).

Herein we report on the meta and para dechlorination of 2,3,4,5-CB and freshly added Aroclor 1260 in estuarine sediment from Charleston Harbor (Charleston, S.C.). A stable PCB-dechlorinating consortium (2,3,4,5-CB enrichment culture) was developed from that source of sediment. The culture attacks only doubly flanked chlorines of PCBs, chlorines that are flanked on each side by another chlorine bound to a carbon atom, indicating a high degree of specificity for PCB dechlorination.

MATERIALS AND METHODS

Sediment sample.

Sediment samples were collected with a petite Ponar grab sampler at a subsurface depth of 4 m in the Ashley River branch of Charleston Harbor (32°47.1′N, 79°57.5′W). Sediments had a black coloration and gelatinous texture. The combined contents of the sampler were transferred to 0.95-liter canning jars (Ball Corporation, El Paso, Tex.). The jars were filled to the top and immediately sealed with dome tops and ring seals to exclude air. The samples were stored at ambient temperature in the dark prior to use.

Culture procedures.

An estuarine medium (E-Cl) was prepared as described by Berkaw et al. (5), except that Na₂S · H₂O was not added. The final pH of the medium was 7.0. Charleston Harbor sediment, when used as a medium component, was dried, ground, and autoclaved twice for 1 h at 121°C on two consecutive days before use. The medium was autoclaved at 121°C for 30 min. Estuarine sediment from Charleston Harbor (2 ml) was inoculated into 8 ml of E-Cl medium. Aroclor 1260, in 10 μl of acetone, was added to a final concentration of 800 μg/g (dry weight) of sediment. The congeners 2,3,4,5-CB and 2,3,5,6-CB, in 10 μl of acetone, were added to a final concentration of 173 μM (50 ppm) or to 350 μM when they were added with the Aroclor. Cultures containing Aroclor were prepared in triplicate, and cultures maintained with single congeners only were prepared in duplicate. The Aroclor-containing cultures were supplemented monthly with 2.5 mM concentrations (each) of sodium acetate, propionate, and butyrate. All cultures were incubated at 30°C in the dark.

Specificity test of PCB dechlorination.

E-Cl medium was inoculated with the 2,3,4,5-CB enrichment culture (1%, vol/vol, transfer from sediment or sediment-free culture). Each culture contained 10 mM fumarate and one of following congeners (final concentration, 100 μM): 3-CB, 4-CB, 2,3-CB, 2,4-CB, 2,5-CB, 3,4-CB, 3,5-CB, 2,3,4-CB, 2,3,5-CB, 2,3,6-CB, 2,4,6-CB, 3,4,5-CB, 2,3,4,5-CB, 2,3,4,6-CB, 2,3,5,6-CB, 2,3,4,5,6-CB, and 2,4,5-2,4,5-CB.

Analytical procedures.

Cell growth in sediment-free enrichment cultures was monitored by measuring the change in optical density at 600 nm with a Spectronic 20D spectrophotometer (Milton Roy, Rochester, N.Y.). Methane concentrations were analyzed on a Hewlett-Packard (Atlanta, Ga.) 5890 series gas chromatograph (GC) equipped with a model RTX-624 capillary column {30 m by 0.53 mm (inside diameter [i.d.]) by 0.3 μm; Restek Corp., Bellefonte, Pa.} and a flame ionization detector. The temperatures of the oven, injector, and detector were 80, 240, and 325°C, respectively.

PCBs were extracted from enrichment cultures with ethyl acetate (high-performance liquid chromatography grade; Fisher Scientific, Pittsburgh, Pa.), and the solvent extracts were passed through a Florisil-copper column (5). PCBs were analyzed with a Hewlett-Packard 5890 series GC equipped with a model RTX-1 capillary column (30 m by 0.25 mm [i.d.] by 0.25 μm; Restek Corp.) and an Ni⁶³ electron capture detector as described previously (5). PCBs were identified by matching their GC retention times with those of authentic standards (99% purity; AccuStandard) and quantified with a 6- to 16-point calibration curve for each congener (5). The congeners 2,4-CB and 2,5-CB or 2,3,5,6-CB and 2,3,4,6-CB could not be separated by this method and so were reported together. Biphenyl was assayed on a Hewlett-Packard 6890 series GC equipped with a model HP-5MS capillary column (30 m by 0.25 mm [i.d.] by 0.25 μm; Hewlett-Packard) and a Hewlett-Packard 6890 series mass selective detector.

Analysis of Aroclor was done according to the method of Wu et al. (19). PCB extraction, preparation, and analysis by GC and electron capture detection were done in a manner similar to that described above. Aroclor 1260 and dechlorination products were identified by matching GC retention times with those of a customized PCB standard prepared by supplementing Aroclor 1260 with the dechlorination products observed in Woods Pond (L. A. Smullen, K. A. DeWeerd, D. L. Bedard, W. A. Fessler, J. C. Carnahan, and R. E. Wagner, p. 45–6, in Twelfth progress report of research and development program for the destruction of PCBs, General Electric Co., Schenectady, N.Y., 1993) or with a standard mixture composed of 3-3-CB, 3-4-CB, 3,5-3-CB, 3,5-4-CB, 2,4-3,5-CB, and 2,5-3,5-CB. Congener assignments were made according to those reported by Frame et al. (9). Each congener in the Aroclor mixture was quantified by use of a piecewise-fit calibration curve generated from standards at 4- to 8-point calibration levels. Congener and homolog distribution for each sample were calculated and reported in units of moles percent, which were used to calculate the chlorine/biphenyl ratios. Congener distributions for each enrichment culture with Aroclor and 2,3,4,5-CB (or 2,3,5,6-CB) were calculated after subtracting the peaks corresponding to 2,3,4,5-CB (or 2,3,5,6-CB) and their potential dechlorination products.

RESULTS

Dechlorination of Aroclor 1260 with Charleston Harbor sediment.

Aroclor 1260 is primarily composed of hexa- and heptachlorobiphenyls. A portion (25.3 ± 1.7% [mean ± standard deviation]) of the chlorines associated with Aroclor 1260 was removed by reductive dechlorination following 6 months of incubation with Charleston Harbor sediment (Fig. 1). The major congeners in Aroclor 1260 include 2,3,4-2,4,5-CB, 2,4,5-2,4,5-CB, 2,3,6-2,4,5-CB, 2,3,4,5-2,3,6-CB, 2,3,4,5-2,4,5-CB, and 2,3,5,6-2,4,5-CB, while the most prevalent dechlorination products were 2,4-2,4-CB, 2,4-2,5-CB, 2,5-2,5-CB, 2,4-2,6-CB, and 2,5-2,6-CB. This activity was heavily dominated by dechlorination of meta chlorines (60.1 ± 2.8%) with a trace of para dechlorination (4.3 ± 0.4%). The addition of congener 2,3,4,5-CB hastened the onset of dechlorination of the Aroclor 1260 by at least 1 month. Total dechlorination increased modestly to 29.8 ± 0.7% when 2,3,4,5-CB was added. Most of this increase was due to enhanced para dechlorination (11.2 ± 0.6%) since only a slight increase in meta dechlorination was observed (66.0 ± 2.1%). The onset of Aroclor dechlorination was also hastened by the addition of congener 2,3,5,6-CB, and meta dechlorination was slightly enhanced, but this congener failed to stimulate para dechlorination. No ortho dechlorination was observed under any conditions with the Charleston sediment. These results stand in contrast to those obtained with Baltimore Harbor sediment (19), where more overall dechlorination occurred due to ortho dechlorination.

FIG. 1.

Dechlorination of Aroclor 1260 in Charleston Harbor sediments incubated in E-Cl medium. Datum points are the means ± standard deviations of results from triplicate cultures. Chlorine/biphenyl ratios are presented for ortho (■)-, meta (□)-, and para (◊)-positioned chlorines. Shown are data from microcosms incubated with 800 μg of Aroclor 1260 per g (dry weight) of sediment (A), with Aroclor plus 350 μM 2,3,4,5-CB (B), and with Aroclor plus 350 μM 2,3,5,6-CB (C).

Establishment of a para-dechlorinating culture.

The Aroclor data clearly established that meta- and para-PCB dechlorination occurred with the Charleston Harbor sediments and that 2,3,4,5-CB could further stimulate this activity. An enrichment series with 10% (vol/vol) Charleston Harbor sediment and 173 μM 2,3,4,5-CB was started in E-Cl medium. Initially, both meta and para dechlorination of 2,3,4,5-CB was observed in these cultures. In the absence of a defined carbon source, congener 2,3,4,5-CB was dechlorinated to 2,3,5-CB, 2,4,5-CB, and 2,4-CB or 2,5-CB (Fig. 2A). The culture clearly favored a combination of meta and para dechlorination, with >75% of the parent congener being converted to 2,4-CB or 2,5-CB and no appearance of ortho-dechlorination products. At this point, 10 μl of actively dechlorinating sediment slurries were transferred into 10 ml of E-Cl medium containing 5% (wt/vol) sediment. After two sequential transfers, 2,3,5-CB became the sole dechlorination product of 2,3,4,5-CB and the meta-dechlorination activity was no longer observed (Fig. 2B). After four additional transfers, the amount of sediment added to the medium was reduced to 0.1% (wt/vol) and the specificity of dechlorination remained the same. When different carbon and energy sources (acetate, butyrate, crotonate, formate, fumarate, lactate, malate, propionate, pyruvate, and succinate) were added to the culture, fumarate enhanced the rate of para dechlorination of 2,3,4,5-CB by an average of 114%. Henceforth, the medium was supplemented with 10 mM fumarate as the sole carbon and energy source. The dechlorination pattern and rate have remained relatively constant under these conditions for more than 12 sequential transfers over the course of 25 months.

FIG. 2.

Dechlorination of 2,3,4,5-CB by a primary enrichment culture with Charleston Harbor sediment (A) and by the second sequential transfer (duplicate cultures) of the primary enrichment culture (B). PCB congeners represented are 2,3,4,5-CB (■), 2,3,5-CB (○), 2,4,5-CB (▵), and 2,4- or 2,5-CB (⧫).

Following six sequential transfers (1%, vol/vol) into E-Cl medium with 0.1% sediment (dry weight), the active culture was sequentially transferred five times into E-Cl medium containing 10 mM fumarate without added sediment. Compared with the dechlorination activity in the presence of sediment, the activity in the absence of sediment was slower (Fig. 3). However, after three sequential transfers, the rate and extent of dechlorination had returned to levels similar to those observed with cultures maintained with sediment (Fig. 4 and 5). No methane gas was detected in the sediment-free cultures. Following the growth of cultures, the results from which are depicted in Fig. 4, the para-dechlorination activity in sediment-free cultures could be retained in dilutions up to 10⁻⁵ (vol/vol), indicating that the microbial catalyst(s) accounted for at least 10⁵ cells per ml in the culture. Figure 4 also shows that most of the increase in optical density had occurred before most of the dechlorination had taken place. High optical densities were not achieved, but the data indicate that growth did continue at a slow pace while dechlorination continued.

FIG. 3.

Numbers of chlorines per biphenyl for the first generation of sediment-free 2,3,4,5-CB enrichment cultures (●) and the cultures with 0.1% sediment (▵). Datum points are the means of results from duplicate cultures.

FIG. 4.

Optical densities (●) and numbers of chlorines per biphenyl (□) of the fourth sequential transfer of the 2,3,4,5-CB-dechlorinating consortium in the absence of sediment. Datum points are the means of results from duplicate cultures.

FIG. 5.

Dechlorination of 2,3,4,5-CB by the sediment-free dechlorinating consortium with 2 mM BES (●), 100 μg of vancomycin per ml (■), or no inhibitor (▵). Datum points are the means of results from duplicate cultures.

Specificity of PCB dechlorination.

The specificity of PCB dechlorination of the 2,3,4,5-CB enrichment culture was investigated after five sequential transfers of the culture in the absence of sediment and in cultures maintained with sediment (Table 1). Of the 17 PCB congeners tested, only 2,3,4-CB, 2,3,4,6-CB, and 2,3,4,5,6-CB were meta dechlorinated to 2,3-CB, 2,4,6-CB, and 2,3,5,6-CB (or 2,3,4,6-CB), respectively. Congener 3,4,5-CB was para dechlorinated to 3,5-CB. None of the other congeners that contained at least one meta or para chlorine were dechlorinated. The results indicate that the 2,3,4,5-CB enrichment culture will remove only chlorines that are flanked on each side (doubly flanked) by other chlorines regardless of meta or para positioning. No difference in specificity was observed between the sediment and sediment-free enrichment cultures (Table 1). However, the extent of dechlorination was greater with sediment, especially with the congeners 2,4- and 2,4,6-CB. These results suggest that the sediment-free cultures selectively dechlorinate doubly flanked para chlorines.

TABLE 1.

Specificities of PCB dechlorination by the 2,3,4,5-CB para-dechlorinating culture^a

PCB congener	Dechlorination product(s) (mol%b)
	Without sediment	With sediment
2,3,4-CB	2,4-CB (22)	2,4-CB (98)
3,4,5-CB	3,5-CB (96)	3,5-CB (97)
2,3,4,5-CB	2,3,5-CB (86)	2,3,5-CB (97)
2,3,4,6-CB	2,4,6-CB (24)	2,4,6-CB (69)
2,3,4,5,6-CB	2,3,5,6-CB or 2,3,4,6-CB (5)	Not tested

^aThe cultures were grown for 90 days in E-Cl medium (3) containing 10 mM fumarate and one of tested congeners (100 μM) with and without 2% sediment (dry weight). No dechlorination was observed under any test conditions with 3-, 4-, 2,3-, 2,4-, 2,5-, 3,4-, 3,5-, 2,3,5-, 2,3,6-, 2,4,6-, 2,3,5,6-, and 2,4,5-2,4,5-CB. 

^bValues are averaged concentrations (in moles percent) from duplicate samples. 

Effects of inhibitors and antibiotics.

The dechlorination by the sediment-free 2,3,4,5-CB enrichment culture was also tested in the presence of bromoethanesulfonate (BES), molybdate, ampicillin, chloramphenicol, neomycin, streptomycin, and vancomycin. Two to ten millimolar sodium molybdate and 100 μg of ampicillin, chloramphenicol, neomycin, or streptomycin per ml inhibited the dechlorination activity and the growth of this consortium (data not shown). Sodium molybdate inhibits the growth of sulfate-reducing microorganisms but can also cause nonspecific inhibitions (12). The mechanism of the molybdate inhibition observed here was not examined further. Dechlorination of 2,3,4,5-CB was observed in the presence of 2 mM BES or 100 μg of vancomycin per ml, although the rate was lower (Fig. 5). These results are consistent with the absence of methanogenesis and further confirm that methanogens are not required for the dechlorination activities observed.

DISCUSSION

A 2,3,4,5-CB para-dechlorinating consortium was enriched from Charleston Harbor sediment. It could be transferred and sustained in sediment-free minimal medium containing 2,3,4,5-CB and fumarate. This is only the second PCB-dechlorinating consortium shown to maintain its dechlorinating activity in a defined medium without the addition of sediment. In contrast to the first sediment-free culture (8), this enrichment culture cannot remove ortho chlorines but can dechlorinate meta and para chlorines. However, this PCB-dechlorinating consortium attacks only doubly flanked chlorines of PCBs, and once established under sediment-free conditions, it selectively dechlorinates doubly flanked para chlorines.

Microbial PCB dechlorination has most often been observed as the removal of meta and para chlorines of PCBs (2), although there have been a few reported cases of ortho dechlorination under anaerobic conditions (5, 8, 15, 18, 19). It has been proposed that discrete dechlorinating microorganisms with distinct dehalogenating enzymes and congener regiospecificities are responsible for the various dechlorination processes (2–4, 6, 7, 13). However, this is only the second report of an anaerobic culture expressing a highly specific PCB dechlorination in a defined medium.

The 2,3,4,5-CB-dechlorinating consortium enriched from Charleston Harbor sediment para dechlorinated 2,3,4,5-CB to 2,3,5-CB, while meta dechlorination of 2,3,4,5-CB to 2,4,5-CB was lost after two transfers. Recently we have enriched for a strict meta-dechlorinating culture from the same Charleston Harbor sediment using 2,3,5,6-CB (our unpublished data). In contrast to the doubly flanked dechlorinating culture described here, the meta-dechlorinating culture does not para dechlorinate 2,3,4,5-CB and does not require flanking of the target chlorine. Cutter et al. were able to select for ortho dechlorination of 2,3,5,6-CB following transfer of a meta- and ortho-dechlorinating enrichment culture developed with Baltimore Harbor sediment (8). These results show that selection for specific dechlorinating activities is dependent upon the choice of congeners and inoculum source. Such an observation is consistent with the hypothesis that distinct consortia or individual species of microorganisms are responsible for specific dechlorination activities.

The combination of highly dilute, frequent transfers (0.1 to 1.0%, vol/vol) and transfer made immediately after 50% of the 2,3,4,5-CB had been para dechlorinated likely led to the selection of a nonmethanogenic, highly specific dechlorinating culture. Such conditions would probably not allow for the proliferation of slow-growing methanogens, and early transfer may have minimized the growth of species that would dechlorinate secondary congener products. The lack of inhibition due to BES, an inhibitor of methanogenesis, is not surprising in light of the culture being nonmethanogenic. However, PCB dechlorination is usually observed under methanogenic conditions (1, 10, 11, 14, 21, 22), and BES has been shown to inhibit dechlorination of certain PCB congeners (14, 16) or dechlorination processes (11). Such inhibition has been hypothesized to be due to the dechlorinating microorganisms using the sulfonic acid moiety of the BES, instead of a PCB, as an electron acceptor (23). The results presented here suggest that BES does not competitively inhibit the dechlorination of doubly flanked chlorines of PCBs by the microorganisms in this culture. Other examples of PCB dechlorination in the absence of methanogenesis include the meta dechlorination of Aroclor 1242 by pasteurized microorganisms (21) and the meta and para dechlorination of a mixture of congeners following growth of a mixed bacterial population on agar medium (10).

The combined effect of discrete PCB-dechlorinating pathways contributes to the overall dechlorination of Aroclors, and such combinations are referred to as dechlorination processes (2). There are at least five processes (Q, H′, H, P, and LP) that involve para dechlorination of PCBs (2, 4). Process Q removes virtually all para chlorines, regardless of the surrounding chlorine configuration, and meta chlorines of 2,3- and possibly 2,3,6-chlorophenyl groups. Process H′ is characterized by the removal of para chlorines in 3,4- and 2,4,5-chlorophenyl and meta chlorines in 2,3-, 2,3,4-, and 2,3,6-chlorophenyl. Process H is very much like H′ except that there is no dechlorination of 2,3-chlorophenyl groups. Process LP dechlorinates PCBs by removal of unflanked para chlorines of PCBs. Process P dechlorinates flanked and doubly flanked para-substituted PCBs containing 3,4-, 2,3,4-, 2,4,5-, and 2,3,4,5-chlorophenyl groups. The specific dechlorination activity of the 2,3,4,5-CB para-dechlorinating culture described here does not match any of these para-dechlorinating processes, but the microbial catalysts within this culture may contribute to process P. The dechlorination of Aroclor 1260 with Charleston Harbor sediment most closely resembles process N plus some process P. Process N is characterized by an almost exclusive loss of flanked and doubly flanked meta chlorines (2, 13). It is conceivable that the microorganisms belonging to the culture described here contributed to the Aroclor dechlorination reported in Fig. 1. However, it is impossible to document a highly specific activity such as the requirement of doubly flanked chlorines within an Aroclor mixture of congeners. Only after sequential transfers in the presence of the single congener, i.e., 2,3,4,5-CB, which stimulated the para dechlorination of Aroclor 1260, could this highly specific activity be characterized.

In summary, a PCB-dechlorinating consortium is sustainable in a defined, sediment-free medium and exhibits a very selective dechlorination activity. This is the first microbial consortium shown to para or meta dechlorinate a PCB congener in a defined and sediment-free medium. Furthermore, the culture exhibits a highly specific form of PCB dechlorination, namely, that of only doubly flanked chlorines, while favoring the para position. This is only the second report of a PCB-dechlorinating culture in a defined minimal medium. The culture clearly possesses a specific dechlorinating activity distinct from that of the other sediment-free PCB-dechlorinating culture, i.e., doubly flanked para or meta dechlorination versus ortho dechlorination.