Diversity and Abundance of Nitrate Reductase Genes (narG and napA), Nitrite Reductase Genes (nirS and nrfA), and Their Transcripts in Estuarine Sediments

Abstract

Estuarine systems are the major conduits for the transfer of nitrate from agricultural and other terrestrial-anthropogenic sources into marine ecosystems. Within estuarine sediments some microbially driven processes (denitrification and anammox) result in the net removal of nitrogen from the environment, while others (dissimilatory nitrate reduction to ammonium) do not. In this study, molecular approaches have been used to investigate the diversity, abundance, and activity of the nitrate-reducing communities in sediments from the hypernutrified Colne estuary, United Kingdom, via analysis of nitrate and nitrite reductase genes and transcripts. Sequence analysis of cloned PCR-amplified narG, napA, and nrfA gene sequences showed the indigenous nitrate-reducing communities to be both phylogenetically diverse and also divergent from previously characterized nitrate reduction sequences in soils and offshore marine sediments and from cultured nitrate reducers. In both the narG and nrfA libraries, the majority of clones (48% and 50%, respectively) were related to corresponding sequences from delta-proteobacteria. A suite of quantitative PCR primers and TaqMan probes was then developed to quantify phylotype-specific nitrate (narG and napA) and nitrite reductase (nirS and nrfA) gene and transcript numbers in sediments from three sites along the estuarine nitrate gradient. In general, both nitrate and nitrite reductase gene copy numbers were found to decline significantly (P < 0.05) from the estuary head towards the estuary mouth. The development and application, for the first time, of quantitative reverse transcription-PCR assays to quantify mRNA sequences in sediments revealed that transcript numbers for three of the five phylotypes quantified were greatest at the estuary head.

Anthropogenic and agricultural activities result in increasing amounts of nitrate entering terrestrial, freshwater, and marine systems (63). In the environment, nitrate levels are influenced by the microbially driven processes of nitrate reduction, whereby nitrate is reduced initially to nitrite, which subsequently can be reduced by denitrification (the stepwise reduction of nitrite to dinitrogen gas) or via dissimilatory nitrate reduction to ammonium (DNRA). Recently, anaerobic ammonium oxidation (anammox) in marine and estuarine systems has been shown, at some sites, to remove significant amounts of nitrite and ammonium, resulting in the formation of dinitrogen gas (12, 35, 61). As a result of microbial activity, nitrate is either removed from ecosystems by reduction to gaseous forms during denitrification and anammox or converted, via DNRA, to biologically available ammonium, which is retained within the system.

Nitrate reduction is mediated by a diverse polyphyletic group of bacteria (66). Consequently, rRNA-based approaches are of limited value for understanding the structure and diversity of nitrate-reducing communities. Instead, genes that encode key enzymes (nitrate, nitrite, nitric oxide, and nitrous oxide reductases) in nitrate reduction have been exploited as molecular markers. The diversity of narG, encoding the membrane-bound nitrate reductase, has been widely studied in a variety of soil ecosystems (11, 13-15, 19, 31, 41, 46, 47), and results have indicated that soil nitrate-reducing communities are dominated by diverse unknown nitrate reducers. In contrast, there has been only one investigation of diversity of the napA gene, which encodes the periplasmic nitrate reductase. In this limited analysis of a freshwater sediment, a tightly clustered napA community was found, with highest sequence identities to napA genes from gamma-proteobacteria (23). Diversity of denitrifiers has been investigated using PCR-based analysis of nirS, encoding the cytochrome cd₁ nitrite reductase (5, 6, 8, 10, 25, 36, 43, 44, 58, 60); nirK, encoding copper-containing nitrite reductase (6, 8, 36, 44, 53, 60); norB, encoding nitric oxide reductase (7); and nosZ, encoding nitrous oxide reductase (19, 52, 60). These studies have focused primarily on a range of marine sediments and soil ecotypes and similarly have shown a phylogenetically diverse group of nitrate-reducing bacteria. For nitrate ammonification, nrfA, which encodes a periplasmic nitrite reductase catalyzing the conversion of nitrite to ammonia, can be used as a marker. Previous PCR-based analysis of nrfA diversity in anammox and sulfate-reducing reactors showed that the majority of clones were most closely related to nrfA genes from Bacteroides spp. (40).

The focus of the current study is the Colne estuary, United Kingdom, a hypernutrified estuary with strong gradients of nitrate and ammonium from the estuary head to the mouth. In the Colne, 20 to 25% of the total N load entering the estuary is removed by denitrification, with highest rates of denitrification recorded at the estuary head site (17, 18). In our earlier molecular analysis, we detected PCR amplicons from the narG, napA, nirS, nirK, and nosZ genes in the estuary sediment and additionally demonstrated expression of both nirS and nosZ by reverse transcription-PCR (RT-PCR) from mRNA (43). Nitrite reductase gene expression has also been detected in pond water (65) and in rhizosphere root soils, where nirK but not nirS gene expression was demonstrated (53). The application of quantitative PCR (Q-PCR)-based approaches to microbial ecology offers the potential for determining gene and transcript abundances in complex environments. Q-PCR-based methods have now become widely established for the quantification of gene copy numbers from environmental samples (3, 54, 55, 57), and a series of Q-PCR assays to quantify nitrate reduction genes have been described. Gruntzig et al. (28) developed a nirS Q-PCR TaqMan assay to quantify Pseudomonas stutzeri-related nirS genes in lake sediment and groundwater, while SYBR green-based Q-PCR approaches have also been used to quantify nirK (30) and nosZ (29) genes in soil and to compare narG copy numbers present in a variety of soil systems and in freshwater sediment (37). To investigate gene expression, Q-PCR can be combined with an initial RT reaction (Q-RT-PCR) to determine gene transcript numbers in environmental systems. Thus far, such studies have been limited to the analysis of gene expression in aqueous systems or to quantify gene expression in individual species (21, 32, 42, 64).

The aims of the current study were twofold: first, to investigate the diversity of key functional genes (narG, napA, and nrfA) along the nitrate gradient in sediments from the Colne estuary, and second, to develop and apply TaqMan-based Q-PCR and Q-RT-PCR assays to quantify the abundance and expression of these genes and also of nirS along this gradient.

MATERIALS AND METHODS

Sample collection.

Triplicate sediment samples (top 1 cm) were taken, from within an area of 1 square meter, at two monthly intervals during 2005 along the Colne estuary, United Kingdom, from the estuary head at the Hythe (51°52.4′N, 0°55.5′E), midway down the estuary at Alresford (51°50.5′N, 0°58.4′E), and from the estuary mouth at Brightlingsea (51°45′N, 1°30′E) (Fig. 1). The estuary has strong gradients of NO₃⁻ and NH₄⁺ decreasing from the estuary head towards the mouth (45, 49), resulting from agricultural inputs into the River Colne and from a sewage treatment plant near the Hythe. Sediments at the Hythe and Alresford are fine silt, while surface sediment at Brightlingsea is muddy sand. Salinity at the Hythe ranges between 2 and 17‰ depending upon tidal flow, that at Alresford ranges between 20 and 32‰, and that at Brightlingsea ranges between 28 and 32‰ (17). Sediment samples were transported on ice and returned to the laboratory within 1 hour of sampling. Aliquots of 0.5 g (wet weight) of homogenized sediment from each replicate were then stored at −70°C until subsequent molecular analysis.

FIG. 1.

Map of the Colne estuary, Essex, United Kingdom, showing the locations of the three sampling sites (Hythe, Alresford, and Brightlingsea).

Total DNA extraction.

Nucleic acids were extracted from sediment samples by using Lysing Matrix B tubes (Bio 101, QBiogene, Cambridge, United Kingdom). To each tube, 0.5 g of sediment, 0.5 ml of 240 mM sodium phosphate buffer (pH 8.0), and 0.5 ml of phenol-chloroform-isoamyl alcohol (25:24:1) (pH 4.0) were added. Samples were lysed by bead beating for 30 seconds at 2,000 rpm using a Mikrodismembrator U (B.Braun Biotech International, Melsungen, Germany). The sediment and lysate were separated by centrifugation at 17,563 × g for 10 min at 4°C. The upper aqueous layer was removed and transferred to an Eppendorf tube containing 0.5 ml of chloroform-isoamyl alcohol (24:1) and centrifuged for a further 10 min. The upper aqueous phase was again transferred to a fresh tube, and total nucleic acids were precipitated with 2.5 volumes of absolute ethanol and 0.1 volume of 3 M sodium acetate (pH 5.2) at −20°C for 1 h and centrifuged for a further 25 min. The resulting pellet was then washed twice with 70% ethanol, air dried, and finally resuspended in 100 μl of sterile RO water. Extraction of DNA was confirmed by gel electrophoresis. Total genomic DNA was further purified by agarose gel electrophoresis using the QIAquick gel extraction kit (QIAGEN Ltd., Crawley, United Kingdom) according to the manufacturer's recommendations.

Total RNA extraction and cDNA synthesis.

Total RNA was extracted from sediment samples in a manner similar to that for DNA, but the phenol-chloroform-isoamyl alcohol was replaced with 0.5 ml of Tri Reagent (Fluka, Seelze, Germany). The Lysing Matrix B tubes (Bio 101) were subjected to two 30-s cycles of bead beating at 2,000 rpm and left at room temperature for 5 min. Sediment and cell debris were pelleted by centrifugation at 17,563 × g for 10 min at 4°C. The supernatant was transferred to a fresh tube, and 200 μl of chloroform was added, mixed by inversion, and left at room temperature for 15 min prior to centrifugation as described above. The upper aqueous layer containing the RNA was transferred to a fresh tube containing 500 μl of ice-cold isopropanol, mixed by inversion, and left at room temperature for 10 min prior to centrifugation as described above. The resulting RNA pellet was washed with 1 ml of 75% ethanol, centrifuged at 5,670 × g for 5 min, and briefly air dried prior to resuspension in 100 μl of RNA storage solution (Ambion, Austin, TX). Extraction of RNA was confirmed by gel electrophoresis. Aliquots of 50 μl of the RNA were digested with TURBO DNase (Ambion) according to the manufacturer's instructions. To determine that the RNA sample was free of contaminating DNA, a control PCR was carried out on the RNA samples using bacterial 16S rRNA gene primers 1369F (5′ CGG TGA ATA CGT TCY CGG 3′) and Prok 1492R (5′ GGW TAC CTT GTT ACG ACT T 3′) (57). One microliter of a 10⁻¹ dilution of the RNA template was added to a 50-μl PCR mixture containing 1× PCR buffer (QIAGEN), 1.5 mM MgCl₂, 0.2 mM of each deoxynucleoside triphosphate (dNTP), 0.25 μM of each primer, and 2.5 units of Taq polymerase (QIAGEN). The reaction was initially denatured at 95°C for 5 min; followed by 30 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s; followed by a final extension step at 72°C for 5 min. To test for the presence of any inhibitors of the RT reaction in the RNA samples, an RT-PCR targeting 16S rRNA was performed as described previously (54).

PCR amplification of nitrate and nitrite reductase gene sequences.

napA amplicons (∼414 bp) were amplified using the primer pair napA V67 F (5′ TAY TTY YTN HSN AAR ATH ATG TAY GG 3′) and napA V67 R (5′ DAT NGG RTG CAT YTC NGC CAT RTT 3′) (23). One microliter of DNA was added to a 50-μl PCR mixture containing 1× PCR buffer (QIAGEN), 1.5 mM MgCl₂, 0.2 mM of each dNTP, 0.5 μM of each primer, and 2.5 units of Taq polymerase (QIAGEN). The reaction was initially denatured at 94°C for 5 min; followed by 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min; followed by a final extension step at 72°C for 10 min. narG amplicons (∼650 bp) were amplified using the primer set narG 1960 F (5′ TAY GTS GGC CAR GAR AA 3′) and narG 2659 R (5′ TTY TCR TAC CAB GTB GC 3′) with a touchdown PCR cycle and conditions as described by Philippot et al. (47). nrfA amplicons (∼520 bp) were amplified using the primers nrfA F1 (5′ GCN TGY TGG WSN TGY AA 3′) and nrfA 7R1 (5′ TWN GGC ATR TGR CAR TC 3′) (40). Reaction contents were the same as those used for the napA PCR, except that MgCl₂ was added to a final concentration of 3.0 mM. Reactions were initially denatured at 94°C for 5 min; followed by 30 cycles of 94°C for 30 s, 60°C for 45 s, and 72°C for 1.5 min; followed by a final extension step at 72°C for 10 min. PCR products were visualized by gel electrophoresis followed by ethidium bromide staining to ensure the correct size fragment was amplified.

Clone library construction and sequencing.

narG, napA, and nrfA amplicons were purified using the QIAquick PCR purification kit (QIAGEN) and cloned using the TOPO TA cloning kit for sequencing (Invitrogen, Paisley, United Kingdom) according to the manufacturer's instructions. Transformants were selected on Luria-Bertani agar plates containing ampicillin (100 μg ml⁻¹) and X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) (20 μg ml⁻¹). White colonies were screened by PCR using the vector primers M13F (5′ GTA AAA CGA CGG CCA G 3′) and M13R (5′ CAG GAA ACA GCT ATG AC 3′) (Invitrogen). Nucleotide sequences of clones containing inserts of the expected size were determined by sequencing both strands using the vector primers T7 and T3 and the BigDye terminator cycle kit v3.1 (Applied BioSystems, Warrington, United Kingdom) according to the manufacturer's recommendations, followed by electrophoresis on an ABI 3100 genetic analyzer (Applied BioSystems).

napA, narG, and nrfA nucleotide sequences were translated into protein sequences using the Translate tool on the ExPASy (Expert Protein Analysis System) proteomics server of the Swiss Institute of Bioinformatics (http://us.expasy.org/tools/dna.html). Nucleotide and protein sequences were compared to entries in GenBank using BlastN and BlastP, respectively (1). Protein sequence alignments were constructed using ClustalX, (59), and distance matrices were calculated using the PROTDIST program in PHYLIP (20). Phylogenetic trees were created from the distance matrices by using the neighbor-joining method (50) and Kimura substitution algorithm (34) using PHYLIP. Consensus trees were calculated after bootstrapping (1,000 replicate trees).

Q-PCR primer and probe construction.

A suite of Q-PCR primers and probes was designed manually, to target divergent sequence phylotypes using alignments of the cloned napA, narG, and nrfA gene sequences from the Colne estuary (this study). Additionally, primer and probe suites were designed to target divergent nirS phylotypes, using sequence alignments of cloned mRNA sequences from the Colne estuary (43). Details of the resulting primer and probe sequences, the expected amplicon length, and the annealing temperatures for each primer and probe set are given in Table 1.

TABLE 1.

Primer and probe sets used for Q-(RT)-PCR

Target gene	Phylotype	Amplicon size (bp)	Primer or probe		Q-PCR cycle annealing temp (°C)
			Namea	Sequence (5′→3′)
napA	napA-1	111	napA-1F	GTY ATG GAR GAA AAA TTC AA	55
			napA-1R	GAR CCG AAC ATG CCR AC
			napA-1 (TM-MGB)	AAC ATG ACC TGG AAG
	napA-2	76	napA-2F	GAA CCK AYG GGY TGT TATG	55
			napA-2R	TGC ATY TCS GCC ATR TT
			napA-2 (TM-MGB)	CTT TGG GGT TCA A
	napA-3	130	napA-3F	CCC AAT GCT CGC CAC TG	60
			napA-3R	CAT GTT KGA GCC CCA CAG
			napA-3 (TM-MGB)	TGG GTT GTT ACG A
narG	narG-1	69	narG-1F	GAC TTC CGC ATG TCR AC	60
			narG-1R	TTY TCG TAC CAG GTG GC
			narG-1 (TM-MGB)	TAY TCC GAC ATC GT
	narG-2	89	narG-2F	CTC GAY CTG GTG GTY GA	55
			narG-2R	TTY TCG TAC CAG GTS GC
			narG-2 (TM-MGB)	AAC TTC CGC ATG GA
nrfA	nrfA-2	67	nrfA-2F	CAC GAC AGC AAG ACT GCC G	60
			nrfA-2R	CCG GCA CTT TCG AGC CC
			nrfA-2 (TM-MGB)	TTG ACC GTC GGC A
nirS	nirS-e	172	nirS-efF	CAC CCG GAG TTC ATC GTC	60
			nirS-efR	ACC TTG TTG GAC TGG TGG G
			nirS-ef (TM-MGB)	TGC TGG TCA ACT A
	nirS-m	162	nirS-mF	GGA AAC CTG TTC GTC AAG AC	60
			nirS-mR	CSG ART CCT TGG CGA CGT
			nirS-m (TM)	TCT GGG CCG ACG CGC CGA TGA AC
	nirS-n	140	nirS-nF	AAG GAA GTC TGG ATY TC	55
nirS-nRb		CGT TGA ACT TRC CGG T
nirS-n (TM-MGB)		ATC CGA AGA TSA

^aFor probes: TM-MGB, TaqMan minor groove binding; TM, TaqMan.

^bAlso known as nirS6r (6).

Q-PCR and Q-RT-PCR standards.

DNA standard curves were constructed as described previously (54) using cloned napA, narG, and nrfA (this study) and nirS (43) sequences for each primer and probe combination. DNA and RNA were quantified using a NanoDrop spectrophotometer (NanoDrop Technologies, Delaware). For RT-Q-PCR standards, standard curves were produced from cDNA following prior in vitro transcription of the target mRNA by using the MEGAshortscript T7 kit (Ambion) as described previously (54).

Q-PCR analysis.

For each primer and probe combination, Q-PCR assays were carried out within a single assay plate, thus permitting direct comparison of absolute numbers between environmental DNA or RNA samples (54). Each assay contained a standard curve containing 10³ to 10⁷ amplicons μl⁻¹ (DNA or cDNA) for amplification by Q-PCR, independent triplicate sediment DNA or cDNA templates from each of the three sites along the Colne estuary, and triplicate no-template controls (NTC). Each of the triplicate DNA or cDNA samples was additionally amplified in triplicate within each Q-PCR assay. Experimental Q-PCR triplicates for each DNA/cDNA sample were then averaged to give a single gene copy number. Q-PCR amplification mixtures contained 1 μl of template, 10.5 μl of TaqMan universal master mix (no Amperase uracil-N-glycosylase) (Applied BioSystems), 900 nM of each primer, and 200 nM of probe made up to a total volume of 20 μl with sterile RO water. Q-PCR amplification and detection for all primer and probe combinations were performed using an ABI 7000 sequence detection system (Applied BioSystems), with an initial denaturation for 10 min at 95°C followed by 40 cycles of 95°C for 15 s and annealing/extension at the temperatures indicated in Table 1 for 1 min. Gene and gene transcript numbers were quantified via comparison to standard curves using the ABI Prism 7000 sequence detection software. Automatic analysis settings were selected to determine the threshold cycle (C_T) values and baseline settings. For each standard curve, the slope, y intercept, and coefficient of determination (r²) were determined. In addition, the efficiency of the amplification (E) was calculated using the equation E = (10^1/m − 1) × 100, where m is the slope of the standard curve.

Q-RT-PCR analysis.

Two-step Q-RT-PCR amplifications were performed. The initial RT reaction mixtures, containing either 1 μl of the RNA standard or 5 μl of environmental RNA, 2 mM of the appropriate reverse primer, and 10 mM of each dNTP and made up to a final volume of 14 μl with sterile diethylpyrocarbonate-treated water, were denatured at 65°C for 5 min and then transferred to ice for 1 min. Four microliters of 5× first-strand buffer, 1 μl of 0.1 mM dithiothreitol, and 200 units of SuperScript III (Invitrogen) were added to the reaction mixture and incubated at 50°C for 50 min, followed by inactivation of the reaction at 70°C for 15 min. One microliter of the first-strand cDNA synthesized in this reaction was amplified by Q-PCR using the same reaction and cycling conditions as described above.

Statistical analysis.

Analysis of variation in the quantification of the environmental samples between sites was done using a one-way analysis of variance followed by a post hoc Tukey test (62) in SPSS v14. Data was first transformed using log10 (x + 1) (2).

Nucleotide sequence accession numbers.

Nucleotide sequences have been deposited in the EMBL database with the following accession numbers: nrfA, AM408253 to AM408282; napA, AM408431 to AM408493; and narG, AM408498 to AM408549.

RESULTS

Diversity of narG, napA, and nrfA genes in sediments from the Colne estuary.

Clone libraries of nitrate and nitrite reductase gene sequences amplified from DNA extracted from sediments from the three sampling sites during 2005 were constructed. In total, 52 narG sequences were obtained from the Hythe and Alresford (June) and from Brightlingsea (August); 63 napA sequences were obtained from the Hythe (February and June) and from Alresford and Brightlingsea (June); and 30 nrfA sequences were generated from the Hythe (April and October), Alresford (April), and Brightlingsea (October). Clone designations were given to represent, in order, the target gene (nar, nap, or nrf), sampling site (H, A, or B), sampling month (F, A, J, Au, or O), and a numerical identifier. Deduced protein sequences from the cloned narG, napA, and nrfA sequences together with sequences from cultured bacteria and other environmental clone libraries were then used to construct phylogenies (see Fig. 2 to 4).

FIG. 2.

Neighbor-joining dendrogram derived from the deduced protein sequences encoded by narG sequences cloned from Colne estuary sediments (in boldface) and from cultured bacteria and other environmental clones. The narG sequence from Mycobacterium tuberculosis was used as the outgroup. Bootstrap values of >70% (from 1,000 replicates), a scale bar representing 10% sequence divergence, and clusters I to VI are indicated. Clone designations are shown by letters (and symbols) as follows: Hythe, H (□); Alresford, A (▵); and Brightlingsea, B (○). The month of sampling is indicated by either Au (August) or J (June). Sequences used to design narG-1 and -2 Q-PCR primer and probe sets are indicated.

FIG. 4.

Neighbor-joining dendrogram derived from the deduced protein sequences encoded by nrfA sequences cloned from Colne estuary sediments (in boldface) and from cultured bacteria and other environmental clones. The nrfA gene from Escherichia coli was used as the outgroup. Bootstrap values of >70% (from 1,000 replicates), a scale bar representing 10% sequence divergence, and clusters I to IV are indicated. Clone designations are shown by letters (and symbols) as follows: Hythe, H (□); Alresford, A (▵); and Brightlingsea, B (○). The month of sampling is indicated by either A (April) or O (October). Sequences used to design the nrfA-2 Q-PCR primer and probe set are indicated.

narG sequences in the Colne sediments were diverse (Fig. 2) and divergent from those of both cultured bacteria and other environmental narG clone libraries. Forty-two of the 52 translated narG sequences from the Colne estuary represented unique phylotypes. Of the remaining 10 clones, six further phylotypes were identified, of which the largest cluster comprised four clones from Alresford (narAJ8, narAJ18, narAJ30, and narAJ50). Twenty-three of the 52 clones formed a discrete cluster (cluster I, with bootstrap support of 100%) of sequences closely related to each other that ranged from 61 to 65% identity at the protein level to NarG from Anaeromyxobacter dehalogenans 2CP-C (51). This cluster was comprised of narG sequences from Hythe and Alresford sediments only. A further 10 clones (cluster II) were most closely related to narG sequences derived from agricultural soils (11, 41). Four clones were related to narG sequences derived from agricultural soils (Fig. 2, cluster III) and to narG from Thermus thermophilus strain HB8 (48) (66 to 69% identity). A further four clones (cluster IV) were most closely related to sequences from rhizosphere soil (14, 15) and to narG from the delta-proteobacterium Geobacter metallireducans, originally isolated from freshwater sediment (38). Of the remaining narG clones, the nine clones in cluster V were most closely related to narG sequences cloned from nickel mine spoils (31) and to the beta-proteobacterium Rhodoferax ferrireducans originally isolated from subsurface sediment collected from Oyster Bay, VA (22), while a further two clones (narHJ58 and narBAu35) were most closely related to narG from Halomonas halodenitrificans (16).

Colne sediment napA sequences were also found to be highly diverse (Fig. 3), and again none of the sequences were identical to those from either cultured bacteria or other environmental clone libraries. Of the 63 cloned napA sequences, 45 represented unique phylotypes, with a further nine phylotypes each comprised of two clones. The translated napA sequences from the Colne were distributed within four of five distinct clusters (Fig. 3, clusters I to IV), with cluster I containing 28 sequences and cluster II containing 32 sequences. The napA sequences in cluster I were related to those from alpha- (Rhodobacter sphaeroides), beta- (Cupriavidus necator), and gamma-proteobacteria (Hahella chejuensis, Shewanella spp., and members of the Vibrionaceae) (Fig. 3). In cluster II, napA sequences clustered with those from gamma- (Saccharophagus degradans) and alpha-proteobacteria (Rhodopseudomonas palustris). There was no discernible site-specific clustering of clones within these two clusters. Of the remaining three clones, one sequence (napHF12) was most closely related to napA from the alpha-proteobacterium Magnetospirillum magneticum AMB1 (39), while two other sequences (napHJ5 and HJ6) were most closely related to the napA sequence from the epsilon-proteobacterium Helicobacter hepaticus ATCC51449 (56). A fifth cluster comprised napA sequences from previously cultured Pseudomonas spp.

FIG. 3.

Neighbor-joining dendrogram derived from the deduced protein sequences encoded by napA sequences cloned from Colne estuary sediments (in boldface) and from cultured bacteria and other environmental clones. The napA gene from Escherichia coli was used as the outgroup. Bootstrap values of >70% (from 1,000 replicates), a scale bar representing 10% sequence divergence, and clusters I to V are indicated. Clones designations are shown by letters (and symbols) as follows: Hythe, H (□); Alresford, A (▵); and Brightlingsea, B (○). The month of sampling is indicated by either F (February) or J (June). Sequences used to design napA-1, -2, and -3 Q-PCR primer and probe sets are indicated.

From the 30 nrfA sequences, 23 phylotypes were identified in total, of which 19 consisted of one clone only (Fig. 4). Of the other four phylotypes, one contained five clones from the Hythe sediment (nrfHO5, nrfHO7, nrfHO15, nrfHO16, and nrfHO29), while the other three phylotypes each comprised two clones. Fifteen of the 30 cloned sequences were most closely related to nrfA from members of the delta-proteobacteria (Fig. 4, cluster III), with six nrfA sequences related to those from epsilon-proteobacteria (Fig. 4, cluster II) and two sequences related to nrfA from the gamma-proteobacterium Shewanella frigidimarina (4) (Fig. 4, cluster IV). The final seven Colne nrfA sequences (Fig. 4, cluster I) were related to that from Chlorobium phaeobacteroides (33) and additionally to a series of environmental nrfA clones from an anammox reactor and a sulfate-reducing sludge reactor (40).

Design of Q-PCR primer and probe sets.

A suite of primer and Q-PCR probes was designed for use in TaqMan-based assays to quantify the abundances of nitrate reductase (narG and napA) and nitrite reductase (nrfA and nirS) genes and transcript numbers. Due to the considerable sequence diversity of the cloned Colne sediment sequences for the narG, napA, and nrfA genes (this study) and additionally of RT-PCR-amplified nirS mRNA Colne sequences from our earlier study (43), “universal” primer and probe combinations could not be designed. Therefore, a suite of Q-PCR primer and TaqMan probe sets was designed, targeting specific subgroups of napA, narG, and nrfA as indicated in Fig. 2 to 4 and, for nirS, targeting four clusters (m, n, e, and f) identified in Fig. 4 of reference 43. The nirS primer and probe sets for clusters m and n targeted cloned mRNA sequences retrieved from the Hythe site, while the combined set for clusters e and f targeted mRNA sequences cloned originally from the Alresford site.

Q-PCR quantification of nitrate and nitrite reductase gene copy numbers in Colne estuary sediments.

Q-PCR assays were used to investigate variation in numbers of narG, napA, nirS, and nrfA genes in sediments taken in October 2005, using the primer and probe sets described in Table 1. For both of the narG subgroups targeted by Q-PCR, the highest gene copy numbers were found at the estuary head (Hythe), with a subsequent decline in gene copy number towards the estuary mouth (Brightlingsea) (Fig. 5A), with a 3.5-fold significant difference (analysis of variance, P < 0.05) between the Hythe and Brightlingsea for the narG-1 subgroup. For the narG-2 subgroup, gene copy numbers were greater than the corresponding narG-1 numbers at each site and corresponded to a gene copy number ratio of 16.5:4.3:1.0 at Hythe, Alresford, and Brightlingsea, respectively. Significant differences in narG-2 gene copy numbers were observed between all three sites (P < 0.001) (Fig. 5A).

FIG. 5.

Variation in gene abundance (copy number g⁻¹ sediment) of nitrate reduction genes along the Colne estuary, October 2006. (A) narG; (B) napA; (C) nirS. Standard errors are indicated (n = 3) for each separate Q-PCR subgroup, with statistical differences (P = 0.05) between sites indicated by different letters for each gene phylotype. Gene copy numbers were calculated from the following standard curves: for narG-1, r² = 0.995, y intercept = 40.52, E (amplification efficiency) = 100.9%, and NTC undetected; for narG-2, r² = 0.999, y intercept = 44.57, E = 90.9%, and C_T cutoff of 32.3; for napA-1, r² = 0.998, y intercept = 44.99, E = 86.9%, and NTC undetected; for napA-2, r² = 0.997, y intercept = 39.00, E = 89.6%, and C_T cutoff of 32.83; for napA-3, r² = 0.996, y intercept = 47.60, E = 89.2%, and NTC undetected; for nirS-e&f, r² = 0.998, y intercept = 35.89, E = 86.0%, and C_T cutoff of 35.00; for nirS-m, r² = 0.996, y intercept = 41.13, E = 86.1%, and NTC undetected; and for nirS-n, r² = 0.998, y intercept = 42.84, E = 82.1%, and NTC undetected.

For napA, gene copy numbers for two of the three subgroups (napA-1 and napA-2) decreased from the estuary head to the estuary mouth (Fig. 5B). Gene copy numbers of the napA-1 subgroup were significantly higher at the Hythe than at the other two sites (P < 0.05), but there was no significant difference between gene copy numbers at Alresford and Brightlingsea (P = 0.57). Differences between gene copy numbers at the three sites along the estuary were less pronounced for the napA-2 and napA-3 subgroups. For the napA-2 subgroup, although gene copy numbers were highest at the Hythe, with decreasing gene abundances along the estuary, there were no significant differences between napA-2 gene copy numbers at the three sites (P < 0.05). Similarly, there was no significant difference between napA-3 gene copy numbers along the estuary (P < 0.05) (Fig. 5B).

Gene copy numbers of the three nirS subgroups were lower at each site than those found for nitrate reductase genes (narG and napA) (Fig. 5C). For all three nirS subgroups, gene copy numbers decreased from the estuary head towards the estuary mouth. Highest nirS gene copy numbers were found for the nirS-n subgroup at the Hythe site (Fig. 5C), which showed a significant 265-fold decrease in gene copy number (P <0.05) from the Hythe to the Alresford site, while nirS-n sequences were not detected at Brightlingsea. Similarly, a decrease in gene copy numbers along the estuary gradient was found for the nirS-m subgroup, with a gene copy ratio of 136.9:4.5:1.0 from the estuary head to the estuary mouth. The nirS-m gene copy numbers in Hythe sediments were significantly greater than those at the other two sites (P < 0.05), but there was no significant difference between nirS-m copy numbers in Alresford and Brightlingsea sediments (P = 0.083). The differences in gene copy numbers along the estuary for the nirS-e&f subgroup were less pronounced, and although nirS-e&f copy numbers declined from the estuary head towards the estuary mouth (with a ratio of 5.7:2.3:1 for Hythe, Alresford, and Brightlingsea, respectively), these differences were not significant (P = 0.128).

For the nrfA gene, a single Q-PCR primer and probe set (nrfA-2) was successfully developed (Fig. 4), with gene copy numbers determined against a standard curve with an r² value of 0.997, y intercept of 39.89, and amplification efficiency (E) of 96.02%. An NTC cutoff value of 28.51 was set. Gene copy numbers for the nrfA-2 subgroup were 1.7 times greater at the Hythe (mean ± standard error, 3.44 × 10⁶ ± 2.82 × 10⁵) than at Alresford (2.01 × 10⁶ ± 2.91 × 10⁵), although there was no significant difference between sites (P = 0.190). nrfA-2 sequences were not detected in the Brightlingsea sediment.

Q-RT-PCR quantification of nitrate and nitrite reductase gene transcript numbers in Colne estuary sediments.

A series of Q-RT-PCR assays were used to investigate variation in the levels of narG, napA, nrfA, and nirS gene expression using RNA extracted from sediments along the estuary in October 2005. For all of the gene transcripts studied, C_T values derived from RNA extracted from sediment samples were low and were detected late in the Q-RT-PCR amplifications. To ensure that the C_T values detected by Q-RT-PCR were “real” amplifications of mRNA and not derived from background noise, a C_T cutoff value corresponding to 3.3 cycles fewer than that determined for the NTC (if detected) was set (54). C_T values from environmental RNA templates that were below this threshold were regarded as being below the limit of detection.

For narG, gene transcripts were detected and quantified for both of the narG subgroups from all three sites (Table 2). Transcript numbers for narG-1 were significantly higher at the Hythe (P < 0.05). Although transcript numbers were higher at Brightlingsea than at Alresford, this difference was not significant (P = 0.971). Highest numbers of narG-2 gene transcripts were found at Alresford (1.6-fold higher than at the Hythe site), although there were no significant differences between narG-2 transcript numbers along the estuary (P = 0.118). napA-3 transcript numbers at the Hythe were 1.6 times higher than those at Alresford (Table 2), although this difference was not significant (P = 0.597). napA-1 and napA-2 gene transcripts were not detected at any site. Gene transcripts were detected from two of the three nirS subgroups. nirS-e&f transcripts were detected in similar numbers at all three sites (Table 2), with no significant difference between transcript numbers (P = 0.227). nirS-n transcripts were detected and quantified in Hythe sediments only (3.29 × 10⁴ ± 2.22 × 10⁴), and nirS-m transcripts were not detected at any site. Finally, nrfA-2 transcripts were not detected in the sediment from any of the sites in October 2005.

TABLE 2.

Nitrate (narG and napA) and nitrite (nirS) reductase gene transcript numbers from triplicate Colne estuary sediment samples, October 2005

Site	Transcripts g−1 sediment (mean ± SE)a
	narG		napA (napA-3)	nirS
	narG-1	narG-2		nirS-e&f	nirS-n
Hythe	1.58 × 106 ± 9.17 × 105a	2.53 × 105 ± 7.08 × 104	6.07 × 103 ± 6.07 × 103	2.28 × 106 ± 2.24 × 105	3.29 × 104 ± 2.22 × 104
Alresford	8.36 × 104 ± 5.83 × 104b	4.16 × 105 ± 8.16 × 104	3.87 × 103 ± 1.96 × 103	2.45 × 106 ± 1.23 × 105	NDb
Brightlingsea	1.87 × 105 ± 1.72 × 105b	2.35 × 105 ± 6.35 × 104	ND	1.52 × 106 ± 8.87 × 105	ND

^anapA-1, napA-2, nirS-m, and nrfA-2 transcript numbers were below detection levels at all three sites. Statistical differences (P < 0.05) are indicated by different letters for narG-1. No significant differences were found between sites for other phylotypes (see text for details). Standard curve descriptors and detection levels: for narG-1, r² = 0.997, y intercept = 37.34, E (amplification efficiency) = 89.2%, and C_T cutoff of 35.73 equivalent; for narG-2, r² = 0.993, y intercept = 40.20, E = 89.2%, and NTC undetected; for napA-3: r² = 0.997, y intercept = 37.86, E = 88.6%, and NTC undetected; for nirS-e&f, r² = 0.996, y intercept = 43.56, E = 94.2%, and NTC undetected; and for nirS-n, r² = 0.997, y intercept = 38.34, E = 90.3%, and NTC undetected.

^bND, not detected.

DISCUSSION

This study has, for the first time, investigated the diversity and abundance of genes (narG, napA, and nrfA) encoding nitrate reduction in estuarine sediments. Moreover, this represents the most comprehensive investigation of napA and nrfA gene diversity in the natural environment undertaken to date. The nitrate-reducing community in the Colne estuary has been found to be both phylogenetically diverse and comprised predominantly of sequences that are only distantly related to those from cultured nitrate-reducing bacteria (9, 24, 26), as was similarly found in soils and marine sediments (6, 8, 10, 14, 15, 31, 47).

For narG, the dominant cluster (Fig. 2, cluster I) showed highest similarity to narG from the aryl-halorespiring delta-proteobacterium Anaeormyxobacter delahogens 2CP-C, which was originally isolated from a Cameroonian rain forest soil and is capable of reducing nitrate to ammonium (51). Earlier studies of narG diversity in soil environments had previously identified sequences related to those from Actinobacteria or other proteobacterial subclasses (alpha-, beta-, or gammaproteobacteria) (14, 31, 47), suggesting that nitrate reduction in estuarine and soil environments may be mediated by different phylogenetic groups. For napA, sequences from the Colne estuary were related to napA genes found in the genome sequences of a diverse selection of alpha-, beta-, and gamma-proteobacteria (Fig. 3). These sequences were all divergent from those found in the only prior investigation of napA diversity in a natural environment (freshwater sediment) (23), suggesting that there are diverse groups of nitrate-reducing bacteria that utilize a napA-encoded nitrate reductase. The majority of nrfA sequences (50%) clustered with nrfA genes from delta-proteobacteria (Fig. 4). The dominance of delta-proteobacterium-related nrfA sequences and potentially also sequences in the narG clone libraries indicates that delta-proteobacteria may play a major role in nitrate reduction in estuaries. Other nrfA sequences from the Colne sediments were related to those found previously in anammox reactors (Fig. 4, cluster I), indicating the potential presence of anammox bacteria in the Colne estuary. Anammox has previously been shown to occur in sediments in the nearby Thames estuary (61).

Having established that there was substantial heterogeneity within the nitrate-reducing communities within the Colne sediments, a suite of Q-PCR assays was developed to determine the abundance and expression of genes encoding nitrate reduction functions in sediments along the estuary. The narG, napA, and nrfA gene phylogenies, together with that from our earlier study describing nirS mRNA sequence variation (43), showed that the nitrate-reducing bacterial communities in the Colne estuary are distinct from those in other (soil and marine) environments. Consequently, PCR primer and TaqMan probe combinations were developed to target and quantify dominant phylotypes in this environment, as opposed to using the existing Q-PCR primer suites developed for other environments (28, 29, 37). In general, a statistically significant decline (Fig. 5) was observed in gene copy numbers (narG, napA, nirS, and nrfA) along the estuary from the Hythe to Brightlingsea (from ∼10⁸ to 10⁵ to ∼10⁷ to 10⁴ genes g⁻¹ sediment, respectively), with the exception of the napA-2 and napA-3 phylotypes. Henry et al. (30) previously reported narG numbers of between 10⁶ and 10⁸ gene copies g⁻¹ from a variety of soil systems. Direct comparisons between these data sets and those generated in this study should be treated cautiously, due to differences between the extractions methods, Q-PCR primer sets, and Q-PCR chemistries that were used in the two studies. The reduction in gene copy numbers mirrored the declining concentrations of nitrate in the sediment and a reduction in rates of denitrification and DNRA from the head to the mouth of the estuary (17, 18).

In our earlier study, nirS mRNA transcripts were detected by RT-PCR in the Hythe and Alresford sediments (43). Isolation of mRNA directly from complex environmental samples can be problematic, as mRNA is a labile molecule with a potentially very short half life (27). The application of Q-RT-PCR to environmental samples has been previously limited to quantification of bacterial gene expression in water (21, 32, 42, 64). In this study, for the first time, the numbers of mRNA transcripts have been quantified by Q-RT-PCR from three functional genes within sediments. While gene transcript numbers in the sediment samples were low (≤10⁶ transcripts g⁻¹ sediment, if detected), transcript numbers for three of the five phylotypes for which transcripts numbers could be determined, were highest at the estuary head, following the same general trend found for gene numbers (Table 2 and Fig. 5). Furthermore, this study also provided evidence for temporal stability in the active nitrate-reducing communities in the Colne, as nirS-e&f and nirS-n mRNA sequences had been found previously in the Colne estuary sediments in January 2001 (43), with these same transcripts detected and quantified by Q-RT-PCR in the sediments over 4 years later (Table 2).

In conclusion, we have shown that the nitrate- and nitrite-reducing communities present in sediments from the Colne estuary are phylogenetically diverse and also divergent both from those in other environments and from cultured bacteria that are considered models for nitrate reduction. In the present research, we have also developed and applied Q-(RT)-PCR to quantify specific nitrate/nitrite-reducing phylotypes along the Colne estuary's nitrate gradient and have demonstrated that gene copy and transcript numbers are, in general, greatest at the estuary head (Hythe), where the rates of denitrification/DNRA are highest.