Identification of Multiple Dehalogenase Genes Involved in Tetrachloroethene-to-Ethene Dechlorination in a Dehalococcoides-Dominated Enrichment Culture

Chloroethenes (CEs) are widespread groundwater toxicants that are reductively dechlorinated to nontoxic ethene (ETH) by members of Dehalococcoides. This study established a Dehalococcoides-dominated enrichment culture (designated “YN3”) that dechlorinates tetrachloroethene (PCE) to ETH with high dechlorination activity, that is, complete dechlorination of 800 μM PCE to ETH within 14 days in the presence of Dehalococcoides species at 5.7 ± 1.9 × 107 copies of 16S rRNA gene/mL. The metagenome of YN3 harbored 18 rdhA genes (designated YN3rdhA1–18) encoding the catalytic subunit of reductive dehalogenase (RdhA), four of which were suggested to be involved in PCE-to-ETH dechlorination based on significant increases in their transcription in response to CE addition. The predicted proteins for two of these four genes, YN3RdhA8 and YN3RdhA16, showed 94% and 97% of amino acid similarity with PceA and VcrA, which are well known to dechlorinate PCE to trichloroethene (TCE) and TCE to ETH, respectively. The other two rdhAs, YN3rdhA6 and YN3rdhA12, which were never proved as rdhA for CEs, showed particularly high transcription upon addition of vinyl chloride (VC), with 75 ± 38 and 16 ± 8.6 mRNA copies per gene, respectively, suggesting their possible functions as novel VC-reductive dehalogenases. Moreover, metagenome data indicated the presence of three coexisting bacterial species, including novel species of the genus Bacteroides, which might promote CE dechlorination by Dehalococcoides.


Introduction
Chloroethenes (CEs) such as tetrachloroethene (PCE) and trichloroethene (TCE) have been used extensively in dry cleaning and as degreasing agents. Consequently, they are commonly detected in groundwater because of improper disposal and accidental spills. Dehalorespiring bacteria, such as Dehalococcoides, which reductively dechlorinate CEs via respiration [1], have been applied as biocatalysts to remedy environments contaminated with CEs [2]. However, the accumulation of toxic compounds such as cis-dichloroethene (cis-DCE) and vinyl chloride (VC) as incomplete dechlorination products due to the limited distribution of Dehalococcoides, which completely dechlorinate CEs to ethene (ETH), has been reported [2]. To date, only members of the genus Dehalococcoides are known to convert these toxic intermediates to nontoxic ETH [2]. So far, no isolated strain can completely dechlorinate PCE to ETH without accumulation of toxic intermediates in pure culture [3]. Complete dechlorination of PCE to ETH has been rarely demonstrated, even in complexed microbial communities, and the time required for PCE-to-ETH dechlorination in consortia is highly variable, depending on the microbial community and PCE concentration [4][5][6][7]. In general, consortia containing non-dehalorespiring bacteria together with Dehalococcoides show faster dechlorination [8,9], and several possible functions of non-dehalorespirators in dehalorespiring consortia have been suggested [8][9][10][11][12].

BioMed Research International
The ability of Dehalococcoides to partially or completely dechlorinate CEs as well as other organohalides depends largely on reductive dehalogenase (Rdh). Rdhs identified in Dehalococcoides and other dehalorespiring bacteria consist of a catalytic subunit (RdhA) encoded by the rdhA gene and membrane-anchoring protein (RdhB) encoded by rdhB neighboring to rdhA [13,14]. In addition, several other genes assumed as Rdh-associated genes are frequently located close to rdhA and rdhB and are suggested to be involved in transcription regulation (i.e., rdhS, rdhP, and rdhR), transcription regulation/electron transport (rdhC), or maturation of Rdhs (i.e., rdhF, rdhG, rdhH, and rdhI) [1].
To date, several hundred rdhAs have been sequenced, although specific substrates have been determined for only a few RdhAs that were successfully partially purified from Dehalococcoides; VcrA [15] and BvcA [16] show apparent dechlorination of TCE to ETH and cis-DCE to ETH, respectively, while PceA [17] and TceA [18] can dechlorinate PCE to TCE and TCE to VC, respectively. Besides the CEs-RdhAs that were identified as RdhA for CEs on the basis of enzymatic activity, several RdhAs have been suggested to dechlorinate CEs on the basis of their high sequence similarity to these experimentally proven CE-RdhAs [19][20][21] and high transcription of rdhA specifically induced by CEs [22,23]. Reverse transcription quantitative PCR (RT-qPCR) is a relatively straightforward and promising screening tool for finding novel functions of rdhA from the clear correlation between high transcription levels of known CE-RdhAs and their actual function [23,24].
In this study, we successfully obtained a consortium containing Dehalococcoides that showed high, complete PCEto-ETH dechlorination activity, that is, dechlorination of 800 M PCE to ETH within 14 days. Four rdhAs are suggested to be involved in PCE-to-ETH dechlorination on the basis of their transcription response to CEs. Notably, two out of the four rdhAs had been never proven as rdhA for CEs and, thus, are novel candidate rdhAs involved in CE dechlorination. In addition, metagenome data showed the presence of three coexisting bacterial species, including novel species of the genus Bacteroides, and suggested they might promote CE dechlorination by Dehalococcoides.

Enrichment of Bacteria Dechlorinating
CEs. Sediment from Arako River (Nagoya city, Aichi prefecture, Japan) was used as the inoculum for the enrichment of CEdechlorinating bacteria. Approximately 10 g (wet weight) of the sediment was introduced into 60 mL serum bottles containing 20 mL of distilled water supplemented with 1 mg/L resazurin. The bottles were purged with N 2 gas for 15 min and sealed with Teflon-lined rubber stoppers and aluminum crimp caps. The bottles were purged once more with a gas mixture of H 2 and CO 2 (4 : 1, v/v) and then supplemented with 20 mM acetate and either PCE (1 mM) or cis-DCE (500 M) as an e − acceptor and incubated at 28 ∘ C. After incubation for 1 month, 100 L of the headspace gas was analyzed in a gas chromatograph (GC) as described below.
Microcosms showing dechlorination of PCE or cis-DCE were transferred to 20 mL of a mineral medium (DHB-CO 3 medium) in a 60 mL serum bottle at a 5% transfer rate. The DHB-CO 3 medium included the following components (per liter): 1 g NaCl; 2.5 g NaHCO 3 ; 0.5 g KCl; 0.5 g NH 4 Cl; 0.1 g CaCl 2 ⋅2H 2 O; 0.1 g MgCl 2 ⋅6H 2 O; 0.2 g KH 2 PO 4 ; 1 mL of 1 mg/L resazurin solution; 1 mL of trace element solution SL10 [25]; 10 mL of vitamin solution [26]; 1 mL of Se/W solution [25]; and 10 mL of titanium (III) trinitriloacetic acid solution [27]. The medium was prepared under anaerobic condition with flashing of a mixture of N 2 and CO 2 (4 : 1, v/v), as described previously [28]. Prior to inoculation, the headspace was exchanged aseptically with a gas mixture of H 2 and CO 2 (4 : 1, v/v). The transferred cultures were repeatedly incubated and transferred every month after observation of the dechlorination. The cultures supplemented with PCE showed weak dechlorination activity, and therefore, further enrichment was stopped after the third transfer. On the other hand, cultures supplemented with cis-DCE showed stable and complete dechlorination activity. Therefore, only the culture supplemented with cis-DCE was maintained and was designated "YN3." Time-course monitoring of the CEs and dechlorinated products in YN3 was carried out in freshly prepared culture (20 mL) inoculated with YN3 from the previous passage (1 mL) and supplemented with either 70 M PCE or 500 M cis-DCE in 60 mL serum bottles. To assay the applicability of YN3 for dechlorination of high concentrations of PCE, YN3 grown with 70 M PCE was additionally spiked with 200-800 M PCE. A portion of YN3 was periodically withdrawn for monitoring of 16S rRNA gene copy number for total bacteria and specifically, for Dehalococcoides, as described below. To test the ability of YN3 to dechlorinate TCE, trans-DCE, or 1,1-DCE, 200 M of these compounds was added to cultures.

Chemical Analysis.
CEs and ETH were detected and quantified by gas chromatography (GC), as described previously [36]. In brief, 100 L of headspace gas was withdrawn using a gastight syringe and injected manually into a GC-2014 (Shimadzu, Kyoto, Japan) equipped with flame ionization detector and Porapak Q column (GL Sciences Tokyo, Japan). Nitrogen was used as a carrier gas at a flow rate of 35 mL/min. The injection, detection, and column temperatures were set at 200 ∘ C.

DNA Extraction.
For genomic DNA extraction, cells from YN3 were collected by centrifugation at 15,000 ×g for 15 min and stored at −20 ∘ C until DNA extraction. Genomic DNA was extracted as described previously [37].  [40] and Rapid Annotation using Subsystem Technology (RAST) (ver. 2.0; http://rast.nmpdr.org) pipelines [41]. The contigs obtained from the metagenomic reads were taxonomically classified based on DNA sequence similarity of the entire contigs by NCBI BLASTN (https://blast.ncbi.nlm.nih.gov). In this study, contigs having at least 90% sequence similarity to a known bacterial sequence were classified at the genus level [42,43]. Contigs showing poor similarity with partial sequences in the known genome were classified based on similarity of the translated sequences of individual genes presenting in the contigs with genes in the published genome by BLASTP. Predicted coding sequences (CDSs) of all contigs were classified into functional categories by matching them to the SEED database provided within the RAST pipeline [44]. The phylogenetic trees of 16S rRNA genes and RdhAs were constructed using MEGA 6 software [45].

Reverse
Transcription-(RT-) qPCR. mRNA copy numbers of transcribed rdhAs were examined by RT-qPCR in YN3 spiked with CEs, using specific primers (Table S1, supplementary material). The YN3 used for the RT-qPCR experiment was incubated for approximately 30 days to complete the dechlorination of PCE or cis-DCE to ETH, and the headspace gas was exchanged with filter-sterilized N 2 to eliminate remaining CEs and ETH. It was then incubated for another 5 days without addition of any CEs as starvation period to attenuate mRNA in the cells. Twentymilliliter aliquots of starved YN3 were transferred into serum bottles and spiked with PCE (100 M), TCE (200 M), cis-DCE (500 M), or VC (100 M). An unspiked culture was prepared in parallel as a negative control. After 9-15 h of culture showing dechlorination activity, cells were collected by centrifugation for RNA and DNA extractions. Cell pellets used for RNA extraction were spiked with mRNA of the luciferase-encoding gene (luc) (Promega, Fitchburg, WI, USA) and then subjected to RNA extraction using ISOGEN II (Nippon Gene, Tokyo, Japan). RT-qPCRs for rdhA and luc quantification were carried out in parallel using the One Step SYBR PrimeScript RT-PCR Kit II (TaKaRa, Otsu, Japan). mRNA copies of rdhA in YN3 were normalized to spikedin luc mRNA. The RNA recovery rates of the samples in this study varied with total RNA concentration and were in the range of 2-17%. The transcription rate of rdhA was calculated by dividing the mRNA copy numbers by the gene copy numbers determined by qPCR in parallel.

Enrichment of PCE-to-ETH Dechlorinating Culture.
A serial transfer culture established from CE-contaminated river sediment, using 1 mM PCE and 500 M cis-DCE, yielded two enrichment cultures: an unstable PCEdechlorinating culture and the stable cis-DCE-to-ETHdechlorinating YN3 ( Figure S1, supplementary material). YN3 maintained activity to dechlorinate 500 M cis-DCE to ETH within 10 days even after 50 transfers over five years. In contrast, the PCE-dechlorinating culture lost its dechlorination activity after the third transfer. Next, YN3 was tested for dechlorination of PCE at a low concentration (70 M) and showed stable PCE-to-ETH dechlorination activity. Figure 1(a) shows the changes in PCE and the dechlorination products in YN3 throughout incubation. The culture supplemented with 70 M PCE started to dechlorinate PCE to TCE at 6 days, reaching a maximum of 38 ± 6 M TCE at day 16. The dechlorination of TCE started at day 11, and the main dechlorination products of TCE were ETH together with cis-DCE and VC, indicating the immediate dechlorination of TCE to ETH via cis-DCE and VC in YN3. At day 25, the concentration of ETH reached a maximum (54 ± 2 M) and accounted for 77% of the added PCE. The results clearly demonstrated the ability of YN3 to completely dechlorinate PCE into ETH via TCE, cis-DCE, and VC. A PCE spiking experiment showed that YN3 could dechlorinate 200-800 M of PCE within 14 days (Figure 1(b)). YN3 also dechlorinated 200 M 1,1-DCE to ETH via VC within 2 weeks without the accumulation of intermediates; however, it showed no dechlorination activity for trans-DCE.

Population of Dehalococcoides and Non-Dehalorespiring
Bacteria in the YN3 Metagenome. The metagenome of YN3 indicated the presence of the single known dehalorespiring species, Dehalococcoides, in the enrichment culture. The 16S rRNA gene designated as phylotype YN3-01 of the genus Dehalococcoides was identical to the 16S rRNA genes of D. mccartyi strains CG5, IBARAKI, CBDB1, and DCMB5 in the Pinellas subgroup of Dehalococcoides ( Figure S2, supplementary material), one of the three subgroups of Dehalococcoides proposed by Hendrickson et al. [47]. The 16S rRNA gene of the phylotype YN3-01 was identical to the 16S rRNA genes of strains CG5, CBDB1, IBARAKI, and DCMB5. Strains CBDB1, CG5, and DCMB5 can dechlorinate PCE to TCE or DCE but not DCE and VC [3,24,48]. The partial inconsistency between the phylogeny of the 16S rRNA gene and the dechlorination spectrum observed in this study can be explained in two ways: multiple strains of Dehalococcoides are present in YN3 or rdhAs for cis-DCE and VC were horizontally acquired independently of the core genes [49,50].
To confirm PCE-to-ETH dechlorination by Dehalococcoides in YN3, the Dehalococcoides population was monitored by 16S rRNA gene-based qPCR analysis. Figure 2 shows the qPCR data for YN3 dechlorinating 70 M PCE (Figure 1(a)) based on 16S rRNA primers specific for Dehalococcoides and total bacteria. YN3 contained 0.86 ± 1.5 × 10 7 copies of the Dehalococcoides 16S rRNA gene/mL of culture at day 0, which increased to 1.6 ± 0.69 × 10 8 copies/mL at day 21, at which PCE was completely dechlorinated to ETH. Meanwhile, the total bacterial population in YN3 showed 3.0 ± 5.2 × 10 7 copies/mL at day 0, which increased to 6.7 ± 4.0 × 10 8 copies/mL at day 21. Throughout the incubation, Dehalococcoides species comprised 15-28% of the total bacterial population. Similar growth trends were observed for Dehalococcoides and total bacteria in YN3 dechlorinating cis-DCE, in which Dehalococcoides accounted for 31-45% of the total bacteria ( Figure S3, supplementary material). Bacteroides species, non-dehalorespiring bacteria that existed together with Dehalococcoides species in YN3, reportedly have five 16S rRNA gene copies [51], while Dehalococcoides spp. harbor a single copy [49,52], suggesting that the Dehalococcoides population in YN3 was potentially underestimated. YN3 spiked with 200-800 M PCE showed increasing Dehalococcoides copy numbers, as summarized in Table  S2 (supplementary material), indicating their dehalorespiring growth at higher concentrations of PCE.  the nearest strains according to the above 16S rRNA gene analysis (Figure 3(a)). The genome size (1.34 Mbp) and CDS number (1420) in the Dehalococcoides metagenome in YN3 were slightly lower than those of the closest strains, which are 1.39-1.43 Mbp, containing 1413-1477 CDSs. The functional distributions of the genes from the Dehalococcoides metagenome and the related strains were very similar, except for minor variation in some categories.

Non-Dehalococcoides Metagenome.
Genomic fragments of non-Dehalococcoides bacteria in YN3 were affiliated to not only Bacteroidetes, but also the phyla Actinobacteria and Firmicutes. In their metagenomes, rpoB, encoding thesubunit of RNA polymerase, and the 16S rRNA gene revealed the presence of strains closely related to Microlunatus phosphovorus NM-1 of the phylum Actinobacteria (86% of rpoB similarity) and Bacteroides thetaiotaomicron VPI-5482 of the phylum Bacteroidetes (97% of 16S rRNA gene similarity). Unfortunately, we were unable to compare the Firmicutes metagenome with isolated genomes because of low read numbers and lack of the 16S rRNA or rpoB gene in this metagenome (Figure 3(d)).  Figure 4: Gene clusters containing reductive dehalogenases in the Dehalococcoides metagenome. The predicted encoded proteins for all genes have been described previously [1] and are as follows: rdhA: catalytic subunit of Rdh, rdhB: membrane anchor protein of Rdh, rdhC: protein that has a putative function in regulation or electron transport, rdhS and rdhP: sensor histidine kinase and response regulator of the two-component regulatory system, respectively, rdhSP: hybrid rdhS and rdhP, rdhR: multiple resistance regulator (MarR) regulator, rdhF: corrinoid-synthetizing protein, rdhG: Rdh-modifying proteolytic protein, rdhH: hypothetical protein of unknown function, and rdhI: corrinoid-modifying protein.
The Bacteroidetes/Actinobacteria metagenomes had smaller genome sizes and lower total gene numbers than the phylogenetically closest strains (Figures 3(b) and 3(c)). This suggests insufficient sequencing depth to complete the genomes of all members in YN3. In contrast to the overall trend, the Actinobacteria metagenome had 7-19 more genes in the functional categories "motility and chemotaxis" and "miscellaneous," and the Bacteroidetes metagenome had up to 10 genes more in the categories "dormancy and sporulation," "secondary metabolism," "protein metabolism," "motility and chemotaxis," and "miscellaneous." Actinobacteria/Bacteroidetes members with those particular functions are potentially selectively cultivated under dehalorespiration by Dehalococcoides.

Reductive Dehalogenase and Associated Genes.
Eighteen pairs of rdhA and rdhB, encoding the catalytic unit and membrane anchor protein of Rdh, respectively, were detected in the Dehalococcoides metagenome ( Figure 4)  twin-arginine translocation (TAT) leader sequence at the N-terminus and two iron-sulfur-binding motifs near the C-terminus [13], with only YN3RdhA10 lacking the TAT sequence. Most rdhAB pairs in the YN3 metagenome were detected with genes of transcription regulators (rdhS/P and rdhR), suggesting their involvement specific to each rdhA.
Other associated genes involved in assembling and maturation of RdhA (rdhF, rdhG, rdhH, and rdhI) were a minority in the YN3 metagenome, as reported for other isolates of Dehalococcoides spp. Surprisingly, rdhA encoded in gene clusters lacking those accessory genes is expressed as an active, mature RdhA, as indicated by a previous study [50], which suggests a possible association of these accessory proteins remotely encoded in the genome to multiple RdhAs. Among the 18 RdhAs in YN3, only YN3RdhA8 and YN3RdhA16 showed significant similarity to the previously experimentally proven CE-dechlorinating RdhAs (Table S4, supplementary material). YN3RdhA8 had 94% amino acid identity with PceA of strain 195, which dechlorinates PCE to TCE [17]. In the phylogenetic tree ( Figure 5 and BTF08) and non-PCE-dehalorespiring strains VS [53], UCH007 [20], and IBARAKI [21]. YN3RdhA16 showed 97% amino acid similarity with VcrA of strain VS, an RdhA that can dechlorinate cis-DCE and VC, and TCE with lower activity [15]. In the phylogenetic tree, YN3RdhA16 clustered with VcrA and RdhAs from cis-DCE-to-ETHdehalorespiring strains UCH007 [20] and IBARAKI [21] and PCE-to-ETH-dechlorinating strain BTF08 [5].

Transcription of rdhA in Response to CEs.
To identify the rdhAs involved in the PCE-to-ETH dechlorination in YN3, RT-qPCR targeting mRNA of all 18 rdhAs was conducted. The rdhAs showing a higher than twofold increase in transcription after spiking of CEs as compared to without spiking by single determination were screened as candidate rdhAs for CEs. A first RT-qPCR screening of YN3 grown with cis-DCE (Table S5, supplementary material) provided three candidate CEs-responsive rdhAs, that is, YN3rdhA6, YN3rdhA12, and YN3rdhA16. Unexpectedly, YN3rdhA8, closely related to pceA of a known PCE-Rdh, was not transcribed at a significant level, although the starved YN3 spiked with PCE produced TCE.
Next, the transcription of three rdhAs significantly responsive to CE-spiking and YN3rdhA8 was confirmed in at least triplicate determinations. YN3rdhA6, YN3rdhA12, and YN3rdhA16 were confirmed to significantly respond to some of the CEs (Figure 6), while no significant transcription of YN3rdhA8 was detected. Specifically, YN3rdhA6 mRNA was greatly increased by spiking of VC (75 ± 38 mRNA copies per gene; Figure 6(a)) and slightly increased by spiking with cis-DCE, TCE, and PCE. YN3rdhA12 responded especially to VC spiking (16 ± 8.6 mRNA copies per gene) and showed a slight response to all other CEs (Figure 6(b)). YN3rdhA16, an rdhA closely related to vcrA, showed increased transcription in response to VC (4.0 ± 2.9 mRNA copies per gene), cis-DCE (7.4 ± 2.7 mRNA copies per gene), and TCE (3.4 ± 1.4 mRNA copies per gene) (Figure 6(c)). In this experiment, two rdhAs (YN3rdhA6 and YN3rdhA12) that had been never previously identified as rdhAs for CEs unexpectedly showed much stronger transcription than the two rdhAs corresponding to known CE-RdhAs, that is, YN3rdhA16 (vcrA) and YN3rdhA8 (pceA).
The observed marginal increases in YN3rdhA6 and YN3rdhA12 mRNA after spiking of PCE and no significant increase in YN3rdhA8 (pceA) transcription motivated us to reexamine the transcription by using YN3 grown with PCE. YN3 grown with PCE was starved similar to the above transcription analysis and then spiked with PCE. As a result, an increase in transcription was observed especially for YN3rdhA8 (pceA), with 2.6±1.8 mRNA copies per gene (Table  S5 (supplementary material) and Figure 6(d)).
Overall, the results suggested the involvement of four rdhAs in different steps of PCE-to-ETH dechlorination. YN3rdhA8 (pceA) is seemingly involved in PCE-to-TCE dechlorination, especially under PCE-growing condition. Subsequently, YN3rdhA16 (vcrA) is transcribed to respond to the produced TCE and is involved in the dechlorination of TCE to ETH. Besides these known rdhAs, YN3rdhA6 and YN3rdhA12 were found to be transcribed in the presence of VC and are likely involved in VC dechlorination.

Discussion
In this study, we successfully obtained a Dehalococcoidesdominated PCE-to-ETH-dechlorinating culture, YN3, and revealed that four rdhAs are involved in the PCE-to-ETH dechlorination. Particularly, we identified a novel candidate rdhA, YN3rdhA6, encoding an RdhA that probably dechlorinates VC, on the basis of its higher transcription than that of other rdhAs in response to VC.
To date, only two rdhAs encoding RdhAs that dechlorinate VC to ETH, that is, vcrA and bvcA, have been identified by enzymatic assays [15,16]. vcrA was originally identified via amino acid sequencing of the partially purified VcrA [15], while bvcA was first suggested as a candidate VC RdhA on the basis of its increased transcription in response to VC [22], and BvcA was recently proven to dechlorinate VC by native gel assay [16]. Both vcrA and bvcA have been applied as authentic biomarkers to estimate VC dechlorination activity in bioremediation sites [2]. This study suggests the potential presence of multiple VCreductive dehalogenases in Dehalococcoides and an additional VC-reductive dehalogenase-encoding gene, YN3rdhA6, far related to vcrA and bvcA, with 31% translated amino acid similarity. However, some Dehalococcoides, either in pure form or in consortia, have rdhAs identical or similar (>94% of the translated similarity) to YN3rdhA6 but are unable to dechlorinate VC (Tables S4 and S6, supplementary material). This contradictory finding might indicate that YN3rdhA6 responds to VC at the transcriptional level but is not involved in the dechlorination. Additional experiments, including enzymatic assays, are required to further assess whether YN3rdhA6 encodes a VC-reductive dehalogenase.
The sequences of YN3rdhA6 and its gene cluster are not unique but are highly conserved in strains of Dehalococcoides ( Figure S4A, supplementary material). However, the substantial transcriptional response of YN3rdhA6-related genes to VC has been missed in previous studies because they generally focused on responses of this gene to PCE [24] (Table  S6, supplementary material). The difference in magnitude of the transcriptional response between YN3rdhA6 and other rdhAs can be attributed to a particular transcription regulator, the hybrid protein (rdhSP) of sensor histidine kinase (rdhS) and response regulator (rdhP) of the two-component regulatory system (Figure 4 and Figure S4 (supplementary material)). In the metagenome of Dehalococcoides in YN3, another transcription regulator, multiple resistance regulator (MarR) regulator (rdhR), and a gene set of rdhS and rdhP are frequently observed in multiple Rdh-associated gene clusters and are suggested to function in the transcription regulation of rdhAs. The hybrid protein rdhSP is particularly detected in only two gene clusters, one of which includes YN3rdhA6.
Another rdhA that particularly responded to VC spiking, YN3rdhA12, is a possible candidate rdhA that dechlorinates VC. The gene cluster harboring the gene is conserved in YN3 and Dehalococcoides ( Figure S4B, supplementary material). The transcription of rdhA identical to YN3rdhA12 has been studied in strain CG5 [24] and a PCE-dechlorinating enrichment culture TUT2264 [54] (Table S6, supplementary material). The YN3rdhA12-related rdhA of CG5 showed lower transcription than other rdhAs in culture, and consequently, this gene has been never focused on as an rdhA for CEs in this strain [24]. On the other hand, three YN3rdhA12-related rdhAs of TUT2264 showed remarkable (>20-fold) increases in transcription in response to spiking of CEs [54]. However, the specific CE affecting transcription was different among YN3rdhA12 and the three rdhAs in TUT2264. Specifically, like YN3rdhA12, one of the YN3rdhA12-related rdhAs of TUT2264 was increasingly transcribed by addition of VC, while the two other genes were particularly transcribed in the presence of other CEs, but not VC. These results indicated that YN3rdhA12 and related rdhAs can be suggested as rdhAs for CEs, although further study is required to identify specifically which CEs can be dechlorinated by the rdhAs.
The PCE-to-ETH dechlorination rate in YN3 seemed to be comparatively higher than those previously reported in other PCE-to-ETH-dechlorinating enrichment cultures. The KB-1/PCE culture was reported to dechlorinate 100-300 M of spiked PCE to ETH within two weeks with the involvement of Dehalococcoides and Geobacter as dechlorinators [4]; BTF08, a highly enriched culture containing Dehalococcoides as a single dehalorespirator, dechlorinated approximately 500 M PCE to ETH in approximately 100 days [5]; AMEC-4P culture dechlorinated 2 mM PCE to ETH within 143 days, also with Dehalococcoides and Geobacter as dechlorinators [6]. Although the dechlorination rates of those microbial communities are variable depending on the culture conditions, YN3 can be suggested as a promising consortium for application in the bioaugmentation of environments contaminated with CEs.
In contrast to the relatively higher dechlorinating activity in YN3, metagenome analysis of YN3 indicated no apparent uniqueness in the genome of Dehalococcoides in this culture. Specifically, Dehalococcoides in YN3 showed 16S rRNA gene sequences identical to those of strains CG5, CBDB1, IBARAKI, and DCMB5 ( Figure S2, supplementary material) in the Pinellas subgroup of Dehalococcoides. Moreover, all the predicted RdhAs detected in YN3 showed 98-100% similarity with some of the RdhAs in those strains (Table  S4, supplementary material). Therefore, we suggest that the high dechlorination activity in YN3 is attributable to other mechanisms.
A plausible explanation for the high dechlorination activity of YN3 is the constitution of its microbial community. The metagenomic analysis indicated the coexistence of members of the phyla Bacteroidetes, Firmicutes, and Actinobacteria. Particularly, a 16S rRNA gene-based phylotype of Bacteroides might be a new species of Bacteroides, on the basis of the 16S rRNA gene similarity (97%) with known strains. Additionally, members of Bacteroides have been frequently detected in dehalorespiring cultures together with Dehalococcoides [11,55,56]. To date, coexisting bacteria belonging to diverse bacterial phyla have been reported to support Dehalococcoides by supplementation of growth factors [8,9], removal of toxic substances [12], and scavenging the harmful oxygen [57]. Considering the variety in function and phylogeny of non-dechlorinating bacteria that support dechlorination, the non-dechlorinating bacteria in YN3 also possibly contribute to enhancing the dechlorination activity of YN3. Additional experiments will be necessary to prove their contribution to dechlorination by YN3.

Conclusion
A Dehalococcoides-dominated enrichment culture, YN3, was newly established from a CE-contaminated river sediment and efficiently dechlorinated 800 M PCE to ETH within 14 days. The high dechlorination rate was attributed to the presence of non-dehalorespiring bacteria of phyla Actinobacteria and Firmicutes and a new species of the phylum Bacteroidetes. Transcription analysis suggested the involvement of four rdhAs in the PCE-to-ETH dechlorination in YN3. Among these four rdhAs, two were closely related to well-known rdhAs involved in the dechlorination of CEs, pceA, and vcrA. The other two rdhAs, YN3rdhA6 and YN3rdhA12, can be novel candidate rdhAs encoding VCreductive dehalogenases, on the basis of their significant increase in transcription in response to VC.

Conflicts of Interest
The authors declare no conflicts of interest.