FEMS Microbiology Ecology 36 (2001) 1^9
www.fems-microbiology.org
E¡ects of mercury contamination on the culturable heterotrophic,
functional and genetic diversity of the bacterial community in soil
Lasse D. Rasmussen, SÖren J. SÖrensen *
Department of General Microbiology, University of Copenhagen, SÖlvgade 83H, DK-1307 Copenhagen K, Denmark
Received 13 October 2000; received in revised form 29 January 2001; accepted 4 February 2001
Abstract
This study investigates the effect of mercury contamination on the culturable heterotrophic, functional and genetic diversity of the
bacterial community in soil. The changes in diversity were monitored in soil microcosms, enriched with 25 Wg Hg(II) g31 soil, over a period
of 3 months. The culturable heterotrophic diversity was investigated by colony morphology and colony appearance on solid LB medium.
Functional diversity was analysed as sole carbon utilisation patterns in ECOplates. Genetic diversity was measured as bands on denaturing
gradient gel electrophoresis (DGGE) gels obtained by purification of total soil DNA and amplification of bacterial 16S rDNA fragments by
polymerase chain reaction. Concentrations of bioavailable and total mercury were measured throughout the experiment. The effect on the
culturable heterotrophic and genetic diversity was very similar, showing an immediate decrease after mercury addition but then slowly
increasing throughout the entire experimental period. Pre-exposure levels were not reached within the time span of this investigation. The
DGGE band pattern indicated that a shift in the community structure was responsible for recovered diversity. When analysed by Shannon^
Weaver indices, functional diversity was found to increase almost immediately after mercury addition and to remain at a level higher than
the control soil for the rest of the experiment. The fraction of culturable heterotrophic bacteria increased from 1% to 10% of the total
bacterial number as a result of mercury addition, and the mercury-resistant population increased to represent the entire heterotrophic
population. ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
Keywords : Soil microcosm; Microbial community ; ECOplate; Denaturing gradient gel electrophoresis ; Colony morphology
1. Introduction
The last decade has witnessed a still increasing interest
in the changes in the community structure and diversity of
microbial communities as a response to environmental
stress. Before the development of molecular techniques
for estimations of genetic diversity, investigations of this
kind were restricted to cultivation-based methods covering
only the aerobic heterotrophic fraction of the total bacterial population capable of forming colonies on solid media. The severity of these limitations was exposed when
investigations of genetic diversity, employing a DNA reassociation technique, estimated the presence of 4000 completely di¡erent bacterial genomes in 1 g of soil, representing as many as 13 000 di¡erent species [1,2].
The development of rapid and e¡ective methods for
* Corresponding author. Tel. : +45 35322053; Fax: +45 35322040;
E-mail : [email protected]
recovery of DNA directly from environmental samples
without prior cultivation and the use of genetic markers,
of which the 16S rRNA genes (rDNA) are the most commonly used [3], have resulted in the development of a vast
number of methods trying to circumvent the limitations of
cultivation-based investigations by `genetic ¢ngerprinting'
of the microbial community.
To investigate the structure and genetic diversity of
complex microbial communities denaturing gradient gel
electrophoresis (DGGE) is commonly used. This method
separates DNA fragments of the same length on the basis
of di¡erences in base composition, and was recently
adapted from detection of point mutations [4^6] to being
used on a mixture of 16S rRNA gene fragments ampli¢ed
by PCR from complex environmental DNA samples [7].
Each band in a DGGE gel is believed to represent a single
species/genus although heterogeneity of rDNA genes within a single species has been reported [8]. It still has to be
investigated how widespread this phenomenon is.
Because of the dominating role of the microbial community in decomposition of organic matter, recycling of
0168-6496 / 01 / $20.00 ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 8 - 6 4 9 6 ( 0 1 ) 0 0 1 1 1 - 8
FEMSEC 1231 17-5-01
2
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
N and C and bioremediation of polluted soil, functional
diversity may be a factor as important as taxonomic and
genetic diversity. The majority of studies of the microbial
functional diversity have been using BIOLOG microtitre
plates (Biolog Inc., Hayward, CA, USA) based on the
sole-source carbon utilisation pattern of 95 di¡erent carbon sources by the bacterial community [9,10].
Despite this increased interest and awareness of the different levels of bacterial community structure and diversity, to the best of our knowledge no studies have investigated changes of these di¡erent diversity types
simultaneously in mercury-contaminated soil. Investigations of this kind could help to increase the understanding
of the functioning and dynamics of the microbial community.
In the present study we investigated the functional and
genetic diversity and the diversity of the culturable heterotrophic fraction (in the following referred to as heterotrophic diversity) of the bacterial community in agricultural soil. Furthermore, we monitored the changes of all three
types of diversity as a response to environmental stress in
soil microcosms. Functional diversity was measured as
sole carbon utilisation using ECOplates (Biolog Inc.),
which contain 31 di¡erent carbon sources, allowing triplicate samples on a single 96-well microtitre plate. Genetic
diversity was analysed by total DNA extraction, PCR ampli¢cation of 16S rDNA fragments followed by separation
by DGGE [7]. Two approaches were taken to measure the
diversity of the heterotrophic population: (i) colony morphology, groupings performed on the basis of colony
shape, size, colour, etc. [11,12] and (ii) colony appearance,
where isolates are grouped according to the time of appearance on solid media [13].
Mercury at a concentration of 25 Wg Hg2‡ g31 soil was
used as a stressor in this experiment and monitored by a
mer-lux biosensor for the estimation of bioavailable mercury [14^16].
2.2. Soil microcosms
The soil used was an agricultural soil where no pesticides or fertiliser had been used for at least 20 years with
crop change every year (soil characteristics: total C %
1.296; total N % 0.189; C/N ratio 6.86; ammonium
0.077 Wg N g31 dry soil; nitrate 7.416 Wg N g31 dry
soil; water holding capacity 24%; pH 6.6).
The soil was collected during the summer of 1998 and
stored at 4³C for a week before use. Prior to setting up
microcosms the soil was sieved (mesh size 2 mm) and airdried at room temperature overnight. Water was added to
42% of the soil water holding capacity. Mercury to a ¢nal
concentration of 25 Wg Hg(II) g31 (as HgCl2† was added to
the soil with the water. Control soil was amended with
water. After addition of mercury and/or water the soils
were placed in ziplock bags and mixed thoroughly. The
soils were kept at room temperature for 24 h prior to ¢rst
sampling. Preliminary experiments (data not shown)
showed that this is necessary to obtain a uniform distribution of mercury and an equilibrium between mercury
and soil binding sites. For each treatment one microcosm,
consisting of 500 g soil, was established in a 500-ml glass
beaker and used for all subsequent samplings. Between
samplings microcosms were incubated at 24³C in a
water-saturated atmosphere (closed plastic box containing
two open water containers) to minimise water evaporation
during the experiment. Soil water content was measured at
every sampling point by incubating 1 g of soil in a microwave oven for 15 min followed by cooling for at least 1 h
in an exicator at room temperature before weighing. Soil
water content was constant during the experiment (data
not showed). Glass beakers, and all other glassware used
in the experiment, were acid-rinsed using 2 N HNO3 .
All experiments were done in triplicate from both mercury-spiked and control soil. Prior to every sampling soils
were mixed thoroughly in ziplock bags.
2.3. Quanti¢cation of mercury
2. Materials and methods
2.1. Bacterial strains, plasmids, growth and cell preparation
The strains used in this investigation were two mer-lux
derivatives of Escherichia coli HMS174, one containing
plasmid pRB28 [17] and the other with a constitutive mutant of pRB28, pRB27 [18]. The constitutive mutant was
used in all assays as a control that light emission was not
inhibited by assay conditions. Cultures were maintained in
LB medium using kanamycin (50 Wg ml31 ) for selection of
plasmids. Growth and preparation of cells for mer-lux
assays were as described by Barkay et al. [16]. The optical
density of cell suspensions in 67 mM phosphate bu¡er (pH
6.8) was adjusted to 0.5 at 660 nm (Ultrospec 2000 ; Pharmacia Biotech) which equals approximately 2U108 cells
ml31 .
Bioavailable mercury was quanti¢ed using a mer-lux
bioassay applied to soil samples [15]. Extraction of bioavailable mercury was performed by water leaching. One
gram of soil was mixed with 10 volumes (w/v) of sterile
ddH2 O in a 300-ml Erlenmeyer £ask and shaken for
15 min at 300 rpm at room temperature. Earlier experiments had shown that shaking times longer than 15 min
reduce the amount of bioavailable mercury dissolved in
the leachate [15]. Prior to assays large particles and indigenous microorganisms were removed from leachates by
centrifugation for 10 min at 12 000Ug at 4³C.
The assay medium was mixed from stock solutions [16]
immediately prior to assays (¢nal volume 180 Wl) in 6-ml
polystyrene tubes (Falcon, New Jersey, USA). The ¢nal
assay medium consisted of pyruvate (5 mM), Na,K-phosphate bu¡er (67 mM PO4 ; 34 mM Na; 33 mM K; pH 6.8)
FEMSEC 1231 17-5-01
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
and (NH4 )2 SO4 (0.091 mM). 1.72 ml of appropriate dilutions of soil leachate in water was added to a ¢nal volume
of 1.9 ml and assays were initiated by addition of 0.1 ml
biosensor cell suspension (¢nal concentration 107 cells
ml31 ). Light emission was monitored as relative light units
per 30 s using a BG-P Portable luminometer (MGM Instruments, Hamden, CT, USA), each sample was measured every 5^10 min over a period of 70^90 min. As a
control of biosensor performance all experiments included
assays performed in distilled water with a known mercury
concentration.
mer-lux expression factors (log quanta min31 ) were calculated from the slopes of light emission curves as described by Barkay et al. [16]. A regression between expression factors and mercury concentration was obtained from
assays performed in distilled water. This regression was
used to calculate bioavailable mercury in soil from expression factors found in assays performed with soil leachates.
Assays employed 107 cells of the biosensor per ml which
gave a linear response between Hg concentration and expression factors in the range of 0.3^1 nM [14]. Leachates
were diluted to give expression factors that fell within this
concentration range.
Total mercury in the soil microcosms was measured using a Jerome 431-x mercury vapor analyser (Arizona Instruments, Phoenix, AZ, USA) using soil method 2 as
described by Kriger and Turner [19].
2.4. Enumeration of colony-forming units (CFU)
One gram of soil was added to a test tube containing
9 ml sterile 1% NaCl in distilled water, this suspension was
vortexed at maximum velocity for 60 s. Appropriate decimal dilutions (0.1 ml) were plated on LB agar plates containing 25 Wg ml31 of the fungicide natamycin (Merck)
[20]. Mercury-resistant CFU were enumerated on plates
also containing 10 Wg Hg(II) as HgCl2 per ml. All plates
were incubated at 24³C for 4 days.
2.5. Diversity of colony-forming populations
Diversity analysis by colony morphology [11,12] was
performed by grouping colonies appearing on both nonselective and mercury-enriched LB agar plates (see above)
according to visual di¡erences, e.g., colour, shape, size,
etc. Plates containing approximately 100 colonies after
4 days incubation were used. For this analysis results
from the triplicates were pooled. Shannon^Weaver (S-W)
indices (HP = 34pi loge pi , where pi is the ratio between the
number in a speci¢c group and the total number [21]) were
calculated on the basis of these groupings.
Colonies were also grouped according to day of appearance [13]. Every day newly appeared colonies were
counted on plates from three consecutive dilutions (the
same three dilutions were used throughout the entire experiment for total CFU while dilutions varied for resistant
3
isolates according to number of colonies). Since selective
pressure of mercury-enriched plates diminishes with time
due to volatilisation of mercury, colonies were counted for
only 9 days. In order to make comparisons between diversity of resistant and total CFU all plates were counted for
9 days. S-W indices were calculated as mentioned above.
2.6. Functional diversity
The functional diversity of the microbial populations
was measured using ECOplates (Biolog Inc.). ECOplates
di¡er from Biolog GN plates in that only 31 di¡erent
carbon sources are used which makes it possible to have
triplicate samples in one 96-well microtitre plate. DAPI
staining was used to count the total number of bacteria
in the dilution series (see above) in order to inoculate each
well with 5U104 cells in 125-Wl samples. Plates were incubated at 24³C in plastic bags containing a water-soaked
paper towel in order to minimise evaporation from the
wells. The optical density in the wells caused by reduction
of tetrazolium dye was read using a Bio Kinetics reader
EL340 (Bio-Tek Instruments, Winooski, USA) after 48 h
incubation. The data were analysed by both principal
component analysis (PCA) [10], using the program SPSS
on a MacIntosh G3, and calculation of S-W indices (see
above). The S-W indices were calculated on the basis of
the ratio between the optical density in the single wells and
the total optical density summed from all the wells [9].
Only wells with an OD higher than the control well were
used.
2.7. Genetic diversity
Genetic diversity analysis of the total microcosm bacterial populations was performed at every sampling point by
total DNA extraction, PCR ampli¢cation of bacterial 16S
rDNA fragments followed by DGGE. DNA extractions
were performed as described by Porteous et al. [22]. The
only modi¢cation was that sonication time was reduced
from 3 min to only 10 s since tests of several di¡erent soils
showed that DNA was rapidly lost with longer sonication
times (unpublished data). Removal of humic acids was
performed by gel electrophoresis (0.7% low-melting SeaPlaque agarose) for 1 h at 125 V. Following electrophoresis gel blocks containing the DNA were cut from the gel
and stored at 320³C in Eppendorf tubes. Immediately
prior to PCR the gel blocks were melted at 68³C for
5 min, three volumes of dH2 O were added and samples
were incubated for 20 min at 68³C [22]. PCR reactions
(100 Wl per reaction) consisted of: `PCR Master' as described by the manufacturer (Roche Diagnostics, Mannheim, Germany), 10 Wl DNA template, 1 pM of each primer [7] and sterile PCR-grade H2 O (Boehringer) to a total
volume of 100 Wl. Preheated PCR Master (80³C) was
added to the tubes after a 4-min hotstart at 94³C. PCR
was performed using Perkin-Elmer GeneAmp PCR Sys-
FEMSEC 1231 17-5-01
4
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
tem 9600 for 35 cycles of: 94³C 1 min, 60³C 1 min, 72³C
1 min. The last cycle was followed by 8 min at 72³C.
DGGE was performed using D-GENE System (BioRad) DGGE equipment. 30 Wl of the PCR product was
loaded onto 7.5% (w/v) polyacrylamide gels made with
denaturing gradients ranging from 40 to 70% (where
100% contains 7 M urea and 40% formamide). Gels
were made using the Bio-Rad DGGE kit as described by
the manufacturer. The electrophoresis was run at 60³C for
16 h at 70 V. After electrophoresis gels were soaked in
SYBR Gold for 1 h (1:10 000 dilution, Molecular Probes,
Eugene, OR, USA) and digital images were obtained using
Gel Doc 1000 (Bio-Rad). Bands were detected, counted
and quanti¢ed using the Quantify One0 computer program on a MacIntosh G3.
3. Results
Fig. 2. Development of total and mercury-resistant CFU in mercuryspiked (25 Wg Hg(II) g31 ) (squares) and control (circles) microcosms.
Open symbols represent total CFU; ¢lled symbols represent mercuryresistant CFU. Data represent mean þ S.D. of triplicate samples.
3.1. Quanti¢cation of mercury
The concentration of bioavailable and total mercury
was followed in soil enriched with 25 Wg Hg(II) g31 soil
for a period of 3 months. The amount of bioavailable
mercury increased from day 1 to day 18 from about 0.18
to 0.25 Wg Hg(II) g31 soil (Fig. 1, in order to ¢t the scale
of the graph the bioavailable mercury results were multiplied by 50). This initial increase was followed by a rapid
decrease in the concentration of bioavailable mercury until
day 53 after which the decrease diminished and the
amount of bioavailable mercury levelled of at about
Fig. 1. Development of total and bioavailable mercury in both mercuryspiked (25 Wg Hg(II) g31 ) (squares) and control (circles) microcosms.
Bioavailable mercury (open symbols) was estimated by a mer-lux biosensor assay. In order for the bioavailable mercury data to ¢t the scale of
the y-axis the data were multiplied by a factor of 50. Total mercury
(¢lled symbols) was estimated after aqua regia digestion. Data represent
means þ S.D. of triplicate samples.
0.1 Wg Hg g31 of soil (Fig. 1). It may seem surprising
that the concentration of bioavailable mercury increases
the ¢rst days after mercury addition. This is probably
due to the fact that leaking organic material from sensitive
microorganisms is binding mercury thus decreasing bioavailability. Dissolved organic carbon has been found to
decrease bioavailability of mercury [18]. As the carbon is
utilised by resistant bacteria bioavailability increases. No
bioavailable mercury was ever found in the control soil.
The total amount of mercury in the soil found by acid
digestion was followed in the same period of time. The
¢rst 18 days after mercury addition no change in the
amount of total mercury was observed. From day 18 the
total mercury gradually decreased until day 53 after which
the mercury concentration stayed constant at about 20 Wg
Hg g31 soil (Fig. 1). This decrease in total mercury is in
good agreement with the period of fast decrease of bioavailable mercury. There will be an equilibrium between
total and bioavailable mercury that depends on whether
non-bioavailable mercury can convert to a bioavailable
form or if it is to tightly bound to, e.g., soil particles.
The decrease in the measured concentration of bioavailable mercury is a measure of this equilibrium rather than
of bioavailable mercury removal. The reduction in the
concentration of total mercury is due mainly to the fact
that the most predominant bacterial mercury resistance
mechanism is reduction of Hg(II) to volatile Hg0 mediated
by the product of the gene merA (mercuric reductase) [24].
This corresponds well with the ¢nding that after day 18
the heterotrophic bacterial population consists entirely of
mercury-resistant bacteria (see below).
In the control microcosm the mercury concentration
stayed at background levels throughout the entire experiment.
FEMSEC 1231 17-5-01
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
5
3.2. Enumeration of total and mercury-resistant CFU
The number of heterotrophic bacteria capable of forming colonies on solid medium were enumerated on both
non-selective and mercury-enriched (10 Wg ml31 ) LB agar
plates. In the control microcosm no changes in the number
of total (V107 CFU g31 soil) or resistant (V103 CFU g31
soil) CFU occurred during the entire experiment (Fig. 2).
In the mercury-enriched soil the number of CFU increased
rapidly reaching about 108 CFU g31 soil on day 18. The
number of CFU remained at this level for the rest of the
experiment. From day 18 and for the rest of the experiment 100% of the total CFU were mercury-resistant (Fig.
2).
The total number of bacteria as enumerated by microscopy after DAPI staining stayed at a constant level of 109
cells g31 soil in both mercury-contaminated and control
soil during the entire experiment.
3.3. Diversity analysis of CFU
The diversity of the culturable heterotrophic fraction of
the bacterial communities was examined by grouping the
isolates according to colony morphology and day of colony appearance.
Diversity analysis based on colony morphology showed
an initial decrease in the diversity of the total CFU in the
mercury-contaminated soil, but already from day 18 and
for the rest of the experiment the diversity gradually increased reaching the original level on day 81 (Fig. 3A).
The number of colony types varied between 18 and 24
but there was no pattern in the variations as a function
of sampling time. The mercury-resistant CFU experienced
an increased diversity in the middle of the experimental
period from day 18 to day 53 but eventually decreased to
about the same level as on day 1 (Fig. 3A). The number of
mercury-resistant colony types increased slightly during
the experiment from 9 on day 1 to 12^13 from day 32
and for the rest of the experiment. In the control soil
the number of colony types was approximately 30
throughout the entire experiment but a slight increase in
the diversity of total CFU was observed (Fig. 3A). A
possible explanation for this is that air-drying the soil
before setting up the microcosms may result in a decrease
in diversity due to, e.g., dormancy or spore formation.
This may be reversed by the addition of water at the
beginning of the experiment. Because of the low number
of mercury-resistant isolates in the control soil their diversity was not calculated.
As another method of measuring diversity of the CFU,
groupings were performed according to the day of appearance on the LB agar plates. These results showed much
greater correspondence between total and resistant CFU
in the mercury-enriched soil. The diversity of both fractions decreased until day 32 from a HP of about 1.5 to 1.
The diversity stayed at this level for the rest of the experi-
Fig. 3. The development of the heterotrophic, functional and genetic diversity was followed in both mercury-spiked (25 Wg Hg(II) g31 ) and
control soil microcosms. A: Changes in the culturable heterotrophic diversity, calculated on the basis of colony morphology, as a result of
mercury contamination in soil microcosm (squares) and in control microcosm (circles). Open symbols indicate total CFU; ¢lled symbols indicate mercury-resistant CFU. Diversity of mercury-resistant CFU is not
included for control soil because there were too few isolates. All colonies from triplicate samplings were pooled and used to calculate a single
S-W value. B: Development of culturable heterotrophic diversity, calculated on the basis of colony appearance, as a result of mercury contamination in soil microcosm (squares) and in control microcosm (circles).
Open symbols indicate total CFU; ¢lled symbols indicate mercury-resistant CFU. Diversity of mercury-resistant CFU is not included for control soil because there were too few isolates. The presented values are
means of triplicate samples. C: Functional diversity of the mercurystressed (squares) and control soil (circles) bacterial community calculated on the basis of sole carbon utilisation in ECOplates0 . S-W indices
are calculated as described in the text using the OD of the individual
wells and the summed OD. Only wells with an OD larger than the
background well were used for the analysis. Each data point represents
the mean of triplicate samples. D: Development of genetic diversity in
mercury-spiked (squares) and control soil (circles) microcosms. S-W indices were calculated on the basis of quanti¢cation of bands of bacterial
16S rDNA fragments in DGGE gels. Each data point represents the
mean of triplicate samples.
FEMSEC 1231 17-5-01
6
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
Fig. 4. The e¡ect of mercury contamination on the fraction of colonies
appearing on the solid growth medium within the ¢rst 2 days after sampling. The columns represent the percent of the total number of CFU
(colonies appearing during the entire 9 days of incubation) that appeared during the ¢rst 2 days of incubation. Results from both mercury-spiked (shaded columns) (25 Wg Hg(II) g31 ) and control (white
columns) microcosms are presented as mean þ S.D. of triplicate experiments.
ment (Fig. 3B). This decrease in diversity was due to the
fact that the appearance of the colonies shifted from being
almost evenly distributed over the 9 days of counting to
almost all appearing on day 1 or 2 after sampling (Fig. 4).
This result shows a shift in the dominating growth strategy
of the mercury-stressed culturable heterotrophic bacterial
community towards fast-growing individuals.
3.4. Functional diversity
The functional diversity of the bacterial communities
was investigated by examining the community potential
for sole carbon utilisation of the 31 di¡erent carbon sources found in ECOplates. The results obtained from the
ECOplates were used in two di¡erent ways. S-W indices
were calculated as described by Zak et al. [9]. Statistical
analysis was performed using PCA to detect di¡erences
between control and mercury-contaminated soil and/or
changes within the soils in the carbon utilisation during
the incubation [10].
In the mercury-contaminated soil the functional diversity calculated as S-W indices was low on day 1 (compared
with control soil) but increased rapidly reaching a level
above the functional diversity of the control soil on day
11 (Fig. 3C). The diversity in the mercury-contaminated
soil remained constant at this level for the rest of the
experiment. In the control soil no changes were observed
in the functional diversity throughout the entire experiment (Fig. 3C). The data of the functional diversity
from day 81 are excluded from the analysis because of
methodological problems.
Fig. 5. Ordination produced from PCA of sole carbon utilisation patterns of both mercury-spiked (25 Wg Hg(II) g31 ) (shaded symbols) and
control microcosms (white symbols). Each replicate of the triplicate
samples was used independently in this analysis.
The statistical analysis by PCA showed that the bacterial community in the mercury-enriched soil had a very
uniform utilisation pattern as soon as the community
was adapted to the presence of mercury. With the exception of day 1 all sampling points were very close together
(Fig. 5). No pattern was observed in the control soil. The
similarity in the utilisation patterns in the presence of
mercury indicates that adaptation results in the selection
of a few abundant species capable of utilising a broad
spectrum of carbon sources.
3.5. Genetic diversity
Genetic diversity was investigated by total DNA extraction, PCR ampli¢cation of 16S rDNA fragments followed
by separation by DGGE. S-W indices were calculated on
the basis of signal quanti¢cations of each band performed
by computerised image analysis.
Mercury contamination was found to decrease the genetic diversity. On day 1 the diversity of the bacterial
community was decreased in the mercury-enriched soil
compared to the control soil. This diversity decrease was
mainly due to a reduction in the number of bands detected
on the DGGE gel in the mercury-spiked soil (approximately 25 bands) compared to the control soil (approximately 50 bands). A gradual increase in the diversity in the
mercury-contaminated soil throughout the remaining part
of the experiment indicates that the recovery of lost genetic diversity is fast and begins immediately after contamination (Fig. 3D). Recovery of genetic diversity in the mercury-spiked soil was found to be primarily due to the
appearance of new 16S rDNA bands not dominating in
the control soil. Even after 81 days of exposure the band
patterns varied greatly between mercury-exposed and
control soil with many bands dominating in the contaminated soil not detected in the control soil (Fig. 6). In
FEMSEC 1231 17-5-01
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
Fig. 6. PCR DGGE of 16S rDNA fragments from both mercury-spiked
(25 Wg Hg(II) g31 (lanes 1^3) and control soil (lanes 4^6) after 81 days
mercury exposure.
the control soil the genetic diversity was found to be more
or less constant over the duration of the experiment (Fig.
3D).
4. Discussion
E¡ects of mercury contamination on the bacterial community in agricultural soil can be observed in all three
types of diversity investigated in this study. The culturable
heterotrophic fraction of the community increased 10-fold
to represent about 10% of the total population found by
direct counting under the microscope. Leaking of nutrients
by cell lysis from mercury-sensitive microorganisms (bacteria, fungi, etc.) is presumably the basis for this growth of
primarily mercury-resistant heterotrophs. The number of
mercury-resistant CFU increased rapidly by ¢ve orders of
magnitude and after just 18 days represented the entire
culturable heterotrophic population (Fig. 3A). An increase
in the resistant fraction is a commonly reported e¡ect of
heavy metal pollution in both aquatic [23^25] and terres-
7
trial environments [26^28]. How large a percentage the
resistant fraction represents depends on the soil type [27]
and the mercury concentration.
The diversity analysis of the culturable heterotrophic
population showed that not only did the diversity of the
total population decrease due to mercury stress (Fig.
3A,B) but also the population dynamics changed towards
very active fast-growing individuals (r-strategists) (Fig. 4).
The divergence between the diversity of the total and
the mercury-resistant CFU was surprisingly large considering that 100% of the total CFU was found to be resistant. There are di¡erent possible explanations for this discrepancy. It could be due to the presence of sensitive subpopulations, which may survive protected inside soil aggregates [29], too small in numbers to be detectable in the
total number but large enough to have an in£uence on the
diversity in these types of calculations. But the large number of isolates used for these calculations (300^400 colonies) makes this explanation unlikely since a few sensitive
colonies will not markedly change the S-W index. Thus,
another more likely possibility is that the mercury in the
growth medium in some way alters the diversity of the
colony morphology, e.g., by inhibition of pigment formation making colonies of di¡erent species indistinguishable.
It has been found that the growth medium can have an
e¡ect on colony morphology [11,12]. The diversity of the
mercury-resistant culturable heterotrophic population was
increased in the middle of the experimental period. This
could be the result of inter-species horizontal gene transfer
of self-transmissible conjugative plasmids, harbouring
mercury resistance genes, from resistant to sensitive bacteria, thus increasing the diversity of the resistant population. Mercury resistance genes are found often to be located on plasmids [30], which have been reported to have
elevated transfer frequencies at high mercury concentrations [26].
The change in community structure from predominantly
K-strategists in the control soil towards r-strategists in the
mercury-contaminated soil corresponds well with other
¢ndings where r-strategists (de¢ned as organisms that are
good at rapid growth under uncrowded, nutrient-rich conditions) are characteristic for unstable environments, while
K-strategists (organisms that are good at exploiting resources under crowded conditions) do better in stable environments [13]. The increased functional diversity observed in the mercury-stressed population (Fig. 3C)
suggests that the r-strategists dominating in this agricultural soil are capable of utilising a broad spectrum of
carbon sources. Changes in the community utilisation of
sole carbon sources due to environmental stress have been
shown in pesticide-enriched soil [31] but to the best of our
knowledge this study is the ¢rst to show increased functional diversity as a result of mercury contamination. The
increased culturable heterotrophic fraction may also have
an in£uence on the functional diversity measured by this
type of plate since it has been shown that even though the
FEMSEC 1231 17-5-01
8
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
rate of colour development is correlated with inoculum
density [10,32], there is a selection of heterotrophs in the
wells during incubation [33] indicating that mainly heterotrophs are responsible for the functional diversity observed
by this procedure. In the mercury-contaminated soil the
amount of available carbon is increased due to cell lysis of
sensitive microorganisms which may be another factor
partly responsible for the enhanced functional diversity.
An increase in the available environmental carbon has
been reported to result in a utilisation enhancement in
agricultural soil [34]. In this experiment the ECOplates
were incubated for a relatively short period (48 h) in order
to get a more `state of the art' picture of what was being
utilised by the population at the time of sampling rather
than the complete utilisation capabilities of the bacterial
community. This may underestimate the functional diversity in communities dominated by K-strategists when compared with r-strategist-dominated communities as seems to
be the case in this study. But not even an increase of the
incubation time to 7 days decreased the di¡erences in
functional diversity between the mercury-stressed and the
control soil bacterial communities (data not shown).
PCA of the sole carbon patterns is a valuable tool for
comparing microbial communities from di¡erent habitats
on the basis of di¡erences in the pattern of sole carbon
source utilisation or for monitoring changes within a microbial community due to exposure to environmental
stress [9,10,32,34,35]. In the present study PCA showed
that exposure to mercury stress resulted in very uniform
utilisation patterns (Fig. 5) indicating low species diversity.
The e¡ect of mercury contamination on genetic diversity
was seen immediately. The largest decrease was observed
1 day after mercury addition (Fig. 3D). Interestingly
enough a slow but instant increase in genetic diversity
was observed that continued during the entire period of
the study in a linear manner. Full recovery of the genetic
diversity was not reached within the time span of this
experiment. Even though many sophisticated molecular
methods have been developed for detection of genetic diversity in recent years, only very few studies have investigated the e¡ects of pollution on genetic diversity. In general these studies observed a decrease in the genetic
diversity as a result of pollution with, e.g., pesticides [31]
and copper [36]. But common to these studies was that the
development of genetic diversity was not monitored after
the initial decrease. The present study showed that even
though a recovery of genetic diversity was found this was
not due to a reversion towards the pre-exposure community but mainly due to the appearance of new dominating
species (Fig. 6). Hence mercury stress at this concentration
may result in permanent changes in the composition of the
soil bacterial community. To con¢rm that this is the case it
would be necessary to monitor genetic diversity at least
until the selective pressure of the mercury contamination
has disappeared.
In this study we monitored the e¡ect of mercury con-
tamination on the culturable heterotrophic, functional and
genetic diversity of the bacterial community. We followed
the development of the diversity of all three types by frequent samplings over a period of 3 months after contamination. When comparing the development in the diversity
of the di¡erent types, the resemblance between the diversity measured by colony morphology (Fig. 3A) and the
genetic diversity (Fig. 3D) is striking. Both begin increasing gradually immediately after an initial decrease. This
resemblance is not surprising if one takes into account
that both methods only measure the diversity of the individuals dominating in numbers. Since the culturable heterotrophic population accounts for 10% of the total bacterial community this fraction probably makes up a large
part of the total genetic diversity measured.
The results of this study indicate that the culturable
heterotrophic population probably plays a very important
role in the development of all three types of diversity during environmental stress. Furthermore, the analysis of genetic diversity indicates that adaptation to mercury stress
may result in a recovery of diversity due to a shift in the
community structure.
Acknowledgements
This study was supported by the `Centre for biological
processes in contaminated soil and sediment' (www.
biopro.dk) under The Danish Environmental Research
Programme. The authors wish to thank Pia Kringelum
for excellent technical assistance.
References
[1] Torsvik, V., GoksÖyr, J. and Daae, F.L. (1990) High diversity in
DNA of soil bacteria. Appl. Environ. Microbiol. 56, 782^787.
[2] Torsvik, V., Salte, K., SÖrheim, R. and GoksÖyr, J. (1990) Comparison of phenotypic diversity and DNA heterogeneity in a population
of soil bacteria. Appl. Environ. Microbiol. 56, 776^781.
[3] Liesack, W., Janssen, P.H., Rainey, F.A., Ward-Rainey, N.L. and
Stackebrandt, E. (1997) Microbial diversity in soil: The need for a
combined approach using molecular techniques in: Modern Soil Microbiology (van Elsas, J.D., Trevors, J.T. and Wellington, E.M.H.,
Eds.), pp. 375^440. Marcel Dekker, New York.
[4] Myers, R.M., Fischer, S.G., Lerman, L.S. and Maniatis, T. (1985)
Nearly all single base substitutions in DNA fragments joined to a
GC-clamp can be detected by denaturing gradient gel electrophoresis.
Nucleic Acids Res. 13, 3131^3145.
[5] Myers, R.M., Maniatis, T. and Lerman, L.S. (1987) Detection and
localization of single base changes by denaturing gradient gel electrophoresis. Methods Enzymol. 155, 501^527.
[6] Cariello, N.F. and Skopek, T.R. (1993) Mutational analysis using
denaturing gradient gel electrophoresis and PCR. Mutat. Res. 288,
103^112.
[7] Muyzer, G., de Waal, E.C. and Uitterlinden, A.G. (1993) Pro¢ling of
complex microbial populations by denaturing gradient gel electrophoresis of polymerase chain reaction-ampli¢ed genes coding for
16S rRNA. Appl. Environ. Microbiol. 59, 695^700.
[8] Nu«bel, U., Engelen, B., Felske, A., Snaidr, J., Wieshuber, A.,
FEMSEC 1231 17-5-01
L.D. Rasmussen, S.J. SÖrensen / FEMS Microbiology Ecology 36 (2001) 1^9
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
Amann, R.I., Ludwig, W. and Backhaus, H. (1996) Sequence heterogeneities of genes encoding 16S rRNA in Paenibacillus polymyxa
detected by temperature gradient gel electrophoresis. J. Bacteriol.
178, 5636^5643.
Zak, J.C., Willig, M.R., Moorhead, D.L. and Wildman, H.G. (1994)
Functional diversity of microbial communities: A quantitative approach. Soil Biol. Biochem. 26, 1101^1108.
Garland, J.L. and Mills, A.L. (1991) Classi¢cation and characterization of heterotrophic microbial communities on the basis of patterns
of community-level sole-carbon-source utilization. Appl. Environ.
Microbiol. 57, 2351^2359.
Palumbo, A.V., Zhang, C., Liu, S., Scarborough, S.P., P¢¡ner, S.M.
and Phelps, T.J. (1996) In£uence of media on measurement of bacterial populations in the subsurface. Appl. Biochem. Biotechnol. 57/
58, 905^914.
Balkwill, D.L. and Chiorse, W.C. (1985) Characterization of subsurface bacteria associated with two shallow aquifers in Oklahoma.
Appl. Environ. Microbiol. 50, 580^588.
De Leij, F.A.A.M., Whipps, J.M. and Lynch, J.M. (1993) The use of
colony development for the characterization of bacterial communities
in soil and roots. Microb. Ecol. 27, 81^97.
Rasmussen, L.D., Turner, R.R. and Barkay, T. (1997) Cell-densitydependent sensitivity of a mer-lux bioassay. Appl. Environ. Microbiol. 63, 3291^3293.
Rasmussen, L.D., SÖrensen, S.J., Turner, R.R. and Barkay, T. (2000)
Application of a mer-lux biosensor for estimating bioavailable mercury in soil. Soil Biol. Biochem. 32, 639^646.
Barkay, T., Turner, R.R., Rasmussen, L.D., Kelly, C.A. and Rudd,
J.W.M. (1998) Luminiescence Facilitated Detection of Bioavailable
Mercury in Natural Waters. In: Methods in Molecular Biology, Vol.
102: Bioluminescence Methods and Protocols (R.A. LaRossa ed.) pp.
231^246 Humana Press Inc., Totowa, NJ.
Selifonova, O., Burlage, R. and Barkay, T. (1993) Bioluminescent
sensors for detection of bioavailable mercury(II) in the environment.
Appl. Environ. Microbiol. 59, 3083^3090.
Barkay, T., Gillman, M. and Turner, R.R. (1997) E¡ects of dissolved
organic carbon and salinity on bioavailability of mercury. Appl. Environ. Microbiol. 63, 4267^4271.
Kriger, A.A. and Turner, R.R. (1995) Field analysis of mercury in
water, sediment and soil using static headspace analysis. Water Air
Soil Poll. 80, 1295^1304.
Pedersen, J.C. (1992) Natamycin as a fungicide in agar media. Appl.
Environ. Microbiol. 58, 1064^1066.
Shannon, C.E. and Weaver, W. (1949) The Mathematical Theory of
Communication. University of Illinois Press, Urbana, IL.
Porteous, L.A., Armstrong, J.L., Seidler, R.J. and Watrud, L.S.
(1994) An e¡ective method to extract DNA from environmental samples for polymerase chain reaction ampli¢cation and DNA ¢ngerprint analysis. Curr. Microbiol. 29, 301^307.
9
[23] Barkay, T. (1987) Adaptation of aquatic microbial communities to
Hg2‡ stress. Appl. Environ. Microbiol. 53, 2725^2732.
[24] Barkay, T. and Olson, B.H. (1986) Phenotypic and genotypic adaptation of aerobic heterotrophic sediment bacterial communities to
mercury stress. Appl. Environ. Microbiol. 52, 403^406.
[25] Rasmussen, L.D. and SÖrensen, S.J. (1998) The e¡ect of longterm
exposure to mercury on the bacterial community in marine sediment.
Curr. Microbiol. 36, 291^297.
[26] Smit, E., Wolters, A. and van Elsas, J.D. (1998) Self-transmissible
mercury resistance plasmids with gene-mobilizing capacity in soil bacterial populations: In£uence of Wheat roots and mercury addition.
Appl. Environ. Microbiol. 64, 1210^1219.
[27] Ranjard, L., Richaume, A., Jocteur-Monrozier, L. and Nazaret, S.
(1997) Response of soil bacteria to Hg(II) in relation to soil characteristics and cell location. FEMS Microbiol. Ecol. 24, 321^331.
[28] Roane, T.M. and Kellogg, S.T. (1996) Characterization of bacterial
communities in heavy metal contaminated soils. Can. J. Microbiol.
42, 593^603.
[29] Ranjard, L., Nazaret, S., Gourbie©re, F., Thioulouse, J., Linet, P. and
Richaume, A. (2000) A soil microscale study to reveal the heterogeneity of Hg(II) impact on indigenous bacteria by quanti¢cation of
adapted phenotypes and analysis of community DNA ¢ngerprints.
FEMS Microbiol. Ecol. 31, 107^115.
[30] Summers, A.O. and Barkay, T. (1989) Metal resistance genes in the
environment. In: Gene Transfer in the Environment (Levy, S.B. and
Miller, R.V., Eds.), pp. 287^308. McGraw-Hill, New York.
[31] Engelen, B., Meinken, K., von Wintzingerode, F., Heuer, H., Malkomes, H.-P. and Backhaus, H. (1998) Monitoring impact of pesticide treatment on bacterial soil communities by metabolic and genetic
¢ngerprinting in addition to conventional testing procedures. Appl.
Environ. Microbiol. 64, 2814^2821.
[32] Haack, S.K., Garshow, H., Klug, M.J. and Forney, L.J. (1995) Analysis of factors a¡ecting the accuracy, reproducibility and interpretation of microbial community carbon source utilization patterns.
Appl. Environ. Microbiol. 61, 1458^1468.
[33] Smalla, K., Wachtendorf, U., Heuer, H., Liu, W.-T. and Forney, L.
(1998) Analysis of BIOLOG GN substrate utilisation patterns by
microbial communities. Appl. Environ. Microbiol. 64, 1220^1225.
[34] Bossio, D.A. and Scow, K.M. (1995) Impact of carbon and £ooding
on the metabolic diversity of microbial communities in soils. Appl.
Environ. Microbiol. 61, 4043^4050.
[35] Zabinski, C.A. and Gannon, J.E. (1997) E¡ects of recreational impacts on soil microbial communities. Environ. Manage. 21, 233^238.
[36] Smit, E., Lee£ang, P. and Wernars, K. (1997) Detection of shifts in
microbial community structure and diversity in soil caused by copper
contamination using ampli¢ed ribosomal DNA restriction analysis.
FEMS Microbiol. Ecol. 23, 249^261.
FEMSEC 1231 17-5-01