OUP user menu

Analysis of bacterial communities in heavy metal-contaminated soils at different levels of resolution

Ruth-Anne Sandaa , Vigdis Torsvik , Øivind Enger , Frida Lise Daae , Tonje Castberg , Dittmar Hahn
DOI: http://dx.doi.org/10.1111/j.1574-6941.1999.tb00652.x 237-251 First published online: 1 November 1999

Abstract

The impact of heavy metal contamination on soil bacterial communities was studied in soils amended for many years with sewage sludge contaminated with heavy metals to varying extents. At the broad level of resolution, DNA reassociation analysis indicated a dramatic decrease in bacterial diversity from 16 000 bacterial genomes (g soil [wet wt])−1 in the non-contaminated soil to 6400 bacterial genomes (g soil [wet wt])−1 in soil with low metal amendments and only 2000 bacterial genomes (g soil [wet wt])−1 in soil with high metal amendments. No differences between bacterial communities of these soils, however, were displayed in the %G+C profiles analysed by thermal denaturation. At a coarse level of characterisation, in situ hybridisation analysing larger phylogenetic groups of bacteria revealed a general decrease in the percentage of cells detected with probes ARCH915, BET42a, GAM42a, SRB385, CF319a, LGCb and HGC69a with increasing metal amendment. Only cells detected with probe ALF1b increased significantly from 3.1±0.8% of the cells detected by the domain-specific probe EUB338 in the non-contaminated soil to 6.5±1.3% in soil with high metal amendments. These shifts in populations of larger phylogenetic groups were largely confirmed by dot blot analysis of 16S and 23S rDNA clone libraries from bacteria in soil with low metal and high metal amendments, respectively. For a fine-level characterisation, 72 clones of 16S rDNA libraries were identified by comparative sequence analysis. A few sequences could not be assigned to the major taxa described. Most of the sequences were assigned to the Gram-positive bacteria with a high DNA G+C content (45%) and the α-subdivision of Proteobacteria (24%). However, only minor differences were seen between bacterial communities from the low and high metal soils. In the soil with high metal amendment, more sequences clustered to the α-subdivision of Proteobacteria, while in the low metal soil, more sequences clustered to the Gram-positive bacteria with a high DNA G+C content.

Keywords
  • Heavy metal contamination
  • Soil
  • Bacterial diversity
  • Reassociation
  • % G+C profile
  • Hybridization
  • 16S rDNA sequence

1 Introduction

Sewage sludge has been used in agriculture as a valuable source of plant nutrients and organic matter in many European countries. Sewage sludge from industrial areas, however, often contains significant quantities of heavy metals such as Cu, Ni, Cd, Zn and Cr, which are chelated by the organic matter in the sludge [1]. Repeated application of sewage sludge to soil may therefore increase the concentration of heavy metals, which persist for extremely long periods of time [1]. Due to the recognised toxicity of elevated concentrations of heavy metals to most organisms, many studies in agriculture have been conducted on the effects of heavy metals on plant yield, on animals grazing on the land and on metal exposure of humans through the food chain [2].

Heavy metals also have large effects on processes important for soil fertility by affecting structure and function of microbial communities [2]. Key functions in soil processes such as mineralisation of organic material and nitrogen fixation may be inhibited even by minor metal contaminations with concentrations still below the upper legal limits set by the European Union [36]. Heavy metal contamination may lead to a reduction of total microbial biomass [7, 8], to a decrease in numbers of specific populations such as infecting rhizobia [4] or mycorrhizae [9], or to shifts in microbial community structure [1013].

Shifts in microbial community structures in heavy metal-contaminated soils have been analysed by cytochemical methods, such as total fatty acid analysis (FAME) [1012], thymidine incorporation [10] and substrate utilisation analysis (BIOLOG) [10, 14, 15], but also by molecular methods such as mole percent guanine+cytosine (%G+C) analysis [13] or amplified ribosomal DNA restriction analysis (ARDRA) [16].

Many of these methods were used in laboratory studies on short-term effects of heavy metals in experimentally contaminated soils preferentially targeting physiologically active microorganisms. The aim of our study was to analyse the long-term effect of heavy metals on total bacterial communities in contaminated samples from field sites. Molecular methods, with different levels of resolution, were used to analyse microbial communities in soils from three experimental field plots in Braunschweig, Germany, which, during the years from 1980 to 1991, received low metal or high metal amendments or remained non-contaminated [4]. The present paper presents broad-scale analysis of the community composition by DNA reassociation kinetics and determination of %DNA G+C profiles. For a coarse-scale characterisation of larger phylogenetic groups, filter and in situ hybridisation techniques were applied. A fine-level characterisation of the uncultured microbial communities in metal-amended soils was performed by comparative sequence analysis of 16S rDNA libraries.

2 Materials and methods

2.1 Soil characteristics

Soil samples, down to 10 cm depth, were collected at the end of November 1994 from an experimental field site situated in Braunschweig, Germany. The plots were characterised by (i) no sludge amendments, but N fertilisation, (ii) low metal amendments and (iii) high metal amendments (Table 1). The soils have a matrix consisting of 5% clay, 50% silt and 45% sand. In the period from 1980 to 1991, the low metal soil received 100 m3 ha−1 year−1 of contaminated sludge, while the high metal soil received 300 m3 ha−1 year−1 of the same sludge spiked with heavy metals (Cd, Cu, Ni and Zn) to increase the heavy metal load to the upper legal limits set by the European Union [4]. At the sampling time, no pronounced differences in organic C (Table 1) were noticed between these soils, although the increase in total concentrations of heavy metals Cd, Cu, Ni and Zn from the non-contaminated original soil to soil with the high metal amendment was accompanied by a decrease in pH from 7.1 to 5.3 (Table 1). All plots had been planted with spring rapes.

View this table:
1

Soil characteristics

Four samples from each of the three plots were sieved, then pooled and mixed before analysis. For reassociation, thermal denaturation and in situ hybridisation studies all three plots were investigated. For bacterial 16S and 23S rDNA library construction, only soils with low metal and high metal amendments were studied.

2.2 Extraction of bacteria from soils

For reassociation and thermal denaturation studies, bacteria were extracted from three portions of 30 g of soil each. The soil portion was homogenised in 100 ml of 1% (w/v) Na-cholate (Sigma) and 30 g of Chelex 100 (Sigma) in a Waring blender in three 1-min bursts at low speed with intermediate cooling on ice for 5 min. The soil homogenates were diluted to 400 ml with distilled water and centrifuged for 15 min at 5°C and 1000×g. Supernatants were pooled and stored on ice until further processing. The soil pellets were homogenised again with 100 ml ice-cold sterile distilled water for 1 min and centrifuged as before. The procedure was repeated once more. The pooled supernatants were concentrated by high speed centrifugation at 10 000×g for 45 min at 5°C. The bacterial pellet was washed once with 100 ml of 2% (w/v) hexametaphosphate (pH 8.5), the bacteria concentrated by centrifugation at 10 000×g for 30 min at 5°C and subsequently washed with Chrombach buffer (330 mM Tris-HCl, 1 mM EDTA, pH 8.0) [17]. After centrifugation, the pellet was resuspended in 100 ml isopropanol using an Ystral tissue homogeniser [18].

2.3 Nucleic acid extraction

For reassociation and thermal denaturation studies, DNA was extracted from bacteria according to Torsvik [19], with some modifications in the lysis procedures [20]. Bacteria were lysed using 5 mg ml−1 lysozyme. RNase IIa (50 mg ml−1) and RNase T1 (9 U ml−1) were added to remove RNA from the lysate. After addition of sodium dodecyl sulfate (SDS) to a final concentration of 1% (w/v), 5 M NaClO4 was added to a final concentration of 1 M and the suspension was vigorously shaken. The lysate was further deproteinised by addition of one volume of chloroform:isoamyl alcohol (1:24) followed by shaking for 10 min. The lysate was subsequently clarified by centrifugation for 5 min at 5000×g and finally mixed with hydroxyapatite [19]. Nucleic acids were purified on a hydroxyapatite column (Bio-Gel HT, Bio-Rad) and concentrated by cetylpyridium bromide (Sigma) precipitation. The DNA used for reassociation was sheared in a French Press at 20 000 psi followed by a second purification on a hydroxyapatite column and concentration by cetylpyridium bromide precipitation [19].

DNA used for PCR amplification and cloning was isolated by direct lysis of the bacteria in the soil. Soil (10 g) was homogenised for 1 min in a Waring blender with 100 ml of Crombach buffer. Cells in 10 ml of the soil homogenate were lysed by addition of 5 mg of lysozyme ml−1 (Sigma) and incubation at 37°C for 1 h. After addition of 0.2 mg ml−1 of proteinase K (Sigma) and further incubation at 37°C for 30 min [18], SDS was added to a final concentration of 1% (w/v) and the mixture was incubated at 65°C for 15 min. Afterwards, the lysate was cleared by centrifugation at 10 000×g for 10 min [18] and nucleic acids were purified using the Wizard total DNA clean up system (Promega, Madison, WI, USA).

2.4 Reassociation and thermal denaturation

Reassociation and thermal denaturation of nucleic acids were analysed in a UV-visible recording spectrophotometer (Varian, Cary 4, software by Svein Norland, Department of Microbiology, University of Bergen, Norway) [18]. The applied software calculates Cot1/2 values based on partial reassociation curves. Escherichia coli B DNA (Sigma D-2001) was used as a control in both reassociation and thermal denaturation experiments. Reassociation of sheared and heat-denatured DNA (400 μg ml−1 in 6×SSC and 30% (v/v) DMSO) was measured as the decrease in absorbance of the DNA kept at a constant temperature 25°C below Tm. The DNA complexity was calculated relative to the size of the genome of E. coli B assuming 4.1×106 base pairs [18]. Thermal denaturation studies were performed with DNA at a concentration of about 25 μg ml−1 in 0.1×SSC, by increasing the temperature 0.1°C min−1 for 14 h. The melting profile was calculated as percent hypochromicity [18]. The mole fraction of guanine and cytosine (GC) in 0.1×SSC was calculated using the formula (Mol fraction GC=[Tm−16.3 log(SSC/SSC0.1)]/50.2−0.990) [21]. The first derivative was calculated and used for analysis of the fine structure of the melting profile.

2.5 In situ hybridisation

Soil samples of 5 g were fixed in 4% (w/v) paraformaldehyde/phosphate buffered saline (PBS, composed of 0.13 M NaCl, 7 mM Na2HPO4 and 3 mM NaH2PO4, pH 7.2 in water) on ice for 3 h [22]. The suspensions were centrifuged at 4000×g for 5 min and the pellets subsequently washed with PBS, resuspended in 20 ml of 96% (v/v) ethanol and stored at −20°C at a concentration of 25 mg ml−1 of soil (wet wt). Before application to slides, 40 μl of the cell suspension was dispersed in 960 μl of 0.1% (w/v) sodium pyrophosphate in distilled water by mild sonication (Branson Sonifier B-12, Danbury, CT, USA) for 30 s at a setting of 8 for homogenising the soil sample [23]. 20 μl was subsequently spotted onto gelatin-coated slides (0.1% (w/v) gelatin, 0.01% (w/v) KCr(SO4)2), dried at 45°C for 30 min and finally dehydrated in 50, 80 and 96% (v/v) ethanol for 3 min each.

Oligonucleotide probes (Table 2) were synthesised with a Cy3 reactive fluorescent dye at the 5′ end (MWG Biotech, Ebersberg, Germany). The labeled oligonucleotides were diluted in distilled water to a concentration of 25 ng μl−1 and stored at −20°C. Hybridisations were performed in 9 μl of hybridisation buffer (0.9 M NaCl, 20 mM Tris-HCl, 5 mM EDTA, 0.01% (w/v) SDS, pH 7.2) in the presence of 10–35% (v/v) formamide (Table 2), 1 μl of the probe (25 ng ml−1) and 1 μl of DAPI (200 ng ml−1) at 42°C for 2 h [23]. After hybridisation, slides were washed in buffer containing 20 mM Tris-HCl, 10 mM EDTA, 0.01% (w/v) SDS and either 440, 308, 102 or 80 mM NaCl depending on the formamide concentration during hybridisation (10, 20, 30 and 35% (v/v), respectively) for 15 min at 48°C, thereafter rinsed in distilled water and air-dried [23].

View this table:
2

Probes and primers used in this study

Slides were mounted with Citifluor solution AF1 (Citifluor, Canterbury, UK) and the preparations were examined with a Zeiss Axiophot microscope fitted for epifluorescence with a high-pressure mercury bulb (50 W) and filter sets Zeiss 02 (G365, FT395, LP420) and HQ-Cy3 (AHF Analysentechnik, Tübingen, Germany; G535/50, FT565, BP610/75) at 1000× magnification. DAPI-stained bacteria and cells hybridising with the probe in 20 fields, selected at random, covering an area of 0.01 mm2 each, were counted. Standard deviations were calculated for all counts.

2.6 16S and 23S rDNA library construction

Part of the 16S or 23S rDNA of bacteria was amplified by PCR using the GeneAmp 9600 Thermocycler from Perkin-Elmer (Norwalk, CT, USA). Three different primer sets were applied separately (Table 2). Each 100 μl reaction mixture contained standard PCR buffer (Perkin-Elmer), 0.16% (w/v) bovine serum albumin (BSA), dNTPs (each 200 μM), primers (each 1 μM), 2.5 U Taq DNA polymerase (Perkin Elmer) and template DNA (1.9 ng). An initial denaturation step at 95°C for 5 min was followed by 30 rounds of temperature cycling (95°C for 1 min, 55°C for 2 min and 72°C for 1.5 min) and a subsequent incubation at 72°C for 6 min. PCR products were purified by preparative gel electrophoresis (1% (w/v) low melting NuSieve Agar, FMC Bioproducts, Rockland, ME, USA) followed by purification using the PCR clean up kit (Promega). Cleaned PCR products were concentrated by ethanol precipitation [24].

Cloning of PCR products into the pMOSBlue T-vector was done according to the manufacturer's instructions (Amersham, Buckinghamshire, UK). Approximately 150 clones were analysed from each cloning reaction with the three different amplifications and two different soil treatments, so that a total of 1200 colonies were finally screened for positive inserts by PCR with T7 and U-19mer primers (Table 2) (Amersham). PCR was performed as described above, but using plasmids in whole bacterial cells as the template in a total volume of 50 μl.

2.7 Dot blot hybridisation

Ten microlitres of PCR products amplified from a total of 1200 cloned 16S and 23S rDNA fragments as templates were blotted to a Hybond-N+ membrane (Amersham) using a slot blot minifold (Schleicher and Schuell). Hybridisation with digoxigenin-labeled oligonucleotide probes generated by coupling digoxigenin-ddUTP to the 3′ end of oligonucleotides by terminal transferase according to the manufacturer's instructions (Boehringer, Mannheim, Germany) was performed in hybridisation buffer containing; 5×SSC, 0.1% (w/v) N-lauroylsarcosine (Sigma), 0.02% (w/v) SDS and 1% (v/v) blocking reagent (Boehringer), for 6 h at different temperatures depending on the probe (Table 2). After washing under high stringency conditions (Table 2), hybrid formation was shown by using the DIG luminescent detection kit following the manufacturer's instructions (Boehringer). The percentage of clones belonging to different phylogenetic groups was calculated relative to the total number of clones analysed.

2.8 Sequence analysis

Plasmid DNA was prepared from 37 and 31 clones from the 16S rDNA libraries from bacteria in soil with the low metal and high metal amendments, respectively, using the Qiaprep Plasmid Miniprep kit (Qiagen Inc., Chatsworth CA, USA). Sequence analyses of the cloned inserts (approx. 500 bp) were performed at the Advanced Biotechnology Centre (Charing Cross and Westminster Centre, London, UK) using the T7 primer as a sequencing primer. Sequences were checked for chimeric artefacts by the CHECK-CHIMERA program of the Ribosomal Database Project (RDP) [25]. Sequences were aligned using the PILEUP program of the GCG package. Distance matrices and phylogenetic trees were made using the CLUSTAL W program [26]. The sequences were taxonomically assigned using the nucleotide database at GenBank BLAST (NCBI).

3 Results

3.1 Broad-level analysis

In the non-contaminated soil, analysis of total DNA reassociation kinetics gave Cot1/2 values corresponding to approximately 16 000 different bacterial genomes (g soil [wet wt])−1, while in soil with low metal amendments, the Cot1/2 value corresponded to approximately 6400 bacterial genomes (g soil [wet wt])−1 (Fig. 1). In soil with high metal amendments, the Cot1/2 value corresponded to approximately 2000 bacterial genomes (g soil [wet wt])−1. No obvious differences in the %G+C profiles of DNA obtained from bacteria of the non-contaminated low metal or high metal soils were observed (Fig. 2). The profiles of all three soils showed a median %G+C content around 65%.

1

Reassociation (Cot plot) of DNA from the non-contaminated soil (▾) and soils with low metal (○) or high metal (●) amendments. The abscissa gives the logarithm of the initial concentration of single-stranded DNA (in mole nucleotides per litre) multiplied by time in seconds. The ordinate gives the per cent reassociated DNA.

2

%G+C distribution in DNA from the non-contaminated soil (solid line;) and soils with low metal (dotted line) or high metal (dashed line) amendments, derived by fitting the formula by Mandel et al. [21].

Coarse-level analysis

Total numbers of bacteria, determined after DAPI staining, were in the same range in all three soils although slight reductions were seen in the low and high metal soils compared to the non-contaminated soil (Table 1). In situ hybridisation with probe EUB338, designed to detect all members of the domain Bacteria, enabled us to detect 42±2.0, 46±2.3 and 41±3.8% of the DAPI-stained cells in the control soil and the soils with low metal and high metal amendments, respectively. The numbers of Archaea, detected with the probe ARCH915, decrease from 1.3±0.3% of the DAPI-stained cells in non-contaminated soil to numbers below the detection limit set at <1% of the DAPI-stained cells in contaminated soils [20, 23].

In the non-contaminated soil, probes ALF1b and SRB385 targeting bacteria of the α- and δ-subdivision of Proteobacteria, respectively, hybridised to 3.0±0.8% and 2.1±0.6% of the cells detected by the domain-specific probe EUB338 (Fig. 3A). All other phylogenetic probes gave hybridisation numbers below 1% of cells detected by the domain-specific probe EUB338. Bacterial communities in the soil with low metal amendment did not show significant differences to those in the non-contaminated soil (Fig. 3A). In the soil with high metal amendment, numbers of cells detected by phylogenetic probes were generally lower than in the control and low metal soils. The values only represented a trend rather than statistically significant data, since most of the numbers were below the detection limit set at 1% of the DAPI count (Fig. 3A) [20, 23]. However, there was a pronounced increase in cells detected with probe ALF1b from 3.0±0.8% and 2.1±0.3% of the cells detected after hybridisation with probe EUB338 in the control and low metal soil, respectively, to 6.5±1.3% of the cells detected after hybridisation with probe EUB338 in the high metal soil (Fig. 3A).

3

Analysis of larger phylogenetic groups with oligonucleotide probes ALF1b, BET42a, GAM42a, SRB385, CF319a, LGCb and HGC69a by in situ hybridisation in soil slurries (A) and by dot blot hybridisation in 16S and 23S rDNA libraries (B). Bars represent bacteria in % of detectable Bacteria (after hybridisation with probe EUB338) and % of the total number of clones analysed in non-contaminated soil (◻) and soils with low metal (▨) and high metal (■) amendments. Error bars represent standard deviation.

3.3 Fine-level analysis

Dot blot analysis of cloned fragments in 16S and 23S rDNA libraries of bacteria from the soil with high metal amendment showed that probes BET42a, GAM42a, SRB385, CF319a, LGCb and HGC69a generally hybridised to a lower percentage of clones than in the libraries of bacteria from the soil with low metal amendment (Fig. 3B). Only hybridisation with probe ALF1b resulted in the detection of much higher numbers of clones in the library of bacteria from the soil with high metal amendment, increasing from 16.4% of the clones in the library of bacteria from the soil with low metal amendment to 39.4% of the clones in the library of bacteria from the high metal soil.

Comparative sequence analysis was performed of 37 and 31 randomly selected clones from 16S rDNA libraries from bacteria from low metal and high metal soil respectively. Seven of the sequences formed two distinct groups, not clustering with any hitherto described taxa (Fig. 4). Group I was represented by three clones (clones 5a, 10a and 94d), whereas group II (69a, 96a, 102d and 1458a) consisted of four clones (Fig. 4). Twenty-four percent of the sequences were assigned to the phylogenetic group α-Proteobacteria while only a few clones clustered with bacteria of the β- and γ-subdivision of Proteobacteria, the Cytophaga-Flavobacterium cluster of the CFB phylum and the Gram-positive bacteria with a low G+C DNA content (Fig. 4). The majority of sequences (45%) were affiliated to the Gram-positive bacteria with a high G+C DNA content (Fig. 5).

4

Neighbour-joining tree assigning 37 partial sequences from 16S rDNA libraries from bacteria in soils with low metal (with suffix a) and high metal (with suffix d) amendment to the α-, β-, γ- and δ subdivision of Proteobacteria, the Cytophaga-Flavobacterium cluster of the CFB phylum and Gram-positive bacteria with a low DNA G+C content. Clades I and II represent groups that could not be assigned to a known group within the domain Bacteria. Small numbers indicate the percentage of 100 bootstrap resamplings that support some of the major topological elements. Scale bar represents number of substitutions per site.

5

Neighbour-joining tree assigning 31 partial sequences from 16S rDNA libraries from bacteria in soils with low metal (with suffix a) and high metal (with suffix d) amendment to the Gram-positive bacteria with a high DNA G+C content. Clades denoted A–E represent the phylogenetic positions of the clones within the actinomycete line of descent. Small numbers indicate the percentage of 100 bootstrap resamplings that support some of the major topological elements. Scale bar represents number of substitutions per site.

Only minor differences were observed between community structures of bacteria in soils with low metal or high metal amendments. The number of sequences assigned to the α-subdivision of Proteobacteria, for example, was slightly higher (10 clones) in the library from bacteria in soil with high metal amendments compared to the low metal soil (seven clones) (Fig. 4). In this soil, two closely related groups belonging to the α-subdivision of Proteobacteria were found which were not observed in the library from bacteria in the soil with low metal amendment. One of the groups, containing clones 8d, 20d, 52d and 61d, displayed 87–96% sequence similarity to the 16S rRNA of Sphingomonas adhaesiva, while the other group, with clones 7d, 129d and 139d, showed 87% sequence similarity to the 16S rRNA of Rhodopseudomonas palustris and Rhodopseudomonas vannielii (data not shown). A third cluster, containing clones 11a, 9a, 114a and 154a, was derived from bacteria in soil with low heavy metal amendment only.

The number of sequences from Gram-positive bacteria with a high DNA G+C content was slightly higher in the library from bacteria in soil with low metal amendments (18 clones) compared to the high metal soil (13 clones) (Fig. 5). A distinct group of sequences belonging to Gram-positive bacteria with a high DNA G+C content (clones 26d, 78d and 79d) was observed in the library from soil with high metal amendment only. All sequences belonging to the Gram-positive bacteria with a high DNA G+C content belonged to the actinomycete line of descent. Many of the clones showed sequence similarities of 85% up to 99% to sequences of unidentified actinomycetes from acid forest soil [27], a peat bog [28] and a paddy field [25]. The clones clustered into five distinct groups namely groups A, B, C, D and E (Fig. 5). Two of the groups (groups A and B) showed sequence similarity to Acidimicrobium ferrooxidans and Microthrix parvicella, while another group clustered together with Terrabacter tumescens (group C). Groups D and E showed no sequence similarity to any hitherto cultured actinomycete.

4 Discussion

Analysis of DNA reassociation kinetics indicated a dramatic decrease in bacterial diversity with increasing contamination of the soils. In the present reassociation experiments, we obtained reliable measurements for a longer time period than in previous experiments [18, 29, 30]. Thus, the reassociation included a larger fraction of rare DNA types. This is reflected by the high Cot1/2 values as compared to previous experiments, where the Cot1/2 were extrapolated from reassociation of the most dominating DNA types [29]. This has made it necessary to modify the equations taking into account that the reassociation reaction is not a true second order reaction. In the non-contaminated soil, Cot1/2 values corresponded to approximately 16 000 different bacterial genomes (g soil [wet wt])−1, while in soil with low metal amendments and high metal amendments the corresponding values were 6400 and 2000 bacterial genomes, respectively. These differences in Cot1/2 values are in accordance with other studies which usually observe a much higher bacterial diversity in pristine than in perturbed environments (approximately 6000 genomes (g soil [dry wt])−1 in pristine forest soils versus 1000–2000 genomes (g soil [dry wt])−1 in two agricultural soils and 12 000 genomes (g sediment [dry wt])−1 in pristine sediment versus 100 genomes (g sediment [dry wt])−1 in adjacent fish farm sediment) [18, 20, 30]. In the soil with low metal amendments, the genetic complexity of the bacterial community decreased by approximately 60% compared to that in the non-contaminated soil, while the complexity in soil with high metal amendments decreased by approximately 90%. The fact that there are small differences in environmental conditions between original soil and soil with low metal amendments, with respect to organic carbon content, pH and total concentrations of heavy metals (Table 1), suggests that the presence of even small amounts of heavy metals caused a substantial reduction in the total bacterial diversity. The reduction in bacterial diversity was most pronounced in soil with high metal amendments. However, in studies of natural ecosystems, it is often difficult to separate the influence of heavy metals from those of pH and other environmental factors, since pH has large effects on the solubility of heavy metals [2]. In our study, the increase in the concentration of heavy metals in the high metal soil compared to the control soil was accompanied by a two-unit decrease in pH. It has been observed in other studies that a one-unit decrease in pH results in a two-fold increase in the concentration of heavy metals (Cd, Ni and Zn) in solution [31, 32].

No pronounced differences were observed in the %G+C profiles of DNA obtained from bacteria of the non-contaminated low metal or high metal soils (Fig. 2). This indicates that bacteria with a relatively high DNA G+C content dominated the community in the original as well as in the contaminated soils. A closer assignment to different bacterial populations might be possible based on the comparative analysis of the %DNA G+C contents of commonly enumerated bacterial genera [33]. Such an assignment, however, remains highly speculative because different species might have the same DNA %G+C content and changes at the species level might occur without influencing the %G+C composition in the bacterial community. Analysis of %G+C base profiles might give an indication of differences in bacterial community structure at the community level. Such analyses have been used to demonstrate differences in bacterial communities of aquatic and terrestrial environments [34], of compartments in an activated sludge system [35] and of compartments in the gastrointestinal tract of chickens [36]. In terrestrial environments, the bacterial communities of two pristine forest soils have been shown to differ in %G+C base profiles [20]. On the other hand, bacterial communities in soils, experimentally contaminated with either Cu, Pb, Ni or Zn, displaying similar %G+C base profiles have been shown to be separable by statistical analyses [13].

Total numbers of bacteria were slightly lower in the low and high metal soils compared to the control soil (Table 1). Assuming the same biovolume of bacteria in these samples, the biomasses decrease with increasing heavy metal contamination, which is in accordance with earlier observations by other methods [12].

Hybridisation with both the EUB338 and ARCH915 probes resulted in values similar to those of earlier studies in pristine environments [20, 23] and did not indicate differences in detectability of bacteria in the three soils under study. The decrease in numbers of Archaea with increasing heavy metal contamination indicates that heavy metals may influence growth or survival of some of the Archaea. The significance of this observation, however, is questionable, since the percentages of DAPI-stained cells are around the detection limit set at 1%[20, 23].

Our failure to detect all cells by hybridisation with probes EUB338 and ARCH915 might be due to restricted breadth of the probes and lack of full coverage of the microbial community by in situ hybridisation due to a restricted permeability of vegetative and dormant cells for probes. For example some clones, assigned to the Gram-positive bacteria with a high DNA G+C content (clones 11d, 17d, 26d and 81d), did not hybridise to probe EUB338. Low levels of rRNA per cell and secondary structure could be other explanations for the observed discrepancy between DAPI counts and the numbers of cells hybridising with the two domain probes [37]. Likewise, only approximately 10% of the cells detected by the EUB338 probe could be affiliated to larger phylogenetic groups such as α-, β-, γ- and δ-subdivisions of Proteobacteria, bacteria of the Cytophaga-Flavobacterium cluster of the CFB phylum and Gram-positive bacteria (Fig. 3A). This percentage is much lower than the approximately 25% obtained in pristine forest soils in which members of the α- and δ-subdivisions of Proteobacteria and of the Planctomycetes were found to be numerically dominant [20, 23]. This phenomenon might be due to restricted breadth of probes and lack of suitable sets of probes covering all organisms within the domain Bacteria. The specificity of the phylogenetic probes is also not absolute [3842]. Further, probes and hybridisation conditions are generally designed on the basis of sequences from culturable bacteria and may therefore not be applicable to all members of the different phylogenetic groups. Another methodological drawback might be that spores of Gram-positive bacteria with a high G+C content are often impermeable to fluorescent probes [43]. In the present paper, sequencing resulted in assignment of approximately 44% of the clones from the low and high metal soil library to Gram-positive bacteria with a high G+C content, while the corresponding value by in situ hybridisation was below 1% of the cells hybridising with the domain bacteria probe.

Only small differences in the percentage of bacteria belonging to most of the different phylogenetic groups were seen when comparing the three different soils (Fig. 3A). However, there was a pronounced increase in cells detected with probe ALF1b (Fig. 3A). This shift in bacterial community structure suggests that bacteria belonging to the α-subdivision of Proteobacteria might have selective advantages over other bacteria in soil with high metal amendments. Other studies on soils from the same plots demonstrated an increase in numbers of transferable plasmids mediating metal resistance in the Rhizobium leguminosarum bv. viciae with increasing metal contamination [44]. The shift in soil bacterial community structure might therefore be caused by an increase in one initially resistant population of a member of the α-subdivision of Proteobacteria in the soils. Another hypothetical explanation is occurrence of plasmid-mediated transfer of heavy metal resistance within members of the α-subdivision of Proteobacteria. Plasmid transfer has been shown to occur more likely between members of the same phylogenetic group than crossing phylogenetic group borders ([45], unpublished data).

Dot blot analysis of cloned fragments in 16S and 23S rDNA libraries from bacteria in soils with low metal or high metal amendments showed essentially the same distribution profiles of the phylogenetic groups as in situ hybridisation (Fig. 3B). These results suggest that clones in the libraries reflected the abundance of the target sequences in the original sample. This might not necessarily always be the case, since the PCR reaction might be biased [46, 47]. However, comparative sequence analysis of 68 randomly selected clones from 16S rDNA libraries from bacteria in both soils only partly confirmed the distribution pattern of phylogenetic groups when analysed by hybridisation. In accordance with the hybridisation data, many sequences were assigned to bacteria of the α-subdivision of Proteobacteria, however, in contrast to the hybridisation data, the majority of sequences were affiliated to the Gram-positive bacteria with a high G+C DNA content. This discrepancy might be due to differences in the libraries analysed (16S versus 23S rDNA libraries) or limited probe specificity and coverage because the design of probe HGC69a was based on a limited number of sequences [41].

Similar to other studies on environmental clone libraries [4851], most sequences in our study were similar to sequences from described organisms, still some sequences could not be assigned to the major taxa described by Olsen et al. [52] (Figs. 4 and 5). A high percentage of bacteria belonging to the α-subgroup of Proteobacteria are also found in other studies of soil diversity [53]. Most of the clones, belonging to the Gram-positive bacteria with high G+C DNA content, showed low sequence similarities to described species, but moderate to high sequence similarities with hitherto uncultured organisms falling within the actinomycete line of descent [25, 27, 28, 54, 55]. It is interesting that these uncultured taxa are detected in habitats of different physicochemical composition, such as forested soil, paddy fields, soybean field and peat, and also from different geographical locations, such as Finland, Germany, Australia and Japan. In these studies the percentage of clones clustering within the actinomycete line of descent were comparable, from 7% to 23%. In the present study, however, 45% of the clones belonged to these taxa. This high percentage may reflect the effect of either sludge or heavy metal amendments. The distribution into five phylogenetic groups (Fig. 5) resembles the finding of Rheims et al. [28] where unculturable members of the actinomycete line of descent formed a distinct pattern of three groups. In the present study, some of the clones formed groups together with known members of the actinomycete line of descent, while the majority of the clones formed three groups with low sequence similarity to known actinomycetes.

Concerns have been raised with respect to the influence of biases in molecular ecological studies [56, 57]. Although these biases undoubtedly influence the taxon composition of the clone library obtained, it is noteworthy that the uncultured taxa of the actinomycetes line of descent have been found in different clone libraries even though different experimental procedures were applied [25, 27, 28, 54, 55].

Our results demonstrate long-term effects of heavy metals on total bacterial communities in contaminated samples from field sites. Compared to non-contaminated soil, a pronounced reduction of bacterial diversity as well as changes in bacterial community structure were obtained even in the presence of low metal concentrations below the upper legal limits set by the European Union. More information on fine-level differences in bacterial populations in soils with low metal or high metal amendments could probably be obtained by focusing the analysis on a specific group of bacteria. The increase in the percentage of bacteria belonging to the α-subdivision of Proteobacteria with increasing metal contamination as well as the prominent occurrence of sequences affiliated to the Gram-positive bacteria with a high DNA G+C content in both soils suggest that both subgroups might be influenced by metal amendments. The biodiversity of both groups might therefore serve as an indicator for effects of metal contamination.

Acknowledgments

This work was supported by grants from the European Commission on ‘Microbial Diversity and Functions in Metal Contaminated Soils’ and ‘High Resolution Automated Microbial Identification-2’. The authors are indebted to Dr Bruce Knight (IACR-Rothamsted, Harpenden, Herts, UK) for supplying physiological and chemical parameters of the soils used in the study.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
  48. [48].
  49. [49].
  50. [50].
  51. [51].
  52. [52].
  53. [53].
  54. [54].
  55. [55].
  56. [56].
  57. [57].
  58. [58].
  59. [59].
  60. [60].
  61. [61].
View Abstract