This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/authorsrights
Transcript
This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies areencouraged to visit:
jou rn al hom ep age: www.elsev ier .com/ locate /aquatox
Transcriptomic responses of European flounder (Platichthys flesus)liver to a brominated flame retardant mixture
Tim D. Williamsa,∗, Amer M. Diabb, Matt Gubbinsc, Catherine Collinsc, Iveta Matejusovac,Rose Kerrc, James K. Chipmana, Raoul Kuiperd, A. Dick Vethaake,f, Stephen G. Georgeg
a School of Biosciences, The University of Birmingham, Birmingham B15 2TT, UKb Glasgow International College, Glasgow University, Glasgow G11 6NU, UKc Marine Scotland Science, Marine Laboratory, 375 Victoria Road, Aberdeen AB11 9DB, UKd Karolinska Institute Phenotyping Core Facility, Department of Laboratory Medicine, FENO, F52,Karolinska University Hospital Huddinge, 141 86 Stockholm, Swedene Deltares, Marine and Coastal Systems, PO Box 177, 2600 MH Delft, The Netherlandsf Institute for Environmental Studies (IVM), VU University, Amsterdam, The Netherlandsg Institute of Aquaculture, The University of Stirling, Stirling FK9 4LA, UK
a r t i c l e i n f o
Article history:Received 24 May 2013Received in revised form 18 July 2013Accepted 19 July 2013
Male European flounder (Platichthys flesus) were exposed to a technical mixture of brominated diphenylethers (PDBEs, DE-71, Pentamix) that had been purified to remove contaminating dioxins. Controls wereexposed to carrier solvent alone. Fish were exposed to decadally increasing concentrations of Pentamixvia both sediment and spiked food. The GENIPOL P. flesus cDNA microarray, differentially expressedgene profiling (DEG) and quantitative PCR were employed to detect hepatic transcriptional differencesbetween exposed fish and controls. Gene transcriptional changes were more sensitive to Pentamix expo-sure than biomarkers measured previously. Pentamix exposure induced transcripts coding for enzymesof xenobiotic metabolism (CYP1A, aldo-keto reductases) and elicited endocrine disruption (vitellogeninand thyroid hormone receptor alpha), with effects on CYP1A and VTG occurring at the highest exposure.Ontology analysis clearly showed dose-responsive changes indicative of oxidative stress, induction ofmitochondrial dysfunction, and apoptosis. We conclude that exposure to PBDEs in both sediment andfood has a significant adverse effect on a broad range of crucial biochemical processes in the livers of thiswidely distributed estuarine fish species, the flounder.
The polybrominated diphenylethers (PBDEs) are brominatedflame retardants (BFRs), formerly widely used to reduce theflammability of furniture, textiles and electronic equipment. Com-mercial PBDE mixtures are classified according to the degree ofbromination; the ‘Pentamix’ (DE-71) contains mainly tetra- andpenta-BDEs. Over the past 20 years there has been progressivecontamination of the aquatic environment and bioaccumulationof lower brominated congeners in aquatic biota at various trophiclevels (Boon et al., 2002a,b; Birnbaum and Staskal, 2004). The mar-keting and use of Pentamix has been banned in the European Unionsince 2004 (Kemmlein et al., 2009), but the lower brominated con-geners are resistant to degradation, thus are likely to persist inthe environment and pose an ongoing risk to aquatic organisms,particularly demersal fish.
Reported effects of acute PDBE exposure in mammalian sys-tems include interactions with sex steroid and thyroid endocrinesystems (Hamers et al., 2006; Pacyniak et al., 2007; Fery et al.,2009) and oxidative stress leading to liver damage (Albina et al.,2010; Bruchajzer et al., 2010), genotoxicity (Barber et al., 2006;Ji et al., 2011) and neurotoxicity (Branchi et al., 2003; Blancoet al., 2011). PBDE exposure has been associated with numeroustranscriptional changes in mammals (Suvorov and Takser, 2010;Dunnick et al., 2012) and fish (Olsvik et al., 2009; Chen et al., 2010;Han et al., 2011; Softeland et al., 2011; Chan and Chan, 2012).Evidence for alteration of cytochrome P450 1A activity is mixed(Holm et al., 1994; Boon et al., 2002a,b; Chen and Bunce, 2003;Olsvik et al., 2009; Chen et al., 2010; Wahl et al., 2010; Softelandet al., 2011) and may be attributable to dioxin contamination ofthe commercial mixtures since purification of commercial Pen-tamix (DE-71) to remove planar components resulted in a lackof DR-CALUX response and only weak CYP1A immunoreactivityin exposed zebrafish (Danio rerio) gills (Kuiper et al., 2006), indi-cating that the majority of the AhR activation potential of thismixture is caused by non-BDE components. In an in vitro study
a Significantly different from solvent control (P < 0.05), TOC denotes total organic carbon. Adapted from our previous publication, Kuiper et al. (2008).
of a range of BFRs, 2,2′,4,4′-tetra-bromodiphenyl ether (BDE47)was found to be a potent inhibitor of estrogen sulfotransferaseactivity (Hamers et al., 2006). In reporter gene assays, other con-geners found in Pentamix were found to be androgen receptor (AR)antagonists and variably, estrogen receptor (ER) and dioxin recep-tor (DR) agonists or antagonists (Hamers et al., 2006). After 3.5months exposure to a PBDE mixture, spawning success in stick-leback (Gasterosteus aculeatus) was reported to be dramaticallyreduced (Holm et al., 1993). Toxicity of BFRs was reviewed byBirnbaum and Cohen Hubal (2006) and their endocrine disruptingeffects by Legler (2008).
In mammals, PBDEs can be metabolised to hydroxylatedforms (Wang et al., 2012). Incubation of BDE47 with rat livermicrosomes resulted in formation of a range of hydroxy-lated metabolites that exhibited increased transthyretin bindingand estradiol-sulfotransferase inhibition; particularly 3-OH-BDE47and 4′-OH-BDE49 (Hamers et al., 2008). Zhong et al. (2011)reported a study employing human L02 cells in which expo-sure to hydroxylated BDE47 metabolites elevated reactive oxygenspecies (ROS) levels and superoxide dismutase (SOD) activ-ity, decreased glutathione (GSH) levels, induced apoptosis andinhibited cell proliferation. Additional metabolites of PBDEs include2,4-dibromophenol, an uncoupler of oxidative phosphorylation(Hempfling, 1970). After acute exposure of zebrafish fibroblaststo 1 �M hydroxylated BDE47 (6-OH-BDE47), transcripts relatedto carbohydrate metabolism and proton transport were inducedand uncoupling of oxidative phosphorylation elicited acute tox-icity (van Boxtel et al., 2008), however, in medaka fish (Oryziaslatipes), Wan et al. (2010) reported that BDE47 was not metabolisedto 6-OH-BDE47.
Because of their high abundance, bottom-dwelling life style,and susceptibility to adverse environmental conditions, the Euro-pean flounder (Platichthys flesus) is one of the species of choice formonitoring the effects of endocrine disruptors and other chemi-cal contaminants in UK and European estuarine and coastal waters(Kirby et al., 2004; Vethaak et al., 2009). Flounders had beensimultaneously exposed to “dioxin-free” Pentamix contaminatedsediment and contaminated feed for three months, detailed ina previous publication (Kuiper et al., 2008). Briefly, this studyshowed that in flounder muscle tissue, BDE47 was the most abun-dant congener (63 ± 6%) with detectable levels of BDEs 49, 99,100, 153 and 154. The profile of congeners detected in floundermuscle was different from that of the parent Pentamix (DE-71), with a notable decrease in the proportion of BDE99 andincrease in proportion of BDE47, and the profile of congenersdiffered between flounder and zebrafish. Whilst there were nosignificant changes in length, weight, or relative liver and gonadweights of the fish there was significant accumulation of BDE47in flounder muscle at higher exposure concentrations. However,the residue levels were extremely variable with differences of upto 20-fold between individuals within an exposure group. No sig-nificant changes in concentrations of plasma thyroid hormones T3
and T4, gonad aromatase activity or microsomal ethoxyresorufinO-dethylase (EROD), pentoxyresorufin-O-deethylase (PROD) orbenzoxy-resorufin-O-deethylase (BROD) activities were found(Kuiper et al., 2008). Liver tissues from these animals were ana-lysed in the present study to determine if gene transcriptionalchanges could be more sensitive indicators of Pentamix exposureafter chronic sub-toxic exposure, to characterise any changes interms of biological pathways potentially affected and thus to relatetoxicogenomic data directly to both a pollutant and an organism ofhigh environmental relevance.
2. Materials and methods
2.1. Fish exposures
Exposure conditions were described previously by Kuiper et al.(2008). Briefly, 159 days old artificially reared flounders obtainedfrom Manx Mariculture Ltd. (Isle of Man, UK) (48 ± 12 g) were heldin aquaria containing 15 kg of (Pentamix spiked) sediment and160 L of water from the Eastern Scheldt (a tidal bay connected to theNorth Sea with a salinity of approximately 30%); (temperature was15 ◦C; renewal rate was twice weekly via continuous flow-through)for 3 months. Fish were fed three times a week at an estimated 1%of body weight with Pentamix-spiked pellets prepared from fishmeal, mussels and fish oil. Groups of ten animals were exposed toa range of sediment and food Pentamix concentrations as shown inTable 1. For convenience these are subsequently referred to as sol-vent control (SC), untreated sediment control (UC), and 0.007, 0.07,0.7, 7, 70 and 700 �g/g groups. Fish metadata were presented pre-viously (Kuiper et al., 2008). For transcriptional analysis, samplesof liver tissue were frozen at −80 ◦C.
These procedures are more fully detailed in SupplementaryFile 1. Briefly, Total RNA was extracted (RNEasy Tissue Mini Kit;Qiagen, Crawley, UK) from liver tissue samples of all fish (n = 10per group). Equal amounts of RNA from each fish were pooledby exposure group, reverse transcribed to cDNA and amplifiedwith arbitrary primers in a two-step procedure using the Gene-fishing DEG Premix Kit (Biogene, Kimbolton, UK) according to themanufacturers’ instructions. Products were visualised by agarosegel electrophoresis, apparently differentially expressed productswere excised and purified (Geneclean III; Biogene or MinElute GelExtraction; Qiagen), then cloned (TOPO TA cloning kit; Invitro-gen, Paisley, UK), plasmids were purified from individual colonies(QIAprep Spin Miniprep kit; Qiagen) and 5–10 clones from eachproduct were sequenced. Sequences were aligned (Sequencher;Ann Arbour, MI, USA) and putatively identified by BLAST (NCBI).Primers were designed from the sequences obtained, RNA fromindividual fish was reverse transcribed to cDNA and used for SYBR-green real-time QPCR (ABI7000, Applied Biosystems, Carlsbad, CA,
Author's personal copy
T.D. Williams et al. / Aquatic Toxicology 142– 143 (2013) 45– 52 47
USA). Beta-actin expression did not vary significantly over theexperiment. Ct values were normalised to those of beta-actin,relative mRNA expression was compared by the 2-deltadeltaCtmethod (Livak and Schmittgen, 2001). After log 2 transformationof the data, the Shapiro–Wilks test showed that data were not nor-mally distributed, therefore the Kruskal–Wallace non-parametrictest was used to identify significantly changing transcripts and theMann–Whitney U test provided P-values for each exposed group incomparison with the solvent controls, these tests were carried outwithin SPSS (SPSS Statistics 20; IBM, New York, NY, USA). Finallya multiple-testing correction was applied to obtain false discoveryrate corrected P-values (q values) (Pike, 2011).
2.3. Microarray experiments
The GENIPOL European flounder microarray (Williams et al.,2006) was used for determination of differential mRNA expression.This microarray consisted of 13,824 flounder clones and controlsspotted in duplicate onto Corning UltraGAPS slides (Corning Inc.,Corning, NY, USA) with an MGII robot (Biorobotics, Cambridge, UK)at Birmingham University Functional Genomics Laboratory, rep-resenting approximately 3336 unique genes. We have previouslyshown details of clones incorporated into the microarray (Williamset al., 2003, 2006, 2007; Diab et al., 2008). Microarray experimentswere carried out as described previously (Diab et al., 2008). Samplesof five male livers each from the control and 0.07 �g/g, 0.7 �g/g,7 �g/g and 700 �g/g Pentamix-exposed groups were assessed bymicroarray analysis. Total RNA samples from individual fish werereverse transcribed to cDNA then labelled with Cy5-dCTP (GE Life-sciences, Amersham, UK). Equal amounts of each cDNA samplewere pooled and labelled with Cy3-dCTP (GE Lifesciences) for useas a reference sample. Each array consisted of Cy5-labelled cDNAfrom one individual fish hybridised versus the Cy3-labeled pooledreference. Hybridisations were carried out for 18 h, before strin-gent washing and scanning with a Scanarray Express HT (PerkinElmer, Beaconsfield, UK). Validation of the microarray by qPCR hasbeen shown previously (Williams et al., 2006, 2007). Data werecaptured using Genepix software (Molecular Devices, Wokingham,UK), and each slide was checked in detail, with spots showing poormorphology or arrays showing experimental artefacts being dis-carded. MIAME compliant data were submitted to ArrayExpressand was assigned the accession E-MTAB-1255. The data used inanalyses consisted of local background-subtracted median intensi-ties. Two samples were discarded due to microarray hybridizationartefacts.
2.4. Microarray data analyses
Microarray data were normalized using Genespring GX 7.3.1software (Agilent Technologies, Berkshire, UK), median expressionover control was brought into concordance using median polish-ing (Agilent). Only data from spots designated as ‘present’ wereused. Data for low intensity, highly variable spots (control inten-sities <10 in any groups, raw intensities <10 or SD > 1.4 in any 2of 5 groups) were removed. Lists of differentially expressed trans-cripts were generated within TMEV (Saeed et al., 2006) using theStatistical Analysis of Microarrays (SAM) package (Tusher et al.,2001), employing a False Discovery Rate (FDR) cut-off of 0.1, cal-culated by permutation. Enrichment analyses were carried outwithin Blast2GO (Gotz et al., 2008) for all identifiable transcriptsand in DAVID (Huang et al., 2009a,b) for transcripts with identifi-able human orthologs, employing FDR cut-offs of 0.1. In both casesall detected identifiable transcripts were used as a background setfor comparison.
3. Results
3.1. Differentially expressed gene profiling (DEG) and real-timePCR
DEG analysis identified PCR products apparently differentiallyexpressed between exposure groups. These were subsequentlycloned, sequenced and identified by homology as shown in Sup-plementary File 2. Sequences were submitted to dbEST at NCBIand were assigned accession numbers JZ167821 to JZ167872. Themajority of clones were identified as encoding partial sequencesof digestive enzyme transcripts including elastase, chymotrypsin,amylase and carboxypeptidase. Additionally, heat shock cognate 71(HSC71), fructose bisphosphate aldolase, a transmembrane protein4 family gene (TM4SF5), kininogen 1, astacin-like metallopeptidaseand translation elongation factor delta were identified as appar-ently differentially expressed. Primers were designed to amplify aselection of these products from all individual samples of floundercDNA and their expression was analysed by SYBR-green real-time PCR. Additional primers were designed for thyroid hormonereceptor alpha as a representative transcript for thyroid hormonesignalling. Details are shown in Supplementary File 2. The resultsof PCR analysis are shown in Table 2. In general, mRNA expres-sion was highly variable between individual fish and the majorityof transcripts were induced at higher exposure concentrations ofPentamix, the exceptions being HSC71 that was repressed onlywith 0.7 �g/g Pentamix and carboxypeptidase A1 that did not sig-nificantly change in expression. Thyroid hormone receptor alphawas significantly induced (FDR < 0.05) with the 700 �g/g exposureand at FDR < 0.1 above 0.7 �g/g. The digestive enzyme transcripts,TM4SF5, aldolase, astacin-like, kininogen and elongation factor 1delta were all significantly induced with 0.7 �g/g Pentamix expo-sure, although not all were statistically significantly induced athigher concentrations due to high inter-individual variability. Withthe highest concentration of Pentamix, 700 �g/g, the extent ofinduction appeared to have reduced.
3.2. Microarray
SAM analyses of the microarray data resulted in identificationof 719 differentially expressed transcripts with 0.07 �g/g Pentamix,271 with 0.7 �g/g, 379 with 7 �g/g and 86 with 700 �g/g (FDR < 0.1).The mean expression of these transcripts relative to the controlgroup is shown in Supplementary File 3. The mean variance ofmRNA expression between individuals in each group was 0.73 forthe controls, 0.18 for 0.07 �g/g, 0.31 for 0.7 �g/g, 0.32 for 7 �g/gand 2.09 for 700 �g/g, though this was only statistically signifi-cantly different from controls for the 0.07 �g/g group (P < 0.039).In comparison with the real-time PCR data, HSC71 and EEF1Dwere assessed as significantly induced by array at 0.07 �g/g butno other data were statistically significant, whereas by PCR HSC71was repressed at 0.7 �g/g and EEF1D induced above 0.7 �g/g; beta-actin was mildly but significantly 1.31-fold induced at 0.07 �g/g bymicroarray (Supplementary File 3). Other genes assessed by real-time PCR were either not represented on the array or did not resultin statistically significant expression changes. From the microarraydata, four transcripts were significantly differentially expressed 2-fold higher in every exposed group relative to the controls; thesewere aldo-keto reductase AKR7A2, proteasome subunit beta type4, PSMB4 and serine protease HTRA2. No transcripts were signif-icantly expressed 2-fold less than the controls in every exposedgroup. The most highly induced transcript was vitellogenin A, rep-resented by two non-contiguous sequences on the microarray,which was induced above 13-fold in response to 700 �g/g Pen-tamix. This was in contrast to the 0.07–7 �g/g exposures where thistranscript was repressed. Vitellogenin B showed a similar profile of
Author's personal copy
48 T.D. Williams et al. / Aquatic Toxicology 142– 143 (2013) 45– 52
Tab
le
2G
ene
exp
ress
ion
by
real
-tim
e
PCR
show
n
nor
mal
ised
to
the
mea
n
of
the
solv
ent
con
trol
grou
p
±
stan
dar
d
dev
iati
on, w
ith
stat
isti
call
y
sign
ifica
nt
chan
ges
in
bold
typ
e
(FD
R
<
0.1)
or
bold
typ
e
wit
h
aste
risk
(FD
R
<
0.05
).
SC: s
olve
nt
con
trol
, UC
:
un
trea
ted
con
trol
, 0.0
7–70
0
�g/
g
Pen
tam
ix
trea
ted
grou
ps.
ID
Puta
tive
targ
et
gen
e
SC
UC
0.00
7
0.07
0.7
7
70
700
THR
A
Thyr
oid
rece
pto
r
alp
ha
1
±
1.97
0.78
±
0.68
0.56
±
0.76
0.84
±
1.73
1.71
±
2.04
2.33
±
2.92
3.09
±
5.93
1.48
±
0.93
*C
C78
Hea
t
shoc
k
cogn
ate
71
kDa
pro
tein
1
±
0.41
1.21
±
0.7
1.4
±
2.09
2.14
±
2.4
0.85
±
0.67
*
0.87
±
0.17
1.67
±
1.07
1.73
±
1.63
CC
41
Tran
smem
bran
e
4
L6
fam
ily
mem
ber
5
1
±
0.37
2.35
±
1.24
*
1.76
±
1.97
1.79
±
1.51
2.36
±
1.13
*
2.32
±
0.77
*
13.9
5
±
33.6
3 2.
56
±
0.72
*C
C24
Tran
slat
ion
elon
gati
on
fact
or
eEF-
1
del
ta
1
±
1.07
2.93
±
3.09
3.07
±
4.78
2.74
±
5.95
3.19
±
3.9
9.64
±
16.6
2*
3.42
± 3.
17*
2.3
±
1.72
C6-
15
Ald
olas
e
B
1
±
0.54
3.62
±
3.75
3.28
±
4.17
3.75
±
7.08
3.47
±
2.54
*
7.79
±
11.8
1
3.89
± 3.
68*
2.24
±
1.59
*G
4-18
Ast
acin
-lik
e
met
allo
pep
tid
ase
1
±
0.82
1.98
±
2.65
3.33
±
5.9
2.04
±
2.83
4.63
±
4.23
*
21.6
7
±
52.2
9*
3.93
± 3.
21*
1.95
±
2.9*
CC
2
Car
boxy
pep
tid
ase
A1
1
±
0.83
0.49
±
0.57
1.3
±
1.12
2.59
±
3.61
2.27
±
1.87
0.69
±
0.98
3.99
±
4.05
*
1.27
±
1.58
CC
9
Alp
ha
amyl
ase
1
±
0.59
1.41
±
1.33
1.14
±
1.16
3.44
±
4.08
3.73
±
4.65
*
0.87
±
1.13
*
3.97
±
3.55
*
3.01
±
4.7*
CC
77
Kin
inog
en
1
1
±
0.52
2.24
±
0.89
*
1.8
±
1.1
1.43
±
0.81
2.44
±
2.27
2.95
±
2.91
*
2.69
±
1.22
*
2.4
±
1.02
*C
C22
Elas
tase
1
1
±
0.84
0.85
±
0.56
2.91
±
3.76
2.77
±
4.51
5.03
±
5.32
*
9.21
±
17.5
9
4.21
±
2.5*
2.33
±
2.88
CC
42
Elas
tase
1
1
±
0.85
1.2
±
0.77
4.18
±
7.36
3.72
±
7.21
5.78
±
5.97
*
12.8
8
±
27.0
7 4.
36
±
2.57
*
2.54
±
3.1
G2-
15
Ch
ymot
ryp
sin
ogen
1
1
±
0.79
1.38
±
1.39
2.48
±
3.91
1.81
±
2.22
3.52
±
3.2
15.5
8
± 36
.66
3.32
±
2.52
*
1.65
±
2.14
D4-
17
Ch
ymot
ryp
sin
ogen
2
1
±
0.85
1.74
±
2.39
2.71
±
3.37
2.31
±
3.67
4
±
3.6*
19.9
2
± 47
.77
3.77
±
2.3*
1.77
±
2.3
Fig. 1. Log expression of vitellogenin A (shaded columns) and vitellogenin B (clearcolumns) mRNA ± SEM in flounder liver in response to Pentamix exposure. Datanormalised to mean of control group. Stars indicate single T-test P-values < 0.05versus control group.
expression but was not statistically significant by SAM due to highinter-individual variability. These vitellogenin expression changesare shown in Fig. 1. The subset of transcripts induced significantlyand at least 2-fold versus controls at the highest concentration isshown in Table 3. In addition to VTG, these fall into a number offunctional classes, including xenobiotic metabolism (e.g. CYP1A),proteasome, haem- and iron-binding, mitochondrial, pro-apoptoticand oxidative stress response.
3.3. Functional analyses of differentially expressed transcripts
Gene ontology analysis, using up-regulated transcripts derivedfrom the SAM analyses at FDR < 0.1, identified a number of GOterms that were significantly enriched versus a background ofall detectable flounder genes on the array. The down-regulatedtranscripts did not show significant enrichment of GO terms. Thecomplexity of the lists was reduced using the ‘find most spe-cific terms’ function of Blast2GO and further reduced by groupingsimilar GO terms. For example, translation, ribosome, structuralconstituent of ribosome and ribosome biogenesis were groupedtogether as ‘translation’. The statistical significance of these terms isillustrated for each exposure concentration in Fig. 2. Although thesetranscripts were generally more highly expressed with Pentamixexposure, differences in expression were concentration depend-ent. Transcripts for translation and cofactor metabolic processeswere induced at low concentration; those for proton transport
Fig. 2. Gene ontology groups significantly enriched (FDR < 0.1) amongst genes sig-nificantly (FDR < 0.1) induced with Pentamix exposures. x-Axis indicates exposuregroup, y-axis the inverse of the FDR enrichment value, z-axis gene ontology groups.
Author's personal copy
T.D. Williams et al. / Aquatic Toxicology 142– 143 (2013) 45– 52 49
Table 3Transcripts induced significantly (FDR < 0.1) and at least 2-fold with 700 �g/g Pentamix treatment. Asterisks indicate significance (FDR < 0.1), bold type indicates apparent2-fold change versus the mean of the control group.
Clone ID Name Symbol 0.07 0.7 7 700
EndocrineContig458 Vitellogenin A VTG 0.8 0.22 0.12 25.31*PfIL257D11 Vitellogenin A VTG 0.39* 0.25 0.23* 13.87*
Xenobiotic metabolismContig504 Cytochrome P450 1A CYP1A 2.33* 1.68 1.32 3.24*PfIL254G11 Alcohol dehydrogenase class III ADH5 1.76* 2.1* 1.65* 2.53*PfIL275F12 Aldo-keto reductase family 7, member A5 AKR7A2 2.66* 2.54* 2.54* 2.49*PfIL248F12 3 beta-hydroxysteroid dehydrogenase type VII HSD3B7 2.18* 1.73* 1.74* 2.17*
Protein degradationPfIL287C12 Proteasome subunit beta type 9 PSMB9 1.24 1.61 1.97* 2.78*Contig250 Proteasome subunit beta type 4 PSMB4 2.21* 2.22* 2.42* 2.22*PfIL226E10 Proteasome subunit beta type 5 PSMB5 1.97* 1.84* 2.42* 2.11*
were induced throughout the range. Transcripts related to pro-teasome, RNA export, golgi and endoplasmic reticulum were mosthighly induced at the 7 �g/g exposure, and those for haem proteinsat the 700 �g/g exposure. DAVID analysis of GO term enrich-ment was similar to that of Blast2GO; DAVID also identifiedenrichment of a number of KEGG pathways amongst inducedgenes (Fig. 3). Ribosome annotation (hsa03010) was enriched atthe lowest exposure concentration, 0.07 �g/g, but no other, andproteasome (hsa03050) at the middle two concentrations. Theenrichment of oxidative phosphorylation annotation (hsa00190)appeared exposure-responsive; its degree of fold enrichment cor-related highly with Pentamix exposure (r2 = 0.97, P-value = 0.014,2-tailed Welch T-test). KEGG annotation for human neurodegen-erative diseases (hsa05010, hsa05012, hsa05016), in which alteredoxidative phosphorylation is a predominant pathogenic factor, wasenriched with similar profiles (Fig. 3).
4. Discussion
DEG and real-time PCR successfully identified a number oftranscripts that were induced in response to Pentamix exposure,though inter-individual expression was highly variable reflectingthe previously published variability in BFR content of fish with cor-responding exposures. The majority of transcripts were related todigestion, glycolysis and translation, implying an increased feed-ing rate, energy production and protein synthesis. However, fishweight did not alter with Pentamix exposure (Kuiper et al., 2008),which may be consistent with an increased energy demand for
synthesis of metabolic proteins that is achieved by an increasein feeding. As both food and sediment were spiked with Pen-tamix, variations in digestive enzyme expression, feeding rate andinternal Pentamix concentrations between individuals are likely to
50 T.D. Williams et al. / Aquatic Toxicology 142– 143 (2013) 45– 52
be inter-related. Digestive enzymes are expressed mainly in thehepatopancreatic tissue that is distributed throughout fish liver,therefore intra-organ tissue heterogeneity may also affect apparentdigestive gene expression. Flatfish are known to form dominancehierarchies that could have contributed to variation in both feedingand stress responses of individuals (Dou et al., 2004).
Thyroid hormone receptor alpha (THRA) mRNA was alsoinduced above 0.7 �g/g Pentamix, consistent with its inductionin zebrafish larvae exposed to BDE-47 (Chan and Chan, 2012). Nochanges in plasma concentrations of T3 and T4 thyroid hormoneswere detected in the flounder used in the present study (Kuiperet al., 2008). Whilst it could be hypothesised that induction of THRAallows an increase in metabolic rate in the absence of increased T3and T4 concentrations, the situation is likely to be more complexgiven the large changes in digestive enzyme expression indicativeof differences on feeding behaviour between individuals.
Surprisingly, the number of transcripts assessed as signifi-cantly changing expression in comparison to the control groupby microarray analysis was not positively correlated with theexposure concentration of Pentamix. However the inter-individualvariation of mRNA expression within the 0.07 �g/g group was sig-nificantly lower than that of the control group and increased withexposure such that the 700 �g/g group showed more apparent vari-ability than the controls. This affected the number of transcriptsthat were assessed as statistically significantly altered in expres-sion. As shown in Table 3, a number of transcripts (e.g. CYP1A)were significantly induced with the lowest and highest exposurebut not at intermediate concentrations. Additionally, muscle tissueaccumulation of BDE47, the major component of Pentamix, washighly variable between individual fish, with means of the threelower exposures falling into a similar range (0.21–0.34 �g/g wetwt) whilst the mean for the highest exposure was over two ordersof magnitude higher (Table 1) and varied nearly 20-fold betweenindividuals (Kuiper et al., 2008). Although muscle tissue concen-tration may not be a good surrogate for liver tissue, this variationdoes imply differential uptake, distribution, metabolism or excre-tion of the PBDEs between individuals. These findings, and the lackof linear dose–response for the majority of differentially expressedtranscripts, imply considerable complexity in the hepatic responseto Pentamix. Additionally the large variability in PBDE accumula-tion and in mRNA expression is entirely in accord with variability ininter-individual feeding behaviour which is indicated by the vari-able digestive enzyme transcript expression observed.
The most dramatic transcriptional change was a significantinduction of vitellogenin A of over 10-fold with 700 �g/g Pentamixexposure, whilst at lower concentrations VTG was significantlyrepressed in comparison to controls (Fig. 1). These apparentlyopposing effects may be attributable to effects of different con-geners in the DE-71 mixture or metabolites of these components,since in rats BDEs 47, 49 and 100 have been characterised as weakto moderate ER agonists, whilst 6-OH-BDE47 was a highly potentER antagonist (Hamers et al., 2006, 2008). The observed repres-sion of VTG expression at low doses of Pentamix in flounder isconsistent with the repression of VTG and estrogen receptor (ER)transcripts reported by Han et al. (2011) when male zebrafish wereexposed for 120 days to 0.5 �g/L Pentamix. It is possible that one ormore hydroxylated metabolites could be responsible for repress-ing ER activity and thus reducing VTG transcription at low andmedium doses of Pentamix and highlights the need for character-isation of BDE metabolism in different species, as it differs evenamongst the mammals, for example between beluga whale (Del-phinapterus leucas) and rat (McKinney et al., 2006). The inductionof VTG expression observed at the highest exposure concentra-tion may also be due to a different proportion of ER agonisticand antagonistic congeners and their metabolites and could indi-cate that the biotransformation capacity of the tissue has been
exceeded. Indeed, BDE49 is a moderate inhibitor of estrogen sul-fotransferase and BDE47 is a highly potent inhibitor (Hamers et al.,2006), whilst some hydroxylated metabolites of BDE47 are evenmore potent than the parent compound (Hamers et al., 2008), also,sulfotransferases can become saturated at low substrate concen-trations (Coughtrie, 2002). Thus different concentrations of parentcongeners and their metabolites may induce differential effects onER target genes. Inhibition of estrogen sulfotransferase indirectlyincreases estradiol bioavailability in target tissues, thus potentiallyactivating ERs and inducing expression of VTG. This is a similarmechanism to that shown for polychlorinated biphenyls (Kesteret al., 2000). Given the severe reduction of spawning in sticklebackchronically exposed to a high dose of PBDE for 3.5 months (Holmet al., 1993) Pentamix might be considered a reproductive toxicantat high concentrations.
Transcripts encoding a variety of xenobiotic metabolisingenzymes, most prominently CYP1A, class III alcohol dehydroge-nase (ADH5) and aldo-keto reductase 7 (AKR7A2) were significantlyinduced at both high and low concentrations of Pentamix and3-beta-hydroxysteroid dehydrogenase type VII (HSD3B7) was sig-nificantly induced at all concentrations (Table 3). Metabolism ofPBDEs can proceed through cytochrome P450 activity, includingP450 superfamilies 1, 2 and 3 (Erratico et al., 2011) but, as forthe estrogen receptor, different PBDEs and their metabolites dis-play different agonistic and antagonistic effects on dioxin responseelement activated transcription via the aromatic hydrocarbonreceptor (Hamers et al., 2006). No increase in CYP-dependentEROD activity was found in this experiment (Kuiper et al., 2008),although CYP1A transcript significantly increased over 3-fold atthe highest PBDE concentration (Table 3), similar to the induc-tion of CYP transcription in salmon hepatocytes acutely exposedto three BDEs (Softeland et al., 2011). Whether P450s are inducedby PBDEs has long been controversial and has been complicatedby the presence of dibenzofuran and dioxin AhR inducers in thePBDE preparations used for early experiments (Peters et al., 2004),but the Pentamix used in this study was free from such con-taminants (Kuiper et al., 2008). Interestingly, Peters et al. (2004)showed that PBDE reduced TCDD-dependent EROD induction with-out a reduction in CYP1A1 mRNA, that was hypothesised to bedue to a post-transcriptional process such as interference withhaem synthesis. In our experiment the highest concentration ofPentamix induced the transcription of a set of metalloproteinsincluding haemoglobins, biliverdin reductase, delta aminolevulinicacid synthase and ferritin (Table 3), implying disruption of iron andhaem homeostasis. Additional responsive genes included ADH5,inducible by lipid peroxides and AKR7A5 that reduces quinones,both may be therefore induced under oxidative stress conditions.HSD3B7 is an enzyme of the bile synthesis pathway; its inductionis consistent with an increase in excretion of conjugated metabo-lites.
In addition to a focus on transcriptional changes in key toxi-cologically relevant pathways such as endocrine disruption andxenobiotic metabolism, annotation enrichment analyses providean overview of the biological processes and pathways altered afterPentamix exposure. At the lowest concentration of Pentamix, andno other, KEGG and GO terms related to ribosome were enrichedamongst induced transcripts (Fig. 3), implying a requirement forincreased protein synthesis. This has been observed during sub-toxic exposure to a variety of toxicants in flounder liver (Williamset al., 2008) and in mammals (Chung et al., 2005), whilst thereis much evidence for repression of translation under more severetoxic stress (Patel et al., 2002). At the medium exposure levels pro-teasomal transcripts were significantly enriched amongst inducedtranscripts (Fig. 3) with some individual transcripts also showinginduction at the highest exposure level (Table 3). These changesimply an increased requirement for degradation of proteins,
Author's personal copy
T.D. Williams et al. / Aquatic Toxicology 142– 143 (2013) 45– 52 51
potentially due to oxidative protein damage. There was little evi-dence of a co-ordinated induction of chaperones, but we havepreviously found that this tends to be a transient response to acutestress in flounder (Williams et al., 2006).
The major response detected by annotation enrichment anal-ysis was over-representation of transcripts encoding proteins ofthe oxidative phosphorylation pathway amongst induced trans-cripts in all exposed groups (Fig. 3). Enrichment of oxidativephosphorylation annotation correlated highly and significantlywith Pentamix exposure, and was mirrored by enrichment ofannotation to neurodegenerative diseases in humans. This isdue to the overlap between these four pathways. The inducedgenes related to oxidative phosphorylation (has:00190) consistedof mitochondrial ATP synthases (ATP5A1, ATP5C1, ATP5F1,ATP5J2), vacuolar ATP synthases (ATP6V0A1, ATP6V0C, ATP6V0E1),cytochrome oxidases (COX4l1, COX5A, COX5B, COX6C, COX6B1,COX7C, COX8A), ubiquinol–cytochrome c reductase (UQCRQ) andNADH dehydrogenases (NDUFB2, NDUFB10) with, in addition; forParkinson’s (hsa05012) mitochondrial membrane porin (VDAC2),pro-apoptotic mitochondrial serine protease (HTRA2) and the apo-ptosis executor caspase 3 (CASP3); for Alzheimer’s (hsa05010)calmodulin 3 (CALM3); and for Huntingdon’s (hsa05016) superox-ide dismutases (SOD1, SOD2) and histone deacetylase 1 (HDAC1).Induction of these mitochondrial transcripts implies mitochon-drial dysfunction, oxidative stress and induction of apoptosis. Theseresponses to PBDE exposure are entirely consistent with thosereported in PBDE exposed cultured chicken (Ji et al., 2011) andhuman cells (Zhong et al., 2011) where there was also an induc-tion of apoptosis and similar effects on electron transport geneexpression (Suvorov and Takser, 2010). In a study with zebrafishexposed to 6-OH-BDE47 van Boxtel et al. (2008) also detected asimilar induction of vacuolar ATPases and determined that thishydroxylated metabolite of BDE was an uncoupler of oxidativephosphorylation. Although it is unclear if 6-OH-BDE47 is a majormetabolite (Wan et al., 2010), other metabolites of BDE47 arealso uncouplers of oxidative phosphorylation, for example 2,4-dibromophenol (Hempfling, 1970; Wang et al., 2012).
5. Conclusions
Pentamix induced transcriptional changes in flounder liverwhen it was detected in fish muscle at a mean concentration of0.2 mg/kg and above. Transcriptional changes were more sensitivebiomarkers of exposure than the enzyme activities and hor-mone levels measured previously. These transcriptional changesmay mainly represent an adaptive response, including xenobio-tic metabolism and translation, whilst some implied potentiallydeleterious effects, including repression of vitellogenin transcrip-tion, oxidative stress, induction of mitochondrial dysfunction andapoptosis. When Pentamix was detected in muscle at a meanconcentration of 45 mg/kg, there were hepatic transcriptional indi-cations of pro-estrogenic endocrine disruption and tissue damage,distinct from the profiles elicited by lower doses.
Given the recognised differential uptake of PBDE congenersand the potential for a wide range of species-specific metabolicproducts, it is not surprising that experiments employing differentdosing strategies, times of exposure, tissues and species have oftenresulted in different responses of the endocrine system and xeno-biotic metabolism between different experimental models. Indeed,in our experiment there was also wide inter-individual variationin BDE accumulation and gene transcription. These diverse pro-cesses imply a complex interplay between absorption, distribution,metabolism and excretion that is highly likely to present intracellu-lar concentrations of congeners and metabolites that vary with doseand time and result in non-linear responses. Pentamix is indeed a
technical mixture but in some respects also appears to be a non-additive toxicological mixture.
Acknowledgements
This work was funded by EU‘FIRE’ grant QLK4-CT-2001-00596and EU ‘GENIPOL’ grant EKV-2001-0057.
Appendix A. Supplementary data
Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.aquatox.2013.07.013.
References
Albina, M.L., Alonso, V., Linares, V., Belles, M., Sirvent, J.J., Domingo, J.L., Sanchez, D.J.,2010. Effects of exposure to BDE-99 on oxidative status of liver and kidney inadult rats. Toxicology 271, 51–56.
Birnbaum, L.S., Cohen Hubal, E.A., 2006. Polybrominated diphenyl ethers: a casestudy for using biomonitoring data to address risk assessment questions. Envi-ronmental Health Perspectives 114, 1770–1775.
Birnbaum, L.S., Staskal, D.F., 2004. Brominated flame retardants: cause for concern?Environmental Health Perspectives 112, 9–17.
Blanco, J., Mulero, M., Lopez, M., Domingo, J.L., Sanchez, D.J., 2011. BDE-99 deregu-lates BDNF, Bcl-2 and the mRNA expression of thyroid receptor isoforms in ratcerebellar granular neurons. Toxicology 290, 305–311.
Boon, J.P., Lewis, W.E., Tjoen, A.C.M.R., Allchin, C.R., Law, R.J., De Boer, J., Ten Hallers-Tjabbes, C.C., Zegers, B.N., 2002a. Levels of polybrominated diphenyl ether(PBDE) flame retardants in animals representing different trophic levels of theNorth Sea food Web. Environmental Science & Technology 36, 4025–4032.
Boon, J.P., van Zanden, J.J., Lewis, W.E., Zegers, B.N., Goksoyr, A., Arukwe, A.,2002b. The expression of CYP1A vitellogenin and zona radiata proteins inAtlantic salmon (Salmo salar) after oral dosing with two commercial PBDE flameretardant mixtures: absence of short-term responses. Marine EnvironmentalResearch 54, 719–724.
Branchi, I., Capone, F., Alleva, E., Costa, L.G., 2003. Polybrominated diphenyl ethers:neurobehavioral effects following developmental exposure. Neurotoxicology24, 449–462.
Bruchajzer, E., Frydrych, B., Sporny, S., Szymanska, J.A., 2010. Toxicity of penta- anddecabromodiphenyl ethers after repeated administration to rats: a comparativestudy. Archives of Toxicology 84, 287–299.
Chan, W.K., Chan, K.M., 2012. Disruption of the hypothalamic-pituitary-thyroid axisin zebrafish embryo-larvae following waterborne exposure to BDE-47, TBBPAand BPA. Aquatic Toxicology 108, 106–111.
Chung, H., Hong, D.P., Jung, J.Y., Kim, H.J., Jang, K.S., Sheen, Y.Y., Ahn, J.I., Lee, Y.S.,Kong, G., 2005. Comprehensive analysis of differential gene expression profileson carbon tetrachloride-induced rat liver injury and regeneration. Toxicologyand Applied Pharmacology 206, 27–42.
Coughtrie, M.W., 2002. Sulfation through the looking glass—recent advances in sul-fotransferase research for the curious. Pharmacogenomics Journal 2, 297–308.
Diab, A.M., Williams, T.D., Sabine, V.S., Chipman, J.K., George, S.G., 2008. The GENIPOLEuropean flounder Platichthys flesus L toxicogenomics microarray: applicationfor investigation of the response to furunculosis vaccination. Journal of FishBiology 72, 2154–2169.
Dou, S.-Z., Masuda, R., Tanaka, M., Tsukamoto, K., 2004. Size hierarchies affectingthe social interactions and growth of juvenile Japanese flounder, Paralichthysolivaceus. Aquaculture 233, 237–249.
Dunnick, J.K., Brix, A., Cunny, H., Vallant, M., Shockley, K.R., 2012. Characterizationof polybrominated diphenyl ether toxicity in Wistar Han rats and use of livermicroarray data for predicting disease susceptibilities. Toxicologic Pathology 40,93–106.
Erratico, C.A., Moffatt, S.C., Bandiera, S.M., 2011. Comparative oxidative metabolismof BDE-47 and BDE-99 by rat hepatic microsomes. Toxicological Sciences 123,37–47.
Fery, Y., Buschauer, I., Salzig, C., Lang, P., Schrenk, D., 2009. Technicalpentabromodiphenyl ether and hexabromocyclododecane as activators of thepregnane-X-receptor (PXR). Toxicology 264, 45–51.
Gotz, S., Garcia-Gomez, J.M., Terol, J., Williams, T.D., Nagaraj, S.H., Nueda, M.J., Robles,M., Talon, M., Dopazo, J., Conesa, A., 2008. High-throughput functional anno-tation and data mining with the Blast2GO suite. Nucleic Acids Research 36,3420–3435.
Author's personal copy
52 T.D. Williams et al. / Aquatic Toxicology 142– 143 (2013) 45– 52
Hamers, T., Kamstra, J.H., Sonneveld, E., Murk, A.J., Kester, M.H., Andersson, P.L.,Legler, J., Brouwer, A., 2006. In vitro profiling of the endocrine-disruptingpotency of brominated flame retardants. Toxicological Sciences 92, 157–173.
Hamers, T., Kamstra, J.H., Sonneveld, E., Murk, A.J., Visser, T.J., Van Velzen, M.J.,Brouwer, A., Bergman, A., 2008. Biotransformation of brominated flame retard-ants into potentially endocrine-disrupting metabolites with special attentionto 2,2′ ,4,4′-tetrabromodiphenyl ether (BDE-47). Molecular Nutrition & FoodResearch 52, 284–298.
Han, X.B., Lei, E.N., Lam, M.H., Wu, R.S., 2011. A whole life cycle assess-ment on effects of waterborne PBDEs on gene expression profile along thebrain–pituitary–gonad axis and in the liver of zebrafish. Marine Pollution Bul-letin 63, 160–165.
Hempfling, W.P., 1970. Studies of the efficiency of oxidative phosphorylation inintact Escherichia coli B. Biochimica et Biophysica Acta 205, 169–182.
Holm, G., Norrgren, L., Andersson, T., Thuren, A., 1993. Effects of exposure to foodcontaminated with PBDE, PCN or PCB on reproduction, liver morphology andcytochrome P450 activity in the three-spined stickleback, Gasterosteus aculeatus.Aquatic Toxicology 27, 33–50.
Holm, G., Lundstrom, J., Andersson, T., Norrgren, L., 1994. Influences of halo-genated organic substances on ovarian development and hepatic EROD activityin the three-spined stickleback, Gasterosteus aculeatus, and rainbow trout,Oncorhynchus mykiss. Aquatic Toxicology 29, 241–256.
Huang, D.W., Sherman, B.T., Lempicki, R.A., 2009a. Bioinformatics enrichment tools:paths toward the comprehensive functional analysis of large gene lists. NucleicAcids Research 37, 1–13.
Huang, D.W., Sherman, B.T., Lempicki, R.A., 2009b. Systematic and integrative anal-ysis of large gene lists using DAVID bioinformatics resources. Nature Protocols4, 44–57.
Ji, K., Choi, K., Giesy, J.P., Musarrat, J., Takeda, S., 2011. Genotoxicity of severalpolybrominated diphenyl ethers (PBDEs) and hydroxylated PBDEs and theirmechanisms of toxicity. Environmental Science & Technology 45, 5003–5008.
Kemmlein, S., Herzke, D., Law, R.J., 2009. Brominated flame retardants in the Euro-pean chemicals policy of REACH-Regulation and determination in materials.Journal of Chromatography A 1216, 320–333.
Kester, M.H., Bulduk, S., Tibboel, D., Meinl, W., Glatt, H., Falany, C.N., Coughtrie, M.W.,Bergman, A., Safe, S.H., Kuiper, G.G., Schuur, A.G., Brouwer, A., Visser, T.J., 2000.Potent inhibition of estrogen sulfotransferase by hydroxylated PCB metabolites:a novel pathway explaining the estrogenic activity of PCBs. Endocrinology 141,1897–1900.
Kirby, M.F., Allen, Y.T., Dyer, R.A., Feist, S.W., Katsiadaki, I., Matthiessen, P., Scott, A.P.,Smith, A., Stentiford, G.D., Thain, J.E., Thomas, K.V., Tolhurst, L., Waldock, M.J.,2004. Surveys of plasma vitellogenin and intersex in male flounder (Platichthysflesus) as measures of endocrine disruption by estrogenic contamination inUnited Kingdom estuaries Temporal trends, 1996 to 2001. Environmental Tox-icology and Chemistry 23, 748–758.
Kuiper, R.V., Murk, A.J., Leonards, P.E., Grinwis, G.C., van den Berg, M., Vos, J.G.,2006. In vivo and in vitro Ah-receptor activation by commercial and fraction-ated pentabromodiphenylether using zebrafish (Danio rerio) and the DR-CALUXassay. Aquatic Toxicology 79, 366–375.
Kuiper, R.V., Vethaak, A.D., Canton, R.F., Anselmo, H., Dubbeldam, M., van denBrandhof, E.J., Leonards, P.E., Wester, P.W., van den Berg, M., 2008. Toxicityof analytically cleaned pentabromodiphenylether after prolonged exposure inestuarine European flounder (Platichthys flesus), and partial life-cycle exposurein fresh water zebrafish (Danio rerio). Chemosphere 73, 195–202.
Legler, J., 2008. New insights into the endocrine disrupting effects of brominatedflame retardants. Chemosphere 73, 216–222.
Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data usingreal-time quantitative PCR and the 2(−Delta Delta C(T)) Method. Methods 25,402–408.
McKinney, M.A., De Guise, S., Martineau, D., Beland, P., Arukwe, A., Letcher, R.J.,2006. Biotransformation of polybrominated diphenyl ethers and polychlori-nated biphenyls in beluga whale (Delphinapterus leucas) and rat mammalianmodel using an in vitro hepatic microsomal assay. Aquatic Toxicology 77, 87–97.
Olsvik, P.A., Lie, K.K., Sturve, J., Hasselberg, L., Andersen, O.K., 2009. Transcriptionaleffects of nonylphenol, bisphenol A and PBDE-47 in liver of juvenile Atlantic cod(Gadus morhua). Chemosphere 75, 360–367.
Pacyniak, E.K., Cheng, X., Cunningham, M.L., Crofton, K., Klaassen, C.D., Guo, G.L.,2007. The flame retardants polybrominated diphenyl ethers, are pregnane Xreceptor activators. Toxicological Sciences 97, 94–102.
Patel, J., McLeod, L.E., Vries, R.G., Flynn, A., Wang, X., Proud, C.G., 2002. Cellularstresses profoundly inhibit protein synthesis and modulate the states of phos-phorylation of multiple translation factors. European Journal of Biochemistry269, 3076–3085.
Peters, A.K., van Londen, K., Bergman, A., Bohonowych, J., Denison, M.S., van denBerg, M., Sanderson, J.T., 2004. Effects of polybrominated diphenyl ethers onbasal and TCDD-induced ethoxyresorufin activity and cytochrome P450-1A1expression in MCF-7, HepG2, and H4IIE cells. Toxicological Sciences 82, 488–496.
Pike, N., 2011. Using false discovery rates for multiple comparisons in ecology andevolution. Methods in Ecology and Evolution 2, 278–282.
Saeed, A.I., Bhagabati, N.K., Braisted, J.C., Liang, W., Sharov, V., Howe, E.A., Li, J., Thi-agarajan, M., White, J.A., Quackenbush, J., 2006. TM4 microarray software suite.Methods in Enzymology 411, 134–193.
Softeland, L., Petersen, K., Stavrum, A.K., Wu, T., Olsvik, P.A., 2011. Hepatic in vitrotoxicity assessment of PBDE congeners BDE47 BDE153 and BDE154 in Atlanticsalmon (Salmo salar L.). Aquatic Toxicology 105, 246–263.
Suvorov, A., Takser, L., 2010. Global gene expression analysis in the livers of rat off-spring perinatally exposed to low doses of 2,2′ ,4,4′-tetrabromodiphenyl ether.Environmental Health Perspectives 118, 97–102.
Tusher, V.G., Tibshirani, R., Chu, G., 2001. Significance analysis of microar-rays applied to the ionizing radiation response. Proceedings of theNational Academy of Sciences of the United States of America 98, 5116–5121.
van Boxtel, A.L., Kamstra, J.H., Cenijn, P.H., Pieterse, B., Wagner, J.M., Antink, M., Krab,K., van der Burg, B., Marsh, G., Brouwer, A., Legler, J., 2008. Microarray analy-sis reveals a mechanism of phenolic polybrominated diphenylether toxicity inzebrafish. Environmental Science & Technology 42, 1773–1779.
Vethaak, A.D., Pieters, J., Jol, J., 2009. Long-term trends in the prevalence of cancerand other major diseases among flatfish in the Southeastern North Sea as indi-cators of changing ecosystem health. Environmental Science & Technology 43,2151–2158.
Wahl, M., Guenther, R., Yang, L., Bergman, A., Straehle, U., Strack, S., Weiss, C.,2010. Polybrominated diphenyl ethers and arylhydrocarbon receptor agonists:different toxicity and target gene expression. Toxicology Letters 198, 119–126.
Wan, Y., Liu, F., Wiseman, S., Zhang, X., Chang, H., Hecker, M., Jones, P.D., Lam, M.H.,Giesy, J.P., 2010. Interconversion of hydroxylated and methoxylated polybromi-nated diphenyl ethers in Japanese medaka. Environmental Science & Technology44, 8729–8735.
Wang, X., Wang, Y., Chen, J., Ma, Y., Zhou, J., Fu, Z., 2012. Computational toxicologicalinvestigation on the mechanism and pathways of xenobiotics metabolized bycytochrome P450: a case of BDE-47. Environmental Science & Technology 46,5126–5133.
Williams, T.D., Diab, A.M., George, S.G., Godfrey, R.E., Sabine, V., Conesa, A., Minchin,S.D., Watts, P.C., Chipman, J.K., 2006. Development of the GENIPOL Euro-pean flounder (Platichthys flesus) microarray and determination of temporaltranscriptional responses to cadmium at low dose. Environmental Science &Technology 40, 6479–6488.
Williams, T.D., Diab, A.M., George, S.G., Sabine, V., Chipman, J.K., 2007. Gene expres-sion responses of European flounder (Platichthys flesus) to 17-beta estradiol.Toxicology Letters 168, 236–248.
Williams, T.D., Diab, A., Ortega, F., Sabine, V.S., Godfrey, R.E., Falciani, F., Chipman, J.K.,George, S.G., 2008. Transcriptomic responses of European flounder (Platichthysflesus) to model toxicants. Aquatic Toxicology 90, 83–91.
Williams, T.D., Gensberg, K., Minchin, S.D., Chipman, J.K., 2003. A DNA expressionarray to detect toxic stress response in European flounder (Platichthys flesus).Aquatic Toxicology 65, 141–157.
Zhong, Y.F., Wang, L.L., Yin, L.L., An, J., Hou, M.L., Zheng, K.W., Zhang, X.Y., Wu,M.H., Yu, Z.Q., Sheng, G.Y., Fu, J.M., 2011. Cytotoxic effects and oxidative stressresponse of six PBDE metabolites on human L02 cells. Journal of Environmen-tal Science and Health. Part A: Toxic/Hazardous Substances & EnvironmentalEngineering 46, 1320–1327.
