Isolation of Acetobacterium sp. Strain AG, Which Reductively Debrominates Octa- and Pentabrominated Diphenyl Ether Technical Mixtures

Abstract

Polybrominated diphenyl ethers (PBDEs) are a class of environmental pollutants that have been classified as persistent organic pollutants since 2009. In this study, a sediment-free enrichment culture (culture G) was found to reductively debrominate octa- and penta-BDE technical mixtures to less-brominated congeners (tetra-, tri-, and di-BDEs) via a para-dominant debromination pattern for the former and a strict para debromination pattern for the latter. Culture G could debrominate 96% of 280 nM PBDEs in an octa-BDE mixture to primarily tetra-BDEs in 21 weeks. Continuous transferring of culture G with octa-/penta-BDEs dissolved in n-nonane or trichloroethene (TCE) yielded two strains (Acetobacterium sp. strain AG and Dehalococcoides sp. strain DG) that retained debromination capabilities. In the presence of lactate but without TCE, strain AG could cometabolically debrominate 75% of 275 nM PBDEs in a penta-BDE mixture in 33 days. Strain AG shows 99% identity to its closest relative, Acetobacterium malicum. In contrast to strain AG, strain DG debrominated PBDEs only in the presence of TCE. In addition, 18 out of 19 unknown PBDE debromination products were successfully identified from octa- and penta-BDE mixtures and revealed, for the first time, a comprehensive microbial PBDE debromination pathway. As an acetogenic autotroph that rapidly debrominates octa- and penta-BDE technical mixtures, Acetobacterium sp. strain AG adds to the still-limited understanding of PBDE debromination by microorganisms.

INTRODUCTION

Polybrominated diphenyl ethers (PBDEs) are a group of brominated flame retardants that have been extensively used in the manufacturing of textiles, plastics, and electronic equipment for the past 3 decades (1). Recent studies revealed an exponential elevation of PBDEs in environmental samples and human tissues, with a doubling time of approximately 5 years (2–4). Several studies suggest that PBDEs are associated with tumors, neurodevelopmental toxicity, and hormonal imbalance (5). Due to concerns about their increasing presence in the environment and potential toxicity, production and usage of penta-BDE and octa-BDE technical mixtures have been banned in the European Union and in several states of the United States (6, 7). Tetra- and penta-BDE congeners present in penta-BDE technical mixtures have recently been classified as persistent organic pollutants (POPs), as are the hexa- and hepta-BDE congeners, which are major components in octa-BDE technical mixtures (8).

Nevertheless, human exposure to PBDEs will continue for the coming decades due to the persistent nature of PBDEs. Interestingly, PBDEs have been shown to be biotransformed by microbes both anaerobically (9–14) and aerobically (15, 16). While aerobic bacterial isolates transform only less-brominated congeners (up to hexa-BDEs) (15, 16), anaerobic bacterial isolates belonging to the Dehalococcoides, Dehalobacter, and Desulfitobacterium genera were reported to debrominate octa-BDE technical mixtures (hexa- to nona-BDEs) to less-brominated congeners (11, 12). However, most of anaerobic microbial debromination processes discovered previously required a primary halogenated electron acceptor (e.g., chloroethenes or chlorophenols) to support cell growth prior to debromination (11, 12). It would be of high value to identify environmental microorganisms capable of debrominating PBDEs without additional halogenated electron acceptors, since PBDEs do not always coexist with contaminants such as chloroethenes or chlorophenols in contaminated sites.

Recently, several sediment-free cultures that debrominate PBDEs were enriched, i.e., two octa-BDE mixture debrominating cultures, C-N-7 and U-N-1 (from Wuhan, China, and San Francisco, CA), and a penta-BDE mixture debrominating culture, GY2 (from Guiyu, China) (14, 17). These three cultures are able to carry out debromination without a primary halogenated electron acceptor. Although Dehalococcoides strains were not detected in cultures C-N-7 and U-N-1, Chloroflexi-like microorganisms seemed to be present (14). In culture GY2, it was demonstrated that Dehalococcoides strains were highly possible to be responsible for PBDE debromination (17). Therefore, while only Chloroflexi and a few other bacterial genera (e.g., Dehalobacter and Desulfitobacterium [11, 12]) are known to debrominate PBDEs, it would be interesting to find out whether bacteria belonging to other genera also catalyze PBDE debromination, which would help to broaden our understanding of PBDE debrominators in nature.

In this study, we obtained an enrichment culture from soil samples from a river bank which showed the capability of debrominating both penta- and octa-BDE technical mixtures. Two debrominating strains with distinct bacterial populations were also obtained by transferring the original culture with PBDE dissolved in either trichloroethene (TCE), a primary electron acceptor, or n-nonane, an inert solvent. Debromination pathways for both penta- and octa-BDE technical mixtures in the enriched cultures were successfully determined using a convenient and low-cost congener identification strategy that requires only a few standards for verification.

MATERIALS AND METHODS

Chemicals.

An octa-BDE technical mixture (consisting of nona-BDE [congener 207], octa-BDEs [congeners 203, 196, and 197], hepta-BDE [congener 183], and hexa-BDE [congener 153]) was purchased from Sigma-Aldrich (St. Louis, MO). A penta-BDE technical mixture dissolved in ethyl acetate (consisting of hexa-BDE [congener 153], penta-BDEs [congeners 99 and 100], tetra-BDE [congener 47], and 2,2′,4,4′5,5′-hexabromobiphenyl [HBB]) and another penta-BDE technical mixture dissolved in iso-octane (consisting of congeners 47, 99, and 100) were obtained from Ultra-Scientific (North Kingstown, RI). Deca-bromobiphenyl (DBB) was purchased from AccuStandard (New Haven, CT). TCE and n-nonane, with minimum purity of 99.5%, were obtained from Sigma-Aldrich (St. Louis, MO).

Culture medium and enrichment.

A sediment-free culture (designated culture G) was derived from an octa-BDE-debrominating microcosm established from soil samples collected from a river bank with no reported PBDE contamination in Wisconsin. All bacterial cultures were grown at 30°C in 160-ml serum bottles containing 100 ml autoclaved bicarbonate-buffered mineral salts medium as described previously (18). Sterile-filtered lactate (10 mM), pyruvate (10 mM), or acetate (10 mM) was added as carbon source, and hydrogen (20 ml or 333,333 ppm by volume [ppmv]) was supplied to acetate-fed cultures only. For strain AG grown in H₂-CO₂ (24 ml:6 ml, equivalent to 9.7 mM:2.4 mM aqueous concentration), all organic compounds, including the organic carbon source, _L-cysteine, dithiothreitol, and the buffering agent 2-[tris(hydroxymethyl)methylamino]-1-ethanesulfonic acid (TES), were omitted from the medium, while concentrations of sodium bicarbonate and sodium sulfide were adjusted to 45 mM and 0.8 mM, respectively. The octa-BDE technical mixture dissolved in either TCE or n-nonane (3.1 g/liter) was added to the medium with initial measured concentrations of 12.0 nM nona-BDEs, 26.6 nM octa-BDEs, 220.4 nM hepta-BDEs, and 9.5 nM hexa-BDEs. The penta-BDE technical mixture was added to the medium with an initial measured concentration of 70.6 nM congener 153, 126.9 nM total congeners 99 and 100, and 76.1 nM congener 47 (stock solution, 0.5 g/liter of each congener in ethyl acetate). When TCE was added as a primary electron acceptor, the initial TCE concentration was 714 μM in the medium (∼6.5 μl per 100 ml medium). When n-nonane was used as a solvent for the octa-BDE mixture, it was added at the same volume as TCE. The effects of ethyl acetate in the penta-BDE technical mixture on culture performance were evaluated by replacing this mixture with another penta-BDE technical mixture in iso-octane at the same initial concentrations. All experiments were conducted with biological triplicates, and abiotic controls (without inoculation of bacteria) were included to monitor for potential nonbiological debrominating activity.

Sample extraction and analyses.

Extraction of PBDEs from culture, analysis of PBDEs by gas chromatography-mass spectrometry (GC-MS), and establishment of calibration curves were performed as previously described (14). The PBDE concentration in samples spiked with the octa-BDE technical mixture was calibrated by the response of a known initial amount of DBB spiked into the samples before extraction, while for the penta-BDE technical mixture, HBB in the substrate was chosen as the internal calibration standard. Chloroethenes were measured as described previously (19). Volatile fatty acids (VFAs) were determined on Agilent 1100 high-pressure liquid chromatography (HPLC) system (Agilent, CA) with a UV detector (210 nm). Separation of VFAs was conducted on a Rezex ROA-organic acid H+ (8%) HPLC column (Phenomenex) at room temperature, with 2.5 mM sulfuric acid as a mobile phase at a flow rate of 0.5 ml min⁻¹ (10-μl injection volume). Detection limits are 50 μM for acetate, 100 μM for lactate, and 2 μM for pyruvate. Protein concentrations were measured using a Bio-Rad DC protein assay kit (Bio-Rad, CA) according to the manufacturer's instructions. Hydrogen in the headspace was monitored by using an Agilent GC7890 equipped with a thermal conductivity detector as described previously, with a detection limit of 8 ppmv (20). Fluorescence microscopy was performed on a Nikon Eclipse 200 using Syto-9 staining dye (Invitrogen).

Molecular analyses.

Genomic DNA was extracted from the cell pellet of a 1-ml culture as described previously (21). For microbial community analysis, 16S rRNA genes were amplified with a universal primer set Cy5-labeled 8F (5′-Cy5-AGA GTT TGA TCC TGG CTC AG-3′) (22) and 1392R (5′-ACG GGC GGT GTG TAC-3′) (23). Amplicons were subjected to restriction enzyme digestion with HhaI and MspI (NEB, Ipswich, MA) and to terminal restriction fragment length polymorphism (T-RFLP) analysis on an ABI 3100 sequencer (Applied Biosystems, CA) according to the manufacturer's instructions. Construction of the 16S rRNA gene clone library and sequencing of representative plasmids were done as described previously (19).

The presence of known reductively dechlorinating populations was screened by targeting genomic DNAs of debrominating cultures with genus-specific primers of Dehalococcoides, Dehalobacter, Desulfitobacterium, Desulfovibrio, Acetobacterium, Anaeromyxobacter, and Desulfuromonas (24–30) (the primers used in this study are given in Table S1 in the supplemental material). Nested PCR was performed by using the universal primers 8F and 1392R in the first PCR, followed by a second-round PCR using the genus-specific primers of the respective dechlorinating populations on the 50-times-diluted first-round PCR DNA template. PCR-denaturing gradient gel electrophoresis (DGGE) was performed by using primers 341FGC and 518R (see Table S1 in the supplemental material) as described previously (19).

Quantitative real-time PCR targeting Acetobacterium was performed on an ABI 7500 Fast real-time PCR system (ABI, Foster City, CA) by using QuantiTect SYBR green PCR master mix (Qiagen, Germany) and Acetobacterium primers Aceto572F and Aceto784R (29). The presence of five copies of 16S rRNA genes per Acetobacterium genome was assumed based on strain AG's closest sequenced genome, that of Acetobacterium woodii DSM 1030, in order to convert gene copies to cell numbers (data from NCBI).

Identification of unknown PBDE congeners.

Based on previous GC retention time databases (31, 32) and a mixed PBDE standard solution containing 39 congeners from mono- to hepta-BDEs, it is possible to identify biotransformation products from PBDEs or at least to narrow down the possible candidates to a number that is affordable when purchasing individual congener standards. The 39-congener mixed standard, which served as retention time reference point in the GC profile, is readily available through various vendors, including AccuStandard, Cambridge Isotope Laboratories (this study), and Ultra Scientific.

To identify unknown PBDE product congeners, the overall strategy is “identification by elimination” from a pool of possible debromination products. The number of bromines for each unknown congener was determined based on mass spectra, and then all possible congeners from debromination of higher-brominated parent congeners were listed as candidates. GC retention times of the unknown congeners were compared with either known congeners in the mixed standard or the retention time databases in the studies by Korytar et al. (32) and La Guardia et al. (31). Finally, for those congeners that still could not be identified, individual standards were purchased and used to match their retention times and mass spectrum profiles.

Nucleotide sequence accession numbers.

The 16S rRNA gene sequences of Acetobacterium sp. strain AG and Dehalococcoides sp. strain DG have been deposited in GenBank under accession numbers JQ627627 and JQ627628, respectively.

RESULTS

Debromination of an octa-BDE technical mixture dissolved in TCE by culture G.

An octa-BDE mixture debrominating anaerobic sediment-free culture G was obtained by sequential transfers of active debrominating microcosms to mineral salts medium amended with lactate and an octa-BDE technical mixture dissolved in TCE (octa/TCE). During the reductive debromination process, lactate (10 mM) was quickly depleted within 4 days (it was undetectable on day 4), producing ∼15 mM acetate solely, which is a typical homoacetic fermentation process as seen in acetogens (33, 34). Visible growth was observed within 5 days of inoculation, and 33% of parent PBDE congeners were debrominated on day 7 (Fig. 1A). Meanwhile, TCE was dechlorinated to vinyl chloride (VC) and ethene via all three dichloroethene (DCE) isomers (cis-DCE was predominant) within 2 to 3 weeks (Fig. 1B). Furthermore, extensive reductive debromination of the octa-BDE technical mixture yielded tri-BDEs (6.4 nM), tetra-BDEs (168.6 nM), and penta-BDEs (39.0 nM) at week 10 (71% removal of PBDEs) (Fig. 1A), which is much higher than the removal percentage reported before for three debrominating cultures, with the highest removal of 17.3% of 27 nM hepta-BDE congener 183 at week 12 (12). Since lactate was consumed within 4 days, an additional 10 mM lactate was added to the cultures on week 12 to promote further possible debromination of residual PBDE congeners. Again, the additional 10 mM lactate was also quickly consumed after only 4 days. At week 21, congeners 196, 197, 203, and 183 in the octa-BDE mixture were mostly debrominated (96% removal), with tetra-BDE congeners (235.8 nM) accumulated as the major end products (Fig. 1A). Throughout the experimental period, PBDE concentrations in abiotic controls remained relatively constant, with no debromination products observed (data not shown). In addition to debrominating octa-BDEs, culture G could also debrominate the penta-BDE technical mixture in a debromination profile similar to that for strain AG (shown in the next section).

Fig 1.

Debromination of octa-BDE technical mixture and dechlorination of trichloroethene by culture G. (A) Debromination of octa-BDE technical mixture and product formation. (B) Dechlorination of trichloroethene and product formation. 1,1-DCE was detected in a negligible amount (below 5 μM) and is not shown. The data shown in panels A and B depict performance from the same culture bottles (triplicates). S and P, substrates and products, respectively.

In an effort to identify the bacterial populations that were involved in reductive debromination of the octa-BDE technical mixture dissolved in TCE, genus-specific primers targeting 16S rRNA genes of known dehalogenating bacterial groups (see Materials and Methods) were tested on genomic DNA of culture G. Amplicons with the expected sizes (∼630 bp and ∼230 bp) were obtained from nested PCR with primers targeting the Dehalococcoides and Acetobacterium groups. However, primers targeting other dehalogenating bacterial groups did not yield any bands. T-RFLP analysis revealed the presence of two dominant peaks with either HhaI (∼198 bp and ∼373 bp) or MspI (∼219 bp and ∼515 bp) (see Fig. S1 in the supplemental material). Comparing the T-RF sizes with in silico digestion results for various bacterial 16S rRNA sequences, we infer that the ∼198-bp peak in the HhaI digestion and the ∼515-bp peak in the MspI digestion probably belong to Dehalococcoides and that the other dominant peak probably belongs to Acetobacterium.

To further enrich Dehalococcoides, a debrominating subculture with TCE-dechlorinating activity was obtained by sequential transfers of culture G in medium amended with acetate-hydrogen and octa/TCE. Clone library examination of this culture found a dominance of Dehalococcoides strains (74 clones of a total 80, while the rest belonged to Acetobacterium). After further enrichment of Dehalococcoides in a medium spiked with 200 mg/liter ampicillin and subsequent transfer to ampicillin-free medium, the culture was found to consist of only Dehalococcoides (designated strain DG). Absence of Acetobacterium was verified by microscopic observation (lack of rod-shaped cells [data not shown]), nested PCR using an Acetobacterium-specific primer pair (see Fig. S2 in the supplemental material), and a single tight band on a DGGE gel after PCR amplification of the 16S rRNA genes (see Fig. S3 in the supplemental material). The 16S rRNA gene sequence of strain DG (accession number JQ627628) is 100% identical to that of the Dehalococcoides mccartyi strains GT and CBDB1 over 1,354 bp.

Debromination of the octa-BDE technical mixture by strain DG was much slower than that by culture G even in the presence of TCE, producing minimal amounts of tetra- and penta-BDEs after 6 weeks (see Fig. S4C in the supplemental material), while debromination of the penta-BDE technical mixture by strain DG produced two tetra-BDE congeners in the presence of TCE after 4 weeks (see Fig. S5C in the supplemental material). TCE dechlorination kinetics in strain DG were similar to those in culture G (data not shown). When TCE was omitted from the medium, strain DG did not show any debromination activities on either octa- or penta-BDEs, suggesting that strain DG debrominated PBDEs cometabolically.

Isolation and debromination kinetics of Acetobacterium sp. strain AG.

To enrich other possibly existing debrominators that do not rely on TCE, culture G was repeatedly transferred and fed with the octa-BDE mixture dissolved in n-nonane (octa/nonane). After five consecutive serial dilutions of culture G in lactate-amended medium spiked with octa/nonane, the subculture generated debromination products similar to those with strain DG (see Fig. S4B in the supplemental material). Therefore, this subculture (designated culture AG) could debrominate PBDEs in the absence of primary electron acceptors such as TCE. Culture AG was further transferred to a medium amended with lactate and octa/TCE in triplicates. No dechlorination of TCE was observed on day 314, although debromination occurred as early as on day 143 (data not shown), indicating that subculture AG does not contain TCE dechlorinators.

To identify the microbes in culture AG, a 16S rRNA gene clone library was constructed. A total of 77 clones were picked, and enzymatic digestion (HhaI and MspI) of the gene inserts revealed only one digestion profile, suggesting the purity of culture AG. Analysis of two representative clones yielded identical sequences (accession number JQ627627), sharing 99% 16S rRNA gene sequence similarity with Acetobacterium malicum (accession number X96957). In order to analyze the debromination ability of this close relative, Acetobacterium malicum DSM4132 was acquired from DSMZ (Germany), but no debromination activities were observed on either the penta- or octa-BDE mixture. In order to confirm that Dehalococcoides populations were absent in culture AG, nested PCR was performed using universal (first-round PCR) and Dehalococcoides-specific (second-round PCR [nested PCR]) primer pairs. No amplicons were found after two rounds of PCR, confirming that Dehalococcoides strains were diluted out in culture AG (see Fig. S6 in the supplemental material). A single tight band on 16S-DGGE also confirms that no other bacterial groups are present in culture AG (see Fig. S3 in the supplemental material). Therefore, culture AG consists of an Acetobacterium isolate (strain AG).

Apart from the octa-BDE mixture, strain AG was also capable of debrominating the penta-BDE technical mixture in the absence of any primary electron acceptor (e.g., TCE) (Fig. 2). Similar to the case for culture G, lactate (10 mM) was quickly consumed in the first 3 to 4 days of incubation (4.1 to 5.3 mM was left on day 3, and it was undetectable on day 4), producing ∼15 mM acetate. After 12 days of incubation, significant removal (∼37%) of the congeners 153, 100, 99, and 47 (275 nM total initial concentration) was observed, with the generation of di-, tri-, and tetra-BDEs (Fig. 2). Debromination became stagnant after day 12. Then, after respiking lactate on day 30 (5 mM final concentration in the culture, which dropped to below the detection limit on day 33), debromination proceeded further, and the total removal of PBDE substrates reached 75% after only 3 days (day 33; all lactate was depleted by then, producing ∼7.5 mM acetate). The increase of Acetobacterium cell numbers as monitored by quantitative PCR (qPCR) was concurrent with the addition and depletion of lactate, accompanying the debromination of penta-BDE mixture, implying that debromination was triggered by lactate addition and cell growth (Fig. 2). No debromination in abiotic controls was observed throughout the experimental period (data not shown).

Fig 2.

Debromination of a penta-BDE technical mixture and product formation by strain AG. S and P, substrates and products, respectively.

Since debromination resumed only when lactate was respiked in the medium, it is highly possible that PBDE debromination by strain AG is a cometabolic process. To verify this, strain AG was inoculated into mineral salts medium with different initial lactate concentrations and the same amount of penta-BDE mixture (∼280 nM) upon inoculation (growing-cell experiment). Simultaneously, pregrown strain AG cells (measured as 38.9 mg/liter protein) were pelleted and resuspended in the same volume of organic-carbon-free medium spiked with penta-BDE mixture (resting-cell experiment, i.e., cell suspension experiment) (Fig. 3, sample R). After incubation for 12 days, debromination rates (determined as bromine removed) showed positive correlation with initial lactate concentrations and cell amounts (measured as protein concentrations) in growing-cell experiments (Fig. 3). In contrast, the resting-cell test exhibited only 10 to 15% debromination activity compared to that for growing cells with similar amounts of proteins (37.2 to 41.3 mg/liter). Such low debromination activity in resting cells further implies that PBDE debromination is a cometabolic process and relies on the presence of lactate.

Fig 3.

PBDE debromination shown as bromine removal during cell growth (growing cells) and in a cell suspension assay (resting cells) of strain AG. Protein and PBDE measurements were performed on day 12. Molar concentrations on the x axis indicate lactate concentrations in growing-cell experiments, while “R” stands for resting cells of strain AG that was pregrown in 10 mM lactate. Bromine removal was calculated from actual concentrations of PBDE congeners for easier comparison.

Reductive debromination of strain AG was also supported by pyruvate (10 mM) and H₂-CO₂ (9.7 mM/2.4 mM equivalent aqueous concentration). Debromination of penta-BDEs in pyruvate occurs at rates similar to that in lactate, while strain AG in H₂-CO₂-amended medium debrominated at a lower rate (75% of the debromination products in lactate or pyruvate on day 12). In organic-carbon-free medium spiked with H₂-CO₂, strain AG consumed H₂ within 10 days (700 ppmv residual H₂ in the headspace) in a typical process for acetogenesis by producing acetate according to an exact 4:1 molar ratio with H₂ (0.249 ± 0.002 mmol acetate from an initial 0.982 mmol H₂ on day 12). Additionally, no debromination activity was observed after 1 month of incubation when culture filtrate of AG was spiked with octa- and penta-BDE mixtures. This observation precludes the possibility that certain metabolites from strain AG (such as vitamin B₁₂) play a role in abiotic debromination of PBDEs.

Debromination of octa-BDE mixture by coculture of AG plus DG.

Compared to those in the original culture G, the debromination rates for the octa-BDE mixture in strains AG and DG are much lower (see Fig. S4 in the supplemental material). Two possible explanations are available: (i) other active debrominators were diluted out during the above-mentioned enrichment, or (ii) there may be synergistic interactions on PBDE debromination between the two strains that enable much faster debromination. In order to evaluate the second hypothesis, strain AG and strain DG were coinoculated in lactate medium spiked with octa/TCE and H₂ (3 ml or 50,000 ppmv). After 5 weeks of incubation, this AG-DG coculture exhibited almost the same debromination rate and profile as the original enrichment culture G (see Fig. S4D and E in the supplemental material), and TCE was dechlorinated to VC and ethene. When H₂ was omitted from the medium, this coculture's debromination performance remained at the same level as that of culture G (data not shown). Actually, in the H₂-free medium, the amount of H₂ reached a maximum of 200 ppmv on day 9 and dropped to below the detection limit (8 ppmv) on day 17, indicating that H₂ produced by Acetobacterium in the coculture could be utilized for reductive dehalogenation by Dehalococcoides.

Debromination pathways of culture G.

Based on a PBDE standard solution containing 39 PBDE congeners and previously published GC retention time databases (31, 32), we successfully identified almost all the debromination products (18 out of 19) from octa- and penta-BDE mixtures produced by culture G with only eight PBDE congener standards purchased. GC retention times for all identified PBDE congeners in our study are given in Table S2 in the supplemental material. Table S3 in the supplemental material shows detailed procedures for identification of congeners in culture G using the previously mentioned methodology.

When fed with the octa-BDE technical mixture, culture G produced 10 PBDE congeners, including three hexa-BDEs, three penta-BDEs, three tetra-BDEs, and one tri-BDE. These PBDE congeners were also observed when the major constituent in the octa-BDE technical mixture, the hepta-BDE congener 183 (82% mol fraction) was fed as a single electron acceptor to culture G. Therefore, the debromination pathways for the octa-BDE technical mixture in culture G are described only for congener 183 (Fig. 4A), though we cannot exclude the possible debromination of octa- and nona-BDE congeners in the octa-BDE mixture. Congener 183 is debrominated to hexa-BDE congeners 144, 149, 154, and possibly 153 (because congener 153 is also present as a substrate in the octa-BDE mixture, its formation from congener 183 cannot be confirmed). These four hexa-BDE congeners are then converted to penta-BDE congeners 95, 101, and 103, which are further converted to tetra-BDE congeners 52 and 53 and another, unidentified tetra-BDE congener (either congener 44 or congener 59). Notably, congener 53 (from congeners 95 and/or 103) accounts for more than 70% of all three tetra-BDEs formed. Finally, congener 53 generates a minimal amount of tri-BDE congener 19, which is the only tri-BDE congener produced. The confirmed debromination steps shown in Fig. 4A (depicted with solid arrows) are dominant by para debromination (i.e., five para, two meta, and one ortho debromination steps), consistent with those observed in the previous study on octa-BDE mixtures (12). Culture G preferentially removes single-flanked para bromines as observed in the debromination of hexa-BDE congener 154 (246-245); instead of removing a nonflanked para bromine to form congener 102 (26-245), congener 154 is debrominated in a single-flanked para bromine and produces congener 103 (246-25).

Fig 4.

Debromination pathways for octa- and penta-BDE technical mixtures. Solid arrows indicate confirmed pathways, while dashed arrows indicate pathways that possibly exist but are not confirmed, usually because multiple possible pathways are present for the product generated. Labels beside arrows indicate positions of bromines removed: o, ortho; m, meta; p, para. (A) Debromination pathways for octa-BDE technical mixture in culture G. (B) Debromination pathways for penta-BDE technical mixture in strain AG/culture G and strain DG. Boxed labels are pathways shared by both strain AG/culture G and strain DG, while other labels are pathways observed only in strain AG/culture G.

Culture G and strain AG exhibit the same debromination pathway for the penta-BDE mixture via a strict para debromination pattern. Nine PBDE congener products were observed (Fig. 4B). All four parent congeners (153 [245-245], 100 [246-24], 99 [245-24], and 47 [24-24]) in the penta-BDE mixture contain two para bromines (either single flanked or nonflanked). They were all debrominated in two consecutive para debromination steps to form congener 52 (25-25), congener 19 (26-2), congener 18 (25-2), and congener 4 (2-2). This particular preference for para bromines in both penta- and octa-BDE mixtures makes culture G/AG a potential para debromination specialist in field application. Strain DG removes only single-flanked para bromines in the penta-BDE mixture (congener 153→congener 101→congener 52, congener 99→congener 49) to form two tetra-BDE congeners as final products. It is thus possible that strains AG and DG employ different mechanisms in the para debromination of penta-BDEs.

DISCUSSION

In this study, an enriched culture G was found to possess unique debromination capability toward octa- and penta-BDE mixtures. Enrichment culture G and its isolates AG and DG exhibited excellent debromination performance by showing (i) rapid PBDE debromination activities (e.g., 33% removal at day 7 and 96% removal after 21 weeks on an octa-BDE technical mixture by culture G and 75% removal at day 33 on a penta-BDE technical mixture by strain AG) and (ii) extensive bromine removal, with culture G generating predominant tetra-BDEs from the octa-BDE mixture and di- to tetra-BDEs from the penta-BDE mixture. The debromination extent in culture G exceeds those of previously reported cultures on the octa-BDE mixture (12); i.e., congener 183 was debrominated beyond hexa-BDEs and penta-BDEs.

Interestingly, Acetobacterium in culture G plays an important role in PBDE debromination, which is confirmed by strain AG's debromination activity with PBDEs in the absence of TCE. The results for culture AG are thus distinguished from those of previous studies that required the presence of a primary electron acceptor (e.g., TCE or pentachlorophenol) (9–14). While the presence of Dehalococcoides (strain DG) in culture G is well expected, as Dehalococcoides is a known PBDE debrominator (11, 12), Acetobacterium sp. strain AG is novel in debromination because it is the first Acetobacterium strain that can remove halogens from aromatic rings. Strain AG debrominates PBDEs when growing on lactate, on pyruvate, and even in an organic-carbon-free medium amended with H₂-CO₂, which is favorable in bioaugmentation sites with various growth conditions.

Judging from the fact that PBDE debromination was enhanced immediately after respiking lactate (Fig. 1), PBDE debromination in strain AG should be a cometabolic process dependent on energy from lactate. This is further confirmed by experiments with growing and resting cells (constant cell numbers), which demonstrate a strong positive relationship of PBDE debromination activities with initial lactate concentration and cell numbers (Fig. 3). Despite being cometabolic, the debromination rates in culture G are among the highest in the reported cases of anaerobic debromination of PBDEs, with a maximum daily debromination rate of ∼30 nM day⁻¹ (e.g., day 2 to 4 and day 30 to 33 with the penta-BDE mixture [Fig. 2]), even at the same magnitude as observed in the metabolic debrominator culture GY2 (78.6 nM day⁻¹) (17). Rapid debromination was observed in the first 1 to 3 days after addition of lactate, suggesting that stepwise dosage of electron donor/carbon source may be preferred when applying culture G and strain AG in PBDE-contaminated sites. However, it is difficult to ultimately determine whether PBDEs are growth supporting. This is because (i) strain AG can grow solely on lactate as well as on other carbon sources added, and (ii) no significant difference in growth can be observed since the amount of PBDE added is too small (<1 μM).

We found that there are synergistic interactions between Acetobacterium and Dehalococcoides in culture G, as suggested by much higher PBDE debromination rates when the two strains were grown together than when they were separated (see Fig. S4D and E in the supplemental material). It is speculated that Acetobacterium may provide certain growth factors for Dehalococcoides for synthesis of necessary dehalogenases that are responsible for PBDE debromination, as implied by previous coculture studies with Acetobacterium/Dehalobacter (35), Sedimentibacter/Dehalobacter (36), and Enterococcus/Desulfitobacterium (37) (nondechlorinating supporter/dechlorinator). Acetobacterium spp. are known for producing vitamin B₁₂ (cyanocobalamin) (38–40), which could not be synthesized by Dehalococcoides, and it has already been proven that Acetobacterium can enhance growth of Dehalococcoides mccartyi strain 195 in lactate medium (18). Although vitamin B₁₂ (added at 50 μg/liter in this study) should not be a limiting factor in the growth of Dehalococcoides, the form of cobalamin added in may not be the one preferred by Dehalococcoides (41). Therefore, Acetobacterium possibly produces a specific form of cobalamin that stimulates debromination by Dehalococcoides. Future studies will try to elucidate the synergistic relationship between these two bacterial groups.

Identification of PBDE congeners and debromination pathways is important in evaluation of overall toxicity change and in determination of debromination patterns. Based on a 39-congener PBDE mixture and previously published PBDE congener retention time databases (31, 32), this study makes it feasible to reliably identify all the debromination products from octa- and penta-BDE mixtures, requiring only a limited number of PBDE congener standards for verification. Out of 19 PBDE congener products, 18 were identified (Fig. 4). The PBDE debromination pathways observed in culture G are different from those in previous studies (12, 17). The penta-BDE mixture was debrominated only through para debromination pathways in culture G and strains AG and DG (Fig. 4B) (culture G and AG remove all single-flanked and nonflanked para bromines, while strain DG removes only single-flanked para bromines). In contrast, the previously reported coculture GY2 of Desulfovibrio and Dehalococcoides is able to remove bromines at all three positions, with a preference toward ortho bromines (17). The octa-BDE mixture was debrominated primarily through para debromination in this study, with only one ortho debromination pathway confirmed, while no ortho debromination is present in the debromination of the penta-BDE mixture by culture G. The findings here are consistent with previous studies showing difficulties in removing ortho bromines in biotransformation processes (11, 12).

It is difficult to assess the impacts of the identified debromination pathways on the overall toxicity of PBDE mixtures, because toxicological data are available only for several common constituents in the penta-, octa-, and deca-BDE mixtures. Given the high specificity toward para debromination by culture G, dioxin-like PBDE congeners, if there are any, are not likely to form, since both para bromines are indispensable in a dioxin-like structure (42). One recent study suggested that single-substituted para bromine is critical for enhancing ryanodine receptor (RyR) activity in human kidney cells, which is relevant for the neurotoxicity of PBDEs (43). Again, this PBDE structure (single para bromine) could also be avoided in the debromination process by culture G because of its rapid removal of both para bromines. Moreover, several of the most environmentally abundant and toxic PBDE congeners, such as congeners 47, 99, and 100, can be removed by culture G. In total, it seems that the debromination of PBDEs by culture G does not yield more toxic products, but this still needs further evidence from toxicological studies.

To sum up, Acetobacterium sp. strain AG, which has unique debromination capabilities, was isolated. Acetobacterium spp. belong to acetogens, whose presence is ubiquitous in the environment (44). Although reductive dehalogenation by acetogens has been implicated for chloroethanes and chloroethenes (45–47), they are rarely shown to dehalogenate brominated/aromatic compounds, with only one such reported case (i.e., Desulfitobacterium chlororespirans strain Co23 on an octa-BDE mixture but with no detailed debromination profile or kinetics [12, 48]). The acetogenic autotroph Acetobacterium sp. strain AG, which is capable of rapid PBDE debromination, has deepened our understanding of the fate and biotransformation of PBDEs in the natural environment and also serves as a promising candidate for the bioremediation of these POPs at contaminated sites.