Identification of active denitrifiers by DNA-stable isotope probing and amplicon sequencing reveals Betaproteobacteria as responsible for attenuation of nitrate contamination in a low impacted aquifer

Identification of active denitrifiers by DNA-stable isotope probing and amplicon sequencing... Abstract Groundwater reservoirs constitute important freshwater resources. However, these ecosystems are highly vulnerable to contamination and have to rely on the resident microbiota to attenuate the impact of this contamination. Nitrate is one of the main contaminants found in groundwater, and denitrification is the main process that removes the compound. In this study, the response to nutrient load on indigenous microbial communities in groundwater from a low impacted aquifer in Uruguay was evaluated. Denitrification rates were measured in groundwater samples from three different sites with nitrate, acetate and pyrite amendments. Results showed that denitrification is feasible under in situ nitrate and electron donor concentrations, although the lack of readily available organic energy source would limit the attenuation of higher nitrate concentrations. DNA-stable isotope probing, combined with amplicon sequencing of 16S rRNA, nirS and nirK genes, was used to identify the active denitrifiers. Members of the phylum Betaproteobacteria were the dominant denitrifiers in two of three sites, with different families being observed; members of the genus Vogesella (Neisseriaceae) were key denitrifiers at one site, while the genera Dechloromonas (Rhodocyclaceae) and Comamonas (Comamonadaceae) were the main denitrifiers detected at the other sites. denitrification, groundwater, stable isotope probing, nirS, nirK, Betaproteobacteria INTRODUCTION Aquifers are key freshwater resources and their water quality represents an important environmental issue worldwide. Domestic, industrial and agricultural activities rely on the quality of groundwater, which is strongly linked to the geochemical processes modulated by the activity and diversity of indigenous microbiota (Griebler and Lueders 2009). Nitrate (NO3−) is one of the key contaminants that affect groundwater quality (Rivett et al.2008). Unlike other pollutants that usually reach the subsurface by defined, well-localised sources, nitrate load in aquifers occurs continuously and in a diffusive way, making both contamination control and removal difficult. Thus, natural attenuation will be a key process in the regulation of nitrate concentration. Denitrification is the metabolic process that consumes the largest amount of nitrate in subsurface (Rivett et al.2008; Weymann et al.2008). Despite its importance, factors that limit denitrification in situ, such as geochemical conditions, remain unclear, especially for aquifers with lower pollution levels (Rivett et al.2008; Eschenbach, Well and Walther 2014). During complete denitrification, nitrate is used as a terminal electron acceptor, and it is reduced to molecular nitrogen (N2) after passing through several reduction states. The reduction of nitrite (NO2−) to nitric oxide (NO) is considered a key step for denitrifiers, since the nitrogen goes from a dissolved anion state to gaseous state and can be easily removed from the system (Tiedje 1988). Several geochemical factors are involved in the regulation of the process, such as electron donor composition and availability, oxygen concentration, pH, temperature, microbial acclimation, salinity, presence of inhibitory substances, although electron donor and oxygen concentration are considered the main limiting factors (Smith and Duff 1988; Santoro, Boehm and Francis 2006; Rivett et al.2008). The variety of electron donors used for denitrification is broad and both organotrophic and lithotrophic processes occur in aquifers (Smith and Duff 1988; Rivett et al.2008; Weymann et al.2008; Torrentó et al.2011). Denitrification is a metabolic trait that is widespread among bacteria, archaea and fungi (Shoun et al.1992; Philippot, Hallin and Schloter 2007). Therefore, analysis of ribosomal RNA genes alone may not be suitable for assessing denitrifier diversity. The use of nitrite-reductase encoding genes, nirS and nirK, as biomarkers in combination with ribosomal genes allow the assessment of both functional diversity and the phylogenetic affiliation of denitrifiers. Stable isotope probing (SIP) of nucleic acids and fatty acids is a technique that has been used since 2000 for the identification of active populations within a microbial community in a specific metabolic process (Neufeld et al.2007b). It has enabled the study of a wide diversity of processes, such as syntrophic interactions for propionate oxidation (Lueders, Pommerenke and Friedrich 2004), methylated compounds assimilation (Neufeld et al.2007a; Antony et al.2010) and identification of acetate-utilising microbes (Osaka et al.2006; Schwarz et al.2007) and pollutants degraders (Singleton et al.2005; Sun and Cupples 2012). Denitrification has also been studied using SIP technique coupled with the analysis of both ribosomal and denitrification genes, with the utilisation of labelled carbon compounds such as acetate, succinate and/or methanolas substrates for nitrate reduction in coastal sediments, paddy soil and activated sludge (Maneesha et al.2004; Ginige, Keller and Blackall 2005; Osaka et al.2006, 2008; Saito et al.2008). However, reports on the use of SIP (both DNA and SIP coupled to Fatty Acids Analysis) to study denitrification in aquifers are scarce and are more frequently linked to the degradation of pollutants such as aromatic or complex compounds (Pelz et al.2001; Fischer, Manefield and Bombach 2016). Our study site, Raigón aquifer, is a sand and gravel groundwater system located in south-west Uruguay. Nitrate levels in this groundwater reservoir will eventually increase due to nearby agricultural and industrial activities. A previous study evaluated the potential for denitrification by groundwater microbial communities from the Raigón aquifer and characterised more than 50 denitrifiers isolated from several sampled wells (Bellini et al.2013). This study describes the use of DNA-SIP combined with amplicon sequencing of 16S rRNA and nirS and nirK genes to determine active indigenous denitrifiers and assess their capacity to remove nitrate from this groundwater system in a scenario of an increasing nutrients load. MATERIALS AND METHODS Sample collection and physicochemical parameters determination Groundwater samples were obtained in July 2012 from three water supply wells (C, D and H) at Raigón aquifer, as previously described by Bellini et al. (2013). Briefly, groundwater samples were collected from water supply wells after three bottle volumes (5 L) were pumped and a constant value of dissolved oxygen was reached. Twenty-five litres of groundwater were collected from each site. Groundwater was transported at 4°C, and was processed within 24 h after sampling for all experiments. Groundwater temperature, pH and dissolved oxygen were measured at the site. The sampling wells lie along a 50-km stretch and have depths that lie between 37 and 50 m. The three selected wells cover the range of nitrate (Table 1) concentrations detected along the aquifer, and they are representative of various human activities that occur on the surface of the aquifer. In the surroundings of site C there is an agricultural stablishment where dairy milking and farming activities occur, while site D is located in a potato cleaning and processing plant. Site H is located in a small town, which relies on septic tanks for sewage treatment. Table 1. Groundwater physicochemical parameters at sampling sites. Site  T (°C)  pH  NO3− (mg L−1/mmol L−1)  Total Fe (μg L−1)  SO42− (mg L−1)  O2 (mg L−1)  C  18.6  7.7  33.6/0.54  <17  49.0  7.4  D  18.0  7.1  10.3/0.17  29  33.2  3.9  H  19.2  7.4  24.9/0.40  65  28.8  6.4  Site  T (°C)  pH  NO3− (mg L−1/mmol L−1)  Total Fe (μg L−1)  SO42− (mg L−1)  O2 (mg L−1)  C  18.6  7.7  33.6/0.54  <17  49.0  7.4  D  18.0  7.1  10.3/0.17  29  33.2  3.9  H  19.2  7.4  24.9/0.40  65  28.8  6.4  View Large Nitrate, nitrite, total iron and sulphate were analysed according to the American Public Health Association ‘Standard Methods for the examination of Water and Wastewater’ methods (APHA 1998), with 4500-NO3- B, 4500-NO2- B, 3500-Fe-B Phenanthroline Method and 4500 methods, respectively. Total organic carbon (TOC) was analysed with a Shimadzu TOC-VSCN analyser and the limit of quantification was 1.1 mg L−1. Denitrification potential of groundwater samples The potential for nitrate attenuation of indigenous microbial community was tested by measuring denitrification rates in groundwater samples with different amendments and compared to endogenous rates (i.e. without amendment). Rates were measured in groundwater obtained from wells C, D and H with the acetylene blockage technique (Mahne and Tiedje 1995). Triplicate sterile 120-mL flasks were flushed with filtered O2-free N2 while being filled aseptically with 40 mL of groundwater. Nitrate plus either acetate or pyrite (crushed and sieved to particles of <65 μm) were added to compare organotrophic and lithotrophic denitrification capacity, respectively. In addition, flasks with only nitrate amendment (containing between 3 and 12 times the nitrate concentration measured in the wells), non-amended flasks and acetate or pyrite-amended sterile controls (autoclaved) were also set up for each well. Acetate was chosen as the carbon amendment since it is a non-fermentable substrate and other parallel processes besides denitrification would be unlikely, due to incubation conditions. In addition, most of the previously isolated bacteria from the Raigón aquifer were able to use it as a carbon and electron source for denitrification (Bellini et al.2013). Acetate and pyrite were added according to a complete NO3− reduction stoichiometry. Every set of triplicate flasks contained the following (final concentration): phosphate buffer (1 mmol L−1, pH 7.1) and either 2 mmol L−1 KNO3 and 1.5 mmol L−1 ammonium acetate (nitrate plus acetate amended), 2 mmol L−1 KNO3, 1200 mg L−1 pyrite and 0.5 mmol L−1 Na2CO3 (nitrate plus pyrite amended) or 2 mmol L−1 KNO3 (nitrate amended). Acetylene was added (10% of the headspace) to every flask, and flasks were incubated in the dark with continuous shaking at 20°C. Samples from the headspace of every flask were removed at several time points for N2O measurement by gas chromatography (Shimadzu GC-2014 gas chromatograph); gas chromatography conditions and calculations of denitrification rates were done as described in Bellini et al. (2013). Nitrate and nitrite concentrations were also measured during the incubations by HPLC as described by Tarlera and Denner (2003). To test if denitrification rates from different treatments or wells were significantly different, a squared root transformation was applied to rates and then compared with an analysis of variance (ANOVA) using P ≤ 0.05 and Tukey correction with InfoStat software (Di Rienzo et al.2009). All transformed data satisfied the ANOVA assumptions. Assessment of groundwater basal communities: 16S rRNA gene amplicon sequencing To estimate the structure of the naturally occurring bacterial communities present in the sampled groundwater, 20 L of groundwater were filtered through a 0.22-μm pore cellulose acetate membrane. DNA was then extracted from biomass using the PowerWater DNA Isolation Kit (MoBio Laboratories, Inc., CA, USA). For amplicon libraries, the 16S rRNA gene was amplified with PCR primers 563F and 802R (Cole et al.2009). PCR amplicons were purified using Agencourt AMPure XP beads (Agencourt Bioscience Corporation, MA, USA) and sequenced at the INDEAR sequencing facilities (Rosario, Argentina) on a Genome Sequencer FLX (454-Roche Applied Sciences) according to the manufacturer's instructions. SIP incubations, gradient centrifugation and DNA extraction SIP incubations were set up in 150-mL microcosms. Two replicates, named A and B, were prepared for each sampled site and treatment; 70 mL of groundwater were aseptically placed in sterile flasks containing the following (final concentration): phosphate buffer (1 mmol L−1) and KNO3 (2 mmol L−1) while flushed with O2-free N2. Organotrophic treatment flasks were amended with 1.5 mmol L−1 (final concentration) of 13C-labelled sodium acetate (Na13CH3COO); granulated pyrite (particle size < 65 μm) and Na213CO3 were added to lithotrophic treatment flasks to a final concentration of 1200 mg L−1 and 0.5 mmol L−1, respectively. In order to confirm isotopic enrichment of DNA in 13C incubations, 12C unlabelled controls were set up for every treatment. Controls with only acetate were not included because other parallel processes (fermentation, heterotrophic anaerobic respiration, lithotrophic anaerobic denitrification) are unlikely or marginal in the incubation conditions. In addition, determination of the denitrification rates and the SIP incubations showed that the amounts of acetate and nitrate consumed were stochiometric corresponding to a denitrification process (see below). Incubations were carried out in the dark at 20°C with continuous shaking. Nitrate and nitrite were monitored in every flask as described before. Acetate consumption at the end of the incubation was determined by HPLC, with an Aminex-87H column and 5 mmol L−1 sulphuric acid as the mobile phase. Runs were carried out at 35°C and UV detection at 210 nm was used. Biomass was harvested by centrifugation when at least 90% of added nitrate and the stoichiometric amount of acetate had been consumed. Pellets were frozen at –20°C until DNA extraction. DNA was extracted from cell pellets using a modification of the protocol described in Neufeld et al. (2007a). Briefly, 200 μL of SET buffer (40 mmol L−1 ethylenediamine tetra-acetic acid, 50 mmol L−1 Tris-HCl pH 9.0 and 0.75 mmol L−1 sucrose) and 22.5 μL of freshly prepared lysozyme solution (9 mg mL−1) were added to the pellets. After incubation at 37°C for 30 min, 25 μL of 10% (w/v) SDS and 7 μL of a fresh solution of proteinase K (20 mg mL−1) were added and the tubes were incubated at 55°C for 2 h. Following the addition of 125 μL of SET buffer, the lysates were transferred to a phase-lock tube (Eppendorf, Hamburg, Germany) and two extractions with 375 μL of phenol:cholorform:isoamylalcohol (25:24:1) were performed. DNA in the aqueous phase was precipitated with 2 μL of glycogen (Roche, Basel, Switzerland), 170 μL of a 7.5 mol L−1 ammonium acetate solution and 700 μL of 95% ethanol, overnight at –20°C. Tubes were centrifuged at 21 000g, and the pellets were washed twice with 500 μL 80% (v/v) ethanol. After drying, DNA was re-suspended in 50 μL of sterile water. DNA was run on an agarose gel, and concentration and quality were analysed with a NanoDrop 2000 spectrophotometer. Ultracentrifugation, fractionation and precipitation of DNA were carried out according to Neufeld et al. (2007b). Twelve or thirteen fractions were collected from 13C duplicates and from the 12C control flask from each well. The buoyant density of each fraction was determined indirectly by measuring the refractive index (nD-TC) with a digital refractometer (Reichert AR200, Reichert Inc., NY, USA). Those fractions with the highest densities for which a positive 16S rRNA gene amplification was obtained (see later) were selected for further analysis and designated heavy fractions (13C-DNA, densities between 1.7211 and 1.7269 g mL−1). Fractions with densities of 1.7035–1.7094 g mL−1 were designated and analysed as light fractions (12C-DNA). SIP incubations follow-up: denaturing gradient gel electrophoresis analysis of 16S rRNA genes In order to confirm that heavy fractions from 13C incubations showed a different community pattern than light fractions from 13C incubations and than heavy fractions from 12C incubations, denaturing gradient gel electrophoresis (DGGE) analysis of 16S rRNA genes was carried out. In addition, DGGE was performed to compare the structure of communities from replicated SIP incubations on non-fractionated DNA from each replicate and well.16S rRNA gene PCR primers for Bacteria were 341F-GC(5΄-CGCCCGCCGCGCGCGG CGGGCGGGGCGGGGGCACGGGGGGCCTACGGGAGGCAGCAG-3΄) and 907R (5΄-CCGTCAATTCMTTTRAGTTT-3΄) (Muyzer, De Waal and Uitterlinden 1993); PCR reactions were carried out in a total volume of 50 μL and contained 2 mmol L−1 MgCl2, 0.2 mmol L−1 dNTP, 0.2 μmol L−1 of each primer, 2.5 U of DreamTaq DNA Polymerase (Fermentas, Burlington, Ontario, Canada), 5 μL 10X Taq buffer, 0.07% BSA and 30 ng DNA. Cycling conditions included an initial step of 5 min at 94°C followed by 30 cycles of 94°C for 30 s, 55°C for 30 s and 72°C for 45 s, with a final extension step of 72°C for 8 min. PCR products were checked for size and purity on agarose gels, and the concentration was estimated by comparison to a 1-kb ladder (Fermentas). Approximately 300 ng of PCR product were loaded in each lane of an 8% (w/v) polyacrylamide gel with a gradient of denaturants (urea and deionised formamide) that ranged from 20% to 60%. Electrophoresis was performed using a D-Code System (Bio-Rad, Hercules, CA, USA) on TAE 1X buffer at 60°C for 16 h at 80 V. The gel was stained with SYBR Gold Nucleic Acid Gel Stain (Invitrogen, Paisley, UK) for 1 h and visualised with an imaging device. The predominant bands from heavy fractions were excised and 16S rRNA gene was amplified as previously described, with the exception that the primers used did not have a GC-clamp. Amplicons were sequenced by Sanger method at Source BioScience, Nottingham, UK. Sequences were used in a BLAST search (Altschul et al.1997) and deposited in GenBank with the accession numbers KY548747 to KY548762. The potential role of archaeal communities in denitrification during SIP incubations was also tested. A nested PCR of the 16S rRNA gene specific for Archaea domain (Cunliffe et al.2008) was performed on heavy and light fractions of DNA isolated from every well, and only weak amplifications were observed in heavy fractions from site H. Amplicon sequencing from SIP fractions DNA from heavy and light fractions from 13C incubations was used as a template for triplicate PCR amplification of 16S rRNA gene with 503F (5΄-GTGCCAGCMGCNGCGG-3΄) and 1100R (5΄-GGGTTNCGNTCGTTR-3΄) primers. Reaction mixtures were performed in a total volume of 50 μL and contained 2 mmol L−1 MgCl2, 0.2 mmol L−1 dNTP, 0.3 μmol L−1 of each primer, 1.25 U of DreamTaq DNA Polymerase (Fermentas) and 5 μL 10X Taq buffer. Cycling conditions included an initial step of 5 min at 94°C followed by 30 cycles of 94°C for 30 s, 55°C for 30 s and 72°C for 45 s, with a final extension step of 72°C for 10 min. DNA from heavy and light fractions from replicates A was also used as a template for triplicate PCR amplification of nirS and nirK genes. The primers and conditions used were as follows: nirS cd3aF/R3Cd (Throbäck et al.2004) and nirK F1aCu/R3Cu (Hallin and Lindgren 1999); reactions mixtures were prepared in a final volume of 50 μL and contained 2 mmol L−1 MgCl2, 0.2 mmol L−1 dNTP, 1.25 U of DreamTaq DNA Polymerase (Fermentas), 5 μL 10X Taq buffer, 0.2 μmol L−1 of each primer for nirS gene and 0.5 μmol L−1 for nirK and 30 ng DNA. Cycling conditions included an initial step of 94°C for 4 min followed by 30 cycles of 30 s at 94°C, 1 min at 57°C and 1 min at 72°C with a final extension step of 10 min at 72°C. PCR products were checked for size and purity on agarose gels. All PCR products were visualised in an agarose gel, replicates were pooled and purified with Agencourt AMPure XP beads (Agencourt Bioscience Corporation) and checked on an agarose gel. PCR products were sent to Mr DNA Molecular Research (Lubbock, TX, USA) for barcoded sequencing where a second 5-cycle PCR was performed to add barcodes and sequencing keys. Following PCR, all amplicons from different samples were mixed in equal concentrations and purified using Agencourt AMPure XP beads (Agencourt Bioscience Corporation). Samples were sequenced using a Roche 454 FLX Titanium instrument and reagents according to the manufacturer's instructions. Bioinformatic analysis of amplicon sequences 16S rRNA data were analysed using the pipeline available through the QIIME platform, version 1.7 (Caporaso et al.2010b). Libraries were split using default parameters except that the number of ambiguous bases allowed was one instead of six and primer mismatches allowed was zero. Chimera checking (de novo and reference-based) and operational taxonomical units (OTU) picking (0.97 cut-off) were performed with usearch61 (Edgar 2010), and taxonomy assignment was done using the Greengenes 13_5 database with the RDP Classifier (Wang et al.2007). The representative set was aligned with Pynast software (Caporaso et al.2010a) against the Greengenes core reference alignment (DeSantis et al.2006). Finally, low-abundance OTUs (less than 1% of the dataset) and those that failed to align were filtered from the OTU table. Similar filtering options were used for naturally occurring (basal) communities’ data, except only singletons and doubletons were removed. Alpha diversity indexes were calculated within QIIME. Rarefaction analysis of the libraries was performed using the free Analytic Rarefaction 1.3 software (http://strata.uga.edu/software/index.html). Data from sequencing of denitrification genes were initially split into libraries, followed by a de novo chimera checking with usearch61 (Edgar 2010) within the QIIME platform. The split and filtered libraries were then screened for frame shifts using the RDP FunGene (Fish et al.2013) FrameBot tool (Wang et al.2013) with a length cut-off of 80 amino acids and an identity cut-off of 0.4. Frame-corrected nucleotide sequences were then clustered (0.03 distance) using uclust (Edgar 2010) within QIIME and an OTU table was constructed. Low-abundance OTUs (less that 1% of the dataset) were filtered from the OTU table. Rarefaction analysis of the libraries was performed as described previously. Representative sequences were found for nirK and nirS libraries and compared with the EMBL/GenBank database using the NCBI Blast algorithm (Altschul et al.1997). Translated sequences were aligned with MUSCLE at amino acid level. The resulting alignments were checked via comparison with aligned HMM training sequences obtained from FunGene repository (Functional gene pipeline and repository, http://fungene.cme.msu.edu/index.spr) and corrected manually when necessary. Phylogenetic trees were constructed with nucleotide sequences based on the neighbour-joining (NJ) method and maximum likelihood (ML) algorithms to confirm overall topology using MEGA6 (Tamura et al.2013) software. A bootstrap confidence analysis was performed with 1000 replicates. A total of 303 informative positions were taken into account for the analysis of the nirS gene, while 272 were used for nirK gene. All datasets have been deposited at MG RAST database under Project ‘Nitrate contamination in a low impacted aquifer’: Clean_X_16S_heavy/light corresponds to 16S rRNA amplicon libraries from SIP heavy/light fraction from site X (i.e. C, D or H); Clean_X_nirS/nirK_heavy corresponds to amplicon libraries from nirS/nirK SIP heavy fractions from site X (ie C, D or H); and Clean_X_16S corresponds to 16S rRNA amplicon libraries from basal groundwater communities from sites C, D or H. Abundance of nirS and nirK genes as determined in DNA-SIP incubations To determine the abundance of nirS and nirK genes in DNA samples from DNA-SIP incubations, qPCR was performed. PCR products used for standard curves in qPCR experiments were obtained from the isolated denitrifiers AR45 (Pseudomonas stutzeri) and AR11 (Achromobacter sp.) for nirS and nirK, respectively (Bellini et al.2013). PCR products were cloned using the TOPO TA cloning kit (Invitrogen®) following the manufacturer's instructions. Clones were subjected to PCR using primers T3–T7 provided with the cloning kit. Correct fragment length and identity of the PCR product were verified by electrophoresis on a 1.5% agarose gel and sequenced at the Macrogen Sequencing Service, Korea, using the T7 primer. Plasmid was extracted using the Purelink Quick Plasmid miniPrep kit (Invitrogen®), and the size of the plasmid fragment was verified by electrophoresis in a 1.5% agarose gel. A triplicate amplification of the cloned fragments with T3 and T7 primers was performed. PCR products were pooled, diluted to 90 μL with MilliQ water and purified with MICROCON®100 columns. Following a fragment size checking in an agarose gel, the product was quantified with Qubit dsDNA HS Assay kit in a Qubit 2.0 Fluorometer (Invitrogen®) and the gene copy number was calculated based on the total fragment length (partial nirS or nirK fragment plus cloning vector fragments flanking the insertion point), the molecular weight of the amplicon and Avogadro's number. nirS and nirK quantification on DNA samples from heavy and light fractions from SIP incubations was performed in duplicate for samples and in triplicates for standard curves. One-in-ten and one-in-one hundred sample dilutions were included in the run to test for any inhibition of PCR reactions. No template controls were included in the run. RESULTS Denitrification activity and groundwater chemistry Physicochemical parameters measured in sampled groundwater are shown in Table 1. TOC and nitrite concentrations were below the quantification limit (1.1 and 0.02 mg L−1, respectively) in samples for all the wells. Denitrification rates in groundwater samples from sites C, D and H were estimated under four different conditions (endogenous, nitrate amended, nitrate plus pyrite amended and nitrate plus acetate amended). Results were analysed statistically to compare the endogenous denitrification potential and the effects of amendments within each site (Table 2). According to the endogenous and nitrate amended rates, denitrification is feasible at all three sites, even in a scenario of increasing nitrate concentration. At sites C and H, nitrate plus acetate amendments significantly stimulated denitrification compared to the respective endogenous rates. Conversely, neither nitrate nor nitrate plus pyrite additions changed denitrification rates significantly at any sample site. Samples revealed differential response to the added nutrients, as denitrification rates with nitrate addition showed significantly higher denitrification rates in well H compared to other wells. Table 2. Mean denitrification rates of indigenous bacterial communities per well (μmol N2O L−1 h–1).     Amendment  Site  Endogenous  Nitrate  Nitrate + pyrite  Nitrate + acetate  C  0.14a,A  0.22a,A  0.91a,AB  4.43b,B  D  0.07a,A  0.14a,A  0.07a,A  0.65a,A  H  1.44a,A  3.09ab,B  2.02a,B  7.42b,B      Amendment  Site  Endogenous  Nitrate  Nitrate + pyrite  Nitrate + acetate  C  0.14a,A  0.22a,A  0.91a,AB  4.43b,B  D  0.07a,A  0.14a,A  0.07a,A  0.65a,A  H  1.44a,A  3.09ab,B  2.02a,B  7.42b,B  P = 0.05, lowercase letters indicate significant differences between rates within one site (lines), uppercase letters indicate differences between sites (columns). View Large Bacterial community structure in basal groundwater samples A total of 19 844 16S rRNA gene sequences were obtained from the three groundwater samples before incubation. This number was reduced to 13 447 after quality checking and removal of chimeras. Sequences were grouped in a final OTU table composed of 936 OTUs and 11 154 reads, which were distributed as 3343, 3094 and 4717 reads for sites C, D and H, respectively. About 97% of the sequences in the final dataset were between 220 and 240 bp long. The rarefaction curves (Fig. S1, Supporting Information) levelled off at the sequenced depth for all the samples, indicating adequate coverage for each site. As shown in Fig. S2, Supporting Information, Proteobacteria was the major phylum in all three samples. The distribution within this phylum however varied between sampling sites. In site C, Alpha and Gammaproteobacteria were the dominant classes, while in sites D and H, sequences affiliated to Betaproteobacteria dominated the bacterial communities. Other phyla that were present in high proportions were Actinobacteria at site C (20%), Nitrospirae at site D (22%), Verrucomicrobia (11%) and Chlamydiae (10%) at sample H. Analysis of SIP incubations by DGGE In order to identify the denitrifiers that would respond to an input of nitrate and an energy source to the groundwater, replicate DNA-SIP incubations were set up with 2 mmol L−1 of nitrate and either 1.5 mmol L−1 of acetate or pyrite plus carbonate (final concentration 1200 mg L−1 and 0.5 mmol L−1, respectively). These were monitored over time by measuring nitrate consumption (Fig. S3, Supporting Information); acetate consumption was measured in acetate amended treatments at the end of incubation. Flasks that were amended with pyrite did not show more than 20% decrease in nitrate concentration over 4 weeks of incubation, and cells were not harvested from these flasks. This was similar to the observations obtained from the denitrification rates experiments, where no increase in the rates was observed with pyrite amendment. Acetate-amended flasks responded to the substrates and consumed the added nitrate and acetate, although in a time course that differed among sites. A shorter incubation time was observed for sample H (7 days), while 10 and 14 days were required for samples C and D to consume at least 90% of the added nitrate, and the stoichiometric amount of acetate, respectively. Bacterial community composition in non-fractionated DNA and in DNA from fractions from 13C and 12C acetate incubations obtained after gradient centrifugation and fractionation were analysed by DGGE fingerprinting of 16S rRNA genes. Biological replicates of 13C incubations and 12C controls displayed reasonable similarity among non-fractionated profiles (Fig. S4, Supporting Information); thus, replicate A was selected for further work. Heavy DNA fractions from 13C incubations in wells C and D displayed different 16S rRNA gene profiles from those obtained with light fractions from 13C incubations, with the presence of specific bands only in the heavy fractions (Fig. 1). This suggests that selective 13C acetate assimilation had occurred after nitrate addition in these wells. Moreover, these specific bands were intense in light fractions from 12C incubations, and faint, or mostly absent, in the heavy fractions from these 12C controls. Taken together, these results reveal successful 13C labelling of DNA in specific microorganisms after acetate consumption and nitrate reduction in wells C and D. Heavy fractions from 13C incubations from site H also showed different profiles from light fractions, suggesting specific labelling. However, the main band observed in heavy fractions from 13C incubation was not absent or faint in heavy fractions from 12C incubations. Therefore, the results obtained for well H do not provide enough evidence to confirm that the main band detected in heavy fractions from 13C incubations correspond to the active denitrifiers. Figure 1. View largeDownload slide Denaturing gradient gel electrophoresis analysis of 16s rRNA gene fragments in unfractionated and in heavy and light fractions from 13C- and 12C-acetate SIP incubations. Density from each fraction is shown (g mL−1). Predominant bands identification is shown (Std: DGGE standard). Figure 1. View largeDownload slide Denaturing gradient gel electrophoresis analysis of 16s rRNA gene fragments in unfractionated and in heavy and light fractions from 13C- and 12C-acetate SIP incubations. Density from each fraction is shown (g mL−1). Predominant bands identification is shown (Std: DGGE standard). Figure 1. View largeDownload slide (Continued). Figure 1. View largeDownload slide (Continued). Identification of active denitrifiers by DNA-SIP with acetate To facilitate preliminary identification of the active denitrifiers responding to a nutrient input, selected DGGE bands were analysed. 16S rRNA gene analysis of predominant bands in heavy fractions from 13C incubations by excision from gels and sequencing (Fig. 1) revealed that members of the genus Vogesella are the main active bacteria at site C, along with Comamonas, Hydrogenophaga and Duganella (all members of Comamonadaceae family) though in lower proportions. In site H on the other hand, bands that belong to Dechloromonas, Comamonas and Ensifer genera were identified in the heavy fraction, while bands excised from site D heavy DNA fractions were affiliated exclusively to the genus Dechloromonas. Datasets obtained after amplicon sequencing of 16 rRNA genes of heavy and light DNA fractions from 13C incubations recovered from acetate SIP incubations with groundwater samples C, D and H contained 99 581 raw reads. This dataset was reduced to 72 109 sequences after quality control and removal of chimeras, and sequences were grouped into 982 OTUs at 97% nucleotide similarity. Removal of low-abundance OTUs and the sequences that failed to align resulted in 66 932 16S rRNA sequences distributed over 18 different OTUs. Ninety percent of DNA sequences in the final 16S rRNA gene dataset were between 450 and 530 bp long. A rarefaction analysis showed that at a depth of 2600 sequences per sample, the number of OTUs reached a plateau for the six samples, indicating adequate sequencing coverage (Fig. S1, Supporting Information). The distribution of 16S rRNA gene sequences retrieved from heavy and light fractions of 13C incubations, and their taxonomic classification is shown in Table 3. The OTUs identified in heavy fractions showed consistent results with DGGE bands sequencing. In most cases, OTUs that arose only in heavy fractions were identified, while other OTUs were considerably enriched from light to heavy fraction. These OTUs were considered as the active consumers of the labelled acetate in the denitrifying conditions imposed, and therefore identified as active denitrifiers. Site C harboured several OTUs that occurred exclusively in the heavy DNA fraction. The predominant sequences belonged to the family Neisseriaceae, which comprised up to 42% of the 16S rRNA gene library at this site. Moreover, the genus Vogesella was the predominant component within this family (36% of the 16S rRNA gene library at site C). Members of the family Comamonadaceae were also abundant (24%) in the heavy DNA fraction at site C. In heavy DNA from site D, the dominant 16S rRNA genes were from the family Rhodocyclaceae (47%) with members of the genus Dechloromonas being a major component (45% of this 16S rRNA gene library). The most abundant OTU in site H was a member of the genus Comamonas (79% of the 16S rRNA gene library). An OTU affiliated with the Rhizobiaceae was also present although in a much lower proportion (7%). Table 3. Taxonomic composition and relative abundance (%) of bacterial communities in 13C SIP incubations for the heavy and light fractions in the three studied sites.   C  D  H  Taxon  Heavy  Light  Heavy  Light  Heavy  Light  Alphaproteobacteria               Rhizobiaceae  3  0  2  0  7  2  Betaproteobacteria               Burkholderiales                Comamonadaceae                 Aquabacterium  0  0  0  0  0  2     Comamonas  9  0  0  0  79  39     Hydrogenophaga  1  0  0  0  2  8     Other Comamonadaceae  14  0  2  0  9  15  Neisseriales               Neisseriaceae                Vogesella  36  0  0  0  0  0    Other Neisseriaceae  6  0  0  0  0  0  Rhodocyclales               Rhodocyclaceae                Dechloromonas  1  0  45  0  1  4    KD1-23  0  0  1  0  0  0    Other Rhodocyclaceae  0  0  1  0  1  2  Other Betaproteobacteria  7  0  0  0  0  0  Gammaproteobacteria               Aeromonadaceae  2  0  0  0  0  0  Pseudomonadales                 Acinetobacter  21  96  0  0  0  0     Pseudomonas  0  0  4  0  0  0  Flavobacteriia               Flavobacteriaceae                Flavobacterium  0  4  0  0  0  0   Weeksellaceae                Chryseobacterium  0  0  2  2  0  0    Other Weeksellaceae  0  0  43  98  0  0  Bacilli                Bacillus  0  0  0  0  1  28    C  D  H  Taxon  Heavy  Light  Heavy  Light  Heavy  Light  Alphaproteobacteria               Rhizobiaceae  3  0  2  0  7  2  Betaproteobacteria               Burkholderiales                Comamonadaceae                 Aquabacterium  0  0  0  0  0  2     Comamonas  9  0  0  0  79  39     Hydrogenophaga  1  0  0  0  2  8     Other Comamonadaceae  14  0  2  0  9  15  Neisseriales               Neisseriaceae                Vogesella  36  0  0  0  0  0    Other Neisseriaceae  6  0  0  0  0  0  Rhodocyclales               Rhodocyclaceae                Dechloromonas  1  0  45  0  1  4    KD1-23  0  0  1  0  0  0    Other Rhodocyclaceae  0  0  1  0  1  2  Other Betaproteobacteria  7  0  0  0  0  0  Gammaproteobacteria               Aeromonadaceae  2  0  0  0  0  0  Pseudomonadales                 Acinetobacter  21  96  0  0  0  0     Pseudomonas  0  0  4  0  0  0  Flavobacteriia               Flavobacteriaceae                Flavobacterium  0  4  0  0  0  0   Weeksellaceae                Chryseobacterium  0  0  2  2  0  0    Other Weeksellaceae  0  0  43  98  0  0  Bacilli                Bacillus  0  0  0  0  1  28  View Large Abundance and functional diversity of responding denitrifiers PCR products were obtained with gene-specific primers targeting nirS and nirK in heavy DNA fractions from 13C incubations. Interestingly, no amplification of any of the denitrification genes was observed in light DNA fractions, indicating that the only denitrifiers that were enriched during SIP incubations were able to use the labelled acetate. The abundance of both nirS and nirK genes was also determined in DNA from heavy fractions from 13C incubations and the relative abundance (nirS copies μL−1/nirK copies μL−1) was calculated. nirS gene copy numbers in the heavy fractions were 1.2–3.6 × 105 copies μL−1 of DNA, whereas nirK abundance was 1.1–1.5 × 104 copies μL−1. The abundance of nirS was at least 10 times higher than nirK for every site, which suggests that the microbial communities responding to nitrate and acetate input in the three sites are dominated by nirS-containing denitrifiers. Amplicon sequencing of nirS and nirK was performed on DNA from heavy fractions from 13C incubations. Total datasets obtained for nirS contained 16 334 reads, which were reduced to 13 481 (7024 for site C, 2862 for site D and 3595 for site H) after quality control, chimera checking and removal of sequences that failed to align. Ninety-five percent of the sequences in the final set had between 360 and 400 bp and 21 OTUs were defined (at 97% sequence identity). Rarefaction analysis was performed, and the number of OTUs reached a plateau for the three samples for both nir genes. The phylogenetic relatedness among representative nirS and nirK sequences from each OTU with sequences from both uncultivated and extant strains recovered from databases was analysed (Figs 2 and 3). Tree topology showed that gene sequences from this study clustered in different groups that were supported by high bootstrap values and consistently obtained with NJ as well as with ML methods. All the closest nirS sequences from extant bacteria found in databases were affiliated to the class Betaproteobacteria. The majority (89%) of representative nirS sequences obtained from site C belonged to the same OTU and clustered with denitrifier gene sequences found in two Vogesella species (Neisseriaceae) which were isolated from the same aquifer in a previous study (Bellini et al.2013). The remaining OTUs clustered with denitrifier gene sequences obtained from Comamonadaceae isolates (8.5%) and from the genus Dechloromonas (2%). All the nirS sequences present in the heavy DNA fraction from site D were closely related to nirS sequences obtained from isolates of the family Rhodocyclaceae. Most of these sequences (97.5%) showed maximum identity with DNA sequences retrieved directly from the environment, mainly ocean sediments, and with sequences from Dechloromonas genus isolates. The remaining 2.5% grouped closely with nirS sequences retrieved directly from the environment. Most of the nirS sequences in the site H library (55%) were closely affiliated with environmental sequences, obtained mostly from aquatic environments. Thirty-two percent of the sequences clustered with nirS sequences from Rhodocyclaceae isolates, which were distributed among the genera Dechloromonas, Azospira, Azoarcus and Zoogloea. The closest nirS sequence from extant bacteria to the remaining 13% of site H library was nirS from soil isolate I-Bh25-7 (Braker, Schwarz and Conrad 2010), a member of the Comamonadaceae family. Figure 2. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirS gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. Sequences obtained from prior culturing from this aquifer are shown in bold. Figure 2. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirS gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. Sequences obtained from prior culturing from this aquifer are shown in bold. Figure 3. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirK gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. Figure 3. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirK gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. In contrast to information on nirS, nirK gene sequences retrieved from heavy fractions from 13C incubations grouped mostly with nirK sequences from cultivated representatives of the Alpha and Gammaproteobacteria. The majority of the nirK OTUs from site C (90%) clustered with environmental clones and with nirK sequences affiliated to Alphaproteobacteria of the genera Sinorhizobuim, Ensifer, Rhizobium, Mesorhizobuim, Bradyrhizobium and Bosea (within the Rhizobiales). Gammaproteobacteria-affiliated nirK sequences represented 10% of the nirK sequences retrieved from site C, with one OTU clustering with nirK of Pseudomonas. All representative nirK sequences from site D were affiliated with nirK from Alphaproteobacteria. Seventy-two percent of the sequences clustered closely to nirK from extant Rhizobium, while the remaining 28% were affiliated with nirK sequences from Bosea. Most (93%) of the nirK sequences from site H clustered in one OTU, most closely related to nirK from members of the genera Ensifer and Sinorhizobium. Other less well-represented OTUs clustered with nirK from Rhizobium (5%), Ochrobactrum and Paracoccus (2%). DISCUSSION Our aim was to evaluate the denitrification potential of indigenous microbial communities from a low impacted aquifer that is likely face an input of nitrate and to identify the microorganisms responsible for the observed activity. The strategy combined kinetic measurements of the denitrification process together with a study of the active microbial communities using DNA-SIP, in order to better predict how denitrifiers in groundwater respond to an increase in nitrate and electron donor concentration. Natural attenuation of dissolved nitrate was feasible in the three sampled sites at in situ nitrate concentrations. However, according to the denitrification rates measured an increase in the nitrate levels without the input of a carbon source would be limited either by abundance or composition of denitrifiers or by insufficient electron donors, since in most of the sampled sites only acetate plus nitrate stimulated denitrification. This suggests that the indigenous denitrifying microbes in water from the Raigón aquifer appear to be adapted to organotrophic denitrification, which is in accordance with previous observations which establish that in general carbon deprivation is the main factor limiting denitrification in aquifers (Rivett et al.2008). However, bacteria capable of using minerals such as pyrite could be present in the aquifer but attached to sediments. Samples reacted differently to the amendments and displayed differences in their basal bacterial communities’ profiles, suggesting intrinsic differences in either the denitrifying population or in geochemical conditions. The results from this study are in agreement with previous observations involving the same aquifer wells (Bellini et al.2013), where a carbon-limited nitrate consumption potential was detected. However, sites C and D did not show an equivalent response to the same amendments seen in the previous study, which could be due to the temporal fluctuation of nutrients and/or denitrification inhibitor concentrations (Lin et al.2012). On the contrary, site H responded similarly to amendments than in the previous study. Considering the activities occurring in the surroundings of each well, groundwater in site H (located in a small town) could be receiving a relatively constant input of nutrients from the surface, while activities on sites D (potato processing plant) and C (dairy farm) would sporadically contribute nutrients to the subsurface. SIP incubations performed on groundwater under the same conditions as the denitrification activity measurements enabled us to identify key denitrifiers responding to increased nitrate contamination of the aquifer. Amplicon sequencing of 16S rRNA genes from 13C-labelled DNA from SIP incubations indicated that Betaproteobacteria play a major role in organotrophic denitrification in two sites examined. Previous molecular and/or cultivation-dependent studies have also shown that Betaproteobacteria are the main denitrifiers in diverse environments such as paddy soil (Saito et al. 2008), wastewater treatment systems (Ginige, Keller and Blackall 2005; Osaka et al.2006; McIlroy et al.2016) as well as aquifers where heterotrophic denitrification occurs (Calderer et al.2014; Zeng et al.2016). Although a high diversity was detected in naturally occurring communities in all three wells, Betaproteobacteria was the predominant class in naturally occurring communities from sites D and H, although it was less represented in site C (Fig. S2, Supporting Information). However, different denitrifying bacteria from this class respond rapidly to nitrate and carbon input in the wells. Members of the genus Vogesella were the predominant denitrifiers in samples from site C. In a previous study, Vogesella was isolated from the same aquifer from wells C and H, and its denitrifying capability with acetate as electron donor was confirmed in the laboratory (Bellini et al.2013). The fact that Vogesella was detected by both molecular and cultivation-based methods, in both samplings, suggests Vogesella is well adapted to this groundwater system and could be relevant in the response to carbon and nitrate input in this site. Interestingly, though Betaproteobacteria have been identified as important denitrifiers in different environments, Neisseriales have scarcely ever been reported as major actors in denitrification processes (Yoshida et al.2012). The ability of Vogesella to denitrify was observed when the genus was first described (Grimes et al.1997). Vogesella has also been shown to be prevalent in other processes that may occur in groundwater such as removal of heavy metals (Vishnivetskaya et al.2010) or improving resistance to disinfectants (Whiteley et al.2001). Together with our findings, this suggests that Vogesella could be relevant not only for denitrification, but also for other environmental processes occurring in groundwater ecosystems. Members of the Comamonadaceae were also identified as active denitrifiers at site C. This is also in agreement with previous findings where Comamonadaceae, along with other bacteria, were found (Wakelin et al.2011) or enriched under denitrifying conditions in groundwater samples (Bellini et al.2013; Calderer et al.2014) and in other systems, such as paddy soils and wastewater treatment (Saito et al.2008; McIlroy et al.2016). In addition, growth and molecular-based studies have shown species of genus Comamonas, a member of Comamonadaceae, to be capable of denitrification (Etchebehere et al.2001; Gumaelius et al.2001) and that they play important roles in denitrification processes in activated sludges (Ginige, Keller and Blackall 2005), paddy soils (Osaka et al.2006) and groundwater systems (Calderer et al.2014; Wang et al.2014). Site D contained only one dominant active denitrifying genus, Dechloromonas, a member of the Rhodocyclaceae. Denitrifying members of this family were isolated from the same well from Raigón aquifer in a previous study (Bellini et al.2013), suggesting that members of this family regularly respond to denitrifying conditions in this site. Rhodocyclaceae have also been depicted as key denitrifiers in systems with a greater load of nitrate and/or organic matter such as activated sludge (Ginige, Keller and Blackall 2005), paddy soil (Osaka et al.2006) and surface freshwater systems (Yu, Yang and Liu 2014). The most abundant taxon detected in site H was genus Comamonas, although according to DGGE results its role as key denitrifier was not confirmed. To further characterise the diversity of active denitrifying communities from the Raigón aquifer, we analysed nirS and nirK genes in 13C-labelled DNA from SIP incubations. Care must be taken when assigning taxonomical information to nir genes sequences since congruence between 16S rRNA and nir genes phylogenies is taxon dependent (Heylen et al.2006; Jones et al.2008). The nirS gene library from site C was dominated by two OTUs that clustered closely with nirS sequences obtained from two Vogesella species isolated from the same aquifer well (Bellini et al.2013). Retrieved nirS sequences that cluster closely with members of Comamonadaceae and Dechloromonas could also be active denitrifiers, although they were less abundant. nirK sequences affiliated with Rhizobiales also appear to represent active denitrifiers at site C. For site D, nirS sequences that clustered with nirS from Dechloromonas were most abundant while most nirK gene sequences affiliated with nirK from Rhizobium and Bosea. These results reinforce the findings obtained with 16S rRNA gene analyses and suggest that Dechloromonas is a key denitrifier in site D, with a minor contribution to the denitrification process by Rhizobiaceae. Analysis of nirS libraries in site H showed that the dominant OTUs clustered with environmental sequences and sequences from the family Rhodocyclaceae, although this family was clearly under-represented in 16S rRNA gene library. The dominant taxon observed in the 16S rRNA gene library (Comamonas) was not observed either in nirS or nirK libraries. However, most of nirS sequences from this site clustered with environmental clones, and 13% of the sequences clustered closely to the Comamonadaceae isolate I-Bh25-7. A BLAST search using 16S rRNA gene from isolate I-Bh25-7 indicates that it could belong to the genus Simplicispira, within the Comamonadaceae. Moreover, most denitrifying Comamonas isolates studied so far possesses nirS (Heylen et al.2006), and these nirS sequences should have been retrieved with the PCR primers used in this study. Accordingly, most of the nirS sequences found in this site could actually belong to the Comamonas that were detected with 16S rRNA gene sequences. Altogether, 16S rRNA, nirS and nirK data suggest that Comamonadaceae, Rhodocyclaceae and Rhizobiaceae were present at site H. In summary, the three sampled sites from the Raigón aquifer differed in microbial communities and in their denitrification rates in response to nutrients amendments. However, the indigenous communities in this aquifer have the potential to remove contaminating nitrate and perform a natural attenuation process through different and taxonomically diverse genera, mainly Betaproteobacteria. SUPPLEMENTARY DATA Supplementary data are available at FEMSEC online. Acknowledgements The authors would like to thank the Agencia Nacional de Investigación e Innovación (ANII), Laboratorio Tencológico del Uruguay (LATU), Comisión Sectorial de Investigación Científica (CSIC),Programa de Desarrollo de Ciencias Báscias (PEDECIBA) and Luciana Pereira. REFERENCES Altschul SF, Madden TL, Schaeffer AA et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res  1997; 25: 3389– 402. Google Scholar CrossRef Search ADS PubMed  American Public Health Association (APHA). Standard Methods for the Examination of Water and Wastewater . 20th ed., Washington DC: American Public Health Association, 1998. Antony CP, Kumaresan D, Ferrando L et al. Active methylotrophs in the sediments of Lonar Lake, a saline and alkaline ecosystem formed by meteor impact. ISME J  2010; 4: 1470– 80. Google Scholar CrossRef Search ADS PubMed  Bellini MI, Gutiérrez L, Tarlera S et al. Isolation and functional analysis of denitrifiers in an aquifer with high potential for denitrification. Syst Appl Microbiol  2013; 36: 505– 16. Google Scholar CrossRef Search ADS PubMed  Braker G, Schwarz J, Conrad R. Influence of temperature on the composition and activity of denitrifying soil communities. FEMS Microbiol Ecol  2010; 73: 134– 48. Google Scholar PubMed  Calderer M, Martí V, de Pablo J et al. Effects of enhanced denitrification on hydrodynamics and microbial community structure in a soil column system. Chemosphere  2014; 111: 112– 9. Google Scholar CrossRef Search ADS PubMed  Caporaso JG, Bittinger K, Bushman FD et al. PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics  2010a; 26: 266– 7. Google Scholar CrossRef Search ADS   Caporaso JG, Kuczynski J, Stombaugh J et al. QIIME allows analysis of high-throughput community sequencing data. Nat Meth  2010b; 7: 335– 6. Google Scholar CrossRef Search ADS   Cole J, Wang Q, Cardenas E et al. The Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Res  2009; 37: 141– 5. Google Scholar CrossRef Search ADS   Cunliffe M, Schäfer H, Harrison E et al. Phylogenetic and functional gene analysis of the bacterial and archaeal communities associated with the surface microlayer of an estuary. ISME J  2008; 2: 776– 89. Google Scholar CrossRef Search ADS PubMed  DeSantis TZ, Hugenholtz P, Larsen N et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microb  2006; 72: 5069– 72. Google Scholar CrossRef Search ADS   Di Rienzo JA, Casanoves F, Balzarini MG et al. InfoStat. Grupo InfoStat, FCA . Argentina: Universidad Nacional de Córdoba, 2009. http://www.infostat.com.ar (19 December 2017, date last accessed). Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics  2010; 26: 2460– 1. Google Scholar CrossRef Search ADS PubMed  Eschenbach W, Well R, Walther W. Predicting the denitrification capacity of sandy aquifers from in situ measurements using push-pull 15N tracer tests. Biogeosci Discuss  2014; 11: 16527– 72. Google Scholar CrossRef Search ADS   Etchebehere C, Errazquin MI, Dabert P et al. Comamonas nitrativorans sp. nov., a novel denitrifier isolated from a denitrifying reactor treating landfill leachate. Int J Syst Evol Micr  2001; 51: 977– 83. Google Scholar CrossRef Search ADS   Fischer A, Manefield M, Bombach P. Application of stable isotope tools for evaluating natural and stimulated biodegradation of organic pollutants in field studies. Curr Opin Biotech  2016; 41, 99– 107. Google Scholar CrossRef Search ADS PubMed  Fish JA, Chai B, Wang Q et al. FunGene: the Functional Gene Pipeline and Repository. Front Microbiol  2013; 4: 291. Google Scholar CrossRef Search ADS PubMed  Ginige MP, Keller J, Blackall LL. Investigation of an acetate-fed denitrifying microbial community by stable isotope probing, full-cycle rRNA analysis, and fluorescent in situ hybridization-microautoradiography. Appl Environ Microb  2005; 71: 8683– 91. Google Scholar CrossRef Search ADS   Griebler C, Lueders T. Microbioal biodiversity in groundwater ecosystems. Freshwater Biol  2009; 54: 649– 77. Google Scholar CrossRef Search ADS   Grimes DJ, Woese CR, MacDonell MT et al. Systematic study of the genus Vogesella gen. nov. and its type species, Vogesella indigofera comb. nov. Int J Syst Evol Micr  1997; 47: 19– 27. Gumaelius L, Magnusson G, Pettersson B et al. Comamonas denitrificans sp. nov., an efficient denitrifying bacterium isolated from activated sludge. Int J Syst Evol Micr  2001; 51: 999– 1006. Google Scholar CrossRef Search ADS   Hallin S, Lindgren P. PCR detection of genes encoding nitrite reductase in denitrifing bacteria. Appl Environ Microb  1999; 65: 1652– 7. Heylen K, Gevers D, Vanparys B et al. The incidence of nirS and nirK and their genetic heterogeneity in cultivated denitrifiers. Environ Microbiol  2006; 8: 2012– 21. Google Scholar CrossRef Search ADS PubMed  Jones CM, Stres B, Rosenquist M et al. Phylogenetic analysis of nitrite, nitric oxide, and nitrous oxide respiratory enzymes reveal a complex evolutionary history for denitrification. Mol Bio Evol  2008; 25: 1955– 66. Google Scholar CrossRef Search ADS   Lin X, McKinley J, Resch CT et al. Spatial and temporal dynamics of the microbial community in the hanford unconfined aquifer. ISME J  2012; 6: 1665– 76. Google Scholar CrossRef Search ADS PubMed  Lueders T, Pommerenke B, Friedrich MW. Stable-isotope probing of microorganisms thriving at thermodynamic limits: syntrophic propionate oxidation in flooded soil. Appl Environ Microb  2004; 70: 5778– 86. Google Scholar CrossRef Search ADS   McIlroy SJ, Starnawska A, Starnawski P et al. Identification of active denitrifiers in full-scale nutrient removal wastewater treatment systems. Environ Microb  2016; 18: 50– 64. Google Scholar CrossRef Search ADS   Mahne I, Tiedje JM. Criteria and methodology for identifying respiratory denitrifiers. Appl Environ Microb  1995; 61: 1110– 5. Maneesha PG, Hugenholtz P, Daimes H et al. Use of stable-isotope probing, full-cycle rRNA analysis, and fluorescence in situ hybridization–microautoradiography to study a methanol-fed denitrifying microbial community. Appl Environ Microb  2004; 70: 588– 96. Google Scholar CrossRef Search ADS   Muyzer G, De Waal EC, Uitterlinden AG. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl Environ Microb  1993; 59: 695– 700. Neufeld JD, Schafer H, Cox MJ et al. Stable-isotope probing implicates Methylophaga spp and novel Gammaproteobacteria in marine methanol and methylamine metabolism. ISME J  2007a; 6: 480– 91. Google Scholar CrossRef Search ADS   Neufeld JD, Vohra J, Dumont MG et al. DNA stable-isotope probing. Nat Protocols  2007b; 2: 860– 6. Google Scholar CrossRef Search ADS   Osaka T, Ebie Y, Tsuneda S et al. Identification of the bacterial community involved inmethanedependent denitri¢cation inactivated sludge using DNA stable-isotope probing. FEMS Microbiol Ecol  2008; 64: 494– 506. Google Scholar CrossRef Search ADS PubMed  Osaka T, Yoshie S, Tsuneda S et al. Identification of acetate- or methanol-assimilating bacteria under nitrate-reducing conditions by stable-isotope probing. Microb Ecol  2006; 52: 253– 66. Google Scholar CrossRef Search ADS PubMed  Pelz O, Chatzinotas A, Andersen N et al. Use of isotopic and molecular techniques to link toluene degradation in denitrifying aquifer microcosms to specific microbial populations. Arch Microbiol  2001; 175: 270– 81. Google Scholar CrossRef Search ADS PubMed  Philippot L, Hallin S, Schloter M. Ecology of denitrifying prokaryotes in agricultural soil. Adv Agron  2007; 96: 249– 305. Google Scholar CrossRef Search ADS   Rivett MO, Buss SR, Morgan P et al. Nitrate attenuation in groundwater: a review of biogeochemical controlling processes. Water Res  2008; 42: 4215– 32. Google Scholar CrossRef Search ADS PubMed  Saito T, Ishii S, Otsuka S et al. Identification of novel betaproteobacteria in a succinate-assimilating population in denitrifying rice paddy soil by using stable isotope probing. Microbes Environ  2008; 23: 192– 200. Google Scholar CrossRef Search ADS PubMed  Santoro AE, Boehm AB, Francis CA. Denitrifier community composition along a nitrate and salinity gradient in a coastal aquifer. Appl Environ Microb  2006; 72: 2102– 9. Google Scholar CrossRef Search ADS   Schwarz IKJ, Lueders T, Eckert W et al. Identification of acetate-utilizing Bacteria and Archaea in methanogenic profundal sediments of Lake Kinneret (Israel) by stable isotope probing of rRNA. Environ Microb  2007; 9: 223– 37 Google Scholar CrossRef Search ADS   Shoun H, Kim D, Uchiyama H et al. Denitrification by fungi. FEMS Microbiol Lett  1992; 94: 277– 81. Google Scholar CrossRef Search ADS   Singleton DR, Powell SN, Sangaiah R et al. Stable-isotope probing of bacteria capable of degrading salicylate, naphthalene, or phenanthrene in a bioreactor treating contaminated soil. Appl Environ Microb  2005; 71: 1202– 9. Google Scholar CrossRef Search ADS   Smith RL, Duff JH. Denitrification in a sand and gravel aquifer. Appl Environ Microb  1988; 54: 1071– 8. Sun W, Cupples AM. Diversity of five anaerobic toluene-degrading microbial communities investigated using stable isotope probing. Appl Environ Microb  2012; 78: 972– 80. Google Scholar CrossRef Search ADS   Tamura K, Stecher G, Peterson D et al.   MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol  2013; 30: 2725– 9. Google Scholar CrossRef Search ADS PubMed  Tarlera S, Denner EBM. Sterolibacterium denitrificans gen. nov., sp. nov., a novel cholesterol-oxidizing, denitrifying member of the β-Proteobacteria. Int J Syst Evol Micr  2003; 53: 1085– 91. Google Scholar CrossRef Search ADS   Throbäck IN, Enwall K, Javis A et al. Reassessing PCR primers targeting nirS, nirK, and nosZ genes for community surveys of denitrifying bacteria with DGGE. FEMS Microbiol Ecol  2004; 49: 401– 17. Google Scholar CrossRef Search ADS PubMed  Tiedje JM. Ecology of denitrification and dissimilatory nitrate reduction to ammonium. In: Zehnder AJB (ed). Biology of Anaerobic Microorganisms . New York, USA: John Wiley and Sons, 1988, 179– 244. Torrentó C, Urmeneta J, Otero N et al. Enhanced denitrification in groundwater and sediments from a nitrate-contaminated aquifer after addition of pyrite. Chem Geol  2011; 287: 90– 101. Google Scholar CrossRef Search ADS   Vishnivetskaya TA, Brandt CC, Madden AS et al. Microbial community changes in response to ethanol or methanol amendments for U(VI) reduction. App Environ Microb  2010; 76: 5728– 35. Google Scholar CrossRef Search ADS   Wakelin SA, Nelson PN, Armour JD et al. Bacterial community structure and denitrifier (nir-gene) abundance in soil water and groundwater beneath agricultural land in tropical North Queensland, Australia. Soil Res  2011; 49: 65– 76. Google Scholar CrossRef Search ADS   Wang Q, Garrity GM, Tiedje JM et al. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microb  2007; 73: 5261– 7. Google Scholar CrossRef Search ADS   Wang Q, Quensen JFIII, Fish JA et al. Ecological patterns of nifH genes in four terrestrial climatic zones explored with targeted metagenomics using FrameBot, a new informatics tool. mBio  2013, DOI: 10.1128/mBio.00592-13. Wang Z, Zhang XX, Lu X et al. Abundance and diversity of bacterial nitrifiers and denitrifiers and their functional genes in tannery wastewater treatment plants revealed by high-throughput sequencing. PLoS One  2014, DOI: 10.1371/journal.pone.0113603. Weymann D, Well R, Flessa H et al. Groundwater N2O emission factors of nitrate-contaminated aquifers as derived from denitrification progress and N2O accumulation. Biogeosciences  2008; 5: 1215– 26. Google Scholar CrossRef Search ADS   Whiteley M, Ott JR, Weaver EA et al. Effects of community composition and growth rate on aquifer biofilm bacteria and their susceptibility to betadine disinfection. Environ Microbiol  2001; 3: 43– 52. Google Scholar CrossRef Search ADS PubMed  Yoshida M, Ishii S, Fujii D et al. Identification of active denitrifiers in rice paddy soil by DNA- and RNA-based analyses. Microbes Environ  2012; 27: 456– 61. Google Scholar CrossRef Search ADS PubMed  Yu Z, Yang J, Liu L. Denitrifier community in the oxygen minimum zone of a subtropical deep reservoir. PLoS One  2014, DOI: 10.1371/journal.pone.0092055. Zeng X, Hosono T, Ohta H et al. Comparison of microbial communities inside and outside of a denitrification hotspot in confined groundwater. Int Biodeter Biodegr  2016; 114: 104– 9. Google Scholar CrossRef Search ADS   © FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Microbiology Ecology Oxford University Press

Identification of active denitrifiers by DNA-stable isotope probing and amplicon sequencing reveals Betaproteobacteria as responsible for attenuation of nitrate contamination in a low impacted aquifer

Loading next page...
 
/lp/ou_press/identification-of-active-denitrifiers-by-dna-stable-isotope-probing-YKk6AIBBCd
Publisher
Blackwell
Copyright
© FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0168-6496
eISSN
1574-6941
D.O.I.
10.1093/femsec/fix181
Publisher site
See Article on Publisher Site

Abstract

Abstract Groundwater reservoirs constitute important freshwater resources. However, these ecosystems are highly vulnerable to contamination and have to rely on the resident microbiota to attenuate the impact of this contamination. Nitrate is one of the main contaminants found in groundwater, and denitrification is the main process that removes the compound. In this study, the response to nutrient load on indigenous microbial communities in groundwater from a low impacted aquifer in Uruguay was evaluated. Denitrification rates were measured in groundwater samples from three different sites with nitrate, acetate and pyrite amendments. Results showed that denitrification is feasible under in situ nitrate and electron donor concentrations, although the lack of readily available organic energy source would limit the attenuation of higher nitrate concentrations. DNA-stable isotope probing, combined with amplicon sequencing of 16S rRNA, nirS and nirK genes, was used to identify the active denitrifiers. Members of the phylum Betaproteobacteria were the dominant denitrifiers in two of three sites, with different families being observed; members of the genus Vogesella (Neisseriaceae) were key denitrifiers at one site, while the genera Dechloromonas (Rhodocyclaceae) and Comamonas (Comamonadaceae) were the main denitrifiers detected at the other sites. denitrification, groundwater, stable isotope probing, nirS, nirK, Betaproteobacteria INTRODUCTION Aquifers are key freshwater resources and their water quality represents an important environmental issue worldwide. Domestic, industrial and agricultural activities rely on the quality of groundwater, which is strongly linked to the geochemical processes modulated by the activity and diversity of indigenous microbiota (Griebler and Lueders 2009). Nitrate (NO3−) is one of the key contaminants that affect groundwater quality (Rivett et al.2008). Unlike other pollutants that usually reach the subsurface by defined, well-localised sources, nitrate load in aquifers occurs continuously and in a diffusive way, making both contamination control and removal difficult. Thus, natural attenuation will be a key process in the regulation of nitrate concentration. Denitrification is the metabolic process that consumes the largest amount of nitrate in subsurface (Rivett et al.2008; Weymann et al.2008). Despite its importance, factors that limit denitrification in situ, such as geochemical conditions, remain unclear, especially for aquifers with lower pollution levels (Rivett et al.2008; Eschenbach, Well and Walther 2014). During complete denitrification, nitrate is used as a terminal electron acceptor, and it is reduced to molecular nitrogen (N2) after passing through several reduction states. The reduction of nitrite (NO2−) to nitric oxide (NO) is considered a key step for denitrifiers, since the nitrogen goes from a dissolved anion state to gaseous state and can be easily removed from the system (Tiedje 1988). Several geochemical factors are involved in the regulation of the process, such as electron donor composition and availability, oxygen concentration, pH, temperature, microbial acclimation, salinity, presence of inhibitory substances, although electron donor and oxygen concentration are considered the main limiting factors (Smith and Duff 1988; Santoro, Boehm and Francis 2006; Rivett et al.2008). The variety of electron donors used for denitrification is broad and both organotrophic and lithotrophic processes occur in aquifers (Smith and Duff 1988; Rivett et al.2008; Weymann et al.2008; Torrentó et al.2011). Denitrification is a metabolic trait that is widespread among bacteria, archaea and fungi (Shoun et al.1992; Philippot, Hallin and Schloter 2007). Therefore, analysis of ribosomal RNA genes alone may not be suitable for assessing denitrifier diversity. The use of nitrite-reductase encoding genes, nirS and nirK, as biomarkers in combination with ribosomal genes allow the assessment of both functional diversity and the phylogenetic affiliation of denitrifiers. Stable isotope probing (SIP) of nucleic acids and fatty acids is a technique that has been used since 2000 for the identification of active populations within a microbial community in a specific metabolic process (Neufeld et al.2007b). It has enabled the study of a wide diversity of processes, such as syntrophic interactions for propionate oxidation (Lueders, Pommerenke and Friedrich 2004), methylated compounds assimilation (Neufeld et al.2007a; Antony et al.2010) and identification of acetate-utilising microbes (Osaka et al.2006; Schwarz et al.2007) and pollutants degraders (Singleton et al.2005; Sun and Cupples 2012). Denitrification has also been studied using SIP technique coupled with the analysis of both ribosomal and denitrification genes, with the utilisation of labelled carbon compounds such as acetate, succinate and/or methanolas substrates for nitrate reduction in coastal sediments, paddy soil and activated sludge (Maneesha et al.2004; Ginige, Keller and Blackall 2005; Osaka et al.2006, 2008; Saito et al.2008). However, reports on the use of SIP (both DNA and SIP coupled to Fatty Acids Analysis) to study denitrification in aquifers are scarce and are more frequently linked to the degradation of pollutants such as aromatic or complex compounds (Pelz et al.2001; Fischer, Manefield and Bombach 2016). Our study site, Raigón aquifer, is a sand and gravel groundwater system located in south-west Uruguay. Nitrate levels in this groundwater reservoir will eventually increase due to nearby agricultural and industrial activities. A previous study evaluated the potential for denitrification by groundwater microbial communities from the Raigón aquifer and characterised more than 50 denitrifiers isolated from several sampled wells (Bellini et al.2013). This study describes the use of DNA-SIP combined with amplicon sequencing of 16S rRNA and nirS and nirK genes to determine active indigenous denitrifiers and assess their capacity to remove nitrate from this groundwater system in a scenario of an increasing nutrients load. MATERIALS AND METHODS Sample collection and physicochemical parameters determination Groundwater samples were obtained in July 2012 from three water supply wells (C, D and H) at Raigón aquifer, as previously described by Bellini et al. (2013). Briefly, groundwater samples were collected from water supply wells after three bottle volumes (5 L) were pumped and a constant value of dissolved oxygen was reached. Twenty-five litres of groundwater were collected from each site. Groundwater was transported at 4°C, and was processed within 24 h after sampling for all experiments. Groundwater temperature, pH and dissolved oxygen were measured at the site. The sampling wells lie along a 50-km stretch and have depths that lie between 37 and 50 m. The three selected wells cover the range of nitrate (Table 1) concentrations detected along the aquifer, and they are representative of various human activities that occur on the surface of the aquifer. In the surroundings of site C there is an agricultural stablishment where dairy milking and farming activities occur, while site D is located in a potato cleaning and processing plant. Site H is located in a small town, which relies on septic tanks for sewage treatment. Table 1. Groundwater physicochemical parameters at sampling sites. Site  T (°C)  pH  NO3− (mg L−1/mmol L−1)  Total Fe (μg L−1)  SO42− (mg L−1)  O2 (mg L−1)  C  18.6  7.7  33.6/0.54  <17  49.0  7.4  D  18.0  7.1  10.3/0.17  29  33.2  3.9  H  19.2  7.4  24.9/0.40  65  28.8  6.4  Site  T (°C)  pH  NO3− (mg L−1/mmol L−1)  Total Fe (μg L−1)  SO42− (mg L−1)  O2 (mg L−1)  C  18.6  7.7  33.6/0.54  <17  49.0  7.4  D  18.0  7.1  10.3/0.17  29  33.2  3.9  H  19.2  7.4  24.9/0.40  65  28.8  6.4  View Large Nitrate, nitrite, total iron and sulphate were analysed according to the American Public Health Association ‘Standard Methods for the examination of Water and Wastewater’ methods (APHA 1998), with 4500-NO3- B, 4500-NO2- B, 3500-Fe-B Phenanthroline Method and 4500 methods, respectively. Total organic carbon (TOC) was analysed with a Shimadzu TOC-VSCN analyser and the limit of quantification was 1.1 mg L−1. Denitrification potential of groundwater samples The potential for nitrate attenuation of indigenous microbial community was tested by measuring denitrification rates in groundwater samples with different amendments and compared to endogenous rates (i.e. without amendment). Rates were measured in groundwater obtained from wells C, D and H with the acetylene blockage technique (Mahne and Tiedje 1995). Triplicate sterile 120-mL flasks were flushed with filtered O2-free N2 while being filled aseptically with 40 mL of groundwater. Nitrate plus either acetate or pyrite (crushed and sieved to particles of <65 μm) were added to compare organotrophic and lithotrophic denitrification capacity, respectively. In addition, flasks with only nitrate amendment (containing between 3 and 12 times the nitrate concentration measured in the wells), non-amended flasks and acetate or pyrite-amended sterile controls (autoclaved) were also set up for each well. Acetate was chosen as the carbon amendment since it is a non-fermentable substrate and other parallel processes besides denitrification would be unlikely, due to incubation conditions. In addition, most of the previously isolated bacteria from the Raigón aquifer were able to use it as a carbon and electron source for denitrification (Bellini et al.2013). Acetate and pyrite were added according to a complete NO3− reduction stoichiometry. Every set of triplicate flasks contained the following (final concentration): phosphate buffer (1 mmol L−1, pH 7.1) and either 2 mmol L−1 KNO3 and 1.5 mmol L−1 ammonium acetate (nitrate plus acetate amended), 2 mmol L−1 KNO3, 1200 mg L−1 pyrite and 0.5 mmol L−1 Na2CO3 (nitrate plus pyrite amended) or 2 mmol L−1 KNO3 (nitrate amended). Acetylene was added (10% of the headspace) to every flask, and flasks were incubated in the dark with continuous shaking at 20°C. Samples from the headspace of every flask were removed at several time points for N2O measurement by gas chromatography (Shimadzu GC-2014 gas chromatograph); gas chromatography conditions and calculations of denitrification rates were done as described in Bellini et al. (2013). Nitrate and nitrite concentrations were also measured during the incubations by HPLC as described by Tarlera and Denner (2003). To test if denitrification rates from different treatments or wells were significantly different, a squared root transformation was applied to rates and then compared with an analysis of variance (ANOVA) using P ≤ 0.05 and Tukey correction with InfoStat software (Di Rienzo et al.2009). All transformed data satisfied the ANOVA assumptions. Assessment of groundwater basal communities: 16S rRNA gene amplicon sequencing To estimate the structure of the naturally occurring bacterial communities present in the sampled groundwater, 20 L of groundwater were filtered through a 0.22-μm pore cellulose acetate membrane. DNA was then extracted from biomass using the PowerWater DNA Isolation Kit (MoBio Laboratories, Inc., CA, USA). For amplicon libraries, the 16S rRNA gene was amplified with PCR primers 563F and 802R (Cole et al.2009). PCR amplicons were purified using Agencourt AMPure XP beads (Agencourt Bioscience Corporation, MA, USA) and sequenced at the INDEAR sequencing facilities (Rosario, Argentina) on a Genome Sequencer FLX (454-Roche Applied Sciences) according to the manufacturer's instructions. SIP incubations, gradient centrifugation and DNA extraction SIP incubations were set up in 150-mL microcosms. Two replicates, named A and B, were prepared for each sampled site and treatment; 70 mL of groundwater were aseptically placed in sterile flasks containing the following (final concentration): phosphate buffer (1 mmol L−1) and KNO3 (2 mmol L−1) while flushed with O2-free N2. Organotrophic treatment flasks were amended with 1.5 mmol L−1 (final concentration) of 13C-labelled sodium acetate (Na13CH3COO); granulated pyrite (particle size < 65 μm) and Na213CO3 were added to lithotrophic treatment flasks to a final concentration of 1200 mg L−1 and 0.5 mmol L−1, respectively. In order to confirm isotopic enrichment of DNA in 13C incubations, 12C unlabelled controls were set up for every treatment. Controls with only acetate were not included because other parallel processes (fermentation, heterotrophic anaerobic respiration, lithotrophic anaerobic denitrification) are unlikely or marginal in the incubation conditions. In addition, determination of the denitrification rates and the SIP incubations showed that the amounts of acetate and nitrate consumed were stochiometric corresponding to a denitrification process (see below). Incubations were carried out in the dark at 20°C with continuous shaking. Nitrate and nitrite were monitored in every flask as described before. Acetate consumption at the end of the incubation was determined by HPLC, with an Aminex-87H column and 5 mmol L−1 sulphuric acid as the mobile phase. Runs were carried out at 35°C and UV detection at 210 nm was used. Biomass was harvested by centrifugation when at least 90% of added nitrate and the stoichiometric amount of acetate had been consumed. Pellets were frozen at –20°C until DNA extraction. DNA was extracted from cell pellets using a modification of the protocol described in Neufeld et al. (2007a). Briefly, 200 μL of SET buffer (40 mmol L−1 ethylenediamine tetra-acetic acid, 50 mmol L−1 Tris-HCl pH 9.0 and 0.75 mmol L−1 sucrose) and 22.5 μL of freshly prepared lysozyme solution (9 mg mL−1) were added to the pellets. After incubation at 37°C for 30 min, 25 μL of 10% (w/v) SDS and 7 μL of a fresh solution of proteinase K (20 mg mL−1) were added and the tubes were incubated at 55°C for 2 h. Following the addition of 125 μL of SET buffer, the lysates were transferred to a phase-lock tube (Eppendorf, Hamburg, Germany) and two extractions with 375 μL of phenol:cholorform:isoamylalcohol (25:24:1) were performed. DNA in the aqueous phase was precipitated with 2 μL of glycogen (Roche, Basel, Switzerland), 170 μL of a 7.5 mol L−1 ammonium acetate solution and 700 μL of 95% ethanol, overnight at –20°C. Tubes were centrifuged at 21 000g, and the pellets were washed twice with 500 μL 80% (v/v) ethanol. After drying, DNA was re-suspended in 50 μL of sterile water. DNA was run on an agarose gel, and concentration and quality were analysed with a NanoDrop 2000 spectrophotometer. Ultracentrifugation, fractionation and precipitation of DNA were carried out according to Neufeld et al. (2007b). Twelve or thirteen fractions were collected from 13C duplicates and from the 12C control flask from each well. The buoyant density of each fraction was determined indirectly by measuring the refractive index (nD-TC) with a digital refractometer (Reichert AR200, Reichert Inc., NY, USA). Those fractions with the highest densities for which a positive 16S rRNA gene amplification was obtained (see later) were selected for further analysis and designated heavy fractions (13C-DNA, densities between 1.7211 and 1.7269 g mL−1). Fractions with densities of 1.7035–1.7094 g mL−1 were designated and analysed as light fractions (12C-DNA). SIP incubations follow-up: denaturing gradient gel electrophoresis analysis of 16S rRNA genes In order to confirm that heavy fractions from 13C incubations showed a different community pattern than light fractions from 13C incubations and than heavy fractions from 12C incubations, denaturing gradient gel electrophoresis (DGGE) analysis of 16S rRNA genes was carried out. In addition, DGGE was performed to compare the structure of communities from replicated SIP incubations on non-fractionated DNA from each replicate and well.16S rRNA gene PCR primers for Bacteria were 341F-GC(5΄-CGCCCGCCGCGCGCGG CGGGCGGGGCGGGGGCACGGGGGGCCTACGGGAGGCAGCAG-3΄) and 907R (5΄-CCGTCAATTCMTTTRAGTTT-3΄) (Muyzer, De Waal and Uitterlinden 1993); PCR reactions were carried out in a total volume of 50 μL and contained 2 mmol L−1 MgCl2, 0.2 mmol L−1 dNTP, 0.2 μmol L−1 of each primer, 2.5 U of DreamTaq DNA Polymerase (Fermentas, Burlington, Ontario, Canada), 5 μL 10X Taq buffer, 0.07% BSA and 30 ng DNA. Cycling conditions included an initial step of 5 min at 94°C followed by 30 cycles of 94°C for 30 s, 55°C for 30 s and 72°C for 45 s, with a final extension step of 72°C for 8 min. PCR products were checked for size and purity on agarose gels, and the concentration was estimated by comparison to a 1-kb ladder (Fermentas). Approximately 300 ng of PCR product were loaded in each lane of an 8% (w/v) polyacrylamide gel with a gradient of denaturants (urea and deionised formamide) that ranged from 20% to 60%. Electrophoresis was performed using a D-Code System (Bio-Rad, Hercules, CA, USA) on TAE 1X buffer at 60°C for 16 h at 80 V. The gel was stained with SYBR Gold Nucleic Acid Gel Stain (Invitrogen, Paisley, UK) for 1 h and visualised with an imaging device. The predominant bands from heavy fractions were excised and 16S rRNA gene was amplified as previously described, with the exception that the primers used did not have a GC-clamp. Amplicons were sequenced by Sanger method at Source BioScience, Nottingham, UK. Sequences were used in a BLAST search (Altschul et al.1997) and deposited in GenBank with the accession numbers KY548747 to KY548762. The potential role of archaeal communities in denitrification during SIP incubations was also tested. A nested PCR of the 16S rRNA gene specific for Archaea domain (Cunliffe et al.2008) was performed on heavy and light fractions of DNA isolated from every well, and only weak amplifications were observed in heavy fractions from site H. Amplicon sequencing from SIP fractions DNA from heavy and light fractions from 13C incubations was used as a template for triplicate PCR amplification of 16S rRNA gene with 503F (5΄-GTGCCAGCMGCNGCGG-3΄) and 1100R (5΄-GGGTTNCGNTCGTTR-3΄) primers. Reaction mixtures were performed in a total volume of 50 μL and contained 2 mmol L−1 MgCl2, 0.2 mmol L−1 dNTP, 0.3 μmol L−1 of each primer, 1.25 U of DreamTaq DNA Polymerase (Fermentas) and 5 μL 10X Taq buffer. Cycling conditions included an initial step of 5 min at 94°C followed by 30 cycles of 94°C for 30 s, 55°C for 30 s and 72°C for 45 s, with a final extension step of 72°C for 10 min. DNA from heavy and light fractions from replicates A was also used as a template for triplicate PCR amplification of nirS and nirK genes. The primers and conditions used were as follows: nirS cd3aF/R3Cd (Throbäck et al.2004) and nirK F1aCu/R3Cu (Hallin and Lindgren 1999); reactions mixtures were prepared in a final volume of 50 μL and contained 2 mmol L−1 MgCl2, 0.2 mmol L−1 dNTP, 1.25 U of DreamTaq DNA Polymerase (Fermentas), 5 μL 10X Taq buffer, 0.2 μmol L−1 of each primer for nirS gene and 0.5 μmol L−1 for nirK and 30 ng DNA. Cycling conditions included an initial step of 94°C for 4 min followed by 30 cycles of 30 s at 94°C, 1 min at 57°C and 1 min at 72°C with a final extension step of 10 min at 72°C. PCR products were checked for size and purity on agarose gels. All PCR products were visualised in an agarose gel, replicates were pooled and purified with Agencourt AMPure XP beads (Agencourt Bioscience Corporation) and checked on an agarose gel. PCR products were sent to Mr DNA Molecular Research (Lubbock, TX, USA) for barcoded sequencing where a second 5-cycle PCR was performed to add barcodes and sequencing keys. Following PCR, all amplicons from different samples were mixed in equal concentrations and purified using Agencourt AMPure XP beads (Agencourt Bioscience Corporation). Samples were sequenced using a Roche 454 FLX Titanium instrument and reagents according to the manufacturer's instructions. Bioinformatic analysis of amplicon sequences 16S rRNA data were analysed using the pipeline available through the QIIME platform, version 1.7 (Caporaso et al.2010b). Libraries were split using default parameters except that the number of ambiguous bases allowed was one instead of six and primer mismatches allowed was zero. Chimera checking (de novo and reference-based) and operational taxonomical units (OTU) picking (0.97 cut-off) were performed with usearch61 (Edgar 2010), and taxonomy assignment was done using the Greengenes 13_5 database with the RDP Classifier (Wang et al.2007). The representative set was aligned with Pynast software (Caporaso et al.2010a) against the Greengenes core reference alignment (DeSantis et al.2006). Finally, low-abundance OTUs (less than 1% of the dataset) and those that failed to align were filtered from the OTU table. Similar filtering options were used for naturally occurring (basal) communities’ data, except only singletons and doubletons were removed. Alpha diversity indexes were calculated within QIIME. Rarefaction analysis of the libraries was performed using the free Analytic Rarefaction 1.3 software (http://strata.uga.edu/software/index.html). Data from sequencing of denitrification genes were initially split into libraries, followed by a de novo chimera checking with usearch61 (Edgar 2010) within the QIIME platform. The split and filtered libraries were then screened for frame shifts using the RDP FunGene (Fish et al.2013) FrameBot tool (Wang et al.2013) with a length cut-off of 80 amino acids and an identity cut-off of 0.4. Frame-corrected nucleotide sequences were then clustered (0.03 distance) using uclust (Edgar 2010) within QIIME and an OTU table was constructed. Low-abundance OTUs (less that 1% of the dataset) were filtered from the OTU table. Rarefaction analysis of the libraries was performed as described previously. Representative sequences were found for nirK and nirS libraries and compared with the EMBL/GenBank database using the NCBI Blast algorithm (Altschul et al.1997). Translated sequences were aligned with MUSCLE at amino acid level. The resulting alignments were checked via comparison with aligned HMM training sequences obtained from FunGene repository (Functional gene pipeline and repository, http://fungene.cme.msu.edu/index.spr) and corrected manually when necessary. Phylogenetic trees were constructed with nucleotide sequences based on the neighbour-joining (NJ) method and maximum likelihood (ML) algorithms to confirm overall topology using MEGA6 (Tamura et al.2013) software. A bootstrap confidence analysis was performed with 1000 replicates. A total of 303 informative positions were taken into account for the analysis of the nirS gene, while 272 were used for nirK gene. All datasets have been deposited at MG RAST database under Project ‘Nitrate contamination in a low impacted aquifer’: Clean_X_16S_heavy/light corresponds to 16S rRNA amplicon libraries from SIP heavy/light fraction from site X (i.e. C, D or H); Clean_X_nirS/nirK_heavy corresponds to amplicon libraries from nirS/nirK SIP heavy fractions from site X (ie C, D or H); and Clean_X_16S corresponds to 16S rRNA amplicon libraries from basal groundwater communities from sites C, D or H. Abundance of nirS and nirK genes as determined in DNA-SIP incubations To determine the abundance of nirS and nirK genes in DNA samples from DNA-SIP incubations, qPCR was performed. PCR products used for standard curves in qPCR experiments were obtained from the isolated denitrifiers AR45 (Pseudomonas stutzeri) and AR11 (Achromobacter sp.) for nirS and nirK, respectively (Bellini et al.2013). PCR products were cloned using the TOPO TA cloning kit (Invitrogen®) following the manufacturer's instructions. Clones were subjected to PCR using primers T3–T7 provided with the cloning kit. Correct fragment length and identity of the PCR product were verified by electrophoresis on a 1.5% agarose gel and sequenced at the Macrogen Sequencing Service, Korea, using the T7 primer. Plasmid was extracted using the Purelink Quick Plasmid miniPrep kit (Invitrogen®), and the size of the plasmid fragment was verified by electrophoresis in a 1.5% agarose gel. A triplicate amplification of the cloned fragments with T3 and T7 primers was performed. PCR products were pooled, diluted to 90 μL with MilliQ water and purified with MICROCON®100 columns. Following a fragment size checking in an agarose gel, the product was quantified with Qubit dsDNA HS Assay kit in a Qubit 2.0 Fluorometer (Invitrogen®) and the gene copy number was calculated based on the total fragment length (partial nirS or nirK fragment plus cloning vector fragments flanking the insertion point), the molecular weight of the amplicon and Avogadro's number. nirS and nirK quantification on DNA samples from heavy and light fractions from SIP incubations was performed in duplicate for samples and in triplicates for standard curves. One-in-ten and one-in-one hundred sample dilutions were included in the run to test for any inhibition of PCR reactions. No template controls were included in the run. RESULTS Denitrification activity and groundwater chemistry Physicochemical parameters measured in sampled groundwater are shown in Table 1. TOC and nitrite concentrations were below the quantification limit (1.1 and 0.02 mg L−1, respectively) in samples for all the wells. Denitrification rates in groundwater samples from sites C, D and H were estimated under four different conditions (endogenous, nitrate amended, nitrate plus pyrite amended and nitrate plus acetate amended). Results were analysed statistically to compare the endogenous denitrification potential and the effects of amendments within each site (Table 2). According to the endogenous and nitrate amended rates, denitrification is feasible at all three sites, even in a scenario of increasing nitrate concentration. At sites C and H, nitrate plus acetate amendments significantly stimulated denitrification compared to the respective endogenous rates. Conversely, neither nitrate nor nitrate plus pyrite additions changed denitrification rates significantly at any sample site. Samples revealed differential response to the added nutrients, as denitrification rates with nitrate addition showed significantly higher denitrification rates in well H compared to other wells. Table 2. Mean denitrification rates of indigenous bacterial communities per well (μmol N2O L−1 h–1).     Amendment  Site  Endogenous  Nitrate  Nitrate + pyrite  Nitrate + acetate  C  0.14a,A  0.22a,A  0.91a,AB  4.43b,B  D  0.07a,A  0.14a,A  0.07a,A  0.65a,A  H  1.44a,A  3.09ab,B  2.02a,B  7.42b,B      Amendment  Site  Endogenous  Nitrate  Nitrate + pyrite  Nitrate + acetate  C  0.14a,A  0.22a,A  0.91a,AB  4.43b,B  D  0.07a,A  0.14a,A  0.07a,A  0.65a,A  H  1.44a,A  3.09ab,B  2.02a,B  7.42b,B  P = 0.05, lowercase letters indicate significant differences between rates within one site (lines), uppercase letters indicate differences between sites (columns). View Large Bacterial community structure in basal groundwater samples A total of 19 844 16S rRNA gene sequences were obtained from the three groundwater samples before incubation. This number was reduced to 13 447 after quality checking and removal of chimeras. Sequences were grouped in a final OTU table composed of 936 OTUs and 11 154 reads, which were distributed as 3343, 3094 and 4717 reads for sites C, D and H, respectively. About 97% of the sequences in the final dataset were between 220 and 240 bp long. The rarefaction curves (Fig. S1, Supporting Information) levelled off at the sequenced depth for all the samples, indicating adequate coverage for each site. As shown in Fig. S2, Supporting Information, Proteobacteria was the major phylum in all three samples. The distribution within this phylum however varied between sampling sites. In site C, Alpha and Gammaproteobacteria were the dominant classes, while in sites D and H, sequences affiliated to Betaproteobacteria dominated the bacterial communities. Other phyla that were present in high proportions were Actinobacteria at site C (20%), Nitrospirae at site D (22%), Verrucomicrobia (11%) and Chlamydiae (10%) at sample H. Analysis of SIP incubations by DGGE In order to identify the denitrifiers that would respond to an input of nitrate and an energy source to the groundwater, replicate DNA-SIP incubations were set up with 2 mmol L−1 of nitrate and either 1.5 mmol L−1 of acetate or pyrite plus carbonate (final concentration 1200 mg L−1 and 0.5 mmol L−1, respectively). These were monitored over time by measuring nitrate consumption (Fig. S3, Supporting Information); acetate consumption was measured in acetate amended treatments at the end of incubation. Flasks that were amended with pyrite did not show more than 20% decrease in nitrate concentration over 4 weeks of incubation, and cells were not harvested from these flasks. This was similar to the observations obtained from the denitrification rates experiments, where no increase in the rates was observed with pyrite amendment. Acetate-amended flasks responded to the substrates and consumed the added nitrate and acetate, although in a time course that differed among sites. A shorter incubation time was observed for sample H (7 days), while 10 and 14 days were required for samples C and D to consume at least 90% of the added nitrate, and the stoichiometric amount of acetate, respectively. Bacterial community composition in non-fractionated DNA and in DNA from fractions from 13C and 12C acetate incubations obtained after gradient centrifugation and fractionation were analysed by DGGE fingerprinting of 16S rRNA genes. Biological replicates of 13C incubations and 12C controls displayed reasonable similarity among non-fractionated profiles (Fig. S4, Supporting Information); thus, replicate A was selected for further work. Heavy DNA fractions from 13C incubations in wells C and D displayed different 16S rRNA gene profiles from those obtained with light fractions from 13C incubations, with the presence of specific bands only in the heavy fractions (Fig. 1). This suggests that selective 13C acetate assimilation had occurred after nitrate addition in these wells. Moreover, these specific bands were intense in light fractions from 12C incubations, and faint, or mostly absent, in the heavy fractions from these 12C controls. Taken together, these results reveal successful 13C labelling of DNA in specific microorganisms after acetate consumption and nitrate reduction in wells C and D. Heavy fractions from 13C incubations from site H also showed different profiles from light fractions, suggesting specific labelling. However, the main band observed in heavy fractions from 13C incubation was not absent or faint in heavy fractions from 12C incubations. Therefore, the results obtained for well H do not provide enough evidence to confirm that the main band detected in heavy fractions from 13C incubations correspond to the active denitrifiers. Figure 1. View largeDownload slide Denaturing gradient gel electrophoresis analysis of 16s rRNA gene fragments in unfractionated and in heavy and light fractions from 13C- and 12C-acetate SIP incubations. Density from each fraction is shown (g mL−1). Predominant bands identification is shown (Std: DGGE standard). Figure 1. View largeDownload slide Denaturing gradient gel electrophoresis analysis of 16s rRNA gene fragments in unfractionated and in heavy and light fractions from 13C- and 12C-acetate SIP incubations. Density from each fraction is shown (g mL−1). Predominant bands identification is shown (Std: DGGE standard). Figure 1. View largeDownload slide (Continued). Figure 1. View largeDownload slide (Continued). Identification of active denitrifiers by DNA-SIP with acetate To facilitate preliminary identification of the active denitrifiers responding to a nutrient input, selected DGGE bands were analysed. 16S rRNA gene analysis of predominant bands in heavy fractions from 13C incubations by excision from gels and sequencing (Fig. 1) revealed that members of the genus Vogesella are the main active bacteria at site C, along with Comamonas, Hydrogenophaga and Duganella (all members of Comamonadaceae family) though in lower proportions. In site H on the other hand, bands that belong to Dechloromonas, Comamonas and Ensifer genera were identified in the heavy fraction, while bands excised from site D heavy DNA fractions were affiliated exclusively to the genus Dechloromonas. Datasets obtained after amplicon sequencing of 16 rRNA genes of heavy and light DNA fractions from 13C incubations recovered from acetate SIP incubations with groundwater samples C, D and H contained 99 581 raw reads. This dataset was reduced to 72 109 sequences after quality control and removal of chimeras, and sequences were grouped into 982 OTUs at 97% nucleotide similarity. Removal of low-abundance OTUs and the sequences that failed to align resulted in 66 932 16S rRNA sequences distributed over 18 different OTUs. Ninety percent of DNA sequences in the final 16S rRNA gene dataset were between 450 and 530 bp long. A rarefaction analysis showed that at a depth of 2600 sequences per sample, the number of OTUs reached a plateau for the six samples, indicating adequate sequencing coverage (Fig. S1, Supporting Information). The distribution of 16S rRNA gene sequences retrieved from heavy and light fractions of 13C incubations, and their taxonomic classification is shown in Table 3. The OTUs identified in heavy fractions showed consistent results with DGGE bands sequencing. In most cases, OTUs that arose only in heavy fractions were identified, while other OTUs were considerably enriched from light to heavy fraction. These OTUs were considered as the active consumers of the labelled acetate in the denitrifying conditions imposed, and therefore identified as active denitrifiers. Site C harboured several OTUs that occurred exclusively in the heavy DNA fraction. The predominant sequences belonged to the family Neisseriaceae, which comprised up to 42% of the 16S rRNA gene library at this site. Moreover, the genus Vogesella was the predominant component within this family (36% of the 16S rRNA gene library at site C). Members of the family Comamonadaceae were also abundant (24%) in the heavy DNA fraction at site C. In heavy DNA from site D, the dominant 16S rRNA genes were from the family Rhodocyclaceae (47%) with members of the genus Dechloromonas being a major component (45% of this 16S rRNA gene library). The most abundant OTU in site H was a member of the genus Comamonas (79% of the 16S rRNA gene library). An OTU affiliated with the Rhizobiaceae was also present although in a much lower proportion (7%). Table 3. Taxonomic composition and relative abundance (%) of bacterial communities in 13C SIP incubations for the heavy and light fractions in the three studied sites.   C  D  H  Taxon  Heavy  Light  Heavy  Light  Heavy  Light  Alphaproteobacteria               Rhizobiaceae  3  0  2  0  7  2  Betaproteobacteria               Burkholderiales                Comamonadaceae                 Aquabacterium  0  0  0  0  0  2     Comamonas  9  0  0  0  79  39     Hydrogenophaga  1  0  0  0  2  8     Other Comamonadaceae  14  0  2  0  9  15  Neisseriales               Neisseriaceae                Vogesella  36  0  0  0  0  0    Other Neisseriaceae  6  0  0  0  0  0  Rhodocyclales               Rhodocyclaceae                Dechloromonas  1  0  45  0  1  4    KD1-23  0  0  1  0  0  0    Other Rhodocyclaceae  0  0  1  0  1  2  Other Betaproteobacteria  7  0  0  0  0  0  Gammaproteobacteria               Aeromonadaceae  2  0  0  0  0  0  Pseudomonadales                 Acinetobacter  21  96  0  0  0  0     Pseudomonas  0  0  4  0  0  0  Flavobacteriia               Flavobacteriaceae                Flavobacterium  0  4  0  0  0  0   Weeksellaceae                Chryseobacterium  0  0  2  2  0  0    Other Weeksellaceae  0  0  43  98  0  0  Bacilli                Bacillus  0  0  0  0  1  28    C  D  H  Taxon  Heavy  Light  Heavy  Light  Heavy  Light  Alphaproteobacteria               Rhizobiaceae  3  0  2  0  7  2  Betaproteobacteria               Burkholderiales                Comamonadaceae                 Aquabacterium  0  0  0  0  0  2     Comamonas  9  0  0  0  79  39     Hydrogenophaga  1  0  0  0  2  8     Other Comamonadaceae  14  0  2  0  9  15  Neisseriales               Neisseriaceae                Vogesella  36  0  0  0  0  0    Other Neisseriaceae  6  0  0  0  0  0  Rhodocyclales               Rhodocyclaceae                Dechloromonas  1  0  45  0  1  4    KD1-23  0  0  1  0  0  0    Other Rhodocyclaceae  0  0  1  0  1  2  Other Betaproteobacteria  7  0  0  0  0  0  Gammaproteobacteria               Aeromonadaceae  2  0  0  0  0  0  Pseudomonadales                 Acinetobacter  21  96  0  0  0  0     Pseudomonas  0  0  4  0  0  0  Flavobacteriia               Flavobacteriaceae                Flavobacterium  0  4  0  0  0  0   Weeksellaceae                Chryseobacterium  0  0  2  2  0  0    Other Weeksellaceae  0  0  43  98  0  0  Bacilli                Bacillus  0  0  0  0  1  28  View Large Abundance and functional diversity of responding denitrifiers PCR products were obtained with gene-specific primers targeting nirS and nirK in heavy DNA fractions from 13C incubations. Interestingly, no amplification of any of the denitrification genes was observed in light DNA fractions, indicating that the only denitrifiers that were enriched during SIP incubations were able to use the labelled acetate. The abundance of both nirS and nirK genes was also determined in DNA from heavy fractions from 13C incubations and the relative abundance (nirS copies μL−1/nirK copies μL−1) was calculated. nirS gene copy numbers in the heavy fractions were 1.2–3.6 × 105 copies μL−1 of DNA, whereas nirK abundance was 1.1–1.5 × 104 copies μL−1. The abundance of nirS was at least 10 times higher than nirK for every site, which suggests that the microbial communities responding to nitrate and acetate input in the three sites are dominated by nirS-containing denitrifiers. Amplicon sequencing of nirS and nirK was performed on DNA from heavy fractions from 13C incubations. Total datasets obtained for nirS contained 16 334 reads, which were reduced to 13 481 (7024 for site C, 2862 for site D and 3595 for site H) after quality control, chimera checking and removal of sequences that failed to align. Ninety-five percent of the sequences in the final set had between 360 and 400 bp and 21 OTUs were defined (at 97% sequence identity). Rarefaction analysis was performed, and the number of OTUs reached a plateau for the three samples for both nir genes. The phylogenetic relatedness among representative nirS and nirK sequences from each OTU with sequences from both uncultivated and extant strains recovered from databases was analysed (Figs 2 and 3). Tree topology showed that gene sequences from this study clustered in different groups that were supported by high bootstrap values and consistently obtained with NJ as well as with ML methods. All the closest nirS sequences from extant bacteria found in databases were affiliated to the class Betaproteobacteria. The majority (89%) of representative nirS sequences obtained from site C belonged to the same OTU and clustered with denitrifier gene sequences found in two Vogesella species (Neisseriaceae) which were isolated from the same aquifer in a previous study (Bellini et al.2013). The remaining OTUs clustered with denitrifier gene sequences obtained from Comamonadaceae isolates (8.5%) and from the genus Dechloromonas (2%). All the nirS sequences present in the heavy DNA fraction from site D were closely related to nirS sequences obtained from isolates of the family Rhodocyclaceae. Most of these sequences (97.5%) showed maximum identity with DNA sequences retrieved directly from the environment, mainly ocean sediments, and with sequences from Dechloromonas genus isolates. The remaining 2.5% grouped closely with nirS sequences retrieved directly from the environment. Most of the nirS sequences in the site H library (55%) were closely affiliated with environmental sequences, obtained mostly from aquatic environments. Thirty-two percent of the sequences clustered with nirS sequences from Rhodocyclaceae isolates, which were distributed among the genera Dechloromonas, Azospira, Azoarcus and Zoogloea. The closest nirS sequence from extant bacteria to the remaining 13% of site H library was nirS from soil isolate I-Bh25-7 (Braker, Schwarz and Conrad 2010), a member of the Comamonadaceae family. Figure 2. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirS gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. Sequences obtained from prior culturing from this aquifer are shown in bold. Figure 2. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirS gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. Sequences obtained from prior culturing from this aquifer are shown in bold. Figure 3. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirK gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. Figure 3. View largeDownload slide Unrooted NJ phylogenetic analysis of partial nucleotide sequences of nirK gene fragments from SIP heavy and light fractions from 13C incubations. Squares, triangles and circles represent sequences from sites C, D and H, respectively. Percentages indicate OTU proportion in sample library. Bootstrap values greater than 70% (1000 replicates) are shown. The scale bar indicates the number of changes per sequence position. In contrast to information on nirS, nirK gene sequences retrieved from heavy fractions from 13C incubations grouped mostly with nirK sequences from cultivated representatives of the Alpha and Gammaproteobacteria. The majority of the nirK OTUs from site C (90%) clustered with environmental clones and with nirK sequences affiliated to Alphaproteobacteria of the genera Sinorhizobuim, Ensifer, Rhizobium, Mesorhizobuim, Bradyrhizobium and Bosea (within the Rhizobiales). Gammaproteobacteria-affiliated nirK sequences represented 10% of the nirK sequences retrieved from site C, with one OTU clustering with nirK of Pseudomonas. All representative nirK sequences from site D were affiliated with nirK from Alphaproteobacteria. Seventy-two percent of the sequences clustered closely to nirK from extant Rhizobium, while the remaining 28% were affiliated with nirK sequences from Bosea. Most (93%) of the nirK sequences from site H clustered in one OTU, most closely related to nirK from members of the genera Ensifer and Sinorhizobium. Other less well-represented OTUs clustered with nirK from Rhizobium (5%), Ochrobactrum and Paracoccus (2%). DISCUSSION Our aim was to evaluate the denitrification potential of indigenous microbial communities from a low impacted aquifer that is likely face an input of nitrate and to identify the microorganisms responsible for the observed activity. The strategy combined kinetic measurements of the denitrification process together with a study of the active microbial communities using DNA-SIP, in order to better predict how denitrifiers in groundwater respond to an increase in nitrate and electron donor concentration. Natural attenuation of dissolved nitrate was feasible in the three sampled sites at in situ nitrate concentrations. However, according to the denitrification rates measured an increase in the nitrate levels without the input of a carbon source would be limited either by abundance or composition of denitrifiers or by insufficient electron donors, since in most of the sampled sites only acetate plus nitrate stimulated denitrification. This suggests that the indigenous denitrifying microbes in water from the Raigón aquifer appear to be adapted to organotrophic denitrification, which is in accordance with previous observations which establish that in general carbon deprivation is the main factor limiting denitrification in aquifers (Rivett et al.2008). However, bacteria capable of using minerals such as pyrite could be present in the aquifer but attached to sediments. Samples reacted differently to the amendments and displayed differences in their basal bacterial communities’ profiles, suggesting intrinsic differences in either the denitrifying population or in geochemical conditions. The results from this study are in agreement with previous observations involving the same aquifer wells (Bellini et al.2013), where a carbon-limited nitrate consumption potential was detected. However, sites C and D did not show an equivalent response to the same amendments seen in the previous study, which could be due to the temporal fluctuation of nutrients and/or denitrification inhibitor concentrations (Lin et al.2012). On the contrary, site H responded similarly to amendments than in the previous study. Considering the activities occurring in the surroundings of each well, groundwater in site H (located in a small town) could be receiving a relatively constant input of nutrients from the surface, while activities on sites D (potato processing plant) and C (dairy farm) would sporadically contribute nutrients to the subsurface. SIP incubations performed on groundwater under the same conditions as the denitrification activity measurements enabled us to identify key denitrifiers responding to increased nitrate contamination of the aquifer. Amplicon sequencing of 16S rRNA genes from 13C-labelled DNA from SIP incubations indicated that Betaproteobacteria play a major role in organotrophic denitrification in two sites examined. Previous molecular and/or cultivation-dependent studies have also shown that Betaproteobacteria are the main denitrifiers in diverse environments such as paddy soil (Saito et al. 2008), wastewater treatment systems (Ginige, Keller and Blackall 2005; Osaka et al.2006; McIlroy et al.2016) as well as aquifers where heterotrophic denitrification occurs (Calderer et al.2014; Zeng et al.2016). Although a high diversity was detected in naturally occurring communities in all three wells, Betaproteobacteria was the predominant class in naturally occurring communities from sites D and H, although it was less represented in site C (Fig. S2, Supporting Information). However, different denitrifying bacteria from this class respond rapidly to nitrate and carbon input in the wells. Members of the genus Vogesella were the predominant denitrifiers in samples from site C. In a previous study, Vogesella was isolated from the same aquifer from wells C and H, and its denitrifying capability with acetate as electron donor was confirmed in the laboratory (Bellini et al.2013). The fact that Vogesella was detected by both molecular and cultivation-based methods, in both samplings, suggests Vogesella is well adapted to this groundwater system and could be relevant in the response to carbon and nitrate input in this site. Interestingly, though Betaproteobacteria have been identified as important denitrifiers in different environments, Neisseriales have scarcely ever been reported as major actors in denitrification processes (Yoshida et al.2012). The ability of Vogesella to denitrify was observed when the genus was first described (Grimes et al.1997). Vogesella has also been shown to be prevalent in other processes that may occur in groundwater such as removal of heavy metals (Vishnivetskaya et al.2010) or improving resistance to disinfectants (Whiteley et al.2001). Together with our findings, this suggests that Vogesella could be relevant not only for denitrification, but also for other environmental processes occurring in groundwater ecosystems. Members of the Comamonadaceae were also identified as active denitrifiers at site C. This is also in agreement with previous findings where Comamonadaceae, along with other bacteria, were found (Wakelin et al.2011) or enriched under denitrifying conditions in groundwater samples (Bellini et al.2013; Calderer et al.2014) and in other systems, such as paddy soils and wastewater treatment (Saito et al.2008; McIlroy et al.2016). In addition, growth and molecular-based studies have shown species of genus Comamonas, a member of Comamonadaceae, to be capable of denitrification (Etchebehere et al.2001; Gumaelius et al.2001) and that they play important roles in denitrification processes in activated sludges (Ginige, Keller and Blackall 2005), paddy soils (Osaka et al.2006) and groundwater systems (Calderer et al.2014; Wang et al.2014). Site D contained only one dominant active denitrifying genus, Dechloromonas, a member of the Rhodocyclaceae. Denitrifying members of this family were isolated from the same well from Raigón aquifer in a previous study (Bellini et al.2013), suggesting that members of this family regularly respond to denitrifying conditions in this site. Rhodocyclaceae have also been depicted as key denitrifiers in systems with a greater load of nitrate and/or organic matter such as activated sludge (Ginige, Keller and Blackall 2005), paddy soil (Osaka et al.2006) and surface freshwater systems (Yu, Yang and Liu 2014). The most abundant taxon detected in site H was genus Comamonas, although according to DGGE results its role as key denitrifier was not confirmed. To further characterise the diversity of active denitrifying communities from the Raigón aquifer, we analysed nirS and nirK genes in 13C-labelled DNA from SIP incubations. Care must be taken when assigning taxonomical information to nir genes sequences since congruence between 16S rRNA and nir genes phylogenies is taxon dependent (Heylen et al.2006; Jones et al.2008). The nirS gene library from site C was dominated by two OTUs that clustered closely with nirS sequences obtained from two Vogesella species isolated from the same aquifer well (Bellini et al.2013). Retrieved nirS sequences that cluster closely with members of Comamonadaceae and Dechloromonas could also be active denitrifiers, although they were less abundant. nirK sequences affiliated with Rhizobiales also appear to represent active denitrifiers at site C. For site D, nirS sequences that clustered with nirS from Dechloromonas were most abundant while most nirK gene sequences affiliated with nirK from Rhizobium and Bosea. These results reinforce the findings obtained with 16S rRNA gene analyses and suggest that Dechloromonas is a key denitrifier in site D, with a minor contribution to the denitrification process by Rhizobiaceae. Analysis of nirS libraries in site H showed that the dominant OTUs clustered with environmental sequences and sequences from the family Rhodocyclaceae, although this family was clearly under-represented in 16S rRNA gene library. The dominant taxon observed in the 16S rRNA gene library (Comamonas) was not observed either in nirS or nirK libraries. However, most of nirS sequences from this site clustered with environmental clones, and 13% of the sequences clustered closely to the Comamonadaceae isolate I-Bh25-7. A BLAST search using 16S rRNA gene from isolate I-Bh25-7 indicates that it could belong to the genus Simplicispira, within the Comamonadaceae. Moreover, most denitrifying Comamonas isolates studied so far possesses nirS (Heylen et al.2006), and these nirS sequences should have been retrieved with the PCR primers used in this study. Accordingly, most of the nirS sequences found in this site could actually belong to the Comamonas that were detected with 16S rRNA gene sequences. Altogether, 16S rRNA, nirS and nirK data suggest that Comamonadaceae, Rhodocyclaceae and Rhizobiaceae were present at site H. In summary, the three sampled sites from the Raigón aquifer differed in microbial communities and in their denitrification rates in response to nutrients amendments. However, the indigenous communities in this aquifer have the potential to remove contaminating nitrate and perform a natural attenuation process through different and taxonomically diverse genera, mainly Betaproteobacteria. SUPPLEMENTARY DATA Supplementary data are available at FEMSEC online. Acknowledgements The authors would like to thank the Agencia Nacional de Investigación e Innovación (ANII), Laboratorio Tencológico del Uruguay (LATU), Comisión Sectorial de Investigación Científica (CSIC),Programa de Desarrollo de Ciencias Báscias (PEDECIBA) and Luciana Pereira. REFERENCES Altschul SF, Madden TL, Schaeffer AA et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res  1997; 25: 3389– 402. Google Scholar CrossRef Search ADS PubMed  American Public Health Association (APHA). Standard Methods for the Examination of Water and Wastewater . 20th ed., Washington DC: American Public Health Association, 1998. Antony CP, Kumaresan D, Ferrando L et al. Active methylotrophs in the sediments of Lonar Lake, a saline and alkaline ecosystem formed by meteor impact. ISME J  2010; 4: 1470– 80. Google Scholar CrossRef Search ADS PubMed  Bellini MI, Gutiérrez L, Tarlera S et al. Isolation and functional analysis of denitrifiers in an aquifer with high potential for denitrification. Syst Appl Microbiol  2013; 36: 505– 16. Google Scholar CrossRef Search ADS PubMed  Braker G, Schwarz J, Conrad R. Influence of temperature on the composition and activity of denitrifying soil communities. FEMS Microbiol Ecol  2010; 73: 134– 48. Google Scholar PubMed  Calderer M, Martí V, de Pablo J et al. Effects of enhanced denitrification on hydrodynamics and microbial community structure in a soil column system. Chemosphere  2014; 111: 112– 9. Google Scholar CrossRef Search ADS PubMed  Caporaso JG, Bittinger K, Bushman FD et al. PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics  2010a; 26: 266– 7. Google Scholar CrossRef Search ADS   Caporaso JG, Kuczynski J, Stombaugh J et al. QIIME allows analysis of high-throughput community sequencing data. Nat Meth  2010b; 7: 335– 6. Google Scholar CrossRef Search ADS   Cole J, Wang Q, Cardenas E et al. The Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Res  2009; 37: 141– 5. Google Scholar CrossRef Search ADS   Cunliffe M, Schäfer H, Harrison E et al. Phylogenetic and functional gene analysis of the bacterial and archaeal communities associated with the surface microlayer of an estuary. ISME J  2008; 2: 776– 89. Google Scholar CrossRef Search ADS PubMed  DeSantis TZ, Hugenholtz P, Larsen N et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microb  2006; 72: 5069– 72. Google Scholar CrossRef Search ADS   Di Rienzo JA, Casanoves F, Balzarini MG et al. InfoStat. Grupo InfoStat, FCA . Argentina: Universidad Nacional de Córdoba, 2009. http://www.infostat.com.ar (19 December 2017, date last accessed). Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics  2010; 26: 2460– 1. Google Scholar CrossRef Search ADS PubMed  Eschenbach W, Well R, Walther W. Predicting the denitrification capacity of sandy aquifers from in situ measurements using push-pull 15N tracer tests. Biogeosci Discuss  2014; 11: 16527– 72. Google Scholar CrossRef Search ADS   Etchebehere C, Errazquin MI, Dabert P et al. Comamonas nitrativorans sp. nov., a novel denitrifier isolated from a denitrifying reactor treating landfill leachate. Int J Syst Evol Micr  2001; 51: 977– 83. Google Scholar CrossRef Search ADS   Fischer A, Manefield M, Bombach P. Application of stable isotope tools for evaluating natural and stimulated biodegradation of organic pollutants in field studies. Curr Opin Biotech  2016; 41, 99– 107. Google Scholar CrossRef Search ADS PubMed  Fish JA, Chai B, Wang Q et al. FunGene: the Functional Gene Pipeline and Repository. Front Microbiol  2013; 4: 291. Google Scholar CrossRef Search ADS PubMed  Ginige MP, Keller J, Blackall LL. Investigation of an acetate-fed denitrifying microbial community by stable isotope probing, full-cycle rRNA analysis, and fluorescent in situ hybridization-microautoradiography. Appl Environ Microb  2005; 71: 8683– 91. Google Scholar CrossRef Search ADS   Griebler C, Lueders T. Microbioal biodiversity in groundwater ecosystems. Freshwater Biol  2009; 54: 649– 77. Google Scholar CrossRef Search ADS   Grimes DJ, Woese CR, MacDonell MT et al. Systematic study of the genus Vogesella gen. nov. and its type species, Vogesella indigofera comb. nov. Int J Syst Evol Micr  1997; 47: 19– 27. Gumaelius L, Magnusson G, Pettersson B et al. Comamonas denitrificans sp. nov., an efficient denitrifying bacterium isolated from activated sludge. Int J Syst Evol Micr  2001; 51: 999– 1006. Google Scholar CrossRef Search ADS   Hallin S, Lindgren P. PCR detection of genes encoding nitrite reductase in denitrifing bacteria. Appl Environ Microb  1999; 65: 1652– 7. Heylen K, Gevers D, Vanparys B et al. The incidence of nirS and nirK and their genetic heterogeneity in cultivated denitrifiers. Environ Microbiol  2006; 8: 2012– 21. Google Scholar CrossRef Search ADS PubMed  Jones CM, Stres B, Rosenquist M et al. Phylogenetic analysis of nitrite, nitric oxide, and nitrous oxide respiratory enzymes reveal a complex evolutionary history for denitrification. Mol Bio Evol  2008; 25: 1955– 66. Google Scholar CrossRef Search ADS   Lin X, McKinley J, Resch CT et al. Spatial and temporal dynamics of the microbial community in the hanford unconfined aquifer. ISME J  2012; 6: 1665– 76. Google Scholar CrossRef Search ADS PubMed  Lueders T, Pommerenke B, Friedrich MW. Stable-isotope probing of microorganisms thriving at thermodynamic limits: syntrophic propionate oxidation in flooded soil. Appl Environ Microb  2004; 70: 5778– 86. Google Scholar CrossRef Search ADS   McIlroy SJ, Starnawska A, Starnawski P et al. Identification of active denitrifiers in full-scale nutrient removal wastewater treatment systems. Environ Microb  2016; 18: 50– 64. Google Scholar CrossRef Search ADS   Mahne I, Tiedje JM. Criteria and methodology for identifying respiratory denitrifiers. Appl Environ Microb  1995; 61: 1110– 5. Maneesha PG, Hugenholtz P, Daimes H et al. Use of stable-isotope probing, full-cycle rRNA analysis, and fluorescence in situ hybridization–microautoradiography to study a methanol-fed denitrifying microbial community. Appl Environ Microb  2004; 70: 588– 96. Google Scholar CrossRef Search ADS   Muyzer G, De Waal EC, Uitterlinden AG. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl Environ Microb  1993; 59: 695– 700. Neufeld JD, Schafer H, Cox MJ et al. Stable-isotope probing implicates Methylophaga spp and novel Gammaproteobacteria in marine methanol and methylamine metabolism. ISME J  2007a; 6: 480– 91. Google Scholar CrossRef Search ADS   Neufeld JD, Vohra J, Dumont MG et al. DNA stable-isotope probing. Nat Protocols  2007b; 2: 860– 6. Google Scholar CrossRef Search ADS   Osaka T, Ebie Y, Tsuneda S et al. Identification of the bacterial community involved inmethanedependent denitri¢cation inactivated sludge using DNA stable-isotope probing. FEMS Microbiol Ecol  2008; 64: 494– 506. Google Scholar CrossRef Search ADS PubMed  Osaka T, Yoshie S, Tsuneda S et al. Identification of acetate- or methanol-assimilating bacteria under nitrate-reducing conditions by stable-isotope probing. Microb Ecol  2006; 52: 253– 66. Google Scholar CrossRef Search ADS PubMed  Pelz O, Chatzinotas A, Andersen N et al. Use of isotopic and molecular techniques to link toluene degradation in denitrifying aquifer microcosms to specific microbial populations. Arch Microbiol  2001; 175: 270– 81. Google Scholar CrossRef Search ADS PubMed  Philippot L, Hallin S, Schloter M. Ecology of denitrifying prokaryotes in agricultural soil. Adv Agron  2007; 96: 249– 305. Google Scholar CrossRef Search ADS   Rivett MO, Buss SR, Morgan P et al. Nitrate attenuation in groundwater: a review of biogeochemical controlling processes. Water Res  2008; 42: 4215– 32. Google Scholar CrossRef Search ADS PubMed  Saito T, Ishii S, Otsuka S et al. Identification of novel betaproteobacteria in a succinate-assimilating population in denitrifying rice paddy soil by using stable isotope probing. Microbes Environ  2008; 23: 192– 200. Google Scholar CrossRef Search ADS PubMed  Santoro AE, Boehm AB, Francis CA. Denitrifier community composition along a nitrate and salinity gradient in a coastal aquifer. Appl Environ Microb  2006; 72: 2102– 9. Google Scholar CrossRef Search ADS   Schwarz IKJ, Lueders T, Eckert W et al. Identification of acetate-utilizing Bacteria and Archaea in methanogenic profundal sediments of Lake Kinneret (Israel) by stable isotope probing of rRNA. Environ Microb  2007; 9: 223– 37 Google Scholar CrossRef Search ADS   Shoun H, Kim D, Uchiyama H et al. Denitrification by fungi. FEMS Microbiol Lett  1992; 94: 277– 81. Google Scholar CrossRef Search ADS   Singleton DR, Powell SN, Sangaiah R et al. Stable-isotope probing of bacteria capable of degrading salicylate, naphthalene, or phenanthrene in a bioreactor treating contaminated soil. Appl Environ Microb  2005; 71: 1202– 9. Google Scholar CrossRef Search ADS   Smith RL, Duff JH. Denitrification in a sand and gravel aquifer. Appl Environ Microb  1988; 54: 1071– 8. Sun W, Cupples AM. Diversity of five anaerobic toluene-degrading microbial communities investigated using stable isotope probing. Appl Environ Microb  2012; 78: 972– 80. Google Scholar CrossRef Search ADS   Tamura K, Stecher G, Peterson D et al.   MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol  2013; 30: 2725– 9. Google Scholar CrossRef Search ADS PubMed  Tarlera S, Denner EBM. Sterolibacterium denitrificans gen. nov., sp. nov., a novel cholesterol-oxidizing, denitrifying member of the β-Proteobacteria. Int J Syst Evol Micr  2003; 53: 1085– 91. Google Scholar CrossRef Search ADS   Throbäck IN, Enwall K, Javis A et al. Reassessing PCR primers targeting nirS, nirK, and nosZ genes for community surveys of denitrifying bacteria with DGGE. FEMS Microbiol Ecol  2004; 49: 401– 17. Google Scholar CrossRef Search ADS PubMed  Tiedje JM. Ecology of denitrification and dissimilatory nitrate reduction to ammonium. In: Zehnder AJB (ed). Biology of Anaerobic Microorganisms . New York, USA: John Wiley and Sons, 1988, 179– 244. Torrentó C, Urmeneta J, Otero N et al. Enhanced denitrification in groundwater and sediments from a nitrate-contaminated aquifer after addition of pyrite. Chem Geol  2011; 287: 90– 101. Google Scholar CrossRef Search ADS   Vishnivetskaya TA, Brandt CC, Madden AS et al. Microbial community changes in response to ethanol or methanol amendments for U(VI) reduction. App Environ Microb  2010; 76: 5728– 35. Google Scholar CrossRef Search ADS   Wakelin SA, Nelson PN, Armour JD et al. Bacterial community structure and denitrifier (nir-gene) abundance in soil water and groundwater beneath agricultural land in tropical North Queensland, Australia. Soil Res  2011; 49: 65– 76. Google Scholar CrossRef Search ADS   Wang Q, Garrity GM, Tiedje JM et al. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microb  2007; 73: 5261– 7. Google Scholar CrossRef Search ADS   Wang Q, Quensen JFIII, Fish JA et al. Ecological patterns of nifH genes in four terrestrial climatic zones explored with targeted metagenomics using FrameBot, a new informatics tool. mBio  2013, DOI: 10.1128/mBio.00592-13. Wang Z, Zhang XX, Lu X et al. Abundance and diversity of bacterial nitrifiers and denitrifiers and their functional genes in tannery wastewater treatment plants revealed by high-throughput sequencing. PLoS One  2014, DOI: 10.1371/journal.pone.0113603. Weymann D, Well R, Flessa H et al. Groundwater N2O emission factors of nitrate-contaminated aquifers as derived from denitrification progress and N2O accumulation. Biogeosciences  2008; 5: 1215– 26. Google Scholar CrossRef Search ADS   Whiteley M, Ott JR, Weaver EA et al. Effects of community composition and growth rate on aquifer biofilm bacteria and their susceptibility to betadine disinfection. Environ Microbiol  2001; 3: 43– 52. Google Scholar CrossRef Search ADS PubMed  Yoshida M, Ishii S, Fujii D et al. Identification of active denitrifiers in rice paddy soil by DNA- and RNA-based analyses. Microbes Environ  2012; 27: 456– 61. Google Scholar CrossRef Search ADS PubMed  Yu Z, Yang J, Liu L. Denitrifier community in the oxygen minimum zone of a subtropical deep reservoir. PLoS One  2014, DOI: 10.1371/journal.pone.0092055. Zeng X, Hosono T, Ohta H et al. Comparison of microbial communities inside and outside of a denitrification hotspot in confined groundwater. Int Biodeter Biodegr  2016; 114: 104– 9. Google Scholar CrossRef Search ADS   © FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

FEMS Microbiology EcologyOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from Google Scholar, PubMed
Create lists to organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off