Novel Firmicutes Group Implicated in the Dechlorination of Two Chlorinated Xanthones, Analogues of Natural Organochlorines

Abstract

Although the abundance and diversity of natural organochlorines are well established, much is still unknown about the degradation of these compounds. Triplicate microcosms were used to determine whether, and which, bacterial communities could dechlorinate two chlorinated xanthones (2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone), analogues of a diverse class of natural organochlorines. According to quantitative-PCR (qPCR) results, several known dechlorinating genera were either not present or not enriched during dechlorination of the xanthones. Denaturing gradient gel electrophoresis, however, indicated that several Firmicutes were enriched in the dechlorinating cultures compared to triplicate controls amended with nonchlorinated xanthones. One such group, herein referred to as the Gopher group, was further studied with a novel qPCR method that confirmed enrichment of Gopher group 16S rRNA genes in the dechlorinating cultures. The enrichment of the Gopher group was again tested with two new sets of triplicate microcosms. Enrichment was observed during chlorinated xanthone dechlorination in one set of these triplicate microcosms. In the other set, two microcosms showed clear enrichment while a third did not. The Gopher group is a previously unidentified group of Firmicutes, distinct from but related to the Dehalobacter and Desulfitobacterium genera; this group also contains clones from at least four unique cultures capable of dechlorinating anthropogenic organochlorines that have been previously described in the literature. This study suggests that natural chlorinated xanthones may be effective biostimulants to enhance the remediation of pollutants and highlights the idea that novel genera of dechlorinators likely exist and may be active in bioremediation and the natural cycling of chlorine.

INTRODUCTION

The biogeochemical importance of naturally occurring chlorinated compounds (commonly referred to as organochlorines or organochlorides) has been garnering attention, as these compounds are important components of the chlorine and carbon cycles (1, 2, 3, 4, 5). More than 4,400 natural organohalogens have been identified (6), and the production of these compounds is becoming well understood (1, 2, 7, 8, 9). Studies on the degradation of naturally occurring organochlorines, however, are generally lacking. Xanthones are a group of naturally occurring compounds that may be particularly abundant in some environments (10, 11, 12, 13). Xanthones are synthesized by plants, bacteria, and fungi (10, 13) and can contain a range of substitutions, including hydroxyl, methyl, and chloro groups (10, 11, 12). A total of 38 xanthones have been identified that contain between 1 and 4 chlorines (11, 12). Interestingly, the basic chemical structure of chlorinated xanthones resembles that of the anthropogenic chlorinated pollutants polychlorinated biphenyls (PCBs) and dibenzo-p-dioxins (dioxins) (Fig. 1).

FIG 1.

The structures of 2,7-dichloroxanthone (A) and 5,7-dichloro-1,3-dihydroxylxanthone (B) compared to the structures of polychlorinated biphenyls (C) and dibenzo-p-dioxins (D).

Several bacteria that reductively dechlorinate organochlorines for energy generation, referred to as organohalide respirers, have been isolated. Research to date has largely focused on organohalide respirers that dechlorinate anthropogenic pollutants and, in particular, on Dehalococcoides mccartyi, largely because of its ability to dechlorinate a large number of pollutants (14, 15, 16, 17). Recently, several organohalide respirers have been detected in uncontaminated soils and sediments (18, 19, 20), and Dehalococcoides-like bacteria have been reported to grow in response to the amendment of enzymatically produced organochlorines (18). Nevertheless, there remains much about organohalide respiration that is unknown, particularly with respect to organisms whose niche lies in uncontaminated environments. Furthermore, despite the clear importance of Dehalococcoides mccartyi, it is possible that the sometimes narrow focus on this genus (i.e., 18) has drawn attention away from other organohalide respirers.

By investigating organohalide respiration in uncontaminated environments, we may better understand chlorine, and perhaps even carbon, cycling. We may also be able to exploit this process for the bioremediation of anthropogenic pollutants. Specifically, because chlorinated xanthones are structurally similar to pollutants such as PCBs and dioxins, these natural compounds may be useful as alternative electron acceptors, or “primers,” at contaminated sites, speeding the remediation of contaminated sediments (21, 22). With this in mind, we investigated whether two analogues of naturally occurring chlorinated xanthones, 2,7-dichloroxanthone and 5,7-chloro-1,3-dihydroxylxanthone, could be reductively dechlorinated in soils and sediments and which organisms (Dehalococcoides-like or other bacteria) were involved in this process.

MATERIALS AND METHODS

Microcosms.

Two chlorinated xanthones, 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone, were purchased from Specs (Delft, The Netherlands) and Princeton Biomolecular Research, Inc. (Princeton, NJ), respectively. Xanthone (unsubstituted) was purchased from Sigma-Aldrich, and 1,3-dihydroxylxanthone was purchased from Princeton Biomolecular Research, Inc. Soils used in experiments were from a pine-dominated forest in the Pine Barrens in New Jersey (Pine Barrens I), a maple-dominated forest in the Pine Barrens in New Jersey (Pine Barrens II), and Tilden Regional Park in California (Tilden). Sediments used in experiments came from Leech Lake in Minnesota (Leech Lake) and Palos Verdes Harbor in California (Palos Verdes). The soils and Leech Lake sediment are not known to be contaminated with anthropogenic chlorinated compounds; the Palos Verdes Harbor sediment has been historically contaminated with dioxins. These soils and sediments were collected for previous studies (18, 19, 23) and stored at room temperature in an anaerobic chamber (Coy) with a 3% H₂–97% N₂ headspace.

Serum bottles (160 ml) capped with Teflon-lined septa and aluminum crimps were used for microcosms. Microcosms contained 130 ml of mineral media (24) reduced with 2 μM titanium citrate, 10 mM potassium acetate, 1 ml of vitamin solution (25) and 5 g of wet soil or sediment. Microcosm headspace contained 3% H₂–97% N₂. Triplicate microcosms were amended with chlorinated xanthones, and triplicate control microcosms were amended with equivalent molar amounts of nonchlorinated analogues (xanthone and/or 1,3-dihydroxylxanthone). Triplicate autoclaved control microcosms were also set up as described for the active microcosms, seeded with 5 g of a wet mixture of forest soil collected for a previous study (18), including the Pine Barrens I, Pine Barrens II, and Tilden soils, and maintained with an amendment of 50 mM sodium azide. The Tilden and Palos Verdes soils and autoclaved controls were maintained for 698 days; the Pine Barren I microcosms were maintained for 557 days, and the Leech Lake and Pine Barrens II microcosms were maintained for 118 days. The Tilden and Palos Verdes microcosms were initially inoculated with 10 μM 2,7-dichloroxanthone and received multiple amendments of both 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone thereafter, while the Pine Barrens I microcosms were initially amended with 3 μM 2,7-dichloroxanthone and then received multiple amendments of only 2,7-dichloroxanthone thereafter. The Pine Barrens II and Leech Lake microcosms were inoculated with 100 μM concentrations each of 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone and did not receive additional amendments. The chemicals were added to the microcosms as an acetone solution (100 mM) or as dry powder. After amendment with dry powder, the microcosms were placed on a shaker table for 10 days to better incorporate the xanthones into the soil matrix. Otherwise, the microcosms were stored statically in an anaerobic glove bag at room temperature. Microcosms were sampled as previously described (23) for microbiological and/or chemical analysis.

Analysis of xanthones.

Silanized glassware was used for analyzing xanthones. Mixed slurry samples (1.5 ml) were taken for chemical analysis from the chlorinated-xanthone-amended microcosms and the autoclaved controls. Samples were mixed with 10 g oven-baked (550°C) Ottawa sand (Sigma-Aldrich) and air dried for 48 h. The samples were then extracted with an ASE 350 accelerated solvent extractor (Dionex) under the EPA guidelines for accelerated solvent extraction of dioxins (EPA method 3545A). The samples were extracted with a 50:50 mixture of acetone and hexane in three 15-min static intervals at 100°C and 1,600 lb/in². The solvent extract was then reduced with roto-evaporation and transferred to a vial, where the remaining extract was blown down to dryness under a stream of nitrogen. Anhydrous sodium sulfate was added to the vial, and 10 ml of hexane was added to resuspend the extract. Exact amounts of slurry and hexane were determined gravimetrically. The chlorinated xanthones and the nonchlorinated products xanthone and 1,3-hydroxylxanthone were separated via gas chromatography (GC) on an HP-5 column (Agilent). The temperature program used for compound separation was 40°C for 2 min, a ramp of 10°C/min to 180°C, a ramp of 15°C/min to 270°C, and 15 min at 270°C for a total run time of 37 min. Substituted and unsubstituted xanthones were quantified with an electron capture detector (GC-ECD). External standards were prepared for each compound and used for quantification; standard concentrations ranged from 0.1 to 10 ng/mg for 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone (in hexane) and from 1 to 10 ng/mg for xanthone and 1,3-dihydroxylxanthone (in hexane), respectively. The detection limits were 0.1 μM, 0.1 μM, 1 μM, and 1 μM for 2,7-dichloroxanthone, 5,7-dichloro-1,3-dihydroxylxanthone, xanthone, and 1,3-dihydroxylxanthone, respectively. Standards and samples were also analyzed via GC coupled with mass spectrometry (GC-MS) to verify elution times (with standards) and to screen samples for the expected singly chlorinated products (M⁺ peaks of 230 and 232 for 2-chloroxanthone and 262 and 264 for 7-chloro-1,3-dihydroxylxanthone and 5-chloro-1,3-dihydroxylxanthone).

Recoveries of 2,7-dichloroxanthone, 5,7-dichloro-1,3-dihydroxylxanthone, xanthone, and 1,3-dihydroxylxanthone were each individually tested. Five vials (21 ml) containing 15 ml of slurry identical in composition to the microcosms described above were autoclaved. Different amounts of chemical were added to each of the five vials covering the range of concentrations expected in this study (1 μM, 5 μM, 10 μM, 20 μM, and 100 μM), and vials were placed on a shaker table for 3 days. Vials were then sampled, extracted, and analyzed as described above. The recoveries of 2,7-dichloroxanthone, 5,7-dichloro-1,3-dihydroxylxanthone, xanthone, and 1,3-dihydroxylxanthone were found to be 92 ± 1.3%, 89 ± 6.3%, 95 ± 3.1%, and 96 ± 2.4%, respectively (means ± standard errors for the five concentrations).

DNA extraction, DGGE analysis, cloning, and sequencing.

DNA was extracted from microcosm slurry samples using a PowerSoil DNA kit (Mo Bio Laboratories). Denaturing gradient gel electrophoresis (DGGE) was performed similarly to what was described previously (23, 26). Briefly, genomic DNA was diluted 100-fold and amplified with primers 338F (5′-ACT CCT ACG GGA GGC AGC AG-3′) (27) and 518R (5′-ATT ACC GCG GCT GCT GG-3′) attached to a GC clamp (28). PCR amplicons were analyzed on a 30-to-55% denaturing gradient gel, and electrophoresis was performed on a D-Code apparatus (Bio-Rad) in 0.5× Tris-acetate-EDTA (TAE) buffer. Electrophoresis was performed at 20 V for 15 min, followed by 200 V until the loading dye migrated to the bottom of the gel (at least 3 h). Gels were stained with SYBR green I (Molecular Probes) and visualized on a UV transilluminator (either an EC3 Bioimaging system by Ultra-Violet Products or a GelDoc system by Bio-Rad).

To determine the full 16S rRNA sequences responsible for the DGGE bands that appeared in the chlorinated-xanthone-amended microcosms and not in the controls, a clone library was generated followed by an enzyme digestion to identify unique clones. Unique clones were analyzed by DGGE to identify the position of the band relative to the banding pattern of the original DNA extract. Clone libraries were developed from the DNA extracts from one of the triplicate reactors from the following chlorinated-xanthone-amended microcosms: Tilden (day 635), Palos Verdes (day 635), Pine Barrens II (day 118), and Leech Lake (day 118). Nearly complete 16S rRNA genes were amplified using the bacterial primers 27F (27) and 1522R (29) as previously described (30). The amplicons were then cloned into electrocompetent Escherichia coli DH5α using the pGEM-T Easy cloning kit (Promega). An enzyme digestion was performed with 2 U of the restrictive enzyme AluI (New England BioLabs) according to the manufacturer's recommendations. The digests were run on 1.5% agarose gels stained with GelRed (Biotium) and imaged on a UV illuminator with QuantityOne Software (Bio-Rad). One or more representative clones from each digestion fingerprint were chosen for DGGE analysis. Clones and the DNA extract from which the clones were generated were analyzed by DGGE in parallel to identify the DGGE position of the cloned 16S rRNA gene. The clones with bands that corresponded to the position of the targeted bands (those unique to microcosms amended with chlorinated xanthones) were sequenced at the University of Minnesota Biomedical Genomics Center (BMGC) or the University of Hawaii at Manoa College of Natural Sciences Advanced Studies of Genomics, Proteomics, and Bioinformatics (ASGPB).

Bands targeted for analysis and selected bands from the control DGGE lanes, often those that were positioned near the targeted bands, were directly excised and sequenced as well (see the supplemental material for further details). Excised bands were moved to microcentrifuge tubes with ethanol-washed tweezers, and 20 μl of DNase/RNase-free water was added. Gel slices were then left overnight, and the supernatant was diluted 1:100 and subjected to repeated rounds of PCR, DGGE, and excision until the fragment appeared clear of other fragments. The final excised fragments were then subjected to PCR (for DGGE) as described above, but without the GC clamp. PCR fragments were cleaned with the GeneClean II kit (MP Biomedical) and subjected to sequencing as described above. Sequences of the excised bands from the chlorinated-xanthone-amended microcosms closely matched (>90% similarity) the sequence of their corresponding clones (see the supplemental material for sequence information). Sequences of the excised bands from the controls were not closely (<90%) related to the sequences corresponding to the 12 bands of enriched bacteria, with the exception of one band excised from an autoclaved control that was 94% similar in sequence to the targeted band Pine Barrens II-2 (see the supplemental material).

qPCR/PCR of 16S rRNA genes.

Previously published PCR and quantitative-PCR (qPCR) methods were used to target the 16S rRNA genes from the reductively dechlorinating Dehalococcoides-like Chloroflexi (18), Dehalogenimonas spp. (31), Desulfitobacterium spp. (32), Dehalobacter spp. (32), Desulfomonile spp. (33), and Anaeromyxobacter spp. (34) (see the supplemental material).

A novel qPCR amplification technique was developed to target a group within the Firmicutes, herein referred to as the Gopher group. The full 16S rRNA sequences from clones retrieved from NCBI's BLAST database and this study that belong to the group were used as a target for primer specificity. These sequences were aligned with the 16S rRNA sequences from several Firmicutes, including the closely related Desulfitobacterium spp., Dehalobacter spp., and Desulfosporosinus spp. as well as representatives from more distant genera and the other 16S rRNA sequences from this study. The two qPCR primers Gfr163F (5′-TGA CCY TGG CAT CAG GGA-3′) and Gfr441R (5′-TAT TTT ACA ACC CGA AGG CCT TCG-3′) are specific for all Gopher group sequences (see Fig. 4) and are nonspecific for all other sequences shown in these figures. A BLAST search for these primers indicates no matches for sequences clearly belonging to other identified genera. The primers were used to form a clone library from the DNA extract of Tilden-A (day 635), and one clone was selected for standards and prepared by an overnight culture, plasmid extraction, and plasmid quantification and dilution as previously described (18). The sequence of the plasmid insert is >95% similar to all Gopher group sequences and is provided in the supplemental material. Thermocycling conditions had an initial denaturation of 2.5 min at 95°C followed by 40 cycles of 15 s at 95°C and 30 s at 52°C. The amplification mixture consisted of 1× SYBR green MasterMix (Bio-Rad Laboratories), 100 nM each primer, and 1 mg/liter of BSA. Melting curves were run at the end of each experiment to confirm specificity of amplification and lack of primer/dimer formation.

FIG 4.

Phylogenetic tree of the nearly full-length 16S rRNA genes identified from the clone libraries and DGGE analysis. Sequences from this study are in bold, and the clone's corresponding DGGE positions from Fig. 3 are in parentheses. Numbers at nodes indicate percent agreement from bootstrap analysis (500 bootstraps) when over 50%. Bacillus subtilis was used as an outgroup to root the tree.

Phylogenetic analysis.

Phylogenetic analysis was performed on nearly full-length 16S rRNA sequences with Mega5 (35) with an evolutionary history inferred using the neighbor-joining method (36) with a bootstrap test (500 bootstraps) (37). Evolutionary distances were computed with the maximum composite likelihood method (38) and, with the scale bar as a reference, correspond to the number of base pair substitutions per site. Phylogenetic analyses included the closest matches from BLAST searches, using only those 16S rRNA sequences that were nearly full length, as well as the full 16S rRNA sequences from selected Firmicutes isolates (all enriched bands corresponded to bacteria within the Clostridiales order of the Firmicutes).

PCR of reductive dehalogenase genes.

The amplification of reductive dehalogenase genes was attempted using previously published primers and methods (39, 40, 41, 42, 43). These attempts were unsuccessful at amplifying putative reductive dehalogenase genes (see the supplemental material).

Nucleotide sequence accession numbers.

The sequences from this study are deposited in NCBI's GenBank database under accession numbers KF275148 to KF275169.

RESULTS

Dechlorination of chlorinated xanthones.

Both 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone were dechlorinated in the Tilden- and Palos Verdes-inoculated microcosms, with the concomitant production of the nonchlorinated daughter products xanthone and 1,3-dihydroxylxanthone (Fig. 2). There was an initial lag prior to 2,7-dichloroxanthone dechlorination and prior to 5,7-dichloro-1,3-dihydroxylxanthone dechlorination in the Palos Verdes microcosms (Fig. 2). There was no lag prior to 5,7-dichloro-1,3-dihydroxylxanthone dechlorination in the Tilden microcosms. Initial dechlorination rates varied by compound and soil/sediment, with relatively slow initial dechlorination (Fig. 2). Subsequent loss of the dechlorinated products was also observed, suggesting that these compounds were also metabolized over time. No singly chlorinated xanthones were detected in this experiment by GC-MS, suggesting that either the detection limit was too high or the dechlorination of the singly chlorinated intermediates occurred faster than, or as fast as, that of the parent compounds. No significant decrease in chlorinated xanthones occurred in the control microcosms (maintained for 698 days) or the Pine Barrens I microcosms (maintained for 557 days) (see Fig. S1 in the supplemental material).

FIG 2.

Concentrations of 2,7-dichloroxanthone (solid diamonds), 5,7-dichloro-1,3-dihydoxyxanthone (solid squares), xanthone (open diamonds), and 1,3-dihydroxylxanthone (open squares) over time for the Tilden microcosms (A) and the Palos Verdes microcosms (B). Errors bars represent standard errors for triplicate microcosms. Arrows point to times at which chlorinated xanthones were reamended.

The abundance of multiple genera of organohalide-respiring bacteria (Dehalococcoides-like bacteria, Desulfomonile, Dehalogenimonas, Desulfitobacterium, Dehalobacter, and Anaeromyxobacter) was measured with qPCR to determine if these populations increased in abundance during chloroxanthone dechlorination. Dehalococcoides-like bacteria were detected at similar levels in the control (fed only xanthone and 1,3-dihydroxylxanthone) and chlorinated-xanthone-amended microcosms during the last 100 days of microcosm operation (see Fig. S2 in the supplemental material). Dehalobacter, Dehalogenimonas, Desulfomonile, and Anaeromyxobacter spp. were not detected in any of the microcosms, and the qPCR method used for Desulfitobacterium amplified nontargeted sequences only (see the supplemental material). Therefore, none of these organohalide-respiring bacteria were likely to be the primary organisms involved in chlorinated xanthone dechlorination.

To determine which organisms might be responsible for the dechlorination of 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone, DGGE was performed on samples from the Tilden and Palos Verdes microcosms (Fig. 3). A total of 7 bands were identified that were both (i) present in the gels run with samples from the chlorinated-xanthone-amended microcosms and (ii) not visible in the gels run with samples from the control microcosms fed only xanthone and 1,3-dihydroxylxanthone. Clone libraries enabled identification of these 7 unique bands via nearly full-length 16S rRNA gene sequences. Of these, the sequences corresponding to bands T-3 and PV-2 were the most interesting because they were the only sequences common to both sets of microcosms and they were the most closely related to the Desulfitobacterium and Dehalobacter spp. (∼91% identical and ∼89% identical, respectively, along ∼1,500 bp of the 16S rRNA genes). Additionally, as determined from BLAST analysis, these sequences were closely related to several clones from other dechlorinating microcosms (44, 45, 46, 47) yet not related to clones found in association with purely fermentative cultures. We named this group the Gopher group (Fig. 4) and developed a qPCR method to target the 16S rRNA genes of this group (see Materials and Methods). Bands in the control microcosms were also excised and sequenced for quality control (see Fig. S4 and Table S2 in the supplemental material), with sequenced bands from the controls having less than 90% similarity to the sequences targeted from the chlorinated-xanthone-amended microcosms.

FIG 3.

DGGE fingerprinting of bacterial communities from Tilden microcosms at day 635 (A), Palos Verdes microcosms at day 635 (B), Pine Barrens II microcosms at day 118 (C), and Leech Lake microcosms at day 118 (D). Triplicate (lanes a, b, and c) microcosms are shown for sets of microcosms amended with chlorinated xanthones (ClX) and nonchlorinated xanthones (control). Numbered bands point to those targeted for phylogenetic analysis by clone library and DGGE analysis. Selected bands from control lanes were excised and sequenced (see Fig. S4 and Table S2 in the supplemental material).

In the final 100 days of Tilden and Palos Verdes microcosm operation, the ratio of Gopher group 16S rRNA genes normalized to the number of total Bacteria 16S rRNA genes (Gopher/Bacteria) was maintained at 9.5% ± 5% and 5.0% ± 3% (mean ± standard deviation of triplicate microcosms), respectively, compared to 0.12 ± 0.08% and 0.9 ± 0.04% in the Tilden and Palos Verdes control microcosms, respectively. The overall abundance of Gopher 16S rRNA genes was also much higher in the chlorinated-xanthone amended reactors than the controls (8.6 ± 0.5 and 8.1 ± 0.2 logarithmic units 16S rRNA genes/g in the Tilden and Palos Verdes chlorinated-xanthone-amended microcosms, respectively, compared to 6.1 ± 0.4 and 7.4 ± 0.3 logarithmic units 16S rRNA genes/g for Tilden and Palos Verdes control microcosms, respectively).

Verification of Gopher group involvement.

The enrichment of Gopher 16S rRNA genes was then tested with two new inoculum sources as described in Materials and Methods: sediment from Leech Lake in Minnesota and additional soil from the New Jersey Pine Barrens (called the Pine Barrens II microcosm). These new microcosms were amended with chlorinated xanthones, which were subsequently dechlorinated over a period of approximately 118 days (Fig. 5A and B).

FIG 5.

Degradation of chlorinated xanthones for microcosms inoculated with Leech Lake sediment (A) and Pine Barrens II soil (B) and the amount of Gopher group 16S rRNA genes per g in microcosms inoculated with Leech Lake sediment (C) and Pine Barrens II soil (D). (A and B) Xanthone data are shown for the reactors amended with chlorinated xanthones. ●, 2,7-dichloroxanthone; ■, 5,7-dichloro-1,3-dihydroxylxanthone; ○, unsubstituted xanthone; □, 1,3-dihydroxylxanthone. (C and D) Dark bars, chlorinated-xanthone-amended microcosms; light bars, nonchlorinated-xanthone-amended controls. Error bars represent standard deviations for triplicate reactors, except for the Gopher group data from the chlorinated-xanthone-amended microcosms (D), in which case the error bars represent standard deviations for duplicate reactors.

Figure 5C shows the enrichment of Gopher group organisms that occurred in the Leech Lake microcosms. Indeed, the Gopher/Bacteria ratio increased from 0.023% ± 0.02% to 14% ± 10%, with the abundance of Gopher 16S rRNA genes increasing from 4.8 ± 0.6 to 8.4 ± 0.1 logarithmic units 16S rRNA genes/g during dechlorination in the Leech Lake microcosms. The proportion and abundance of these organisms remained essentially constant at ≤0.04 ± 0.03% or ≤5.8 ± 0.3 logarithmic units 16S rRNA genes/g in the nonchlorinated-xanthone-amended controls (Fig. 5C).

In the Pine Barrens II microcosms, the Gopher group organisms were enriched over time when amended with chlorinated xanthones in two of the three replicate microcosms, growing from 5.0 ± 0.5 to 8.0 ± 0.2 logarithmic units 16S rRNA genes/g (Fig. 5D). In the third Pine Barrens II microcosm, the Gopher group organisms failed to grow (see Fig. S3 in the supplemental material); the reason for this lack of growth is unknown. Growth also occurred, though to a lesser extent (1.3 ± 0.8 logarithmic units [16S rRNA genes/g]), in the triplicate Pine Barrens II control microcosms (Fig. 5D).

The Leech Lake and Pine Barrens II microcosms were also subjected to DGGE to determine if unique organisms were enriched in addition to the Gopher group organisms (Fig. 3). The phylogenetic analysis of the full 16S rRNA genes (obtained from clone libraries) of bands uniquely enriched in the chlorinated-xanthone-amended microcosms is shown in Fig. 4. Other Firmicutes were identified by DGGE that appeared to be enriched in the chlorinated-xanthone-amended Leech Lake and Pine Barrens II microcosms (Fig. 4). Again, bands in the control microcosms were also excised and sequenced for quality control (see Fig. S4 and Table S2 in the supplemental material), and all but one of the sequenced bands from the controls were found to be less than 90% similar to the sequences targeted from the chlorinated-xanthone-amended microcosms. One band, Ctrl-PB3, was 94.2% similar to the sequences corresponding to PB-2 (Fig. 3). This suggests that the 16S rRNA gene associated with PB-2 is not likely to be unique to the microcosms amended with chlorinated xanthones.

DISCUSSION

Phylogenetically, the Gopher group bacteria form a newly described group of putative organohalide respirers, distinct from but related to the Dehalobacter and Desulfitobacterium in the Firmicutes (Fig. 4). Members of the Gopher group could be found in literature reports of at least four other cultures active in the dechlorination of chlorinated ethene (KB1 culture) (45), chlorobenzenes (SJA culture) (44), PCBs (Er-MLAYS and Er-LLAYS cultures) (46), and pentachlorophenol (PCP culture) (47). As mentioned above, the Gopher group also lacks representatives from more generic fermentative studies, and therefore, their detection and presence thus far appear to be solely dependent on active dechlorination. Our research suggests that previously overlooked organisms belonging to the Gopher group may have participated in organohalide respiration in previously investigated cultures (44, 45, 46, 47). These Gopher group organisms should be the focus of further directed research to clarify their niche in both contaminated and uncontaminated environments.

Additional strains of Firmicutes were also enriched during the dechlorination of 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone (Fig. 3 and 4). Their role, if any, in the dechlorination of these two compounds is unknown. Two of the targeted DGGE bands do not appear to be associated with any previously reported dechlorinating cultures (Tilden-2 and Palos Verdes-3). Eight targeted bands were also closely related to bacteria from both dechlorinating and nondechlorinating cultures (Tilden-4, Palos Verdes-1, Pine Barrens II-2, Pine Barrens II-4, the sequences closely related to Pine Barrens II-3 and Tilden-1, Pine Barrens II-1, and Leech Lake-1). Interestingly, sequences corresponding to band Pine Barrens II-1 are closely associated with the genera Pelosinus; two Pelosinus species were recently isolated from chlorinated-solvent-contaminated aquifers but, upon isolation, were not capable of dechlorinating selected solvents (48). A third Pelosinus isolate from a dechlorinating culture (strain BXM) was deposited in GenBank but has not yet been described, and a Pelosinus isolate was identified to be a dominant strain in a trichloroethene-dechlorinating enrichment culture containing Dehalococcoides (49). These isolates may be involved in reductive dechlorination or may be nondechlorinating bacteria associated with organohalide respirers. Again, further work is needed to clarify what role, if any, these non-Gopher group Firmicutes strains play in the dechlorination process.

Relevance to bioremediation and application.

This research presents two interesting findings with relevance to bioremediation. First, the possibility that natural compounds such as chlorinated xanthones may be useful as biostimulants or primers to speed contaminant dechlorination exists and is intriguing. Priming with chlorinated or brominated compounds has previously been demonstrated to speed the bioremediation of chlorinated pollutants, even highly weathered pollutants (21, 22); nevertheless, currently identified primers are also considered toxic and therefore cannot be used. For chlorinated xanthones, though the dechlorination rates and lag times were relatively long at low concentrations, they were short when higher doses of 100 μM 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone were amended (Fig. 2 and 5). These compounds appear promising as primers; nevertheless, their toxicity is unknown and must first be determined. The enrichment of the Gopher group bacteria on chlorinated xanthones is also particularly interesting, as similar organisms have been found in cultures dechlorinating anthropogenic pollutants. This further suggests that chlorinated xanthones may have the potential to enrich organisms relevant for bioremediation.

Second, this research highlights the possibility that there may exist vast and still largely undiscovered biodiversity with respect to organohalide-respiring organisms. Indeed, recently a Shewanella sediminis strain was found to contain five putative reductive dehalogenases, one of which was confirmed to be functional in the dechlorination of tetrachloroethene, despite the fact that the genus has never been associated with organohalide respiration (50). Isolates of the genera Dehalococcoides, Dehalogenimonas, “Dehalobium,” Dehalobacter, Desulfitobacterium, Anaeromyxobacter, Geobacter, Desulfomonile, Desulfuromonas, Desulfovibrio, Sulfurospirillum, Propionibacterium, Clostridium, and Desulfoluna have been linked to the specific dechlorination activities of several compounds (51, 52, 53, 54), yet dechlorination unattributable to any of these genera has also been frequently observed (23, 30, 46, 55). Our results implicate a group within the Firmicutes, the Gopher group, in the dechlorination of 2,7-dichloroxanthone and 5,7-dichloro-1,3-dihydroxylxanthone and point to the need for additional research on the biodiversity of organohalide respirers. Indeed, as research continues, previously unidentified organisms may be found to play a role in organohalide respiration, increasing our knowledge of elemental cycling and improving bioremediation as well.