• Denitrification;
  • Nitrogen fixation;
  • Soil DNA isolation;
  • Soil bacterium;
  • DNA probing with soil bacterium


  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and methods
  5. 3. Results
  6. 4. Discussion
  7. Acknowledgments
  8. References

The populations of N2-fixing and denitrifying bacteria in an acid forest soil near Cologne were characterized by gene probing. The DNA isolated from the soil for this purpose was suitable for DNA–DNA hybridization using 0.4–0.7-kb probes targeting denitrification enzymes, dinitrogenase reductase (nifH) and eubacterial 16S rRNA. The densitometrical comparison of band intensities obtained in these Southern hybridizations indicated that the highest number of total bacteria, of denitrifying and N2-fixing microorganisms always occurred in the upper (∼5 cm) soil layer. The concentration of all these organisms decreased in parallel with the soil depth. The soil investigated was rich in nitrate in all layers, and the availability of nitrate apparently did not govern the distribution of denitrifying and N2-fixing bacteria in this soil. Soil cores investigated in the laboratory formed N2O on addition of nitrate irrespective of the presence of C2H2. Hybridization intensities, with a gene probe for the 16S rRNA, and MPN numbers were generally higher in soil samples taken from the roots of plants than in the bulk soil. There was no selective enrichment of denitrifying or N2-fixing bacteria at the roots. The data obtained by hybridizing isolated soil DNA generally matched previous results obtained with culturable bacteria.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and methods
  5. 3. Results
  6. 4. Discussion
  7. Acknowledgments
  8. References

N2 fixation and denitrification are essential processes of the nitrogen cycle. It has been stated that most of the global conversion of the dinitrogen molecule to ammonia proceeds by nitrogenase of prokaryotic microorganisms [1]. In denitrification, nitrate is reduced via nitrite to gaseous nitrogen compounds (NO, N2O and N2). The two opposite reactions, N2 fixation and denitrification, keep the global nitrogen budget largely in balance. The relative contributions of the different microbial species to these two reactions are scarcely known for soils and other habitats. This is mainly due to the fact that only a small percentage (1% or less; [2–4]) of soil bacteria can be grown in culture. The major part of the bacterial community can, therefore, not be examined for their N2-fixing and denitrifying capabilities under laboratory culture conditions.

Gene probing has become a new avenue for assessing the distribution of bacteria in diverse environments. Probes which allow monitoring of the total population of bacteria in a sample are mainly based on DNA sequences encoding 16S or 23S rRNA. Such rRNA-based probes cannot be used to differentiate between N2-fixing or denitrifying bacteria and non-performing species, since these physiological traits are widespread in microorganisms of unrelated taxonomic affinities [5,6]. For N2-fixing bacteria, nifH, the gene encoding dinitrogenase reductase (the smaller subunit of the nitrogenase complex) is relatively conserved in all known organisms [7] and is, therefore, suitable for developing probes to screen for the occurrence of nitrogenase in bacteria. For denitrification, probes have mainly been developed from the DNA encoding nitrite reductase. This enzyme catalyzes the conversion of nitrite to nitric oxide and is therefore diagnostic for differentiating denitrification from bacterial nitrate ammonification and assimilatory nitrate reduction. Denitrifying bacteria can possess either a cytochrome cd1 or a Cu-containing nitrite reductase with dissimilar prosthetic groups and amino acid sequences [8]. Gene probes which show no cross hybridization can readily be developed [5]. The Cu-nitrite reductase occurs in bacteria of unrelated affinities, but is fairly conserved among the different organism. The cytochrome cd1-containing enzyme appears to be more widespread among microorganisms [8], but DNA probes or antibodies developed for this enzyme can recognize its occurrence only in a more limited range of bacteria [6,9]. Hybridization with probes from other genes (narG, dissimilatory nitrate reductase, or nosZ, nitrous oxide reductase) have rarely been employed [10–12].

Probing for denitrification and N2 fixation genes has been done either by PCR amplification using 15–25 oligomers of highly conserved regions, or by DNA–DNA hybridization with 400-bp to 1-kb sized probes. Experiments have been performed with mixed bacterial populations, with sediments, activated sludge [13] or aquatic habitats [14]. Molecular biological analysis of soil community DNA may be problematic. Shearing of the DNA has to be prevented during the isolation procedure, and preparations often contain humic acid or other interfering substances [15–17]. Several improved protocols for the isolation of DNA from soils have recently been published, e.g. [18–24]. In the present study, a further DNA preparation was developed from such protocols. The DNA was amenable to hybridization with probes for denitrification, N2 fixation and 16S rRNA and was screened for the distribution of denitrifying and N2-fixing bacteria in different soil layers of a forest soil in the vicinity of Cologne.

2. Materials and methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and methods
  5. 3. Results
  6. 4. Discussion
  7. Acknowledgments
  8. References

2.1 Parameters of the soil for which the distribution of microorganisms has been investigated

The alluvial gleysol of the Chorbusch forest (51°02′50″N, 06°48′40″E), about 20 km west of the city of Cologne, supports a largely undisturbed oak–hornbeam forest. The tree layer is mainly composed of Quercus robur L. and Quercus petraea (Mattuschka) Liebl., Tilia cordata Miller, Carpinus betulus L. and a relatively low percentage of Fagus sylvatica L. The scrub layer consists of the same trees and, in addition, plants like Crataegus monogyna Jacq. and Crataegus laevigata (Poiret) DC. The herb layer is fairly rich and contains the two acid indicators Oxalis acetosella L. and Deschampsia cespitosa (L.) P.B. Other plants like Lamiastrum galeobdolon (L.) Ehrend. and Polatschek, Ranunculus ficaria L., Stellaria holostea L. or Anemone nemorosa L. indicate a nutrient richer, neutral soil.

Samples were taken out of about 40-cm deep square holes dug into the earth, and approximately 2-cm segments were taken at soil depths of 5 cm, 10 cm and 25 cm. Samples were immediately used in the laboratory for DNA isolation and activity measurements without any storage in a refrigerator. The Chorbusch soil is a fairly homogeneous, acid clay loam in the upper 30 cm. A typical B-horizon is absent. The soil parameters (Table 1) determined by standard techniques [25,26] have the following main features: average concentrations of cations, low organic carbon content, rapid decomposition of leaves and other litter on the soil surface, probably due the vigour of microorganisms, moderate to high nitrogen content in all three soil layers. Consequently, the low C/N ratio might indicate that the N content is not limiting for microbial life. With respect to the phosphorus, values <1 g kg−1 soil indicate limitations, and the values in the Chorbusch soil layers are at best only 1/10 of that value.

Table 1.  Characterization of the Chorbusch soil, a gleysol, used for the present study
ParameterSoil depth
5 cm10 cm25 cm 
  1. The soil parameters were determined by standard techniques [25,26] in March 1999 (CECe=effective cation exchange capacity).

pH (KCl)
pH (H2O)
Organic C content (%)
Total nitrogen (%)
P2O5 (mg kg−1)100.133.417.4
C/N ratio9.68.26.4
C/P ratio5499581183
NO3-N (mg kg−1)
NH4+-N (mg kg−1)0.120.450.36
Total inorganic nitrogen (mg kg−1)5.97.712.8
NO3 (% of the total inorganic N)
Ca2+ (meq kg−1)
Mg2+ (meq kg−1)2.114.515.5
K+ (meq kg−1)
Na+ (meq kg−1)
Mn2+ (meq kg−1)
Fe2+ (meq kg−1)
Al3+ (meq kg−1)54.750.345.7
H+ (meq kg−1)16.112.723.6
CECe (meq kg−1)89.687.993.7
Specific weight (g cm−3)
Stone (skeleton) content (%)

2.2 Reference organisms

The following organisms were used for reference purposes: Azospirillum brasilense Sp7 (ATCC 29145), grown in the AB medium supplemented with malate [27], Alcaligenes eutrophus (=Ralstonia eutropha) H16 (ATCC 17699) and Escherichia coli K12 (ATCC 23716), both grown in LB [28]. Cells were grown to the late exponential phase.

2.3 Extraction of DNA from soil layers and from pure cultures

The method described is based on published protocols [16], including some alterations: soil samples (5 g) were suspended in 10 ml 0.1% Na4P2O7 dissolved in 10 mM Tris–HCl buffer pH 8.0/1 mM EDTA (=TE buffer), incubated at room temperature (10 min), followed by centrifugation (10 min, 6000×g). Such a pretreatment, suggested in [29], removed most of the free DNA and part of the humic acids and of other interfering substances but left the soil bacteria intact (as indicated by separation on agarose gels loaded with DNA isolated with or without the pyrophosphate treatment). To get rid of residual pyrophosphate, samples were washed with 10 ml TE buffer and centrifuged (10 min, 6000×g). The pellet was dissolved in 5 ml TE containing 25 mg lysozyme and incubated in a shaking water bath (1 h, 37°C). After the addition of 1 ml of 10% sodium dodecyl sulfate (SDS), the samples were further incubated in the shaking water bath (1 h, 65°C) and centrifuged (10 min, 6000×g). The collected supernatants were then mixed with an equal volume of phenol followed by centrifugation (15 min, 12 000×g). The aqueous phase was transferred to tubes containing 5 ml chloroform/isoamyl alcohol (24/1, v/v) to remove traces of phenol. The DNA-containing upper phase (∼5 ml) was transferred to centrifuge tubes, mixed with 300 μl 3 M K+-acetate plus 3 ml isopropanol and incubated (16 h, 4°C) to complete the DNA precipitation. After centrifugation (10 min, 10 000×g), purification with 300 μl 70% ethanol and a further centrifugation (5 min, 10 000×g), the pellet was dried by vacuum centrifugation (Speed Vac) and dissolved in 1 ml 1.2 M NaCl. After 2 h at room temperature, the preparation was transferred to Eppendorf tubes already containing 1/10 of the volume of hexadecyltrimethylammonium bromide dissolved in 1.2 M NaCl and incubated (10 min, 65°C). Then 1 ml of chloroform was added, the tubes were centrifuged in the Sigma centrifuge at maximal speed (1830×g, 5 min), and the supernatant was removed with a pipet. This washing step was repeated once more, and the DNA in the aqueous supernatant (∼1 ml) was then precipitated with 50 μl 3 M potassium acetate and 0.6 ml isopropanol, incubated (16 h, 4°C) and centrifuged (20 min, 18 000×g). After a further washing step with 70% ethanol, the pellet was dried in the vacuum centrifuge and suspended in 20 μl TE buffer.

To get an estimate for the DNA extraction efficacy, seeding experiments were performed with A. brasilense Sp7 (α-proteobacteria), A. eutrophus (β-subgroup) and E. coli K12 (γ-subgroup). Bacterial suspensions in the late exponential growth phase (2 ml with 1.5×109 cells ml−1) were thoroughly mixed with 2 g of Chorbusch soil in three independent experiments. DNA was then isolated both from soil/bacterial mixtures and from the bacterial control cultures. The DNA recoveries were 69.5±8.5% in the case of A. brasilense Sp7, 47.5±9.1% for E. coli K12 and 69.8±1.4% for A. eutrophus H16 (n=3, always).

2.4 Southern hybridization with soil DNA and gene probes for denitrification and N2 fixation

Soil DNA (5 μl with 2 μg DNA) was electrophoresed on a 0.7% agarose gel. The upper part containing DNA but no humic acids or RNA was transferred in 0.4 N NaOH to a nylon+ membrane (PALL Gelman Sciences, D-Rossdorf). Hybridization was then performed at 68°C overnight with 200–300 ng labelled probe per 100 cm2 membrane in 5×SSC, 0.5% blocking reagent, 0.1% Na-lauroylsarcosine, 0.02% SDS [32]. The filters were rinsed twice with 2×SSC, 0.1% SDS for 10 min at 68°C. Immunological detection with the non-radioactive hybridization kit (Boehringer-Roche, Mannheim, Germany) and the labelling of the gene probes with digoxigenin were as described [10].

2.5 Gene probes used in the hybridization experiments

The probes (narG, nirK, nirS, nosZ and nifH) used for hybridization were obtained by PCR amplification, cloning and sequencing [10–12]:

  • 1
    narG (dissimilatory nitrate reductase, Mo-pterin binding subunit): a 414-bp segment of amplified DNA from E. coli K12, showing 100% sequence identity, both on the DNA and the amino acid level, to the published sequence ([30]; EMBL accession no. X15181, position of the probe within the DNA sequence: N28–N441).
  • 2
    nirS (cytochrome cd1-containing nitrite reductase): a 691-bp segment from A. brasilense Sp7 ATCC 29143, showing 64% identity, on the amino acid level, to the sequence published for the same segment from Pseudomonas aeruginosa NCTC 6750 ([31]; X16452, N226–924).
  • 3
    nirK (Cu-containing nitrite reductase): a 576-bp segment from Alcaligenes xylosoxidans NCIMB 11015, showing 67% identity, on the amino acid level, to the sequence published for Alcaligenes faecalis S-6 ([32]; D13155, N526–1101).
  • 4
    nosZ (N2O reductase): a 598-bp segment from Pseudomonas stutzeri ZoBell, showing only an identity of 88% (DNA) or of 91% (amino acid basis), respectively, to the sequence published for nosZ from the same organism ([33]; M 22628, N630–1227).
  • 5
    nifH (nitrogenase reductase): a 435-bp segment of this nitrogenase gene of A. brasilense Sp7, showing 100% identity to the published sequence of the same organism ([34]; L08050, N19–453).
  • 6
    as general probe: a 693-bp segment of the 16S rRNA gene of A. brasilense Sp7, showing 85% identity to the published sequence of E. coli ([35]; N8–701).

2.6 Quantification of the signals obtained in Southern hybridizations

From one hole in February 1998 and two holes in March 1999, probes from the Chorbusch soil were sampled at 5, 10 and 25 cm depth. For each depth, the samples were pooled in the experiment of March 1999, and DNA was isolated. 2 μg DNA of each layer was blotted onto each of three filters, and hybridization with the gene probes was performed. In the case of the experiment of February 1998, hybridization was done only with the physiological gene probes (nifH, narG, nirS, nirK or nosZ). In March 1999, hybridization was first performed with these physiological gene probes. The signal intensities were determined densitometrically (NIH Image 1.61 picture analyzer program for MacIntosh). The filters were then stripped, hybridized with the 16S rRNA probe and signal intensities were quantified. The total signals (sum of the signals obtained with DNA from the three layers) were set to 100%. The data in Table 3 refer to % signal activity in each layer per g soil.

Table 3.  DNA isolated from the bulk soil and from roots of different plants: hybridization with the different gene probes
Gene probe16S rRNAnifH16S RNAnirS16S rRNAnirK16S rRNAnosZ16S RNAnarG
  1. The experiments were performed as described in the legend to Table 2 and under Section 2. Filters were first hybridized with the physiological gene probe and then with the 16S rRNA probe. The data obtained with the bulk soil were set to 100%. Standard deviations are from three independent filters. Experiment of March 1999.

Bulk soil100%100%100%100%100%100%100%100%100%100%
D. cespitosa115±3130±14110±19101±16119±4249±29107±1590±58155±13134±18
T. cordata114±4110±24121±10112±22161±5049±30129±185±24150±4157±46
O. acetosella138±50137±33169±62133±10131±557±5144±9124±24130±34138±23
S. holostea154±28146±38138±58131±48140±867±14108±295±19180±21167±40
L. Galeobdolon152±40135±71121±4110±33180±3174±38155±14133±27183±28185±10
A. nemorosa107±22104±28118±18130±4151±3579±27120±3150±4134±2124±12

2.7 Activity measurements in samples from different soil layers

To get an estimate for their denitrification activity, soil samples (5 g, wet weight) from Chorbusch were tested for their potential N2O formation activity in the laboratory in ∼7 ml Fernbach flasks covered with a Suba Seal. The flasks were supplemented with 0.2 mM Na-nitrate. The gas phase was exchanged by argon, and C2H2 (∼15%, v/v) was added to stop N2O reductase activity. After 1 h of incubation at 30°C, the amount of N2O formed was determined by gas chromatography [10–12].

2.8 Other determinations

The determinations of the colony forming units (cfu) on agar plates containing LB medium and the MPN counts using the liquid LB medium were done exactly as in the preceding publication [11].

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and methods
  5. 3. Results
  6. 4. Discussion
  7. Acknowledgments
  8. References

3.1 The quality of the DNA isolated from the soil for DNA probing and its distribution in different soil layers

The methods employed allowed reproducible isolation of genomic DNA from the three layers 5, 10, 25 cm of the Chorbusch soil (Fig. 1). The molecular mass of the isolated DNA was >20 kb with no smear on the slants at the lower molecular mass range, indicating that the molecules were intact. The 260/280 nm ratio of the isolated DNA (after electrophoresis on a low melting agarose gel) was 1.7±0.2 (standard deviation from 10 different extractions). The yield from the upper (5 cm) soil layer varied between 3 and 20 μg DNA isolated g−1 dry weight of soil. In the Chorbusch soil, the DNA amount of the upper 5 cm sample was approximately double of that of the 10-cm layer, and the DNA concentration was lowest at 25 cm (Fig. 1). The genomic DNA isolated from the three Chorbusch soil layers was amenable to restriction by the enzymes BamHI, SmaI and SacI but not by HindIII or EcoRI (data not shown).


Figure 1. Genomic DNA isolated from three different layers of the Chorbusch soil. DNA was isolated from the soil layer (in each case from 0.5 g of soil fresh weight) as described under Section 2 and electrophoresed on a 0.7% agarose gel. Lanes: (a) 1-kb ladder (from Gibco), (b) DNA from 5 cm depth (=3.35 μg DNA g−1 dry weight of soil), (c) DNA from 10 cm depth (=1.65 μg DNA g−1 dry weight), (d) DNA from 25 cm depth (1.12 μg DNA g−1 dry weight).

Download figure to PowerPoint

The DNA isolated from the Chorbusch soil (5-cm layer) hybridized with the bacterial DNA probe encoding the 16S rRNA segment (in the following termed 16S rRNA probe). To quantify bacterial DNA, the signal was scanned (Fig. 2). A direct proportionality existed between the signal intensity and up to 2 μg DNA isolated from the Chorbusch soil (Fig. 2A,B) or to 3 μg from E. coli as a control (Fig. 2A). Three concentrations of the 16S rRNA probe gave the same signal intensities (Fig. 2B), indicating that the probe did not limit the hybridization under the assay conditions employed.


Figure 2. A: Signals obtained by hybridizing different amounts of DNA from the Chorbusch soil or from E. coli with the digoxigenin-labelled DNA probe encoding a segment of the 16S rRNA. DNA–DNA hybridization was performed under standard conditions using 125 ng digoxigenin-labelled 16S rRNA probe. The DNA was isolated from soil of the upper (5 cm) layer of the Chorbusch forest. B: Densitometric determination of the signal strength obtained in the DNA–DNA hybridizations using the soil DNA and the digoxigenin-labelled 16S rRNA probe. The experimental conditions were as in A. Hybridization was performed with DNA from the upper soil layer and the three concentrations of the 16S rRNA probe: 100 ng ♦, 125 ng ▴, 250 ng •.

Download figure to PowerPoint

To examine the distribution of the bacterial DNA in different layers of the Chorbusch soil, two samples ∼10 m apart were assayed (Fig. 3), and hybridization was performed with the 16S rRNA probe. In both samples, the highest amount of hybridizing DNA was at 5 cm, and the intensity at 10 cm was consistently lower than that seen at 5 cm. Lowest signal intensities were obtained with the DNA from 25 cm. This was consistent with the total DNA assessments (Fig. 1). The data also indicated that the absolute amount of hybridizing DNA was variable within 10 m, because the intensity of the 5-cm band on the same filter was much higher in the b than in the a sample (Fig. 3).


Figure 3. Hybridization of the genomic DNA isolated from the Chorbusch soil layers with the digoxigenin-labelled 16S rRNA probe. DNA was isolated from the three layers of the soils at two places, separated from each other by ∼10 m. The amount of DNA in each lane was from 0.5 g soil fresh weight.

Download figure to PowerPoint

3.2 Distribution of denitrifying and N2-fixing bacteria in the Chorbusch soil assessed by DNA probing

In control experiments (Fig. 4), the nifH and nirS probes did not hybridize with DNA from E. coli. Positive scores with these probes were, however, obtained with DNA from A. brasilense Sp7 which is a known N2-fixing and denitrifying bacterium [36]. Further control experiments with isolated DNA of a wide range of bacteria are listed in [10,11]. In the experiment of February 1998, isolated DNA (2 μg) from the Chorbusch soil was blotted onto each lane of the filters, and hybridization was performed only with the physiological gene probes. As the dilution factor is different to obtain 2 μg out of each of the three layers, a comparison of the signal intensities, referred to 1 g soil, e.g., already indicated that the highest relative signal intensity was always detected with DNA from the upper (5 cm) layer and the lowest one was consistently seen with DNA from the lowest (25 cm) soil layer with all physiological gene probes (Table 2A). To draw conclusions about the distribution of denitrying and N2-fixing bacteria in the different soil layers, it had to be assured that the signal intensity was always the same when blotting 2 μg DNA of the same probe onto one filter and the next. This was, however, not the case in these blottings performed manually (e.g. Fig. 3). The intensities of the DNA–DNA hybridization signals depended on experimental details such as efficiency of the DNA transfer onto the filters, hybridization strength of the buffer, stripping of the filter, conditions during photoimaging, staining with ethidium bromide, quality of the digoxigenin-labelled probe, etc. Therefore signal intensities can be compared only on a relative basis on each filter. In the experiment of March 1999, soil DNA on each filter was first hybridized with the physiological gene probe and after the densitometric determination and stripping, hybridization with the same filter was performed with the 16S rRNA probe (Fig. 4, Table 2Bα,β). Such a procedure allowed to evaluate the differences seen in the different filters statistically. Any rise in the quotient in the signal intensities between physiological gene/16S rRNA probe (Table 2Bγ) would indicate a selective enrichment of bacteria with a physiological gene in any layer. However, such a statistically significant increase in this ratio was not observed for any layer and any gene probe, indicating that there was a proportional decrease of total DNA, bacterial DNA hybridizing with the 16S rRNA probe and of bacteria possessing the genes for denitrification and N2 fixation with the depth of the soil.


Figure 4. Hybridizations of the genomic DNA isolated from the Chorbusch soil with the gene probes nifH and nirS. Two μg of isolated DNA was blotted onto each lane. Hybridization was first performed with the digoxigenin-labelled nifH (A) or the nirS probe (B). The filters were then stripped, followed by hybridization with the labelled 16S rRNA probe. For experimental details see Section 2. DNA was blotted onto the following lanes: 1 and 4, from 5 cm depth of the Chorbusch soil; 2 and 5, from 10 cm depth; 3 and 6, from 25 cm depth; 7, from the rhizophere of A. nemorosa; 8, of the fern Athyrium filix-femina (L.) Roth; 9, of Galeobdolon luteum; 10 and 11, from E. coli K12 (as negative control); 12 and 13, from A. brasilense Sp7 (as positive control).

Download figure to PowerPoint

Table 2.  Hybridization of the total genomic DNA isolated from the three different layers of the Chorbusch soil with the gene probes for nitrogenase and denitrification
Soil depth (cm)nifHnarGnirKnirSnosZ
  1. From each depth (5, 10 and 25 cm), 2 μg DNA was blotted onto each of three filters, and hybridization with the gene probes was performed. In the case of the experiment of February 1998, hybridization was performed only with the physiological gene probes. In March 1999, hybridization was first performed with the physiological gene probes, and then, after stripping, with the 16S rRNA probe. The data in the table refer to % signal intensity in each layer per g soil (total sum obtained in all three layers=100%). Standard deviations are given for the recordings on the three filters.

(A) Relative hybridization intensities in the different soil layers (experiment of February 1998)
(B) Relative hybridization intensities in the different soil layers (experiment of March 1999)
(α) Hybridization with the gene probes
(β) The same filter, but after stripping hybridization now with the 16S rRNA probe
(γ) Quotient physiological gene probes/16S rRNA probe (=α:β)

Signal intensities were also compared when the DNA isolated from the soils surrounding the roots of plants was hybridized with the different gene probes (Table 3). The signal obtained with DNA from the bulk, root-free soil (taken from ∼5 cm depth) was always set to 100%. Previous MPN determinations had indicated that the total number of bacteria in the vicinity of roots was significantly higher than in the bulk soil [10,12]. This is also seen here in the Southern blots with the 16S rRNA probe which recognizes bacterial DNA (Table 3, with the possible exceptions of soil from A. nemorosa and O. acetosella). The signals obtained with DNA from the soil of the roots of almost all plants showed statistically higher signal intensities with virtually all physiological gene probes with the exception of nirK. This might only reflect that the total number of bacteria and, in parallel with this, the bacteria hybridizing with the physiological gene probes were increased in the vicinity of roots probably due to the higher availability of organic carbon there. In the case of the nirK probe, however, signal intensity was statistically higher in the root-free soil (Table 3).

3.3 Bacterial counts and determination of potential denitrification activities

When bacteria from the three soil layers were isolated and grown on LB medium on plates, the number of cfu was highest in the upper zone, and the lowest amounts of colonies were obtained in the 25-cm layer (Table 4). A similar pattern was obtained by MPN determinations, although the absolute numbers were 4–11-fold higher in the MPN than in the cfu counts. In two different probes sampled from the same location, highest N2O formation (denitrification) activity was always observed in the upper soil layer and lowest in the 25-cm zone. This activity was, however, dependent on the addition of 0.2 mM nitrate to the assays. Acetylene is a specific inhibitor of N2O reductase in denitrification [37]. The addition of C2H2 to the samples only partly enhanced N2O formation (Table 4) which might indicate that a proportion of the bacteria in this soil were not able to convert N2O to N2.

Table 4.  Cell counts and potential denitrification activities in the Chorbusch soil
Parameter determined5 cm10 cm25 cm
  1. The experimental conditions for determining MPN, cfu and N2O-formation have been described under Section 2. Experiments a and b are for two soils samples taken ∼10 m apart from each other.

1. Cell number×106 (March 1999)
(a) MPN107028169
(b) cfu904714
2. N2O formation activity (nmol g−1 dry weight soil day−1)
(a) In the presence of C2H2
March 99, Experiment a2805927
Experiment b149110
October 99, Experiment a2165183
Experiment b3895050
(b) In the absence of C2H2
March 99, Experiment a2404933
Experiment b2841911
October 99, Experiment a19313373
Experiment b4027099

4. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and methods
  5. 3. Results
  6. 4. Discussion
  7. Acknowledgments
  8. References

4.1 General problems to study microbial populations in soils

From a molecular ecological perspective, investigators working with soils are faced with three major problems: soils are rich in interfering substances like humic acid, tannins and polysaccharides [20], the bacterial community is complex [3,38] and its culture-dependent community structure analysis is said to produce heavily biased results [39,40], soils are generally non-homogeneous and their elemental, microbial and plant compositions vary within very short distances.

As to point (a), bacterial DNA amenable for molecular characterization can only be obtained with great difficulty because of such contaminants. The DNA isolation protocol used in the present study takes ∼2 days and is thus time-consuming. Otherwise the protocol has several advantages: the initial washing step with pyrophosphate helped largely to remove interfering substances but left the soil bacteria intact. The DNA finally obtained showed a 260/280 nm ratio of ∼1.7 and had a molecular mass of >20 kb. It should be stressed that impurer preparations did not give reproducible DNA hybridization signals with the probes for denitrification and nifH. A good indication for the quality of DNA preparation from a soil is reflected by its accessibility to restriction enzymes [16,17,19,24]. Our preparation was suitable to restriction by several enzymes and also for PCR amplification using 25–30 oligomer primers for nirS, narG and nifH (A. Mergel, unpublished).

(b) As said, a general probe, based on 16S rRNA sequences, cannot be developed for monitoring denitrifying and likely not for N2-fixing bacteria. In contrast to the DNA encoding rRNAs, denitrification and N2 fixation genes occur in low copy numbers. In soil cores, the detection by in situ hybridizations of genes present in bacteria in single or low copy numbers is difficult [41]. Therefore we had to restrict our previous studies [11,12,42] to culturable bacteria. These investigations can be criticized on the grounds of relevance because culturable bacteria reflect only a low percentage of the soil bacterial population. The successful isolation of soil DNA in the present study allowed assessment of the relative distribution of bacteria with specific physiological traits in different soil layers and in the vicinity of plants. The signal strength attained with the probes for denitrification and nifH may appear to be surprising. However, the DNA from the upper layer of the Chorbusch soil has recently been used to amplify segments of the denitrification genes and of nifH by PCR and with specific oligonucleotide primers (C. Rösch, A. Mergel, H. Bothe, to be published). Cloning and sequencing of some 100 of these PCR products revealed that most sequences are new, but rather similar to those of the data banks, which is in accordance with the strong signal strength seen in the present study. While the present results can be compared on a relative basis between samples, it is hardly feasible to draw conclusion as to the absolute number of denitrifying and N2-fixing bacteria in the soils and their activities in these processes on a single cell basis. As no better methods are available, soil seeding experiments with cultured cells, as in the present study, are often performed to get an estimate about the DNA extraction efficacy. However, the endogenous microorganisms may attach more tightly to soil particles than do seeded bacteria; they may not so readily release their DNA in the lytic step; they may hybridize with different intensities to the 16S rRNA probe and their genes encoding denitrification and N2 fixation occur in unknown copy numbers. It should also be mentioned that the 16S rRNA probe used cannot differentiate between bacterial DNA and any contribution of plastid and mitochondrial DNA in the soil samples.

(c) The large variability of the DNA content within a few meters (Fig. 3) of a soil of homogeneous appearance reflects the general difficulty in studying soil organisms. In spite of this, the present study gave clear-cut and statistically reproducible differences in the relative abundance of bacteria in the different layers of the Chorbusch soil. The upper soil layer contained the highest number of culturable bacteria and showed the highest potential denitrification activity. The DNA probes also indicated that the most intensive hybridization bands (thus the highest number of bacteria) obtained with the 16S rRNA, as well as with the nifH and the denitrification probes, were seen in the upper soil layer.

4.2 General conclusions drawn from the present results

The present study provided pieces of information which were unexpected to us:

  • 1
    One reason for selecting the Chorbusch soil for the present study was its high nitrate content, even in the lower (25 cm) soil layer. At this depth, O2 is presumably limiting (exact concentrations cannot be determined in situ, but the vivid activity of the microorganisms and plant roots in the upper layer presumably consumes most of the O2 available). Therefore conditions should be favorable for denitrifying bacteria in 25 cm depth. In spite of this, a selective enrichment of denitrifying bacteria in the lower layer was not observed in any of the Chorbusch soil samples. A certain percentage of the total bacteria in all soil layers possesses the genes for denitrification and N2 fixation, but they apparently enjoy no selective advantage unless denitrification in microsites plays a role in such soil profiles [43]. The Chorbusch soil was selected by us because of its homogeneity within the upper 30 cm, so different extraction efficacies of the DNA from the three layers, from the bulk soil or from the soil adjacent to roots are fairly unlikely to us.
  • 2
    The soil is fairly acid. Therefore we had expected a low number of bacteria, but just the opposite was observed. The bacterial cell number, determined by the MPN method, of acid soils generally amounts to ∼107 cells g−1 dry weight of soil but reached 2×109 cells in the upper soil layer of the Chorbusch soil and is thus in the range of the numbers normally detected in neutral pH soils. In contrast, both ecto- and arbuscular mycorrhizal fungi occurred in unexpectedly low numbers (F. Ouziad, H. Bothe, unpublished data) despite the fact that phosphorus was limiting in this soil.
  • 3
    The results of previous investigations using culturable bacteria [11] matched those of the present study on the relative abundance of DNA hybridization signals and were also corroborated by the data from the denitrification activity measurements in soil cores. This may well have happened by chance. However, as recently expressed [44], the possibility exists that culturable bacteria represent the metabolically active cells, while DNA may be retrieved from a huge background of inactive cells. In previous studies with culturable isolates, denitrifying bacteria were consistently found to be enriched in the vicinity of the plant roots [11,12], however this was not seen in the present study (Table 3).
  • 4
    The Chorbusch soil may contain unusual microorganisms. When nitrate was added to the soil cores, N2O was produced to some extent independently of the addition of C2H2 to the medium. Denitrifying bacteria exist, though exceptionally, which cannot express N2O reductase [45] and these may be responsible for N2O formation in the Chorbusch soil. Such bacteria were presumably not present, e.g., in a Norway spruce forest soil near Villingen in the Black Forest where this gas production was strictly dependent on the addition of C2H2 to the assays [46]. The observation that bacteria with a Cu-nitrite reductase preferentially occupy the root-free soil was similarly unexpected. Further work has to show whether this is a special trait of the Chorbusch soil.


  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and methods
  5. 3. Results
  6. 4. Discussion
  7. Acknowledgments
  8. References

This work was kindly supported by a grant from the Deutsche Forschungsgemeinschaft within the priority program ‘Microbial Ecology’. The authors are indebted to Dr. M.G. Yates, Brighton, UK, for correcting the English and to one anonymous referee for helpful comments.


  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and methods
  5. 3. Results
  6. 4. Discussion
  7. Acknowledgments
  8. References
  • [1]
    Burns, R.C. and Hardy, R.W.F. (1975) Nitrogen Fixation in Bacteria and Higher Plants, pp. 43–60. Springer, Berlin.
  • [2]
    Amann, R., Springer, N., Ludwig, W., Görtz, H.-D., Schleifer, K.H. (1991) Identification in situ and phylogeny of uncultured bacterial endosymbionts. Nature (Lond.) 251, 161164.
  • [3]
    Torsvik, V., Goksoyr, J., Daae, F.L. (1990) High diversity in DNA of soil bacteria. Appl. Environ. Microbiol. 56, 782787.
  • [4]
    Ward, D.M., Weller, R., Bateson, M.M. (1990) 16S rRNA sequences reveal numerous uncultured inhabitants in a natural community. Nature (Lond.) 345, 6365.
  • [5]
    Hallin, S., Lindgren, P.-E. (1999) PCR detection of genes encoding nitrite reductase in denitrifying bacteria. Appl. Environ. Microbiol. 65, 16521657.
  • [6]
    Ward, B.B. (1995) Diversity of culturable denitrifying bacteria. Limits of rDNA RFLP analysis for the functional gene, nitrite reductase. Arch. Microbiol. 163, 167175.
  • [7]
    Ruvkun, G.B., Ausubel, F.M. (1980) Interspecies homology of nitrogenase genes. Proc. Natl. Acad. Sci. USA 77, 191195.
  • [8]
    Coyne, M., Arunakumari, A., Averill, B., Tiedje, J.M. (1989) Immunological identification and distribution of dissimilatory heme cd1 and nonheme copper nitrite reductases in denitrifying bacteria. Appl. Environ. Microbiol. 55, 29242931.
  • [9]
    Smith, G.B., Tiedje, J.M. (1992) Isolation and characterization of a nitrite reductase gene and its use as a probe for denitrifying bacteria. Appl. Environ. Microbiol. 58, 376384.
  • [10]
    Kloos, K., Fesefeldt, A., Gliesche, C.G., Bothe, H. (1995) DNA-probing indicates the occurrence of denitrifying and nitrogen fixation genes in Hyphomicrobium. Distribution of denitrifying and nitrogen fixing isolates in a sewage treatment plant. FEMS Microbiol. Ecol. 18, 205213.
  • [11]
    Kloos, K., Hüsgen, U., Bothe, H. (1999) DNA-probing for genes coding for denitrification, N2-fixation and nitrification in bacteria isolated from different soils. Z. NaturForsch. 53c, 6981.
  • [12]
    Linne von Berg, K.-H., Bothe, H. (1992) The distribution of denitrifying bacteria in soils monitored by DNA-probing. FEMS Microbiol. Ecol. 86, 331340.
  • [13]
    Neef, A., Zaglauer, H., Amann, R., Lemmer, H., Schleifer, K.-H. (1996) Population analysis in a denitrifying sand filter: conventional and in situ identification of Paracoccus spp. in methanol-fed biofilms. Appl. Environ. Microbiol. 62, 43294339.
  • [14]
    G. BrakerA. FesefeldtK.-P. Witzel Development of PCR primer systems for amplification of nitrite reductase genes (nirK and nirS) to detect denitrifying bacteria in environmental samples, Appl. Environ. Microbiol., 64 1998 3969.
  • [15]
    Smith, S.E. and Read, D.J. (1997) Mycorrhizal Symbiosis, 2nd edn. Academic Press, San Diego, CA.
  • [16]
    Tebbe, C.C., Vahjen, W. (1993) Interference of humic acids and DNA extracted directly from soil in detection and transformation of recombinant DNA from bacteria and yeast. Appl. Environ. Microbiol. 59, 26572665.
  • [17]
    Trevors, J.T., van Elsas, J.D. (1989) A review of selected methods in environmental microbial genetics. Can. J. Microbiol. 35, 895901.
  • [18]
    Clegg, C.D., Ritz, K., Griffith, B.S. (1997) Direct extraction of microbial community DNA from humified upland soils. Lett. Appl. Microbiol. 25, 3033.
  • [19]
    Jacobsen, C.S., Rasmussen, O.F. (1992) Development and application of a new method to extract bacterial DNA from soil based on separation of bacteria from soil with cation-exchange resin. Appl. Environ. Microbiol. 58, 24582463.
  • [20]
    Porteous, L.A., Armstrong, J.L., Seidler, R.L., Waltrud, L.S. (1994) An effective method to extract DNA from environmental samples for polymerase chain reaction amplification and DNA fingerprint analysis. Curr. Microbiol. 29, 301307.
  • [21]
    Sandaa, R.-A., Torsvik, V., Enger, O., Daae, F.L., Castberg, T., Hahn, D. (1999) Analysis of bacterial communities in heavy metal-contaminated soils at different levels of resolution. FEMS Microbiol. Ecol. 30, 237251.
  • [22]
    Schwieger, F., Tebbe, C.C. (1997) Efficient and accurate PCR amplification and detection of a recombinant gene in DNA directly extracted from soil using the Expand™ High Fidelity PCR System and T4 Gene 32 Protein. Biochem. Newsl. 2, 7375.
  • [23]
    Wikström, P., Wiklund, A., Andersson, A.-C., Forsman, M. (1996) DNA recovery and PCR quantification of catechol 2,3-dioxygenase genes from different soil types. J. Biotechnol. 52, 107120.
  • [24]
    Zhou, J., Bruns, M.A., Tiedje, J.M. (1996) DNA recovery from soils of diverse compositions. Appl. Environ. Microbiol. 62, 316322.
  • [25]
    Rowell, D.L. (1994) Soil Science: Methods and Application, Longman, London.
  • [26]
    Schlichting, E., Blume, H.P. and Stahr, K. (1995) Bodenkundliches Praktikum, Pareys Studientexte 81, Blackwell Wissenschafts-Verlag, Berlin.
  • [27]
    Bothe, H., Klein, B., Stephan, M.P., Döbereiner, J. (1981) Transformation of inorganic nitrogen by Azospirillum spp.. Arch. Microbiol. 130, 96100.
  • [28]
    Sambrook, J., Fritsch, E. and Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual, Vol. 1–3. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
  • [29]
    Orlov, D.S. (1974) Humic Acids of Soils (in Russian). Moscow State University (MGU), self-publisher, Moscow.
  • [30]
    McPherson, M.J., Baron, A.J., Pappin, D.C.D., Wootton, J.C. (1984) Respiratory nitrate reductase of Escherichia coli. Sequence identification of the large subunit gene. FEBS Lett. 177, 260264.
  • [31]
    Silvestrini, M.C., Galeotti, C.L., Gervais, M., Schinina, E., Barra, D., Bossa, F., Brunori, M. (1989) Nitrite reductase from Pseudomonas aeruginosa: Sequence of the gene and the protein. FEBS Lett. 254, 3338.
  • [32]
    M. NishiyamaJ. SuzukiM. KukimotoT. OhnukiS. HorinouchiT. Beppu Cloning and characterization of nitrite reductase gene from Alcaligenes faecalis and its expression in Escherichia coli, J. Gen. Microbiol., 139 1993 25.
  • [33]
    Viebrock, A., Zumft, W.G. (1988) Molecular cloning, heterologous expression, and primary structure of the structural gene for the copper nitrous oxide reductase from denitrifying Pseudomonas stutzeri. J. Bacteriol. 170, 46584668.
  • [34]
    De Zamaroczy, M., Delorme, F., Elmerich, C. (1989) Regulation of transcription and promoter mapping of the structural genes for nitrogenase (nifHDK) of Azospirillum brasilense Sp7. Mol. Gen. Genet. 220, 8894.
  • [35]
    Brosius, J., Dull, T.J., Sleeter, D.D., Noller, H.F. (1981) Gene organization and primary structure of a ribosomal RNA operon from Escherichia coli. J. Mol. Biol. 148, 107127.
  • [36]
    Döbereiner, J. and Pedrosa, F.O. (1987) Nitrogen Fixing Bacteria in Nonleguminous Crop Plants. Brock/Springer, Science Tech Publishers, Madison, WI.
  • [37]
    Yoshinari, T., Knowles, R. (1976) Acetylene inhibition of nitrous oxide reduction by denitrifying bacteria. Biochem. Biophys. Res. Commun. 69, 705710.
  • [38]
    Borneman, J., Skroch, P.W., O'Sullivan, K.M., Palus, J.M., Rumjanek, N.G., Jansen, J.L., Nienhuis, J., Triplett, E.W. (1996) Molecular microbial diversity of an agricultural soil in Wisconsin. Appl. Environ. Microbiol. 62, 19351943.
  • [39]
    Amann, R., Glöckner, F.-O., Neef, A. (1997) Modern methods in subsurface microbiology: in situ identification of microorganisms with nucleic acid probes. FEMS Microbiol. Rev. 20, 191200.
  • [40]
    Wagner, M., Amann, R., Lemmer, H., Schleifer, K.H. (1993) Probing activated sludge with oligonuclides specific for proteobacteria: Inadequacy of culture-dependent methods for describing microbial community structure. Appl. Environ. Microbiol. 59, 15201525.
  • [41]
    Hodson, R.E., Dustman, W.A., Garg, R.P., Moran, M.A. (1995) In situ PCR for visualization of microscale distribution of specific genes and gene products in prokaryotic communities. Appl. Environ. Microbiol. 61, 40744082.
  • [42]
    Fesefeldt, A., Kloos, K., Bothe, H., Lemmer, H., Gliesche, C.G. (1998) Distribution of denitrification and nitrogen fixation genes in Hyphomicrobium spp. and other budding bacteria. Can. J. Microbiol. 44, 181186.
  • [43]
    Philippot, L., Renault, P., Sierra, J., Hénault, C., Clays-Josserand, A., Chenu, C., Chaussod, R., Lensi, R. (1996) Dissimilatory nitrite reductase provides a competitive advantage to Pseudomonas RTC01 to colonise the centre of soil aggregates. FEMS Microbiol. Ecol. 21, 75185.
  • [44]
    Felske, A., Wolterink, A., van Lis, R., de Vos, W.M., Akkermans, A.D.L. (1999) Searching for predominant soil bacteria: 16S rDNA cloning versus strain cultivation. FEMS Microbiol. Ecol. 30, 137145.
  • [45]
    Zumft, W.G. (1992) The denitrifying procaryotes. In: The Procaryotes (Balows, A., Trüper, H.G., Dworkin, M., Harder, W. and Schleifer, K.-H. (Eds.), pp. 443–582. Springer, Heidelberg.
  • [46]
    Bothe, H., Kloos, K., Kaiser, K., Schmitz, B. and Nawrath, K. (1997) Seasonal fluctuations of N2O-producing bacteria in an acid soil of a spruce forest. In: Proceedings of the 7th International Workshop on Nitrous Oxide Emissions, Cologne, Berichte der Physikalischen Chemie, no. 41, pp. 315–319. Wuppertal.