Phylogenetically Distinct Bacteria Involve Extensive Dechlorination of Aroclor 1260 in Sediment-Free Cultures

Microbial reductive dechlorination of the persistent polychlorinated biphenyls (PCBs) is attracting much attention in cleanup of the contaminated environment. Nevertheless, most PCB dechlorinating cultures require presence of sediment or sediment substitutes to maintain their dechlorination activities which hinders subsequent bacterial enrichment and isolation processes. The information on enriching sediment-free PCB dechlorinating cultures is still limited. In this study, 18 microcosms established with soils and sediments were screened for their dechlorination activities on a PCB mixture – Aroclor 1260. After one year of incubation, 10 out of 18 microcosms showed significant PCB dechlorination with distinct dechlorination patterns (e.g., Process H, N and T classified based on profiles of PCB congeners loss and new congeners formation). Through serial transfers in defined medium, six sediment-free PCB dechlorinating cultures (i.e., CW-4, CG-1, CG-3, CG-4, CG-5 and SG-1) were obtained without amending any sediment or sediment-substitutes. PCB dechlorination Process H was the most frequently observed dechlorination pattern, which was found in four sediment-free cultures (CW-4, CG-3, CG-4 and SG-1). Sediment-free culture CG-5 showed the most extensive PCB dechlorination among the six cultures, which was mediated by Process N, resulting in the accumulation of penta- (e.g., 236-24-CB) and tetra-chlorobiphenyls (tetra-CBs) (e.g., 24-24-CB, 24-25-CB, 24-26-CB and 25-26-CB) via dechlorinating 30.44% hepta-CBs and 59.12% hexa-CBs after three months of incubation. For culture CG-1, dechlorinators mainly attacked double flanked meta-chlorines and partially ortho-chlorines, which might represent a novel dechlorination pattern. Phylogenetic analysis showed distinct affiliation of PCB dechlorinators in the microcosms, including Dehalogenimonas and Dehalococcoides species. This study broadens our knowledge in microbial reductive dechlorination of PCBs, and provides essential information for culturing and stimulating PCB dechlorinators for in situ bioremediation applications.


Introduction
The use of harmful PCB mixtures (e.g., Aroclor 1260) in electrical transformers, hydraulic fluids and other areas has resulted in PCB contamination in soils and sediments of lakes, rivers, and harbors worldwide [1][2][3][4]. Due to their low vapor pressure and water solubility, the refractory PCBs may enter the food-chains through bioaccumulation [5], thus posing a threat to the health of human beings and ecosystems [6]. Removal of halogenated compounds from the contaminated sites can be achieved using various remediation technologies including natural attenuation, physical methods (e.g. capping), thermal and chemical treatment (e.g. microwave-generated steam technology and photocatalytic dechlorination), and biological processes employing bacteria or yeast [5,7]. Among them, bioremediation is an effective and economical approach to remove PCBs from contaminated environments, and the microbial removal of highly chlorinated PCB congeners is conducted anaerobically through microbial reductive dechlorination [8].
After the first report of in situ dechlorination of PCBs by anaerobes [9,10], evidence to date have demonstrated that in situ PCB dechlorination is widespread in many anaerobic PCB contaminated environments, including freshwater (pond, lake, and river), estuarine, and marine sediments [11]. To understand the potential for bioremediation at these sites, it is important to evaluate the possible indigenous dechlorinators, which is normally conducted by culturing microbes in the laboratory through establishing microcosms with indigenous soil or groundwater samples [12]. Thus far, many enrichment cultures have been obtained by using either single PCB congeners or PCB mixtures [11,13]. For example, a mixed culture enriched from Dutch sediment can dechlorinate 234-234-CB and 236-236-CB to pentaand tetra-CBs by attacking ortho, meta, or para-chlorines [14]. Two anaerobic PCB-dechlorinating enrichments with para-dechlorination specificities were obtained with sediments derived from Aroclor 1260-contaminated Woods Pond and PCB-free Sandy Creek Nature Center sediments, respectively [15]. Enriched cultures with extensive Aroclor 1260 dechlorination activity were also obtained by using effective primer compounds, e.g., 23456-CB [16] and brominated biphenyls [17]. Based on profiles of PCB congeners loss and new congeners formation by these enriched cultures and in situ field studies, eight distinct microbial dechlorination processes (i.e., Processes M, Q, H9, H, P, N, LP and T) have been identified [11,18].
However, most of previously developed enrichment cultures require the presence of sediments to maintain PCB dechlorination activities [8,11]. To further elucidate PCB dechlorination processes, isolation of PCB dechlorinators is necessary, which requires the development of sediment-free cultures [8]. To date, only a few sediment-free cultures exhibiting PCB dechlorination activity have been reported, e.g., two cultures were enriched on 2356-CB and 2345-CB by removing ortho-chlorines [19] and meta-/para-chlorines [20], respectively. Their PCB dechlorinators were subsequently identified to be Chloroflexi bacterium o-17 [21] and DF-1 [22] by using the PCR-DGGE method. Another tetrachloroethylene (PCE) -dechlorinating Dehalococcoides mccartyi strain 195 can also dechlorinate 23456-CB to 2346-, 2356-and 246-CBs [23]. However, 2356-CB, 2345-CB and 23456-CB are PCB congeners chlorinated on a single ring, which are not the major PCBs present at the contaminated sites [24]. For microbial reductive dechlorination of PCB mixtures (e.g. Aroclor 1260), it is more complicated to develop sediment-free cultures. For example, pure culture bacterium DF-1 isolated from 2345-CB dechlorination was also shown to extensively dechlorinate weathered PCB mixtures, nevertheless, in the presence of soil [25,26]. Until now, only culture JN has been successfully established for PCB mixture dechlorination in sediment-free form [27], in which Dehalococcoides species was identified to couple their growth with Aroclor 1260 dechlorination by Process N [28]. This study made a major progress to show that PCB dechlorinating bacteria can maintain their Aroclor 1260 dechlorination activities in defined mineral medium. In culture JN, the activities required the presence of silica powder as a carrier for Aroclor 1260 to increase the bioavailability [27]. By using the same strategy, Dehalococcoides sp. CBDB1, a pure culture dechlorinating chlorobenzenes, also showed the capability to dechlorinate Aroclor 1260 in a co-metabolic process after pregrowing on trichlorobenzenes [29]. The requirement of silica powder as a sediment-substitute may explain one key role of sediments in PCB dechlorinating cultures is to improve the bioavailability of PCB mixtures. However, employing silica sediment-substitute shall complicate the medium preparation and operation processes for enriching and isolating PCB dechlorinators. Thus far, all the PCB dechlorinating bacteria were identified to be either Dehalococcoides species (i.e. Pinellas subgroup) or phylogenetically related but distinct Chloroflexi bacteria (e.g., DF-1 and o-17) [8,30]. Therefore, information still remains limited on both maintaining Aroclor 1260 dechlorination activities in sediment-free cultures without the aid of sedimentsubstitutes and identification of their PCB dechlorinators.
In this study, 18 soil and sediment samples were collected from different locations to set up microcosms for screening of Aroclor 1260 dechlorinating microbes. From them, six sediment-free cultures with distinct dechlorination specificities were successfully established without amending any sediments/sediment-substitutes. Initial phylogenetic insights into key dechlorinators were also gained by using 16S rRNA gene-based techniques. In the sediment-free cultures, the growth of PCB dechlorinators coupled with Aroclor 1260 dechlorination was quantified by using realtime PCR.

Ethics Statement
No specific permits were required for the described field studies. The sampling activities did not involve any endangered or protected species, and sampling locations are not privately-owned or protected in any way.

Chemicals
Unless stated otherwise, chemicals were purchased from Sigma-Aldrich at the highest purity available. All PCBs were purchased from AccuStandard (New Haven, CT, USA). H 2 was obtained from a hydrogen generator (NM-H250, Schmidlin-DBS AG, Neuheim, Switzerland). The DNA extraction kits were obtained from Qiagen (QIAGEN, Hilden, Germany), and the GoldTaq DNA polymerase and related PCR reagents were purchased from Applied Biosystems (Foster City, CA, USA).
Microcosm preparation, culture transferring, and growth conditions A total of 18 samples were collected in four Asian countries, China (Wuhan in Hubei and Guiyu in Guangdong, China), Indonesia (West Java, Indonesia), Malaysia (Penang, Malaysia), and Singapore (Jurong Island, Singapore). The characteristics of samples are shown in Table 1. The sublayer soil and sediment samples (sampling depth, 5-20 cm) were acquired directly by filling sterile 50-ml plastic Falcon tubes that were capped and transported to the laboratory at an ambient temperature. Concentrations of PCBs in these samples were below detection limit. To control exposure of the samples to air, Falcon tubes were sealed with Parafilm (Pechiney Plastic Packaging Company, Chicago, IL, USA), and microcosm setup was conducted in anaerobic chamber as described soon after their arrivals [31]. Briefly, 90 ml of bicarbonate-buffered mineral salts medium amended with 10 mM of lactate were dispensed into 160 ml serum bottles containing ,10 grams of collected samples [32,33]. The mineral salts medium contains salts, trace elements and vitamins as shown in Table S1. L-cysteine and Na 2 S?9H 2 O (0.2 mM each) were added to the medium to achieved reduced conditions. The bottles were sealed with black butyl rubber septa (Geo-Microbial Technologies, Inc, Ochelata, OK, USA) and secured with aluminum crimp caps. After that, a 60 ml of Aroclor 1260 stock solution (50 mg of total PCBs per ml) in GC grade isooctane was spiked into the medium to a final nominal concentration of 30 ppm (or 81 mM). The microcosms were incubated in the dark at 30uC. PCB dechlorination activities were measured every four weeks by gas chromatograph equipped with an electron capture detector (GC-ECD) as described in the following section. Cultures were transferred when observing obvious PCB dechlorination activity. After three years, a total of six sediment-free cultures were obtained after at least 5 times of transferring supernatant of the active microcosm to the same medium (5%, v/v) as described above. Cultures amended with two individual PCB congeners (i.e., 2345-245-CB and 234-245-CB) were also prepared to determine their dechlorination pathways in the sediment-free cultures. All experiments were set up in triplicates. Duplicate abiotic controls and non-PCB controls were also set up for each experiment under the same conditions but without bacterial inocula and PCBs injection, respectively.

Analytical methods
PCBs in the culture medium were extracted with the solvent isooctane. Mixed liquor of 1 ml was withdrawn and subjected to a liquid-liquid extraction with an equal volume of isooctane in a  PCB congeners were listed from more to less decrease, but all are above 2 mol% decrease. c PCB congeners were listed from more to less increase, but all are above 2 mol% increase. ND, no PCB congeners experienced more than 2 mol% decrease or increase. doi:10.1371/journal.pone.0059178.t001 4 ml amber vial. The vial was vigorously shaken for 2 h and then centrifuged at 14,000 rpm for 10 min. The solvent phase of 0.5 ml was transferred to a 2-ml amber glass vial for subsequent GC analysis. PCBs were measured with a gas chromatograph (GC) 6890 N equipped with an electron capture detector (GC-ECD) (Agilent, Santa Clara, CA, USA) and a DB-5 capillary column (30 m60.32 mm60.25 mm film thickness; J&W Scientific, Folsom, CA, USA). The temperature program was initially held at 170uC for 5 min, increased at 2.5uC min 21 to 260uC, and held for 10 min. Injector and detector temperature were set at 250uC and 300uC, respectively. Nitrogen was used as a carrier gas at a flow rate of 1.2 ml min 21 . Sample of one ml was injected into the GC inlet in a splitless mode. The elution time of all 209 PCB congeners was determined with PCB congener mixtures 1 through 9 from AccuStandard. The relative elution time of the PCBs in these mixes were published for DB-5 column [34]. PCBs were quantified by using a customized calibration standard prepared from Aroclor 1260 plus the congeners that are known to be frequent dechlorination products as described [35]. Additional congeners were quantified from standards prepared from the AccuStandard PCB congener mixtures. In all cases we used a seven-point calibration curve. The total moles of PCBs kept roughly the same throughout the whole dechlorination process. The microbial reductive dechlorination of PCBs result in profile changes of mol% of PCB congeners, which can be quantified based on their weight percent distributions and molecular weight as described previously [13,27,36]. Mole percent value for each congener, total number of chlorines per biphenyl and the PCB homolog distribution were calculated as described previously [17].

DNA extraction, PCR, and sequencing
Total community DNA was extracted from 1 ml of cell pellets collected from dechlorinating cultures as well as the controls according to the manufacturer' instructions but with minor modifications [37]. The concentration of the nucleic acid was determined by a Nanodrop-1000 instrument (NanoDrop Technologies, Wilmington, DE, USA). PCR amplifications of 16 S rRNA gene sequences were conducted on a MastercyclerHcycler (Eppendorf, Hamburg, Germany) under conditions as described previously [33]. The primers used in this study are shown in Table  S2. PCR products were sequenced and aligned as described previously [38].

Two step DGGE (2S-DGGE)
2S-DGGE was developed to obtain full-length 16 S rRNA gene sequences of minor populations which might be difficult to be recovered from clone libraries [39]. PCR amplifications were first conducted with 8F and genus-specific reverse primer (i.e., DEB630R, DHC710R and DHCG812R) on community DNA of each culture. Then the diluted PCR products (650 dilutions) of Dehalobacter/Dehalococcoides/Dehalogenimonas 16 S rRNA gene sequences were subjected to subsequent 2S-DGGE procedures [39]. PCR products (12 ml) amplified with the GC-clamped universal primer sets were separated on an 8% polyacrylamide gel with a gradient range of 30-60% (100% denaturant consisted of 7 M urea and 40% deionized formamide) in 0.5 6 TAE buffer. Gradient gels were cast with Bio-Rad's Model 475 gradient delivery system (Bio-Rad, Hercules, CA, USA). The electrophoresis was performed for 15 h at a constant electric current of 30 mA and a temperature of 60uC with the D-Code Mutation Detection System (Bio-Rad, Hercules, CA, USA). Gel images of DNA stained with SYBRH Gold (Invitrogen, Carlsbad, CA, USA) were taken by using a Molecular Imager Gel Doc XR System (Bio-Rad, Hercules, CA, USA). The DNA bands were excised and their DNA fragments were extracted by using the QIAEX II Gel Extraction Kit (QIAGEN, GmbH, Germany). The captured DNAs were then PCR re-amplified, and re-analyzed by DGGE to confirm that single bands were obtained before sending the PCR re-amplified products for sequencing. Based on sequencing data (from base 8-529, E. coli numbering), strain-specific forward primers would be designed or chosen from currently available primers for PCR amplification of the left 16 S rRNA gene sequence (from base 529-1392, E. coli numbering). Second round DGGE was employed to confirm the PCR amplification specificity. Lastly, the second round specifically amplified PCR products were sent for sequencing, and the full-length 16 S rRNA gene sequences can be obtained by assembling the two partial sequences.

Illumina high throughput sequencing analysis of 16 S rRNA genes
To analyze the taxonomic composition of the sediment-free PCB dechlorinating cultures, the V3 region of the 16 S rRNA gene (from base 334-533, E.coli numbering) was chosen for PCR amplification with the eubacteria primer sets containing barcode sequences, 341F (59-Fusion A-Barcode-ACTCCTACGGGAGG-CAGCAG-39) and 533R (59-TTACCGCGGCTGCTGGCAC-39). A total of six bar-coded [40] forward primers were used to differentiate the individual samples. Amplified PCR products were purified by using QIAquick PCR purification kit (QIAGEN, GmbH, Germany) according to the manufacturer's instructions. Then the six PCR samples with equal amounts were mixed together for subsequent Illumina high throughput sequencing. Illumina (Highseq2000, Illumina, San Diego, CA, USA) high throughput sequencing services were provided by BGI (Hongkong, China). Raw sequencing reads were checked for their quality through elimination of sequences that did not perfectly match the proximal PCR primer and that with short sequencing length (,130 nt). A total of 287,069 pair-end reads were obtained for these PCB dechlorinating cultures, and each sequence has an average read length of 150 bp. Pair-end reads were combined to form longer composite reads (170,200 bp) by using the SHERA [41] software package. Sequence alignments were conducted with each subset reads based on NAST [42], and with other settings kept at their default values as described [43]. After NAST alignment, aligned subsets were merged into one Microsoft Excel file, in which sequences were clustered (based on 97% of sequence similarity) according to template ID. Manual adjustments were performed to improve the alignment and clustering whenever necessary. Representative sequences for each cluster were identified through Classifier [44] and BLAST analysis [45], which were further utilized to construct phylogenetic tree by using MEGA4 [46]. Relative abundances of predominant bacterial genera were showed by using BAND [47]. qPCR A TaqManHquantitative real-time PCR (qPCR) (ABI 7500 Fast real-time PCR system; ABI, Foster City, CA, USA) assay was performed in triplicates for PCB dechlorinating cultures by using Bacteria and Dehalococcoides 16 S rRNA gene-targeted primers/ probes, respectively [37,38]. Dehalogenimonas species was quantified by using SYBRH green assays and Dehalogenimonas 16 S rRNA gene-targeted primers [48]. The primer and probe sequences used in this study were shown in Table S2. A calibration curve was obtained by using 10-fold serial dilutions of known plasmid DNA concentrations. The standard curves spanned a range of 10 2 to 10 8 gene copies per ml of template DNA. Nuclease-free water or plasmid without an insert was used as the negative control. The cell growth supported by chlorine removal was calculated using cell-growth (cells/ml)/total chlorine removal (nmol/ml).

Sample nomenclature
In this study, the samples were denoted based on the locations from which they were collected. The abbreviation used for locations were as follows: CW, Wuhan, China; CG, Guiyu, China; ID, Indonesia; MY, Malaysia; SG, Singapore. For example, CW-4 indicates a sample from Wuhan, China.

Nucleotide sequence accession numbers
The nucleotide sequence data obtained in this study were submitted to the Genbank with the following accession numbers: JQ990318-JQ990328.

Development of sediment-free PCB dechlorinating cultures
To obtain sediment-free PCB dechlorinating cultures, the 12 active microcosms were transferred in defined medium amended with Aroclor 1260 (30 ppm or 81 mM) and lactate (10 mM). Subsequently, six sediment-free PCB dechlorinating cultures (i.e., CW-4, CG-1, CG-3, CG-4, CG-5 and SG-1) were obtained after five serial transfers (5% inocula, v/v). The PCB dechlorination activities can be sustained and transferred in defined medium amended with lactate. However, upon replacing lactate with acetate (10 mM) and H 2 (5610 4 Pa or 0.40 mM), no PCB dechlorination activities were observed in the 6 sediment-free cultures after two transfers. For the six sediment-free cultures, two PCB congeners (i.e., 2345-245-CB and 234-245-CB, two of the most abundant congeners in Aroclor 1260 and comprising ,20% by mole of total PCBs) were then spiked to the cultures to study their PCB dechlorination pathways. The data of mole percent value for each congener and PCB homolog distribution were collected after 6 months of incubation for cultures CW-4, CG-1, CG-3, CG-4 and SG-1; and after 3 months for culture CG-5.
(i) In culture CW-4 ( Figure 1B Figure 1E). The difference between these two cultures is that culture CG-4 is unable to effectively dechlorinate 245-245-CB into lower chlorinated PCB congeners and further dechlorination of the produced penta-CBs into tetra-CBs proceeded at a slow rate.

Identification of PCB dechlorinators
To acquire information on the possible PCB dechlorinators present in the active microcosms, PCR amplifications were conducted with primers (Table S2) specific for the 16 S rRNA genes of known PCB dechlorinating bacteria (i.e., Dehalococcoides and o-17/DF-1-type Chloroflexi) and the obligate dehalogenator, Dehalobacter species [27,53]. To have a broader match, a new primer set (i.e., DEB165F and DEB630R in Table S2) was utilized to screen Dehalobacter species from these cultures, of which the specificity was verified by using RDP's ProbeMatch [54]. PCR amplification with o-17/DF-1 specific primers (8F/Dehal1265R in Table S2) cannot detect the involvement of o-17/DF-1-type Chloroflexi in the microcosms. All known PCB dechlorinators and the Dehalobacter species were absent in the two microcosms of ID-1 and MY-1 based on PCR amplification with genus-specific primers. Among the other 10 microcosms, Dehalobacter species were detected only in microcosm CW-1, whereas Dehalococcoides species were detected to be present in the other nine microcosms.
To have a comprehensive understanding of the microbial community in sediment-free PCB dechlorinating cultures, Illumina high throughput sequencing analysis was conducted to reveal their phylogenetic compositions. Results showed that the cultures were comprised mainly of the phyla Chloroflexi, Synergistetes, Firmicutes, Proteobacteria, Spirochaetes, Bacteroidetes, and Thermotogae Bacteria (Figure 4). Dehalococcoides within Chloroflexi phylum was the only known PCB dechlorinator present in these sediment-free cultures. Normal DGGE analysis with Dehalococcoides-specific GC-clamped primers (1FGC/259R in Table S2) also showed that only single 16 S rRNA genotype Dehalococcoides bacteria present in the sediment-free cultures ( Figure S1). In particular, Dehalococcoides populations were present at abundances ranging from 0.37% to 14.85% of the total bacterial community, which is comparable with Dehalococcoides population (3.74%) in highly enriched JN culture [27]. In addition, the highest Dehalococcoides population percentage (14.85%) in culture CG-5 corroborated the observed fast and extensive PCB dechlorination activities. Interestingly, Illumina high throughput sequencing analysis suggested the involvement of populations from a newly characterized genus, Dehalogenimonas (2.16%), in reductive dechlorination of Aroclor 1260 in culture CG-3. Subsequent 2S-DGGE analysis showed the full-length 16 S rRNA gene sequence of Dehalogenimonas in culture CG3 (Dhg-CG3) shares 96% similarity (55 bp difference over 1493 bp) with that of Dehalogenimonas lykanthroporepellens BL-DC-9 ( Figure 3B). qPCR analysis showed coupled growth of PCB dechlorinating bacteria (i.e. Dehalococcoides and Dehalogenimonas) with chlorine removal from PCBs in the sediment-free cultures ( Figure 5). Along with an average chlorine removal of 36.49 nmol per mL (from 17.97 to 64.23 nmol per mL), the cell number of PCB dechlorinating bacteria grew to 1.1610 7 cells per mL (average number) in the six cultures. The resulting average growth of 3.30610 14 cells per mole of chlorine released is comparable to that (9.25610 14 cells per mole of chlorine released) in sediment-free JN cultures [28]. No obvious cell growth of PCB dechlorinators was observed in their corresponding controls that were not spiked with Aroclor 1260 (data not shown). The Desulfovibrio species reported to be imperative for PCB dechlorination by bacteria DF-1 [25] were also ubiquitously present in the PCB-dechlorinating sediment-free cultures (ranging from 0.01% to 13.86% of the total bacterial community).

Discussion
In this study, sediment-free PCB-dechlorinating cultures (i.e., CW-4, CG-1, CG-3, CG-4, CG-5 and SG-1) have been successfully obtained from microcosms set up with soils, sediments and sludge, which are different from previous Aroclor1260dechlorinating cultures requiring the presence of sediments or sediment-substitutes [8,11,15,27,29,56]. In the six sediment-free cultures, PCB dechlorinating bacteria are able to grow on Aroclor 1260 in low PCB bioavailability without the aid of sedimentsubstitutes. Recently, a pure culture was also reported to debrominate octa-brominated diphenyl ether (octa-BDE) mixture in even lower bioavailability (based on concentration and water solubility of octa-BDE mixture) [57]. Therefore, some dechlorinating bacteria are able to maintain dehalogenation activities in low substrate bioavailability, which shall be easier to be isolated by using the traditional serial-dilution-to-extinction method compared to dechlorinators requiring sediment-substitutes. During our cultivation process, microcosms ID-1 and MY-1 lost their PCB dechlorination activities in the second transfer due to their low dechlorination activities, while CW-1, CW-2, CW-3 and CG-2 gradually lost their PCB dechlorination capabilities with the elimination of sediments in the serial transfers. This phenomenon may be explained by the distinct roles of sediments, e.g., supply specific nutrients for PCB dechlorinators [11,56], and improve PCB bioavailability [8,27]. For example, when serving to improve PCB bioavalability, sediments were replaced by silica powder to cultivate the JN culture to extensively dechlorinate Aroclor 1260 [27]. Similarly, Dehalococcoides sp. strain CBDB1 can also extensively dechlorinate Aroclor 1260 in the presence of silica powder when pregrown on trichlorobenzenes. However, in many PCB dechlorinating cultures, sediments likely play both roles simultaneously, since sediments can not be simply substituted with silica powder in the defined medium, e.g., PCB dechlorination activities in silica powder -dosed JN cultures can not be sustained at the third transfer unless yeast extract was added to the medium [27]. Nevertheless, during our cultivation of sediment-free cultures, appropriate carbon source selection appears to be a key factor. Comparing with acetate or/and formate used in previous studies [27][28][29], lactate serves as an excellent carbon source to support the formation of robust PCB-dechlorinating microbial consortia in the absence of sediments or sediment substitutes. This possibly is due to the lactate-supporting syntrophic bacteria (Figure 4) for maintaining PCB dechlorination by supplying metabolic elements, such as amino acids and vitamins by Clostridium and fatty acids by Syntrophus. Actually, mixed carbon sources (e.g., glucose, lactic acid, pyruvic acid and acetic acid) have been applied to defined medium in order to boost the growth of PCB dechlorinators in previous studies [11,14]. In all, the successful cultivation of the six Aroclor 1260 -dechlorinating cultures (i.e., CW-4, CG-1, CG-3, CG-4, CG-5 and SG-1) without the presence of any sediments or sediment substitutes can facilitate future isolation and characterization of PCB dechlorinators.
Thus far, PCB dechlorination pathway has been categorized into eight different dechlorination patterns (i.e., Processes M, Q, H9, H, P, N, LP and T) [8,10,11,18,49]. In the six sediment-free cultures, three distinct PCB dechlorination patterns (e.g., H, N and T) were observed based on PCB congener profile changes from dechlorination of Aroclor 1260 and the dechlorination products of two individual PCB congeners (i.e., 2345-245-CB and 234-245-CB). Impressively, Process H is the dominant PCB dechlorination pattern observed in four of six sediment-free cultures (i.e., cultures CW-4, CG-3, CG-4 and SG-1), which implies the wide distribution of PCB dechlorination activities by Process H in the environment. Dechlorination Process H removes flanked paraand double flanked meta-chlorines (i.e., 34, 234, 245, 2345), which was first observed in situ both in Acushnet Estuary and in some parts of the Hudson River [45], and subsequently observed in sediment microcosms originated from Hudson River sediment [51] and in a pure culture of Dehalococcoides sp. CBDB1 [29]. Among the six sediment-free cultures, CG-5 showed the most extensive dechlorination of Aroclor 1260 via dechlorination Process N, showing its great potential for bioremediation applications at PCB contamination sites. For the last sediment-free culture CG-1, PCB dechlorinators mainly attack double flanked meta-chlorines and partially ortho-chlorines, which represent a novel PCB dechlorination pattern. Currently, the microbial dechlorination of orthochlorines from PCBs has not been categorized into anyone of the eight summarized PCB dechlorination processes. However, ortho dechlorination was observed in many cultures spiked with single PCB congeners [19,58,59]. A similar dechlorination pattern were also reported in anaerobic slurries of estuarine sediments amended with Aroclor 1260 [36].
Dechlorinating bacteria are normally present in PCB dechlorinating microcosms or sediment-free cultures as minor populations, e.g. Dehalococoides in JN cultures (3.74% of total population abundance) [28] and in sediment-free culture CG-3 (0.36% of total population abundance in this study). This poses a challenge to traditional genotyping methods (e.g., DGGE and T-RFLP) for identification of functional microbes in a mixed community [60]. Hence, this study employed Illumina high throughput sequencing together with 2S-DGGE to capture microbial community profiles and further phylogenetically characterize the key functional bacteria with full-length 16 S rRNA gene sequences. The 16 S rRNA gene-based analysis showed phylogenetically distinctive bacteria involved in reductive dechlorination of Aroclor 1260, i.e., Dehalogenimonas species and all three subgroups of Dehalococcoides (i.e., Cornell, Victoria and Pinellas). The Dehalococcoides bacteria are obligate dechlorinators and can only grow on halogenated compounds. Strains of the newly identified Dehalogenimonas genus also showed their dependent growth on halogenated compounds [61,62]. Thus, the presence of these obligate dechlorinators suggests their involvement in PCB dechlorination. To date, knowledge about the phylogenetic information of PCB dechlorinators is still limited. All PCB dechlorinators were identified to be either Dehalococcoides or o-17/DF-1 like bacteria through culture enrichment and molecular techniques (e.g., DGGE). Furthermore, the previously identified Dehalococcoides bacteria for Aroclor 1260 dechlorination were all affiliated to the Pinellas subgroup, e.g., DEH10 in an Aroclor1260-dechlorinating microcosm [63], Dehalococcoides in JN cultures [27,28], and Dehalococcoides sp. CBDB1 [29]. This study shows the involvement of all three subgroups (i.e., Cornell, Victoria and Pinellas) of Dehalococcoides and Dehalogenimonas bacteria in reductive dechlorination of Aroclor 1260. qPCR analysis further confirmed the coupled growth of Dehalococcoides and Dehalogenimonas with PCB dechlorination in sediment-free cultures. Such information would serve to broaden our knowledge of the phylogenetic diversity of PCB dechlorinators.
In conclusion, six sediment-free cultures were successfully established for reductive dechlorination of Aroclor 1260, which opened the door for following-up isolation and characterization of PCB dechlorinators. Phylogenetic analysis of the dechlorinating bacteria also expanded the diversity of known PCB dechlorinators. In the sediment-free cultures, both Dehalococcoides and Dehalogenimonas bacteria can couple their growth with PCB dechlorination. These observations would have a profound impact on cultivation and enrichment of PCB dechlorinators for future in situ bioremediation.

Supporting Information
Figure S1 DGGE profile of Dehalococcoides (Dhc) species present in the six sediment-free cultures with Dhc-specific GC-clamped primers 1FGC/259R. (TIFF)