ARTICLE
pubs.acs.org/est
Assessing the Toxicity of Naphthenic Acids Using a Microbial Genome
Wide Live Cell Reporter Array System
Xiaowei Zhang,*,† Steve Wiseman,‡ Hongxia Yu,† Hongling Liu,† John P. Giesy,†,‡,||,^,#,r,O and
Markus Hecker§
†
State Key Laboratory of Pollution Control and Resource Reuse & School of the Environment, Nanjing University, Nanjing, China
Toxicology Centre, University of Saskatchewan, Saskatoon, Canada
§
ENTRIX, Inc., Saskatoon, Saskatchewan
Department of Veterinary, Biomedical Sciences, University of Saskatchewan, Saskatoon, Canada
^
Department of Biology and Chemistry, City University of Hong Kong, Kowloon, Hong Kong SAR, China
#
School of Biological Sciences, the University of Hong Kong, Hong Kong SAR, China
r
Department of Zoology, and Center for Integrative Toxicology, Michigan State University, East Lansing, Michigan, USA
O
Zoology Department, College of Science, King Saud University, P.O. Box 2455, Riyadh 11451, Saudi Arabia
)
‡
bS Supporting Information
ABSTRACT: Mixtures of naphthenic acids (NAs), which include cyclopentyl and cyclohexyl carboxylic acids, have been suggested
to be toxic components in oils spills, effluents from the petrochemical industry and in oil sands process waters (OSPW). The present
study demonstrated, for the first time, an application of a high throughput live bacterial cell array in a genome-scale investigation of
the toxic mechanisms of environmental chemicals, a commercial NAs technical mixture extracted from crude oil. Real time gene
profiling of time- and concentration- dependent responses of live cells exposed to NAs for three hours was conducted using a library
of 1800 fluorescent transcriptional reporters for Escherichia coli (E. coli) growing in 384-well plates. The response patterns obtained
after exposure to NAs suggested that the primary cellular responses were up-regulation of genes in the pentose phosphate pathway,
involved in the molecular function of NADP or NADPH binding, and down-regulation of the ATP-binding cassette (ABC)
transporter complex. Transcriptional networks that were significantly modulated by NAs included those that were regulated by
transcriptional factors such as CRP-, RecA-, and GadE. Down-regulation of the SOS response pathway suggested that DNA damage
might not be the direct result of NAs within the first three hours of exposure. However, CRP-dependent genes modulated by
exposure to NAs indicated that the cellular level of cyclic AMP was altered immediately upon exposure of cells to NAs. Furthermore,
the linear range of the concentration-response curve of the selected gene reporters encompassed a range of concentrations between
10 and 1000 mg NAs/L, which covers concentrations typically observed in the environment and makes this assay system ideal for the
detection of environmental NAs.
’ INTRODUCTION
Naphthenic acids (NAs) are a group of alkyl-substituted
saturated cyclic and noncyclic carboxylic acids that have been
identified as naturally occurring toxic components in effluents
discharged from offshore oil production platforms in North Sea of
Europe and in the oil sands process-affected water (OSPW)
discharged in Western Canada.1-3 Because NAs can account for
up to 4% of raw petroleum by weight4 and are more soluble in water
than alkanes, alkenes, and polycyclic aromatic hydrocarbons (PAH),
they are suspected to be one of the most toxic components in
deepwater-oils spills, such as the spill that occurred in the Gulf of
Mexico beginning on April 20, 2010. Such large-scale releases of
NAs increase public and government concerns about the risk to
marine and freshwater organisms.
Naphthenic acids range in molecular weight from approximately 120 to over 700 atomic mass units. The main fractions of
NAs are carboxylic acids with a carbon skeleton of 9 to 20
carbons. The composition of these mixtures differs with the crude
oil composition, the conditions during refining and oxidation, and
r 2011 American Chemical Society
changes over time as the result of environmental weathering.4 The
complex nature of the NA mixture presents challenges for their
toxicological characterization.3
NAs have been reported to be acutely toxic to various
organisms. However, the critical mechanism of toxicity remains
largely unknown. NAs are persistent in aquatic environments and
are acutely toxic to aquatic bacteria, invertebrates, fish, and
plants.5 NAs cause acute toxicity in the Vibrio fischeri bacterial
bioluminescence assay (Microtox) with a half-maximum effect
concentration (EC50) range of 42 mg/L to 65 mg/L as a function
of molecular weight.6 Exposure to a commercial mixture of NAs
resulted in an increase in the incidence of deformity and a decrease
in length at hatch of yellow perch (Perca flavescens) and Japanese
medaka (Oryzias latipes).7 Narcosis has been suggested to be the
Received: September 27, 2010
Accepted: January 11, 2011
Revised:
December 27, 2010
Published: February 10, 2011
1984
dx.doi.org/10.1021/es1032579 | Environ. Sci. Technol. 2011, 45, 1984–1991
Environmental Science & Technology
probable mode of acute toxicity by NAs, particularly for lower
molecular weight (MW) NAs. Higher MW NAs are less acutely
toxic than lower MW NAs. Toxicity of NAs is inversely proportional
to carboxylic acid content within NA structures of higher MW NAs.1
Toxicity of NAs is also related to the amount of NA that can be
accumulated into the organisms as well as their inherent toxic
potency. Recent studies also found other mechanisms of toxicity of
NAs. NAs extracted from North Sea offshore produced water
discharges were shown to be estrogen receptor (ER) agonists and
androgen receptor (AR) antagonists in vitro.2 Furthermore, NAs
from OSPW have been found to modulate sex steroid production in
H295R Steroidogenesis Assay by decreasing testosterone (T) and
increasing 17β-estradiol (E2) concentrations.8
In general, the complex and changing nature of mixtures of
NAs make it difficult to predict toxicity. By determining the
critical mechanism of toxicity of NAs, it might be possible to
develop more effective predictive relationships to account for the
toxic effects observed in living organisms exposed to NAs. By
examining the specific pathways affected by NAs, it might also be
possible to develop diagnostic biomarkers of exposure that can be
applied in monitoring programs.
Gene expression analysis represents a powerful approach to
investigate the molecular and cellular responses of organisms to
chemicals and other environmental stressors.9,10 Conventional
transcriptional profiling techniques, such as microarray and real
time PCR, can simultaneously examine the mRNA expression of
multiple genes or the whole transcriptome expressed within a
cell, tissue, or whole organism, which can help elucidating
mechanisms of toxicity. Recent developments in live cell array
(LCA) technology have demonstrated LCA is an accurate and
versatile method of determining gene expression in bacteria in
a high-throughput format.11 Utilizing transcriptional fusions
of specific promoters with fast-folding fluorescent proteins,
such as GFP, this technology allows real time monitoring of gene
expression at successive time points during exposure. The approach
differs from cDNA microarrays which report mRNA concentration
as a balance of mRNA production and degradation. In LCA,
accumulation of GFP fluorescence in live cells represents a readily
measurable protein product of gene expression. The two most
significant advantages of the LCA approach over complementary transcriptomic methodologies are that 1) the cost is
significantly less because no sample preparation is required
prior to measurement, 2) simplified assay procedures are applied,
and 3) the kinetics of gene expression in response to particular
stimuli can be established in high resolution.
Recently, there has been increasing application of microbial
cell-based arrays to detect toxicity of environmental chemical(s)
or classify them based on different modes of action.12-14 These
microbial array systems consist of genetically engineered microorganisms tailored to respond in a dose-dependent manner to
environmental stimuli. The fusion of stress promoters to reporter
genes, such as bioluminescence or fluorescence, is the basis for
the approach for determining which genes are differentially expressed during responses to toxicants.12,13 The combination of
whole cell bioreporter assays and recent development in array
technologies are the basis for promising tools for both assessment
of toxicity and elucidating potential mechanisms of toxic action.15,16
However, the current applications are normally limited to several or
a group of well-established stress-responsive promoters.14,15
The present study investigated possible mechanisms of NAinduced effects on living cells by use of a genome-scale examination
of gene expression which utilizes fluorescent transcriptional repor-
ARTICLE
ters for Escherichia coli (E. coli). A high throughput toxicity test
approach was developed by integrating a 384-well assay format with
a library of fluorescent transcription fusions that includes over 75%
of the promoters of the bacterium E. coli.11 Gene expression was
represented by the transcriptional activity of its promoter, which was
measured by the accumulation of the GFP fluorescence. For these
genes which were from the same multigene operon, their gene
expressions were measured by the same promoter construct. The
results of these assays are applicable to not only bacteria but also
to metazoans due to the fact that many common pathways are
conserved from bacteria to vertebrates such as stress responsive
pathways.14,15 Consequently, this assay is not bacteria-specific and
can represent responses of systems that are conserved in multiple
organisms, including metazoans.
’ MATERIALS AND METHODS
Chemicals. A technical mixture of naphthenic acids was purchased from Sigma Aldrich (#70340, St. Louis, MO, USA). A stock
solution (20 mg/mL) was prepared by dissolving 1.0 g NAs in 50 mL
of 0.08 N NaOH. Other stock solutions of 2.0 and 0.20 mg NAs/mL
were made by serial dilution with nanopure water. Stock solutions
were then further diluted 20 times in microbial culture medium prior
to initiation of the assay. The final pH values of the NAs exposure
microbial culture medium were measured at the beginning of
exposure and were confirmed to be 7.4 ( 0.2.
Microbial Gene Profiling System. The microbial gene reporter collection acquired from Open Biosystems Thermo Fisher
Scientific (Huntsville, AL, USA) was developed by researchers at
the Weizmann Institute of Science and includes more than 1900
promoters (out of 2500 in the entire genome) for the E. coli K12
strain MG1655.11 Each of the reporter strains has a bright, fastfolding green fluorescent protein (GFP) fused to a full-length
copy of an E. coli promoter in a low-copy plasmid which enables
measurement of gene expression within minutes with high
accuracy and reproducibility.
All clones were grown at 37 C in 2X LB-Lennox (low salt)
media plus 25 μg/mL kanamycin. Assay plates were prepared by
adding 71.25 uL of LB medium to each well in black 384-well
optical bottom plates (NUNC, Rochester, NY, USA). E. coli
strains were inoculated in the 384-well plate from a 96-well stock
plate by disposable replicators (Genetix, San Jose, CA, USA).
Cells were incubated at 37 C for 3.5 h before exposure to NAs.
At the end of the initial incubation, the OD of each well was
measured at 620 nM using a Fluostar OPTIMA microplate reader
(Offenburg, Germany). Then 3.75 μL of nanopure water as a
negative control or NAs stock solutions were added into individual
wells on the 384-well plate to make final concentrations of 0, 10,
100, or 1000 mg/L. GFP intensity of each well was consecutively
monitored every 10 min for three hours.
Data Processing and Statistical Analysis. A linear regression model was applied to assess whether a gene reporter responded significantly to the exposure with NAs. For each gene
reporter strain, a linear model where the response measured as
GFP fluorescence was fitted to a function of time (eq 1)
GFP=OD ¼ Rþβ TimeðhrÞþγ NAsþδ
ðTimeðhrÞ NAsÞ
ð1Þ
in which R is the basal expression; β is the coefficient of time; γ is
coefficient of NAs (mg/L); and δ is the interaction coefficient of
time and NAs. A gene reporter was considered significantly
responsive to NAs when the effect of NAs was dependent on
1985
dx.doi.org/10.1021/es1032579 |Environ. Sci. Technol. 2011, 45, 1984–1991
Environmental Science & Technology
time, i.e. δ was statistically significant. To reduce the rate of false
positives in multiple tests, a p value less than 0.001 was
considered significant. Effects of NAs on gene expression were
expressed as fold-change relative to the corresponding control.
Hierarchical clustering analysis on concentration- and timedependent gene expression patterns was performed for the
selected genes by use of ToxClust,17 which is a method for
evaluating multivariate responses programmed in R (http://
www.r-project.org/). Briefly, the distance between any two
genes was calculated by summing their Manhattan distance of
gene expression at all the concentration vs time combinations.
The advantage of this method, relative to comparing point
estimates, such as the LC50, is that it makes use of the discriminatory
power of all of the data from each concentration and each time
point. Thus, the ToxClust method is more efficient and powerful in
making discriminations between responses to treatments. Dendrograms of genes were calculated by the “complete” agglomeration
method and vertically plotted at the left of the graph. Visualization of
both concentration- and time-dependent gene expression was
implemented by use of gradient graph methods and plotted in
a 3 M (gene number) N (time) matrix format. The (3(i-1):3i, j)
element of the matrix corresponds to the fold change in expression
of the ith gene at the jth time point in the cells exposed to three
different concentrations, 10, 100, or 1000 mg/L, respectively.
The color gradient from left to right displayed the time-dependent
response curves. An R script was written to conduct this analysis and
the code is available upon request.
Pathway Analysis. Gene lists were developed for further
analysis based on statistical significance and 1.5 or 2.0 fold-change
cutoff. Selected genes were analyzed by gene ontology (GO) term
searching by use of ClueGO (v1.3) with default settings.18
Construction of the E. coli K-12 transcriptional regulation
model was done by integrating the transcriptional factor (TF)gene relationship database from RegulonDB (v 6.0).19 Visualization of biological networks and gene expression data was conducted by use of Cytoscape.20 Highly connected network regions
were identified as function modules by use of the cytoscape plug-in
jActiveModules.20 The annotation of E.coli genes was using the
EcoCyc database.21
’ RESULTS AND DISCUSSION
Performance of the High Throughput Fluorescent microbialF Reporter Array. Concentration-dependent responses of
1820 different fluorescent transcriptional reporters in E.coli
exposed to a commercial NAs mixture were assayed in a 384
well high throughput format. The entire data set consists of 1820
strains, four different treatment groups, and 18 different time
points, which resulted in more than 130,000 data points. Eightyone percent of the gene reporter strains in the library (1471
reporter stains out of a total number of 1820 strains) showed
detectable fluorescence greater than background (promoter-less
reporter strains) throughout all time points within the 3 h of
exposure in the control incubations. This is consistent with the
results of previous results that reported 60% of the reporter stains
showed detectable fluorescence greater than background in 96
well format.11
NAs Modulation of the Transcription Activation of 1820 Genes.
NAs modulated expression of E. coli genes in a concentrationdependent manner during a 3 h exposure (Figure 1). Of the
1820 strains, 320 displayed significant changes after exposure to
NAs, compared to controls. Using a 1.5-fold change as a cutoff,
ARTICLE
Figure 1. NAs induced concentration-dependent response in the
number of differentially expressed genes. Venn diagram displaying the
differentially expressed genes selected by 1.5 or 2.0 fold change cutoff at
three different NAs concentrations, 10, 100, and 1000 mg/L.
exposures to 10, 100, or 1000 mg NAs/L significantly altered
the activity of 21, 29, and 80 genes, respectively. Expression of
a total of 83 strains was altered at least 1.5-fold during the three
hour exposure. Of these, 13 strains were responsive to all three
concentrations of NAs. Using 2.0-fold change as a cutoff, exposures to 10, 100, or 1000 mg NAs/L significantly altered the
activity of 3, 9, and 26 genes, respectively, during the same time.
All the strains that responded to 10 mg NAs/L and 8 out of the 9
that responded to 100 mg NAs/L were responsive to 1000 mg
NAs/L as well. Expression of 27 reporter stains changed at least 2
fold during the three hours exposure.
Exposure to NAs resulted in fewer up-regulated gene reporter
stains than down-regulated strains within the three-hour exposure. The 27 gene reporter stains selected by application of a
2-fold cutoff were further classified into an up-regulated group
and a down-regulated group by their time course of concentrationdependent gene expression response to NAs (Figure 2). Twelve
and 15 strains displayed general up-regulation and down-regulation, respectively, when exposed to NAs, relative to the control.
Of the 83 gene reporter stains selected by application of a 1.5-fold
cutoff, 25 and 58 strains were further classified into the up- and
down-regulated groups, respectively (see Figure S1, Supporting Information).
Transcriptional Response in Toxicity Assessment. Transcriptional response of a large number of genes can be effectively
integrated to characterize the concentration-dependent relationship of chemical-induced effects in toxicological studies. In some
applications, transcriptional gene expression profiles are used to
identify concentrations that do not elicit a change in gene
expression, like the concept of a No Observed Transcriptional
Effect Level (NOTEL).22 Two different approaches, adopting
discrete or continuous response variables, have been used to
describe concentration-dependent transcriptional responses.22,23
In the discrete variable approach, the number or the percentage of
genes affected are used to describe the degree of chemical-induced
effects. The number of affected genes is normally selected by
statistical significance and arbitrary fold-change cutoff. The
magnitude of individual gene expression is normally not
considered in this case. In the second approach, a continuous
variable integrating the actual expression level of all the selected
genes is used to differentiate the degree of effect induced by
different concentrations of chemicals. In the latter approach, both
the number of genes changing and the magnitude of that change
are used to indicate the chemical-induced effect. As an example, we
had previously developed a hepatic transcriptional index (HTI) as
1986
dx.doi.org/10.1021/es1032579 |Environ. Sci. Technol. 2011, 45, 1984–1991
Environmental Science & Technology
ARTICLE
Figure 2. Clustering of the concentration- and time-dependent expression of the 27 genes altered at least 2-fold change over background by NAs.
Classification and visualization of the gene expression were derived by use of ToxClust.17 The dissimilarity between genes was calculated by the
Manhantan distance between the gene expression at all the concentration vs time combinations. The fold change of gene expression was indicated by
color gradient, and the time course of expression changes were indicated from left to right. As shown on the top of the figure, time-dependent gene
expression in the three exposed concentrations, 10, 100, and 1000 mg NAs/L, were displayed by the low, middle, and high bands in the rectangle area of
labeled by the gene name.
the sum of log-transformed expression levels of a cluster of hepatic
genes weighted by their principal component factor to represent
the overall expression level of the hepteic estrogenic response.23
The medaka HTI represents values in the quantitative assessment
of chemical-induced effects on reproduction of Japanese medaka
during a short-term exposure. In the present study, both approaches were applied and both effectively depicted the linear
range of the concentration-dependent transcriptional response of
E. coli to NAs (Figure 3). As calculated by the continuous variable
approach, the transcriptional response of the least concentration,
10 mg NAs/L, were 8.4% and 7.6% in the fold change cutoff of
1.5 and 2.0, respectively. However, in the discrete approach, the
transcriptional responses and 10 mg NAs/L were 26.25% and
11.5%, respectively, when the 1.5 and 2.0 fold change cut-offs were
applied, respectively. These results suggested that in the discrete
variable approach, the level of statistical significance and fold change
cutoff could significantly affect the calculated response at lesser
concentrations of NAs, and thus affect calculation of toxicant
threshold values, such as the NOEL and NOAEL. Alternatively,
the response at lesser NA concentrations in the continuous
variable approach is less susceptible to the arbitrary parameter
chosen, which render it the preferable approach to assess
chemical-induced toxicity.
Biological Pathways Involved in NAs Effects. Multiple
biological pathways were identified as being responsive to NAs
during a three hour exposure. The pentose phosphate pathway
is the significant Kyoto Encyclopedia of Genes and Genomes
(KEGG) pathway identified by ClueGO analysis of genes of
E. coli altered by NAs. In this pathway, the transcriptional activity of
gluconate-6-phosphate dehydrogenase (Gnd) and ribosephosphate
isomerize (RpiA) were both up-regulated more than 2-fold by NAs.
The pentose phosphate pathway is a process that generates
nicotinamide-adenine dinucleotide phosphate (NADPH) and pentoses (5-carbon sugars). Generation of NADPH is used in reductive
biosynthesis reactions within cells and to prevent oxidative stress.
NADPH reduces glutathione via glutathione reductase. The molecular function of NADP or NADPH binding was one of the
significant GO terms observed to be altered by NAs. The two
genes, Gnd and ketol-acid reductoisomerase (IlvC), involved
in this function of NADP or NADPH binding were also up-regulated
more than 2-fold. The enzyme Gnd has a high degree of specificity
for NADP. Up-regulation of these genes could increase the capability
1987
dx.doi.org/10.1021/es1032579 |Environ. Sci. Technol. 2011, 45, 1984–1991
Environmental Science & Technology
ARTICLE
Figure 3. Concentration-dependent transcriptional response to NAs. In the discrete variable approach, the number or the percentage of genes affected
was used to describe the degree of chemical-induced effects. In the continuous variable approach, the actual expression level of all the selected genes was
integrated to differentiate the degree of effect induced by different concentrations of chemical. The gene percentage was calculated by dividing the
number of genes or expression magnitude of a gene affected at any concentrations by that in the 1000 mg NAs/L group.
of selective, noncovalently interaction with NADPH and facilitate
generation of NADPH, a coenzyme involved in many redox and
biosynthetic reactions. These responses might be part of basic
adaptive mechanisms in the living organism to compensate for the
cellular oxidative stress resulted from the NAs exposure.
NAs significantly down-regulated some membrane associated
transporter proteins, especially those involved in the ATP-binding
cassette (ABC) transporter complex. The ABC transporter complex was the GO term significantly altered by NAs. Transcriptional
activity of both the ABC super family of dipeptide transport
proteins (DppA) and oligopeptide transport protein (YejA) was
down-regulated more than 1.5 fold during the 3-h exposure. Other
transporter genes that were down-regulated by NAs but are not
annotated by GO term included putative ABC superfamily
(peri_bind) transport protein (YdcS), putative major facilitator superfamily (MFS) transport protein (YcaD), and branched chain amino acid transporter (BrnQ). ABC transporters
are trans-membrane proteins that utilize energy from hydrolysis of adenosine triphosphate (ATP) to perform certain biological processes, including translocation of various substrates across
membranes and nontransport-related processes such as translation of RNA and repair of DNA.24 Inhibition of these transporter proteins might be an adaptive response to NAs that
reduces usage of energy or might slow down the translocation
of substrates across membranes and other processes. ABC
transporters are also involved in bacterial multidrug resistance.
However, NAs also caused more than 2-fold up-regulation of the
transcriptions of two transport proteins, arabinose ABC transporter (araF) and a MFS family membrane subunit of hexose phosphate transport protein (UhpT). The AraF protein is an ATPdriven transporter involved in arabinose uptake. The UhpT can
enable the cell to acquire phosphorylated sugars from its environment that can be used as carbon and/or energy sources.25 In
S. Typhi, uhpA and uhpB were up-regulated at early stage of an
osmotic upshift stress.
Exposure to NAs also led to inhibition of other molecular
functions. For example, genes with GO molecular function of
intramolecular transferase activity were down-regulated by NAs.
This includes the membrane-associated D-ribose high-affinity
transport system protein (RbsD) and the putative ribosomal large
subunit pseudouridine synthase (RluE). Intramolecular transferase
is responsible for catalysis of the transfer of a functional group from
one position to another within a single molecule.26 Down-regulation
of expression of these two genes might be an adaptive response in
the exposure to NAs to reduce the usage of energy.
Stress Responsive Pathway Affected by NAs Exposure.
NAs elicited transcriptional alteration of a group of stress
responsive genes, which can be divided into three categories:
redox-response, SOS-response, and osmotic-response (see Table
S1, Supporting Information). Transcriptional repressor MarR
and superoxide dismutase/manganese (SodA) were the two
redox responsive genes that have been altered over 2-fold
compared to the controls. MarR participates in controlling
several genes involved in resistance to antibiotics, multidrug
efflux, oxidative stress, organic solvents, and heavy metals.27
The SOS-response genes, ybfE and ssb, were down-regulated
more than 2-fold by NAs. YbfE and ssb are up-regulated for
DNA repair when cells are exposed to DNA-damaging agents
such as mitomycin C.28 However, both genes were down-regulated
by NAs during the three-hour exposure. Other general stress genes
including ybaY, yjbJ, bola, and inaA, are known to be responsive to
osmotic stress.
Transcriptional Networks Involved in NAs-Induced Effects.
NAs modulated genes are predominately regulated through
several transcriptional factors. Of the 320 genes significantly
modulated by NAs (p < 0.001), 32 can be directly regulated by transcriptional factor cAMP receptor protein (CRP)
(Figure 4). Transcriptional regulation of CRP-dependent genes
requires the binding of cAMP and CRP protein to DNA. Genes
that are activated by cAMP-CRP can be grouped into two categories. The first category are the CRP-dependent genes, which
require one cAMP-CRP for activation, while genes in the second
category require multiple activator molecules in which two or more
CAP dimers or one CRP dimer and additional activator proteins
synergistically activate transcription.29 CRP represses transcription
by promoter occlusion in which an activator protein is excluded by
cAMP-CRP through the interaction with a repressor protein in an
antiactivation mechanism or by hindering promoter clearance.
Furthermore, transcriptional factors UxuR and exuR, which are
subject to catabolic repression in the presence of glucose and at low
levels of cyclic AMP, were also down regulated by NAs. Together
with the CRP-dependent genes that were modulated by NAs, these
results suggested that exposure to NAs changes levels of cellular
cAMP. Increased activity of transcriptional repressor lexA might be
related to down-regulation of 8 different genes involved in the
cellular response to DNA damage or inhibition of DNA replication,
1988
dx.doi.org/10.1021/es1032579 |Environ. Sci. Technol. 2011, 45, 1984–1991
Environmental Science & Technology
ARTICLE
Figure 4. Active functional modules of a transcriptional network of patters of gene response in E. coli exposed to NAs. Each gene is displayed by circular
node, and the transcriptional factor (TF)-target gene interaction is indicated by arrow edge. The level of gene expression in cells exposed to 1000 mg
NAs/L is indicated by color gradient. Deep red: >2 fold up regulation; gradient from red to white: from 2 to1 fold up regulation; gradient from white to
blue: from 1 to 2 fold down regulation; gray: >2 fold down regulation. For the three TFs (crp, lexA, and gadE) that displayed no significant change in
response to NAs, their roles in the network modules were highlighted in aquamarine.
including ybfE, ssb, uvrA, rpsU recA, ftsQ, dinG, and ftsK. LexA blocks
access of RNA polymerase to target promoters to repress their
transcription. When DNA is damaged, the association of LexA and
its DNA targets will be disrupted by the RecA coprotease, and the
SOS regulon will be induced for repair of broken DNA.28 Although
the actual mechanism of how NAs caused down-regulation of the
SOS response pathway is unclear, DNA damage might not be the
direct result of NAs exposure for three hours. Three transcriptional
regulators of the so-called “acid resistance system” were also altered
by NAs include GadE, GadX, and GadW.30 The transcriptional
activator GadE, glutamic acid decarboxylase, is positively autoregulated and controls the transcription of genes involved in the maintenance of pH homeostasis. GadE also controls expression of two
transcription factors related to acid resistance, GadW and GadX, and
for this reason it is considered the central activator of the acid
response system. NAs did not alter the pH of the culture medium in
this study. However, since the NAs stock solution was neutralized
with high concentration of NaOH, the presence of sodium ion in
the NAs stock solution might be related to the response of the
GadE pathway.
Potential Biosensors for Environmental NAs Detection.
While there are pathways in vertebrates with no analogs in
bacteria, many common pathways are conserved in both bacteria
and vertebrates, thus the assay is applicable to not only bacteria
but can be used to investigate chemical-induced mechanisms of
cellular signaling. Furthermore, the NA-responsive gene reporter
strains selected in this study are ideal candidates for the biodetection of NAs in waters. The fluorescent gene reporter library not
only presents a powerful tool for toxicogenomic investigations of
chemical-induced effects at the cellular level but also allows
screening of potential chemical-specific response patterns that
can be developed into specific biosensors for screening purposes.
The specific combination of gene reporter stains that were
responsive to NAs could provide real-time functional monitoring
of the presence of NAs in environment. Genes that responded to
NAs included some so-called “general stress response” genes that
have been previously characterized in environmental stress related
exposure studies. However, changes in the expression of genes
responding to NAs were different than the pattern or responses
typically obtained after exposure to metals or genotoxicants.15
1989
dx.doi.org/10.1021/es1032579 |Environ. Sci. Technol. 2011, 45, 1984–1991
Environmental Science & Technology
Also, those genes that had not been previously examined in stress
exposure studies, such as uhpT and araF, could augment the
unique signature for exposure to NAs. Furthermore, the selected
library of reporter gene strains could offer desirable sensitivity for
on-site detection of environmental NAs. The concentrations of
NAs in waters affected by OSPW in Alberta, Canada, have
historically ranged between 20 and 100 mg/L, including successfully remediated water.31 The linear range of concentration
response curve of the selected gene reporters is well within the
concentration between 10-1000 mg/L (Figure 3), which makes
them ideal biosensor candidates for the detection of environmental NAs. Also, further work to demonstrate the reproducibility of the profiles, particularly in environmental samples
containing a similar mixture of NAs would be useful. Finally,
responses in the microbial library need to be calibrated to
ecologically relevant responses in economically important or
ecologically relevant species.
’ ASSOCIATED CONTENT
bS
Supporting Information. Table of differentially expressed
genes and figure of gene clustering analysis. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: 86-25-83593649. Fax: 86-25-83707304. E-mail:
howard50003250@yahoo.com. Corresponding author address:
School of the Environment, Nanjing University, Nanjing 210089,
China.
’ ACKNOWLEDGMENT
The research was supported by a grant from Major State Basic
Research Development Program (No. 2008CB418102) and a
grant from Nanjing University Talent Development Foundation.
This project was also supported by a grant from the Western
Economic Diversification Canada (Projects # 6578 and 6807).
The authors wish to acknowledge the support of an instrumentation grant from the Canada Foundation for Infrastructure. Prof.
Giesy was supported by the Canada Research Chair program and
an at large Chair Professorship at the Department of Biology and
Chemistry and State Key Laboratory in Marine Pollution, City
University of Hong Kong, the Einstein Professor Program of the
Chinese Academy of Sciences and the Distinguished Visiting
Professor Program of King Saud University.
’ REFERENCES
(1) Frank, R. A.; Fischer, K.; Kavanagh, R.; Burnison, B. K.;
Arsenault, G.; Headley, J. V.; Peru, K. M.; Van Der, K. G.; Solomon,
K. R. Effect of carboxylic acid content on the acute toxicity of oil sands
naphthenic acids. Environ. Sci. Technol. 2009, 43 (2), 266–271.
(2) Thomas, K. V.; Langford, K.; Petersen, K.; Smith, A. J.; Tollefsen,
K. E. Effect-directed identification of naphthenic acids as important in
vitro xeno-estrogens and anti-androgens in North sea offshore produced
water discharges. Environ. Sci. Technol. 2009, 43 (21), 8066–8071.
(3) Giesy, J. P.; Anderson, J. C.; Wiseman, S. B. Alberta oil sands
development. Proc. Natl. Acad. Sci. U. S. A. 2010, 107 (3), 951–952.
(4) Headley, J. V.; McMartin, D. W. A review of the occurrence
and fate of naphthenic acids in aquatic environments. J. Environ. Sci.
Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 2004, 39 (8),
1989–2010.
ARTICLE
(5) Clemente, J. S.; Fedorak, P. M. A review of the occurrence,
analyses, toxicity, and biodegradation of naphthenic acids. Chemosphere
2005, 60 (5), 585–600.
(6) Frank, R. A.; Kavanagh, R.; Kent, B. B.; Arsenault, G.; Headley,
J. V.; Peru, K. M.; Van Der, K. G.; Solomon, K. R. Toxicity assessment of
collected fractions from an extracted naphthenic acid mixture. Chemosphere 2008, 72 (9), 1309–1314.
(7) Peters, L. E.; MacKinnon, M.; Van, M. T.; van den Heuvel, M. R.;
Dixon, D. G. Effects of oil sands process-affected waters and naphthenic
acids on yellow perch (Perca flavescens) and Japanese medaka (Orizias
latipes) embryonic development. Chemosphere 2007, 67 (11), 2177–2183.
(8) He, Y.; Wiseman, S. B.; Zhang, X.; Hecker, M.; Jones, P. D.;
El-Din, M. G.; Martin, J. W.; Giesy, J. P. Ozonation attenuates the
steroidogenic disruptive effects of sediment free oil sands process water
in the H295R cell line. Chemosphere 2010, 80 (5), 578–584.
(9) Zhang, X.; Hecker, M.; Park, J. W.; Tompsett, A. R.; Jones, P. D.;
Newsted, J.; Au, D. W.; Kong, R.; Wu, R. S.; Giesy, J. P. Time-dependent
transcriptional profiles of genes of the hypothalamic-pituitary-gonadal
axis in medaka (Oryzias latipes) exposed to fadrozole and 17betatrenbolone. Environ. Toxicol. Chem. 2008, 27 (12), 2504–2511.
(10) Zhang, X.; Yu, R. M.; Jones, P. D.; Lam, G. K.; Newsted, J. L.;
Gracia, T.; Hecker, M.; Hilscherova, K.; Sanderson, T.; Wu, R. S.; et al.
Quantitative RT-PCR methods for evaluating toxicant-induced effects
on steroidogenesis using the H295R cell line. Environ. Sci. Technol. 2005,
39 (8), 2777–2785.
(11) Zaslaver, A.; Bren, A.; Ronen, M.; Itzkovitz, S.; Kikoin, I.;
Shavit, S.; Liebermeister, W.; Surette, M. G.; Alon, U. A comprehensive
library of fluorescent transcriptional reporters for Escherichia coli. Nat.
Methods 2006, 3 (8), 623–628.
(12) Elad, T.; Lee, J. H.; Belkin, S.; Gu, M. B. Microbial Whole-cell
arrays. Microb. Biotechnol. 2008, 1, 137–148.
(13) Belkin, S. Microbial whole-cell sensing systems of environmental
pollutants. Curr. Opin. Microbiol. 2003, 6 (3), 206–212.
(14) Lee, J. H.; Mitchell, R. J.; Kim, B. C.; Cullen, D. C.; Gu, M. B. A
cell array biosensor for environmental toxicity analysis. Biosens. Bioelectron. 2005, 21 (3), 500–507.
(15) Onnis-Hayden, A.; Weng, H.; He, M.; Hansen, S.; Ilyin, V.;
Lewis, K.; Guc, A. Z. Prokaryotic real-time gene expression profiling for
toxicity assessment. Environ. Sci. Technol. 2009, 43 (12), 4574–4581.
(16) Nobels, I.; Dardenne, F.; Coen, W. D.; Blust, R. Application of a
multiple endpoint bacterial reporter assay to evaluate toxicological
relevant endpoints of perfluorinated compounds with different functional groups and varying chain length. Toxicol. in Vitro 2010in press.
(17) Zhang, X.; Newsted, J. L.; Hecker, M.; Higley, E. B.; Jones,
P. D.; Giesy, J. P. Classification of chemicals based on concentrationdependent toxicological data using ToxClust. Environ. Sci. Technol. 2009,
43 (10), 3926–3932.
(18) Bindea, G.; Mlecnik, B.; Hackl, H.; Charoentong, P.; Tosolini,
M.; Kirilovsky, A.; Fridman, W. H.; Pages, F.; Trajanoski, Z.; Galon, J.
ClueGO: a Cytoscape plug-in to decipher functionally grouped gene
ontology and pathway annotation networks. Bioinformatics 2009, 25 (8),
1091–1093.
(19) Gama-Castro, S.; Jimenez-Jacinto, V.; Peralta-Gil, M.; SantosZavaleta, A.; Penaloza-Spinola, M. I.; Contreras-Moreira, B.; Segura-Salazar,
J.; Muniz-Rascado, L.; Martinez-Flores, I.; Salgado, H.; Bonavides-Martinez,
C.; et al. RegulonDB (version 6.0): gene regulation model of Escherichia
coli K-12 beyond transcription, active (experimental) annotated promoters
and Textpresso navigation. Nucleic Acids Res. 2008, 36 (Database issue),
D120–D124.
(20) Cline, M. S.; Smoot, M.; Cerami, E.; Kuchinsky, A.; Landys, N.;
Workman, C.; Christmas, R.; Avila-Campilo, I.; Creech, M.; Gross, B.
et al. Integration of biological networks and gene expression data using
Cytoscape. Nat. Protoc. 2007, 2 (10), 2366–2382.
(21) Karp, P. D.; Keseler, I. M.; Shearer, A.; Latendresse, M.;
Krummenacker, M.; Paley, S. M.; Paulsen, I.; Collado-Vides, J.; GamaCastro, S.; Peralta-Gil, M.; et al. Multidimensional annotation of the
Escherichia coli K-12 genome. Nucleic Acids Res. 2007, 35 (22), 7577–
7590.
1990
dx.doi.org/10.1021/es1032579 |Environ. Sci. Technol. 2011, 45, 1984–1991
Environmental Science & Technology
ARTICLE
(22) Lobenhofer, E. K.; Cui, X.; Bennett, L.; Cable, P. L.; Merrick,
B. A.; Churchill, G. A.; Afshari, C. A. Exploration of low-dose estrogen
effects: identification of No Observed Transcriptional Effect Level
(NOTEL). Toxicol. Pathol. 2004, 32 (4), 482–492.
(23) Zhang, X.; Hecker, M.; Tompsett, A. R.; Park, J. W.; Jones,
P. D.; Newsted, J.; Au, D.; Kong, R.; Wu, R. S.; Giesy, J. P. Responses of
the medaka HPG axis PCR array and reproduction to prochloraz and
ketoconazole. Environ. Sci. Technol. 2008, 42 (17), 6762–6769.
(24) Saurin, W.; Hofnung, M.; Dassa, E. Getting in or out: early
segregation between importers and exporters in the evolution of ATPbinding cassette (ABC) transporters. J. Mol. Evol. 1999, 48 (1), 22–41.
(25) Ma, S.; Selvaraj, U.; Ohman, D. E.; Quarless, R.; Hassett, D. J.;
Wozniak, D. J. Phosphorylation-independent activity of the response
regulators AlgB and AlgR in promoting alginate biosynthesis in mucoid
Pseudomonas aeruginosa. J. Bacteriol. 1998, 180 (4), 956–968.
(26) Pan, H.; Ho, J. D.; Stroud, R. M.; Finer-Moore, J. The crystal
structure of E. coli rRNA pseudouridine synthase RluE. J. Mol. Biol.
2007, 367 (5), 1459–1470.
(27) Alekshun, M. N.; Levy, S. B. The mar regulon: multiple
resistance to antibiotics and other toxic chemicals. Trends Microbiol.
1999, 7 (10), 410–413.
(28) Fernandez De Henestrosa, A. R.; Ogi, T.; Aoyagi, S.; Chafin, D.;
Hayes, J. J.; Ohmori, H.; Woodgate, R. Identification of additional genes
belonging to the LexA regulon in Escherichia coli. Mol. Microbiol. 2000,
35 (6), 1560–1572.
(29) Busby, S.; Ebright, R. H. Transcription activation by catabolite
activator protein (CAP). J. Mol. Biol. 1999, 293 (2), 199–213.
(30) Hommais, F.; Krin, E.; Coppee, J. Y.; Lacroix, C.; Yeramian, E.;
Danchin, A.; Bertin, P. GadE (YhiE): a novel activator involved in the
response to acid environment in Escherichia coli. Microbiology 2004,
150 (Pt 1), 61–72.
(31) Quagraine, E. K.; Peterson, H. G.; Headley, J. V. In situ
bioremediation of naphthenic acids contaminated tailing pond waters
in the athabasca oil sands region--demonstrated field studies and
plausible options: a review. J. Environ. Sci. Health, Part A: Toxic/Hazard.
Subst. Environ. Eng. 2005, 40 (3), 685–722.
1991
dx.doi.org/10.1021/es1032579 |Environ. Sci. Technol. 2011, 45, 1984–1991
Supporting Information
Authors:
Xiaowei Zhang†*, Steve Wiseman‡, Hongxia Yu†, Honglin Liu†, John
Giesy†,‡,║,#,††,‡‡,§§, Markus Hecker§
†
State Key Laboratory of Pollution Control and Resource Reuse & School of the
Environment, Nanjing University, Nanjing, China
‡
§
Toxicology Centre, University of Saskatchewan, Saskatoon, Canada
ENTRIX, Inc., Saskatoon, Saskatchewan
║
Department of Veterinary, Biomedical Sciences, University of Saskatchewan,
Saskatoon, Canada
#
Department of Biology and Chemistry, City University of Hong Kong, Kowloon,
Hong Kong SAR, China
††
School of Biological Sciences, the University of Hong Kong, Hong Kong SAR,
China
‡‡
Department of Zoology, and Center for Integrative Toxicology, Michigan State
University, East Lansing, MI, USA
§§
Zoology Department, College of Science, King Saud University, P. O. Box
2455, Riyadh 11451, Saudi Arabia
Title:
Assessing the Toxicity of Naphthenic Acids Using a Microbial Genome Wide
Live Cell Reporter Array System
Page:
3
Table:
1
Figure:
1
Submitted to: Environmental Science and Technology
Tables S1. Microbial genes modulated over 2 fold by the exposure of naphthenic acids (NAs) in the
concentration range of 10-1000 mg NA/L.
Gene
ycaD
dppA
ydcS
araF
brnQ
uhpT
gnd
ilvC
rpiA
marR
sodA
inaA
ybfE
ssb
insA_7
ybaY
yjbJ
bolA
Description
putative MFS family transport protein (1st module)
ABC superfamily (peri_bind) dipeptide transport protein (1st module)
putative ABC superfamily (peri_bind) transport protein
arabinose ABC transporter
LIVCS family, branched chain amino acid transporter system II (LIV-II)
MFS family, hexose phosphate transport protein, possible membrane
subunit (1st module)
gluconate-6-phosphate dehydrogenase, decarboxylating (1st module)
ketol-acid reductoisomerase, NAD(P)-binding
ribosephosphate isomerase, constitutive
transcriptional repressor for antibiotic resistance and oxidative stress
superoxide dismutase, manganese
pH inducible protein involved in stress response, protein kinase-like
LexA regulated, possible SOS response
ssDNA-binding protein controls activity of RecBCD nuclease
IS1 protein InsA
glycoprotein/polysaccharide metabolism
unknown CDS (predicted stress response protein)
activator of morphogenic pathway (BolA family), important in general
stress response
Categories
Transporter
proteins
Generation of
NADPH
Stress response
alaS
ybeB
ymcC
alanyl-tRNA synthetase
conserved hypothetical protein
putative synthetase
hemC
crl
hydroxymethylbilane synthase (porphobilinogen deaminase)
transcriptional regulator of cryptic genes for curli formation and fibronectin
Others
binding
Sugar Specific PTS family, cellobiose/arbutin/salicinsugar specific enzyme
IIA component
putative enzyme (MutT-like)
Unknown
unknown CDS
function
conserved hypothetical protein
chbA
yfcD
ybjP
somA
Ribosome related
proteins
Figure S1. Clustering of the time-dependent expression of the NAs altered genes selected by 1.5fold change cut-off. Gene expression in cells exposed to 1000 mg NAs/L are displayed.
Classification and visualization of the gene expression were derived by use of ToxClust. The
dissimilarity between genes was calculated by the Manhantan distance between the gene
expression observed at all time points.