Previous Article | Next Article ![]()
Applied and Environmental Microbiology, May 2007, p. 3009-3018, Vol. 73, No. 9
0099-2240/07/$08.00+0 doi:10.1128/AEM.02958-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Center of Marine Biotechnology, University of Maryland Biotechnology Institute, Baltimore, Maryland,1 Department of Microbiology and Immunology, Marine Biomedicine & Environmental Science Center, Medical University of South Carolina, Charleston, South Carolina2
Received 20 December 2006/ Accepted 2 March 2007
|
|
|---|
|
|
|---|
Historically, their use as dielectric fluid of liquid-filled transformers represented the second largest usage of PCBs (approximately 30%), of which the predominant commercial form between 1930 and 1971 was Aroclor 1260 (53). Aroclor 1260 is a mixture of highly chlorinated PCB congeners and is less susceptible to loss by volatilization and transformation by microbial activity than less-chlorinated Aroclor mixtures (4, 44). This might be due to a combination of factors, including (i) lower bioavailability caused by greater hydrophobicity, (ii) greater toxicity of individual more highly chlorinated congeners, and (iii) the lack of less-chlorinated PCB congeners associated with the stimulation of microbial transformation (44). However, despite this lower susceptibility to biotransformation, microbial transformation of Aroclor 1260 by anaerobic reductive dechlorination was reported as early as 1987 (14, 15), and several investigators since then have shown reductive dechlorination of Aroclor 1260 in sediment, as well as laboratory microcosms (4, 6, 8, 11, 33, 44, 45, 51, 52, 61, 62). Brown and coworkers (15) proposed that microorganisms could use PCBs as electron acceptors for respiration, enabling them to occupy a unique niche in anaerobic environments where other electron acceptors are limiting.
Aroclor 1260 is reductively dechlorinated through diverse patterns of congener transformations, depending on the contaminated sediment source and, presumably, the community of PCB-dechlorinating bacteria present (4, 8, 10, 14, 15, 44). Several investigators have attempted to isolate or identify microorganisms responsible for the reductive dechlorination of PCBs (36, 39, 47, 64), and although earlier studies suggested that dechlorination supported microbial growth (18, 31, 61), identification of the microbial catalysts by enrichment and isolation remained elusive. The first PCB-reducing bacteria were identified when the microbial communities in two sediment-free cultures with different dechlorination specificities were characterized by comparative sequence analysis of PCR-amplified 16S rRNA genes (21, 43, 56, 58). The dehalogenating microorganism o-17, which selectively dechlorinates PCB congeners with flanked ortho- and flanked meta-chlorines (22), and DF-1, which selectively dechlorinates congeners with double-flanked chlorines (59), were shown to belong to a deep branch of the phylum Chloroflexi, with their 16S rRNA gene sequences approximately 90% identical to that of the chloroethene-dechlorinating microorganism Dehalococcoides ethenogenes (38). Species within the Dehalococcoides group have been shown to reductively dechlorinate a number of chlorinated compounds (3, 16, 20, 27, 28). Indeed, Dehalococcoides ethenogenes 195 has since been shown to dechlorinate 2,3,4,5,6-pentachlorobiphenyl and other aromatic organochlorines when grown with tetrachloroethene (25); however, the authors did not investigate whether PCB alone could support growth of D. ethenogenes 195. Using 16S rRNA gene primers, Bedard and coworkers (6) identified phylotypes similar to Dehalococcoides spp. in a sediment-free culture dechlorinating Aroclor 1260, further suggesting that Dehalococcoides spp. and related microorganisms within the phylum Chloroflexi are the likely catalysts for the reductive dehalogenation of PCBs in the environment. However, there are currently no reports on how many different microorganisms are required to reductively dechlorinate a commercial PCB mixture such as Aroclor 1260 into less-chlorinated congeners containing unflanked chlorines, which are the terminal products of the dechlorination process.
All prior reports on the selective activities of PCB-dechlorinating bacteria have been conducted with individual PCB congeners. Previously, we showed (24) that two phylotypes, DEH10 and SF1, with high sequence similarity to Dehalococcoides spp. sequentially dechlorinated 2,2'3,3',4,6'-hexachlorobiphenyl (abbreviated as 234-236-CB or PCB 132) to PCB 91 (236-24-CB) and to PCB 51 (24-26-CB). This was the first report to demonstrate the sequential dechlorination of a PCB congener by the synergistic activities of two PCB-dechlorinating microorganisms. Here, we report that these two phylotypes, in addition to a third phylotype, SF2, which is similar to o-17, are selectively enriched during the reductive dechlorination of Aroclor 1260 in Baltimore Harbor (BH) sediment microcosms. Using sediment microcosms containing the 11 most predominant individual PCB congeners of this Aroclor, we show that most steps in these processes are linked to the growth of these phylotypes, which suggests that PCB-impacted environments can sustain these PCB-dechlorinating organisms. Individual PCB dechlorination pathways, terminal end products, and the microorganisms responsible for each step in the pathways are also reported for the first time. Interestingly, these three phylotypes were responsible for the anaerobic dechlorination of Aroclor 1260 to congeners containing unflanked chlorine congeners, which could be subsequently oxidized by aerobic PCB-degrading bacteria (1). The characterization of the selective and synergistic activities of Aroclor 1260-dechlorinating microorganisms reported here is an essential development for designing effective in situ treatment strategies of PCB-impacted sites.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Pathways and rates of reductive dechlorination for individual PCB congeners
|
Calculation of dechlorination rates.
Dechlorination curves were made for all PCB congeners in 50-day intervals over the course of incubation. Dechlorination rates for each congener were determined by calculating number of moles dechlorinated over time within the linear slope of the dechlorination curve. In instances where a congener was dechlorinated in several positions, the dechlorination rate was calculated from the total increase in concentration of each of the daughter products. The dechlorination rate was calculated by dividing the mole amount dechlorinated by the total number of moles present in the culture and the time elapsed in days. The average rate and the standard deviation were calculated from triplicate cultures (Table 1).
Bacterial community 16S rRNA gene analyses.
DNA from pooled samples (0.5 ml from each culture replicate) was extracted every 50 days using a Fast DNA SPIN for soil kit (MP Biochemicals, Solon, OH) or UltraClean soil DNA kit (Mo Bio, Carlsbad, CA) according to manufacturers' protocols. The concentration was determined using a DU 650 spectrophotometer (Beckman, Fullerton, CA), and DNA extracts were diluted with Tris-EDTA (TE) buffer to 10 µg/ml. Diluted DNA (1 µl) was used in all subsequent PCRs for a total of 1 ng of DNA.
Microbial community DNA from Aroclor 1260 and individual PCB congener microcosms was amplified by PCR with either universal 16S rRNA gene primers (40), which resulted in PCR fragments of approximately 160 bp, or primers specific for 16S rRNA genes of a monophyletic group within the phylum Chloroflexi, Chl348FGC and Dehal884R, as described previously (24), which resulted in PCB products of approximately 470 bp. Denaturing gradient gel electrophoresis (DGGE) was performed as described by Watts et al. (56) using the D-Code universal mutation detection system (Bio-Rad, Hercules, CA.). Briefly, 6% (wt/vol) polyacrylamide gels (Sigma, St. Louis, MO) containing a 30 to 70% denaturing gradient and fragments were separated by electrophoresis for 18 h at 75 V. The gels were stained with SYBR green 1 DNA stain (Molecular Probes, Eugene, OR) and visualized using a Storm PhosphorImager (GE Healthcare, Piscataway, NJ). DGGE bands of interest were excised and DNA was eluted by incubation in 30 µl TE buffer overnight at 4°C. PCR and DGGE were repeated until purity was confirmed for DNA fragments in each eluted band.
DNA sequencing and analysis.
PCR products from excised bands were used as templates for dye terminator cycle sequencing using the Big Dye 3.1 kit (Applied Biosystems, Foster City, CA) and an AB3100 genetic analyzer (Applied Biosystems). Sequences were examined for errors and assembled using the software Pregap4 and Gap4 of the Staden software package (http://sourceforge.net/projects/staden). Chimera formation was examined using Chimera Check (19). Sequences similarities were analyzed using the Basic Local Alignment Search Tool (BLAST) (5). In order to taxonomically classify sequences, we used the "classify" program by the Joint Genomic Institute (JGI) Greengenes server (http://greengenes.lbl.gov) after sequences were aligned using the align tool from the Greengenes NAST server (23).
Enumeration of PCB-dechlorinating phylotypes.
Putative dehalogenating members of the phylum Chloroflexi were enumerated by a competitive PCR (cPCR) assay (B. Kjellerup, X. Sun, U. Ghosh, H. May and K. Sowers, submitted for publication). A competitor was constructed based on the primers 348F/884R by using the DNA template supplied in the competitive DNA construction kit (RR017; TaKaRa Bio, Inc., Japan). Briefly, the numbers of 16S rRNA gene copies per µl of normalized DNA sample (1 ng DNA per µl) were determined according to the manufacturer's instructions (TaKaRa Bio). DNA samples (10 µg/ml) were amplified as described above for 35 PCR cycles with 1/10 dilutions of a competitor template with a known concentration. The ratio of the target PCR product to the competitor PCR product (T/C) measured by densitometry was determined using the image analysis software Quantity One (Bio-Rad, Hercules, CA) and log T/C was plotted against the log of copies of the competitor. The regression equation was solved for log C/T = 0 (i.e., equal amounts of target and competitor).
Nucleotide sequence accession numbers.
The 16SrRNA gene sequences for DGGE fragments have been submitted to GenBank under accession no. EF150839 to EF150845.
|
|
|---|
![]() View larger version (38K): [in a new window] |
FIG. 1. PCB congener distribution of the 12 most predominant congeners in Aroclor 1260 microcosms. Bars represent congener distributions at days 0 (white bars), 100 (gray bars), and 400 (black bars). Congeners that represent less than 0.05 wt% in Aroclor 1260 are not included (26). The 12 most dominant congeners are underlined, and the major congener end products are shown in bold text.
|
Dechlorination of individual Aroclor 1260 congeners.
BH sediment microcosms incubated with each of the 12 most predominant PCB congeners in Aroclor 1260 lagged between 3 and 6 months before dechlorination was detected in the initial microcosms, but the lag time generally decreased to less than 50 days by the fourth transfer (data not shown). The PCB congeners used in this experiment are listed in Table 1, except PCB 194 and PCB 94, which were not dechlorinated 500 days and 200 days, respectively, after the initial enrichment inoculation. Figure 2 shows the dechlorination pathways from each of the starting congeners to the final products. Table 1 describes the positions of the target chlorines, the rate of each reaction, and the end mol% for each of the starting congeners. The dechlorination rates of the parent compounds were lower than the dechlorination rates of the daughter compounds, with one exception: PCB 183 (2346-245) was dechlorinated more rapidly than PCB 154 (245-246). Also, the products in the single-congener experiments were the same products observed in the Aroclor 1260 mixture (Fig. 1 and 2), confirming that dechlorination pathways with the individual congeners were representative of the activities observed with the Aroclor.
![]() View larger version (38K): [in a new window] |
FIG. 2. PCB dechlorination pathways of the predominant PCB congeners in Aroclor 1260. Parent congeners are shown in bold text. The pathways are shown with large arrows that indicate different phylotypes: black solid arrows, DEH10; open arrows, SF1; hatched arrows, SF2; gray solid arrows, both DEH10 and SF1. Small arrows indicate minor pathways. The predominant end products are boxed. Reactions in which there was at least a twofold increase in the number of dechlorinating phylotypes relative to the no-PCB control are indicated with asterisks.
|
![]() View larger version (52K): [in a new window] |
FIG. 3. DGGE of Aroclor 1260-dechlorinating microcosms and no-PCB controls. Panel A is total DNA PCR amplified with universal primers (40), and panel B is total DNA PCR amplified with primers specific for PCB-dechlorinating phylotypes (24). DNA fragments labeled A to N were sequenced, and comparative sequence analyses are shown in Table 2.
|
|
View this table: [in a new window] |
TABLE 2. Phylogenetic assignment of microorganisms in Aroclor 1260 microcosms
|
Enumeration of Aroclor 1260-dechlorinating phylotypes.
To determine whether dechlorination of Aroclor 1260 supported microbial growth, microorganisms were enumerated in microcosms by cPCR using a selective primer specific set for PCB-dechlorinating phylotypes. The number of 16S rRNA gene copies of putative dechlorinators per µl of normalized DNA from microcosms dechlorinating Aroclor 1260 shows putative dechlorinators increase in numbers as Aroclor 1260 is dechlorinated (Fig. 4). In contrast, the control culture incubated without added Aroclor 1260 showed only a slight increase over 400 days [(2.46 ± 0.18) x 104 to (4.47 ± 0.27) x 104], which could be accounted for by PCB carried over in the transfer from the Aroclor 1260 microcosm. After 400 days of incubation, the number of PCB-dechlorinating phylotypes increased 25-fold in microcosms with Aroclor 1260 added compared to the no-PCB control.
![]() View larger version (19K): [in a new window] |
FIG. 4. Enumeration of PCB-dechlorinating phylotypes in Aroclor 1260 microcosms. Numbers of 16S rRNA gene copies per µl of normalized DNA of putative dechlorinators are shown from Aroclor 1260 microcosms () and no-PCB controls ( ). Dechlorination activity is shown as the number of chlorines per biphenyl in Aroclor 1260 microcosms ( ). Error bars (not shown) were smaller than the symbols.
|
|
View larger version (26K): [in a new window] |
FIG. 5. Composite DGGE of cultures dechlorinating individual PCB congeners. The primers used are specific for a monophylogenetic group within the phylum Chloroflexi that includes known PCB-dehalogenating microorganisms (24). Numbers indicate PCB congeners. 151m, meta dechlorination of PCB 151; 151°, meta and ortho dechlorination of PCB 151. The control microcosm (C) was incubated without PCB congener.
|
|
View this table: [in a new window] |
TABLE 3. Enumeration of dechlorinating phylotypes by cPCR
|
DEH10 dechlorinated the double-flanked chlorine in 234-substituted chlorophenyl rings, except with PCB 130. DEH10 showed a preference for para-flanked meta-chlorines when no double-flanked chlorines were available and thus dechlorinated 245-substituted chlorophenyl rings in the meta position, with one exception: PCB 154 (245-246), which contains a 246-substitution on the other ring. DEH10 also dechlorinated in the ortho-flanked meta position (PCBs 151 and 95), when the other ring contained 25-substitutions.
|
|
|---|
All 12 major parent congeners, which account for over 50 wt% of Aroclor 1260, were dechlorinated in Aroclor 1260 microcosms, including PCB 194, which was not dechlorinated when incubated as an individual congener in BH sediment microcosms. The dechlorination of PCB 194, as well as the accumulation of tri- and dichlorobiphenyls in Aroclor 1260 microcosms, may be due to the presence of multiple congeners that have been shown to have both stimulatory and inhibitory effects on congener specificity by PCB-dehalogenating microorganisms (37, 44). In the present study, the stimulatory effect of specific or multiple congeners in the Aroclor 1260-dechlorinating microcosm likely promoted the reductive dechlorination of PCB 194.
Several investigators have used single PCB congeners to infer PCB dechlorination pathways (2, 12, 13, 24, 42, 46, 49, 52, 57, 60). However, this is the first comprehensive report on the dechlorination of all major PCB congeners present in Aroclor 1260 by single-congener experiments. Several of the pathways in Fig. 2 have been proposed previously from inference of Aroclor 1260 dechlorination products. For example, a previous report with BH Aroclor 1260 microcosms (62) predicted the ortho dechlorination pathway of PCB 151 (2356-25) to PCB 72 (25-35). However, the previously predicted dechlorination pathway from PCB 170 (2345-234) to PCB 68 (24-35) was only partly consistent with our results (Fig. 2), as most of PCB 170 was dechlorinated to either 26 mol% PCB 47 (24-24) or 13 mol% PCB 49 (24-25) at day 400. Several observations were consistent with the dechlorination pathways for PCBs 101, 132, 138, 153, 170, and 180 in "process N," proposed in a comprehensive review by Bedard and Quensen (10). We determined the dechlorination pathways for PCB 183, which is, to our knowledge, the first report of this pathway. In addition, we defined the pathways of PCB 174 and PCB 151, which were ambiguous in previous reports (6, 7).
In a previous report on dechlorination of Aroclor 1260 in Housatonic River microcosms (6), several proposed dechlorination pathways for 2345-substituted congeners were based upon the assumption that 60% of these congeners were dechlorinated in the double-flanked para position and 40% in the double-flanked meta position. Woods Pond sediments have also been shown to have para dechlorination activity (7, 8). Although we observed some examples of double-flanked para dechlorination, our results showed that dechlorination in BH microcosms more often occurred in the double-flanked meta position. This was especially true for PCB 174 (2345-236), where all of the dechlorination occurred in the double-flanked meta position and for PCB 180 (2345-245), where 67% occurred in the double-flanked meta position and 33% of the dechlorination occurred in the double-flanked para position. On the other hand, PCB 137 (2345-24) was dechlorinated equally in the double-flanked meta and para positions.
Congeners with 2356-substitutions (i.e., PCBs 149, 151, and 187) were dechlorinated in the ortho-flanked meta position. However, we also observed some minor ortho dechlorination of PCB 151 (2356-25), PCB 90 (235-24), and PCB 92 (235-25), which is consistent with prior reports of ortho dechlorination activity of both Aroclor 1260 and single congeners in BH microcosms (12, 21, 22, 37, 62). Based on observations with Aroclor 1260 and the individual congeners, the general sequence of dechlorination in BH microcosms is as follows: double-flanked meta or para position of 2345-substituted chlorophenyl rings, double-flanked meta position of 234- or 2346-substituted chlorophenyl rings, ortho-flanked meta position of 2356-substituted chlorophenyl rings, para-flanked meta position of 245-substituted chlorophenyl rings, ortho-flanked meta position of 236-substituted chlorophenyl rings, ortho-flanked meta position of 235-substituted chlorophenyl rings, and flanked ortho position of 2356-substituted chlorophenyl rings.
Effect of congener characteristics on dehalogenation.
An analysis of (i) the differences in the estimated Gibbs free energy of formation between parent and daughter congeners (29), (ii) the differences in relative retention time (17), (iii) the aqueous solubilities (29), and(iv) the number of total chlorines and the number of ortho-, meta-, and para-chlorines on the ring subjected to dechlorination, as well as on the opposite ring, showed that there were no significant relationships (P < 0.05) between these parameters and the dechlorination rates, end mol%, or the number of putative dechlorinators in BH microcosms. However, we observed weak relationships (with low R2 values) between the dechlorination rate and the aqueous solubility of the PCB congener (r2 = 0.25), the total amount of chlorines (r2 = 0.22), as well as the number of meta-chlorines (r2 = 0.28). Our results indicate that dechlorination rates increased in microcosms with more-soluble PCB congeners and decreased with both the number of total chlorines and the number of meta-chlorines.
Diversity of Aroclor 1260-dehalogenating phylotypes.
PCB-dehalogenating microorganisms have been previously identified as belonging to either Dehalococcoides spp. or the o-17/DF-1 clade within a deep branch of the Chloroflexi (6, 22, 24, 25, 43, 55, 58, 59, 63), and these phylotypes have been detected in microcosms dechlorinating Aroclor 1260 (6). Although the microbial community in the Aroclor 1260 microcosms was diverse (Fig. 3), several lines of evidence indicate that only two phylotypes, DEH10 and SF1, were the predominant biocatalysts of Aroclor 1260 dechlorination: (i) they were only detected in Aroclor 1260 and individual congener microcosms and not detected using universal PCR primers with the no-PCB controls, (ii) they increased in numbers only during active dechlorination of Aroclor 1260 and individual congeners, and (iii) the phylotypes have high sequence similarity to phylotypes and isolates previously shown to reductively dechlorinate PCBs (24, 25, 55). Phylotypes belonging to Bacteriodetes and Spirochaetes (Fig. 3, DNA fragments B and E) were also only present in the dechlorinating cultures, and although similar phylotypes have been previously detected in BH microcosms (43), to date, there is no evidence they reductively dechlorinate PCBs.
When using PCR-based assays of microbial communities, microorganisms with more than one 16S rRNA gene copy (54) and the inherent biases with PCR (50) can influence the results. In our experiments, we compared the diversity and the growth of putative dehalogenators between cultures with and without PCB to detect microbial phylotypes within the microbial community that are specifically associated with active dechlorination of PCBs. Also, since we directly compare the presence of the same phylotypes between dechlorinating microcosms and no-PCB controls, any bias due to multiple 16S rRNA copies or PCR would be the same in both treatments. In some cases, we detected putative dehalogenators in the no-PCB controls (Fig. 3 and Table 3), but this is likely due to enrichment resulting from traces of PCB cotransferred into no-PCB controls during inoculation from active cultures. Dehalococcoides species with very similar or even identical 16S rRNA sequences can have different dechlorination activities (30). Although the 16S rRNA gene sequence of phylotypes DEH10, SF1, and SF2 detected in this study are 100% identical to the phylotypes detected previously in BH sediment microcosms, we cannot confirm that they are each the same species or strain since they came from different microcosms. However, the high identity combined with the fact they each came from the same source and each have the same selective dechlorination activities as previously described phylotypes indicates a high likelihood that they are the same microorganisms.
SF1 and DEH10 exhibited specific activities towards the PCB congeners we tested. The combined activities of SF1 were different from those previously reported for either o-17 (flanked ortho and ortho-flanked meta-chlorines) or DF-1 (double-flanked chlorines). Since phylotype SF2 was only unequivocally associated with two pathways in the present study, this does not represent an adequately comprehensive overview of its specific dechlorinating activities. BH sediment microcosms have been shown previously (62) and here (data not shown) to dechlorinate ortho-chlorines in Aroclor 1260 mixtures. However, our inability to detect SF2 in Aroclor 1260 microcosms suggests that it may have a relatively minor role in the dechlorination process.
When grown with the single congeners or Aroclor 1260, the populations of PCB-dechlorinating microorganisms increased only 1 to 2 orders of magnitude during the course of reductive dehalogenation. These results are consistent with prior reports that showed a similar range of increases for congener mixtures by other methods, including most probable number enumeration and most probable number PCR of 16S rRNA gene copies (18, 24, 31). However, we show unequivocally that phylotypes within Dehalococcoides spp. and the DF-1/o-17-group of the dehalogenating members of the phylum Chloroflexi are directly responsible for the reductive dechlorination of an Aroclor. The relative growth of the individual dechlorinating phylotypes varied among different PCB congeners, increasing up to 2 orders of magnitude, with an average 13.5-fold increase for the 24 congeners tested. Although this supports a conclusion by Kim et al. (31) that the size of the dechlorinating population might be an indicator for PCB dechlorination potential in a site, the results also suggest that other factors, including the types of congeners and the indigenous dechlorinating phylotypes, will have an impact on the size of the population.
Prior reports have shown that Aroclor 1260 is reductively dechlorinated in sediment microcosms, and PCB-dechlorinating microorganisms have been identified from individual congener enrichments. The results of this study show for the first time that the synergistic activities of only three Chloroflexi phylotypes (SF1, SF2, and DEH10) reductively dechlorinated the 11 major PCB congeners in Aroclor 1260 to unflanked tetra- and trichlorobiphenyls in our microcosms. Two of the phylotypes (SF1 and DEH10) were also the predominant phylotypes in Aroclor 1260 enrichments, indicating they likely have a significant role in the dechlorination of other Aroclor 1260 congeners. Demonstration that dechlorination of Aroclor 1260, as well as most of the individual congeners, supports growth of SF1, SF2, and DEH10 suggests that PCB-impacted environments can sustain populations of these PCB-dechlorinating organisms. This is particularly relevant for the development of biostimulation or bioaugmentation strategies for the bioremediation of PCBs. The final products of Aroclor 1260 dechlorination by these three phylotypes (unflanked tetra- and trichlorobiphenyls) can potentially be further transformed by bioaugmentation with microcosms that have been shown to dechlorinate unflanked congeners (9) or could serve directly as substrates for aerobic mineralization by PCB-degrading bacteria (35). Further characterization of the selective and synergistic activities of PCB-dechlorinating microorganisms with different Aroclor mixtures and sediment microcosms is essential for generating models to predict the dechlorination potential at specific PCB-impacted sites and design effective in situ treatment strategies.
We thank Kimberly Anderson, Marcelino T. Suzuki, and Joy E. M. Watts for helpful comments.
Published ahead of print on 9 March 2007. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»