Journal of Microbiological Methods 58 (2004) 335 – 349

www.elsevier.com/locate/jmicmeth

Identification and quantification of arsC genes in environmental

samples by using real-time PCR
Yongmei Sun, Elena A. Polishchuk, Una Radoja, William R. Cullen *
Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, BC, Canada V6T 1Z1

Received 16 April 2004; received in revised form 24 April 2004; accepted 26 April 2004
Available online 26 June 2004

Abstract

The arsC gene is responsible for the first step in arsenate biotransformation encoding the enzyme arsenate reductase. The
quantitative real-time PCR method was developed to quantify the abundance of the arsC genes in environmental samples
contaminated with arsenic. Two sets of primers that showed high specificity for the target arsC gene were designed based on
consensus sequences from 13 bacterial species. The arsC gene was used as an external standard instead of total DNA in the
calibration curve for real-time PCR, which was linear over six orders of magnitude and the detection limit was estimated to be
about three copies of the gene. Soil samples from arsenic contaminated sites were screened for arsC genes by using PCR and
showed the presence of this gene. The copy numbers of the gene ranging from 0.88  104 to 1.56  105 per ng total DNA were
found in eight arsenic contaminated samples. Soil samples from a bioreactor containing pulp mill biomass and high
concentration of arsenate showed a tenfold higher count of arsC gene copies than soil samples collected underground from an
arsenic-rich gold mine.
D 2004 Elsevier B.V. All rights reserved.

Keywords: arsC gene; Real-time PCR; SYBRR Green I; Group-specific primers; Arsenate reduction

1. Introduction The arsenic-resistance operon in bacteria is formed

Knowledge of the genetic systems involved in
arsenic mobilization, transformation and resistance
can significantly contribute to practical strategies for
the bioremediation of arsenic contamination. For
example, the arsC and gamma-ECS genes have been
doubly transformed to engineer hybrids of Arabidop-
sis thaliana plants (Dhankher et al., 2003) that accu-
mulate arsenic from soil.
* Corresponding author. Tel.: +1-604-822-4435; fax: +1-604-
822-2847.
E-mail address: wrc@chem.ubc.ca (W.R. Cullen).
by a cluster comprising three to six genes, which are
located in chromosomes or in plasmids (Mukhopad-
hyay et al., 2002; Rosen, 2002a). The following genes
are in the operon: the arsR gene controls operon
expression by encoding a repressor protein and the
arsD gene encodes a protein controlling the upper
level of operon expression. The arsA and arsB genes
encode, respectively, an ATPase and a protein, which
forms a transmembrane efflux channel, resulting in an
anion transporting arsenite pump. The function of the
arsH gene required for As resistance of group H
plasmids, is as yet unknown (Butcher et al., 2000;
Ryan and Colleran, 2002; Shi et al., 1999).

0167-7012/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.mimet.2004.04.015
336 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349
The arsC gene, a part of the ars operon that
encodes the enzyme arsenate reductase, is the focus
of this study. Arsenate reductase carries out the first
step in arsenate metabolism in living organisms before
its efflux from the cell, reducing arsenate (AsV) to
arsenite (AsIII) (Silver et al., 2002).
Preliminary strategies in bioremediation studies
include the identification of microorganisms in envi-
ronmental samples enriched with species that are
known to be degrading targeted contaminants. Identi-
fication of bacterial and fungal strains might be
conducted by using 16S rDNA and 18S or 26S rDNA
methods respectively to detect microorganisms asso-
ciated with specific contaminant degradation. This
method has been applied to the detection of ammo-
nia-oxidizing bacteria in soil (Hermansson and Lindg-
ren, 2001).
A more fruitful approach would be to directly
identify and quantify genes responsible for the deg-
radation reaction in an environmental sample. Beller
et al. (2002) reported a method for monitoring envi-
ronmental samples by using real-time quantitative
polymerase chain reaction (PCR) to determine the
quantity of the bssA gene associated with the first
step of anaerobic toluene and xylene degradation.
Braker et al. (2000) used nitrite reductase genes nirK
and nirS as functional markers to investigate the
diversity of denitrifying bacteria in marine sediments
by using real-time PCR.
Real-time PCR is used in clinical research for the
development of new screening procedures for cancer,
such as the rapid assessment of anticancer target
expression in patients at risk of developing malignan-
cies (Humar et al., 2001); in the diagnosis of a
multitude of bacterial infections (O’Mahony and Hill,
2002); and in the quantitation of viral DNA in serum
(Chen et al., 2001). Its high sensitivity is one of the
reasons why this method is preferred for the detection
and quantification of introduced DNA elements in
genetically modified food (Berdal and Holst-Jensen,
2001; Permingeat et al., 2002).
Real-time PCR allows amplification of low abun-
dance genes to a sufficient concentration that they
can be quantified. One of the fluorescent reporter
systems that have been developed for real-time
quantitative PCR, is SYBRR Green I (SG I) (Filion
et al., 2003; Stubner, 2002). It is a fluorescent dye
that binds to double-stranded DNA. After primer
attachment to the DNA template during PCR exten-
sion cycles, increasing amounts of double-stranded
DNA are produced and the intensity of the fluores-
cent signal increases. The basis of the analysis is the
quantitative relationship between the amount of start-
ing target sequence and the amount of PCR product
at any given cycle. Quantification of the starting
amount of target gene copies in unknown samples is
accomplished by measuring the cycle number (Ct)
and by using a standard curve. The cycle number,
which corresponds with fluorescence readings above
the threshold is proportional to the starting amount
of template DNA.
Low specificity is one of the drawbacks of using
the SG I in real-time PCR because of the formation of
non-specific products, such as primer-dimers, which
contribute to the fluorescent signal and produce Ct = 1
readings that appear much earlier in the amplification
process than they otherwise should. Therefore special
care should be taken for optimization of PCR con-
ditions to prevent the formation of primer-dimers,
taking it into account a possible reduction of PCR
detection sensitivity.
The goal of this study was to develop a real-time
PCR method for quantification of the arsC gene so that
it can be applied to evaluate the mobilization potential
of arsenic contaminated environmental samples.
The nucleotide sequences of the arsC genes from
120 bacteria were aligned in order to find conservative
regions. A group of 13 bacteria that had the most
similar oligonucleotide composition of the arsC gene
was selected. Corresponding primers were designed
and then used in the real-time PCR method to quantify
the abundance of the arsC genes in soil contaminated
with arsenic.
2. Material and methods
2.1. Strains and plasmids
Bacterial strains and plasmids used in this study are
described in Table 1.
2.2. Environmental samples
Soil and biomass samples used in this work are
listed in Tables 2 and 7. All of them, except the
 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349
Table 1
Bacterial strains and plasmids used in this study
Strain arsC genotype
GenBank accession #
Klebsiella oxytoca plasmid pMH12 AF168737
Escherichia coli K12 W3110 NC_000913
Escherichia coli plasmid R64 AP005147
Serratia marcescens plasmid R478 AJ288983
control sample, were collected from arsenic-rich envi-
ronments. Seven biomass samples were collected
underground from the Giant gold mine (Canada,
Yellowknife, Northwest Territory). Three soil samples
were received from an arsenic-containing pilot scale
bioreactor in Trail (British Columbia, Canada) and
seven samples were collected from a laboratory scale
bioreactor. The bioreactors contained soil enriched
with biomass from a pulp mill and they processed
water with high arsenate concentration. A control soil
sample, not contaminated with arsenic, was obtained
from the University of British Columbia campus
(Vancouver, B.C., Canada).
2.3. DNA Preparation
Extraction of genomic DNA was performed by
using the QIAampR DNA Mini Kit (Qiagen, Mis-
sissauga, ON, Canada). QIAprepR Spin Miniprep kit
(Qiagen) was used for plasmid DNA preparation.
Both were used in accordance with the manufacturer’s
instructions. A protocol from Zhou et al. (1996) was
used for DNA extraction from soil samples. The
purification procedure utilized Sepharose 4B (Sigma,
St. Louis, MO, USA) for removal of humic acids as
described by Jackson et al. (1997).
2.4. Primers
Primer design was based on the multiple sequence
alignment of the arsC genes in a variety of arsenic-
resistant bacteria. When this study was initiated,
sequences of the arsC gene from 120 organisms were
available from the GenBank database (National Cen-
ter for Bioinformatics, website http://www.ncbi.nih.
gov (May 2003)). The multiple sequences of the arsC
genes were aligned by using ClustalW software
(website http://www.ebi.ac.uk/clustalw/) in order to
find the longest homologous part of arsC.
337
Reference Source
Unpublished M. Kelly, University of British Columbia
(Blattner et al., 1997) B. Rosen, Wayne State University
(Komano et al., 1987) T. Komano, Tokyo Metropolitan University
(Ryan and Colleran, 2002) D. Taylor, University of Alberta
A group of 13 bacteria with the best aligned
homologous regions of the arsC gene was chosen
from 120 arsC gene sequences. To amplify these
selected genes, two primer pairs (listed in Table 3)
were designed by using the Saccharomyces Genome
Database (SGDTM, Stanford University). This pro-
gram was chosen from a number of others because it
permits options for amplifying a region with exact
endpoints of the DNA sequence entered. The ‘‘Exact
Endpoints’’ option was critical for this study because
it allowed precise control of the targeted conservative
sequence regions for which primers were to be
selected.
These primers should amplify a fragment of 353
base pairs of the arsC sequence, which includes the
conserved regions on both ends.
2.5. PCR and real-time PCR conditions
Both general and real-time PCR experiments were
conducted in an iCycler (Bio-Rad Laboratories).
General PCR reactions were performed in 50
Al reaction volumes that contained 20 mM Tris –
HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM
each dNTPs (dATP, dCTP, dGTP, dTTP), 0.05 unit/
Al Taq DNA polymerase, a mixture of 250 nM each
of reverse and forward primers pair A (amlt-42-f/
amlt-376-r) and pair B (smrc-42-f/smrc-376-r) and
various concentrations of template DNA. dNTPs and
Taq DNA polymerase were purchased from Invitro-
gen (Burlington, ON, Canada), primers were synthe-
sized by Nucleic Acid and Protein Services (UBC,
Vancouver, Canada). The mixture of two primer
pairs was evaluated for absence of primer-dimer
and hairpin PCR products by conducting the PCR
with distilled deionized water instead of the DNA
template. Real-time PCR was performed using iQk
SYBRR Green I Supermix (Bio-Rad Laboratories,
Hercules, CA, USA) in accordance with the manu-
338 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349

Table 2
Identification and quantification of the arsC gene in environmental samples
Name Description DNA General PCR Presence of Presence of Quantification of the
400 bp signal the arsC gene the arsC gene arsC gene from original
(after incubation)a (after incubation) from original samples by real-time PCR
samples by (copies/ng total DNA)c
general PCRb
YK-1d A black/orange Genomic + YES NO ND
biomass on rock Plasmid
face, Yellowknife,
Giant gold mine,
250 ft level
YK-5d A pink biomass Genomic + YES NO ND
from the rock face, Plasmid
Yellowknife,
Giant gold mine,
425 ft level
YK-6ad Biomass from Genomic + YES NO ND
the rock face, Plasmid
Yellowknife,
Giant mine,
425 ft level
YK-6bd Darker layer of Genomic YES NO ND
biomass slime, Plasmid +
extensively
growing under
the lighter
colored surface,
Yellowknife,
Giant mine,
425 ft level
#406 Biomass from Genomic YES NO 0.88  104 F 0.44  104
vent door in Plasmid +
1100B shaft,
Yellowknife,
Giant mine,
425 ft level
#407 Beige-green Genomic + YES NO ND
biomass, Plasmid +
Yellowknife,
Giant mine,
100 ft level
#408 Brown/ Genomic + YES NO 2.38  104 F 0.17  104
Orange Soil, Plasmid
Yellowknife,
Giant mine,
250 ft level
Soil CC2 Trail, BC Genomic YES NO ND
Plasmid +
Soil CC3 Trail, BC Genomic + YES NO ND
Plasmid
Soil creek Trail, BC Genomic + YES NO 1.17  104 F 0.21  104
Plasmid
 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349 339

Table 2 (continued )
Name Description DNA General PCR Presence of Presence of Quantification of the
400 bp signal the arsC gene the arsC gene arsC gene from original
(after incubation)a (after incubation) from original samples by real-time PCR
samples by (copies/ng total DNA)c
general PCRb
Control#1e Autoclaved Genomic + YES YES 55.5  104 F 3.5  104
soil samples
were spiked
with E. coli
Control#2f E. coli culture Genomic + YES
a
The samples were placed on nutrient agar (enriched with 10 ppm As (V) and 10 Ag/ml cycloheximide) for overnight growth at room
temperature. Then all growing colonies were transferred to nutrient broth with 10 ppm As(V) and incubated at room temperature for 3 days. The
Qiagen Genomic and Plasmid DNA Extraction kit was used for DNA preparation. The arsC gene was amplified by using mixture of two primer
sets amlt-42-f/amlt-376-r and smrt-42-f/smrt-376-r.
b
DNA extracted directly from soil samples was used as template in general PCR. The arsC gene was amplified by using mixture of two
primer sets as described in a.
c
Results were from duplicate samples, n = 2. Error was given as standard deviation. ND is below the detection limit.
d
Arsenic reduction by samples YK-1, YK-5, YK-6a and YK-6b was found in other experiments (unpublished). The samples were incubated
in potato dextrose broth (PDB) with 100 ppm arsenate at room temperature for 38 days. GC-Hydride-Generation-AAS was used to detect
arsenite and the total arsenic concentration in the medium. The results showed that the percentage of arsenite out of total arsenic of YK-1, YK-5,
YK-6a and YK-6b was 0.1%, trace, 0.1% and 0.1%, respectively.
e
A soil sample from the UBC campus was used to prepare a control containing only one bacterial strain known to carry the arsC gene. The
sample was autoclaved (121 jC, 17 psi, 20 min) then spiked with a culture of E. coli W3110 (12 h growth, 1 ml of culture per 10 g of soil) and
incubated overnight at room temperature on a petri plate with nutrient agar containing the same concentration of As (V) and cycloheximide as
described in b.
f
Another petri plate with pure E. coli culture without soil (Control #2) was incubated in the same medium and under the same conditions.
After incubation genomic DNA from both control samples was extracted by using Qiagen Genomic DNA kit.

facturer’s instructions. The reactions were set up
with 1  SYBRR Green I Supermix, the same
mixture of primer pairs as in general PCR, and
various amount of template DNA. Premix SYBRR
Green (2  ) contained 100 mM KCl, 40 mM Tris –
HCl pH 8.4, 0.4 mM dNTPs, 6 mM MgCl2, 20 mM
SYBRR Green I fluorescein, 50 Units/ml iTaqDNA
polymerase.
Real-time PCR conditions were as follows (the
same as for general PCR): 95 jC for 3 min, 40 cycles
of 95 jC for 15 s (denaturation), 60 jC for 15 s
(annealing), and 72 jC for 15 s (elongation). Deion-
ized sterile water and DNA from Klebsiella pneumo-
niae were used for negative control instead of
template DNA.
2.6. Real-time PCR and a calibration curve for the
arsC gene
The arsC target gene of Escherichia coli W3110
was extracted from an agarose gel band and then
used as external standard for quantification. The fol-
lowing tenfold dilution series were prepared: 6.8 
10 1, 6.8  10 2, 6.8  10 3, 6.8  10 4, 6.8 

Table 3
Characteristics of primers
Primer namea Primer sequence (5V – 3V) Length, bp Tm,jC % GC
amlt-42-f TCGCGTAATACGCTGGAGAT 20 54 50
amlt-376-r ACTTTCTCGCCGTCTTCCTT 20 54 50
smrc-42-f TCACGCAATACCCTTGAAATGATC 24 59 41
smrc-376-r ACCTTTTCACCGTCCTCTTTCGT 23 59 47
a
Primer names include an abbreviated description of the specific gene cluster, followed by a number indicating the nucleotide position and f
(forward) or r (reverse) (Alm et al., 1996).
340 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349

10 5, 6.8  10 6 ng of external standard in 12 (six
in duplicates) real-time PCR assays performed as
described above. Sterile deionized water (2 Al) was
used as a negative control instead of template
DNA. The real-time PCR profiles from a series of
arsC target DNA template concentrations are plot-
ted in Fig. 1. The base line, threshold and Ct values
were auto calculated by iCycler software (Version
3, BioRad).
A calibration curve (Fig. 2) was obtained by
plotting the Ct value, which is defined as the cycle
number crossing the threshold point, versus the log-
arithm of starting quantity of gene copy numbers or
concentration of template DNA.
The calculation of the number of arsC gene copies
in a sample was based on the assumption (Gruntzig et
al., 2001) that only one copy of the arsC gene is
present per genome.
The conversion formula relating DNA weight to
gene copies was derived as follows (Ausubel et al.,
1992): the average molecular weight of a double-
stranded DNA base pair is 649 Da and the arsC
gene has 353 base pairs, so the molecular weight of
the gene is 649  353 = 229,097 Da. Thus 1 arsC
gene weights 229,097/6.023  1023 = 3.8  10 19 g,
where Avogadro’s number = 6.023  1023 molecules
(bp)/3mol.
2.7. Analysis of the PCR product: electrophoresis,
melting curve, sequencing
The product of the general PCR amplification of
target DNA was analyzed by using 2% agarose gel
electrophoresis. The gel was visualized under a UV
transilluminator after staining with 0.5 Ag/ml ethidium
bromide.
The melting curve was used to measure the melting
temperature (Tm) of the double-stranded DNA after
PCR amplification to identify the number of amplified
products. At temperatures below the Tm of the ampli-
fied product, SYBRR Green I will bind to the PCR
products and fluoresce brightly. As the Tm is reached,
the DNA denatures and releases SYBRR Green I,
causing a sharp decline in fluorescence. Plotting the
negative first derivative of relative fluorescence units
d(RFU) versus temperature change dT versus tem-

Fig. 1. Real-time PCR amplification plots of relative fluorescence units (RFU) for SYBRR Green I vs. cycle number with mixture of primer set
A amlt-42-f/amlt-376-r and primer set B smrc-42-f/smrc-376-r. The threshold was automatically set at 1610 RFU and it was defined as 10 times
the mean standard deviation of fluorescence in all wells over the baseline cycles.
 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349 341

Fig. 2. Calibration curve plotting Log Starting Quantity of the arsC target gene in terms of copy number versus threshold cycle. Data were
generated from Fig. 1 with a correlation coefficient is 0.999, PCR efficiency is 89.3%, Y = 3.609 X + 41.92.

perature, permits the evaluation of a melting peak and
Tm for the amplified product.
A melting curve (Fig. 3) was obtained after the
real-time PCR amplification, via the following proto-
col: 95 jC for 1 min, 55 jC for 1 min and 80 cycles of
55 jC for 10 s with 0.5 jC increase of temperature.
Sequencing was performed by the Nucleic Acid
and Protein Services (UBC), using the ‘BigDye’
fluorescent-labeled DyeDeoxy and an automated Ap-
plied Biosystems DNA sequencer. The DNA sequenc-
ing results were aligned and searched in the BLAST
database.

Fig. 3. Melting curve plotting the negative derivative of relative fluorescence units with respect to temperature as d(RFU)/dT versus
temperature.
342 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349

3. Results and discussion
3.1. Multiple arsC sequence alignment and selection
of primers
A number of studies have recently been published
on the design of primers based on multiple sequence
alignments. For example, Borneman and Hartin
(2000) used PILEUP (Genetic Computer Group,
Madison, WI) to design two PCR primer pairs to
amplify rDNA from four major phyla of fungi, based
on the alignment of 213 fungal small-subunit rDNA
sequences. In the study of bacterial populations asso-
ciated with arsenic contaminated environmental sam-
ples, Macur et al. (2004) used four sets of primers and
their design was based on arsC genes corresponding
to five different bacterial strains: E. coli, Bacillus
subtilis; Pseudomonas aeruginosa and P. putida;
Agrobacterium tumefaciens. Each primer pair was
used separately to amplify the arsC genes from
DNA extracted from separate isolates, and the PCR
products were then used in dot-blot hybridization. The
authors concluded that the failure of most arsC
probes, except one for A. tumefaciens, to hybridize
with DNA from soil isolates could be explained by the
diversity of known arsC sequences and this diversity
could restrict detection of arsC genes when using
probes designed from phylogenetically different
organisms.
It is certainly unreasonable to expect a single
primer to target every sequence in an environmental
sample, given that variation exists even in the most
highly conserved regions of the gene. Our study was
aimed at designing a primer based on the most diverse
group of bacteria possible, not just organisms from a
similar genus. The approach was to divide the arsC
sequences into groups using fragments of homologous
alignments and to design specific primers for each of
these groups.
A group of 13 arsC sequences was selected from
an alignment of 120 sequences (Table 4) by eliminat-
ing the sequences with low alignment scores. This
group consists of bacterial strains belonging to the
same family of Enterobacteriaceae apart from Acid-
iphilium multivorum, which belongs to Acetobacter-
aceae. The lengths of the arsC sequences retrieved
from the GenBank were in a range of 414 bp for
Klebsiella and 492 bp for Yersinia enterocolitica. The
results of alignment of 13 sequences of arsC genes
gave two subgroups of relatively homologous re-
gions, which were long enough to design the forward
and reverse primer pair at the positions described in
Table 5.
Subgroup I consists of 10 strains and Subgroup II
has three strains.
Subgroup I:
(a) A. multivorum plasmid pKW301, Klebsiella oxy-
toca plasmid pMH12. They are 100% identical in
the arsC position 43 to 62 and they are the 100%
match for the both forward and reverse primer pair
amlt-42-f/amlt-376-r.

Table 4
Selected group of bacteria or plasmids with homologous arsC genes and corresponding PCR primers
Microorganism GenBank accession Location of arsC Primer pair
(1) Yersinia enterocolitica Tn2502 U58366 plasmid amlt-42-f/smrt-376-r
(2) Yersinia enterocolitica pYVe227 NC_002120 plasmid amlt-42-f/smrt-376-r
(3) Salmonella typhimurium plasmid R64 AP005147 plasmid amlt-42-f/amlt-376-r
(4) Plasmid R773 J02591 plasmid amlt-42-f/amlt-376-r
(5) Plasmid R46 U38947 plasmid amlt-42-f/amlt-376-r
(6) Shigella flexneri 2a str. 301 NC_004337 genome amlt-42-f/amlt-376-r
(7) Shigella flexneri 2a str. 2457T NC_004741 genome amlt-42-f/amlt-376-r
(8) Escherichia coli K-12 (sub strain MG1655) NC_000913 genome amlt-42-f/amlt-376-r
(9) Escherichia coli 0157:H7 NC_002655 genome amlt-42-f/amlt-376-r
(10) Escherichia coli CFT073 NC_004431 genome amlt-42-f/amlt-376-r
(11) Serratia marcescens plasmid R478 SMA288983 plasmid smrt-42-f/smrt-376-r
(12) Klebsiella oxytoca plasmid pMH12 AF168737 plasmid amlt-42-f/amlt-376-r
(13) Acidiphilium multivorum plasmid pKW301 AB004659 plasmid amlt-42-f/amlt-376-r
Table 5
Multiple arsC gene sequence alignments used for primers design

Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349

Star (*) means that the nucleotides in that column are identical in all sequences (not including the two primer sets) in the alignment.
Underlined bases represent mismatches with primer pair amlt-42-f/aml-376-r.

343
344 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349

(b) Escherichia coli CFT073, Shigella flexneri 2a
strain 301, Shigella flexneri 2a strain 2457T, E.
coli K-12 (sub strain MG1655), E. coli 0157:H7.
The arsC sequences are different from the forward
primer amlt-42-f only by one nucleotide in
position 45 bp for E. coli CFT073 and one in
the position of 384 for the amlt-376-r reverse
primer for all five strains.
(c) Salmonella typhimurium plasmid R64, plasmid
R773, plasmid R46 are 100% similar for the amlt-
42-f primer, and there are two mismatches for the
amlt-376-r primer in position 387 and 390.
Subgroup II:
(a) Serratia marcescens plasmid R478 is different in
only one nucleotide with only amrt-376-r primer
in the position of 390.
(b) Y. enterocolitica Tn2502 and Y. enterocolitica
pYVe227 are different by two nucleotides from the
primer smrt-376-r and have one nucleotide
difference with the amlt-42-f primer.
otide sequence analysis of PCR products was
performed. A band of 400 bp on agarose gel (Fig.
4), which is the approximate length of the arsC
gene, was cut from the gel for further confirmation
that the PCR product is the arsC gene, and a
sequencing reaction was performed. The results of
a BLAST search confirmed that the PCR product had
a sequence corresponding to the arsC gene of each
strain (Table 6).
The arsC sequence of E. coli W3110 not registered
in the GenBank, and for confirmation of this sequence
the one from E. coli K-12 substrain MG1655, which
has been reported to be similar (http://www.ecoli.aist-
nara.ac.jp) was used instead.
Two negative controls were also set up to confirm
primer specificity. Sterile deionized water was used as
a template for one, and for a second DNA from K.
pneumonia known to be free of the arsC gene. The
400 bp PCR products were absent for both of them
(Fig. 4).

Two pairs of primers were designed with a minimal

numbers of mismatches (one or a maximum of two
nucleotides) from arsC homologous parts of 13
strains. Both pairs were combined when used in the
general PCR and real-time PCR assays.

3.2. Specificity of the arsC primers

Most studies use total DNA as a standard in real-

time PCR experiments. Known arsC sequences were
used as a standard in this work. In order to confirm
primer specificity, four strains of bacteria and plas-
mids known to possess arsC genes were used as
positive controls for PCR. All of them were from
the group of bacteria with the closest arsC alignment
results, listed in Table 4: K. oxytoca plasmid pMH12;
plasmid R64 in E. coli K12 host; S. marcescens
plasmid R478; E. coli K12 W3110.
DNA from each of these four strains was prepared
and used as a template in PCR with the cor-
responding primer pair (Table 3). General PCR was
performed with PCR conditions as described above. Fig. 4. Agarose gel electrophoresis of the PCR products of the
DNAs selected as the positive and negative controls. Lane 1: 100bp
Variability in PCR-based techniques can occasionally DNA ladder; lane 2: S. marcescens plasmid R478; lane 3: K.
result in amplification products that are not com- oxytoca plasmid pMH12; lane 4: IncN plasmid R46; lane 5: E. coli
prised of the intended DNA; consequently a nucle- W3110; lane 6: K. pneumonia; lane 7: Sterile deionized water.
 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349
Table 6
Sequences alignment results of PCR products of arsC
Strains/sequencing Primer pair Match,
results % bp
K. oxytoca plasmid amlt-42-f/amlt-376-r 87
pMH12
E. coli K12 W3110 amlt-42-f/amlt-376-r 99
E. coli plasmid R64 amlt-42-f/amlt-376-r 85
S. marcescens smrt-42-f/smrt-376-r 98
plasmid R478
a
The extent to which two nucleotide sequences are invariant.
Therefore, the results of sequence alignments with
the GenBank database confirmed that the selected
primers are specific for the arsC gene.
3.3. Accuracy and detection limit for real-time PCR
To generate more complete data about the presence
of the arsC gene belonging to different strains in soil
samples, the next step was to make a calibration
curve, not with total DNA, but with arsC gene
prepared from the 400 bp band on an agarose gel.
Despite the length, and the sequence variations among
the arsC genes in 13 chosen strains from 414 to 492
bp, the designed primer was appropriate for all 13
homologous arsC genes in this selected group.
Total DNA is not suitable as a standard template
for establishing a calibration curve in this work
because there could be a significant difference in
total genomic DNA lengths from strain to strain.
This would affect the calculation of the arsC copy
numbers. Furthermore, not all strains have their
complete genome sequenced, and the GenBank da-
tabase has limited information on most of these
strains.
The isolated arsC target sequence was used as an
external standard in establishing the calibration curve
for real-time PCR (Figs. 1 and 2). The linear range
covered six orders of magnitude, ranging from
6.8  10 1 to 6.8  10 6 ng arsC DNA, cor-
responding to 1.8  109 to 1.8  104 copies of the
gene.
A hot start mechanism was used to obtain more
accurate results by decreasing non-specific products,
such as primer-dimer, because SG I dye is non-
specific and binds to any double-stranded DNA.
Real-time PCR results are reliable only under opti-
345
mized conditions for each target gene. In addition,
special attention was paid to primer concentration.
Identitya, In the melting curve (Fig. 3) experiment with a
high template concentration (6.8  10 1, 6.8  10 2,
245/281 6.8  10 3, 6.8  10 4 ng arcC DNA), there was
only one peak with a melting temperature of 89 jC,
396/397
228/267
and it was the arsC PCR product. Two peaks were
322/326 identified in the melting curve with Tm of 89 and 78
jC for the negative control and low amounts of
template. Multiple peaks resulted from multiple am-
plification products, and the additional non-specific
peak could be due to the presence of primer-dimers.
Primer-dimers are formed when concentration of the
designated template is low, and they typically melt at
a lower Tm than the specific product because of their
shorter length. The presence of primer-dimers resulted
in competition for reagents and a reduction of the
dynamic range of the calibration curve from 6 orders
of magnitude to 4. The calibration curve, obtained
from four points without the presence of primer-dimer
had a correlation coefficient of 0.998, a slope of
3.354, an intercept of 39.72, and a PCR efficiency
of 98.7%. There was no significant difference be-
tween the four and six point calibration curves.
Therefore the contribution of primer-dimer was neg-
ligible in calculating the arsC PCR product. The
primer concentration used as described above is
suitable for the quantification of arsC target gene.
The detection limit of the arsC target gene was
estimated to be ca. three arsC gene copies from Fig. 2,
based on the blank signal plus three times the standard
deviation. However, this is very optimistic limit in
view of primer-dimer formation being detected at
around 104 copies. Beller et al. (2002) reported a
detection limit of five copies of a catabolic bssA gene
of hydrocarbon-degrading bacteria in sediment sam-
ples of fuel-contaminated groundwater.
3.4. Incubation of soil samples for identification of
the arsC gene
Our first goal in these experiments was the detec-
tion of the arsC gene in soil samples. In order to select
arsenic-resistant bacteria from soil samples, and re-
move soil particles which could contain humic mate-
rials, the samples listed in Table 2 were placed on
nutrient agar enriched with 10 ppm As (V) and 10 Ag/
ml cycloheximide (to inhibit fungal growth) for over-
346 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349

night incubation at room temperature. Nutrient agar is
suitable for growth of all 13 bacterial strains that have
been used for primer design (Table 4). Next, all
growing colonies were transferred to nutrient broth
with 10 ppm As(V) and incubated at room tempera-
ture for 3 days. Genomic and plasmid DNA were
extracted from the broth.
A control containing only one bacterial strain
known to carry the arsC gene was prepared as
follows: a soil sample from the UBC campus was
autoclaved (121 jC, 17 psi, 20 min), then spiked with
a culture of E. coli W3110 (12 h growth, 1 ml of
culture per 10 g of soil) and incubated overnight at
room temperature on a petri plate with nutrient agar
containing the same concentration of As (V) and
cycloheximide. Another petri plate with pure E. coli
culture without soil (Control #2) was incubated in the
same medium and under the same conditions. After
incubation genomic DNA from both control samples
was extracted.
Following amplification, the arsC gene was found
in the following genomic DNA samples: YK-1, YK-5,
YK-6a, 408 which were collected from the gold mine,
and in the CC3 and Creek soil sample from Trail.
Samples YK-6b and 406 from the gold mine and soil
CC2 had the arsC gene in plasmid DNA. Only one
sample from the gold mine, #407, had the arsC in
both genomic and plasmid DNA.
The results of the control experiment with E. coli
spiked UBC soil showed the presence of the arsC
gene (a 400 bp band on the agarose gel) in genomic
DNA, similar to the one from pure E. coli culture.
Thus the results of the first stage of the study showed
that applied PCR conditions were suitable for detec-
tion of the arsC gene in environmental samples and
all samples tested in this experiment from arsenic-
contaminated sites had the arsC gene in genomic and/
or plasmid DNA.
3.5. Quantification of the arsC gene in soil samples
by real-time PCR
The goal of the second phase of the experiment
was to quantify the arsC genes in samples from
arsenic-contaminated sites. The soil samples used in
this phase of study are described in Tables 2 and 7. To
ensure that the extracted DNA was of sufficient
quality to be used in real-time PCR, amplification
with universal 16S rDNA primers (Lane, 1991) was
performed. Following amplification, the presence of
16S rDNA was confirmed in all DNA templates
(Table 7).
Two samples from the gold mine, #406 and #408
showed 0.88  104 and 2.38  104 arsC gene copies
per ng total DNA, respectively, and one sample from
Creek soil had 1.17  104 copies per ng total DNA
(Table 2). The arsC abundance was below the
detection level in the other samples. When general
PCR of the arsC gene was conducted with DNA
extracted directly from all 10 samples (Table 2), there

Table 7
Identification and quantification of the arsC gene in soil samples from bioreactor
Namea Presence of 16S rDNA Presence of the arsC gene Quantification of the arsC gene
by general PCRb from original samples by from original samples by real-time
general PCRc PCR (copies/ng total DNA)d
Soil #1 YES NO ND
Soil #2 YES YES 12.50  104 F 4.85  104
Soil #3 YES YES 8.20  104 F 1.97  104
Soil #4 YES YES 15.60  104 F 2.21  104
Soil #5 YES NO ND
Soil #6 YES YES 5.43  104 F 1.35  104
Soil #7 YES YES 1.92  104 F 1.57  104
a
Soil samples were obtained from a laboratory-scale bioreactor, which was packed with mixture of active biomass and by-products
discarded from pulp mill (Kingston, Ontario).
b
16S rDNA was amplified by using 63f forward primer (CAGGCCTAACACATGCAAGTC) and 1387r reverse primer
(GGGCGGWGTGTACAAGGC) (Lane, 1991).
c
arsC gene was amplified by using mixture of two primer sets amlt-42-f/amlt-376-r and smrt-42-f/smrt-376-r.
d
Results were from duplicate samples, n = 2. Error was given as standard deviation. ND is below the detection limit.
 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349
were no bands on the agarose gel, although when the
same samples were incubated in the nutrient broth,
all of them had an arsC PCR product on the gel.
Obviously, the arsC gene appears in very low numb-
ers in a soil and additional enrichment by incubation
of the soil samples in nutrient-rich medium increases
the number of bacteria and consequently brings the
arsC number to a detectable level. It has been also
reported for E. coli 0157:H7 16S stx1 and eae genes
that their copy numbers increases with prolongation
of enrichment time of the soil samples (Ibekwe and
Grieve, 2003).
Soil samples from a bioreactor containing pulp mill
biomass (Table 7) showed high numbers of the arsC
gene copies. Five of seven soil samples had the arsC
band on agarose gel and showed approximately a
tenfold higher count of arsC gene copies, in a range
from 1.92  104 to 15.60  104 per ng total DNA,
than samples described in Table 2.
In reality, the number could be higher because of
the presence of other strains, for which specific
primers were not designed.
Gruntzig et al. (2001) described an approach to
quantify denitrifying nitrite reductase gene (nirS) in
Pseudomonas stutzeri by using real-time PCR and the
TaqMan probe. The detection limit was 100 copies or
greater and copy numbers varied from 1  102 to
1.1  1011 per Ag total DNA in different environmen-
tal samples. Beller et al. (2002) used a similar method
for the quantification of hydrocarbon-degrading bac-
teria. The target was the bssA gene, associated with
the first step of anaerobic toluene and xylene degra-
dation. The primer-TaqMan probe set was based on
consensus regions of the bssA gene from four bacterial
strains. In all these studies it was apparent that a
critical limitation to an accurate quantification of a
specific gene in environmental samples is our incom-
plete knowledge of the gene sequence of all repre-
sentatives of the microbial community contained in
the sample. Another major obstacle lies in the diver-
sity of sequences of the target gene in different
microorganisms, which restricts primer design
options.
Certainly many more than 120 bacteria have the
arsC gene. In some soil samples fungi could play a
dominant role in reduction of arsenic. Cullen et al.
(1994) and Lehr et al. (2003) reported that several
fungi, such as Scopulariopsis brevicaulis and Scopu-
347
lariopsis koningii are able to reduce arsenate to
arsenite. The arsC gene could not be amplified from
DNA obtained from the fungi Sclerotinia verartri
which reduce arsenate in high yield (Cullen et al.,
2003). Unfortunately, the genetic system of fungi that
are involved in arsenic biotransformation remains
unknown.
The identification of the arsC gene in contaminat-
ed samples could provide information on the nature of
the contamination process. In some parts of the world
such as Bangladesh, arsenic is being mobilized from
sediment into drinking water, causing grief to millions
of people (Christen, 2001; McArthur et al., 2001).
Various mechanisms have been proposed for this
release, including biological reduction of sediment-
bound arsenate. Information about gene copy numbers
would be helpful to decide if this contamination was
due to biological or geological activity, or both. In
addition, some important environmental cleanup deci-
sions might be made easier if quantitative information
about the presence of the arsC gene was available.
Comparison of the amount of the arsC genes in
different samples could be valuable for evaluation of
the relative mobility of arsenic contamination, since
arsenate is generally believed to be more mobile than
arsenite (Cullen and Reimer, 1989). Mobilization of
arsenate from a soil is less likely to happen if
reduction to arsenite cannot occur and knowledge of
the presence or absence of the arsC gene would
provide some basis for making a choice for decon-
tamination strategies.
Acknowledgements
We thank Dr. B. Rosen (Wayne State Universi-
ty), Dr. M. Kelly (University of British Columbia),
Dr. T. Komano (Tokyo Metropolitan University)
and Dr. D. Taylor (University of Alberta) for
providing E. coli K12 W3110, K. oxytoca plasmid
pMH12, E. coli plasmid R64 and S. marcescens
plasmid R478, respectively. We thank Dr. I. Koch
(Royal Military College) for bioreactor samples and
Dr. D. Ryan (Institute of Technology, Carlow) for
helpful comments.
This work was supported in part by the Natural
Sciences and Engineering Research Council of
Canada.
348 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349
References
Alm, E.W., Oerther, D.B., Larsen, N., Stahl, D.A., Raskin, L., 1996.
The Oligonucleotide probe database. Appl. Environ. Microbiol.
62, 3557 – 3559.
Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman,
J.D., Smith, J.A., Struhl, K., 1992. Short protocols in molecular
biology. Greene Publishing Associates and John Wiley and
Sons, A2 – A4.
Beller, H.R., Kane, S.R., Legler, T.C., Alvarez, P.J.J., 2002. A real-
time polymerase chain reaction method for monitoring anaero-
bic hydrocarbon-degrading bacteria based on a catabolic gene.
Environ. Sci. Technol. 36, 3977 – 3984.
Berdal, K.G., Holst-Jensen, A., 2001. Roundup readyR soybean
event-specific real-time quantitative PCR assay and estimation
of the practical detection and quantification limits in GMO anal-
yses. Eur. Food Res. Technol. 213, 432 – 438.
Blattner, F.R., Plunkett III, G., Bloch, C.A., Pema, N.T., Burland,
V., Rilev, M., Collado-Vides, J., Glasner, J.D., Rode, C.K.,
Mayhew, G.F., Gregov, J., Davis, N.W., Kirkpatrick, H.A.,
Goeden, M.A., Rose, D.J., Mau, B., Shas, Y., 1997. The com-
plete genome sequence of Escherichia coli K-12. Science 277,
1453 – 1474.
Borneman, J., Hartin, R.J., 2000. PCR primers that amplify fungal
rRNA genes from environmental samples. Appl. Environ.
Microbiol. 66, 4356 – 4360.
Braker, G., Zhou, J., Wu, L., Devol, A.H., Tiedje, J.M., 2000.
Nitrite reductase genes (NirK and NirS) as functional markers
to investigate diversity of denitrifing bacteria in Pacific North-
west marine sediment communities. Appl. Environ. Microbiol.
66, 2096 – 2104.
Butcher, B.G., Deane, S.M., Rawlings, D.E., 2000. The chromo-
somal arsenic resistance genes of Thiobacillus ferrooxidans
have an unusual arrangement and confer increased arsenic and
antimony resistance to Escherichia coli. Appl. Environ. Micro-
biol. 66, 1826 – 1833.
Chen, R.W., Piiparinen, H., Seppanen, M., Koskela, P., Sarna, S.,
Lappalainen, M., 2001. Real-time PCR for detection and
quantitation of hepatitis B virus DNA. J. Med. Virol. 65,
250 – 256.
Christen, K., 2001. The arsenic threat worsens. Environ. Sci. Tech-
nol. 35, 288A – 291A.
Cullen, W.R., Reimer, K.J., 1989. Arsenic speciation in the envi-
ronment. Chem. Rev. 89, 713 – 764.
Cullen, W.R., Li, H., Hewitt, G., Reimer, K.J., Zalunardo, N., 1994.
Identification of extracellular arsenical metabolites in the growth
medium of the microorganisms Apiotrichum humicola and Sco-
pulariopsis brevicaulis. Appl. Organomet. Chem. 8, 303 – 311.
Cullen, W.R., Polishchuk, E., Reimer, K.J., Sun, Y., Wang, L., Lai,
V.W.-M., 2003. In: Chappell, W.R., Abernathy, C.O., Calderon,
R.L., Thomas, D.J. (Eds.), Arsenic Exposure and Health Effects
V. Elsevier, Amsterdam, pp. 69 – 78.
Dhankher, O.P., Li, Y., Rosen, B.P., Shi, J., Salt, D., Senecoff, J.F.,
Sashti, N.A., Meagher, R.B., 2003. Engineering tolerance and
hyperaccumulation of arsenic in plants by combining arsenate
reductase and gamma-glutamylcysteine synthetase expression.
Nat. Biotechnol. 20, 1140 – 1145.
Filion, M., St-Arnaud, M., Jabaji-Hare, S.H., 2003. Direct quanti-
fication of fungal DNA from soil substrate using real-time PCR.
J. Microbiol. Methods 53, 67 – 76.
Gruntzig, V., Nold, S.C., Zhou, J., Tiedje, J.M., 2001. Pseudomo-
nas stutzeri nitrite reductase gene abundance in environmental
samples measured by real-time PCR. Appl. Environ. Microbiol.
67, 760 – 768.
Hermansson, A., Lindgren, P.-E., 2001. Quantification of ammonia-
oxidizing bacteria in arable soil by real-time PCR. Appl. Envi-
ron. Microbiol. 67, 972 – 976.
Humar, B., D’Orazio, D., Albrecht, C., Bauerfeind, P., Muller,
H., Dobbie, Z., Bendik, I., 2001. Expression of putative an-
ticancer targets in familial adenomatous polyposis and its
association with the APC mutation status. Int. J. Oncol. 19,
1179 – 1186.
Ibekwe, A.M., Grieve, C.M., 2003. Detection and quantification of
Escherichia coli O157:H7 in environmental samples by real-
time PCR. J. Appl. Microbiol. 94, 421 – 431.
Jackson, C.R., Harper, J.P., Willoughby, D., Roden, E.E., Churchill,
P.F., 1997. A simple, efficient method for the separation of
humic substances and DNA from environmental samples. Appl.
Environ. Microbiol. 63, 4993 – 4995.
Komano, T., Kubo, A., Nisioka, T., 1987. Shufflon: multi-inver-
sion of fav contiguous DNA segments of plasmid R64 creates
seven different open reading frames. Nucleic Acids Res. 15,
1165 – 1172.
Lane, D.J., 1991. 16S/23S rDNA sequencing. In: Stackebrandt, E.,
Goodfellow, M. (Eds.), Nucleic Acid Techniques in Bacterial
Systematics. John Wiley & Sons Ltd., New York.
Lehr, C.R., Polishchuk, E., Delisle, M.-C., Franz, C., Cullen,
W.R., 2003. Arsenic methylation by micro-organisms isolated
from sheepskin bedding materials. Hum. Exp. Toxicol. 22,
325 – 334.
Macur, R.E., Jackson, C.R., Botero, L.M., Mcdermott, T.R., Ins-
keep, W.P., 2004. Bacterial populations associated with the ox-
idation and reduction of arsenic in an unsaturated soil. Environ.
Sci. Technol. 38, 104 – 111.
McArthur, J.M., Ravenscroft, P., Safiulla, S., Thirlwall, M.F.,
2001. Arsenic in groundwater. Testing pollution mechanisms
for sedimentary aquifers in Bangladesh. Water Resour. Res.
37, 109 – 117.
Mukhopadhyay, R., Rosen, B.P., Phung, L.T., Silver, S., 2002.
Microbial arsenic: from geocycles to genes and enzymes. FEMS
Microbiol. Rev. 26, 311 – 325.
O’Mahony, J., Hill, C., 2002. A real time PCR assay for the detec-
tion and quantitation of Mycobacterium avium subsp. paratu-
berculosis using SYBR Green and the Light Cycler. J.
Microbiol. Methods 51, 283 – 293.
Permingeat, H.R., Reggiardo, M.I., Vallejos, R.H., 2002. Detection
and quantification of transgenes in grains by multiplex and real-
time PCR. J. Agric. Food Chem. 50, 4431 – 4436.
Rosen, B.P., 2002. Biochemistry of arsenic detoxification. FEBS
Lett. 529, 86 – 92.
Ryan, D., Colleran, E., 2002. Arsenical resistance in the IncHI2
plasmids. Plasmids 47, 234 – 240.
Shi, J., Vlamis-Gardikas, A., Aslund, F., Holmgren, A., Rosen,
B.P., 1999. Reactivity of glutaredoxins 1,2, and 3 from
 Y. Sun et al. / Journal of Microbiological Methods 58 (2004) 335–349 349

Escherichia coli shows that glutaredoxin 2 is the primary
hydrogen donor to arsC-catalyzed arsenate reduction. J. Biol.
Chem. 274, 36039 – 36042.
Silver, S., Phung, L.T., Rosen, B., 2002. In: Frankenberger, W.T.
(Ed.), Environmental Chemistry of Arsenic. Marcel Dekker,
New York, pp. 247 – 272.
Stubner, S., 2002. Enumeration of 16S rDNA of Desulfotomaculum
lineage 1 in rice field soil by real-time PCR with SybrGreen
detection. J. Microbiol. Methods 50, 155 – 164.
Zhou, J., Bruns, M.A., Tiedje, J.M., 1996. DNA recovery from
soils of diverse composition. Appl. Environ. Microbiol. 62,
316 – 322.