Enhancement of microbial PCB dechlorination by chlorobenzoates,chlorophenols and chlorobenzenes

Young-Cheol Cho a, Ellen B. Ostrofsky b, Roger C. Sokol c, Robert C. Frohnhoefer a,G-Yull Rhee a;b;�

a Wadsworth Center, New York State Department of Health, Empire State Plaza, Albany, NY 12201, USAb School of Public Health, State University of New York at Albany, Albany, NY 12201, USA

c Center for Environmental Health, New York State Department of Health, Troy, NY 12180, USA

Received 31 January 2002; received in revised form 28 May 2002; accepted 29 May 2002

First published online 3 August 2002

Abstract

We investigated the effects of chlorobenzoates (3-, 2,3-, 2,4-, 2,5-, 2,3,5- and 2,4,6-chlorobenzoate), chlorophenols (2,3-, 3,4-, 2,5-, 2,3,6-and penta-chlorophenol), and chlorobenzenes (1,2-, 1,2,3-, 1,2,4- and penta-chlorobenzene) on polychlorinated biphenyl (PCB)dechlorination and on the enrichment of PCB-dechlorinating microorganisms. When the natural microbial populations eluted fromSt. Lawrence River sediments were enriched with each of the 15 haloaromatic compounds (HACs) in PCB-free sediments, PCB-dechlorinating microorganisms were found in all but pentachlorophenol-amended sediments. Similarly, dechlorinating microorganismswere also found in PCB-spiked sediments amended with all HACs, except for those with pentachlorophenol. In HAC-amended PCBsediments there was a long lag in PCB dechlorination until the HACs were reduced to a plateau level. Despite this lag, once PCBdechlorination started it was faster in the HAC-amended sediments compared to the unamended controls. The overall extent of PCBdechlorination was significantly enhanced by all HACs except pentachlorophenol and pentachlorobenzene, but the extent as well as thepattern of the enhancement varied. Of the 13 effective HACs, six (2,3-, 2,4- and 2,4,6-chlorobenzoates; 3,4- and 2,3,6-chlorophenols; and1,2,3-chlorobenzene) enhanced only meta-dechlorination, whereas five (3-chlorobenzoate; 2,3- and 2,5-chlorophenols; and 1,2- and 1,2,4-chlorobenzenes) increased both meta- and para-dechlorination, and two (2,5- and 2,3,5-chlorobenzoates) promoted overall, substitutionnon-specific dechlorination. When the maximum extent of dechlorination was plotted against the highest number of PCB-dechlorinatingmicroorganisms for each HAC, there was a linear relationship (P6 0.01), suggesting that dechlorination enhancement was related to theincrease in their population size. However, there was also evidence to suggest that different dechlorinating microorganisms wereselected. ; 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.

Keywords: PCB dechlorination; Dehalogenation; Dechlorinating microorganism; Chlorobenzoate; Chlorobenzene; Chlorophenol

1. Introduction

The risk of exposure to polychlorinated biphenyls(PCBs) stems largely from the highly chlorinated conge-ners which are recalcitrant to aerobic degradation [1,2]. Itmay be possible to overcome this problem of recalcitranceif the level of chlorination can be reduced. Investigationsin recent years have demonstrated the wide-spread oc-currence of microorganisms in anaerobic sediments thatcan remove chlorines from PCBs [3^8]. Dechlorination is

mainly observed at meta and para positions and is deter-mined by the pattern of chlorine substitution on the bi-phenyl ring [3].There are various kinds of dechlorinating microorgan-

isms with di¡erent capabilities as evidenced by the di¡er-ent dechlorination patterns produced when PCB-contami-nated sediments were treated with ethanol, antibiotics,pasteurization [4,5], or di¡erent temperatures [6]. Usinga combination of the serial dilution and most probablenumber (MPN) methods, we were able to separate twodi¡erent types of dechlorinating populations in St. Law-rence River sediments. Each contributed approximatelyhalf the total PCB dechlorination. One group, which pref-erentially dechlorinated mostly meta-substituted congenerssuch as 2,5,2P,5P-, 2,3,2P,5P- and 2,5,2P-chlorobiphenyls,

0168-6496 / 02 / $22.00 ; 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.PII: S 0 1 6 8 - 6 4 9 6 ( 0 2 ) 0 0 3 2 2 - 7

* Corresponding author. Tel. : +1 (518) 473 8035;Fax: +1 (518) 486 2697.

E-mail address: rhee@wadsworth.org (G.-Y. Rhee).

FEMSEC 1398 16-9-02

FEMS Microbiology Ecology 42 (2002) 51^58

www.fems-microbiology.org

was about two orders of magnitude less abundant than theother [7]. We also identi¢ed two di¡erent dechlorinatingpatterns when sediments were treated with the metabolicinhibitor of methanogens, 2-bromoethanesulfonate. Whenmethanogenesis was inhibited, PCB dechlorination wassigni¢cantly reduced, mainly due to the absence of meta-dechlorination in certain congeners [8]. The results of thesestudies clearly demonstrate that there are many dechlori-nating microorganisms with di¡erent capabilities.There is evidence that enrichment with non-PCB haloar-

omatic compounds (HACs) induced PCB-dechlorinatingactivities in sediment microbial communities. When thenatural microbial communities were enriched with 2,3,6-trichlorobenzoate in Dutch sediments, they were able todechlorinate hexachlorobiphenyls without a lag [9]. Amethanogenic microbial consortium enriched with tri-chlorobenzene in polluted sediments was also capable ofdechlorinating PCBs [10]. Chlorophenol-adapted consortiadechlorinated PCBs, but the rate varied with reducingconditions [11]. Such cross-enrichment is probably dueto the fact that many dehalogenating enzymes are notspeci¢c to a single compound.Although PCB-dechlorinating microorganisms have not

been isolated so far, dehalogenating isolates for otherhalogenated compounds have demonstrated a broad de-grading capability. For example, an anaerobic spore-form-ing microorganism enriched and isolated from a compostsoil on the basis of its ability to grow with 2,3-dichloro-phenol as its electron acceptor, was able to grow using 3-chloro-4-hydroxybenzoate as an alternate electron accep-tor [12]. Many species belonging to the genus Desul¢tobac-terium can also use more than one halogenated compoundas the electron acceptor, with low molecular mass fattyacids as electron donors. A Desul¢tobacterium sp. thatwas isolated from a tetrachloroethene-dechlorinating cul-ture was capable of dechlorinating tetrachloroethene,chlorophenols and 3-chloro-4-hydroxy-phenyl acetatewhen grown with pyruvate, butyrate, formate and succi-nate as electron donors [13,14].In the present study, we investigated various isomers of

chlorobenzoate, chlorobenzene and chlorophenol for theire¡ects on PCB dechlorination and for their ability to en-rich PCB-dechlorinating microbial populations.

2. Materials and methods

2.1. Sediment preparation

PCB-contaminated native sediments were collected fromthe St. Lawrence River (NY, USA), adjacent to the indus-trial outfall of the Reynolds Metal Company (see [15] forsite description). This river is principally contaminatedwith Aroclor 1248 from three industrial sources: GeneralMotors Central Foundry, Reynolds Metal Company andthe Aluminum Corporation of America where they used it

as a hydraulic or heat transfer £uid [15]. PCB-free sedi-ment was collected from its tributary, the Grasse River(NY, USA). Sediment cores were collected and returnedto the laboratory on ice and stored at 2‡C until used.For laboratory studies, PCB-contaminated sediments

were prepared in serum vials by spiking PCB-free sedi-ments with Aroclor 1248 in hexane to yield a concentra-tion of 300 Wg (g sediment)31 (dry weight basis). The hex-ane was removed by evaporation. The sediments were thenmade into slurries by adding reduced synthetic mineralmedium [16] in an anaerobic chamber (N2 :H2 :CO2 {85:10:5} atmosphere; Coy Laboratory Products, Ann Arbor,MI, USA). The slurry contained 10 g sediments (dryweight) in a 50 ml volume and resazurin (0.0001%, w/v)as a redox indicator. All vials were sealed with Te£on-lined stoppers and aluminum crimp seals and autoclavedat 121‡C for 40 min on three successive days.Sediment slurries in serum vials were autoclaved and

then amended with each of the non-PCB HACs (3-, 2,3-,2,4-, 2,5-, 2,3,5- and 2,4,6-chlorobenzoate; 2,3-, 3,4-, 2,5-,2,3,6- and penta-chlorophenol ; and 1,2-, 1,2,3-, 1,2,4- andpenta-chlorobenzene) in acetone stock solutions (0.5%, v/v) to yield a ¢nal concentration of 200 Wg (g sediment)31.To obtain an inoculum of natural sediment populations,10 g of the Reynolds site sediments was ¢rst made intoslurries in 100 ml of sterilized and reduced synthetic min-eral medium, and then stirred for 1 h. After allowing theseslurries to settle for 20 min, the supernatant was used asthe inoculum. After inoculation, the vials were incubatedstatically at room temperature in the dark. All experimentswere set up in duplicate.

2.2. Experimental set-up

To investigate the potential enrichment of PCB-dechlo-rinating microorganisms by non-PCB HACs, the elutedsediment microorganisms were inoculated into PCB-freesediments spiked with each HAC. These cultures were ac-climated by three successive transfers into fresh sedimentsafter 5-week incubation periods. This transfer period wasbased on our preliminary studies, which showed that formost HACs, the HAC concentrations decreased to a pla-teau level in 5 weeks with no further reduction. HAC-freesediments, with and without acetone, were inoculated withthe same inoculum to serve as the controls. These accli-mated mixed cultures were then analyzed for PCB-dechlo-rinating microorganisms by the MPN technique usingPCB-spiked sediments [17] as described below.To determine whether direct enrichment with HACs en-

hanced PCB dechlorination, Aroclor 1248-spiked sedi-ments were amended with each HAC. These sedimentswere then inoculated with sediment microorganisms elutedfrom contaminated river sediments from the Reynolds sitein the St. Lawrence River. HAC-free sediments spikedwith Aroclor 1248, with and without sediment microor-ganisms (sterilized by autoclaving), served as the controls.

FEMSEC 1398 16-9-02

Y.-C. Cho et al. / FEMS Microbiology Ecology 42 (2002) 51^5852

2.3. MPN determination

The number of PCB-dechlorinating microorganisms wasestimated by a ¢ve-vial MPN series. The MPN vials con-tained sediments spiked with 2,3,4- and 2,5,3P,4P-chlorobi-phenyls and these congeners were extracted and analyzedafter 15 weeks of incubation as described previously [17].MPN vials showing dechlorination were counted as pos-itive when dechlorination removed greater than 5% of thetotal chlorines. MPN values were determined from thefrequency of positive vials using the Microsoft Excel pro-gram [18] and then normalized to g dry weight of sedi-ment. The 95% con¢dence interval was calculated by thealgorithm of estimating the MPN. We estimated the num-ber of PCB-dechlorinating microbial populations in onlyone of the replicates. This was because of logistical di⁄-culties: to estimate the MPN for a single time point foreach replicate, it required 400 vials for PCB extraction andanalysis.

2.4. PCB extraction and analysis

For congener-speci¢c analysis, sediments were extractedusing an Accelerated Solvent Extractor system (US Envi-ronmental Protection Agency Method 3545; Dionex, Sun-nyvale, CA, USA). Sulfur was removed by tetrabutylam-monium hydrogen sulfate. Samples were cleaned througha Florisil column as described previously [19]. QuantitativePCB analysis was performed on a Hewlett-Packard 5890II gas chromatograph (Hewlett-Packard, Avondale, PA,USA) equipped with a 63Ni electron capture detectorwith a Rtx0-5 fused silica capillary column (Restek, Bel-lefonte, PA, USA) for Aroclor samples; an Apiezon-Lcolumn (Restek) was used to analyze the 2,3,4- and2,5,3P,4P-chlorobiphenyl samples from the MPN vials.The gas chromatographic conditions were identical tothose described previously [19].All of the chromatography data were collected and pro-

cessed on a microcomputer using a ChromPerfect chroma-tography data system (Justice Innovations, MountainView, CA, USA). The PCB congeners were identi¢edand quantitated using a calibration standard containinga 1:1:1:1 mixture of Aroclors 1016, 1221, 1254 and 1260(0.2 Wg ml31 of each in hexane). Peaks were identi¢ed andcalibrated as previously described [7,15,17,19]. Our analy-sis resolved 98 peaks, representing 127 congeners. Thisstandard mixture was used for quality assurance and con-trol, checking instrument performance, reproducibility andsensitivity. PCB extracts run on the Apiezon-L columnwere identi¢ed and quantitated with a calibration standardcomposed of a mixture of 47 individually weighed authen-tic single congener standards (99% purity; AccuStandard,New Haven, CT, USA). A dilute-to-match procedure [20]was used to ensure that all samples were analyzed withinthe linear range of the calibration standard. UninoculatedPCB-spiked sediment controls, set up at the beginning of

the experiment and sampled at every time point, were usedto monitor extraction e⁄ciency. The precision of the gaschromatograph was also monitored by running duplicatesamples. The PCB congeners in each sample were calcu-lated and expressed as mole percent. The average numberof total chlorines per biphenyl and the average number ofortho-, meta- and para-chlorines were individually calcu-lated from the product of the average number of chlorinesand the molar concentration for each peak divided by thetotal molar concentration summed over all peaks. Themaximum extent of dechlorination was calculated as anaverage of four values that were duplicate measurementsat two di¡erent time points during the plateau phase inwhich there was no further dechlorination. All calculationswere based on the conservation of the biphenyl moietyand the assumption that coeluting congeners were presentin equal proportions [15,21].

2.5. Analysis of HACs

For analysis of chlorobenzoates, the samples were cen-trifuged at 10 000Ug for 10 min to remove particles [22].Chlorophenols were extracted with an equal volume ofacetonitrile by shaking overnight on an orbital shaker.The extract was then centrifuged at 10 000Ug for 10 min[23]. Chlorobenzoates and chlorophenols in the extractswere analyzed using a high-pressure liquid chromatograph(Model 625; Waters, Milford, MA, USA) equipped with aphotodiode array detector (Model 996; Waters) and areverse-phase C18 column (3.9U150 mm; NovaPak C18 ;Waters). Chlorobenzoates were eluted using a linear gra-dient of two solvents, 50 mM sodium acetate (pH 4.5) andacetonitrile [22]. For chlorophenols, acetonitrile^water^acetic acid bu¡er was used as a mobile phase with nogradient [23]. All of the high-pressure liquid chromato-graphic data were collected and processed on a micro-computer using a Waters Millennium32 software system(Waters). Chlorobenzenes were extracted with acetone^hexane (1:1, v/v) by shaking on an orbital shaker over-night. Distilled water was added for phase separation andthe residual water in hexane was removed with anhydroussodium sulfate. Chlorobenzenes were analyzed using a gaschromatograph (5890II; Hewlett-Packard) equipped witha 30m Rtx0-5 column and an electron-capture detector.The initial oven temperature of 60‡C was maintained for 3min and then increased at a rate of 5‡C min31 to 180‡C.The ¢nal temperature was maintained for 3 min [10]. TheChromPerfect chromatography data system was used toobtain and process gas chromatographic data.

3. Results

Whether HACs are capable of enriching PCB-dechlori-nating microorganisms was determined by the MPN tech-nique using PCB-free sediments amended with HACs.

FEMSEC 1398 16-9-02

Y.-C. Cho et al. / FEMS Microbiology Ecology 42 (2002) 51^58 53

Similar studies were also carried out using the same cleansediments spiked with Aroclor 1248 and HACs. All cul-tures acclimated to each of the 15 HACs in PCB-free sedi-ments, except the one adapted to pentachlorophenol, werecapable of enriching PCB-dechlorinating microorganisms(Table 1). Although the MPN values in this table do notrepresent the relative e¡ectiveness of each of the HACs toenrich dechlorinating microorganisms because these valueswere measured at the same time point and disregard po-tential di¡erences in their growth stages, these resultsindicated that HACs were capable of enriching PCB-de-chlorinating microorganisms even when PCBs were notpresent. It is interesting to note that dechlorinating micro-organisms were also found, albeit in lower numbers, inclean sediments (free of both PCBs and HACs) thatwere initially inoculated with the same elutriate fromPCB-contaminated sediments and successively transferredinto the clean sediments in the same fashion as for theHAC-enriched sediments. Acetone had no e¡ect on PCBdechlorination or enriching dechlorinators, since the MPNvalues in clean sediments with and without acetoneshowed no di¡erences and nor did the values for Aroclor1248-spiked sediments with and without acetone (Table 1).To investigate the e¡ect of HACs on PCB dechlori-

nation, sediments spiked with Aroclor 1248 were amendedwith each HAC. HAC-free Aroclor 1248 sediments withinoculum were used as live controls. The same sedimentsthat had been amended with acetone or autoclaved servedas the acetone and sterile controls, respectively. The ace-tone control showed little di¡erence from the live controland no dechlorination was detected in the sterile control.Of the 15 HACs tested, all but pentachlorophenol andpentachlorobenzene enhanced the overall extent of Aro-clor 1248 dechlorination (Fig. 1). In pentachlorophenolsediments, no dechlorination was found, whereas in penta-chlorobenzene sediments, the level of PCB dechlorinationat the end of a 42-week incubation period was not di¡er-ent from the level in sediments with Aroclor 1248 alone

(P=0.09). With the 13 other HACs (3-, 2,3-, 2,4-, 2,5-,2,3,5- and 2,4,6-chlorobenzoates; 2,3-, 3,4-, 2,5- and2,3,6-chlorophenols; and 1,2-, 1,2,3- and 1,2,4-chloroben-zenes), the overall PCB dechlorination was signi¢cantlyenhanced (P6 0.05; Table 2), ranging from 28 to 37%of the total chlorines, in contrast to 24% in those withAroclor 1248 alone.No ortho-dechlorination was observed with or without

HAC amendments. When the HAC-enhanced PCB de-chlorination was analyzed with respect to the chlorinesubstitution position, it became quite clear that the en-hancement e¡ects varied according to the compound (Ta-ble 2). Of the 13 e¡ective HACs, six (2,3-, 2,4- and 2,4,6-chlorobenzoates; 3,4- and 2,3,6-chlorophenols; and 1,2,3-chlorobenzene) enhanced only meta-dechlorination,

Table 1MPN values of PCB-dechlorinating microorganisms in PCB-free sediment enriched with various non-PCB HACsa

Treatment Dechlorinator number(U105 MPN (g sediment)31)

Treatment Dechlorinator number(U105 MPN (g sediment)31)

Aroclor 1248 1.7 (0.50^5.4)b 2,4,6-trichlorobenzoate 2.5 (0.75^8.1)Aroclor 1248+acetonec 1.2 (0.35^3.8) 2,3-dichlorophenol 17 (5.0^54)Clean sediment 0.25 (0.075^0.81) 3,4-dichlorophenol 2.5 (0.75^8.1)Clean sediment+acetone 0.40 (0.12^1.3) 2,5-dichlorophenol 1.7 (0.50^5.4)

2,3,6-trichlorophenol 0.68 (0.21^2.3)3-chlorobenzoated 1.2 (0.35^3.8) pentachlorophenol b.d.e

2,3-dichlorobenzoate 3.5 (1.1^12) 1,2-dichlorobenzene 2.5 (0.75^8.1)2,4-dichlorobenzoate 1.2 (0.35^3.8) 1,2,3-trichlorobenzene 0.54 (0.16^1.8)2,5-dichlorobenzoate 2.5 (0.75^8.1) 1,2,4-trichlorobenzene 2.5 (0.75^8.1)2,3,5-trichlorobenzoate 6.5 (2.0^22) pentachlorobenzene 54 (16^180)

aThe initial number of PCB-dechlorinating microbial populations was 6.5U103 MPN (g sediment)31 in all samples.bValues in parentheses indicate 95% con¢dence limits, estimated by the algorithm of calculating the MPN.c0.5% (v/v).dSediments with non-PCB HACs contained 0.5% (v/v) acetone.eBelow the detection limit (103 MPN (g sediment)31).

Fig. 1. Percent chlorine removed in Aroclor 1248-spiked sedimentsamended with non-PCB HACs. The bars indicate standard deviation.

FEMSEC 1398 16-9-02

Y.-C. Cho et al. / FEMS Microbiology Ecology 42 (2002) 51^5854

whereas ¢ve (3-chlorobenzoate; 2,3- and 2,5-chlorophe-nols ; and 1,2- and 1,2,4-chlorobenzenes) signi¢cantly in-creased both meta- and para-dechlorination (P6 0.05) andtwo (2,5- and 2,3,5-chlorobenzoates) increased overall de-chlorination (P6 0.01). No HAC promoted para-dechlo-rination alone.The level of meta enhancement was not statistically sig-

ni¢cant between the 11 HACs, except for 2,3-chlorophe-nol. The extent of para enhancement by the ¢ve HACswas, however, signi¢cantly di¡erent.Of the e¡ective HACs, 2,3-dichlorophenol enhanced de-

chlorination most. With its amendment, 70% of metachlorines were removed, as opposed to just 49% in sedi-ments without it. Similarly, 39% of para chlorines wereremoved compared to only 13% in the Aroclor 1248 con-trol.During the incubation, HAC concentrations were mea-

sured at 0, 3, 5 and 11 weeks after inoculation. Fig. 2shows a typical time course of PCB dechlorination andHAC utilization for each of the three HAC classes. Thechange in HAC concentrations showed that they wererapidly utilized before reaching a plateau, which appearedto coincide with a lag in PCB dechlorination. In contrast,there was no such lag in PCB dechlorination in HAC-freesediments. These results suggested that the dechlorinationlag in HAC-amended sediments was due to the preferen-tial utilization of the enrichment compounds. However,once PCB dechlorination started, its rate was much fasterin HAC-amended sediments than in HAC-free sediments.The plateau level of dechlorination was also signi¢cantlylower than that in HAC-free sediments, indicating greaterdechlorination of PCBs (Fig. 2).

When the dechlorinator MPN was plotted against themaximum extent of PCB dechlorination at the observedplateau phase with each HAC amendment (Fig. 3), astrong and signi¢cant statistical correlation was found(r=0.70, P6 0.01). Due to the overlap in the 95% con¢-dence intervals surrounding the dechlorinator MPN, wewere unable to determine individual di¡erences betweenthe di¡erent amendments. However, there was no overlapin the 95% con¢dence intervals for the MPN values in theAroclor 1248 sediments and the same sediments with 2,3-dichlorophenol, the HAC amendment that exhibited thegreatest extent of dechlorination. Given this and thestrong statistical association between dechlorinator num-ber and the extent of dechlorination, the results suggestthat the extent of dechlorination may be a function of thepopulation size and that the dechlorination enhancementobserved in sediments amended with HACs may be dueto increased numbers of PCB-dechlorinating microorgan-isms.Despite this statistical association, the composition of

dechlorinating communities also appears to be importantbecause the extent of dechlorination can be quite di¡erenteven at the same MPN value. For example, the extent ofdechlorination was signi¢cantly di¡erent (P6 0.005) be-tween 2,3-dichlorophenol and 3-chlorobenzoate enrich-ment even though the MPN values were the same.

4. Discussion

In the present study, many non-PCB HACs enhancedPCB dechlorination. This was not surprising, since many

Table 2Enhancement of PCB dechlorination by HACs (numbers in parentheses are P values)

Treatment Meta Cls/BP Para Cls/BP Total Cls/BP

Autoclaved control 1.29U 0.004a 1.01 U 0.002 3.96U 0.001Aroclor 1248 0.68U 0.012 (0.461b) 0.89 U 0.011 (0.498) 3.01U 0.005 (0.742)Aroclor 1248+acetone 0.66U 0.025 0.88 U 0.019 3.01U 0.0013-chlorobenzoatec 0.48U 0.048 (0.041) 0.76 U 0.029 (0.044) 2.72U 0.015 (0.002)2,3-dichlorobenzoate 0.50U 0.049 (0.049) 0.74 U 0.069 (0.114) 2.73U 0.025 (0.004)2,4-dichlorobenzoate 0.48U 0.011 (0.011) 0.86 U 0.027 (0.638) 2.82U 0.033 (0.016)2,5-dichlorobenzoate 0.47U 0.086 (0.091) 0.73 U 0.058 (0.079) 2.69U 0.025 (0.003)2,3,5-trichlorobenzoate 0.50U 0.053 (0.058) 0.85 U 0.002 (0.199) 2.82U 0.023 (0.008)2,4,6-trichlorobenzoate 0.52U 0.036 (0.042) 0.86 U 0.001 (0.373) 2.80U 0.024 (0.007)2,3-dichlorophenol 0.39U 0.005 (0.004) 0.62 U 0.003 (0.003) 2.52U 0.009 (0.000)3,4-dichlorophenol 0.52U 0.008 (0.017) 0.84 U 0.019 (0.219) 2.86U 0.025 (0.016)2,5-dichlorophenol 0.47U 0.043 (0.031) 0.74 U 0.019 (0.018) 2.64U 0.098 (0.034)2,3,6-trichlorophenol 0.52U 0.017 (0.022) 0.77 U 0.032 (0.062) 2.72U 0.017 (0.002)pentachlorophenol 1.27U 0.001 (n.d.)d 0.96 U 0.015 (n.d.) 3.93U 0.001 (n.d.)1,2-dichlorobenzene 0.50U 0.001 (0.012) 0.68 U 0.016 (0.008) 2.68U 0.032 (0.005)1,2,3-trichlorobenzene 0.50U 0.007 (0.013) 0.70 U 0.059 (0.056) 2.69U 0.043 (0.009)1,2,4-trichlorobenzene 0.47U 0.011 (0.010) 0.77 U 0.002 (0.016) 2.75U 0.032 (0.008)Pentachlorobenzene 0.67U 0.021 (0.789) 0.83 U 0.032 (0.257) 2.95U 0.025 (0.091)

aStandard deviation.bP values are the t-test results of residual meta, para, or the total chlorines biphenyl31 (Cls/BP) between sediments with Aroclor 1248+acetone and thesame sediments amended with HACs. Bold face indicates P6 0.05.cSediments with non-PCB HACs contained 0.5% (v/v) acetone and Aroclor 1248 (300 Wg (g sediment)31).dNot determined because the Aroclor 1248 in the sediments with pentachlorophenol was not dechlorinated.

FEMSEC 1398 16-9-02

Y.-C. Cho et al. / FEMS Microbiology Ecology 42 (2002) 51^58 55

enzymes, including those involved in dehalogenation, arenot restricted to a single substrate [24^26]. For example,cell extracts of Desulfomonile tiedjei DCB-1, a 3-chloro-benzoate-dechlorinating bacterium, dehalogenate tetra-chloroethylene [24] and this strain can also dehalogenatechlorophenols [25]. Anaerobic bacteria in pond sediment

acclimated to dehalogenate chlorophenols rapidly de-chlorinate chloroanilines [26]. Additionally, Desul¢tobacte-rium sp. strain PCE1, isolated from a tetrachloroethene-dechlorinating enrichment culture, can obtain energy fromthe reductive dechlorination of ortho-chlorinated phenols[13].Earlier studies showed that ‘priming’ with certain PCB

congeners or bromobiphenyls enhanced certain types ofdechlorination [27^30]. However, chlorinated and £uori-dated benzoates did not prime Aroclor 1260 dechlori-nation; only certain brominated and iodiated benzoatesexhibited any e¡ects [22]. Although these observations ap-pear to be contradictory to our present results, it is pos-sible that the enrichment e¡ects vary between Aroclorsand also with the kinds of microorganisms indigenous toa given site [8,15].We did not determine the e¡ects of PCBs, HACs or

acetone on the total sediment microbial populations inthe present study. However, our previous study [17]showed that there was no di¡erence in the total microbialbiomass (estimated by phospholipid phosphate concentra-tion) between PCB-free and PCB-spiked sediments, despitea signi¢cant increase in dechlorinating microbial popula-tions in the latter. This was due to the fact that the de-chlorinator populations, even at their maximum, were twoorders of magnitude less than the total sediment popula-tions, accounting for less than 1%.The pattern of no growth and no dechlorination found

with pentachlorophenol was probably due to its toxicity tosediment microorganisms. Pentachlorophenol has beenfound to completely inhibit the growth of D. tiedjeiDCB-1 at 27 mg l31 [25], half the concentration used inthe present study (40 mg l31).

Fig. 2. A typical time course of PCB dechlorination and non-PCB HACutilization in sediments amended with chlorobenzoates (A), chlorophe-nols (B) and chlorobenzenes (C). The extent of dechlorination (chlorinesbiphenyl31) and the concentration of HAC are average values ( U stan-dard deviation).

Fig. 3. Correlation between the highest number of dechlorinating micro-organisms and the maximum extent of PCB dechlorination in Aroclor1248-spiked sediments amended with non-PCB HACs.

FEMSEC 1398 16-9-02

Y.-C. Cho et al. / FEMS Microbiology Ecology 42 (2002) 51^5856

Although the lag between the decrease of HACs and theonset of PCBs could have been due to the competitiveinhibition of PCB-dechlorinating microorganisms byHAC utilizers, it appears that the same microorganismsutilizing HACs are also involved in PCB dechlorination,because the number of PCB-dechlorinating microorgan-isms in HAC-amended sediments appeared to be higherthan those in sediments with Aroclor alone (data notshown). A similar lag in the onset of PCB dechlorinationhas also been reported when PCB dechlorination was‘primed’ using bromobenzoate [22].The statistical association we observed between the

maximum extent of dechlorination and the total numberof PCB-dechlorinating microorganisms (Fig. 3) is intrigu-ing in light of our previous ¢ndings. We reported thatPCB dechlorination is directly coupled to the growth ofdechlorinating microorganisms [17] and that the maximumdechlorination and the dechlorination rate were correlatedto growth rate at various PCB concentrations [19]. How-ever, the statistical relationship needs further investigationfor two reasons. Firstly, most data points in Fig. 3 areclustered in the mid range. If the points for 2,3-dichloro-phenol, pentachlorobenzene and Aroclor 1248 with andwithout acetone were excluded, the relationship would be-come insigni¢cant. Secondly, as mentioned above, we needto better understand the composition of dechlorinatingcommunities, because it appears that the enhancement ofdechlorination can be quite di¡erent even at the sameMPN numbers, as seen in Fig. 3.The present ¢nding that di¡erent HACs may promote

di¡erent types of dechlorination could re£ect the enrich-ment of microorganisms with di¡erent dechlorinating ca-pabilities. Earlier investigations showed that conditionssuch as the metabolic inhibitor, temperature, or PCB con-centration, select di¡erent dechlorinating microbial popu-lations, revealing di¡erent patterns of PCB dechlorinationdepending on the chlorine-substitution pattern of conge-ners [6,8,31]. For example, when methanogenesis was in-hibited by a metabolic inhibitor [8], the level of dechlori-nation was reduced, mainly owing to the absence of meta-dechlorination in certain congeners. At low concentrationsof Aroclor 1248, meta-rich congeners, such as 2,5,2P-,2,5,2P,5P- and 2,5,2P,3P-chlorobiphenyls, are intermediateproducts of dechlorination accumulated without furtherdechlorination [31]. Therefore, if we can selectively enrichappropriate dechlorinating organisms, it may be possibleto target recalcitrant congeners speci¢c to individual sitesusing the appropriate enrichment. For such site-speci¢c ortargeted remediation, we need to de¢ne the selectivity orenrichment characteristics of various HACs. It is also im-portant to understand whether or not this selectivity ofenrichment compounds is site-dependent, since the compo-sition of di¡erent indigenous microbial communities canbe quite variable.Although the enrichment technique is di⁄cult to use in

most open systems, the contained environment of a con-

¢ned disposal facility may prove to be an ideal site for itsapplication. Either the direct on-site enrichment of indig-enous microorganisms or the application of enriched mi-croorganisms produced in mass culture could be utilized.For any compound to be considered for enrichment, it isimportant that it meets certain criteria, including: (a) thee¡ective concentration for enrichment should be low, (b) itshould be readily degraded without producing toxic inter-mediates or residues, and (c) the residual concentration ofthe compound should be below the level acceptable fromthe standpoint of public and ecosystem health.

Acknowledgements

This work was supported by grants from the Environ-mental Protection Agency (R825449), Hudson RiverFoundation (R005/97A), NYS Great Lakes ProtectionFund (012545), and National Institute of EnvironmentalHealth Science Superfund Basic Research Program(ES04913). We thank Charlotte M. Bethoney for her ex-cellent technical assistance. We thank Ms. Leslie Eiseleand the Biochemistry Core Facility of Wadsworth Centerfor HPLC analysis.

References

[1] Safe, S.H. (1994) Polychlorinated biphenyls (PCBs): environmentalimpact, biochemical and toxic responses, and implications for riskassessment. Crit. Rev. Toxicol. 24, 87^149.

[2] Mohn, W.W. and Tiedje, J.M. (1992) Microbial reductive dehaloge-nation. Microbiol. Rev. 56, 482^507.

[3] Rhee, G-Y., Sokol, R.C., Bethoney, C.M. and Bush, B. (1993) De-chlorination of polychlorinated biphenyls by Hudson River sedimentorganisms: speci¢city to the chlorination pattern of congeners. Envi-ron. Sci. Technol. 27, 1190^1192.

[4] Ye, D., Quensen III, J.F., Tiedje, J.M. and Boyd, S.A. (1992) Anaer-obic dechlorination of polychlorobiphenyls (Aroclor 1242) by pas-teurized and ethanol-treated microorganisms from sediments. Appl.Environ. Microbiol. 58, 1110^1114.

[5] Ye, D., Quensen III, J.F., Tiedje, J.M. and Boyd, S.A. (1995) Evi-dence for para dechlorination of polychlorobiphenyls by methano-genic bacteria. Appl. Environ. Microbiol. 61, 2166^2171.

[6] Wu, Q.Z., Bedard, D.L. and Wiegel, J. (1997) E¡ect of incubationtemperature on the route of microbial reductive dechlorination of2,3,4,6-tetrachlorobiphenyl in polychlorinated biphenyl (PCB)-con-taminated and PCB-free freshwater sediments. Appl. Environ. Micro-biol. 63, 2836^2843.

[7] Cho, Y.-C., Kim, J., Sokol, R.C. and Rhee, G-Y. (2000) Biotrans-formation of polychlorinated biphenyls in St. Lawrence River sedi-ments reductive dechlorination and dechlorinating microbial popula-tions. Can. J. Fish. Aquat. Sci. 57, 95^100.

[8] Kim, J. and Rhee, G-Y. (1999) Interactions of polychlorinated bi-phenyl-dechlorinating microorganisms with methanogens and sulfatereducers. Environ. Toxicol. Chem. 18, 2696^2702.

[9] Hartkamp-Commandeur, L.C.M., Gerritse, J., Govers, H.A.J. andParsons, J.R. (1996) Reductive dehalogenation of polychlorinatedbiphenyls by anaerobic microorganisms enriched from Dutch sedi-ments. Chemosphere 32, 1275^1286.

[10] Middeldorp, P.J.M., de Wolf, J., Zehnder, A.J.B. and Schraa, G.

FEMSEC 1398 16-9-02

Y.-C. Cho et al. / FEMS Microbiology Ecology 42 (2002) 51^58 57

(1997) Enrichment and properties of a 1,2,4-trichlorobenzene-de-chlorinating methanogenic microbial consortium. Appl. Environ. Mi-crobiol. 63, 1225^1229.

[11] Chang, B.V., Chou, S.W. and Yuan, S.Y. (1999) Dechlorination ofpolychlorinated biphenyls by an anaerobic mixed culture. J. Environ.Sci. Health A34, 1299^1316.

[12] Sanford, R.A., Cole, J.R., Lo¥er, F.E. and Tiedje, J.M. (1996) Char-acterization of Desul¢tobacterium chlororespirans sp. nov., whichgrows by coupling the oxidation of lactate to the reductive dechlori-nation of 3-chloro-4-hydroxybenzoate. Appl. Environ. Microbiol. 62,3800^3808.

[13] Gerritse, J., Renard, V., Gomes, T.M.P., Lawson, P.A., Collins,M.D. and Gottschal, J.C. (1996) Desul¢tobacterium sp. strainPCE1, an anaerobic bacterium that can grow by reductive dechlori-nation of tetrachloroethene or ortho-chlorinated phenols. Arch. Mi-crobiol. 165, 132^140.

[14] Holliger, C., Wohlfarth, G. and Diekert, G. (1999) Reductive dechlo-rination in the energy metabolism of anaerobic bacteria. FEMS Mi-crobiol. Rev. 22, 383^398.

[15] Sokol, R.C., Kwon, O.S., Bethoney, C.M. and Rhee, G-Y. (1994)Reductive dechlorination of polychlorinated biphenyls (PCBs) in St.Lawrence River sediments and variations in dechlorination character-istics. Environ. Sci. Technol. 28, 2054^2064.

[16] Balch, W.E., Fox, G.E., Magrum, L.J., Woese, C.R. and Wolfe, R.S.(1979) Methanogens: reevaluation of a unique biological group. Mi-crobiol. Rev. 43, 260^296.

[17] Kim, J. and Rhee, G-Y. (1997) Population dynamics of polychlori-nated biphenyl-dechlorinating microorganisms in contaminated sedi-ments. Appl. Environ. Microbiol. 63, 1771^1776.

[18] Briones Jr., A.M. and Reicharst, W. (1999) Estimating microbialpopulation counts by ‘most probable number’ using Microsoft Excel.J. Microbiol. Methods 35, 157^161.

[19] Rhee, G-Y., Sokol, R.C., Bethoney, C.M., Cho, Y.-C., Frohnhoefer,R.C. and Erkkila, T. (2001) Kinetics of polychlorinated biphenyldechlorination and growth of dechlorinating microorganisms. Envi-ron. Toxicol. Chem. 20, 721^726.

[20] Kimbrough, D.E., Chin, R. and Wakakuwa, J. (1994) Wide-spreadand systematic errors in the analysis of soils for polychlorinatedbiphenyls. Part 3. Gas chromatography. Analyst 119, 1293^1301.

[21] Quensen III, J.F., Boyd, S.A. and Tiedje, J.M. (1990) Dechlorination

of four commercial polychlorinated biphenyl mixtures (Aroclors) byanaerobic microorganisms from sediments. Appl. Environ. Micro-biol. 56, 2360^2369.

[22] DeWeerd, K.A. and Bedard, D.L. (1999) Use of halogenated ben-zoates and other halogenated aromatic compounds to stimulate themicrobial dechlorination of PCBs. Environ. Sci. Technol. 33, 2057^2063.

[23] Liu, S.M., Kuo, C.E. and Hsu, T.B. (1996) Reductive dechlorinationof chlorophenols and pentachlorophenol in anoxic estuarine sedi-ments. Chemosphere 32, 1287^1300.

[24] Townsend, G.T. and Su£ita, J.M. (1996) Characterization of chloro-ethylene dehalogenation by cell extracts of Desulfomonile tiedjei andits relationship to chlorobenzoate dehalogenation. Appl. Environ.Microbiol. 62, 2850^2853.

[25] Mohn, W.W. and Kennedy, K.J. (1992) Reductive dehalogenation ofchlorophenols by Desulfomonile tiedjei DCB-1. Appl. Environ. Mi-crobiol. 58, 1367^1370.

[26] Struijs, J. and Rogers, J.E. (1989) Reductive dehalogenation of di-chloroanilines by anaerobic microorganisms in fresh and dichloro-phenol-acclimated pond sediment. Appl. Environ. Microbiol. 55,2527^2531.

[27] Bedard, D.L., Van Dort, H.M., May, R.J. and Smullen, L.A. (1997)Enrichment of microorganisms that sequentially meta, para-dechlori-nate the residue of Aroclor 1260 in Housatonic River sediment. En-viron. Sci. Technol. 31, 3308^3313.

[28] Bedard, D.L., Van Dort, H.M. and DeWeerd, K.A. (1998) Bromi-nated biphenyls prime extensive microbial reductive dehalogenationof Aroclor 1260 in Housatonic River sediment. Appl. Environ. Mi-crobiol. 64, 1786^1795.

[29] Van Dort, H.M., Smullen, L.A., May, R.J. and Bedard, D.L. (1997)Priming microbial meta-dechlorination of polychlorinated biphenylsthat have persisted in Housatonic River sediments for decades. En-viron. Sci. Technol. 31, 3300^3307.

[30] Wu, Q.Z., Bedard, D.L. and Wiegel, J. (1999) 2,6-Dibromobiphenylprimes extensive dechlorination of Aroclor 1260 in contaminatedsediment at 8^30‡C by stimulating growth of PCB-dechlorinatingmicroorganisms. Environ. Sci. Technol. 33, 595^602.

[31] Sokol, R.C., Bethoney, C.M. and Rhee, G-Y. (1998) E¡ect of Aro-clor 1248 concentration on the rate and extent of PCB dechlori-nation. Environ. Toxicol. Chem. 17, 1922^1926.

FEMSEC 1398 16-9-02

Y.-C. Cho et al. / FEMS Microbiology Ecology 42 (2002) 51^5858