Environ Sci Pollut Res (2015) 22:10206–10213
DOI 10.1007/s11356-015-4216-1
RESEARCH ARTICLE
Development of a set of bacterial biosensors for simultaneously
detecting arsenic and mercury in groundwater
Chi-Wei Huang & Shih-Hung Yang & Man-Wai Sun &
Vivian Hsiu-Chuan Liao
Received: 4 November 2014 / Accepted: 5 February 2015 / Published online: 21 February 2015
# Springer-Verlag Berlin Heidelberg 2015
Abstract There is a growing need for effective and inexpen-
sive environmental monitoring strategies for assessing heavy
metal contamination levels. We developed a set of bacterial
biosensors to simultaneously detect multiple bioavailable
heavy metals (As(III) and Hg(II)). The biosensors provide a
choice of the two reporter systems, luxCDABE and gfp, com-
bined with metal responsive regulatory elements (ars and mer
for As(III) and Hg(II), respectively). The results showed that
the induction of the luxCDABE-based constructs was more
sensitive than that of the gfp-based constructs for the detection
of As(III) and Hg(II). In addition, both the luminescent and
fluorescent biosensors readily distinguished As and Hg con-
centrations in groundwater samples to meet the groundwater
quality standards. Due to the potentially complicated
chemicals present in environmental samples, using a set of
bacterial biosensors with different reporter genes to simulta-
neously determine the bioavailable proportions of heavy
metals is desirable.
Keywords Heavy metals . Biosensors . Bioavailable .
Simultaneous detection
Introduction
Heavy metal contamination is a worldwide problem. Metals
including mercury (Hg) and arsenic (As) have caused
Responsible editor: Philippe Garrigues
Chi-Wei Huang and Shih-Hung Yang contributed equally to this work.
C.<W. Huang : S.<H. Yang : M.<W. Sun : V. H.<C. Liao (*)
Department of Bioenvironmental Systems Engineering, National
Taiwan University, No. 1, Sec. 4, Roosevelt Rd., Taipei 106, Taiwan
e-mail: vivianliao@ntu.edu.tw
hazardous effects on human health and the environment in
many parts of the world. Therefore, there is a growing need
for effective and inexpensive environmental monitoring ap-
proaches to assess heavy metal contamination levels and to
serve as a warning system against future heavy metal
pollution.
Traditionally, heavy metals discharged into the environ-
ment have been analyzed using chemical assays, which
often require sample pretreatment and expensive equip-
ment. In addition, chemical methods yield little informa-
tion regarding the bioavailability of heavy metals and their
potential toxicity. To detect the bioavailable fraction of
specific metals in complex environments, the use of
bacteria-based biosensors with inducible reporter genes
has been investigated and described (Daunert et al. 2000;
Girotti et al. 2008). Bacterial biosensors typically combine
a metal promoter/operator element, which acts as the sens-
ing element, and a reporter gene coded for an easily mea-
surable protein (Daunert et al. 2000). Thus, the expression
of the reporter gene produces a measurable signal when
the bacteria respond to bioavailable metal ions (Daunert
et al. 2000; Ramanathan et al. 1997).
Although different bacterial biosensors have been studied
for the detection of bioavailable heavy metals (Hansen and
Sørensen 2000; Liao and Ou 2005; Liao et al. 2006;
Stocker et al. 2003), most studies have examined single
metals within a sample using biosensors. Limited stud-
ies have examined the presence of multiple heavy
metals simultaneously in environmental samples. Hence,
in this report, we developed a set of bacterial biosensors
to simultaneously detect multiple bioavailable heavy
metals, including As(III) and Hg(II). The biosensors de-
veloped here provide a choice of the two reporter sys-
tems, luxCDABE and gfp, combined with metal respon-
sive regulatory elements (ars and mer for As(III) and
Hg(II), respectively).
Environ Sci Pollut Res (2015) 22:10206–10213
Materials and methods
Construction of the ars::luxCDABE recombinant plasmid
For the ars::luxCDABE recombinant plasmid, the DNA frag-
ment corresponding to the promoter/operator of the ars oper-
on and the arsR gene was PCR-amplified from the pI258
plasmid isolated from Staphylococcus aureus (NCTC
50581; National Collection of Type Cultures, Colindale, Lon-
don, UK). The PCR primers were designed as follows: for-
ward primer (5′-CCGGGATCCTAAAATAACATAGACAAT
AATCT-3′) and reverse primer (5′-CGCGAATTCCATCAA
CAGTCACCTGATT-3′) carrying the BamHI and EcoR1 re-
striction sites (underlined). The amplified DNA fragment was
purified and digested with BamHI and EcoRI. The BamHI-
EcoRI DNA fragment was subsequently inserted upstream of
the promoterless luxCDABE gene in the pUCD615 vector
(Rogowsky et al. 1987), generating the pAs-lux fusion plas-
mid (Fig. 1a). The recombinant plasmid pAs-lux was then
transformed to Escherichia coli (E. coli) DH5α.
Construction of the mer::luxCDABE and mer::gfp
recombinant plasmids
The construction of the mer::luxCDABE recombinant plasmid
was similar to the pAs-lux recombinant plasmid described
Fig. 1 Schematic organization of (A) pAs-lux
the biosensor plasmids developed
in this study. a pAs-lux, b pHg-
lux, and c pHg-gfp. Genes are
represented as arrows. Relevant
restriction sites for cloning are
denoted by their usual
abbreviations
(B) pHg-lux
(C) pHg-gfp
10207
above. For plasmid containing luxCDABE as the reporter
gene, the DNA fragment corresponding to the promoter/
operator of the mer operon and the merR gene was PCR-
amplified from the pI258 plasmid with the forward primer
(5′-CCGGGATCCATGACTTGACCGTGTACT-3′) and re-
verse primer (5′-CGCGGAATTCCCCCCATTTATTTATC
AGGC-3′) carrying the BamHI and EcoRI restriction sites
(underlined). The resulting recombinant plasmid was desig-
nated pHg-lux (Fig. 1b). Recombinant plasmid pHg-gfp
(Fig. 1c) was constructed analogously to that of pHg-lux ex-
cept that the mer promoter sequence was inserted into the
EcoRI and BamH1 sites of the promoterless pPROBE-NT’
vector (Miller et al. 2000).
Bacterial growth and luminescence induction assays
A single colony of E. coli harboring pAs-lux or pHg-lux was
grown overnight in Luria-Bertani (LB) media containing
100 μg/mL of ampicillin at 37 °C. The overnight culture
was diluted in fresh LB medium containing 100 μg/mL am-
picillin at a 1:100 ratio and incubated at 37 °C in an orbital
shaker at 225 rpm until an OD600 of 0.6 was reached. The
cultures were then added in 180-μL aliquots to triplicate wells
of a 96-well plate already containing 20 μL of the appro-
priate concentration of the metals to be tested (As(III)
and Hg(II)) in LB.
10208
Subsequently, the 96-well plates were placed in a
temperature-controlled microplate fluorometer and
luminometer (Thermo Labsystems, Helsinki, Finland). The lu-
minescence was measured at intervals for the duration of the
experiment. Bacterial cell numbers were monitored by measur-
ing the optical density at 600 nm, and the luminescence of the
luxCDABE-producing cells was also measured. Raw lumines-
cence intensities were expressed in the instrument’s arbitrary
relative unit (RLU). The specific bioluminescence intensity
(SLI) is defined as the RLU divided by the number of bacterial
cells measured at each metal concentration and time point. The
induction ratios (IRs) were calculated using the formula IR=Li/
Lb, where Li is the luminescence (in SLI) value of the test
sample and Lb is the luminescence (in SLI) value of the cells
containing no metal (blank). All experiments were repeated at
least three times.
Bacterial growth and fluorescence induction assays
The bacterial growth and fluorescence induction assays for the
metals were adapted from previous studies (Liao and Ou 2005;
Liao et al. 2006). Briefly, a single colony of E. coli harboring
pVLAS1 (Liao and Ou 2005) or pHg-gfp was grown overnight
in LB medium containing 50 μg/mL of kanamycin at 37 °C.
The overnight culture was diluted in fresh LB medium contain-
ing kanamycin (50 μg/mL) at a 1:100 ratio and incubated at
37 °C in an orbital shaker at 225 rpm until an optical density of
0.6 at 600 nm (OD600) was reached. The cultures were then
added in 180-μL aliquots to triplicate wells of a 96-well plate
already containing 20 μL of the appropriate concentration of
the metals to be tested (As(III) and Hg(II)) in LB.
Subsequently, the 96-well plates were placed in a
temperature-controlled microplate fluorometer and
luminometer (Thermo Labsystems). The fluorescence was
measured at intervals for the duration of the experiment. Bac-
terial cell numbers were monitored by measuring the OD600,
and the fluorescence of the GFP-producing cells that were
grown in culture was also measured with excitation at
485 nm and emission at 535 nm. Raw fluorescence intensities
were expressed in the instrument’s arbitrary relative unit (RFU).
The specific fluorescence intensity (SFI) is defined as the RFU
divided by the number of bacterial cells measured at each metal
concentration and time point. The induction ratios (IRs) were
calculated using the formula IR=Li/Lb, where Li is the fluores-
cence (in SFI) value of the test sample, and Lb is the fluores-
cence (in SFI) value of cells containing no metal (blank). All
experiments were repeated at least three times.
Testing of multiple heavy metal contaminants
in environmental samples
The groundwater was spiked with both As(III) (500 μg/L) and
Hg(II) (20 μg/L). These concentrations are groundwater
Environ Sci Pollut Res (2015) 22:10206–10213
quality standards (nondrinking) in Taiwan. The groundwater
was then tested by adding 100 μL of metal-spiked groundwa-
ter sample and 100 μL of one of the DH5α strains harboring
the pAs-lux, pHg-lux, pAs-gfp, or pHg-gfp plasmid at an
OD600 of 0.6 to each well of a 96-well plate. Thus, both As(III)
and Hg(II) were simultaneously analyzed in the groundwater
using a set of As and Hg biosensors on the same 96-well plate.
The bacterial cells were incubated for 2–3 h at 37 °C, and then,
the SFI and SLI for the gfp-based biosensors and the lux-based
biosensors, respectively, were measured using the procedures
described above. For standard curves, known concentrations
of each metal were tested in parallel with 100 μL of deionized
distilled laboratory water in place of the groundwater sample.
A standard curve was generated from a linear regression of the
average SFI or SLI values at each particular metal concentra-
tion, and then, the concentration of each metal equivalent in
the groundwater samples was calculated from the derived
standard curve.
Data analysis
The experiments were performed at least three times for error
analyses. The results are presented as the mean±standard de-
viation (SD). A Student’s t test at α=0.05 was used to deter-
mine statistical significance. Standard curves were fit via a
linear regression analysis.
Results and discussion
Characterization of the arsenic bacterial biosensor
To construct the lux-based biosensor for the detection of arse-
nic, the ars promoter and the arsR gene in the pI258 plasmid
of S. aureus were inserted into the promoterless pUCD615
vector containing the luxCDABE gene (Rogowsky et al.
1987), thus creating a Pars–luxCDABE transcriptional fusion.
The recombinant plasmid was designated pAs-lux, as shown
in Fig. 1a. In the presence of As(III), recombinant plasmid
pAs-lux in the DH5α E. coli strain resulted in a significant
increase in the luminescence intensity relative to that of con-
trol cells with no As(III) (Fig. 2).
The time-dependent induction of the bacterial biosen-
sor in response to As(III) was analyzed by incubating
the bacteria with As(III) for various time intervals. The
results showed that As(III) induced luminescence of the
E. coli DH5α (pAs-lux) strain in a time-dependent man-
ner (Fig. 2a). The specific luminescence intensity of
E. coli DH5α (pAs-lux) continuously increased over
the background during the first 3 h of As(III) exposure
(6.7 μM, 500 μg/L), after which the luminescence dra-
matically decreased (Fig. 2a). In addition, the kinetic
profile of As(III) induction of E. coli DH5α (pAs-lux)
Environ Sci Pollut Res (2015) 22:10206–10213
(A)
(B)
Fig. 2 Characterization of lux-based arsenic bacterial biosensors. a
Time-dependent induction of the DH5α (pAs-lux) biosensor due to
As(III). The DH5α cells harboring the pAs-lux plasmid were exposed
to 6.7 μM As(III), and the specific luminescence intensity (in SLI) was
determined after various exposure periods. b Dose-dependent induction
of the DH5α (pAs-lux) biosensor due to As(III). The luminescence of the
DH5α cells harboring the pAs-lux plasmid was determined after 2-h
incubation periods with various concentrations of As(III). The induction
ratio is the SLI of the As(III)-exposed sample divided by the SLI of a no-
As(III) control, and the no-As(III) control was arbitrarily set to 1.0. The
data presented here represent the mean and standard deviation of at least
three independent experiments
was similar to toluene induction reported of E. coli
DH5α (pTOLLUX) (Li et al. 2008), which may be
due to the biochemical property of the lux reporter gene
(Dollard and Billard 2003; Hakkila et al. 2002).
E. coli DH5α (pAs-lux) was further examined for its
dose-dependent induction by As(III). A 2-h induction
time was selected for the dose–response curve, as it
produced a sufficiently high luminescence signal while
maintaining a short assay time. Figure 2b shows that the
intensity of the luminescence increased with increasing
concentrations of As(III). As(III) at concentration of
0.1 μM (7.5 μg/L) induced a significant increase (2.5-
fold) in the signal compared the control with no As(III).
10209
The luminescence signal increased with increasing con-
centrations of As(III) up to 10 μM (750 μg/L), after
which the luminescence began to decrease (Fig. 2b).
This result may be due to the toxicity of As(III) to
the bacteria.
Bacterial biosensors for As(III) have been previously de-
veloped using various reporter genes, such as lacZ, luxAB,
luxCDABE, luc, and gfp with various induction patterns and
sensitivities for detecting As(III) (Liao and Ou 2005; Sharma
et al. 2013; Stocker et al. 2003; Tauriainen et al. 1997). The
lucFF biosensor was shown to detect 0.1, 3.3, 3.3 μM As(III)
when utilized in the host strains S. aureus RN4220, Bacillus
subtilis BR151, and E. coli MC1061, respectively (Tauriainen
et al. 1997). Stocker et al. (2003) reported that 4 μg/L
(0.05 μM) As(III) was accurately measured by both gfp
and lux biosensors, and the laz biosensor produced vis-
ible blue color at As(III) concentrations above 8 μg/L
(0.1 μM) (Stocker et al. 2003). The detection limit of
As(III) using the gfp biosensor was 0.4 μM for a 2-h
exposure period and 0.1 μM for an 8-h exposure period
(Liao and Ou 2005). The luxCDABE biosensor was re-
ported to detect As(III) between 0.05 and 0.8 μM (0.74
to 60 μg of As/L) (Sharma et al. 2013).
Several factors may influence the sensitivity of the biosen-
sor for detecting As(III). These factors include the exposure
time, the bacterial culture medium composition, the growth
phase of the harvested bacteria, and the amount of bacteria
per measurement (Hansen and Sørensen 2000; Hynninen
and Virta 2010; Tauriainen et al. 1997). Moreover, factors
including the ars plasmid, the vector, the analytical detection
method, the host strain, the reporter gene, and the methods
used to define the detection limit in the biosensor assays
may also contribute to varying sensitivities of As(III) detec-
tion. The sensitivity of the bacterial biosensor for As(III) in
this study is comparable with other reports (Hansen and
Sørensen 2000; Hynninen and Virta 2010; Tauriainen et al.
1997). However, it is difficult to directly compare the biosen-
sor developed in the current study with other biosensor con-
structs due to the factors described above.
Characterization of the mercury bacterial biosensors
The mercury bacterial biosensors were constructed using two
reporter genes, gfp and luxCDABE, and characterized by
green fluorescence and luminescence, respectively. The mer
promoter and the merR gene of the pI258 plasmid of S. aureus
were inserted into the promoterless vector, pUCD615, which
contains the luxCDABE gene (Rogowsky et al. 1987) and
pPROBE-NT’, which contains the gfp gene (Miller et al.
2000), respectively. The two constructed mercury biosensor
plasmids were designated pHg-lux (Fig. 1b) and pHg-gfp
(Fig. 1c), respectively.
10210
When the E. coli DH5α cells were transformed with the
pHg-lux recombinant plasmid, luminescence was observed in
response to Hg(II), resulting in a significant increase in the
luminescence intensity compared to that of control bacteria
without Hg(II) (Fig. 3). Similarly, the E. coli DH5α bacteria
harboring the pHg-gfp recombinant plasmid showed a signif-
icant increase in the fluorescence intensity compared to that of
control bacteria without Hg(II) (Fig. 4).
The time-dependent induction of the bacterial biosensors in
response to Hg(II) was analyzed by incubating the bacteria
with Hg(II) for various time intervals. The induction of the
luminescence and fluorescence under Hg(II) exposure oc-
curred in a time-dependent manner (Figs. 3a and 4a). For
(A)
(B)
Fig. 3 Characterization of lux-based mercury bacterial biosensors. a
Time-dependent induction of the DH5α (pHg-lux) biosensor due to
Hg(II). The DH5α cells harboring the pHg-lux plasmid were exposed
to 5 μM Hg(II), and the specific luminescence intensity (in SLI) was
determined after various exposure periods. b Dose-dependent induction
of the DH5α (pHg-lux) biosensor due to Hg(II). The luminescence of the
DH5α cells harboring the pHg-lux plasmid was determined after 2-h
incubation periods with various concentrations of Hg(II). The induction
ratio is the SLI of the Hg(II)-exposed sample divided by the SLI of a no-
Hg(II) control, and the no-Hg(II) control was arbitrarily set to 1.0. The
data presented here represent the mean and standard deviation of at least
three independent experiments
Environ Sci Pollut Res (2015) 22:10206–10213
(A)
(B)
Fig. 4 Characterization of gfp-based mercury bacterial biosensors. a
Time-dependent induction of the DH5α (pHg-gfp) biosensor due to
Hg(II). The DH5α cells harboring the pHg-gfp plasmid were exposed
to 2.5 μM Hg(II), and the specific fluorescence intensity (in SFI) was
determined after various exposure periods. b Dose-dependent induction
of the DH5α (pHg-gfp) biosensor due to Hg(II). The fluorescence of the
DH5α cells harboring the pHg-gfp plasmid was determined after 2-h
incubation periods with various concentrations of Hg(II). The induction
ratio is the SFI of the Hg(II)-exposed sample divided by the SFI of a no-
Hg(II) control, and the no-Hg(II) control was arbitrarily set to 1.0. The
data presented here represent the mean and standard deviation of at least
three independent experiments
E. coli DH5α harboring the pHg-lux plasmid, there was an
increase in the luminescence intensity during the first 3 h of
the incubation period, after which the luminescence began to
decrease (Fig. 3a). In contrast, E. coli DH5α (pHg-gfp) bac-
teria exhibited a different kinetic profile under Hg(II) expo-
sure. The specific fluorescence intensity continuously in-
creased over the background during the first 4 h of incubation
(Fig. 4a).
Dose-dependent induction under Hg(II) exposure was
further analyzed for the two mercury biosensors. A 2-h
exposure time with Hg(II) was chosen for the dose–re-
sponse curves, as the luminescence and fluorescence
Environ Sci Pollut Res (2015) 22:10206–10213 10211

Table 1 Summary of E. coli DH5α biosensor constructs

Reporter gene Metal detection Recombinant plasmid Induction time Detection limit Reference

luxCDABE As(III) pAs-lux 2h 0.1 μM This study

Hg(II) pHg-lux 2h 0.025 μM This study
gfp As(III) pVLAS1 2 or 8 h 0.4 μM (2 h); Liao and Ou 2005
0.1 μM (8 h)
Hg(II) pHg-gfp 2h 0.025 μM This study

signals obtained during the 2-h period were sufficiently
high. Both the luminescence and fluorescence intensities
increased with the increasing concentration of Hg(II) up
to a certain level (Figs. 3b and 4b). The dose–response
curves for the bioluminescent and fluorescent biosensors
slightly differed, and the DH5α (pHg-lux) bacterial sen-
sor strain appeared more sensitive than the DH5α (pHg-
gfp) bacterial sensor strain.
Figure 3b shows that the intensity of the luminescence
increased with increasing concentrations of Hg(II). Hg(II)
at a concentration of 0.025 μM (5 μg/L) was able to
induce a significant increase (1.5-fold) in the signal com-
pared to that of control (no Hg(II)). The luminescence
signal increased with increasing concentrations of Hg(II)
to a concentration of 5 μM (1000 μg/L), after which the
luminescence began to decrease (Fig. 3b). Interestingly,
5 μM (1000 μg/L) Hg(II) caused a decrease of the fluo-
rescence intensity (Fig. 4b). There is an apparent differ-
ence in the induction signals of the two biosensors. It
appears that the signal intensity of E. coli DH5α (pHg-
lux) was higher than that of DH5α (pHg-gfp) (Figs. 3
and 4), suggesting that the catalytic nature of the biolu-
minescent biosensor system allows a more rapid accu-
mulation of signals.
Several bacterial biosensors using various reporter genes
for mercury have been described (Hakkila et al. 2004;
Hansen and SÖrensen 2000; Ivask et al. 2007; Priyadarshi
et al. 2012) with various induction patterns and sensitivities
for As(III) detection. As described above for the arsenic bio-
sensor, in addition to the genetic constructs, several factors
may affect the sensitivities and induction coefficients of these
biosensors. The sensitivities of the bacterial biosensors for
Hg(II) developed in this study are comparable to other studies
(Hakkila et al. 2004; Hansen and SÖrensen 2000; Ivask et al.
2007; Priyadarshi et al. 2012). However, it is difficult to di-
rectly compare Hg biosensors in the current study with various
biosensor constructs due to the factors described above.
Testing for multiple heavy metals in groundwater using
the biosensors
Heavy metals are considered to be important pollutants due to
their environmental prevalence and toxicity to living organ-
isms. Many studies have demonstrated the proof-of-concept
that whole-cell bacterial biosensors can be used to assess the
bioavailability and toxicity of metals in various sample matri-
ces (Hynninen and Virta 2010; Xu et al. 2013). Despite this,
biosensors are not yet widely used as an environmental mon-
itoring strategy. In addition, contaminated sites often contain
multiple heavy metals; yet, studies that have simultaneously
examined the presence of multiple heavy metals in environ-
mental samples are limited. To examine this potentially

Table 2 Groundwater (GW) samples spiked with both As(III) (500 μg/L) and Hg(II) (20 μg/L) measured using the lux-based biosensors
and the gfp-based biosensors

Recombinant plasmid Asa Hga

GW #1 (μg/L) GW #2 (μg/L) GW #1 (μg/L) GW #2 (μg/L)

luxCDABE
pAs-lux Ob (424±179) O (525±157)
pHg-lux O (18±11) O (6±3)
gfp
pVLAS1 O (426±98)c O (418±93)
pHg-gfp Xb X
a
The groundwater quality standards (for nondrinking purposes) in Taiwan for As and Hg are 500 and 20 μg/L, respectively
b
The bacterial biosensors that produced a significant signal in the As and Hg-spiked groundwater (vs nonspiked groundwater) are marked with O; X not
significant
c
The numbers in the parentheses indicate the concentration of the specific metal calculated by the bacterial biosensor using the derived standard curves
described in BMaterials and methods^
10212
complex situation in environmental samples, a set of bacterial
biosensors with the two different reporter systems, luxCDABE
and gfp, was used to simultaneously detect bioavailable mul-
tiple heavy metals, including As(III) and Hg(II), and the re-
sults of the biosensor assays were compared.
Groundwater samples were collected and used as the envi-
ronmental samples for multiple heavy metal tests. The natural
content of heavy metals in the groundwater samples was very
low. Therefore, to study the usefulness of the bacterial biosen-
sors for multiple heavy metal detection, the groundwater sam-
ples were spiked with heavy metals (As(III) and Hg(II)) to
obtain approximately 500 μg/L As and 20 μg/L Hg. These
concentrations are groundwater quality standards
(nondrinking) in Taiwan.
The bioavailable fraction of the metals was determined
from groundwater samples using the luminescent E. coli
DH5α (pAs-lux) and the fluorescent E. coli DH5α
(pVLAS1) (Liao and Ou 2005) for As(III) and the lumines-
cent E. coli DH5α (pHg-lux) and the fluorescent E. coli
DH5α (pHg-gfp) for Hg(II). A summary of the E. coli
DH5α biosensor constructs is presented in Table 1. A standard
curve was generated with known concentrations of each met-
al, and the resulting equations were used to calculate the
equivalent metal concentrations in the groundwater samples.
The presence and estimated concentrations of As(III)-equiva-
lent and Hg(II)-equivalent metals (As and Hg) in the spiked
groundwater samples are presented in Table 2.
For As detection, both luminescent E. coli DH5α (pAs-lux)
and fluorescent E. coli DH5α (pVLAS1) were able to distin-
guish the As concentration at 500 μg/L, which is the ground-
water quality standard (nondrinking) in Taiwan (Table 2). The
estimated As(III)-equivalent concentrations in the groundwa-
ter samples (GW#1 and GW#2) also compare well to the
spiked As concentration (500 μg/L) (Table 2).
For Hg detection, the luminescent E. coli DH5α (pHg-lux)
biosensor was able to distinguish the Hg concentration at
20 μg/L, which is the groundwater quality standard
(nondrinking) in Taiwan (Table 2). In contrast, the E. coli
DH5α (pHg-gfp) biosensor was unable to distinguish the Hg
concentration at 20 μg/L in the groundwater samples. In ad-
dition, the estimated Hg(II)-equivalent concentration in the
GW#1 groundwater sample also compared well to the spiked
Hg concentration (20 μg/L), whereas the estimated Hg(II)-
equivalent concentration in the GW#2 groundwater sample
was lower than the actual spiked concentration (Table 2). This
result may be due to the inhibitory effects of other chemicals,
in addition to the inducer, on E. coli DH5α (pHg-gfp) in the
GW#2 sample. At contaminated sites, various chemicals other
than the inducer chemicals could be toxic to or interact with
E. coli, thereby causing harmful effects. Because of the com-
plexity of these chemicals, it is impossible to determine the
potential effects of these chemicals, even if the chemicals have
been identified by chemical analyses.
Environ Sci Pollut Res (2015) 22:10206–10213
Conclusions
In summary, to simultaneously determine the bioavailability
of multiple heavy metals, such as As and Hg, in groundwater,
a set of bacterial biosensors with two reporter systems,
luxCDABE and gfp, are likely to provide critical data that
can be useful in environmental monitoring. The induction of
the luxCDABE-based constructs was much more sensitive
than the induction of the gfp-based constructs for the detection
of As(III) and Hg(II).
The choice of a reporter gene for the analysis of metals
must be case-specific. Due to the potential complexity of
chemicals present in environmental samples, using a set of
bacterial biosensors with different reporter genes to simulta-
neously determine the bioavailable portion of heavy metals is
desirable.
Acknowledgments This study was funded in part by the research pro-
ject supported by the Taiwan EPA. The views or opinions expressed in
this article are those of the writers and should not be construed as opinions
of the Taiwan EPA. The mention of trade names, vendor names, or com-
mercial products does not constitute endorsement or recommendation by
the Taiwan EPA.
References
Daunert S, Barrett G, Feliciano JS, Shetty RS, Shrestha S, Smith-Spencer
W (2000) Genetically engineered whole-cell sensing systems: cou-
pling biological recognition with reporter genes. Chem Rev 100:
2705–2738
Dollard MA, Billard P (2003) Whole-cell bacterial sensors for the mon-
itoring of phosphate bioavailability. J Microbiol Methods 55:221–
229
Girotti S, Ferri EN, Fumo MG, Maiolini E (2008) Monitoring of envi-
ronmental pollutants by bioluminescent bacteria. Anal Chim Acta
608:2–29
Hakkila K, Maksimow M, Karp M, Virta M (2002) Reporter genes lucFF,
luxCDABE, gfp, and dsred have different characteristics in whole-
cell bacterial sensors. Anal Biochem 301:235–242
Hakkila K, Green T, Leskinen P, Ivask A, Marks R, Virta M (2004)
Detection of bioavailable heavy metals in EILATox-Oregon samples
using whole-cell luminescent bacterial sensors in suspension or
immobilized onto fibre-optic tips. J Appl Toxicol 24:333–342
Hansen LH, Sørensen SJ (2000) Versatile biosensor vectors for detection
and quantification of mercury. FEMS Microbiol Lett 193:123–127
Hynninen A, Virta M (2010) Whole-cell bioreporters for the detection of
bioavailable metals. Adv Biochem Eng Biotechnol 118:31–63
Ivask A, Green T, Polyak B, Mor A, Kahru A, Virta M, Marks R (2007)
Fibre-optic bacterial biosensors and their application for the analysis
of bioavailable Hg and As in soils and sediments from Aznalcollar
mining area in Spain. Biosens Bioelectron 22:1396–1402
Li YF, Li FY, Ho CL, Liao VH (2008) Construction and comparison of
fluorescence and bioluminescence bacterial biosensors for the detec-
tion of bioavailable toluene and related compounds. Environ Pollut
152:123–129
Liao VH, Ou KL (2005) Development and testing of a green fluorescent
protein-based bacterial biosensor for measuring bioavailable arsenic
in contaminated groundwater samples. Environ Toxicol Chem 24:
1624–1631
Environ Sci Pollut Res (2015) 22:10206–10213
Liao VH, Chien MT, Tseng YY, Ou KL (2006) Assessment of heavy
metal bioavailability in contaminated sediments and soils using
green fluorescent protein-based bacterial biosensors. Environ
Pollut 142:17–23
Miller WG, Leveau JH, Lindow SE (2000) Improved gfp and inaZ broad-
host-range promoter-probe vectors. Mol Plant Microbe Interact 13:
1243–1250
Priyadarshi H, Alam A, Gireesh-Babu P, Das R, Kishore P,
Kumar S, Chaudhari A (2012) A GFP-based bacterial biosen-
sor with chromosomally integrated sensing cassette for quan-
titative detection of Hg(II) in environment. J Environ Sci
(China) 24:963–968
Ramanathan S, Shi W, Rosen BP, Daunert S (1997) Sensing
antimonite and arsenite at the subattomole level with geneti-
cally engineered bioluminescent bacteria. Anal Chem 69:
3380–3384
10213
Rogowsky PM, Close TJ, Chimera JA, Shaw JJ, Kado CI (1987)
Regulation of the vir genes of Agrobacterium tumefaciens plasmid
pTiC58. J Bacteriol 169:5101–5112
Sharma P, Asad S, Ali A (2013) Bioluminescent bioreporter for assess-
ment of arsenic contamination in water samples of India. J Biosci
38:251–258
Stocker J, Balluch D, Gsell M, Harms H, Feliciano J, Daunert S, Malik
KA, van der Meer JR (2003) Development of a set of simple bacte-
rial biosensors for quantitative and rapid measurements of arsenite
and arsenate in potable water. Environ Sci Technol 37:4743–4750
Tauriainen S, Karp M, Chang W, Virta M (1997) Recombinant lumines-
cent bacteria for measuring bioavailable arsenite and antimonite.
Appl Environ Microbiol 63:4456–4461
Xu T, Close DM, Sayler GS, Ripp SA (2013) Genetically modified
whole-cell bioreporters for environmental assessment. Ecol Indic
28:125–141