Population Dynamics of Iron-Oxidizing Communities in Pilot Plants for

The investigations revealed that the iron-oxidizing community was clearly dominated by two groups of Betaproteobacteria affiliated with the poorly kno...
0 downloads 0 Views 2MB Size
Environ. Sci. Technol. 2009, 43, 6138–6144

Population Dynamics of Iron-Oxidizing Communities in Pilot Plants for the Treatment of Acid Mine Waters ELKE HEINZEL,† EBERHARD JANNECK,‡ FRANZ GLOMBITZA,‡ ¨ MANN,† AND MICHAEL SCHLO J A N A S E I F E R T * ,†,§ Interdisciplinary Ecological Center, TU Bergakademie Freiberg, Leipziger Strasse 29, 09599 Freiberg, Germany, and Department of Biotechnology, G.E.O.S. Freiberg Ingenieurgesellschaft m.b.H., 09633 Tuttendorf, Germany

Received January 9, 2009. Revised manuscript received June 23, 2009. Accepted July 8, 2009.

The iron-oxidizing microbial community in two pilot plants for the treatment of acid mine water was monitored to investigate the influence of different process parameters such as pH, iron concentration, and retention time on the stability of the system to evaluate the applicability of this treatment technology on an industrial scale. The dynamics of the microbial populations were followed using T-RFLP (terminal restriction fragment length polymorphism) over a period of several months. For a more precise quantification, two TaqMan assays specific for the two prominent groups were developed and the relative abundance of these taxa in the iron-oxidizing community was verified by real-time PCR. The investigations revealed that the iron-oxidizing community was clearly dominated by two groups of Betaproteobacteria affiliated with the poorly known and not yet recognized species “Ferrovum myxofaciens” and with strains related to Gallionella ferruginea, respectively. These taxa dominated the microbial community during the whole investigation period and accelerated the oxidation of ferrous iron despite the changing characteristics of mine watersflowingintotheplants.Thus,itisassumedthatthetreatment technology can also be applied to other mine sites and that these organisms play a crucial role in such treatment systems.

Introduction Mining activities bring about the exposure of pyrite and other sulfidic minerals to oxygen and water and thus accelerate the oxidation of these minerals (1). The results of this oxidation process are waters with huge amounts of sulfate, dissolved iron, and other metals as well as high acidity (2). Since these pollutants can be rinsed out of abandoned and active mines, the drainage waters pose a serious risk to the environment and efforts must be made to counteract this problem (3). Conventionally, the acid mine water is neutral* Corresponding author phone: 49 341 2351352; fax: 49 341 2351786; e-mail: [email protected]. † TU Bergakademie Freiberg. ‡ G.E.O.S. Freiberg Ingenieurgesellschaft m.b.H. § Present Address: Department of Proteomics, Helmholtz Centre for Environmental Research-UFZ, Permoser Strasse 15, 04318 Leipzig, Germany. 6138

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 16, 2009

ized by the addition of lime which results in the precipitation of iron hydroxides (4). The treatment of mine water by biological ferrous iron oxidation at a pH around 3 has been reported as an innovative treatment technology to lower the load of iron and sulfate of drainage waters by the precipitation of iron hydroxysulfate minerals (5). This technology requires the acceleration of the ferrous iron oxidation by iron-oxidizing microorganisms, because the rate of abiotic ferrous iron oxidation is very low at pH below 4 (2). Thus, information about the microbial community is of crucial importance for process optimization to achieve high oxidation rates in the treatment system and to avoid or overcome process failures. The investigation of the iron-oxidizing community in such a treatment plant at a single time point revealed that a poorly known group of Betaproteobacteria affiliated with “Ferrovum myxofaciens” dominated the microbial community (6). So far “Ferrovum myxofaciens”, isolated from a stream draining an abandoned copper mine in Wales (accession number EF133508, unpublished data), has not formally been described and only little information about its physiology is available. However, it is known that this species can solely use ferrous iron as an electron donor, has an obligately autotrophic metabolism, and appears to be less acid tolerant than the well-studied species Leptospirillum ferrooxidans and Acidithiobacillus ferrooxidans (7). The apparent dominance of the hardly known “Ferrovum myxofaciens”-related organisms in a plant for the treatment of acidic mine waters at a single time point raised the question of whether this community is typical for the plant or whether the snapshot of the former study gave an atypical picture. Moreover, the application of the innovative iron oxidation technology on an industrial scale requires the stability of the microbial community under seasonal changes in temperature and within an adequate range of other variable process parameters, such as the relative flow rate, retention time, pH, and ferrous iron concentration. Since the microbial community was investigated only for a single time point in the first version of the pilot plant (6), information was needed about the population dynamics over a longer period under varying operation conditions. Therefore, in the present study, we monitored the microbial community in two treatment plants, which differed in the chemical composition of the inflowing drainage water and in their capacity of mine water. Over several months, two different molecular techniques were used to investigate whether the conspicuous dominance of the “Ferrovum” species holds up under various operation conditions. The fingerprint technique terminal restriction fragment length polymorphism (T-RFLP) is an effective and rapid method for the monitoring of substantial changes in microbial populations and has been successfully applied in numerous studies (8, 9). A more precise quantification of selected phylogenetic groups was performed by real-time polymerase chain reaction (PCR), since fingerprint techniques such as T-RFLP disregard species with a low relative abundance in the community (10). By the combination of both molecular techniques, we obtained an extensive data set that contributed valuable information to the assessment of the practicability of this mine water treatment technology on an industrial scale.

Materials and Methods Pilot Plants, Sampling, and DNA Preparation. The microbial communities in two different pilot plants located on a lignite opencast pit in Lusatia in Eastern Germany (see detailed 10.1021/es900067d CCC: $40.75

 2009 American Chemical Society

Published on Web 07/21/2009

TABLE 1. Sequences of Primers and TaqMan Probes Used in This Study and Their Final Concentrations in Real-Time PCR Reactionsa target 16S rRNA gene

primer/probe sets (5′-3′)

Gallionella ferruginea relatives

122f: ATATCGGAACATATCCGGAAGT 384r: GGTATGGCTGGATCAGGC 203Probe: HEX-GGATCGCAAGACCTCTCGCTTTCG-BHQ1 “Ferrovum myxofaciens” relatives 643f: ACTGGCAAGCTAGAGTCTGT 847r: TCGCGTTAGCTTCGTTACTGA 822Probe: FAM-CAAGTTGCCCAACAACCAGTTGAC- BHQ1 bacteria 27f: AGAGTTTGATCCTGGCTCAG 357r: CTGCTGCSTCCCGTA

final concentration (nM) 400 900 160 400 900 250 300 300

reference

this study this study 12

a

The target sequences of the TaqMan primer/probe sets were based on relevant sequence data obtained from an extensive clone library analysis of the sampling site in a recent study (6) and on sequence data of closely related strains retrieved from the GenBank Database. FAM and HEX are the fluorophores of the two different TaqMan probes. BHQ1, Black Hole Quencher.

FIGURE 1. Phylogenetic tree of 16S rRNA gene sequences illustrating the relatedness of selected clones from the recent study of the pilot plant (6) affiliated with the Betaproteobacteria and the fragment lengths (nt) of their TRFs and of their pseudo-TRFs. The highlighted TRFs were selected for T-RFLP analysis. Cluster FV 1 comprised three clone sequences (EU360488, EU360489, EU360494); cluster FV 2 comprised six clone sequences (EU360490, EU360495, EU360496, EU360501, EU360503, EU360504). The tree was calculated with the parsimony algorithm on the basis of 16S rRNA gene sequences with a length of approximately 1360 nt using the ARB software. Acidiphilium angustum and Desulfurella acetivorans were used as outgroups. description in the Supporting Information, Figure S1) were studied. The first pilot plant was inoculated with an enriched culture, which was derived from a water sample from a lignite mine and cultivated on 9K medium (11) for several weeks selecting for autotrophic acidophilic iron oxidizers like Acidithiobacillus ferrooxidans. The second plant, in contrast, did not receive an inoculum. From both pilot plants, 1 L water samples were collected weekly from near the inflow, from the middle of the oxidation basin, and from near the effluent pipe, respectively (Figure S1 in the Supporting Information). Additionally, 10 g of precipitated iron hydroxysulfate minerals was collected from biomass carriers at irregular intervals and chemical parameters such as pH and temperature as well as concentration of oxygen, sulfate, ferrous iron, and ferric iron were measured. Due to the high loads of iron and the low pH, the preparation of the DNA from the solid and water samples required an adapted protocol (Figure S2 in the Supporting Information). The detailed protocol for DNA preparation and details about the analysis of chemical parameters are provided in the Supporting Information. T-RFLP Analysis. Bacterial 16S rRNA gene sequences were amplified from extracted microbial community DNA using a PCR premix from Thermo Fisher Scientific (see PCR condition in the Supporting Information), 0.2 µg/µL bovine serum albumin and 270 nM Cy5-labeled primer 27f (Table 1) plus 270 nM unlabeled primer 907r (5′-CCGTCWAT-

TYMTTTRAGTTT-3′) (12). Restriction digestion was separately performed with HaeIII, HhaI, Eco47I, and AluI (Fermentas). (Detailed description of T-RFLP analysis is provided in the Supporting Information.) Fragments were analyzed with a Beckman Coulter CEQ8000 genetic analysis system. The 16S rRNA genes cloned in the recent study (6) were analyzed likewise, and additionally, TRF (terminal restriction fragment) lengths of the clones were estimated by virtual digestion using the webcutter software (version 2.0, http://rna.lundberg.gu.se/cutter2/). The obtained TRF database which was specific for the investigation area was used for the assignment of TRFs detected in the DNA extracts of solid and water samples. According to Hallberg et al. (13), the relative abundance of one TRF signal was calculated as the peak area of the respective TRF divided by the total peak area of the T-RFLP pattern. Due to the occurrence of pseudo-TRFs and the observed overlap of TRFs and pseudo-TRFs (Figure 1), certain TRFs were not considered for the analysis of T-RFLP data. The relative abundance of Acidithiobacillus ferrooxidans was based exclusively on the TRFs resulting from the digestion with AluI. The relative abundance of the Gallionella relatives was based exclusively on TRFs resulting from the restriction with HhaI. For the determination of the relative abundances of the species related to “Ferrovum myxofaciens”, of the Alphaproteobacteria and of the Deltaproteobacteria, TRFs resulting from the digestion with AluI and HhaI were VOL. 43, NO. 16, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6139

FIGURE 2. Results of T-RFLP of the water samples from the middle of the oxidation basin of the first pilot plant and chemical parameters measured in the oxidation basin of the plant. Due to the occurrence of pseudo-TRFs in several DNA extracts, T-RFLP analysis was restricted to those TRFs described in the Materials and Methods. In those cases in which the relative abundance of the phylogenetic group could be analyzed by the TRFs resulting from AluI and HhaI, the relative abundance according to AluI is illustrated by a filled symbol and the relative abundance according to HhaI is illustrated by an unfilled symbol of the same color. Data points illustrating the relative abundance of the different phylogenetic groups with a value of zero were not included in the diagram. considered. In these cases, the relative abundances of these three groups according to the digestion with AluI were shown by filled data points while the relative abundances according to the digestion with HhaI were illustrated by unfilled symbols. Pseudo-TRFs were excluded from the analysis. Some T-RFLP patterns revealed prominent signals, which could not be assigned to any individual sequence known from the clone library analysis (Figure S3 in the Supporting Information). The identification of these unknown signals is described in the Supporting Information. Real-Time PCR. The two dominant taxonomic groups in the iron-oxidizing community which were affiliated with “Ferrovum” and with Gallionella-related species were quantified by real-time PCR using TaqMan assays. Two primer/ probe sets specific for these groups (Table 1) were designed using the ARB software package, which comprises tools for the handling of sequence databases and the analysis of sequence data (14). The specificity of primers and probes was checked using the Probe Match tool of the Ribosomal Database Project II (15). Furthermore, unspecific amplification was tested by PCR and real-time PCR experiments applying the primers and probe specific for the Gallionellarelated bacteria or for the relatives of “Ferrovum myxofaciens” to clones from the recent study of the pilot plant (6), which carried the 16S rRNA gene sequences of different phylogenetic groups. Both TaqMan assays were performed using the Absolute QPCR Mix (Thermo Fisher Scientific). The determination of the relative abundances of the Gallionella relatives and the “Ferrovum”-related bacteria by real-time PCR also required the quantification of the total bacterial 16S rRNA genes in the DNA extracts. Since the quantification of total bacterial 16S rRNA genes with the 6140

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 16, 2009

TaqMan assay of Yu et al. (16) resulted in a poor PCR efficiency, although PCR conditions had been varied over a wide range, bacterial 16S rRNA genes were quantified using a SYBR Green assay with the Absolute QPCR SYBR Green Mix (Thermo Fisher Scientific). Due to inhibitory contaminants, DNA extracts were applied in dilutions of 1:100 (corresponding to 0.4-1.0 ng sample DNA), 1:200, and 1:400. The standard curves were determined using serial dilutions of amplified 16S rRNA gene sequences from clone libraries recently constructed (6), which were affiliated with the respective phylogenetic groups. Further details on real-time PCR and PCR protocols are provided in the Supporting Information.

Results Chemical Parameters of the Drainage Waters Flowing into the Pilot Plants. The drainage waters pumped into the plants were characterized by high loads of iron and sulfate (Table S1 in the Supporting Information). The chemical composition of the drainage water that flowed into the first pilot plant was relatively constant, because the water was pumped out of a single dump well. In contrast, the drainage water that flowed into the second pilot plant originated from several dump wells, and thus, the chemical composition of the inflow changed. Due to the microbial iron oxidation in the plants, the pH decreased to about 3 (Figures 2 and 3) and ferrous iron concentration declined to different degrees. The intensive aeration of the drainage water near the inflow resulted in an oxygen concentration of 8 to 9 mg/L in these zones of the plants decreasing to 5-7 mg/L near the effluent pipe. Detection and Treatment of Pseudo-TRFs in the T-RFLP Analysis. Instead of the expected TRF signals of the dominant

FIGURE 3. Results of T-RFLP of the water samples from the middle of the oxidation basin of the second pilot plant and chemical parameters measured in the plant. Due to the occurrence of pseudo-TRFs in several DNA extracts, T-RFLP analysis was restricted to those TRFs described in the Materials and Methods. In those cases in which the relative abundance of the phylogenetic group could be analyzed by the TRFs resulting from AluI and HhaI, the relative abundance according to AluI is illustrated by a filled symbol and the relative abundance according to HhaI is illustrated by an unfilled symbol of the same color. Symbols illustrating the relative abundance of phylogenetic groups with a value of zero were not included in the diagram. Betaproteobacteria, the T-RFLP patterns of several DNA extracts contained prominent signals, which could not immediately be identified. A repetition of the T-RFLP analysis yielded identical results. The sequence analysis of these fragments revealed that the signals could be assigned to the “Ferrovum” or to the Gallionella-related organisms. Via webcutter analysis of these sequences, we discovered that the unexpected T-RFLP signals resulted from the digestion at the second, third, or fourth restriction site of the respective enzyme. TRFs arising from the digestion at a site other than the first restriction site are designated pseudo-TRF in the literature (17). For further interpretation of the T-RFLP data, the potential pseudo-TRFs of those clones from a recent study (6), which were affiliated with “Ferrovum” or Gallionellarelated strains, were determined using the webcutter software (Figure 1). Microdiversity in several sequence positions was detected among the “Ferrovum”-related sequences from a recent study (6) leading to variations in the length of the TRFs (Figure 1). Due to the precision of the capillary gelelectrophoresis of (1 nt (8) and due to the microdiversity occurring within the “Ferrovum” group, several pseudo-TRFs as well as several TRFs could not unambiguously be assigned to the “Ferrovum” or to the Gallionella-related group. Therefore, with the aid of the webcutter software, we looked for a restriction enzyme, which clearly distinguished between the “Ferrovum” and the Gallionella-related group and found the enzyme Eco47I. However, the resolution of Eco47I was conspicuously lower than the resolution of the other enzymes resulting in an overestimation of the relative abundance of the Betaproteobacteria. As a result of the overlap of several TRFs with pseudo-TRFs and the low resolution of the enzyme Eco47I, the T-RFLP analysis of the “Ferrovum” group was exclusively based on the fluorescence intensity of the significant TRFs with a length of 169-171 nt after digestion with AluI and on

the fluorescence intensity of the significant TRF with a length of 205-209 nt resulting from the digestion with HhaI. Correspondingly, for the analysis of the Gallionella-related group, only the fluorescence intensity of the TRFs with the length of 366 nt or 566 nt resulting from the digestion with HhaI were considered (Figure 1). T-RFLP Analysis of the Microbial Community in the First Pilot Plant. T-RFLP analyses were performed with water samples from the first pilot plant addressing the dynamics in the microbial population over a period of 11 months (Figure 2). The microbial community appeared to be clearly dominated by two groups of Betaproteobacteria, which were related to “Ferrovum myxofaciens” or to Gallionella ferruginea, respectively. The composition of the microbial community appeared to be uniform in the water samples taken from different sampling points in the plant (Table S2 in the Supporting Information). After the operation had been started in November 2005, drainage water was discontinuously pumped into the plant until continuous operation started in the end of December 2005. In this initial period, the oxidation process of ferrous iron started, which could be observed by a decrease of the pH, an increase of the redox potential, and a decline of the ferrous iron concentration (Figure 2). Due to the discontinuous inflow of mine water and the long retention time of the water in the oxidation basin, the water temperature of the pilot plant decreased to 5 °C. In this initial operation period, relatives of Gallionella formed the most abundant group (Figure 2). In four samples, relatives of Acidithiobacillus ferrooxidans were detected presumably arising from the inoculation of the pilot plant with the enrichment culture. After the continuous operation had been started and so increased feeding with ferrous iron was provided, the “Ferrovum” relatives became more important. However, until April 2006, the ferrous iron concentration as well as the ferric iron concentration alternated considerably VOL. 43, NO. 16, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6141

FIGURE 4. (A) Relative abundances of “Ferrovum”- and Gallionella-related groups with respect to total 16S rRNA genes determined in DNA extracts from different sampling points of the second pilot plant by T-RFLP using AluI and HhaI and by real-time PCR in two separate runs. While the total amount of 16S rRNA encoding DNA, which was detected by the universal Sybr Green assay, was set to 100% for real-time PCR analysis, the settings of T-RFLP considered all signals which accounted for at least 10% of the total fluorescence intensity and the total peak area of the T-RFLP pattern was set to 100%. (B) Comparison of absolute amounts of DNA encoding the target 16S rRNA in a 1 ng DNA extract determined in two separate real-time PCR runs. and the microbial community appeared to be dominated by the two groups of Betaproteobacteria with varying frequencies and rarely by heterotrophic Alphaproteobacteria affiliated with Acidiphilium or Acidocella. During the following period, the “Ferrovum” relatives clearly dominated the microbial community under quite constant ferrous iron feeding, pH, and ferric iron concentration. The Gallionella relatives regained in importance in the final operation period. This shift in the community was accompanied by a slight increase in pH as well as by a decrease of the ferric iron concentration. During continuous operation, the water temperature of the oxidation basin turned out to be within a range of 14-20 °C due to the relatively constant temperature of the mine water flowing into the plant. T-RFLP Analysis of the Microbial Community in the Second Pilot Plant. The microbial community in the second pilot plant was dominated by the same two groups of Betaproteobacteria as in the first pilot plant (Figure 3), and again, the composition of the microbial community appeared to be uniform in the whole oxidation basin (Table S3 in the Supporting Information). The Gallionella relatives constituted the most abundant group in the initial phase from October to November 2006 as well as in January 2007, when the turnover of drainage water was considerably enhanced and the oxidation rate dropped. The “Ferrovum” relatives appeared to clearly dominate the microbial community over the longest operation period from February to October 2007, when a constant feeding of drainage water containing ferrous iron was provided and an adequate throughput of drainage water was adjusted. During this period, the oxidation rate of ferrous iron alternated considerably without a correlation to the relative abundance of the “Ferrovum” group. In the final period from October 2007 to January 2008, organisms which could not be identified by T-RFLP were additionally found in the microbial community, but the occurrence of these unknown species was not associated with the ferrous iron concentration or with the oxidation rate. Furthermore, Deltaproteobacteria, which were distantly 6142

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 16, 2009

related to Desulfobacca species, and Acidocella relatives were detected in individual samples over the whole operation period. T-RFLP Analysis of Solid Samples. The relatives of “Ferrovum” were detected as the dominant group in all solid samples. The Gallionella-related strains accounted only for 0-6%, while other taxa could not be found in these samples (Table S4 in the Supporting Information). Real-Time PCR and Comparison of the Results with T-RFLP Data. The dominant Betaproteobacteria affiliated with “Ferrovum” and Gallionella-related strains were quantified more precisely in several samples using two specific TaqMan assays. In replicate runs, the total amount of DNA encoding the target 16S rRNA genes in a 1 ng DNA extract was confirmed demonstrating the reproducibility of these assays (Figure 4B, Table S5 in the Supporting Information). The relative abundances of the Betaproteobacteria with respect to total bacterial 16S rRNA genes detected by realtime PCR turned out to be in the same range as the results of the T-RFLP analysis, and thus, the dominance of these taxa in the microbial community was confirmed (Figure 4 (A)). In concordance with the findings of other authors, the amplification of undiluted DNA extracts was inhibited (18). The compromise between a high dilution of the DNA extract to avoid inhibition of amplification and a lower dilution of the DNA extract to provide enough target 16S rRNA gene sequences for detection resulted in a narrow range of acceptable template concentration for real-time PCR. A few DNA extracts could not be analyzed by real-time PCR, because the PCR reaction was inhibited even in reactions with the highest dilution of template DNA and a further dilution of the respective DNA extract yielded a concentration of target DNA below the detection limit.

Discussion The investigation of the population dynamics in the treatment plants confirmed the leading role of the poorly known species “Ferrovum myxofaciens” in the iron-oxidizing community,

which had already been disclosed in the former study of the first pilot plant focused on a single time point (6). Moreover, the investigations revealed that Gallionella-related strains appeared to be a dominant species in the plants at some time points, although they had been detected in the recent study only with a very low frequency (6). In literature, Gallionella ferruginea is described as a stalkforming, autotrophic iron oxidizer of which habitats are sharply limited to a slightly acidic pH of 5.0-6.5 and to microaerophilic conditions (19, 20). Nevertheless, relatives of Gallionella ferruginea were also detected in several mine water communities with relative abundances of up to 47% (13, 21). With respect to the high oxygen concentrations and the acidic pH in the pilot plants, the habitats investigated in the present study differed considerably from the range of physiochemical conditions required by known representatives of Gallionella ferruginea. Interestingly, the composition of the microbial community in the different pilot plants was very similar. Both plants were clearly dominated by relatives of “Ferrovum myxofaciens” and of Gallionella despite the different size of the plants and the different characteristics of the influent mine water. Thus, these parameters seem to have selected the same type of microbial community. Furthermore, the stability of the microbial community and the continuous oxidation of ferrous iron in the second pilot plant, in which mine water of changing characteristics was treated, indicated that this treatment technology can be applied to a variety of different mine waters and that relatives of Gallionella and of “Ferrovum myxofaciens” should be the dominant organisms in the system. In both pilot plants’ initial operation periods, characterized by unstable operation conditions with respect to ferrous iron concentration, throughput of drainage water, and redox potential, relatives of Gallionella seemed to thrive better than the “Ferrovum” relatives. Acidithiobacillus ferrooxidans, which was used as inoculum of the first pilot plant, was supposed to be well adapted to the acidic conditions in the pilot plant (22), but this species did not prevail in the ironoxidizing community. The “Ferrovum” group dominated the community under proper process conditions with constant ferrous iron feeding and stable operation conditions. During the whole operation period, the microbial diversity appeared to be considerably lower than the diversity reported in studies of three other mine waters in which 4-12 T-RFLP signals had been recorded per sample (13, 23, 24). The microbial diversity also appeared to be lower than the diversity observed in the former study of the plant in which representatives of 5-10 different taxonomic classes had been detected by clone library analyses (6). Only in a few samples further taxonomic groups, besides “Ferrovum” and Gallionella-related strains, were detected. These belonged to the Alphaproteobacteria or to the Deltaproteobacteria, which are unlikely to have an impact on the oxidation rate of ferrous iron. However, it has to be taken into account that T-RFLP with the settings used in this study has only detected signals that reach 10% of the total fluorescence intensity. Thus, there was a bias against bacteria of low relative abundance, but these bacteria should not be major players in microbial iron oxidation. Most iron-oxidizing isolates are described in the literature as mesophilic or thermophilic species (25). Even though biological ferrous iron oxidation was reported over a wide range of climatic conditions, a low temperature may become a limiting factor due to the significantly lower activity of most iron-oxidizing bacteria at temperatures approaching 5 °C (25, 26). Surprisingly, in this study, the measurement of pH, redox potential, and ferrous iron concentration in the first pilot plant disclosed that the iron oxidation process

started quite fast in the initial operation period even at a very low temperature of approximately 5 °C. In contrast to the temperature, the throughput of drainage water and the retention time appeared to be more crucial process parameters, because the decrease of the retention time of mine water from average values of 7-10 h to approximately 4 h in January 2007 resulted in a dramatic decline in the oxidation rate. Thus, the treatment system seemed to be limited to a retention time of >4 h, due to an elution of the biomass suspended in the oxidation basin. In this study, the application of T-RFLP and real-time PCR to environmental samples containing heavy loads of iron was afflicted with a couple of problems. The pseudoTRFs, which occurred in several samples, posed a problem in the interpretation of T-RFLP data, due to the overlap of fluorescent signals of some TRFs with those of some pseudoTRFs. All tested enzymes with an adequate resolution yielded such overlaps of fluorescent signals because of the high phylogenetic relationship between the two dominant groups and the resultant sequence similarity in most of the sequence position. Therefore, the analysis of T-RFLP data was only based on a few TRFs accepting that the exclusion of the pseudo-TRFs from the analysis apparently affected the conclusions. The impact of the pseudo-TRFs on the interpretation of the T-RFLP patterns is indicated by the differences in the relative abundances of the taxonomic groups after the digestion with AluI and HhaI which are illustrated by filled and unfilled symbols in Figures 2 and 3. Egert and Friedrich (17) described the occurrence of pseudo-TRFs as a PCR artifact resulting from the formation of partly singlestranded amplicons, but the influence of PCR inhibitors in the DNA extracts, as it occurred in the present study, on the formation of pseudo-TRF is currently unclarified, and we observed no correlation between the occurrence of pseudoTRFs and specific DNA sequences or sample characteristics. Real-time PCR is very sensitive for PCR inhibitors, which are often coextracted with the DNA from environmental samples and affect the amplification efficiency in real-time PCR (27-29). A difference of 5% in the amplification efficiency between two samples with an identical amount of template can result in twice as much PCR product in one sample after 26 cycles (30). Thus, applying real-time PCR techniques to environmental samples requires special attention to find an adequate range of DNA extract dilution to circumvent inhibition by PCR inhibitors and to provide enough target DNA for accurate detection (18). These parameters restrict the detection limit of real-time PCR assays applied to environmental samples to orders of magnitude higher than the positive control standards (31). In the present study, we obtained reproducible results by real-time PCR applying a narrow range of template DNA concentration (Table S6 in the Supporting Information) and we could confirm our interpretation of the T-RFLP data (Figure 4). The results of the present study gave a new insight into water treatment by biological ferrous iron oxidation and indicated that poorly known species affiliated with “Ferrovum myxofaciens” and with Gallionella were the predominant iron oxidizers in the treatment system. The overall composition of the iron-oxidizing community and the oxidation capacity of the pilot plants proved to be robust under seasonal variations and different compositions of drainage water. However, due to the unknown copy number of 16S rRNA genes in the detected species, which can vary from one to fifteen copies in the prokaryotic genome independent from the phylogenetic affiliation (32), and due to the undefined DNA extraction efficiency, we obtained no information about the cell numbers of the different taxa. Therefore, further investigations should deal with the determination of total cell numbers in the pilot plant to find a correlation between the oxidation rate under various operation conditions and VOL. 43, NO. 16, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6143

the amount of biomass present and active in the pilot plant. In addition, the application of this technology to waters from other mine sites will give further information about the role and dominance of Gallionella and “Ferrovum” relatives in the treatment technology.

Acknowledgments We are grateful to the BMBF for funding this study (Project 01RI05014), the Max-Buchner research foundation for the sponsorship (Number 2721), and Vattenfall Europe Mining & Generation for the support of the project. We greatly thank Daniel Terno, Mario Kohl, Gu ¨ nter Ra¨tsel, and Klaus-Dieter Herbach for the technical service at the pilot plants.

Supporting Information Available Detailed description of the pilot plants and DNA preparation, details about the analysis of chemical parameters, additional information about real-time PCR and T-RFLP, chemical characteristics of drainage water, representative data about the detection of pseudo-TRFs, and tables with the results of real-time PCR and T-RFLP are provided in the Supporting Information. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Lottermoser, B. G. Mine wastes: Characterization, treatment and environmental impacts; Springer: Berlin, 2003. (2) Appelo, C. A. J.; Postma, D. Geochemistry, groundwater and pollution; A. A. Balkema: Rotterdam, The Netherlands, 1993. (3) Johnson, D. B.; Hallberg, K. B. Acid mine drainage remediation options: a review. Sci. Total Environ. 2005, 338 (1-2), 3–14. (4) Sengupta, M. Environmental impacts of mining: monitoring, restoration, and control; Lewis publishers: London, 1993. (5) Glombitza, F.; Janneck, E.; Arnold, I.; Rolland, W.; Uhlmann, W. Eisenhydroxysulfate aus der Bergbauwasserbehandlung als Rohstoff. In Consulting-Erfahrungen und Kontakte fu ¨ r Neuanfa¨nge-Unterta¨giger Bergbau auf Industrieminerale in Deutschland; GDMB Gesellschaft fu ¨ r Bergbau, M., Rohstoff-und Umwelttechnik e.V., Ed.; GDMB Medienverlag: ClausthalZellerfeld, Germany, 2007; pp 31-40. (6) Heinzel, E.; Hedrich, S.; Janneck, E.; Glombitza, F.; Seifert, J.; Schlo¨mann, M. Bacterial diversity in a mine water treatment plant. Appl. Environ. Microbiol. 2009, 75 (3), 858–861. (7) Rowe, O. F.; Johnson, D. B. Comparison of ferric iron generation by different species of acidophilic bacteria immobilized in packed-bed reactors. Syst. Appl. Microbiol. 2008, 31 (1), 68–77. (8) Schu ¨ tte, U. M.; Abdo, Z.; Bent, S. J.; Shyu, C.; Williams, C. J.; Pierson, J. D.; Forney, L. J. Advances in the use of terminal restriction fragment length polymorphism (T-RFLP) analysis of 16S rRNA genes to characterize microbial communities. Appl. Microbiol. Biotechnol. 2008, 80 (3), 365–380. (9) Giraffa, G.; Neviani, E. DNA-based, culture-independent strategies for evaluating microbial communities in food-associated ecosystems. Int. J. Food. Microbiol. 2001, 67 (1-2), 19–34. (10) Blackwood, C. B.; Hudleston, D.; Zak, D. R.; Buyer, J. S. Interpreting ecological diversity indices applied to terminal restriction fragment length polymorphism data: insights from simulated microbial communities. Appl. Environ. Microbiol. 2007, 73 (16), 5276–5283. (11) Silverman, M. P.; Lundgren, D. G. Studies on the chemoautotrophic iron bacterium Ferrobacillus ferrooxidans. I. An improved medium and a harvesting procedure for securing high cell yields. J. Bacteriol. 1959, 77 (5), 642–647. (12) Lane, D. J. 16S/23S rRNA sequencing. In Nucleic acid techniques in bacterial systematics; Stackebrandt, E., Goodfellow, M., Eds.; John Wiley & Sons: Chichester, U.K., 1991; pp 115-175. (13) Hallberg, K. B.; Coupland, K.; Kimura, S.; Johnson, D. B. Macroscopic streamer growths in acidic, metal-rich mine waters in North Wales consist of novel and remarkably simple bacterial communities. Appl. Environ. Microbiol. 2006, 72 (3), 2022–2030. (14) Ludwig, W.; Strunk, O.; Westram, R.; Richter, L.; Meier, H.; Yadhukumar; Buchner, A.; Lai, T.; Steppi, S.; Jobb, G.; Forster,

6144

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 16, 2009

(15)

(16)

(17)

(18)

(19) (20)

(21)

(22) (23) (24) (25)

(26)

(27)

(28) (29) (30) (31) (32)

W.; Brettske, I.; Gerber, S.; Ginhart, A. W.; Gross, O.; Grumann, S.; Hermann, S.; Jost, R.; Ko¨nig, A.; Liss, T.; Lu ¨ βmann, R.; May, M.; Nonhoff, B.; Reichel, B.; Strehlow, R.; Stamatakis, A.; Stuckmann, N.; Vilbig, A.; Lenke, M.; Ludwig, T.; Bode, A.; Schleifer, K. H. ARB: a software environment for sequence data. Nucleic Acids Res. 2004, 32 (4), 1363-1371. Cole, J. R.; Chai, B.; Marsh, T. L.; Farris, R. J.; Wang, Q.; Kulam, S. A.; Chandra, S.; McGarrell, D. M.; Schmidt, T. M.; Garrity, G. M.; Tiedje, J. M. The Ribosomal Database Project (RDP-II): previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy. Nucleic Acids Res. 2003, 31 (1), 442–443. Yu, Y.; Lee, C.; Kim, J.; Hwang, S. Group-specific primer and probe sets to detect methanogenic communities using quantitative real-time polymerase chain reaction. Biotechnol. Bioeng. 2005, 89 (6), 670–679. Egert, M.; Friedrich, M. W. Formation of pseudo-terminal restriction fragments, a PCR-related bias affecting terminal restriction fragment length polymorphism analysis of microbial community structure. Appl. Environ. Microbiol. 2003, 69 (5), 2555–2562. Harms, G.; Layton, A. C.; Dionisi, H. M.; Gregory, I. R.; Garrett, V. M.; Hawkins, S. A.; Robinson, K. G.; Sayler, G. S. Real-time PCR quantification of nitrifying bacteria in a municipal wastewater treatment plant. Environ. Sci. Technol. 2003, 37 (2), 343– 351. Hallbeck, L. E.-L.; Pederson, K.; Genus, I. Gallionella. In Bergey’s manual of systematic bacteriology, 2nd ed.; Brenner, D. J., Krieg, N. R., Staley, J. T., Eds.; Springer: New York, 2005; pp 880-886. Hanert, H. H. The genus Gallionella. In The Prokaryotes. A handbook in habitats, isolation and identification of bacteria; Starr, M. P., Tru ¨ per, H. G., Balows, A., Schlegel, H. G., Eds.; Springer: Berlin, 1981; pp 509-515. He, Z.; Xiao, S.; Xie, X.; Zhong, H.; Hu, Y.; Li, Q.; Gao, F.; Li, G.; Liu, J.; Qiu, G. Molecular diversity of microbial community in acid mine drainages of Yunfu sulfide mine. Extremophiles 2007, 11 (2), 305–314. Jensen, A. B.; Webb, C. Ferrous sulphate oxidation using Thiobacillus ferrooxidans: a Review. Process Biochem. 1995, 30 (3), 225–236. Bruneel, O.; Duran, R.; Koffi, K.; Casiot, C.; Fourcans, A.; ElbazPoulichet, F.; Personne´, J.-C. Microbial diversity in a pyrite-rich tailings impoundment. Geomicrobiol. J. 2005, 22, 249–257. Coupland, K.; Johnson, D. B. Geochemistry and microbiology of an impounded subterranean acidic water body at Mynydd Parys, Anglesey, Wales. Geobiology 2004, 2 (2), 77–86. Kupka, D.; Rzhepishevska, O. I.; Dopson, M.; Lindstro¨m, E. B.; Karnachuk, O. V.; Tuovinen, O. H. Bacterial oxidation of ferrous iron at low temperatures. Biotechnol. Bioeng. 2007, 97 (6), 1470– 1478. Dopson, M.; Halinen, A. K.; Rahunen, N.; Ozkaya, B.; Sahinkaya, E.; Kaksonen, A. H.; Lindstro¨m, E. B.; Puhakka, J. A. Mineral and iron oxidation at low temperatures by pure and mixed cultures of acidophilic microorganisms. Biotechnol. Bioeng. 2007, 97 (5), 1205–1215. Stults, J. R.; Snoeyenbos-West, O.; Methe, B.; Lovley, D. R.; Chandler, D. P. Application of the 5′ fluorogenic exonuclease assay (TaqMan) for quantitative ribosomal DNA and rRNA analysis in sediments. Appl. Environ. Microbiol. 2001, 67 (6), 2781–2789. Teng, F.; Guan, Y.; Zhu, W. A simple and effective method to overcome the inhibition of Fe to PCR. J. Microbiol. Methods. 2008, 75 (2), 362–364. Mitchell, J. I.; Zuccaro, A. Sequences, the environment and fungi. Mycologist 2006, 20 (2), 62–74. Peirson, S. N.; Butler, J. N.; Foster, R. G. Experimental validation of novel and conventional approaches to quantitative real-time PCR data analysis. Nucleic Acids Res. 2003, 31 (14), e73. Chandler, D. P. Redefining relativity: quantitative PCR at low template concentration for industrial and environmental microbiology. J. Ind. Microbiol. Biotechnol. 1998, 21, 128–140. Klappenbach, J. A.; Dunbar, J. M.; Schmidt, T. M. rRNA operon copy number reflects ecological strategies of bacteria. Appl. Environ. Microbiol. 2000, 66 (4), 1328–1333.

ES900067D