|Home | About | Journals | Submit | Contact Us | Français|
Quantitative analysis of genes that code for Dehalococcoides 16S rRNA and chloroethene-reductive dehalogenases TceA, VcrA, and BvcA was done on groundwater sampled from 150 monitoring wells spread over 11 chlorinated ethene polluted European locations. Redundancy analysis was used to relate molecular data to geochemical conditions. Dehalococcoides 16S rRNA- and vinyl chloride (VC)-reductase genes were present at all tested locations in concentrations up to 106 gene copies per ml of groundwater. However, differences between and also within locations were observed. Variation in Dehalococcoides 16S rRNA gene copy numbers were most strongly correlated to dissolved organic carbon concentration in groundwater and to conditions appropriate for biodegradation of chlorinated ethenes (U.S. Environmental Protection Agency score). In contrast, vcrA gene copy numbers correlated most significantly to VC and chlorinated ethene concentrations. Interestingly, bvcA and especially tceA were more correlated with oxidizing conditions. In groundwater microcosms, dechlorination of 1 mM VC was correlated to an increase of vcrA and/or bvcA gene copies by 2 to 4 orders of magnitude. Interestingly, in 34% of the monitoring wells and in 40% of the active microcosms, the amount of individual VC-reductase gene copies exceeded that of Dehalococcoides 16S rRNA gene copies. It is concluded that the geographical distribution of the genes was not homogeneous, depending on the geochemical conditions, whereby tceA and bvcA correlated to more oxidized conditions than Dehalococcoides 16S rRNA and vcrA. Because the variation in VC-reductase gene numbers was not directly correlated to variation in Dehalococcoides spp., VC-reductase genes are better monitoring parameters for VC dechlorination capacity than Dehalococcoides spp.
Chlorinated ethenes, such as tetrachloroethene (PCE) and trichloroethene (TCE), are persistent groundwater pollutants (15, 22). Because these compounds are toxic and mobile in groundwater systems, they form a serious risk for human health and the environment. PCE and TCE can be dechlorinated by microorganisms under anaerobic conditions by reductive dehalogenation to dichloroethene (DCE), vinyl chloride (VC), and ethene (20). Bioremediation strategies for chloroethene-contaminated sites are often based on (stimulation of) reductive dechlorination of the chlorinated ethenes to ethene (7, 12, 14). In practice, reductive dechlorination of PCE and TCE can be incomplete, resulting in accumulation of DCE or VC. Since VC is much more mobile, toxic, and carcinogenic than PCE and TCE (9), monitoring and stimulation of VC dechlorination are essential steps in bioremediation strategies.
Only members of Dehalococcoides spp. are known to be able to reductively dechlorinate VC. Therefore, 16S rRNA genes of these species are often used as molecular target to indicate and monitor DCE and VC dechlorination capacity at contaminated sites. However, previous studies showed different dechlorination capacities for individual Dehalococcoides species, and only a few strains are known to metabolically dechlorinate VC (6, 8, 10, 17, 21). As a consequence, 16S rRNA gene-based detection can lead to overestimation of VC dechlorination capacity. In contrast, although metabolic reductive dechlorination of VC has mostly been linked to Dehalococcoides spp., it cannot be excluded that other microbial species that perform this dechlorination exist. Genes coding for DCE and VC reductases may be exchangeable between different microbial species via horizontal gene transfer. This is plausible since it has been shown that the metabolic genes for VC dechlorination, vcrA and bvcA, have a different evolutionary history than most other Dehalococcoides genes (16). Consequently, Dehalococcoides 16S rRNA gene-based detection can also lead to underestimation of VC dechlorination capacity.
To more precisely determine VC dechlorination capacity, genes directly involved in reductive dechlorination of VC should be used as a molecular target, in addition to Dehalococcoides 16S rRNA genes. A quantitative method was described to detect genes coding for VC-reductases VcrA and BvcA identified in Dehalococcoides sp. strains VS and GT and in Dehalococcoides sp. strain BAV1, respectively (10, 17, 21). Different studies showed direct correlation of vcrA and bvcA gene copy numbers with reductive dechlorination of VC in batch cultures, soil columns, and contaminated sites (2, 11, 19).
Quantification of genes that encode VC-reductases can be a useful method to monitor reductive dechlorination of VC in chloroethene-contaminated groundwater during enhanced natural attenuation activities (4, 19). However, little is known about the presence, dispersion, and importance of specific dehalogenase genes in chlorinated ethene polluted groundwater and their correlation to biogeochemical conditions and reductive dechlorination.
The objective of the present study was therefore to identify the relative importance of TCE-reductase gene tceA and VC-reductase genes vcrA and bvcA in chloroethene-polluted groundwater and to identify geochemical parameters that contribute to variation in copy numbers of these genes. To this end, groundwater of 150 monitoring wells from 11 European polluted sites was analyzed. Furthermore, microcosms with groundwater from 6 locations were started to test whether VC dechlorination is directly correlated to an increase of vcrA or bvcA genes.
Eleven sample locations (Table (Table1)1) were selected based on the availability of groundwater for the present study, the presence of chlorinated organic pollutants, the presence of different monitoring wells at the same location, and the availability of geochemical data from the monitoring wells. All locations are situated in the Netherlands except location H, which is in the United Kingdom. Pollution with chlorinated ethenes originated from (former) industrial activities and was present for several decades. At locations A, C, F, and H, other pollutants such as 1,2-dichloroethane, pesticides, and mineral oil were also present. The natural bioremediation process of chlorinated ethenes was enhanced at locations H, I, J, and K by the addition of electron donor into the groundwater. For all locations, monitoring wells within and outside the polluted zone were included in the study. An overview of the geochemical parameters in the monitoring wells is presented in Table S1 in the supplemental material.
Groundwater was collected between 2007 and 2008 from monitoring wells by using a peristaltic pump and Marprene tubing. Before duplicate samples were taken, three water volumes of the monitoring well were discarded to refresh the well water. Sterilized 1-liter glass bottles were filled completely and stored at 4°C. The samples were used within 1 week for DNA extraction and within 1 month for chemical analyses and preparation of microcosms.
Anaerobic microcosms Bmw421, Cmwdb2, Emw403, Emw100, Fmw401, Fmw54, GmwRH05, GmwRH48, Hmw68, and Hmw267 were prepared with groundwater from monitoring wells pb-421 (location B); pb-db2-23,5 (location C); pb-403 and pb-100-61 (location E); pb-401 and pb-54 (location F); pb-RH05-22 and pb-RH48 (location G); and pb-mwsa68 and pb-BH267 (location H), respectively. Monitoring wells with both a high or low VC concentration, a high or low concentration of vinyl chloride reductase genes, and a high, low, or unknown biodegradation potential (U.S. Environmental Protection Agency [EPA] score) were selected to minimize a potential bias of site selection criteria. Samples of 200 ml of groundwater were incubated at 20°C and 150 rpm for 4 months in 220-ml serum bottles, closed with a Viton stopper. Protamylasses, which is a product from the potato industry (1 ml; Avebe, Veendam, the Netherlands), was added as an electron donor. The protamylasses solution contained 180 mg total organic carbon per liter. This corresponded to 75 mM organic carbon added to the bottles. Assuming an average composition as sugar [CH2O]n, this accounts for 300 mM added “electron equivalents.” For each monitoring well, two microcosms were started, one of which received 1.5 mM VC. Headspace analyses (500 μl) for VC concentration and liquid samples (5 ml) for DNA extraction were taken monthly under sterile and anaerobic conditions with a syringe.
Biomass from 50- to 250-ml groundwater samples was concentrated by vacuum filtration in triplicate on 0.2-μm-pore-size filters (ME 24 membrane filters, 25 mm in diameter; Schleicher & Schuell, Den Bosch, the Netherlands). Biomass was concentrated from 5-ml microcosm samples in duplicate by centrifugation (20 min at 4,000 rpm), and the supernatant was filtrated through 0.2-μm-pore-size filters (Schleicher & Schuell). All filters were crushed with a sterilized wooden toothpick. Total DNA was extracted from the concentrated biomass with Bio 101 Fast DNA kit for soil (Qbiogene, Inc.). Other than the addition of one extra one-quarter-inch ceramic sphere (Qbiogene) to Lysing Matrix E of the DNA isolation kit, no modifications to the manufacturer's protocol were applied. The quality and quantity of the extracted DNA was determined by using a Nanodrop ND-1000 spectrophotometer (Isogen, de Meern, the Netherlands) and by electrophoresis on a 1% agarose gel. DNA was stored at −20°C.
Copy numbers of the VC-reductase genes vcrA and bvcA, tceA and the Dehalococcoides 16S rRNA genes were quantified in 25-μl quantitative PCR (qPCR) assays by using IQ-Supermix (Bio-Rad, Veenendaal, the Netherlands) and 3 μl of template DNA isolated from the filters. For all filters, 10- and 100-fold-diluted DNA extracts were analyzed in duplicate. Sterilized MilliQ water and an environmental sample without target genes were used as negative controls. As positive controls, genes of interest cloned into a pGEM-T vector (Promega) were used. The coefficient of variance was below 5% for all qPCR assays. The specific primers and probes used for detection of the vcrA, bvcA, tceA, and Dehalococcoides 16S rRNA genes were described by Ritalahti et al. (18). The temperature program for all PCR assays on an IQ5 real-time PCR system (Bio-Rad) was performed as follows: 3 min at 95°C; followed by 35 cycles of 30 s at 95°C, 30 s at 58°C, and 1 min at 72°C; with a final elongation step of 5 min at 72°C. The number of gene copies was calculated by comparing the threshold cycle (CT) values of the individual samples to the CT values of a calibration curve. Dehalococcoides 16S rRNA, vcrA, bvcA, and tceA gene calibration was done by using 10-fold serial dilutions of 101 to 107 genes/ml of pGEM-T Easy vectors (Promega, Madison, WI) containing a single copy of the gene of interest. The amount of gene copies in the DNA samples was extrapolated to the amount of genes copies per ml of groundwater or microcosm culture. The lower detection limit in groundwater of Dehalococcoides 16S rRNA and reductase genes differed because of different PCR efficiencies and were 10 and 100 gene copies/ml of groundwater, respectively. The detection limits were determined based on the gene concentration present in the most strongly diluted standard of the calibration curve, which gave a positive signal in the qPCR analysis. This concentration was divided by the amount of groundwater used for DNA extraction. Every single qPCR run had its own calibration curve. The detection limits were 100-fold higher for microcosm samples due to a lower sample volume used for DNA extraction. Regularly, the size of the obtained amplicons were checked on a 2% agarose gel. The standard error for quantification the individual genes differed from a 0.52 to a 0.75 order of magnitude for the Dehalococcoides spp. 16S rRNA and VC or TCE reductase genes.
Chlorinated hydrocarbons (PCE, TCE, DCE, and 1,2-dichloroethane) were analyzed in 8-ml diluted groundwater samples on a Varian 3800 gas chromatographic (GC) system equipped with a mass spectrometry (MS) detector and a Porabond-Q column (0.32 mm by 25 m) (Varian, Middelburg, the Netherlands). The GC settings were as follows: injector temperature, 200°C; detector temperature, 300°C; oven temperature, 3 min at 40°C, followed by an increase of 10°C/min to 70°C, followed by an increase of 15°C/min to 250°C for 7 min; and carrier gas helium with a flow rate of 2 ml/min. External standards at six different concentrations from 0 to 30 μM were used for calibration.
VC, ethene, ethane, and methane in groundwater samples were quantified in 1-ml samples on a Varian 3800 GC equipped with a flame ionizing detector (FID) and a Porabond-Q column (0.32 mm by 25 m) (Varian). The GC settings were as follows: injector temperature, 200°C; detector temperature, 300°C; oven temperature, 3 min at 40°C, followed by an increase of 10°C/min to 70°C, followed by an increase of 15°C/min to 250°C for 7 min; and a flow rate of 2 ml/min. For microcosm analyses, 500-μl headspace samples, obtained by using a 1-ml Pressure-Lock gas syringe (Alltech, Breda, the Netherlands) with a sterile needle, were analyzed on the GC-FID described above. External standards at six different concentrations from 0 to 3 mM were used for calibration. The lower detection limit was 1 μM for all analyzed compounds.
Chloride, nitrate, and sulfate ion analyses were done on 1 ml of diluted groundwater and microcosm samples on a Dionex ICS-1500 equipped with an Ionpac AS14 anion-exchange column and an A SRS-Ultra 14-mm suppressor (Dionex Corp., Sunnyvale, CA). The eluent (2.0 mM Na2CO3 and 0.75 mM NaHCO3) flow rate was 1.3 ml/min. The injection needle was preflushed with 100 μl of MilliQ water, and 50-μl samples were injected. External standards at six different concentrations from 0 to 250 μM were used for calibration. Groundwater and microcosm samples were diluted 10-, 100-, and 500-fold before analysis.
Dissolved organic carbon (DOC) content was determined in 0.5 ml of groundwater by flash combustion, followed by CO2 analysis by an IR detector (SC-632; Leco, Mönchengladbach, Germany). The iron content of the samples was determined by using ICP-OES (inductively coupled plasma-optical emission spectrometer) analysis (Spectro CirosCCD (Spectro, Kleve, Germany), after destruction with a mixture of HF, HClO4, and HNO3.
Hydrogen concentration in the groundwater was analyzed onsite using a gas chromatograph with a reduction gas detector (Trace Analytical, Bester, Amstelveen, the Netherlands), which was equipped with a Carbosieve II column. The carrier gas was N2 with a flow rate of 20 ml/min. The column and detector temperatures were 104°C and 265°C, respectively.
Temperature, pH, redox potential, and salinity were measured in groundwater that was pumped through a flow cell (Eijkelkamp B.V., Groesbeek, the Netherlands) onsite, with a pH/mV/EC/Sal/TDS/T/O2 multiparameter (Eijkelkamp) and corresponding probes (Eijkelkamp).
Biodegradation potential of the individual monitoring wells was quantified by the EPA score, based on 9 to 10 critical geochemical parameters, as described by the U.S. EPA in the Protocol for Evaluating Natural Attenuation (23). For the EPA score, positive or negative values are provided for geochemical parameters that are presumed critical for reductive dechlorination of chlorinated ethenes (e.g., an oxygen concentration greater than 0.5 mg/liter is provided with a score of −3 and a sulfate concentration less than 20 mg/liter is provided with a score of 2). The sum of the scores for the individual parameters results in the EPA score. In groundwater with an EPA score greater than 15, reductive dehalogenation of chloroethenes is likely.
Redundancy analysis (RDA) is a multivariate regression analysis, whereby variation in one set of variables is explained by another set of variables. In the present study, gene copy numbers were selected as one set of variables (species data), and geochemical data were selected as the explaining set of variables (environmental data). The RDA was performed with Canoco 4.5 software (Microcomputer Power, Ithaca, NY) on data of all monitoring wells from locations A, B, E, F, and K and monitoring wells pb103, pb104, pb105, and pb1022 from location C, since for these monitoring wells an almost complete data set was available. For the RDA, a linear distribution of the “species” (in this case, gene copy numbers) was assumed to explain their variation by environmental variables. The data were log transformed [Y = log(AY+B), where A = 1 and B = 1] for parameters with more than 10-fold variation within the data. In the analysis, data below the detection limit were substituted by values of 70% of the detection limit. To test the significance of the relationship of gene copy number variance with geochemical parameters, unrestricted Monte Carlo permutation tests were performed with 499 random permutations and a significance level (P) of 0.05.
VC-reductase genes (vcrA and bvcA), TCE reductase gene (tceA), and Dehalococcoides 16S rRNA genes were detected in 150 groundwater samples obtained from 11 contaminated locations. At all tested locations, Dehalococcoides 16S rRNA genes and VC-reductase genes were present in the groundwater, albeit not in all monitoring wells. Moreover, differences between and within locations were observed with respect to the ratio of vcrA and bvcA with Dehalococcoides 16S rRNA genes in the individual monitoring wells (Fig. (Fig.1).1). The concentration of Dehalococcoides 16S rRNA genes ranged for most monitoring wells from below the detection limit (101 gene copies/ml) up to 104 gene copies/ml but exceeded 105 gene copies/ml at locations H, I, and K, where enhanced bioremediation techniques were applied.
Of the two VC-reductase genes, vcrA was usually more abundant than bvcA. At five different locations, the vcrA concentration was greater than 105 gene copies/ml, although for most locations concentrations up to 104 gene copies/ml were found. Except for monitoring wells pb-307 (location D) and pb-605 (location J), in all samples vcrA genes were detected when the concentration of Dehalococcoides 16S rRNA gene copies was greater than 103/ml. For samples with more than 103 Dehalococcoides 16S rRNA gene copies, the average ratio of vcrA to Dehalococcoides 16S rRNA gene copies was found to be 1.3 ± 0.5, with a correlation coefficient of 0.62. For monitoring wells with Dehalococcoides 16S rRNA concentrations less than 103 gene copies/ml, vcrA gene copy numbers were 11.5 ± 6.9 times higher than Dehalococcoides 16S rRNA genes. However, the difference between both genes is smaller than the standard error for quantification of the individual genes, which differed from 0.52 to 0.75 order of magnitude for the Dehalococcoides 16S rRNA and VC- or TCE-reductase genes. Remarkably, at location B, the amount of vcrA gene copies was approximately 10 to 100 times greater than the detected amount of Dehalococcoides 16S rRNA genes (P < 0.05).
bvcA genes could not be detected in 71% of the analyzed monitoring wells. In most other monitoring wells the bvcA concentration was lower than that of the Dehalococcoides 16S rRNA genes (P < 0.05). At locations G and J, bvcA was below detection limit in all monitoring wells. Only in 15 monitoring wells were bvcA copy numbers greater than 103 copies/ml. Of these 15 wells, 11 were from locations where bioremediation was stimulated by addition of electron donor. At locations C, F, and J, bvcA but no vcrA was detected. At locations A, B, E, H, I, and K, vcrA and bvcA were simultaneously present. Here, the vcrA concentration was always 1 to 2 orders of magnitude higher than the bvcA concentration (P < 0.01).
tceA genes were most abundant in monitoring wells of locations F and G at concentrations of greater than 103 copies/ml. For the other locations, tceA genes were present in one or more monitoring wells at concentrations of up to 104 copies/ml; however, the genes were not detected at locations D and H. No correlation was observed between tceA and the Dehalococcoides 16S rRNA or VC-reductase genes.
RDA showed that variation in Dehalococcoides 16S rRNA, VC-reductase, and TCE-reductase gene copy numbers was explained by different geochemical parameters (Fig. (Fig.2),2), since the vectors representing the different genes show different angles with the vectors that represent specific geochemical conditions. In general, the smaller the angle between vectors in the RDA-plot, the better the correlation between the parameters these vectors represent. Vectors that point in opposite direction indicate that corresponding variables are negatively correlated. Along the first canonical axis (i.e., the x axis in Fig. Fig.2),2), the Dehalococcoides 16S rRNA and vcrA genes were both positively correlated with pH, EPA score, and the methane, ethene, and VC concentrations. Along the second canonical axis (y-axis) both VC-reductase genes correlated negatively with Dehalococcoides spp. 16S rRNA. Dehalococcoides 16S rRNA correlated best with Fe2+, NO3, DOC and biodegradation potential. Variation in VC and CAH concentrations and conductivity of the groundwater had the highest correlation with vcrA genes. BvcA genes were on the y axis correlated to the same environmental parameters as vcrA genes, but on the x axis the bvcA genes were more positively correlated to Fe2+, NO3, and Eh (a high redox potential). Sulfate concentration was strongly negatively correlated to Dehalococcoides 16S rRNA and vcrA genes. tceA genes correlated to different, more oxidized, geochemical parameters than the other analyzed genes and even opposite to vcrA genes.
In microcosms with groundwater from locations B, C, and E, incubated with protamylasses and VC, dechlorination of VC to ethene was observed (Table (Table2).2). No VC dechlorination was observed in groundwater from locations F and G during the incubation period of 4 months. In microcosm HMW68, VC was dechlorinated to ethene, whereas in microcosm HMW267 no dechlorination occurred. In the absence of VC dechlorination, no increase of Dehalococcoides 16S rRNA, vcrA, or bvcA gene copy numbers was found. In contrast, an increase of 2 to 4 orders of magnitude in the gene concentrations was observed in VC-dechlorinating microcosms during the incubation period. After incubation, total gene concentrations for both Dehalococcoides 16S rRNA and VC-reductase genes greater than 107 copies/ml were observed. In the VC-dechlorinating microcosms, vcrA gene copy numbers were 1 to 2 orders of magnitude greater than the bvcA gene copy numbers, as in the groundwater. However, in microcosm Cmwdb2 a significant increase of bvcA occurred, whereas no vcrA was detected. This corresponded to the absence of vcrA and presence of bvcA in groundwater of location C. After incubation, the vcrA/Dehalococcoides 16S rRNA gene ratio in dechlorinating microcosms was 0.49 ± 0.06 and thus lower than that observed in the groundwater.
The aim of the present study was to identify the relative importance of the Dehalococcoides TCE-reductase TceA and VC-reductases VcrA and BvcA in chlorinated ethene polluted groundwater and to identify the geochemical parameters that most significantly contribute to variation in corresponding 16S rRNA, tceA, vcrA, and bvcA genes. Although a few environmental studies reported the presence of tceA, vcrA, or bvcA genes in combination with Dehalococcoides spp. for one or two locations (2, 3, 11), this is the first study that described quantities of all those genes at multiple locations and in combination with geochemical data, allowing for robust multivariate analysis. At all tested locations, VC-reductase-encoding genes were detected, and the highest gene copy numbers were observed at locations where enhanced bioremediation techniques were applied. In general, vcrA genes were detected in more monitoring wells and at higher concentrations than bvcA genes. This is in accordance with the results of Carreon-Diazconti et al. (3). At only one of the locations investigated in the present study (location C) was bvcA present whereas vcrA apparently was not. This contrasts to other environmental studies (2, 11), where the amount of bvcA gene copies numerically exceeded the amount of vcrA genes. Thus, since our results showed that the geographical distribution of vcrA and bvcA is heterogeneous, it can be concluded in combination with other studies that at different locations, the dominant VC-reductase need not be the same. Therefore, both vcrA and bvcA genes should be monitored when VC-reductase-encoding genes are used to identify VC dechlorination capacity.
In monitoring wells with more than 103 Dehalococcoides 16S rRNA gene copies/ml, the correlation coefficient with vcrA gene copy numbers was 0.62 (Fig. (Fig.1A).1A). This indicates that both genes are correlated to each other in the tested monitoring wells. Different factors can be the reason for this relatively low correlation coefficient, e.g., variation in numbers of gene copies per Dehalococcoides cell. This should be addressed in future studies, e.g., by assessing 16S rRNA/VC-reductive dehalogenase gene ratios at the single-cell level. The relative amount of VC-reductase genes to Dehalococcoides 16S rRNA genes is relatively high (>1) compared to the studies of Lee et al. (11) and Scheutz et al. (19), which reported ratios of less than 1.
The amount of Dehalococcoides 16S rRNA genes correlated well with the EPA score for the biodegradation potential of chlorinated ethenes (Fig. (Fig.2).2). The EPA score was developed to identify for a polluted location whether it is likely that chlorinated ethenes are transformed via reductive dechlorination, for example, by Dehalococcoides spp. Our study confirms that for predicting the presence or absence of Dehalococcoides spp. at a location, the EPA score is a robust and better parameter than individual geochemical parameters, such as the total concentration of CAHs and the sulfate or nitrate concentrations. In contrast, TCE- and VC-reductase genes appeared to have a weaker correlation with the EPA score than the Dehalococcoides 16S rRNA genes. This indicates that under favorable conditions for reductive dechlorinating bacteria, Dehalococcoides strains with genes encoding for vcrA or bvcA, such as strains VS, BAV1, and GT, are not necessarily enriched. Thus, the capacity for TCE and VC dechlorination is not necessarily high when the EPA score is high or when high numbers of Dehalococcoides spp. are present. Dehalococcoides 16S rRNA genes are an effective predictor of overall reductive dechlorination capacity; however, these phylogenetic marker genes cannot predict the biodegradation potential for a specific chlorinated ethene, such as VC. Rather, the specific reductase genes are more direct predictors of, respectively, TCE and VC dechlorination capacity. However, it should always be kept in mind that groundwater samples might not fully represent the in situ conditions, since bacteria can have different affinities for the groundwater or soil matrix.
Variation in the presence and abundance of VC-reductase vcrA appeared to be more strongly related with concentrations of VC, CAH, and methane than Dehalococcoides 16S rRNA genes. However, both genes had a positive correlation with ethene. Carreon-Diazconti et al. (3) also found a positive relation between high concentrations of ethene and vcrA but not with bvcA. However, since ethene can be produced from other contaminants than VC in the groundwater, it cannot be concluded that ethene alone is a good indicator for VC dechlorination. Dehalococcoides spp. and vcrA correlated with a low redox potential. In contrast, bvcA and tceA correlated with a high redox potential, sulfate, nitrate, and iron, and least with methane and the EPA score. In particular, tceA had a more positive correlation with more oxidized conditions than Dehalococcoides 16S rRNA genes and vcrA. Possibly, Dehalococcoides strains containing tceA, which dechlorinate higher chlorinated ethenes (PCE and TCE), tolerate more oxidized conditions than VC dechlorinating Dehalococcoides strains. This is in agreement with the findings of Amos et al. (1), who showed that Dehalococcoides strains respond differently to oxygen exposure: strains with VC-reductase genes are more susceptible to oxygen inhibition than others (1). Lu et al. (13) showed that for dechlorination of VC, the hydrogen threshold values were in the range of methanogenesis (2 to 24 nM), whereas these threshold values for PCE and TCE were in the range of denitrification and ferric iron reduction (0.1 to 0.4 nM) (13). Unfortunately, hydrogen data were only available for locations A, I, and J and could therefore not be included in the RDA. However, these data also showed that in monitoring wells with a hydrogen concentration above 1 nM, VC-reductase gene concentrations greater than 103/ml were observed.
In the monitoring wells, a relatively low correlation between concentrations of vcrA and bvcA genes was found. RDA suggested that microorganisms that contain vcrA or bvcA genes have different niches, where Dehalococcoides spp. with bvcA genes prefer higher redox potential. Niche differentiation of these genes was also observed in soil column experiments, where specific enrichment of vcrA or bvcA genes was observed after the addition of different carbon sources (unpublished data).
In line with previous studies (11, 19), VC dechlorination in microcosms was found to be correlated to an increase of vcrA and bvcA gene copies. The mere presence of VC-reductase or Dehalococcoides 16S rRNA genes in low concentrations (103 gene copies/ml) was not in all cases related to the occurrence of dechlorination of VC. Rather, an increase of the VC reductase gene copy numbers with an order of magnitude or more during microcosm incubation correlated with biodegradation. In the present study, in situ monitoring of VC dechlorination over time was only possible at one location. In this case, VC dechlorination also correlated with increase of vcrA and bvcA genes. As a robust indicator for VC dechlorination, an increase of VC reductase gene concentrations with an order of magnitude should be observed. In combination with other in situ parameters, such as ethene formation, redox potential, or a decrease in the VC concentration, the confidence in in situ VC dechlorination can be reinforced.
In groundwater and VC-dechlorinating microcosms, gene copy numbers for vcrA were structurally 10 to 100 times higher than for bvcA. This difference is significantly higher than the standard error within the replicate analyses. Thus, incubation in the presence of VC under anaerobic conditions did not influence the average ratio between vcrA and bvcA gene copies. In groundwater from location C, where vcrA genes appeared to be absent, bvcA genes were enriched to the same levels as vcrA in other microcosms. These results suggest that bacteria containing VC-reductase vcrA are in competition with bacteria containing VC-reductase bvcA, whereby species with vcrA are more efficient under tested groundwater conditions and therefore more readily enriched. In the literature, similar doubling times for Dehalococcoides strains VC and BAV1 have been reported (5, 8). It should be noted, however, that these values were determined under artificial laboratory conditions (e.g., culture medium), whereas in our study the populations were enriched in the original groundwater. This supports the idea of niche differentiation between vcrA and bvcA gene containing Dehalococcoides, as proposed above.
In the present study, we demonstrated that Dehalococcoides 16S rRNA and the specific reductase genes tceA, vcrA, and bvcA are widespread within different groundwater systems. The geographical distribution of the genes was not homogeneous, depending on the geochemical conditions, whereby tceA and bvcA correlated more to oxidized conditions than Dehalococcoides 16S rRNA and vcrA genes. Because variation in the VC-reductase gene numbers was not directly correlated to variation in Dehalococcoides spp., VC-reductase genes are better monitoring parameters for VC dechlorination capacity than Dehalococcoides 16S rRNA genes. To indicate VC dechlorination at contaminated locations, high concentrations (>103 copies/ml) of vcrA or bvcA, and an increase of more than an order of magnitude should be detected over time.
This study was supported by funding from the Stichting Kennisontwikkeling Kennisoverdracht Bodem (SKB) and the European Union through the Aquaterra Integrated project of the Sixth Framework Programme (FP6-SUSTDEV-2003-505428).
We thank Cajo ter Braak for assistance with the statistical analysis and Arian Koeken for practical assistance with the qPCR analyses.
Published ahead of print on 11 December 2009.
†Supplemental material for this article may be found at http://aem.asm.org/.