Direct molecular biological analysis of ammonia oxidising bacteria populations in cultivated soil plots treated with swine manure


Corresponding author. Tel.: +44 (151) 7944413; fax: +44 (151) 7944401; e-mail:


The application of pig slurry, high in ammonia-nitrogen concentration, to agricultural land is a practice whose effect on soil microbial communities is poorly understood. The autotrophic ammonia-oxidising bacteria are an integral component of the nitrogen cycle in soil, and their activity will be affected by addition of nitrogenous fertilizer. Molecular biological techniques were applied to the direct detection and identification of ammonia-oxidiser populations in cultivated soil plots treated with different amounts of pig slurry. Members of the genus Nitrosospira were shown by 16S rDNA-directed PCR to be present in both unamended and amended soils, regardless of the quantity of pig slurry applied. In contrast, members of the genus Nitrosomonas were detected by the same approach only in those soil plots that had received high loadings of slurry. The fidelity of amplification products was always confirmed by oligonucleotide probing. In addition, we used high stringency PCR and confirmatory gene probing to detect the presence of the ammonia monooxygenase gene (amoA) of Nitrosomonas europaea directly in all amended soil samples, with hybridization signal intensities that increased with the amount of pig slurry applied to plots. Nitrosomonas europaea amoA DNA could not be detected in soil from the untreated plot. These data support the view that nitrosospiras are ubiquitous as important members of nitrifying populations in the environment. The direct detection of nitrosomonad DNA only in amended soils supports the hypothesis that these nitrosomonads become highly competitive under conditions analogous to laboratory enrichment cultures.


The biological oxidation of ammonia is the primary step in environmental nitrification and is largely mediated by chemolithotrophic bacteria that assimilate carbon dioxide via the Calvin cycle [1]. These organisms are difficult to isolate and cultivate, such that the application of molecular biological methods to elucidation of their population structure and activity in situ has become an important research topic. Chemolithotrophic ammonia-oxidisers belong to the β and γ subdivisions of the Proteobacteria [2, 3]; near complete 16S rDNA sequence analysis led to the reclassification of the β-subdivision species into two genera, Nitrosomonas and Nitrosospira, the latter encompassing organisms previously classified as ‘Nitrosolobus’ and ‘Nitrosovibrio[4]. There are fewer examples of γ-subdivision ammonia-oxidisers, and these are currently viewed as a group of marine bacteria to which the genus name Nitrosococcus can be applied [4, 5]. The taxonomic framework of the β-subdivision ammonia-oxidisers has been expanded on the basis of 16S rDNA sequences recovered from both environmental samples and from enrichment cultures [6–9]. Interrogation of sequence information has provided sets of PCR amplification primers and confirmatory oligonucleotide probes for direct detection of these bacteria in various environments which has in some cases provided information on the abundance and activity of different groups [7, 10–13]. The phylogenetic coherence of the β-subdivision ammonia-oxidisers lends itself to the application of 16S rDNA or rRNA as a utilisable target. In addition, determination of the DNA sequence of ammonia monooxygenase (AMO) in Nitrosomonas europaea[14] has provided a functional gene target for direct detection of ammonia-oxidisers. amoA-Directed primers have been designed from the single published AMO gene sequence of Nm. europaea to amplify virtually the complete nucleotide sequence of AMO. This gene is believed to encode the 27-kDa functional component (amoA) of the AMO enzyme [15]. More information on amoA sequence variation amongst ammonia-oxidisers has recently been produced [16, 17] and amplification and recovery of this gene from environmental samples has been achieved [18].

The application of molecular biological techniques to microbial ecology can reveal a population structure that is not apparent from studies based on the recovery of viable organisms. It may proceed via in situ oligonucleotide probing to differentiate active and inactive components and expose spatial relationships between species [19], including nitrifying bacteria [13]. These techniques can also be applied to study changes in population structure that result from environmental perturbation, which is the aim of the work reported here. In soil, the amount of ammonium available to microorganisms is limited and nitrifier activity depends on the decomposition of organic nitrogen compounds [20]. Swine waste has a high ammonia-nitrogen content and its application to land might be expected to stimulate nitrifier activity. Ammonia oxidation and nitrification of swine waste has been studied using conventional (MPN and nitrogen determination) methods [21, 22], but not after introduction to soil. Application of swine manure as fertilizer during intensive agricultural practices is increasing, with the possible side effects of environmental pollution, loss of nitrogen and changes in the indigenous microbial community. High ammonium levels in swine waste have been eliminated by spreading on land, but in circumstances where the available land cannot contain all of the waste applied, water courses are often polluted.

In this paper, we report the application of molecular biological techniques for the detection of autotrophic ammonia-oxidisers in cultivated soil plots subjected to different regimes of pig slurry application. Both 16S rRNA and AMO genes are used as targets to provide information on the relative abundance of nitrosospiras and nitrosomonads under different environmental conditions.

2Materials and methods

2.1Bacterial strains

Pure cultures of ammonia-oxidising bacteria were maintained in the medium described by Watson and Mandel [23], with adjustment of salinity as required. Heterotrophic bacteria were grown in L.B. broth (Oxoid, Basingstoke, UK) incubated at the growth temperature appropriate to each strain. The strains of ammonia-oxidising bacteria used in this study were obtained from the University of Liverpool Culture Collection and were Nitrosomonas europaea (C-31; strain A and strain C), Nm. eutropha (C-91), Nm. marina (C-56), Nitrosospira multiformis (C-71), Nsp. briensis (C-128), Nsp. tenuis (Nv-1), Nitrosospira sp. (Nv12), Nitrosospira sp. (Nv141), Nitrosococcus mobilis (Nc-2) and Nc. oceanus (Nc-27). The origins and further details on these strains are given by Head et al. [4] with the exception of Nm. europaea strains A and C which are authentic strains of uncertain origin, used here as positive controls in some experiments. In addition, strains representing the following species, also obtained from the University of Liverpool Culture Collection, were included as controls where appropriate: Escherichia coli, Proteus vulgaris, Morganella morganii, Bacillus subtilis, Aeromonas hydrophila, A. salmonicida, Pseudomonas putida, Ps. stutzeri and Ps. fluorescens. For DNA preparations, cultures were grown to late exponential phase, cells harvested by centrifugation and pellets lysed by the freeze-thaw method [4]. Chromosomal DNA preparations from the methanotrophic bacteria Methylosinus trichosporium, Methylobacter albus, Methylococcus capsulatus and Methylocystis parvus were kindly supplied by Dr. A.J. Holmes, University of Warwick.

2.2Soil samples

Soil samples were obtained from the San Prospero Experimental Centre of I.S.A. Modena, Italy. Experimental corn cultivated plots (11.2 m by 20 m) were treated with different amounts of stored pig slurry; the quantities and composition of swine manure applied are given in Table 1. Pig slurry applied to the soil had been stabilized by storage for six months, then spread on the surface soil prior to seed-bed preparation and sowing. Plots were sown with corn two days after the slurry application, and samples collected seven days after sowing (May, 1994) at points along a diagonal of each plot and at 0–15 cm depth. The soil is alluvial and classified as Vertic-Calcic Gleyic Cambisol 3a (FAO); the soil composition is: N tot.% 1.88; organic matter % 2.43; C/N: 7.50; P tot. (% P2O5): 24.6; K exchangeable (ppm K2O): 389; pH (H2O): 7.9 [24].

Table 1.  Details of experimental plots used to provide soil samples for DNA extractions
PlotPig slurryDry matterN tot.P tot.K tot.Cu tot.Zn tot.
  1. The method used to determine total N content of the soils was that of Kjeldahl by the Kjeltec Auto 1030 Analyzer Tecato [27], and other minerals by the method of Olsen and Sommers [28].


2.3Oligonucleotide primers and probes

Oligonucleotides used as amplification primers and probes, their sequences, target genes and corresponding regions on their target genes are detailed in Table 2.

Table 2.  Oligonucleotides used for PCR amplifications and probe hybridizations of 16S rDNA and amoA genes of ammonia-oxidising bacteria
PrimerSequence (5′-3′)Target gene
  1. Oligonucleotides are designated forward (f) or reverse (r) according to their complementarity and numbers in parentheses give the base numbers in the corresponding E. coli sequence [29]. Numbers associated with amoA-directed primers are the bases from the beginning of the gene [14].

pFr (1073–1053)ACGAGCTGACGACAGCCATGEubacterial 16S rDNA
pHr (1542–1522)AAGGAGGTGATCCAGCCGCAEubacterial 16S rDNA
ProbeSequence (5′-3′)Target gene
AAO258fGCCTTGGTAAGCCTTTACCNon-marine ammonia-oxidiser 16S rDNA
Nlm459rACGGTTAATACCCGTGACTANsp. multiformis and
Nitrosospira sp. (Nv141) 16S rDNA

2.4Extraction and amplification of DNA from soil samples

The protocol of Bruce et al. [25] was used for the extraction of DNA from soil samples. Humic substances were removed from DNA preparations using the polyvinylpyrolidone-agarose gel electrophoresis method of Young et al. [26]. All reactions were performed in 0.2 ml thin-walled MJ Research microtubes under a layer of sterile paraffin oil in a PTC-100 thermal cycler (MJ Research, Watertown, MA, USA). Reaction mixtures (100 μl) contained: 1×PCR buffer (100 mM Tris-HCl [pH 8.8], 25 mM (NH4)2SO4, 0.015% Tween 20, 1.5 mM MgCl2), 200 μM each deoxynucleoside triphosphate, 20 pM each primer, 4 μl (10 ng) of template DNA, and 1 U of Taq polymerase (Polymed Biotechnology Division, Firenze, Italy). The reaction thermal profile for amplification with eubacterial 16S rDNA-directed primers was: 95°C for 7 min then 80°C for the addition of Taq polymerase followed by 25 cycles of 94°C for 1 min, 55°C for 1 min and 72°C for 2 min. A 15 min extension at 72°C was performed after the final cycle. Nested primer amplification was performed on DNA products of eubacterial-specific PCR using primers Nm75f/1007r specific for Nitrosomonas spp. and primers Ns85f/1009r specific for Nitrosospira spp. Annealing temperatures of 63°C and 62°C were used for these genus-specific primers respectively, with denaturation and extension temperatures as described above. For nested amplification, products from the initial PCR were appropriately diluted in HiPerSolv water (BDH, Poole, UK) and used as template for a second round of thermal cycling employing primers that annealed internally to the original primers. In addition to 16S rDNA-directed amplification, DNA preparations were also subjected to an amoA-directed, nested PCR employing first primers AMOF1/R2, then AMOF2/R2R using annealing temperatures of 57°C and 55°C respectively, with other thermal parameters as described above. Amplification products were resolved by electrophoresis of 10-μl aliquots of the reaction mixtures on a 0.8% (w/v) horizontal agarose gel run in 1×TAE buffer (40 mM Tris-acetate, 1 mM EDTA).

2.516S rDNA Oligonucleotide probe hybridization

Oligonucleotide probes were end-labelled with digoxigenin and used for hybridization according to instructions for the Oligonucleotide 3′-End Labelling and Chemioluminescence Detection Kit (Boehringer Mannheim). Electrophoresed PCR products were transferred to Hybond N+ nylon membrane (Amersham, Bucks., UK) by Southern blotting according to the DIG System users guide for membrane hybridization (Boehringer). Membranes were prehybridized at 45°C for 60 min.

The prehybridization solution was 2% [w/v] Blocking Reagent solution (Boehringer Mannheim), 5×SSPE (20×SSPE is 3.6 M NaCl, 0.2 M NaPO4 [pH 7.7], 20 mM EDTA), 20% [v/v] deionized formamide, 0.02% [w/v] sodium dodecyl sulphate (SDS) and 0.1% [w/v] N-lauroyl sarcosine prepared in distilled water. At least 0.2 ml of the solution was used per cm2 of membrane. Hybridization of specific oligonucleotide probes (10 pM in 20 ml of hybridization solution (prehybridization solution with blocking reagent omitted) was at the appropriate temperature overnight. After hybridization, membranes were washed three times at hybridization temperature for 5 min in 100 ml of fresh hybridization solution, wrapped in cling-film and exposed to X-Ray film (Fuji RX-100) for an appropriate period.

2.6amoA gene probe labelling and hybridization

Approximately 50 ng of genomic DNA from a pure culture of Nm. europaea (ATCC 25978) was amplified under the stringent conditions detailed above, using primers AMOF2/R2R in a 50 μl reaction mix containing 2.5 μl (370 kBq μl−1) [α-32P] CTP (ICN Pharmaceuticals, Irvine, CA, USA). In order to determine adequate incorporation of radiolabel into synthesized DNA fragments, 1 μl of PCR products was electrophoresed in 0.8% (w/v) agarose, transferred to nylon membrane by Southern blotting and exposed to X-Ray film for 6 h at −80°C. Unincorporated [α-32P] CTP was removed from the remaining reaction mix using Sephadex G-50 NICKTM-Columns (Pharmacia Biotech), and radiolabelled amplification products heat denatured for use as an amoA gene probe. Membranes for gene probing were prehybridized in 25 ml of prehybridization solution (6×SSC [1×SSC is 0.15 M NaCl, 0.015 M sodium citrate, pH 7.0], 5× Denhardt's solution, 0.5% SDS made up to 25 ml with distilled water containing 0.5 ml of a 1 mg ml−1 solution of calf thymus DNA, denatured by heating to 100°C for 5 min and chilled on ice) for at least 1 h at 65°C with gentle shaking. Probe incubation was performed overnight in 100 ml hybridization buffer (6×SSC, 5× Denhardt's, 0.5% SDS) with shaking at 65°C. After hybridization, membranes were washed in 50 ml 2×SSC at 65°C for 15 min. A second wash was performed in 50 ml of 2×SSC containing 0.1% SDS at 65°C for 30 min, and a final wash in 50 ml of 0.1×SSC at 65°C for 10 min. Membranes were air-dried and exposed to X-Ray film for autoradiography.

3Results and discussion

3.1Pure culture DNA amplifications and oligonucleotide probe hybridizations

Ammonia-oxidiser oligonucleotide sequences (Table 2) were compared with sequences available on the RDP database [30] using the FASTA search program [31]. The suitability of primer pairs for specific gene amplifications was confirmed in control reactions with genomic DNA from pure cultures of reference strains. Table 3 lists amplification products obtained from various combinations of control DNA preparations and the PCR primers used in this study. Amplifications were considered specific when product size was as expected and hybridization was obtained to a suitable internal oligonucleotide probe. Hybridization temperatures for probes were; Nlm459r, Nm75r and Ns85r (45°C); AAO258f (55°C). Wash temperatures were the same as hybridization temperatures for each probe [32]. Under these hybridization conditions, each oligonucleotide was shown to be specific for its target DNA, and non-specific hybridization was not observed with any reference strain DNA tested. Thus, DNA preparations from the heterotrophic and methanotrophic species listed in methods (Section 2.1) gave amplification products with eubacterial rDNA primers, but not with any of the nitrifier-specific primers. The amplification product and oligonucleotide hybridization patterns obtained with ammonia-oxidiser strains were as anticipated [4, 10]. We use oligonucleotide AAO258 as a confirmatory internal probe for nitrifier rDNA amplified with specific primers on the basis of its universal occurrence in the non-marine β-subdivision ammonia-oxidiser cultures originally sequenced [4] The utility of this probe is supported by the work of Stephen et al. [9] who stated that AAO258 performed well when challenged with new sequence data from soil samples.

Table 3.  Results of DNA amplifications and hybridizations with the oligonucleotides used as PCR primers and probes in this study against reference and environmental DNA preparations
OrganismPCR amplificationProbe hybridization
EUBa primersNm. primersNs. primersamoAb primersAAO258fcNm75rcNs85rcNlm459rd
  1. aControl strain eubacterial 16S rDNA amplification results were obtained using primers pAf/pHr and pBf/pFr. Soil eubacterial 16S rDNA amplification results refer to primers pBf/pFr only.

  2. bResults of PCR with primers AMOF1/R2 and F2/R2R were according to the stringent amplification conditions detailed in the methods section.

  3. cApplication of these oligonucleotide probes was to amplification products from eubacterial PCR primers pBf/pFr.

  4. dApplication of this oligonucleotide probe was to amplification products from primers Ns85f/1009r.

Nm. europaea+++++
Nm. eutropha++++
Nm. marina+
Ns. multiformis+++++
Ns. briensis++++
Ns. tenuis (Nv1)++++
Ns. sp. (Nv12)++++
Ns. sp. (Nv141)+++++
Nc. mobilis+
Nc. oceanus+
Soil sample L1++++
Soil sample L2++++
Soil sample L3++++++
Soil sample L4++++++

Under the stringent amplification conditions detailed in 2.4, both amoA primer pairs (Table 2) demonstrated specificity for Nm. europaea DNA, i.e. an amplification product of the predicted size that also hybridized to the amoA gene probe was obtained, but not from any of the other nitrifiers or the methanotrophs. Using a low stringency annealing temperature of 37°C, primer pairs AMOF1/R2 and AMOF2/R2R yielded amplification products of the correct size from DNA of some other β-subdivision ammonia-oxidisers. By the same PCR, Methylocystis parvus DNA yielded similar-sized amplification products by primer pair AMOF1/R2, but was the only strain of methanotroph where this was observed (data not shown). As the genes encoding ammonia monooxygenase in ammonia-oxidising bacteria and particulate methane monooxygenase (pMMO) in methanotrophic bacteria have been shown to share high sequence identity [15], it is probable that this low-stringency amoA-directed PCR amplified the pmoA gene of Mcy. parvus. While these data support observations of sequence homology amongst ammonia and methane monooxygenase genes [15, 18], we have demonstrated that at high stringency, there is sufficient specificity to amplify only Nm. europaea amoA DNA with confidence such that it can be used as a target for direct detection of this species and its close relatives in environmental samples.

3.2Amplification of 16S rDNA from soil samples

DNA extracted from the soil samples (plots L1–L4), was transferred to the PCR mix by the ‘bandstab’ method [33], subsequent to PVP-agarose gel purification. Amplification of 16S rDNA from this template DNA using the eubacteria-specific primers pAf and pHr failed to yield visible products upon agarose gel electrophoresis. This initial attempt at PCR amplification was considered unsuccessful due to the low concentrations of template DNA retrieved from PVP-gel electrophoresis in this way. A nested PCR employing primers pBf and pFr on pAf/pHr amplified DNA, diluted (1:10) was successful in yielding strong bands of eubacterial 16S rDNA (Fig. 1a). Products from these amplifications were transferred to nylon membrane by Southern blotting and hybridized with oligonucleotide AAO258f in order to demonstrate the presence of ammonia-oxidiser 16S rDNA in the eubacterial PCR products. Positive signals were obtained from all soil samples (Fig. 1b). The same eubacterial PCR products, after the removal of oligonucleotide AAO258f, were reprobed with the ammonia-oxidiser genus-specific oligonucleotides Nm75r and Ns85r, in an attempt to differentiate the group-specific signal of AA0258f into its constituent genera, Nitrosomonas and Nitrosospira. Only signals from probe Ns85r were demonstrable in all cases, indicating the presence of Nitrosospira 16S rDNA (Fig. 1c). No hybridization signals were obtained from these amplification products with probe Nm75r. This was a first indication of Nitrosospira ubiquity in these soil samples, as observed previously in other environments using this approach [10].

Figure 1.

Agarose gel (a) and Southern blot hybridisation (b and c) analysis of 16S rDNA PCR products amplified from soil DNA. Amplification primers used to generate products shown in Fig. 1 a were pB and pF′. Lanes: 1, plot L1; 2, plot L2; 3, plot L3; 4, plot L4. Lanes in the Southern blots correspond to those in the gel. Oligonucleotide hybridisations were with probes AAO258 (b) and Ns85′ (c).

These pBf/pFr amplification products were diluted 1:100 and used as template for two separate nested amplifications employing primer pairs Nm75f/1007r and Ns85f/1009r. Amplification products from these reactions were obtained in all cases with the Nitrosospira-specific primers, but only from samples L3 and L4 with the Nitrosomonas-specific primers (Fig. 2a). Hybridization of oligonucleotide AAO258f with all these PCR products confirmed their identity as ammonia-oxidiser 16S rDNA (Fig. 2b).

Figure 2.

Agarose gel (a) and Southern blot hybridisation (b) analysis of 16S rDNA PCR products amplified from soil DNA. Amplification primers used to generate products shown in lanes 1 to 4 were Nm75 and Nm1007′. Primers used to generate products shown in lanes 5 to 8 were Ns85 and Ns1009′. Lanes: 1 and 5, plot L1; 2 and 6, plot L2; 3 and 7, plot L3; 4 and 8, plot L4. Lanes in the Southern blot correspond to those in the gel. Oligonucleotide hybridisation was with probe AAO258 (b).

This nested amplification approach again demonstrates the universal presence of nitrosospiras, but we have now been able to detect nitrosomonad DNA directly in those soil samples that received the higher loadings of swine waste. Previous work in this laboratory [10] failed directly to detect nitrosomonad DNA in environmental samples, although such DNA could normally be recovered from enrichment cultures. Despite this observation it has been suggested that the failure to detect Nitrosomonas DNA was due to insufficient lysis [13] or inappropriate primers [34]. The data presented in this study enable us to refute these suggestions. Therefore, we argue that the soil plots that received nutrient input function analogously to laboratory enrichment cultures in that nitrosomonad growth is encouraged to the point where sufficient DNA template becomes available for detection by PCR. Nitrosomonad DNA has been directly detected in environmental samples previously [13], but in that case activated sludge samples were examined, again representative of an environment where nutritional enrichment encourages proliferation of Nitrosomonas spp. Although the classification of microbial populations in soil as autochthonous or zymogenous in relation to nutrient availability is equivocal [35, 36], the molecular genetic data reported here and elsewhere [12, 15] support the view that nitrosomonad populations are markedly responsive to ammonia inputs [37]. It is unlikely that detection of nitrosomonad DNA in these amended soils is due simply to inoculation with stabilized swine manure that may itself contain a population of nitrosomonads. Nevertheless, the ubiquitous detection of nitrosospiras independent of nutrient amendation or enrichment is again confirmed.

Finally, we have not attempted to examine the 16S rDNA sequence diversity within the nitrosospira group detected in these soils, but have previously been able to detect routinely rDNA of some cultured soil Nitrosospira strains [38] in freshwater samples by application of a specific probe (unpublished data). It was therefore of interest to probe the products of Nitrosospira 16S rDNA amplification (Ns85f/Ns1009r) from all four soils with this oligonucleotide Nlm459r, and we can report that no signals were obtained.

3.3Amplification of ammonia monooxygenase DNA from soils

Recovery and analysis of 16S rDNA from soil provides information on the composition of phylogenetically coherent microbial groups such as the β-subdivision ammonia-oxidisers, but it is also desirable to apply functional gene targets to such studies. We demonstrated in Section 3.1 that the ammonia monooxygenase gene of Nm. europaea can be selectively amplified by PCR, provided that stringent conditions are used. amoA directed amplification was therefore applied to the soil samples. No amplification products (visible on agarose gels) were obtained from DNA extracted from any soil sample after an initial round of thermal cycling using primer pair AMOF1/R2. The putative DNA from this initial PCR amplification was diluted (1:10) with HiPerSolv water, and subjected to a nested PCR using primer pair AMOF2/R2R. Visible amplification products of the expected size were obtained, but only from soil samples L3 and L4 (Fig. 3a). Further evidence for the fidelity of these products was provided by the observation of strong hybridisation signals in Southern blots with the labelled 1.8-kb amoA gene as probe (Fig. 3b). In addition, amended soil sample L2 from which visible amoA amplification products were not obtained, gave a weak hybridisation signal (Fig. 3b) confirming the increased sensitivity of gene probing over ethidium bromide staining in the application of PCR to DNA recovery from the environment. Therefore, the detection of nitrosomonad DNA only in soil plots to which swine waste had been added is established using two independent gene targets. Furthermore, increase in both agarose gel band and hybridisation signal intensity (Fig. 3b) in proportion to the amount of swine waste added to the plots supports the argument for nitrosomonad enrichment in situ.

Figure 3.

Agarose gel (a) and Southern blot hybridisation (b) analysis of amoA PCR products amplified form soil DNA. Primers used in (a) were AMOF2 and AMOR2R. Lanes: 1, plot L1; 2, plot L2; 3, plot L3; 4, plot L4; 5, Nm. europaea strain A; 6, Nm. europaea strain C. Lanes in the Southern blot correspond to those in the gel. Hybridisation was with the amoA gene probe.


This work was supported by the Italian Ministry of Agriculture (MiRAAF), Finalized Project PANDA subproject 3, Series 1, Paper No.43 and the U.K. Natural Environment Research Council. M.T. Ceccherini was supported by a visiting fellowship from the European Environment Research Organisation.