Soil Biology & Biochemistry 38 (2006) 1430–1437

www.elsevier.com/locate/soilbio

Response of microbial activity and microbial community composition

in soils to long-term arsenic and cadmium exposure
Nicola Lorenz a,*, Therese Hintemann b, Tanja Kramarewa c, Arata Katayama d,
Tsuyoshi Yasuta d, Petra Marschner e, Ellen Kandeler a
a
Institute of Soil Science and Land Evaluation, University of Hohenheim, Emil-Wolff-Str. 27, 70599 Stuttgart, Germany
b
Institute of Plant Nutrition, University of Bonn, Karlrobert-Kreiten Straße 13, 53115 Bonn, Germany
c
Institute of Soil Science and Agrochemistry, University of Voronesh, 394693 Voronesh, Russian Federation
d
Research Center for Advanced Waste and Emission Management, University of Nagoya, 464-8603 Nagoya, Japan
e
Soil and Land Systems, School of Earth and Environmental Sciences, The University of Adelaide, SA 5005, DP 636, Australia
Received 7 December 2004; received in revised form 6 October 2005; accepted 26 October 2005
Available online 4 January 2006

Abstract
Arsenic (As) and cadmium (Cd) in soils can affect soil microbial function and community composition and, therefore, may have effects on soil
ecosystem functioning. The aim of our study was to assess the effects of long-term As and Cd contamination on soil microbial community
composition and soil enzyme activities. We analyzed soils that have been contaminated 25 years ago and at present still show enhanced levels of
either As, 18 and 39 mg kgK1, or Cd, 34 and 134 mg kgK1. Soil without heavy metal addition served as control. Polymerase chain reaction (PCR)
followed by denaturing gradient gel electrophoresis (DGGE) showed that bacterial community composition in As and Cd contaminated soils
differed from that in the control soil. The same was true for the microbial community composition assessed by analysis of respiratory quinones.
Soil fungi and Proteobacteria appeared to be tolerant towards As and Cd, while other groups of bacteria were reduced. The decline in alkaline
phosphatase, arylsulphatase, protease and urease activities in the As- and Cd-contaminated soils was correlated with a decrease of respiratory
quinones occuring in Actinobacteria and Firmicutes. Xylanase activity was unaffected or elevated in the contaminated soils which was correlated
with a higher abundance of fungal quinones, and quinones found in Proteobacteria.
q 2005 Elsevier Ltd. All rights reserved.

Keywords: Arsenic; Cadmium; Microbial community composition; Quinones; Denaturing gradient gel electrophoresis; Enzyme activities

1. Introduction Microbial community composition can also be character-

ized by the analysis of respiratory quinones (Fujie et al.,
Concerns about ecotoxicological impacts of heavy metals 1998; Hiraishi et al., 1991; Katayama et al., 2001). These are
and metalloids on soil microorganisms resulted in several essential components in the electron transport systems of most
studies on their effects on soil microbial community organisms, and are present in the cell membrane of
composition and function (Giller et al., 1998; Kandeler et al., prokaryotes and the mitochondrial membrane of eukaryotic
2000; Renella et al., 2005). Using denaturing gradient gel cells. Respiratory quinones are chemically composed of
electrophoresis (DGGE), Renella et al. (2004) showed that benzoquinone (or naphthoquinone) with an isoprenoid side
long-term exposure with Cd had little effect on soil bacterial chain. There are two major groups of quinones in soils,
diversity. The effects of single contamination with As on namely ubiquinones (1-methyl-2-isoprenyl-3,4-dimethoxypar-
terrestrial microbial communities have not been studied so far abenzoquinone) and menaquinones (1-isoprenyl-2-methyl-
by DGGE. naphthoquinone). In contrast to DGGE, quinone analysis
allows both qualitative and quantitative determination of soil
bacterial and fungal communities. Quinones can be used to
* Corresponding author. Address: School of Natural Recources, The Ohio
State University, 2021 Coffey Road, Columbus, OH 43210, USA. Tel.: C1 614 assess the taxonomic diversity because many microbial
247 6127; fax: C1 614 292 7432. groups are characterized by one major quinone type (Fujie
E-mail address: lorenz.64@osu.edu (N. Lorenz). et al., 1998). The effects of single contamination with As and
0038-0717/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. Cd on microbial communities have not been studied by
doi:10.1016/j.soilbio.2005.10.020 analysing respiratory quinones.
 N. Lorenz et al. / Soil Biology & Biochemistry 38 (2006) 1430–1437 1431

Studies evaluating soil microbial functional diversity in As replicates of the control and the treatments Cd (I), Cd (II), As
and Cd contaminated soils have indicated a varying sensitivity (I) and As (II). The soil was sieved (!2 mm) and an aliquot
of soil enzyme activities (Speir et al., 1999; Wilke, 1988). No dried at room temperature for soil chemical analyses. The
effects of As on the activities of invertase and protease were remaining soil was stored at K20 8C for microbial analyses.
reported, whereas activities of alkaline phosphatase and Prior to the microbial analyses, the soil was thawed by storing
arylsulphatase were reduced (Speir et al., 1999; Wilke, for 2 days at 4 8C, and kept at this temperature not longer than 2
1988). Arsenic toxicity may be explained by competitive weeks during the analyses.
inhibition due to its similarity with phosphate (Juma and
Tabatabai, 1977; Speir et al., 1999). Furthermore, As(III) 2.3. Soil organic carbon and pH
covalently bonds to sulfur, and, thus, uncouples oxidative
phosphorylation. This interferes with protein synthesis and Total soil C was determined by dry combustion at 1250 8C
enzyme activities (Tamaki and Frankenberger, 1992). using a LECO 2000 CN elemental analyzer. Since the soil
Exposure to Cd reduced activities of soil alkaline phospha- samples were free of carbonates, the total carbon content was
tase, arylsulphatase and protease (Renella et al., 2005; Wilke, regarded as a measure of soil organic carbon. Soil pH was
1988). According to Renella et al. (2005), microbial adaptation determined in 0.01 M CaCl2 at a soil:solution ratio of 1:2.5
processes to Cd result in higher energy requirement for after 2 h using a glass membrane electrode.
microbial growth and/or the conversion of substrates. This
might explain the decline of both enzyme syntheses and 2.4. Heavy metals
activities. Enzyme activities may also be reduced by binding
of Cd2C to sulphydryl groups (Sanadi, 1982). For the determination of total heavy metal contents, 3 g soil
The aim of this study was to assess the effect of single doses was boiled for 16 h in a mixture of 21 ml 37% HCl (v/v) and
of As and Cd on microbial community composition and 7 ml 65% HNO3 (v/v) (aqua regia) according to DIN ISO
enzyme activities. 11466: 06.97 (1995). Heavy metal concentration in the extracts
was determined by ICP-MS.
2. Materials and methods
2.5. DNA extraction, polymerase chain reaction (PCR) and
2.1. Study site denaturing gradient gel electrophoresis (DGGE)

In 1975, a field experiment was established at the Federal Soil DNA was extraction using the Fast DNAe SPIN Kit
Biological Research Centre for Agriculture and Forestry for Soil (Qbiogene, Carlsbad, CA, USA) following the
(Biologische Bundesanstalt), Berlin (West), Germany, to study manufacturer’s protocol. PCR for bacterial 16S rDNA genes
the effects of inorganic pollutants on soils and plants. The trial was performed using the primer set F984-968GC (5 0 -CGC
consists of 112 plots, 96 of them contaminated with 12 different CCG GGG CGC GCC CCG GGC GGG GCG GGG GCA CGG
pollutants (As, Be, Br, Cd, Cr, F, Pb, Hg, Ni, Se, Sn and V), and 16 GGG GAA CGC GAA GAA CCT TAC-3 0 ) and R1378-1401
were not contaminated. For our study, we selected the plots (5 0 -CGG TGT GTA CAA GGC CCG GGA ACG-3 0 ) (Heuer
contaminated with cadmium chloride (Cd) and sodium arsenate et al., 1997). A GC clamp was attached to the forward primer to
(As), and a control plot. The contaminated plots received single prevent complete separation of the strands during DGGE
elements at two concentrations. In 1975 and 1976, the salts were (Muyzer et al., 1993).
applied to the soil surface and mixed manually into the top 30 cm. For PCR, 2 ml DNA extract (w10 ng) was added to 23 ml
In total, 50 mg Cd kg soilK1 (Cd I) and 250 mg Cd kg soilK1 (Cd PCR Mastermix. The latter contained 17.5 ml ultrapure water,
II), and 50 mg As kg soilK1 (As I) and 300 mg As kg soilK1 (As 2 ml of 2 mm dNTPs (Qbiogene, Carlsbad, CA, USA), 2.5 ml
II) were applied. The experimental site has a split-plot design, and 10!PCR-buffer (Qbiogene, Carlsbad, CA, USA), 0.4 ml of
each treatment consists of four replicates. Each replicate is each primer (5 mM) (Metabion, Martinsried, Germany) and
surrounded by concrete walls that extend from 0.2 m above 0.2 ml Taq Polymerase (equivalent to 1 U) (Qbiogene,
ground to 1 m-depth below ground (Wilke, 1988). Carlsbad, CA, USA). The PCR was performed with a T3
The soil is a Sandy Cambisol (FAO), free of carbonates thermocycler (Biometra, Göttingen, Germany). Cycling started
(!1%). Soil texture is a loamy sand. Mean annual temperature with 5 min at 94 8C, followed by 34 cycles of 1 min at 94 8C,
at the site is 9 8C and the mean annual precipitation is 450 mm. 1 min at 55 8C and 2 min at 72 8C. The 34th cycle was followed
Since 1976, crop rotations included wheat, rye, carrots, beans, by a 10 min extension at 72 8C. Amplification was verified
potatoes, and lettuce, with manual tillage and manual weed by agarose gel electrophoresis (1.5% w/v agarose gel). The
control. Mineral fertilizers were applied each year depending PCR-products were purified prior to further analysis with the
on the crop species. QIAquick PCR purification kit, according to the manufac-
turer’s protocol (Qiagen, Valencia, CA, USA).
2.2. Soil sampling and sample preparation The DGGE was performed with 6% acrylamide gels
containing a linear denaturing gradient from 35 to 55%
At the time of sampling, the plots were not vegetated (bare [100% denaturant contained 7 M urea and 40% (v/v)
fallow). Soil (0–30 cm) was sampled in spring 2000 from four formamide]. Electrophoresis was carried out at 60 8C and
1432 N. Lorenz et al. / Soil Biology & Biochemistry 38 (2006) 1430–1437

90 V for 16 h in 1!TAE buffer with a DCode System (Bio- reducing sugars were determined as described for xylanase
Rad, Laboratories, Hercules, CA, USA). activity.
After electrophoresis, the gels were silver-stained (Merril Measurement of protease activity was based on the method
et al., 1981), scanned and analyzed with the Geldoc Quantity of Ladd and Butler (1972), with slight modifications. Briefly,
One software (vers. 4.01, Bio-Rad Laboratories, Hercules, CA, 0.3 g field moist soil was incubated in 5 ml 2% (w/v) casein
USA). After average background subtraction for the entire gel solution (casein dissolved in 0.05 M Tris buffer, pH 8.1) and
and background subtraction of single lanes (rolling disk 50), 5 ml 0.05 M Tris buffer (pH 8.1), at 50 8C for 2 h. The released
banding patterns were digitized. The normalized intensity and aromatic amino acids were extracted with 5 ml 0.9 M
position of individual bands relative to those of a standard were trichloroacetic acid, and then measured colorimetrically at
transferred into spreadsheets for statistical analysis. 700 nm after addition of 5 ml 0.2 M Folin–Ciocalteu reagent.
For the determination of urease activity, 1 g field moist soil was
2.6. Quinone extraction and determination incubated with 1.5 ml 0.08 M urea solution at 37 8C for 2 h.
The activity was determined colorimetrically at 660 nm by a
Quinones were extracted from soils according to the method modified Berthelot reaction (Kandeler and Gerber, 1988).
described by Fujie et al. (1998). Briefly, 20 g field moist soil Measurement of alkaline phosphatase activity was based on
was suspended in 80 ml of a chloroform–methanol mixture the method of Hoffmann (1968), with slight modifications.
(2:1, v/v), sonicated on ice for 10 min (20 kHz; output power, Briefly, 0.3 g field moist soil was incubated with 2 ml 0.2 M
100 W), and centrifuged at 5000g for 10 min. The supernatant borate buffer (pH 10.0) and 1 ml 0.1 M phenylphosphate
(lipid layer) was collected and passed through No. 2 filter paper solution at 37 8C for 3 h. The release of phenol was determined
(Advantec, Tokyo, Japan) to remove small soil particles. colorimetrically at 614 nm, using 2,6-dibromchinone-
The residue was extracted twice with 80 ml of chloroform– chlorimide.
methanol mixture (2:1, v/v). The extracts were combined and Determination of arylsulphatase activity was performed
evaporated under vacuum to a volume of 60 ml. The extracts according to Tabatabai and Bremner (1970), but slightly
were mixed with 30 ml of saline solution (1% v/v CaCl2, 10% modified. Briefly, 0.3 g field moist soil was incubated in 0.3 ml
v/v NaCl) and re-extracted twice with 50 ml n-hexane. The 0.02 M p-nitrophenyl-sulfate and 1.2 ml 0.5 M Na-acetate
hexane extract (100 ml) was evaporated under vacuum to 5 ml, buffer (pH 5.8) at 37 8C for 1 h. The release of p-nitrophenol
and then transferred to a chromatography column with Sep-Pak was determined colorimetrically at 420 nm after addition of
Plus Silicae cartridges to separate the menaquinone (MK) and 0.5 M NaOH solution.
ubiquinone (Q) fractions. The MK and Q fractions were eluted
by 20 ml of a 2% diethylether–hexane mixture (v/v) and 20 ml
2.8. Statistical analysis
of a 10% diethylether–hexane mixture (v/v), respectively.
Finally, the eluates were evaporated to dryness under N2 gas
The results of the soil chemical and biochemical analyses
stream and stored at K20 8C.
were calculated on the basis of dry mass and presented as
For the determination of the quinones, the samples were
arithmetic means, with standard deviations, of four replicates.
dissolved in 0.2 ml acetone and analyzed by a high
The results of soil chemical data and enzyme activities were
performance liquid chromatograph (HPLC) equipped with a
tested for normal distribution (K–S test) prior to statistical
photodiode array detector. The analytical procedure has been
analysis. Significant differences between control and As or Cd
described previously (Fujie et al., 1998; Katayama et al., 2001).
treatment means were identified using the U-test. Multiple
The HPLC-column was a Zorbax-ODS (Agilent Technologies,
comparisons between all treatments were performed with the
Tokyo, Japan). Eighteen quinone species were detected. The
Tukey-HSD test. Correlations were calculated by Spearman’s
quinones Q-8, Q-9, Q-10, and all 13 MK may be assigned to
correlation coefficient (SPSS for Windowsq, version 11.5.1).
soil bacteria whereas Q-9 and Q-10(H2) are indicators of soil
Community analysis based on quinones and DGGE patterns
fungi (for details see Katayama et al., 2001).
was performed by principle component analysis (PCA), with
focus on inter-species correlations (CANOCO, version 4.0 for
2.7. Enzyme activities
Windowsq, Microcomputer Power, Ithaca, NY, USA). For
further details see ter Braak and Smilauer (1998).
Xylanase activity was determined after incubating 0.3 g
Dissimilarities (%) of the quinone profiles were calculated
field moist soil with 5 ml substrate (1.7% w/v xylan-suspension
according to Hiraishi et al. (1991) as
in 2 M Na-acetate buffer, pH 5.5) and 5 ml 2 M Na-acetate
buffer (pH 5.5) at 50 8C for 24 h (Schinner et al., 1996). After
incubation, potassium hexacyanoferrate (III) and hexacyano-
ferrate (II) were added. Reducing sugars were measured 1X n
Dði; jÞ Z jP KPkj j !100;
colorimetrically at 690 nm by the Prussian Blue reaction 2 kZ1 ki
according to Schinner et al. (1996). Invertase activity was
determined by incubating 0.3 g field moist soil with 5 ml
50 mm sucrose-solution and 5 ml 2 M Na-acetate buffer where Pki, Pkj are the mole fractions of the k quinone species
(pH 5.5) at 50 8C for 3 h (Schinner et al., 1996). The released for the samples i and j, respectively.
 N. Lorenz et al. / Soil Biology & Biochemistry 38 (2006) 1430–1437 1433

Table 1 3.2. Bacterial community composition determined by DGGE

Soil pH, content of organic carbon (Corg) and aqua regia-extractable heavy
metals of control and contaminated soils (nZ4; different letters indicate
significant differences between the treatments; Tukey-HSD-test, P!0.05)
Two representative DGGE gels are shown in Fig. 1. In the
control and As II soils, similar numbers of DGGE bands were
Treatment pH Corg (%) Cd (mg As (mg detected (21 and 20, respectively). In treatment As I, the
Cd kgK1) As kgK1)
number of DGGE bands was significantly higher (23). On the
Control 6.2 a 1.40 a 5.1 a 2.6 a other hand, the number of bands in Cd I and Cd II soils were
As I 6.3 a 1.43 a 5.5 a 18.1 b
As II 6.2 a 1.38 a 5.5 a 39.3 c
significantly reduced to 16 and 15, respectively. In both Cd
Cd I 6.3 a 1.37 a 34.2 b 2.5 a treatments, only few bands were found in the upper part of the
Cd II 6.2 a 1.27 b 134.4 c 2.3 a gel, containing DNA fragments with a low GC-content. The
PCA analysis of the DGGE banding patterns showed three
main clusters (Fig. 2). The control and As-contaminated soils
3. Results were separated from the Cd soils along axis 1, explaining
27.7% of the total variance. Furthermore, control and As-
3.1. Soil organic carbon (Corg), pH, total arsenic and cadmium polluted soils separated along axis 2. However, this axis
explained only an additional 12.5% of the variance. In total, the
Soil pH did not differ between the treatments, with values two axes of the PCA plot explained 40.2% of the total variance
between 6.2 and 6.3 (Table 1). The control soil and the of the data. Dissimilarity values revealed that DGGE patterns
treatments As I and As II, and Cd I had similar Corg contents, differed between the control and the contaminated soils
ranging from 1.37 to 1.43% Corg. However, the organic C (Table 3).
content in treatment Cd II was significantly lower (1.27%).
Arsenic contents in the treatments As I and As II were 3.3. Microbial community composition determined by quinones
significantly different (18.1 and 39.3 mg kgK1, respectively).
The As contents in the control ranged from 2.3 to 2.6 mg kgK1. The total concentration of quinones was highest in the
Treatments Cd I and Cd II had significantly higher total control soil, and decreased with increasing soil pollution with
cadmium contents than the control soil (34.2 and Cd having a greater effect than As (Table 2). For both metals, a
134.4 mg kgK1 vs. 5.1 mg kgK1, Table 1). The relatively decrease of the following eight menaquinones was observed
high Cd content in the control as well as in the As I and II with increasing soil contamination: MK-7(H4), MK-8, MK-
treatments (O5 mg kgK1) is due to a previous contamination 8(H2), MK-8(H4), MK-9, MK-9(H2), MK-9(H4) and MK-
(Wilke, 1988). Levels of total Pb, Cr, Ni, and Hg (30, 29, 9, and 9(H8). The decrease of MK-9(H8) was more pronounced in the
0.3 mg kgK1, respectively) in all treatments were not con- Cd-contaminated than in the As-contaminated soils. On the
sidered as elevated (Klärschlammverordnung, 1992). other hand, concentrations of Q-8, Q-9, Q-10, Q-10(H2) and

Fig. 1. Silver stained DGGE profiles of 16S rDNA gene fragments from the control, As and Cd contaminated soils. S, standard.
1434 N. Lorenz et al. / Soil Biology & Biochemistry 38 (2006) 1430–1437

In total, 39.1% of the total variance was explained by the two-

dimensional PCA plot. Dissimilarity values of quinone patterns
indicated significant differences between the control and As-
and Cd-contaminated soils (Table 3).

3.4. Enzyme activities

Arsenic contamination did not significantly affect the

activities of xylanase, invertase, protease and alkaline
phosphatase (Fig. 4). Only arylsulphatase activity was
significantly lower in As-contaminated soils compared to the
control. In contrast, urease activity was higher in As I than in
the control soil, but did not differ from the control for As II.
Contamination with Cd resulted in a decrease of enzyme
activities by 30% and up to 70% for protease, urease, alkaline
phosphatase and arylsulphatase but had no significant effect on
invertase activity (Fig. 5). Activities of invertase, protease and
arylsulphatase were lower in the Cd II soils compared to Cd I.
However, xylanase activity was 50% higher in both Cd-
Fig. 2. PCA from DGGE profiles of 16S rDNA gene fragments amplified from contaminated soils (I and II) than in the control.
DNA extracted from As (As I and As II) and Cd (Cd I and Cd II) contaminated
and control soils (nZ4). Values on the x and y axis indicate percent of variation
explained by the axis. 3.5. Correlations of enzyme activities with quinones

the menaquinone MK-7 did not differ between the control and Correlations between enzyme activities and quinone species
the contaminated soils. The relative abundance (%Q) of were only shown for quinones that indicated at least one
ubiquinones Q-8, Q-9, Q-10, and Q-10(H2) was higher in the significant correlation to enzyme activities (Table 4). The
contaminated plots than in the control plot, but this increase significant correlations were all positive. Protease, alkaline
was only significant in treatments As II and Cd I (Table 2). phosphatase, urease, arylsulphatase and invertase activities
Similar to DGGE, the PCA plot of quinone patterns showed were correlated with the total quinone concentration (TQ).
that microbial communities of Cd-contaminated soils were Arylsulphatase was correlated with all menaquinones listed
separated from the control along axis 1 (Fig. 3). Control and As (MK-7, MK-7(H4), MK-8, MK-9, MK-9(H2), MK-9(H4),
soil communities were separated along axis 2. However, one MK-9(H6), MK9(H8)), and alkaline phosphatase was
replicate of Cd I had a community composition similar to As II. correlated with all menaquinones listed except MK-7 and
Table 2
Average quinone concentrations of individual quinones, total quinones (pmol gK1) and percentage of ubiquinones (% Q) (nZ4; different letters indicate significant
differences between treatments; Tukey-HSD-test, P!0.05)

Quinones Control As I As II Cd I Cd II
Q-8 35.7 a 30.0 a 37.4 a 52.6 a 30.3 a
Q-9 26.1 a 30.0 a 44.2 a 27.3 a 24.7 a
Q-10 75.3 a 60.6 a 53.7 a 67.6 a 53.0 a
Q-10(H2) 10.8 a 8.8 a 6.5 a 11.4 a 9.6 a
MK-7 79.2 a 74.0 a 64.5 a 71.0 a 67.8 a
MK-7(H4) 50.3 a 45.5 ab 37.8 bc 35.0 b 40.0 bc
MK-8 139.6 a 114.0 b 98.5 b 114.0 b 106.3 b
MK-8(H2) 53.1 a 41.5 b 34.5 b 44.3 ab 35.0 b
MK-8(H4) 69.0 a 65.3 a 48.3 ab 41.0 b 36.0 b
MK-9 32.5 a 27.8 a 9.3 b 11.8 b 16.8 b
MK-9(H2) 27.0 a 23.0 a 12.5 b 22.3 a 18.8 ab
MK-9(H4) 37.7 a 31.8 ab 19.8 b 15.8 b 17.0 b
MK-9(H6) 6.4 a 7.0 a 4.0 a 7.0 a 5.8 a
MK-9(H8) 85.0 a 78.0 ab 67.3 b 65.3 b 45.8 c
MK-10 3.1 a 0 0 0.3 b 0
MK-10(H2) 0.4 a 0 0 0.3 a 0
MK-10(H4) 2.0 a 2.8 a 1.3 a 1.8 a 2.8 a
MK-10(H6) 0 0.3 a 0.5 a 1.5 a 0
Total quinones 733.2 a 640.1 ab 539.7 bc 589.8 ab 509.5 c
%Q 19.5 a 20.1 a 26.2 b 26.9 b 23.1 ab

% QZ(Q-8CQ-9CQ-10CQ-10(H2))!100/total quinones.
 N. Lorenz et al. / Soil Biology & Biochemistry 38 (2006) 1430–1437 1435

Fig. 4. Enzyme activities in As contaminated soils. Control soilZ100% (alk.

Fig. 3. PCA of quinone profiles from As (As I and As II) and Cd (Cd I and Cd II) phosph.: alkaline phosphatase; arylsulph.: arylsulphatase) (nZ4). Significant
contaminated and control soils (nZ4). Values on the x and y axis indicate different from the control *P!0.05, different letters indicate significant
percent of variation explained by the axis. differences between treatments (U-test).

MK-9(H6). Protease, urease and invertase each were correlated bacterial community composition only to a minor extent. This
with 5 menaquinones. Arylsuphatase was correlated with Q-10, may be explained by the lower Cd content in the soils studied
invertase with Q-8, Q-10 and Q-10(H2), and xylanase with the by Renella et al. (2005) compared to our study.
sum of fungal quinones (FQ), Q-8, and Q-10(H2). Shifts in microbial community composition due to As and
Cd were indicated by quinone data, shedding new light on the
response of specific groups of microorganisms, including
4. Discussion bacteria (e.g. Proteobacteria, Firmicutes, Actinobacteria,
Cytophaga-Flavobacterium group) and fungi. The shifts in
4.1. Response of the soil microbial community composition
the overall composition of quinones were due to either
to As and Cd
reduction or increase of single quinones. For both metals, a
decrease of most menaquinones was observed with increasing
Soil bacterial community composition was altered by As
soil contamination. MK-8, characteristic for Firmicutes or
and Cd, as indicated by DGGE patterns. The effect, however,
Actinobacteria in aerobic conditions, was strongly reduced.
was greater in the Cd than in As contaminated soil. This might
In addition, MK-7(H4), MK-8(H2), MK-8(H4), MK-9,
be due to higher Cd than As concentrations, hence our results
do not allow a direct comparison of Cd and As toxicity. Both MK-9(H2), MK-9(H4) and MK-9(H8) decreased, which are
the PCA plot and dissimilarities of the DGGE profiles revealed main components of Actinobacteria (Hiraishi et al., 1998). The
that As and Cd contaminated soils were clearly separated from most obvious effect was the decrease of MK-9(H8) in the
the control. This was due to differences in band positions, band Cd-contaminated soil, indicating a negative effect of Cd
intensities and numbers of bands. The number of bands in As I pollution on Actinobacteria containing MK-9(H8), i.e. Micro-
soil did not differ from the control, but more DGGE bands were monosporineae, Streptmycineae and Streptosporangineae
visible in the As II soil. The number of DGGE bands was (Katayama et al., 2001). Kelly et al. (2003) reported that
significantly reduced in Cd contaminated soils and the Actinobacteria are reduced in Zn contaminated soils.
dissimilarity between control and Cd soils was high. Fewer
bands were visible in the upper part of the DGGE gel of the Cd
contaminated soil, containing DNA fragments of bacteria with
a low GC-content. DNA fragments were not cloned and
sequenced, therefore bacterial community shifts remained
unspecific. Renella et al. (2005) showed that Cd altered the
Table 3
Dissimilarity values of quinones and DGGE bands in control and contaminated
soils (nZ4, U-test, n.s.Znot significant, **P!0.01, ***P!0.001)

Treatment Quinones DGGE

Dissimilarity P Dissimilarity P
Same treatment 7.7 n.s. 54 n.s.
Control–As I 6.7 n.s. 67 ** Fig. 5. Enzyme activities in Cd contaminated soils. Control soilZ100% (alk.
Control–As II 11.7 ** 81 *** phosph.: alkaline phosphatase; arylsulph.: arylsulphatase) (nZ4). Significant
Control–Cd I 12.6 ** 76 *** different from the control *P!0.05, different letters indicate significant
Control–Cd II 11.1 ** 81 *** differences between treatments (U-test).
1436 N. Lorenz et al. / Soil Biology & Biochemistry 38 (2006) 1430–1437

Table 4
Correlations between enzyme activities and quinones (nZ20, Spearman’s correlation coefficient, *P!0.05, **P!0.01)

Protease Alkaline phosphatase Urease Arylsulphatase Invertase Xylanase

TQ 0.66** 0.63** 0.63** 0.75** 0.53* K0.16
FQ 0.03 K0.18 K0.37 0.34 0.33 0.58**
Q-8 0.19 K0.02 K0.12 0.38 0.61** 0.49*
Q-10 0.25 0.30 0.10 0.51* 0.75** 0.30
Q-10(H2) 0.15 0.04 0.01 0.03 0.55* 0.45*
MK-7 0.35 0.43 0.30 0.64** 0.48* K0.00
MK-7(H4) 0.29 0.54* 0.46* 0.46* 0.06 K0.27
MK-8 0.37 0.47* 0.20 0.77** 0.38 0.07
MK-8(H2) 0.48* 0.57** 0.27 0.79** 0.53* 0.17
MK-8(H4) 0.70** 0.82** 0.81** 0.66** 0.40 K0.41
MK-9 0.48* 0.70** 0.58* 0.59** 0.19 K0.13
MK-9(H2) 0.35 0.55* 0.35 0.74** 0.48* 0.11
MK-9(H4) 0.58** 0.77** 0.73** 0.65** 0.20 K0.33
MK-9(H6) 0.39 0.30 0.19 0.53* 0.56** 0.34
MK-9(H8) 0.56* 0.88** 0.79** 0.78** 0.55* K0.38

TQ, total quinones; FQ, fungal quinones, Q-9CQ-10(H2).

Furthermore, Hiroki (1992) showed that Actinobacteria were out that the high sensitivity of phosphatase activity towards As
the microbial group mostly affected by Cd, Cu and Zn. is likely to be due to structural similarity of phosphate and
In contrast to the menaquinones, the content of several arsenate (Juma and Tabatabai, 1977). Furthermore, As (III)
ubiquinones (Q) did either not change or even increased covalently bonds to sulfur and, thus, uncouples oxidative
indicating resistance or tolerance mechanisms of several phosphorylation and interferes with protein synthesis (Tamaki
groups of microorganisms (Sandaa et al., 1999; Silver and and Frankenberger, 1992). In contrast, activities of invertase,
Phung, 1996). Q-8, Q-9, and Q-10 occur mainly in xylanase, and urease were not affected by As. Speir et al.
Betaproteobacteria, Gammaproteobacteria and Alphaproteo- (1999) and Tabatabai (1977) also found no inhibitory effect of
bacteria, respectively. Gammaproteobacteria (Q-9) include As on urease activity. The reason for this response remains
genera such as Acinetobacter and Pseudomonas. Quinone Q-9 unclear. The activities of invertase and xylanase in As- and Cd-
is also assigned to soil fungi, whereas Q-10(H2) is found contaminated soils were unaffected or even higher compared to
exclusively in fungi. Turpeinen et al. (2004) observed a the control. This effect on enzymes involved in breakdown of C
positive relationship between the bioavailability of As and the compounds may be the result of a higher C requirement for
proportion of As (III) resistant species of Acinetobacter, microbial maintenance and repair processes under heavy metal
Edwardsiella, Enterobacter, Pseudomonas, Salmonella and toxicity (Fließbach et al., 1994; Kandeler et al., 1996). The
Serratia, which are classified into Betaproteobacteria or high level of Cd contamination in our study significantly
Gammaproteobacteria. Other studies have shown that Proteo- reduced the activities of protease, urease, alkaline phosphatase
bacteria (Sandaa et al., 1999) and fungi (Fließbach et al., 1994; and arylsulphatase. This is in accordance with studies by
Hiroki, 1992) were not affected by either As or Cd. Resistance Renella et al. (2005) and Wilke (1988) who also reported
of fungi to heavy metals in soil may be explained by spatial reduced activities for enzymes involved in N-, P-, and
separation. Fungi are mainly located in the sand fraction (Poll S-cycling due to Cd pollution. Reduced enzyme activities
et al., 2003), whereas Cd has been shown to be less adsorbed on may be explained by binding of Cd2C to sulphydryl groups
sand particles compared to clay and silt (Andersen et al., 2002). (Sanadi, 1982).
In summary, DGGE and quinones indicated that long-term The observed changes in enzyme activities were
contamination with As and Cd altered the soil microbial accompanied by changes in microbial community composition,
community composition. However, treatment separation was as indicated by correlations between enzyme activities and
indicated clearly by DGGE profiles. DGGE patterns character- quinone species. The decline in enzyme activities involved in
ized the bacterial community composition and showed a distinct N-, P- and S-cycling was correlated to a decrease in
reduction of bacterial diversity. Treatment separtion was some- Actinobacteria and Firmicutes. The high xylanase activities
how reduced by analyzing quinone profiles. This can be explained observed in the contaminated soils were also associated with a
i.a. due to the fact that the responses of bacteria and fungi were shift in the microbial community composition. In fact, a
analyzed by respiratory quinones, and quinones occuring in fungi significant positive correlation was found between xylanase
and Proteobacteria were not reduced due to As or Cd. activity and fungal ubiquinones. It has been shown previously,
that xylanases are mainly produced by saprophytic fungi
4.2. Impact of As and Cd on soil enzyme activities (Alexander, 1977).
In conclusion, molecular fingerprinting techniques as well
The activities of soil arylsulphatase and alkaline phospha- as classical soil microbiological methods have shown that
tase activities were reduced by As. Speir et al. (1999) pointed long-term exposure of soils to Cd and As did not lead to a
 N. Lorenz et al. / Soil Biology & Biochemistry 38 (2006) 1430–1437 1437

specialized microbial population with a low structural Kandeler, E., Gerber, H., 1988. Short-term assay of soil urease activity using
diversity. Fungi and Proteobacteria appear to be more tolerant colorimetric determination of ammonium. Biology and Fertility of Soils 6,
68–72.
to Cd and As than several other groups of bacteria. Although Kandeler, E., Kampichler, C., Horak, O., 1996. Influence of heavy metals on
the present study helped to understand the link between the functional diversity of soil microbial communities. Biology and
abundance and function of soil microorganisms, further Fertility of Soils 23, 299–306.
research should focus on how growth of specific microorgan- Kandeler, E., Tscherko, D., Bruce, K.D., Stemmer, M., Hobbs, P.J.,
isms, protein synthesis and enzyme expression are linked under Bardgett, R.D., Amelung, W., 2000. Structure and function of the soil
microbial community in microhabitats of a heavy metal polluted soil.
long-term pollution of heavy metals.
Biology and Fertility of Soils 32, 390–400.
Katayama, A., Funasaka, K., Fujie, K., 2001. Changes in the respiratory
Acknowledgements quinone profile of a soil treated with pesticides. Biology and Fertility of
Soils 33, 454–459.
We thank the German Research Foundation (DFG) for Kelly, J.J., Häggblom, M.M., Tate, R.L., 2003. Effects of heavy metal
contamination and remediation on soil microbial communities in the
financial support. Thanks to B.-D. Traulsen from the Federal
vicinity of a zinc smelter as indicated by analysis of microbial community
Biological Research Centre for Agriculture and Forestry in fatty acid profiles. Biology and Fertility of Soils 38, 65–71.
Berlin, Germany, for the kind permission to sample the soils of Klärschlammverordnung, 1992. Bundesgesetzblatt Jg. 1992, Teil I, 912–924.
the long-term experiment. Thanks also to B.-M. Wilke for Ladd, J.N., Butler, J.H.A., 1972. Short-term assay of soil proteolytic enzyme
cooperation. Excellent technical support in the lab was activities using proteins and dipeptide derivatives as substrates. Soil
Biology & Biochemistry 4, 19–30.
provided by Sabine Rudolph (University of Hohenheim) and
Merril, C.R., Dunau, M.L., Goldmann, D., 1981. A rapid and sensitive silver
Karen Baumann (Technical University Cottbus). Thanks to J. stain for polypeptides in polyacrylamide gels. Analytical Biochemistry 110,
Breuer (Federal Institute for Agricultural Chemistry, Univer- 201–207.
sity of Hohenheim) for the analysis of the aqua regia Muyzer, G., de Waal, E.C., Uitterlinden, A.G., 1993. Profiling of complex
extractable fraction of heavy metals in soils. We thank Tom microbial populations by denaturing gradient gel electrophoresis analysis of
Speir, Klaus Lorenz, and the two anonymous reviewers for polymerase chain reaction-amplified genes coding for 16S rRNA. Applied
and Environmental Microbiology 59, 695–700.
valuable comments on the manuscript. Poll, C., Thiede, A., Wermbter, N., Sessitsch, A., Kandeler, E., 2003. Micro-
scale distribution of microorganisms and microbial enzyme activities in a
References soil with long-term organic amendment. European Journal of Soil Science
54, 715–724.
Alexander, M., 1977. Introduction to Soil Microbiology. Wiley, New York. Renella, G., Mench, M., van der Lelie, D., Pietramellara, G., Ascher, J.,
Andersen, M.K., Refsgaard, A., Raulund-Rasmussen, K., Strobel, B.W., Ceccherini, M.T., Landi, L., Nannipieri, P., 2004. Hydrolase activity,
Hansen, H.C.B., 2002. Content, distribution, and solubility of cadmium in microbial biomass and community structure in long-term Cd-contaminated
arable and forest soils. Soil Science Society of America Journal 66, 1829– soils. Soil Biology & Biochemistry 36, 443–451.
1835. Renella, G., Mench, M., Landi, L., Nannipieri, P., 2005. Microbial diversity
DIN ISO 11466: 06.97, 1995. Bodenbeschaffenheit—Extraktion in Königs- and hydrolase synthesis in long-term Cd-contaminated soils. Soil Biology
wasser löslicher Spurenelemente. & Biochemistry 37, 133–139.
Fließbach, A., Martens, R., Reber, H.H., 1994. Soil microbial biomass and Sanadi, D.R., 1982. Mitochondrial coupling factor B. Properties and role in
microbial activity in soils treated with heavy metal contaminated sewage ATP synthesis. Biochimica et Biophysica Acta 683, 39–56.
sludge. Soil Biology & Biochemistry 26, 1201–1205. Sandaa, R.-A., Torsvik, V., Enger, Ø., Daae, F.L., Castberg, T., Hahn, D., 1999.
Fujie, K., Hu, H.-Y., Tanaka, H., Urano, K., Saitou, K., Katayama, A., 1998. Analysis of bacterial communities in heavy metal-contaminated soils at
Analysis of respiratory quinone profile in soil for characterization of different levels of resolution. FEMS Microbiology Ecology 30, 237–251.
microbiota. Soil Science and Plant Nutrition 44, 393–404. Schinner, F., Öhlinger, R., Kandeler, E., Margesin, R., 1996. Methods in Soil
Giller, K.E., Witter, E., McGrath, S.P., 1998. Toxicity of heavy metals to Biology. Springer, Berlin.
microorganisms and microbial processes in agricultural soils: a review. Soil Silver, S., Phung, L.T., 1996. Bacterial heavy metal resistance: new surprises.
Biology & Biochemistry 30, 1389–1414. Annual Review of Microbiology 50, 753–789.
Heuer, H., Krsek, M., Baker, P., Smalla, K., Wellington, E., 1997. Analysis of Speir, T.W., Kettles, H.A., Parshotam, A., Searle, P.L., Vlaar, L.N.C., 1999.
Actionomycete communities by specific amplification of genes encoding Simple kinetic approach to determine the toxicity of As(V) to soil
16S rRNA and gel-electrophoretic separation in denaturing gradients. biological properties. Soil Biology & Biochemistry 31, 705–713.
Applied and Environmental Microbiology 63, 3233–3241. Tabatabai, M.A., 1977. Effects of trace elements on urease activity in soils. Soil
Hiraishi, A., Morishima, Y., Takeuchi, J., 1991. Numerical analysis of Biology & Biochemistry 9, 9–13.
lipoquinone patterns in monitoring bacterial community dynamics in Tabatabai, M.A., Bremner, J.M., 1970. Arylsulfatase activity of soils. Soil
waste-water treatment systems. Journal of General and Applied Micro- Science Society of America Proceedings 34, 225–229.
biology 37, 57–70. Tamaki, S., Frankenberger Jr., W.T., 1992. Environmental biochemistry of
Hiraishi, A., Ueda, Y., Ishihara, J., 1998. Quinone profiling of bacterial arsenic. Reviews of Environmental Contamination and Toxicology 124,
communities in natural and synthetic activated sludge for enhanced 79–110.
phosphate removal. Applied and Environmental Microbiology 64, 992– ter Braak, C.J.F., Smilauer, P., 1998. CANOCO Reference Manual and User’s
998. Guide to Canoco for Windows: Software for Canonical Community
Hiroki, M., 1992. Effects of heavy metal contamination on soil microbial Ordination (version 4.0). Microcomputer Power, Ithaca, NY, USA.
population. Soil Science and Plant Nutrition 38, 141–147. Turpeinen, R., Kairesalo, T., Häggblom, M.M., 2004. Microbial community
Hoffmann, G., 1968. Eine photometrische Methode zur Bestimmung der structure and activity in arsenic-, chromium- and copper-contaminated
Phosphataseaktivität in Böden. Zeitschrift für Pflanzenernährung, Düngung soils. FEMS Microbiology Ecology 47, 39–50.
und Bodenkunde 118, 161–172. Wilke, B.-M., 1988. Langzeitwirkung potentieller anorganischer Schadstoffe
Juma, N.G., Tabatabai, M.A., 1977. Effects of trace elements on phosphatase auf die mikrobielle Aktivität einer sandigen Braunerde. Zeitschrift für
activity in soils. Soil Science of America Journal 41, 343–346. Pflanzenernährung und Bodenkunde 151, 131–136.