Chemosphere xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Induction of gene responses in St. Lawrence River northern pike(Esox lucius) environmentally exposed to perfluorinated compounds

0045-6535/$ - see front matter Crown Copyright � 2013 Published by Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.chemosphere.2013.01.099

⇑ Corresponding author. Tel.: +1 514 496 6774; fax: +1 514 496 7398.E-mail addresses: magali.houde@ec.gc.ca (M. Houde), melanie.douville@ec.gc.ca

(M. Douville), simon.despatie@ec.gc.ca (S.-P. Despatie), amila.desilva@ec.gc.ca (A.O.De Silva), christine.spencer@ec.gc.ca (C. Spencer).

Please cite this article in press as: Houde, M., et al. Induction of gene responses in St. Lawrence River northern pike (Esox lucius) environmentally eto perfluorinated compounds. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.01.099

Magali Houde a,⇑, Mélanie Douville a, Simon-Pierre Despatie a, Amila O. De Silva b, Christine Spencer b

a Environment Canada, Centre Saint-Laurent, 105 McGill Street, Montreal, QC, Canada H2Y 2E7b Environment Canada, National Water Research Institute, 867 Lakeshore Road, Burlington, Ontario, Canada L7R 4A6

h i g h l i g h t s

" Municipal effluent may be a source of PFCs in the St. Lawrence River." PFECHS is detected in St. Lawrence northern pike tissues." Gene expression is higher in northern pike captured downstream a municipal effluent." Possible links may exist between PFCs, Vtg gene expression and Vtg protein activity.

a r t i c l e i n f o

Article history:Received 19 December 2012Accepted 31 January 2013Available online xxxx

Keywords:Northern pikeGene expressionBiomarkerMunicipal waste water effluentPerfluorinated compoundsSt. Lawrence River

a b s t r a c t

Municipal waste water effluents (MWWEs) are important sources of chemical contamination for aquaticenvironments. This study investigated the presence and effects of perfluorinated compounds (PFCs) inenvironmentally exposed northern pike (Esox lucius) collected upstream and downstream a major muni-cipal waste water treatment plant (WWTP) in the St. Lawrence River, Canada. Twelve PFCs, including thenewly detected perfluoroethylcyclohexane sulfonate (PFECHS), were quantified in fish muscle, liver, andplasma. Additionally, the expression of eight genes and the activity of three biomarkers were analyzed infish tissues at both sites. Mean

PPFC concentration in fish plasma collected upstream the WWTP was

185 ng/g w.w. compared to 545 ng/g w.w. downstream the point of release. PFECHS was quantified forthe first time in St. Lawrence River fish (mean plasma concentration in MWWE fish: 5.07 ± 4.72 ng/gw.w.). Results of transcriptomic responses were tissue-specific and indicated significant up-regulationfor metallothionein (MT) in blood and MT, glutathion-S-transferase (GST), superoxide dismutase (SOD),and cytochromes P450 1A1 (CYP1A1) in gill tissue of fish collected in the MWWE suggesting greaterstress responses for organisms at this location. Significant relationships were found between vitellogenin(Vtg) gene expression in liver, Vtg activity in plasma and perfluorotridecanoic acid (PFTrA), perfluorotet-radecanoic acid (PFTeA), and perfluorodecane sulfonate (PFDS) plasma concentrations. The possibleendocrine effects of these PFCs should be further investigated.

Crown Copyright � 2013 Published by Elsevier Ltd. All rights reserved.

1. Introduction

The continuous release of chemicals in aquatic ecosystems re-mains an issue of environmental management and a threat to thesustainability of aquatic resources. Municipal wastewater effluents(MWWEs) are complex mixtures of anthropogenic contaminantsand microorganisms and important direct sources of pollutantsfor aquatic environments. Enhanced analytical techniques now en-able the detection and quantification of an increasing number of

substances. However, very little is known about the toxicity ofthose pollutants in wildlife particularly in naturally-exposed aqua-tic organisms. Perfluorinated compounds (PFCs) have been underthe scope of scientists for a number of years as contaminants ofconcern because of their global distribution, their persistence,and their bioaccumulation potential (Houde et al. 2006; Fujiiet al. 2007; Houde et al. 2011). The integration of genomic toolsin ecotoxicological assessment may help understand the impactsof chemical stress to wildlife. Indeed, studying changes in geneexpression, the first line of response to contaminants, is a sensitivetool that may help understand the modes of action of chemicals inorganisms. Integrating related protein expression changes is a keyfactor to increase the functional relevance of genomic analyses(Nikinmaa and Rytkönen, 2011).

xposed

un

e20

11)

9Li

ver

N=

11M

usc

leN

=11

Mea

n±

std

dev

n<

MD

LM

ean

±st

dde

vn

<M

DL

Mea

n±

std

dev

<0.1

69;

(2n

d)<0

.16

All

nd

All

nd

10.2

±7.

260

1.81

±0.

77e

20.

14±

0.08

34.5

±19

.00

6.48

±2.

90e

00.

49±

0.24

e

35.4

±18

.10

7.37

±3.

00e

00.

54±

0.25

11.6

±5.

260

3.16

±1.

17e

40.

23±

0.15

9.51

±2.

630

3.25

±1.

09e

10.

27±

0.09

1.76

±0.

680

1.00

±0.

29e

11<0

.20

0.32

±0.

192

0.24

±0.

135

0.02

±0.

009

1.57

±2.

061

0.38

±0.

148;

(3n

d)<0

.16

410

±20

4e0

127

±74

.7e

09.

70±

4.78

e

15.4

±12

.30

7.25

±4.

63e

00.

32±

0.18

5.07

±4.

72e

00.

54±

0.25

eA

lln

dA

lln

d

545.

516

1.6

12.0

103.

4(2

1%)

23.4

(17%

)1.

75(1

6%)

430.

5(7

9%)

134.

8(8

3%)

10(8

4%)

2 M. Houde et al. / Chemosphere xxx (2013) xxx–xxx

Montreal’s primary wastewater treatment plant (WWTP) is oneof the largest in North America processing more than 2 million m3

of raw sewage daily and discharging its final MWWE directly intothe St. Lawrence River. Information on the contamination of fish inthe St. Lawrence River is scarce and therefore very little informa-tion exists on the potential effects of persistent contaminants infish at this location. Hence northern pike (Esox lucius) were usedto quantify PFC concentrations and to evaluate the possible im-pacts of these chemical stressors on fish health using genomic tools(i.e., qPCR) and biological indicators (i.e., biomarkers). Northernpike, present in most freshwater system of North America, areopportunist carnivores, feeding primarily on small fish, insects,crayfish, and sometimes small vertebrates (Bernatchez and Giroux,2000). The ubiquity and trophic position of this territorial predatorpromote success in captures and presence of PFCs in tissues. Lowlevel of genetic variations have been reported for northern pike(Seeb et al. 1987; Senanan and Kapuscinski, 2000) which representone of the most valuable commercial freshwater species in thenorthwest and a prized game fish (Senanan and Kapuscinski,2000). Habitat changes, however, have greatly affected their pro-duction (Casselman and Lewis, 1996).

on)

ofPF

Cs(n

g/g

wei

ght

wet

)in

nort

hern

pike

(Eso

xlu

cius

)pl

asm

a,liv

er,a

ndm

uscl

e.

Lac

des

Deu

x-M

onta

gnes

(Ju

ne

2011

)Il

etV

ert

(J

Plas

ma

N=

10Li

ver

N=

11M

usc

leN

=11

Plas

ma

N=

sM

DL

n<

MD

LM

ean

±st

dde

van

<M

DL

Mea

n±

std

dev

n<

MD

LM

ean

±st

dde

vn

<M

DL

0.16

All

nd

bA

lln

d7;

(4n

d)<0

.16

All

nd

All

nd

4;(4

nd)

0.08

07.

67±

3.21

01.

12±

0.54

10<0

.08

00.

160

23.2

±5.

690

3.49

±1.

032

0.24

±0.

100

0.19

029

.5±

6.51

04.

83±

1.35

10.

42±

0.14

00.

170

8.57

±1.

920

1.77

±0.

4510

<0.1

70

0.19

08.

70±

1.94

01.

83±

0.52

70.

21±

0.07

00.

200

1.87

±0.

480

0.50

±0.

1711

<0.2

00

0.02

00.

45±

0.19

00.

17±

0.10

10.

04±

0.01

e0

0.16

4<0

.16

11<0

.16

All

nd

All

nd

00.

450

96.9

±20

.40

25.1

±6.

100

1.74

±0.

520

0.13

08.

11±

3.06

02.

46±

1.07

00.

25±

0.14

00.

21A

lln

dA

lln

dA

lln

dA

lln

dA

lln

dA

lln

d0

185.

441

.33.

1579

.9(4

3%)

13.7

(33%

)1.

11(3

5%)

105

(57%

)27

.6(6

7%)

2.04

(65%

)

eth

eco

mpo

un

dw

asde

tect

ed.

,PFT

rA,P

FTeA

,PFP

eDa.

ted

site

for

the

sam

eti

ssu

e.

2. Material and methods

2.1. Environmental fish samples

To minimize biological variation in genomic expression, fishwere collected post-reproduction and a range of sizes was deter-mined before captures. Sampling of adult northern pike (lengthranging from 495 to 845 mm) was conducted at two sites in thevicinity of Montréal, Quebec, Canada in June 2011. A total of 11 fishwere captured at Ilet Vert (10 females; 1 male) downstream Mon-treal’s WWTP whereas 11 fish were taken from an upstream site atLac-des-Deux-Montagnes (9 females; 2 males; SupplementaryMaterial, SM Fig. 1). The low number of males is coincidental andno analyses were conducted between genders. Temperature, dis-solved oxygen, pH, turbidity, and specific conductivity of the waterwere measured at each site using a YSI™ 6600V4 multi parameterprobe. All fish were captured from a boat using standard anglingequipment. Pike were euthanized in a 250 mg/L solution of cloveoil. Fresh blood was drawn and 400 lL of whole blood transferredin RNase/DNase free tubes containing RNAlater� (Sigma–Aldrich,Canada). Remaining blood was kept on ice until centrifugation inthe laboratory to obtain plasma. Gill lamellae, liver, and muscle tis-sue samples were put in RNAlater� and frozen at �20 �C until RNAextraction. Fish carcasses were frozen (�20 �C) until microscopicgender determination using gonads and chemical analyses. Clei-thra were removed for age determination by counting the numberof growth annuli. Fish age ranged between 2 and 8 years-old.

Tabl

e1

Ari

thm

etic

mea

nco

ncen

trat

ions

(±st

anda

rdde

viat

i

Com

pou

nds

Nam

esA

cron

ym

Perfl

uor

ooct

anoi

cac

idPF

OA

Perfl

uor

onon

anoi

cac

idPF

NA

Perfl

uor

odec

anoi

cac

idPF

DA

Perfl

uor

oun

deca

noi

cac

idPF

Un

APe

rflu

orod

odec

anoi

cac

idPF

DoA

Perfl

uor

otri

deca

noi

cac

idPF

TrA

Perfl

uor

otet

rade

can

oic

acid

PFTe

APe

rflu

orop

enta

deca

noi

cac

idPF

PeD

aPe

rflu

oroh

exan

esu

lfon

ate

PFH

xSPe

rflu

oroo

ctan

esu

lfon

ate

PFO

SPe

rflu

orod

ecan

esu

lfon

ate

PFD

SPe

rflu

oroe

thyl

cycl

ohex

ane

sulf

onat

ePF

ECH

S

Sum

ofal

lPF

Cs

Sum

PFC

Asc

and

%of

PFC

sSu

mPF

SAsd

and

%of

PFC

s

aM

ean

con

cen

trat

ion

s±

SDin

all

sam

ples

wh

erb

nd

=N

otde

tect

ed.

cSu

mPF

CA

s:PF

OA

,PFN

A,P

FDA

,PFU

nA

,PFD

oAd

Sum

PFSA

s:PF

HxS

,PFO

S,PF

DS,

PFEC

HS.

eC

once

ntr

atio

ns

sign

ifica

ntl

yh

igh

erat

desi

gna

2.2. Contaminant analyses

Twelve perfluorinated compounds (Table 1), including perflu-oroethylcyclohexane sulfonate (PFECHS), were analyzed in plasma(n = 19; blood could not be collected for three fish), liver (n = 22),and muscle (n = 22) of northern pike using methods described byDe Silva et al. (2011). The method detection limits (MDLs) werecalculated using mean method blank response +3 � standard devia-tions (Table 1). For compounds which were not detected in blanks,a value of half the LC/MS/MS limit of detection was substituted.The limit of detection is based on the concentration yielding a sig-nal to noise ratio of 3.

Please cite this article in press as: Houde, M., et al. Induction of gene responses in St. Lawrence River northern pike (Esox lucius) environmentally exposedto perfluorinated compounds. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.01.099

M. Houde et al. / Chemosphere xxx (2013) xxx–xxx 3

2.3. RNA extraction and reverse transcription

RNA from liver, gill, muscle, and blood was extracted with theRNeasy� plus mini kit (QIAGEN, Canada). Reverse transcriptionwas performed with the QuantiTect� Reverse transcription kit(QIAGEN, Canada) following manufacturer’s instructions for a40 lL total volume of cDNA (all details available in the Supplemen-tary Material).

2.4. Quantitative PCR

All qPCR analyses were performed using iQ™ SYBR� GreenSupermix and Mastercycler� ep realplex2 (details in the Supple-mentary Material). Quantification cycle (Cq) values were importedinto GenEx 5.3.1 Enterprise software (MultiD Analyses, AB). Rela-tive gene expression was calculated using reference genes (TBAand UBIQ for blood, UBIQ and PPIA for liver, PPIA and RPS20 forgills and muscle) and efficiency of Cq values corrected. The com-parative threshold method (DDCt; Livak and Schmittgen, 2001)was used for relative quantification and results from fish collectedat the upstream site (Lac des Deux-Montagnes) used as reference.

2.5. Primer design

Expressed gene sequences for Esox lucius were selected basedon their availability in the GenBANK database. Analysis of salmo-nid RNA putative transcripts also indicated a 89% similarity be-tween northern pike and Atlantic salmon (Koop et al. 2008)therefore allowing the use of Salmo salar expressed sequences forprimer design. Genes selected for this study, and respective prim-ers, are presented in Table 2. All primers were designed by theauthors using Primer-BLAST from NCBI (Primer3 with Blast; Rozenand Skaletsky, 2000). The presence of secondary structures wasevaluated using Netprimer (PREMIER Biosoft, CA, USA). For eachgene, two or more primer pairs were evaluated. Primers were man-ufactured by IDT (Coralville, USA).

2.6. Biomarker analyses

Biomarkers were used in order to investigate possible links be-tween gene expression and protein activity. The choice of bio-marker was based on the selected gene sequences and therelevance to the compounds studied. Glutathione S-transferase(GST) and catalase (CAT) activities were quantified in gills. Asemi-quantitative method was also used to estimate vitellogenin(Vtg) protein in plasma. Gill tissue was homogenized in 10 mMHepes-NaOH buffer, pH 7.4, containing 0.1 mM EDTA, 140 mMNaCl, 5 mM KCl. Homogenates were centrifuged and the superna-tant stored at �80 �C. Methods for GST, CAT, and Vtg are detailedin the Supplementary Material. Total protein content was mea-sured by the Bio-Rad protein assay (Bio-Rad, Canada) using theCoomassie Brilliant Blue G-250 dye (absorbance reading at600 nm; Bradford 1976). Standard curve of serum bovine albuminwas used for calibration.

2.7. Data treatment

PFC concentrations were corrected for method blanks. Statisti-cal tests were only performed for compounds with >50% detection.In these cases, concentrations < MDLs were substituted with a ran-domly generated number between 0 and analyte-specific MDL. Siteand tissue comparisons were performed using ANOVAs (JMP 9, SASInstitute Inc., USA). Spearman’s correlations were used to quantifyrelationships between gene expression, biomarker, and chemicaldata. Significance levels were adjusted within each hypothesis

Please cite this article in press as: Houde, M., et al. Induction of gene responsesto perfluorinated compounds. Chemosphere (2013), http://dx.doi.org/10.1016/

using the step-down Bonferroni (Holm) procedure (SAS, version9.2).

3. Results

3.1. Length and age of fish

Fish biological/morphological analyses indicate significant age-related differences between sampling sites. Fish captured in theplume of Montreal’s MWWE (i.e., Ilet Vert) were significantlyyounger than those captured upstream at Lac des Deux-Montagnes(ANOVA; p = 0.0091). Length at age of pike indicated a fastergrowth at Ilet Vert compared to the upstream site (homogeneityof slopes p = 0.022; SM Fig. 2).

3.2. PFC concentrations

Concentrations were significantly lower for all PFCs, with theexception of PFPeDa and PFHxS, in northern pike liver collectedin Lac des Deux-Montagnes compared to fish sampled at Ilet Vert(Table 1). At both sites, PFC concentrations were statistically higherin plasma (p < 0.05) compared to liver and muscle. Contaminationpattern for perfluorinated carboxylic acids (PFCAs) were very sim-ilar between sites with an increase in concentrations from PFNA toPFUnA followed by a decrease with more elevated chain-lengths(Fig. 1). Concentrations of PFOS and PFECHS were significantlyhigher in fish collected in the MWWE. Results indicated a predom-inance of perfluorinated sulfonates (PFSAs) in all fish tissues fromLac des Deux-Montagnes (% of PFCAs ranged from 33% to 43%;PFSAs: 57–67%) and Ilet Vert (PFCAs: 16–21%; PFSAs: 79–84%).PFECHS was not detected in plasma of northern pike from Lacdes Deux-Montagnes and a mean concentration of 5.07 ± 4.72 ng/g w.w. was measured at Ilet Vert.

3.3. Genomic responses

Genomic analyses were carried out for selected genes (Table 2)in blood, liver, muscle, and gill tissue. The relative genomic expres-sion (fold change) in fish tissue is reported in Fig. 2 and SM Table 1.In terms of fold changes, MT, GST, SOD, and CYP1A1 in gill and MTand SOD expression in fish blood were significantly different in fishcollected in the MWWE (ANOVA, p < 0.05; Fig. 2) with SOD in bloodbeing the only down-regulated gene. Vtg, ERa, and CYP3A27expression could only be analyzed in liver tissue. Significant differ-ences in gene expression between gill and muscle tissues were ob-served for CAT, CYP1A1, SOD, GST, and MT. Analyses of geneexpression and contaminant concentrations in tissues indicatedsignificant correlations (p < 0.05) between Vtg gene expression inliver and concentrations of PFTrA, PFTeA, and PFDS in plasma(respective Spearman’s coefficient of 0.850, 0.683, and 0.783).However, once adjusted for multiple testing (step-down Bonfer-roni procedure), p-values were above 0.05.

3.4. Biomarkers

Gill GST and CAT activities in fish at Lac des Deux-Montagnesand Ilet Vert (respectively: 80.9 ± 41.8 and 99.8 ± 28.2 CDNB conju-gate/min/mg prot; 1219 ± 422 and 1188 ± 649 mmol H2O2 con-sumed/min/mg prot) were not significantly different betweensites. Plasma Vtg concentrations were estimated and did not varysignificantly between sites (2.59 ± 2.41 lg Vtg/mg prot at Deux-Montagnes; 1.75 ± 1.00 lg Vtg/mg prot at Ilet Vert). Relationshipanalyses between gene expression and protein activities indicatedthat gene response for CAT in liver and its activity in gills were sig-nificantly correlated (step-down Bonferroni procedure; p < 0.05).

in St. Lawrence River northern pike (Esox lucius) environmentally exposedj.chemosphere.2013.01.099

Table 2Genes, acronyms, and primers used for qPCR analyses in northern pike blood, liver, muscle, and gill tissues.

Acronyms Species Accession number Primers Amplicon size (bp)

Target genesMetallothionein MT Esox lucius X59392.1 Forward AGTTGCTGCTCCTGCTGTCCT 82

Reverse AGCTGGTATCACAGGTCTTGCCCTGlutathione-S-transferase GST Esox lucius BT079891.1 Forward CACCATCACCTATTTTGCTGTCCG 84

Reverse ACCTCCTTCCATTCCTGACCCTCytochrome P450 1A1 CYP1A1 Salmo salar BT045666.1 Forward ACCAAAGCCCCTGCCCATCA 149

Reverse CGCTGCCACTCAGAACAACCACACytochrome P450 3A27 CYP3A27 Salmo salar BT045599.1 Forward TCCGTTGCCGAGGACGATACA 138

Reverse TGCCTGCTTCTTCATTCCGCTCACatalase CAT Salmo salar BT045615.1 Forward TGGTGTGGGACTTCTGGAGCCT 117

Reverse AGGTGTGAGAGCCGTAGCCGTTSuperoxide dismutase SOD Esox lucius BT079033.1 Forward TGGGGCTGGCTGGGTTTGGA 70

Reverse GTGGGTCTTGGTTAGGGCAGGCVitellogenin Vtg Salmo salar AY049952.1 Forward ATGCCCCAAGCCACGGTCTC 115

Reverse TCGCCATCAGCCTTTCCACAGEstrogen receptor a2 ERa Salmo salar NM_001123592.1 Forward GGAAACTCATCTTTGCCCAGGACC 70

Reverse TCAGCCATACCCTCCACACAGTReference genesb2 microglobulin b-micro Esox lucius BT079123.1 Forward TGTTGCCCTTGTTTTCTGCGTGGT 87

Reverse TGGCCGTATTTGCCAGGGTTGCRibosomal protein S20 RPS20 Salmo salar BT060032.1 Forward CCCCTGTTGAGGCTGAGGTTGC 151

Reverse TTGGTGGGCATACGGACTGGTPeptidyl-prolyl cis-trans-peptidylisomerase PPIA Esox lucius BT080015.1 Forward TGCTAAAACTGCTTGGCTGGATGG 116

Reverse TCGCTTTCGTCTTGCCGCTGTubulin a TBA Esox lucius BT079618.1 Forward GGTCACTACACCATCGGCAAGGA 157

Reverse ACAGGCGTTCCATCAGCAGGGAUbiquitin UBIQ Esox lucius BT079424.1 Forward CAGCAGGCGAAAGGAGTGGAGT 89

Reverse TGCCAGGACGGTGGACAAAGT

Fig. 1. PFC mean concentrations (±std dev) detected in northern pike plasmacollected at two sites in the St. Lawrence River, Canada. MDLs (ng/g w.w.) were 0.16(PFOA), 0.08 (PFNA), 0.16 (PFDA), 0.19 (PFUnA), 0.17 (PFDoA), 0.19 (PFTrA), 0.20(PFTeA), 0.02 (PFPeDa), 0.16 (PFHxS), 0.45 (PFOS), 0.13 (PFDS), and 0.21 (PFECHS).�Concentrations significantly different between sites, p < 0.05.

4 M. Houde et al. / Chemosphere xxx (2013) xxx–xxx

Significant correlations were also observed for Vtg gene expressionin liver and Vtg protein activity in plasma (p = 0.046).

4. Discussion

Municipal effluents are composed of household, industrial,commercial, and institutional waste, and also in this case, of storm-waters (CCME, 2006). Results from this study indicated a more ra-pid growth in fish captured in the plume of the urban effluent.Several studies have reported similar observations in various fishspecies (e.g., longnose dace, three-spined and brook sticklebacks)sampled downstream municipal WWTPs (Jeffries et al. 2008; Pott-inger et al. 2011; Tetreault et al. 2012). Elevated productivity asso-ciated with increased water temperature or oxygen availability,and considerable nutrient and organic carbon outputs have beensuggested to explain the pronounced fish growth in effluents (Pott-inger et al. 2011). The morphological differences between sitescould also be the results of genetic adaptation or physiologicalacclimations of fish to chronic chemical exposure (Wirgin andWaldman, 2004).

4.1. PFC concentrations in fish

PFC concentrations detected in northern pike tissues are in therange of results reported in fish from industrial regions of theworld (Houde et al. 2006, 2011). To our knowledge, this is the firstpublished PFC quantification in fish from the St. Lawrence River.This is also the first quantification of PFECHS in St. Lawrence biota.PFECHS is a cyclic PFC used as an erosion inhibitor in aircrafthydraulic fluid recently detected in Great Lakes surface waterand fish samples (De Silva et al. 2011) as well as water, inverte-brates, and snapping turtles collected from sites adjacent anddownstream of an international airport (de Solla et al. 2012). Ef-fects of this newly detected PFC are still unknown in aquaticorganisms.

Significantly higher levels of PFCs, particularly PFOS, werefound in fish from Ilet Vert, located about 4 km downstream of

Please cite this article in press as: Houde, M., et al. Induction of gene responsesto perfluorinated compounds. Chemosphere (2013), http://dx.doi.org/10.1016/

the Montreal’s WWTP point of release in the river. PFC concentra-tions in the eggs of gulls collected from breeding sites from thesame geographical area were among the highest (

PPFSA 299 ng/

g w.w.) reported across Canada (Gebbink et al. 2011). The in-creased presence of PFCs in northern pike and gulls at this locationsuggest that the MWWE may be a source of PFC contamination inthis aquatic ecosystem. The presence of PFECHS in fish tissuedownstream the WWTP also suggest its discharge by the effluent.PFCs are now detected in WWTP effluents on a global scale (Loga-nathan et al. 2007; Furl et al. 2011; Sun et al. 2011; Arvaniti et al.2012; Kim et al. 2012). Differences in perfluorinated sulfonate con-centrations between sites could also be linked to different watermasses. Lac des Deux-Montagnes is located at the mouth of the Ot-

in St. Lawrence River northern pike (Esox lucius) environmentally exposedj.chemosphere.2013.01.099

(A) Blood

(B) Liver

(C) Gill

(D) Muscle

Genes

0

1

2

3

4

5

MT GST CAT SOD

Lac des Deux-MontagnesIlet Vert

Rel

ativ

e ab

unda

nce

0

1

2

3

MT GST CYP1A1 CYP3A CAT SOD Vtg ER

Rel

ativ

e ab

unda

nce

0

1

2

3

4

MT GST CAT SOD CYP1A1

Rel

ativ

e ab

unda

nce

0

1

2

MT GST CAT SOD

Rel

ativ

e ab

unda

nce

Fig. 2. Relative abundance and standard deviation of selected genes analyzed byqPCR in northern pike in (A) blood, (B) liver, (C) gill, and (D) muscle tissue. Catalase(CAT), cytochromes P450 1A1 and 3A27 (CYP1A1, CYP3A27), estrogen receptor(ERa), glutathion-S-transferase (GST), metallothionein (MT), superoxide dismutase(SOD), vitellogenin (Vtg). ⁄Gene expression significantly different between sites(ANOVA, p < 0.05). Deux-Montagnes was used as reference for relative abundancecalculation using the DDCt method.

M. Houde et al. / Chemosphere xxx (2013) xxx–xxx 5

tawa River and Ilet Vert in the water mass coming from the GreatLakes. Numerous studies have reported PFC concentrations inGreat lakes fish (Giesy et al. 2006; Furdui et al. 2007; Houdeet al. 2008; De Silva et al. 2011). Analyses of surface water fromthe Ottawa River have indicated lower PFC concentrations thanthe St. Lawrence River (Scott et al. 2009). No PFC data have beenreported in the literature for biota at these latter sites.

4.2. Variations in genomic expression

The increased transcription of genes in blood (MT) and gills(MT, GST, SOD, and CYP1A1) suggest that northern pike may beresponding to greater environmental stressors downstream theWWTP. Induction of CYP1A1 gene expression in rainbow trout ex-posed to MWWE has been previously reported (Ings et al. 2011)and suggests the presence of substances that can activate the arylhydrocarbon receptor (AhR). The induction of MT is usually associ-ated with the presence of toxic elements. However, the assessmentof metals in these same northern pike has indicated no relation-

Please cite this article in press as: Houde, M., et al. Induction of gene responsesto perfluorinated compounds. Chemosphere (2013), http://dx.doi.org/10.1016/

ships between the element concentrations in muscle and MT geneexpression in gill (Houde, Unpublished). Results also show a signif-icant down-regulation for SOD, an important indicator of oxidativestress, in blood of fish captured in the MMWE. These results illus-trate the complex tissue-specific responses of aquatic organisms toexternal stressors and the importance of tissue selection. Increasedgenomic responses in gills may results from the direct contact ofthis organ with water and toxicants. Gills could constitute an inter-esting tool for ecotoxicological assessment or environmental mon-itoring given the genomic responses and the possibility ofcollecting these samples from fish in a non-invasive manner.

4.3. Relationship between gene expression, protein activities, and PFCs

There are no direct functional correlations between geneexpression and their corresponding proteins as protein productionis not solely regulated by transcription and RNA levels (Bozinovicand Oleksiak, 2010). Furthermore, external stressors such as toxi-cants may increase gene expression and at the same time inhibitthe production or the transport of the related protein (Nikinmaaand Rytkönen, 2011). It is therefore important to examine relation-ships between gene expression and protein levels in order tounderstand how gene function affects the organism.

Statistical analyses between gene expression, protein activities,and contaminant concentrations in fish tissues have indicated pos-sible correlations between Vtg gene expression in liver, Vtg proteinactivity in plasma, and PFTrA, PFTeA, and PFDS plasma concentra-tions. Vitellogenin, a protein produced in liver and transportedthrough blood, is a precursor of lipoproteins and phosphoproteinsthat make up most of the protein content of yolk. Endocrine-dis-ruptive effects have previously been reported for PFCs. PFOS andperfluorooctanoic acid (PFOA) have been reported to induce Vtgin cultured tilapia hepatocytes (Liu et al. 2007) and PFOA to disturbestrogen activity in mature male rare minnows by inducing hepaticVtg (Wei et al. 2008). Hepatic Vtg gene expression was also up-reg-ulated in zebrafish long-term exposure to PFOS (Du et al. 2009) andVtg concentration was induced in common carp experimentallyexposed to PFOA for 4 d (Kim et al. 2010). Increases of Vtg-like pro-teins have been reported in freshwater mussels exposed for 14 d tothe Montreal’s effluent (Gagné et al. 2011). Conclusions from thatstudy suggest that the continuous exposure of mussels to the Mon-treal effluent leads to neuroendocrinial alterations and to oxidativestress. Elsewhere, rainbow trout exposed to MWWE exhibited ele-vated plasma Vtg levels (Ings et al. 2011) and fathead minnows ex-posed to MWWE exhibited significant induction of the Vtg geneexpression associated with increased Vtg plasma levels and reduc-tion of competitive behavior (Garcia-Reyero et al. 2011). Relation-ships observed between PFOS and gene expression (CYP1A1 andGST) in tissue of Atlantic salmon fed a PFOS/PFOA diet were not ob-served in this study (Mortensen et al. 2011). Results from the pres-ent study suggest that other PFCs, such as PFTrA, PFTeA, and PFDS,may also be estrogenic and this should further be investigated.

In conclusion, differences in genomic responses have beenfound in northern pike residing downstream a major WWTP. Re-sults indicated relationships between Vtg gene expression in liver,estimated Vtg protein activity in plasma, and PFTrA, PFTeA, andPFDS plasma concentrations recommending more in-depth inves-tigation. The use of non-model organisms, such as northern pike,is of high value for ecotoxicological studies but limits the possiblegenomic analyses. Next-generation sequencing may soon allowextensive genomic analyses of non-model aquatic species (Mehin-to et al. 2012). Additional well-designed experiment with pheno-type anchoring, using endocrine and histopathologicalparameters for example, would enable to scale up from geneexpression and cellular indicators to organism level effects. Com-plementary experimental exposures are planned to better under-

in St. Lawrence River northern pike (Esox lucius) environmentally exposedj.chemosphere.2013.01.099

6 M. Houde et al. / Chemosphere xxx (2013) xxx–xxx

stand impacts and mode of action of individual and mixtures ofPCFs in fish.

Acknowledgements

The authors would like to thank Claude Lessard, ChristianBlaise, Timothée Debenest, Niraj Maulkhan, and Pierre Gagnonfor their support during the course of this project. This projectwas funded by the Strategic Technology Application of Genomicsin the Environment (Environment Canada).

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.chemosphere.2013.01.099.

References

Arvaniti, O.S., Ventouri, E.I., Stasinakis, A.S., Thomaidis, N.S., 2012. Occurrence ofdifferent classes of perfluorinated compounds in Greek wastewater treatmentplants and determination of their solid–water distribution coefficients. J.Hazard. Mater. 239–240, 24–31.

Bernatchez, L., Giroux, M., 2000. Les poissons d’eau douce du Québec et leurrépartition dans l’Est du Canada. Édition Broquet, Ottawa. p. 350.

Bozinovic, G., Oleksiak, M.F., 2010. Embryonic gene expression among pollutantresistant and sensitive Fundulus heteroclitus populations. Aquat. Toxicol. 98,221–229.

Bradford, M.M., 1976. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye binding.Anal. Biochem. 72, 248–254.

CCME – Canadian Council of Ministers of the Environment, 2006. Municipalwastewater in Canada, Municipal wastewater effluent development committee,pp. 1–9.

Casselman, J.M., Lewis, C.A., 1996. Habitat requirements of northern pike (Esoxlucius). Can. J. Fish. Aquat. Sci. 53, 161–174.

De Silva, A.O., Spencer, C., Scott, B.F., Backus, S., Muir, D.C.G., 2011. Detection of acyclic perfluorinated acid, perfluoroethylcyclohexane sulfonate, in the GreatLakes of North America. Environ. Sci. Technol. 45, 8060–8066.

de Solla, S.R., De Silva, A.O., Letcher, R.J., 2012. Highly elevated levels ofperfluorooctane sulfonate and other perfluorinated acids found in biota andsurface water downstream of an international airport, Hamilton, Ontario,Canada. Environ. Int. 39, 19–26.

Du, Y., Shi, X., Liu, C., Yu, K., Zhou, B., 2009. Chronic effects of water-borne PFOSexposure on growth, survival and hepatotoxicity in zebrafish: a partial life-cycletest. Chemosphere 74, 723–729.

Fujii, S., Polprasert, C., Tanaka, S., Lien, N.P.H., Qiu, Y., 2007. New POPs in the waterenvironment: distribution, bioaccumulation and treatment of perfluorinatedcompounds – a review paper. J. Wat. Sup. Res. Technol. AQUA 56, 313–326.

Furdui, V.I., Stock, N.L., Ellis, D.A., Butt, C.M., Whittle, D.M.I., Crozier, P.W., Reiner,E.J., Muir, D.C.G., Mabury, S.A., 2007. Spatial distribution of perfluoroalkylcontaminants in lake trout from the Great Lakes. Environ. Sci. Technol. 41,1554–1559.

Furl, C.V., Meredith, C.A., Strynar, M.J., Nakayama, S.F., 2011. Relative importance ofwastewater treatment plants and non-point sources of perfluorinatedcompounds to Washington state rivers. Sci. Total Environ. 409, 2902–2907.

Gagné, F., André, C., Cejka, P., Hausler, R., Fournier, M., 2011. Evidence ofneuroendocrine disruption in freshwater mussels exposed to municipalwastewaters. Sci. Total Environ. 409, 3711–3718.

Garcia-Reyero, N., Lavelle, C.M., Escalon, B.L., Martinovic, D., Kroll, K.J., Sorensen,P.W., Denslow, N.D., 2011. Behavioral and genomic impacts of a wastewatereffluent on the fathead minnow. Aquat. Toxicol. 101, 38–48.

Gebbink, W.A., Letcher, R.J., Burgess, N.M., Champoux, L., Elliott, J.E., Hebert, C.E.,Martin, P., Wayland, M., Weseloh, D.V.C., Wilson, L., 2011. Perfluoroalkylcarboxylates and sulfonates and precursors in relation to dietary source tracersin the eggs of four species of gulls (Larids) from breeding sites spanning Atlanticto Pacific Canada. Environ. Int. 37, 1175–1182.

Please cite this article in press as: Houde, M., et al. Induction of gene responsesto perfluorinated compounds. Chemosphere (2013), http://dx.doi.org/10.1016/

Giesy, J.P., Mabury, S.A., Martin, J.W., Kannan, K., Jones, P.D., Newsted, J.L., Coady, K.,2006. Perfluorinated compounds in the great lakes. Hdb Env. Chem., vol. 5, PartN, Springer-Verlag, Berlin, pp. 391–438.

Houde, M., De Silva, A.O., Muir, D.C.G., Letcher, R.J., 2011. Monitoring ofperfluorinated compounds in aquatic biota: an updated review. Environ. Sci.Technol. 45, 7962–7973.

Houde, M., Gertje, C., Small, J., Backus, S., Wang, X., Alaee, M., Muir, D.C.G., 2008.Fractionation and bioaccumulation of perfluorooctane sulfonate (PFOS) isomersin a Lake Ontario food web. Environ. Sci. Technol. 42, 9397–9403.

Houde, M., Martin, J.W., Letcher, R.J., Solomon, K.R., Muir, D.C.G., 2006. Biologicalmonitoring of polyfluoroalkyl substnaces: a review. Environ. Sci. Technol. 40,3463–3473.

Ings, J.S., Servos, M.R., Vijayan, M.M., 2011. Hepatic transcriptomics and proteinexpression in rainbow trout exposed to municipal wastewater effluent. Environ.Sci. Technol. 45, 2368–2376.

Jeffries, K.M., Jackson, L.J., Peters, L.E., Munkittrick, K.R., 2008. Changes inpopulation, growth, and physiological indices of longnose dace (Rhinichthyscataractae) in the Red Deer River, Alberta, Canada. Arch. Environ. Contam.Toxicol. 55, 639–651.

Kim, S.-K., Im, J.-K., Kang, Y.-M., Jung, S.-Y., Kho, Y.L., Zoh, K.-D., 2012. Wastewatertreatment plants (WWTPs)-derived national discharge loads of perfluorinatedcompounds (PFCs). J. Hazard. Mat. 201–202, 82–91.

Kim, W.-K., Lee, S.-K., Jung, J., 2010. Integrated assessment of biomarker responsesin common carp (Cyprinus carpio) exposed to perfluorinated organiccompounds. J. Hazard. Mat. 180, 395–400.

Koop, B.F., von Schalburg, K.R., Leong, J., Walker, N., Lieph, R., Cooper, G.A., Robb, A.,Beetz-Sargent, M., Holt, R.A., Moore, R., Brahmbhatt, S., Rosner, J., Rexroad III,C.E., McGowan, C.R., Davidson, W.S., 2008. A salmonid EST genomic study:genes, duplications, phylogeny and microarrays. BMC Genomics 9, 545.

Liu, C., Du, Y., Zhou, B., 2007. Evaluation of estrogenic activities and mechanism ofaction of perfluorinated chemicals determined by vitellogenin induction inprimary cultured tilapia hepatocytes. Aquat. Toxicol. 85, 267–277.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data usingreal-time quantitative PCR and the 2-DDCT method. Methods 25, 402–408.

Loganathan, B.G., Sajwan, K.S., Sinclair, E., Senthil Kumar, K., Kannan, K., 2007.Perfluoroalkyl sulfonates and perfluorocarboxylates in two wastewatertreatment facilities in Kentucky and Georgia. Water Res. 41, 4611–4620.

Mehinto, A.C., Martyniuk, C.J., Spade, D.J., Denslow, N.D., 2012. Applications of next-generation sequencing in fish ecotoxicogenomics. Front. Gene. 3, 1–10.

Mortensen, A.S., Letcher, R.J., Cangialosi, M.V., Chu, S., Arukwe, A., 2011. Tissuebioaccumulation patterns, xenobiotic biotransformation and steroid hormonelevels in Atlantic salmon (Salmo salar) fed a diet containing perfluorooctanesulfonic or perfluorooctane carboxylic acids. Chemosphere 83, 1035–1044.

Nikinmaa, M., Rytkönen, K.T., 2011. Functional genomics in aquatic toxicology – donot forget the function. Aquat. Toxicol. 105, 16–24.

Pottinger, T.G., Cook, A., Jürgens, M.D., Rhodes, G., Katsiadaki, I., Balaam, J.L., Smith,A.J., Matthiessen, P., 2011. Effects of sewage effluent remediation on body size,somatic RNA: DNA ratio, and markers of chemical exposure in three-spinedsticklebacks. Environ. Int. 37, 158–169.

Rozen, S., Skaletsky, H.J., 2000. Primer3 on the WWW for general users and forbiologist programmers. Bioinformatics Methods and Protocols Methods inMolecular Biology. In: Misener, S., Krawetz, S.A. (Eds.). Humana Press, NJ, USA,pp. 365–386.

Scott, B.F., Spencer, C., Lopez, E., Muir, D.C.G., 2009. Perfluorinated alkyl acidconcentrations in Canadian rivers and creeks. Water Qual. Res. J. Can. 44, 263–277.

Seeb, J.E., Seeb, L.W., Oates, D.W., Utter, F.M., 1987. Genetic variation and postglacialdispersal of populations of northern pike (Esox lucius) in North America. Can. J.Fish. Aquat. Sci. 44.

Senanan, W., Kapuscinski, A.R., 2000. Genetic relationships among populations ofnorthern pike (Esox lucius). Can. J. Fish. Aquat. Sci. 57, 391–404.

Sun, H., Li, F., Zhang, T., Zhang, X., He, N., Song, Q., Zhao, L., Sun, L., Sun, T., 2011.Perfluorinated compounds in surface waters and WWTPs in Shenyang, China:mass flows and source analysis. Water Res. 45, 4483–4490.

Tetreault, G.R., Bennett, C.J., Cheng, C., Servos, M.R., McMaster, M.E., 2012.Reproductive and histopathological effects in wild fish inhabiting an effluent-dominated stream, Wascana Creek, SK, Canada. Aquat. Toxicol. 110–111, 149–161.

Wei, Y., Liu, Y., Wang, J., Tao, Y., Dai, J., 2008. Toxicogenomic analysis of the hepaticeffects of perfluorooctanoic acid on rare minnows (Gobiocypris rarus). Toxicol.App. Pharmacol. 226, 285–297.

Wirgin, I., Waldman, J.R., 2004. Resistance to contaminants in North American fishpopulations. Mutat. Res. 552.

in St. Lawrence River northern pike (Esox lucius) environmentally exposedj.chemosphere.2013.01.099