Aerobic nitrate respiration in a nitrite-oxidising bioreactor


*Corresponding author. Tel.: +61 (7) 3365 4878; Fax: +61 (7) 3365 4620, E-mail address:


The ability of heterotrophic bacteria in a nitrite-oxidising bioreactor to respire with nitrate as an electron acceptor was examined. Approximately 70% of 1000 heterotrophic isolates were able to express a nitrate reductase. A detailed survey of 15 isolates showed that five expressed the azide-insensitive nitrate reductase encoded by the napA gene. A two-round PCR amplification of the napA gene using degenerate PCR primers and DNA sequence analysis of these products confirmed the presence of this gene in the positive isolates. Partial 16S rDNA products and napA products were amplified from the biomass in the bioreactor and denaturing gradient gel electrophoresis of these products identified 21 distinct ribotypes and 12 distinct napA sequences. The results show that the ability to respire with nitrate as an electron acceptor under aerobic conditions is widespread among the heterotrophic population of this bioreactor.


Over the last decade it has been established that nitrate can be used as an electron acceptor by bacteria under anoxic conditions and in the presence of oxygen (reviewed in [1]). The molecular properties of the nitrate respiratory systems involved in these processes have been described in detail [2]. Under anaerobic conditions a membrane-bound nitrate reductase, exemplified by the enzymes from Escherichia coli and Paracoccus denitrificans, dominates. This enzyme has an active site that faces the cytoplasm; electron flow from quinol to nitrate is linked to generation of a proton electrochemical gradient and so the membrane-bound nitrate reductase has a clear role in energy generation. The nar operon encodes this membrane-bound nitrate reductase and its expression requires the oxygen-responsive transcriptional regulator Fnr; even when nitrate is available under aerobic conditions the level of expression of the nar operon is low because Fnr is not active [1]. A nitrate respiratory system that is produced under aerobic conditions and is active in the presence of oxygen has been characterised in P. denitrificans strain GB17 (formerly known as Thiosphera pantotropha) [3]. It is composed of a periplasmic nitrate reductase that is connected to the quinone pool of the respiratory chain via c-type cytochromes [3]. This periplasmic nitrate respiratory system is encoded by the nap gene cluster (EMBL Z36773 for P. denitrificans). Very little is known about the regulation of expression of the nap gene cluster and the physiological function of the periplasmic nitrate respiratory system is uncertain. However, there is increasing evidence that the nap genes are widely distributed amongst bacterial species [4].

Recently, a nested PCR for the amplification of a specific napA sequence was developed [5] and used to analyse DNA extracted from a fresh water community. In this paper we describe an analysis of heterotrophic bacteria growing in a nitrite-oxidising bioreactor. The presence of the periplasmic nitrate reductase in individual isolates and in the bacterial community was examined using a combination of biochemical and molecular methods, including analysis of napA sequences.

2Materials and methods

2.1Bacterial strains and culture conditions

P. denitrificans strain GB17 (also known as T. pantotropha) and Rhodobacter sphaeroides strain 2.4.1 were grown on RCV medium supplemented with vitamins required for growth of R. sphaeroides[6]. A nitrite-oxidising sequencing batch reactor and its operation has been described previously [7]. This bioreactor was the source of bacterial isolates that were used in this study. Bacterial isolates were cultured on nutrient agar (Oxoid) or solid TYS media [8].

2.2Molecular biological methods

DNA was released from individual bacterial isolates by collecting a small amount of biomass in a microcentrifuge tube containing 10 μl of 10×Tth plus reaction buffer, 6 μl 1.5 mM MgCl2 and 60 μl of sterile MilliQ water. This reaction mix was overlayed with about 20 μl of mineral oil. The reaction mix was then heated to 96°C for up to 30 min to lyse the cells and release DNA. Amplification of 16S rRNA genes was carried out essentially as described in [7] using the bacterial conserved primers 27f and 1492r [9] in a PCR. For amplification of napA DNA a two-round PCR was used. In the first round DNA was denatured at 96°C for 120 s and the samples were then amplified for 36 cycles of 60 s at 55°C, 120 s at 72°C and 60 s at 94°C, before a final extension step of 600 s at 72°C. The second round of PCR required 3 μl of first PCR product to be added to a PCR reaction mix, denatured for 120 s at 96°C and amplified for 42 cycles of 60 s at 60°C, 120 s at 72°C and 60 s at 94°C before a final extension step of 600 s at 72°C. The primer sequences used are shown in Table 1 and have been described in [5].

Table 1. napA primer sequences
  1. aThe codes for the degenerate bases employed were D (A, G, or T); H (A, C, or T); R (A or G); Y (C or T); M (A or C); S (G or C), K (G or T).

PrimerSequenceaPosition in napA nucleotide sequence

Amplicons were purified with the Wizard PCR Cleanup Kit and sequenced with the ABI dideoxy sequencing kit according to the manufacturers’ instructions and with primer 530f for 16S rDNA sequences [10] and V66 for napA sequences (Table 1). The partial 16S rDNA and napA sequences were compared with those on accessible databases by using the program basic local alignment search tool (BLAST) [11].

2.3Denaturing gradient gel electrophoresis (DGGE)

Amplification of DNA for DGGE analysis was performed essentially as described above except that the V66 and 27f primers that were used had a GC-clamp attached to the 5′ end (Table 1) in conjunction with normal V67 and 907r [9] primers respectively. Double-stranded DNA fragments were separated on a denaturing gradient acrylamide gel. The denaturing gradient gel had a vertical denaturant gradient concentration, of urea and formamide, of 10% to 70% and contained 0.01% (w/v) ammonium persulfate and 0.001% (v/v) TEMED. A stacking gel of 0% denaturant was layered on the top of the denaturing gel. Samples were electrophoresed in 1×TAE buffer at 60°C at 200 V for 6 h in a Denaturing Gel Electrophoresis Tank Model #2000 (CBS Scientific Co., USA). The fragment banding patterns were visualised by silver staining.

2.4Nitrite overlay procedure

Cultures of organisms were grown overnight at 28°C on modified TYS solid medium containing 20 mM NaNO3. The plates were first overlayed with 200 μl 25% (w/v) NaNO3 in 1.5 ml of molten agar and incubated aerobically for 30 min. The plates were then overlayed with 2 ml of 4% (w/v) sulfanilamide, 1 ml of 0.08% (w/v) N-1-NED, and 3 ml of 1.5% (w/v) molten agar. The plates were then observed for a pink/purple colour change indicating a positive reaction for nitrate reductase activity. This procedure was also modified to detect only the periplasmic nitrate reductase. This was achieved by incorporating 100 mM NaN3 into the first overlay which was then incubated for 30 min aerobically. The second overlay was then performed as described above and the plates observed for the development of colour.

2.5Biochemical methods

Pure cultures were grown to mid to late log phase and then pelleted by centrifugation at 4000×g for 10 min at 4°C. the cells were resuspended in 5 ml of phosphate buffer (20 mM NaH2PO4 and 1 mM Na2EDTA, pH 7.2) and centrifuged for a further 10 min at 4°C. The pellet was resuspended in 5 ml of phosphate buffer and passed twice through a French press pressure cell at approximately 700 Psi, unless otherwise stated, to break the cell walls. The unbroken cells and large cell fragments were removed by centrifugation at 4000×g for 10 min. The cell-free extracts were then stored at −20°C. Nitrate reductase activity was assayed by measuring the nitrate-dependent oxidation of dithionite-reduced methylviologen in a spectrophotometer [12] and protein concentration was determined as in [13].


3.1Identification of nitrate respiratory activity in heterotrophic isolates

Heterotrophic bacteria were isolated from the nitrite-oxidising bioreactor by growth on nutrient agar plates. 1000 colonies were randomly selected and tested for the presence of nitrate reductase activity using the nitrite overlay procedure. 713 of these isolates tested positive for nitrate respiratory activity. A more detailed study was carried out on 15 isolates whose colonies appeared to be morphologically distinct. The nitrite overlay procedure was refined to allow a reaction with only the periplasmic nitrate reductase, as it is insensitive to micromolar concentrations of azide whilst the membrane-bound nitrate reductase is inhibited. Table 2 shows that nitrite was produced by 13 of the 15 isolates, indicating the presence of one or both nitrate reductases. In isolates 5, 10, 11, 13 and 14 the nitrate reductase was inhibited by azide, indicating that these isolates possess the membrane-bound nitrate reductase alone. Azide did not inhibit nitrite production in isolates 1, 2, 4, 7, 8, 9, 12 and 15, indicating the presence of the periplasmic nitrate reductase, although the presence of the membrane-bound nitrate reductase could not be ruled out.

Table 2.  Identification of isolates from the nitrite-oxidising bioreactor
  1. aIdentification of the presence of the indicated nitrate reductase is denoted as positive (+) whilst no detected nitrate reductase by the specified enzyme is recorded as a negative result (−).

  2. bNumber of nucleotides compared.

  3. cAccession number of the closest sequence by BLAST.

IsolateBLAST Identity by 16S rDNA analysisBLAST matchNaraNapa16S rDNA nucleotidesbAccession no.c
1Unclassified gamma proteobacterium98%+300UG12217
2Pseudomonas species99%++275PSD103
3Alcaligenes radioresistens99%350AR16SRRNB
4Unclassified gamma proteobacterium97%++350UG12217
5Unclassified gamma proteobacterium91%+285UG12217
6Unclassified gamma proteobacterium97%345UG12217
7Unidentified beta proteobacterium95%+300UB34035
9Pseudomonas cichorii97%++360PCZ76658
10Stenotrophomonas africae97%+133SAU62647
11Unclassified gamma proteobacterium94%+244UG12217
12Acinetobacter species100%++127ASU87126
14Acinetobacter species97%+340ASU87119
15Pseudomonas cichorii100%++306PCZ76658

The periplasmic nitrate reductase has been identified in a phylogenetically diverse range of proteobacteria [4]. Thus, it was of interest to determine identity of those bacteria that had been isolated from the bioreactor. A part of the 16S rRNA gene from the isolates was amplified by PCR and sequenced. The results of the BLAST analysis are shown in Table 2. The majority of the isolates were gamma proteobacteria.

3.2Amplification and analysis of napA sequences in heterotrophic isolates

A nested PCR was used to amplify napA DNA, using the degenerate primers described in Section 2. The predicted size of this product, based on the sequence of P. denitrificans napA was 430 bp. Fig. 1 shows that a PCR product corresponding approximately to the predicted size of the napA PCR product was observed in four isolates (1, 4, 7 and 12). The putative napA PCR products were sequenced and the sequences from isolates were shown by BLAST analysis to have 72%, 88%, 80% and 84% sequence identity respectively to the napA gene from Pseudomonas sp. st. S3.6.

Figure 1.

Products arising from a two-round PCR reaction for the amplification of the napA gene on a 2% agarose gel. Lane 1, molecular mass marker (ΦX174 DNA HaeIII digest); lane 2, isolate 1; lane 3, isolate 4; lane 4, isolate 7; lane 5, isolate 12; lane 6, positive control; lane 7, negative control.

3.3Analysis of the diversity of napA and 16S rDNA sequences in the nitrite-oxidising bioreactor

The observation that a number of different heterotrophic isolates possessed the napA gene prompted us to investigate the diversity of napA sequences present in the bacterial population of the bioreactor. The diversity of napA genes was analysed using DGGE and DGGE analysis of partial 16S rRNA genes was also performed in order to ascertain the diversity of the bacterial population (ribotype diversity). Fig. 2, Lane B shows PCR products obtained using 16S rDNA primers resolved by DGGE. Twenty-one distinct 16S rDNA sequences were identified and this indicates the minimum number of bacterial species present in the bioreactor. The DGGE analysis of the napA gene within the bioreactor bacterial population is shown in Fig. 2, Lane A. This revealed 12 resolvable PCR products and indicated that slightly greater than half of the total bacterial population in the bioreactor possessed the napA gene.

Figure 2.

Diagrammatic representation of the banding pattern of napA genes (Lane A) and 16S rRNA genes (Lane B) as amplified by DGGE PCR from a sample of DNA extracted from an enriched nitrite-oxidising bioreactor.

3.4Nitrate reductase activity in the bacterial population of the bioreactor

The detection of azide-insensitive nitrate reductase activity in isolates by use of the modified nitrite overlay procedure together with the identification of nap genes in isolates and in the bioreactor via DGGE analysis clearly showed that nitrate respiration via the periplasmic nitrate reductase had the potential to take place in the nitrite-oxidising bioreactor. To test this possibility, the nitrate reductase activity of the biomass in the bioreactor was measured during an aerobic cycle of 12 h. Fig. 3 shows that azide-insensitive periplasmic nitrate reductase could be measured throughout the growth cycle. During this period all nitrite in the bioreactor was converted to nitrate.

Figure 3.

Nitrate reductase activity in samples taken from the nitrite-oxidising bioreactor over a 12 h time period. Nitrate reductase specific activity (▪), protein concentration (♦). (1 μkat defined as 1 μmol MV+ oxidised s−1.)


Over the last decade it has become established that the ability to respire using nitrate as an electron acceptor under aerobic conditions is widespread among bacteria [1]. In a recent survey 104 to 107 bacteria per g of soil and sediment were shown to be able to respire with nitrate under aerobic conditions [4]. Thus, it has become clear that the periplasmic nitrate respiratory system is a significant bioenergetic system in cells growing in aerobic environments. Unlike soils and sediments that can be subject to wide fluctuations in oxygen tension and nutrient levels we have investigated a system in which bacteria were grown in a relatively stable continuously aerobic environment. The presence of nitrite as the sole energy source ensured that nitrifying bacteria (mainly Nitrospira) were dominant within the bacterial population [7]. Despite the lack of added carbon, heterotrophic bacteria that were not nitrite-oxidisers were isolated from the bioreactor.

The ability to respire with nitrate was a property of about 70% of the culturable heterotrophs. In the detailed study of 15 isolates it was shown that five of these bacterial isolates possessed the membrane-bound nitrate respiratory system encoded by the nar operon while another eight possessed either the periplasmic nitrate reductase or both the nitrate respiratory systems. The observation that about 50% of the culturable heterotrophs in the bioreactor could express the periplasmic nitrate reductase is consistent with DGGE analysis of the diversity of napA sequences that could be amplified from DNA extracted from the bioreactor.

The ability to respire anaerobically using nitrate is a property of a wide variety of facultative aerobes [14,15]. In principle, the presence of anoxic microzones within flocs could have allowed the membrane-bound nitrate reductase to be expressed and anaerobic nitrate respiration to occur, even in the aerobic bioreactor. However, the lack of azide sensitivity of the nitrate activity in bacterial biomass directly isolated from the bioreactor suggests that anaerobic nitrate respiration is not significant in the bioreactor, although it is clear that within the population there are bacteria that have the potential to express this respiratory system. In contrast, it appears that aerobic nitrate respiratory system is expressed in the bacterial population of the bioreactor. However, the nitrate reduction rate of the periplasmic respiratory system must be relatively small in comparison to the rate of nitrite oxidation within the bioreactor because essentially all added nitrite was converted to nitrate.

The functional significance of nitrate respiration via the periplasmic nitrate reductase under aerobic conditions remains uncertain. It has been suggested that co-respiration of nitrate and oxygen might offer a physiological advantage under conditions of fluctuating oxygen tension and a related advantage might be the disposal of excess reducing equivalents generated during the catabolism of reduced carbon substrates [16]. Thus, it has been suggested that nitrate respiration by the periplasmic nitrate reductase may be important in environments that are rich in reduced carbon and oxygen limited [16]. Our data are not entirely consistent with this model since we have observed periplasmic nitrate reductase activity in a bacterial population that is carbon limited. The only carbon in the bioreactor available to heterotrophic bacteria must come from the nitrite-oxidising bacteria in the form of dead cell material of extracellular polymeric material, which constitutes part of the floc structure [17], or soluble carbon products secreted by the autotrophic bacteria [18]. In contrast to previous proposals this suggests that the periplasmic nitrate reductase may also be expressed under conditions of carbon limitation. Support for this view comes from the observation that in R. sphaeroides strain 2.4.1 (a strain which possesses only the periplasmic nitrate reductase) and in some of the 15 isolates used in this study the periplasmic nitrate reductase was active in colonies grown on minimal medium but not in colonies grown on rich medium (McDevitt, Horne, Blackall, McEwan, unpublished observations).


We thank Stephen Spiro and David Richardson for making the information of the nested primer design available.