Sequence and In Vitro Function of Chicken, Ring-Necked Pheasant,and Japanese Quail AHR1 Predict In Vivo Sensitivity to DioxinsReza Farmahin,†,‡ Dongmei Wu,‡ Doug Crump,‡ Jessica C. Herve,́‡ Stephanie P. Jones,‡ Mark E. Hahn,§

Sibel I. Karchner,§ John P. Giesy,∥,⊥,∇ Steven J. Bursian,○ Matthew J. Zwiernik,○

and Sean W. Kennedy*,†,‡

†Centre for Advanced Research in Environmental Genomics, Department of Biology, University of Ottawa, Ottawa, Ontario K1N6N5, Canada‡Environment Canada, National Wildlife Research Centre, Ottawa, Ontario K1A 0H3, Canada§Department of Biology, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, United States∥Department of Veterinary Biomedical Sciences and Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan S7N5B3, Canada⊥Department of Zoology and Center for Integrative Toxicology, Michigan State University, East Lansing, Michigan 48824, UnitedStates∇Department of Biology & Chemistry, and State Key Laboratory for Marine Pollution, City University of Hong Kong, Kowloon,Hong Kong, SAR, China○Department of Animal Science, Michigan State University, East Lansing, Michigan 48824, United States

*S Supporting Information

ABSTRACT: There are large differences in sensitivity to the toxic and biochemicaleffects of dioxins and dioxin-like compounds (DLCs) among vertebrates. Previously,we demonstrated that the difference in sensitivity between domestic chicken (Gallusgallus domesticus) and common tern (Sterna hirundo) to aryl hydrocarbon receptor 1(AHR1)-dependent changes in gene expression following exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) is based upon the identities of the amino acidsat two sites within the ligand binding domain of AHR1 (chicken - highly sensitive;Ile324_Ser380 vs common tern - 250-fold less sensitive than chicken;Val325_Ala381). Here, we tested the hypotheses that (i) the sensitivity of otheravian species to TCDD, 2,3,4,7,8-pentachlorodibenzofuran (PeCDF), and 2,3,7,8-tetrachlorodibenzofuran (TCDF) is also determined by the amino acids at sites thatare equivalent to sites 324 and 380 in chicken, and (ii) Ile324_Ala380 andVal324_Ser380 genotypes confer intermediate sensitivity to DLCs in birds. Wecompared ligand-induced transactivation function of full-length AHR1s from chicken,common tern, ring-necked pheasant (Phasianus colchicus; Ile324_Ala380) andJapanese quail (Coturnix japonica; Val324_Ala380), and three Japanese quail AHR1mutants. The results support our hypothesis that avian species can be grouped intothree general classes of sensitivity to DLCs. Both AHR1 genotype and in vitrotransactivation assays predict in vivo sensitivity. Contrary to the assumption that TCDD is the most potent DLC, PeCDF wasmore potent than TCDD at activating Japanese quail (13- to 26-fold) and common tern (23- to 30-fold) AHR1. Our resultssupport and expand previous in vitro and in vivo work that demonstrated ligand-dependent species differences in AHR1 affinity.The findings and methods will be of use for DLC risk assessments.

■ INTRODUCTIONPolychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans(PCDFs), and coplanar polychlorinated biphenyls (PCBs) arechemicals collectively referred to as dioxins or dioxin-likecompounds (DLCs) due to their structural similarities andcommon mechanisms of biochemical and toxic action. Thereare large difference in sensitivity to DLCs among species andstrains of animals. For approximately 35 years there has beeninterest in discovering the fundamental mechanisms that

account for differences in sensitivity among species. In ouropinion, an understanding of the reasons for differentialsensitivity to DLC toxicity among species will lead todevelopment of better methods for environmental and

Received: December 8, 2011Revised: January 17, 2012Accepted: January 23, 2012Published: January 23, 2012

Article

pubs.acs.org/est

© 2012 American Chemical Society 2967 dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−2975

pubs.acs.org/est

human health risk assessments of this class of environmentalcontaminants.There is considerable evidence that most, if not all, toxic

effects of DLCs are initiated by activation of the arylhydrocarbon receptor (AHR), a ligand-dependent transcriptionfactor, and member of the basic-helix−loop−helix (bHLH)-PAS family of genes. AHR-deficient mice are highly resistant to2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) toxicity,1 andearlier studies with mouse strains,2 mice,3 rats,4 and birds5,6

indicate that the molecular properties of the AHR itself andAHR-mediated gene expression provide insight into mecha-nisms underlying differential sensitivity to DLCs.7 Some of thelargest differences in sensitivity to DLCs occur among speciesof birds, with the domestic chicken (Gallus gallus domesticus)being considerably more sensitive than other avian species toseveral AHR-mediated effects, including induction of cyto-chrome P4501A (CYP1A), hepatotoxicity, edema, andembryolethality.5

There are at least two isoforms of the AHR (AHR1 andAHR2) in birds. The expression of AHR1 was reported to beseveral-fold greater than that of AHR2 in avian liver,8,9 andAHR1 appears to be more transcriptionally active than AHR2.9

Karchner and colleagues discovered that the identities of twoamino acids within the ligand-binding domain (LBD) of AHR1(Ile324 and Ser380 in chicken; Val325 and Ala381 in commontern [Sterna hirundo]) confer differences in sensitivity toTCDD between the more sensitive chicken and the relativelyinsensitive common tern.6 Using site-directed mutagenesis, thatstudy showed that replacing Val and Ala in the common ternAHR1 with the corresponding chicken amino acids impartedhigh-affinity binding of TCDD to the tern AHR1. In addition, areporter gene assay that used cells transfected with wild-type ormutant AHR1 constructs confirmed the importance of theidentities of amino acids at sites 324/325 and 380/381 forAHR1 function. Subsequently, the AHR1 LBDs of several avianspecies were sequenced.5 Based on these sequences, and datafrom in vitro and egg injection studies, the authorshypothesized that birds could be grouped into three generalclasses of DLC sensitivity: type 1 (chicken-like; Ile324_S-er380), type 2 (ring-necked pheasant [Phasianus colchicus]-like;Ile324_Ala380), or type 3 (Japanese quail [Coturnix japonica]-like; Val324_Ala380).Here we report the results of in vitro studies that were

designed to examine, in greater detail, the role of amino acids atLBD sites 324/325 and 380/381. An efficient reporter geneassay was developed for measuring AHR1-dependent geneexpression in monkey kidney cells (the COS-7 line, which haslow endogenous AHR1 expression) that were transientlytransfected with seven different avian AHR1 constructs andexposed to TCDD, 2,3,4,7,8-pentachlorodibenzofuran(PeCDF), or 2,3,7,8-tetrachlordibenzofuran (TCDF). TheAHR1 constructs were (a) full-length chicken, ring-neckedpheasant, Japanese quail, or common tern AHR1 or (b) threemutants of Japanese quail AHR1 that contained amino acidmutations at sites 324 and/or 380. Comparison of the resultsobtained with the reporter gene assay with results from arecently published in ovo study of toxicity support thehypothesis that the genotype of the AHR1 LBD is predictiveof sensitivity to DLCs in birds. The present work providesfurther evidence that the classification of birds into groupsbased on sensitivity to DLCs would be of considerable use forenvironmental risk assessments.

■ EXPERIMENTAL SECTIONCloning and Sequence Analysis of AHR1 cDNA. The

methods for cloning and sequence analysis of ring-neckedpheasant and Japanese quail AHR1 cDNA are indicated below;the methods for chicken and common tern AHR1 cloning andsequencing are described elsewhere.6

i. Total RNA and mRNA Isolation. Livers of ring-neckedpheasant and Japanese quail were obtained from birds bred andhoused at Couvoir Simetin (Mirabel, QC, Canada). Immedi-ately after the birds were euthanized, livers were collected,frozen in liquid nitrogen, and stored at −80 °C. Total RNA wasextracted using RNeasy kits (Qiagen, Mississauga, ON,Canada), and mRNA was isolated by use of Oligotex kits(Qiagen) according to procedures provided by the manufac-turer. Approximately 0.5 mg of total RNA was used to obtainthe concentrated mRNA solutions used subsequently.

ii. cDNA Synthesis. Single-stranded cDNA was synthesizedfrom mRNA (1 μg) by use of a modified lock-docking oligo(dT) primer that contained two degenerate nucleotides at the3′ end. These nucleotides position the primer at the beginningof the poly A+ tail, thus eliminating the 3′ heterogeneityinherent with conventional oligo (dT) priming. Double-stranded cDNA for the AHR1 of ring-necked pheasant andJapanese quail was synthesized by use of Marathon cDNAamplification kits (Clontech, Foster City, CA). Followingdouble-stranded cDNA synthesis, blunt ends were created, andMarathon adaptors were ligated to both ends of the double-stranded cDNA.

iii. 3′ and 5′ Rapid Amplification of cDNA Ends (RACE). 3′and 5′ RACE procedures were performed using oligonucleotideprimers (Table S1) designed to recognize pheasant andJapanese quail sequences coupled with adaptor primers (AP1or AP2). PCR products were run on 1.2% agarose gels,followed by band excision and purification with Nucleotrap gelextraction kits (Clontech).

iv. Ligation, Transformation, Sequencing, and Phyloge-netic Analysis. All DNA fragments were cloned into thepGEM-T Easy vector (Promega, Madison, WI, USA) andtransformed into JM109 competent cells. Plasmid DNA waspurified with QIAprep spin miniprep kits (Qiagen) andsequenced on an Applied Biosystems 3730 DNA Analyzer atthe Ontario Hospital Research Institute (Ottawa, ON, Canada).Reported sequences indicate consensus between at least threedifferent colonies at the nucleotide level. Sequences wereanalyzed using Sequencher 4.8 (Gene Code Corporation, AnnArbor, MI, USA) and aligned using ClustalW. The tree wasconstructed using the minimum evolution criterion inPAUP*4.0 with the neighbor-joining tree as the starting tree.Only the N-terminal half of the protein (∼400 amino acids)was used because the alignment is less certain in the C-terminalhalf.

Expression Constructs. Full-length chicken, ring-neckedpheasant, and Japanese quail AHR1 cDNAs were amplified withPfu Ultra High-Fidelity DNA polymerase (Stratagene, La Jolla,California, USA) by use of CS-forward and CS-reverse primers(Table S1). For chicken AHR1 amplification, an AHR1construct (pcDNA-GgAHR)6 was used as the template. Atthe 5′-end of the forward primer, a CACC 5′-overhang and theKozak consensus sequence (CACCATGA) were added toensure directional ligation into the pENTR/D-TOPO vector(Invitrogen, Burlington, ON, Canada) and efficient initiation oftranslation, respectively. At the 3′-end of the reverse primer, the

Environmental Science & Technology Article

dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−29752968

TGA stop codon was removed to allow downstream V5 epitopeexpression. PCR products were run on 1.2% agarose gels andbands were excised. Bands were purified with the Nucleotrapgel extraction kit, ligated into pENTR/D-TOPO vectors, andsubcloned into pcDNA 3.2/V5-DEST vectors (Invitrogen).Chicken, ring-necked pheasant, and Japanese quail AHR1constructs were sequenced to ensure accuracy. Otherconstructs used were as follows: common tern AHR1(pcDNA-ShAHR; Karchner et al. (2006)), Renilla luciferasevector (phRL-CMV; Promega), CYP1A5 reporter vector (GL4-ccCYP1A5), and cormorant ARNT1 (pcDNA-ccArnt1). GL4-ccCYP1A5 and pcDNA-ccArnt1 were kindly provided by Dr.Hisato Iwata, Ehime University, Japan.9,10

Site-Directed Mutagenesis of Japanese Quail AHR1.Three mutant Japanese quail AHR1 constructs were preparedusing site-directed mutagenesis to change the amino acidresidues at sites 324 and 380 (schematic shown in Figure S1);the mutant forms were V324I_A380S (chicken-like mutant),A380S (unknown mutant; to our knowledge, no avian speciessequenced to date have this genotype), and V324I (pheasant-like mutant). Primer sequences are indicated in Table S1; allmutations were constructed by use of Pfu Ultra High-FidelityDNA polymerase. PCR products were treated with DpnIendonuclease (New England Biolabs, Beverly, MA, USA) andthen transformed to MAX Efficiency DH5α Competent Cells(Invitrogen). Full-length sequencing of each of the threemutant constructs was conducted to ensure that each containedthe desired sequence.AHR1 Expression in COS-7 Cells. i. Preparation of

TCDD, PeCDF, and TCDF Solutions. A detailed description ofthe preparation of TCDD, PeCDF, and TCDF solutions isprovided elsewhere.11 In brief, serial dilutions of TCDD,PeCDF, and TCDF were prepared from stock solutions indimethyl sulfoxide (DMSO), and concentrations weredetermined by isotope dilution following EPA method 161312

by high-resolution gas chromatography high-resolution massspectrometry.ii. Cell Culture, Transfection, and Luciferase Assay. COS-7

cells were provided by Dr. R. Hache ́ (University of Ottawa,Ottawa, ON, Canada) and were maintained at 37 °C under 5%CO2 in Dulbecco’s modified Eagle’s medium (DMEM,Invitrogen) supplemented with 10% fetal bovine serum (FBS;Wisent, St. Bruno, QC, Canada), 1% MEM nonessential aminoacids (Invitrogen), and 1% penicillin-streptomycin (Invitrogen;10 000 unit/mL penicillin, 10 000 μg/mL streptomycin). Cellswere plated at a concentration of 10 000 cells/well in 96-wellplates in dextran-coated and charcoal-treated DMEM (In-vitrogen) supplemented with 10% charcoal stripped FBS and 1xpenicillin-streptomycin (Invitrogen). Transfection was carriedout 18 h after plating. DNA and Fugene 6 transfection reagent(Roche, Laval, QC, Canada) were diluted in Opti-MEM(Invitrogen). A total of approximately 50 ng of DNA wascomplexed with 0.15 μL of Fugene 6 transfection reagent, andthe mixture (5 μL) was then added to each well. A Renillaluciferase vector (phRL-CMV), carrying the cytomegalovirus(CMV) promoter, was used as a transfection control. The ratioof firefly luciferase luminescence units to Renilla luciferaseluminescence units is defined throughout this study as theluciferase ratio. The amounts of transfected expression vectorswere 8 ng of chicken, ring-necked pheasant, Japanese quail,common tern, or mutated avian AHR1, 1.55 ng of cormorantARNT1, 5 ng of CYP1A5 reporter construct, and 0.75 ng ofRenilla luciferase vector. The total amount of transfected DNA

was kept constant at 50 ng by adding salmon sperm DNA(Invitrogen). For each 96-well plate, COS-7 cells weretransfected with (a) one of the four wild-type constructs(chicken, ring-necked pheasant, Japanese quail, or commontern) or one of the three mutant Japanese quail constructs, (b)the ARNT construct, (c) the CYP1A5 reporter construct, and(d) Renilla luciferase vector. Cells were dosed 5 h aftertransfection with DMSO or DMSO solutions of TCDD,PeCDF, or TCDF at 0.05% final DMSO concentration. Cellswere removed from the incubator 18−20 h after dosing tomeasure luciferase activity. Luminescence was measured usingDual-Glo luciferase assay kits (Promega) in a LuminoSkanAscent luminometer (Thermo Fisher Scientific Inc., Waltham,MA, USA).

iii. AHR1 Protein Expression. COS-7 cells were plated in 6-well plates and transfected with wild-type or mutant avianAHR1 by use of Fugene 6 (Roche). Forty-eight hours aftertransfection, cells were lysed in IPH buffer (50 mM Tris [pH7.4], 150 mM NaCl, 0.5% Nonidet P-40, 5 mM EDTA, 2 mMdithiothreitol, and protease inhibitor cocktail; Roche). Proteinconcentration was determined using Bradford reagent (Bio-Rad, Mississauga, ON, Canada). Cell lysates (30 μg aliquots)were resolved by SDS-PAGE and transferred to nitrocellulosemembranes (Invitrogen). Anti-V5-HRP antibody (Invitrogen)was used at 1:5000 dilution for detecting V5-AHR1. Anti-β-actin-peroxidase (Sigma-Aldrich, Oakville, ON, Canada) wasused at 1:25000 as a loading control. The blots were thenvisualized by enhanced chemiluminescence.

Concentration−Response Curves. Duplicate concentra-tion−response curves were obtained for each DLC and eachAHR1 construct, and the duplicate curves are referred to asStudy 1 and Study 2. For each 96-well plate, the luciferase ratiosobtained from four replicate DMSO-treated wells and fourwells treated with each concentration of TCDD, PeCDF, orTCDF were determined. The mean luciferase ratio obtained forDMSO-treated wells was subtracted from the luciferase ratiovalues obtained for DLC-treated wells. GraphPad (GraphPadPrism 5.0 software; San Diego, CA, USA) was used for curvefitting, and data were fit to a four parameter logistic model.13

Relative Sensitivity (Interspecies Comparisons). Theslopes and maximum responses (efficacies) of the concen-tration−response curves (luciferase ratio vs log DLCconcentration) were not usually identical among the AHR1constructs, and relative sensitivity (ReS) values weredetermined at three points (EC20, EC50, and EC80) on theconcentration−response curves. Because the chicken is moresensitive than most, if not all, avian species to the toxic andbiochemical effects of potent DLCs, ReS values for wild-typeand mutant AHR1 constructs were expressed relative to thesensitivity observed in cells containing the chicken AHR1construct. For example, the ReS for the 50% response (ReS50)of Japanese quail AHR1 was as follows: chicken TCDD EC5 0(0.21 nM) ÷ quail TCDD EC50 (21 nM) = 0.01. The generaltypes of concentration−response curves obtained and thelocations of EC20, EC50, and EC80 values are shown in FigureS2.

Relative Potency (Intercompound Comparisons). Thesystematic framework proposed by Villeneuve et al.14 was usedto obtain ReP values. For each construct, the luciferase ratiodata were first converted to the percent maximum luciferaseratio elicited by TCDD (% TCDDmax). The normalizedluciferase ratio data were then plotted against log DLCconcentration and curves were fitted using a four parameter

Environmental Science & Technology Article

dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−29752969

logistic equation.13 Three points (Y20,50,80), which correspondedto EC20, 50 and 80 values that were observed for TCDD, were usedto calculate ReP values for PeCDF and TCDF. For example,the PeCDF ReP50 value for the Japanese quail construct was asfollows: quail TCDD EC50 ÷ quail PeCDF Y50. The generaltypes of concentration−response curves obtained and thelocations of EC20, EC50, EC80, Y20, Y50, and Y80 values are shownin Figure S3.In Vitro Transcription and Translation (IVTT) of Avian

AHR1. FluoroTect GreenLys tRNA-labeled AHR1 proteins forchicken, ring-necked pheasant, Japanese quail, and mutant(V324I, A380S, and V324I_A380S) constructs were synthe-sized by use of TnT Quick-Coupled Reticulocyte LysateSystems kits (Promega) following the manufacturer’s instruc-tions. SDS-PAGE was performed using 20 μL of fluorescentlylabeled TnT reaction products and 10% SDS polyacrylamidegels. A Typhoon 9210 Imager (Molecular Dynamics Inc.,Sunnyvale, CA, USA) was used to detect the fluorescent bands(excitation: 532 nm; emission: 580 nm).

■ RESULTSAHR1 Characterization. The full-length sequences for

ring-necked pheasant AHR1 (RNP-AHR1; accession number:HM053553) and Japanese quail AHR1 were determined(Figure S4). Two allelic variants were found in Japanese quailAHR1 (JQ-AHR1*1; accession number: HM053555 and JQ-AHR1*2; accession number: HM053554). The proteinsencoded by RNP-AHR1, JQ-AHR1*1, and JQ-AHR1*2contained 860 aa (96.6 kDa), 857 aa (96.1 kDa), and 859 aa(96.3 kDa), respectively, with JQ-AHR1*2 containinginsertions of two glutamine residues in the transactivationdomain (positions 640 and 641) that were not present in JQ-AHR1*1. The amino acid sequences of chicken, ring-neckedpheasant, and Japanese quail AHR1 have 96% or greateridentity overall (Table S2). The full-length AHR1 sequencesreported here are consistent with the partial sequences reportedpreviously.5,6 The amino acid residues at two key positionswithin the LBD were as follows: chicken: Ile324_Ser380; ring-necked pheasant: Ile324_Ala380; Japanese quail: Val324_A-

Figure 1. Concentration-dependent effects of TCDD, PeCDF, and TCDF on AHR1 response element (AHRE)-driven luciferase reporter geneactivity in COS-7 cells that were transfected with either (a) full-length AHR1 constructs of chicken, ring-necked pheasant, Japanese quail, commontern or (b) mutant constructs of Japanese quail AHR1. COS-7 cells were also transfected with cormorant ARNT1, CYP1A5 reporter vector, andRenilla luciferase vector. Cells were exposed to DMSO or graded concentrations (0−300 nM) of TCDD, PeCDF, or TCDF for 18−20 h andluciferase ratios (the ratio of firefly luciferase luminescence units to Renilla luciferase luminescence units) were determined. Two studies were donefor each DLC and each construct. Data points indicate the mean luciferase ratio determined in four wells; bars represent SE of the mean values.

Environmental Science & Technology Article

dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−29752970

la380; common tern: Val325_Ala381. IVTT-synthesized full-length chicken, ring-necked pheasant, Japanese quail AHR1proteins, and mutant Japanese quail AHR1 proteins migrated tosimilar positions on SDS polyacrylamide gels and wereexpressed to approximately the same degree (Figure S5), andAHR1 protein expression levels in COS-7 cells transfected withwild-type or mutant AHR1 constructs were similar (Figure S6).Finally, phylogenetic analysis of the avian AHR1 sequenceswith AHRs of selected vertebrate species indicates that theyform a monophyletic group with other AHR1 forms (FigureS7).AHR1 Expression in COS-7 Cells. i. Relative Sensitivity

(Interspecies Comparisons). The concentration-dependenteffects of TCDD, PeCDF, and TCDF on AHR1-associatedreporter gene expression in COS-7 cells that were transfectedwith (a) constructs encoding full-length chicken, ring-necked

pheasant, Japanese quail, common tern or (b) mutant AHR1sare shown in Figure 1. The ECx and ReSx values for each DLCand each AHR1 construct are indicated in Table 1; threegeneral sensitivity classes of avian species (chicken-like,pheasant-like, or quail-like) are highlighted in light, moderate,or dark gray, respectively. The range of ReS values (ReS20−80) isof particular value for comparing the AHR1 constructs becauseit accounts for differences in individual EC values (i.e., EC20,EC50, and EC80) that occurred because the concentration−response curves did not always have identical slopes andmaximum responses. Among the wild-type AHR1s, the rank-order of ReS in TCDD-exposed cells was chicken > pheasant >quail ≈ common tern (the ReS average values were as follows:chicken = 1.0, pheasant = 0.13, quail = 0.011, tern = 0.014).Based on these ReS average values, pheasant AHR1 was 8-foldless sensitive, quail AHR1 was 90-fold less sensitive, and tern

Table 1. COS-7 Cells Transfected with Chicken, Ring-Necked Pheasant, Japanese Quail, Common Tern, or Three JapaneseQuail Mutant AHR1s Were Exposed to Graded Concentrations of TCDD, PeCDF, or TCDFa

aLuciferase ratio values were determined and data were fit to a four-parameter logistic model using GraphPad Prism 5. Concentration-responsecurves were plotted for each curve, and the 20%, 50%, and 80% effective concentrations (EC20, 50, and 80) were calculated. Two separate experiments,each of which used a 96-well plate to generate a concentration-response curve, were conducted for each DLC; four replicates of each concentrationof each DLC were used. Relative sensitivity (ReS) is defined as EC20, 50, or 80 (compound A) in chicken ÷ EC20, 50, or 80 (compound A) in the species of interest.Three general classes of avian species (chicken-like, pheasant-like, or quail-like) are highlighted in light, moderate, or dark grey, respectively. TheA380S genotype (unknown mutant) is not highlighted.

Environmental Science & Technology Article

dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−29752971

AHR1 was 70-fold less sensitive than chicken AHR1 of chickenexposted to TCDD (Figure S8). The difference in sensitivity toPeCDF among the constructs was not as large as that observedfor TCDD. For example, the quail AHR1 was only 5-fold lesssensitive than chicken AHR1 to PeCDF (the ReS value forchicken was 1.0; the ReSaverage value for quail was 0.19). Therange in sensitivity to TCDF among the four wild-type AHR1constructs was more similar to the range in sensitivity observedfor TCDD (e.g., quail and tern were 33- and 34-fold lesssensitive than chicken to TCDF, respectively).The ReSaverage values for the chicken-like mutant

(V324I_A380S) for each DLC (0.7 for TCDD; 1.1 forPeCDF; 1.0 for TCDF) were similar to the ReS in chicken(1.0). Similarly, the ReSaverage values for the pheasant-likemutant (V324I) of 0.13 for TCDD, 0.42 for PeCDF, and 0.1for TCDF were similar to the ReSaverage values in pheasant (0.13for TCDD; 0.49 for PeCDF; 0.09 for TCDF). The ReSaveragevalues for the unknown mutant (A380S) appeared to be slightlyhigher than the ReSaverage values for the pheasant-like mutant(V324I), but further studies would be needed to determine ifthere are indeed differences in ReS values between the A380Smutant and the V324I mutant.ii. Relative Potency (Intercompound Comparisons). ReP

values were obtained from data that were normalized to themaximum luciferase ratios elicited by TCDD for each AHR1construct (Figures S3 and S9), and ReP values are indicated inTable 2 and Figure S10. Perhaps the most interesting resultswere with PeCDF because it was considerably more potentthan TCDD in cells transfected with Japanese quail AHR1(RePaverage = 20) or common tern AHR1 (ReP = 26). PeCDFwas slightly more potent than TCDD in the pheasant(RePaverage = 3.7), pheasant-like mutant (RePaverage = 3.6), andunknown mutant (RePaverage = 3.6). The RePaverage values forthe DLCs in COS-7 cells transfected with the chicken AHR1

(1.0 for TCDD; 0.84 for PeCDF; 0.45 for TCDF) weregenerally similar to the chicken like mutant AHR1 RePaveragevalues (1.0 for TCDD; 1.1 for PeCDF; 0.91 for TCDF).

In Vitro−In Vivo Comparisons. An important goal of ourresearch on avian AHR1 is to develop in vitro methods that canbe used to predict the in vivo potencies of DLCs among avianspecies. The in vitro results obtained in the present study werecompared to the LD20, LD50, and LD80 values of TCDD,PeCDF, and TCDF determined in a companion study15 wherethese compounds were injected into chicken, pheasant, orJapanese quail eggs. As shown in Figure 2 and Figure S11, there

was a significant correlation between in ovo LD20, LD50, and

LD80 values and in vitro EC20, EC50, and EC80 values obtained

with the luciferase reporter gene assay. We also determined the

correlations between in ovo LD50 values and ethoxyresorufin O-

Table 2. Relative Potency (ReP) Values Were Calculated Based on the ReP Estimation Methods Similar to Methods Used byVilleneuve et al.,14 Kim et al.,28 and Leena et al.8a

aReP is defined as EC20, 50, or 80 (species A) of TCDD ÷ three points of responses (Y20, 50 or 80) of the compound of interest that correspond to 20, 50 and80% of TCDD maximum response. For compounds that elicited maximum responses that were less than 80% TCDD, ReP at the observedmaximum response was calculated and replaced by the aReP75,

bReP65,cReP60,

dReP70, oreReP72. Three general classes of avian species (chicken-like,

pheasant-like, or quail-like) are highlighted in light, moderate, or dark gray, respectively. The A380S genotype (unknown mutant) is not highlighted.

Figure 2. Correlation between LD 50 data from egg injection studies15

and EC 50 data from the luciferase reporter gene assay (LRG assay; thispaper) for chicken (C), ring-necked pheasant (P), and Japanese quail(Q). To our knowledge there are no LD50 values for common tern(the other species studied using the LRG assay) exposed in ovo toTCDD, PeCDF, or TCDF.

Environmental Science & Technology Article

dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−29752972

deethylase (EROD) EC50 values in chicken, pheasant, andJapanese quail hepatocyte cultures (Figure S12). Thecoefficient of correlation between in ovo LD20, LD50, andLD80 values and in vitro EC20, EC50, and EC80 values was higherfor the LRG assay (r2 = 0.71, p < 0.005; r2 = 0.77, p < 0.005; r2

= 0.70, p < 0.005, respectively) than for the EROD assay (r2 =0.51, p < 0.05). The relationship between the LRG and ERODassays was significant (r2 = 0.72, p < 0.005; Figure S12). TheReP and ReS values derived from egg injection studies, primaryhepatocyte studies, and the present study with transfectedCOS-7 cells are summarized in Table S3.

■ DISCUSSIONThe first report demonstrating that key amino acids within theLBD of AHR1 elicit differential sensitivity to TCDD amongavian species included studies on the in vitro functional analysisof AHR1 from one sensitive species (domestic chicken), oneinsensitive species (common tern), and selected mutants ofcommon tern AHR1.6 In the present report, the findings of theinitial study with TCDD were confirmed, and the work wasbroadened to include the effects of PeCDF and TCDF onAHR1-dependent gene expression for (a) full-length domesticchicken, ring-necked pheasant, Japanese quail, and commontern AHR1 constructs and (b) three mutant constructs ofJapanese quail AHR1. The concentration-dependent effects ofTCDD, PeCDF, and TCDF on in vitro expression of chicken,ring-necked pheasant, and Japanese quail AHR1s were studiedin order to compare the results from the present study to theresults from our companion studies on the effects of TCDD,PeCDF, and TCDF on embryotoxicity15 and CYP1A inductionin primary cultures of embryonic hepatocytes.11,16

The present study not only confirms the earlier finding6 thatthe chicken AHR1 is more sensitive than the common ternAHR1 to activation by TCDD - it extends that finding byproviding a quantitative measure of the difference in sensitivity.For example, the concentration−response curves indicate thatthe chicken AHR1 is approximately 90- and 70-fold moresensitive to TCDD than the Japanese quail and common ternAHR1s, respectively (Table 1; ReSaverage values are 1.0, 0.011,and 0.014 for chicken, quail, and tern, respectively). Similarly,the ring-necked pheasant AHR1 is approximately 8-fold lesssensitive (based upon comparisons of ReSaverage values) toTCDD than the chicken AHR1.The LBDs of common tern and Japanese quail AHR1s both

contain valine at the chicken site 324 and alanine at the chickensite 380 (chicken sites 324 and 380 are equivalent to tern sites325 and 381). The present finding that common tern andJapanese quail AHR1s were relatively insensitive to AHR1-dependent gene expression upon exposure to TCDD isconsistent with the hypothesis that Val324 and Ala380 impartcharacteristics to the AHR1 which confer a reduced sensitivityto activation by DLCs.5,6 The reduced sensitivity might becaused by a conformational change that affects the bindingaffinity of DLCs to the LBD. AHR1-binding assays andmolecular modeling studies could be used to test thishypothesis. The amino acid at LBD site 257 is threonine incommon tern AHR1 and alanine in Japanese quail AHR1, butthe identity of the amino acid at site 257 appears to have noeffect on AHR1-dependent gene activation by TCDD.6 Thereare several different amino acids within the DNA bindingdomain (DBD) and transactivation domain (TAD) of AHR1(Figure S4), but Karchner and colleagues6 found thatdifferences in amino acids within those domains had no

influence on differential sensitivity of chicken or common ternto TCDD.The importance of amino acids at AHR1 LBD sites 324 and

380 in birds is consistent with reports on the key role ofcorresponding sites within the AHR LBD in other vertebrates.For example, mutation of isoleucine at site 319 in mouse AHRto alanine (mouse site 319 corresponds to chicken site 324)resulted in the loss of transactivation activity.3 The importanceof the amino acid at position 375 of mouse AHR and 381 ofhuman AHR (mouse site 375 and human site 381 areequivalent to chicken site 380) has been shown in severalstudies. For example, mutation of valine 381 to aspartatecompletely abolished the ligand binding capacity of humanAHR.17 A study of two alleles of murine AHR showed that thereduced ligand binding affinity of one of the alleles could beattributed to valine at position 375. Changing this amino acidto alanine enhanced ligand binding affinity 4-fold.2 The humanAHR has been shown to have a lower affinity for binding aphotoaffinity-labeled TCDD analogue compared with themouse AHRb‑1 allele. A single amino acid (valine 381) inhuman AHR is responsible for the lower ligand binding affinity;mutating this residue to alanine restored high affinity ligandbinding.18 When alanine 375 in mouse AHR was substitutedwith isoleucine or phenylalanine, both mutants failed to bind toligands or to exhibit AHR activation in cells exposed to AHRligands.19 Mutation of mouse AHR at position 375 from alanineto valine reduced AHR DNA binding in gel retardation analysis,and the A375L mutant was not able to bind to DNA.20,21

Finally, results of site-directed mutagenesis studies in rat AHRrevealed the importance of the site that is equivalent to chickensite 380 in transactivation activity.22

The concentration-dependent effect of TCDD on luciferaseactivity in COS-7 cells transfected with chicken AHR1 or thechicken-like AHR1 mutant (V324I_A380S) yielded similar(

cultures of Japanese quail hepatocytes.11 PeCDF is also morepotent (20- to 30-fold) than TCDD in herring gull (Larusargentatus) hepatocytes.24 Japanese quail, common tern, andherring gull (accession number: DQ371287) AHR1s all containvaline at site 324 and alanine at site 380. As such, it will be ofinterest to determine if PeCDF is more potent than TCDD inall species with this AHR1 genotype. It would also be ofinterest to conduct AHR1 binding studies and homologymodeling studies to attempt to determine the molecularproperties that cause PeCDF to be more potent than TCDD insome birds. Overall, the comparison of potencies of PeCDFand TCDD as activators of AHR1s from different avian speciesprovides an example of ligand-dependent species differences inAHR affinities, similar to those found recently for other ligandsin comparisons between human and mouse AHRs.25

The results of the present study with the LRG assay indicatea good correlation between in vivo (egg injections) and in vitroresults (r2 = 0.77; p < 0.005 for LD50). If the TCDF data areremoved from Figure 2, the correlation is even stronger (r2 =0.93; p < 0.005). We draw attention to this point because onemight not expect as good a relationship when comparing invivo and in vitro results for TCDF. TCDF is more rapidlyeliminated from avian embryos compared to TCDD orPeCDF15 which could potentially reduce the predictive abilityof in vitro studies for TCDF. The LRG assay might be an equalor better predictor of LD50 values than the primary hepatocyteEROD assay, but more in vitro and in vivo studies with otheravian species and other DLCs would be needed to confirm thishypothesis.This study was possible due to development of a rapid

method that allows for determination of concentration-dependent effects of DLCs on activation of AHR1 in transientlytransfected COS-7 cells. Unlike other reporter gene assays thatuse COS-7 cells to measure avian AHR activity,6,9,10 we (a)seeded cells in 96-well plates rather than in Petri dishes or 48-well plates, (b) added the DLCs directly to the wells withoutchanging the cell culture medium, and (c) used the same 96-well plates to measure luminescence without lysing the cellsand transferring cells to another plate. The method allowsgeneration of approximately 32 concentration−response curvesper person-week, and it could be easily automated to allowincreased throughput and efficiency. Despite the efficiency andhigh throughput nature of our assay, it was less sensitive thanthe assay used by Iwata and colleagues.8,10 For example, wedetermined the EC50 of TCDD in COS-7 cells expressingchicken AHR1 to be 0.21 nM, but Iwata et al. reported aTCDD EC50 of 0.034 nM. We conducted several studies withthe goal of improving the sensitivity of our assay (e.g., usingserum-free medium, changing the time that cells were exposedto DLCs, changing the concentration of reporter andexpression vectors, adding a cell lysis step to the procedure),but improvements in sensitivity were not obtained. Regardlessof the reduced sensitivity, our assay was able to measure fullconcentration−response curves for all AHR1 constructs for thethree compounds studied. One can prepare AHR1 constructsfrom any avian species (including wild species) and studysensitivity to any DLC using the approach described in thepresent paper.The CYP1A5 promoter constructs and ARNT1 constructs

(both from cormorant, and kindly provided by Dr. Iwata,Ehime University, Japan) were kept consistent for allexperiments. The rationale for this decision was based ontwo factors: (a) Iwata and colleagues demonstrated that

structural differences between chicken and cormorantCYP1A5 did not appear to affect the response to TCDD10

and (b) it was necessary to keep the ARNT1 constant to testhypotheses regarding the importance of amino acids at sites324 and 380 within AHR1 without introducing other forms ofARNT and CYP1A5 to the experiments. AHR1 might not bethe only factor that affects the biochemical and toxic effects ofDLCs in avian species, and more studies are needed toelucidate the role(s) of the multiple isoforms of AHR or ARNT(i.e., AHR2, ARNT2). While future studies may indicatecomplex roles between AHRs and ARNTs in transfected cells,such work was beyond the scope of the present work. Ourfindings support and expand the understanding of theimportance of the amino acid residues at sites 324 and 380within AHR1.A practical goal of our research on avian AHR1 is to develop

methods that can be used to predict the sensitivity of any avianspecies to the toxic effects of any DLC. This goal is ofimportance because the differences in sensitivity of wild avianspecies to DLCs make it difficult to make accurate assessmentsand predictions of adverse effects of DLCs in wild birds.Currently, the default assumption in risk assessments is to usethe domestic chicken as a surrogate species for birds when riskassessments of DLCs are conducted. To our knowledge, thechicken is the most sensitive avian species to DLCs. Thus, theconventional approach may overestimate the potential risks ofadverse effects and could lead to inappropriate decisions aboutremedial options at contaminated sites.26,27 The results of thepresent study suggest the possibility of an alternative approachto risk assessments that would involve the determination of theAHR LBD genotype of species of concern. Using suchmolecular knowledge could help guide decision makingregarding the potential impacts of DLC exposure with respectto avian wildlife.

■ ASSOCIATED CONTENT*S Supporting InformationTables S1−S3 and Figures S1−S12. This material is availablefree of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*Phone: 613-998-7384; fax: 613-998-0458; e-mail: sean.kennedy@ec.gc.ca.

Funding Information

This research was supported by an unrestricted grant from theDow Chemical Company to the University of Ottawa andEnvironment Canada’s STAGE program and, in part, by aDiscovery Grant from the National Science and EngineeringResearch Council of Canada (Project # 326415−07). Theauthors wish to acknowledge the support of an instrumentationgrant from the Canada Foundation for Infrastructure. ProfessorGiesy was supported by the Canada Research Chair programand an at large Chair Professorship at the Department ofBiology and Chemistry and State Key Laboratory in MarinePollution, City University of Hong Kong, the EinsteinProfessor Program of the Chinese Academy of Sciences, andthe Visiting Professor Program of King Saud University. M.Hahn and S. Karchner were supported by NOAA Sea Grant(grant number NA06OAR4170021 (R/B-179)) and by theWalter A. and Hope Noyes Smith endowed chair.

Environmental Science & Technology Article

dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−29752974

http://pubs.acs.orgmailto:sean.kennedy@ec.gc.camailto:sean.kennedy@ec.gc.ca

NotesThe authors declare no competing financial interest.

■ REFERENCES(1) Fernandez-Salguero, P. M.; Hilbert, D. M.; Rudikoff, S.; Ward, J.M.; Gonzalez, F. J. Aryl-hydrocarbon receptor-deficient mice areresistant to 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced toxicity. Tox-icol. Appl. Pharmacol. 1996, 140 (1), 173−179.(2) Poland, A.; Palen, D.; Glover, E. Analysis of the four alleles of themurine aryl hydrocarbon receptor. Mol. Pharmacol. 1994, 46 (5), 915−921.(3) Goryo, K.; Suzuki, A.; Del Carpio, C. A.; Siizaki, K.; Kuriyama, E.;Mikami, Y.; Kinoshita, K.; Yasumoto, K.; Rannug, A.; Miyamoto, A.;Fujii-Kuriyama, Y.; Sogawa, K. Identification of amino acid residues inthe Ah receptor involved in ligand binding. Biochem. Biophys. Res.Commun. 2007, 354 (2), 396−402.(4) Boutros, P. C.; Yao, C. Q.; Watson, J. D.; Wu, A. H.; Moffat, I.D.; Prokopec, S. D.; Smith, A. B.; Okey, A. B.; Pohjanvirta, R. Hepatictranscriptomic responses to TCDD in dioxin-sensitive and dioxin-resistant rats during the onset of toxicity. Toxicol. Appl. Pharmacol.2011, 251 (2), 119−129.(5) Head, J. A.; Hahn, M. E.; Kennedy, S. W. Key amino acids in thearyl hydrocarbon receptor predict dioxin sensitivity in avian species.Environ. Sci. Technol. 2008, 42 (19), 7535−7541.(6) Karchner, S. I.; Franks, D. G.; Kennedy, S. W.; Hahn, M. E. Themolecular basis for differential dioxin sensitivity in birds: role of thearyl hydrocarbon receptor. Proc. Natl. Acad. Sci. U. S. A. 2006, 103(16), 6252−6257.(7) Okey, A. B. An aryl hydrocarbon receptor odyssey to the shoresof toxicology: the Deichmann Lecture, International Congress ofToxicology-XI. Toxicol. Sci. 2007, 98 (1), 5−38.(8) Leena Mol, T.; Kim, E. Y.; Ishibashi, H.; Iwata, H. In vitrotransactivation potencies of black-footed albatross (Phoebastrianigripes) AHR1 and AHR2 by dioxins to predict CYP1A expressionin the wild population. Environ. Sci. Technol. 2011, DOI: 10.1021/es2028263.(9) Yasui, T.; Kim, E. Y.; Iwata, H.; Franks, D. G.; Karchner, S. I.;Hahn, M. E.; Tanabe, S. Functional characterization and evolutionaryhistory of two aryl hydrocarbon receptor isoforms (AhR1 and AhR2)from avian species. Toxicol. Sci. 2007, 99 (1), 101−117.(10) Lee, J. S.; Kim, E. Y.; Iwata, H. Dioxin activation of CYP1A5promoter/enhancer regions from two avian species, commoncormorant (Phalacrocorax carbo) and chicken (Gallus gallus):association with aryl hydrocarbon receptor 1 and 2 isoforms. Toxicol.Appl. Pharmacol. 2009, 234 (1), 1−13.(11) Herve, J. C.; Crump, D.; Jones, S. P.; Mundy, L. J.; Giesy, J. P.;Zwiernik, M. J.; Bursian, S. J.; Jones, P. D.; Wiseman, S. B.; Wan, Y.;Kennedy, S. W. Cytochrome P4501A induction by 2,3,7,8-tetrachlor-odibenzo-p-dioxin and two chlorinated dibenzofurans in primaryhepatocyte cultures of three avian species. Toxicol. Sci. 2010, 113 (2),380−391.(12) U.S.EPA. Method 1613 tetra- through octa-chlorinated dioxinsand furans by isotope dilution HRGC/HRMS. U. S. EnvironmentalProtection Agency Office of Water. 1994.(13) Head, J. A.; Kennedy, S. W. Same-sample analysis ofethoxyresorufin-O-deethylase activity and cytochrome P4501AmRNA abundance in chicken embryo hepatocytes. Anal. Biochem.2007, 360 (2), 294−302.(14) Villeneuve, D. L.; Blankenship, A. L.; Giesy, J. P. Derivation andapplication of relative potency estimates based on in vitro bioassayresults. Environ. Toxicol. Chem. 2000, 19 (11), 2835−2843.(15) Cohen-Barnhouse, A. M.; Zwiernik, M. J.; Link, J. E.; Fitzgerald,S. D.; Kennedy, S. W.; Herve, J. C.; Giesy, J. P.; Wiseman, S.; Yang, Y.;Jones, P. D.; Wan, Y.; Collins, B.; Newsted, J. L.; Kay, D.; Bursian, S. J.Sensitivity of Japanese quail (Coturnix japonica), common pheasant(Phasianus colchicus) and white leghorn chicken (Gallus gallusdomesticus) embryos to in ovo exposure to 2,3,7,8-tetrachlorodiben-zo-p-dioxin (TCDD), 2,3,4,7,8- pentachlorodibenzofuran (PeCDF)

and 2,3,7,8-tetrachlorodibenzofuran (TCDF). Toxicol. Sci. 2011, 119(1), 93−103.(16) Herve, J. C.; Crump, D.; Giesy, J. P.; Zwiernik, M. J.; Bursian, S.J.; Kennedy, S. W. Ethoxyresorufin O-deethylase induction by TCDD,PeCDF and TCDF in ring-necked pheasant and Japanese quailhepatocytes: Time-dependent effects on concentration-responsecurves. Toxicol. In Vitro 2010, 24 (4), 1301−1305.(17) Ema, M.; Ohe, N.; Suzuki, M.; Mimura, J.; Sogawa, K.; Ikawa, S.;Fujii-Kuriyama, Y. Dioxin binding activities of polymorphic forms ofmouse and human aryl hydrocarbon receptors. J. Biol. Chem. 1994, 269(44), 27337−27343.(18) Ramadoss, P.; Perdew, G. H. Use of 2-azido-3-[125I]iodo-7,8-dibromodibenzo-p-dioxin as a probe to determine the relative ligandaffinity of human versus mouse aryl hydrocarbon receptor in culturedcells. Mol. Pharmacol. 2004, 66 (1), 129−136.(19) Murray, I. A.; Reen, R. K.; Leathery, N.; Ramadoss, P.; Bonati,L.; Gonzalez, F. J.; Peters, J. M.; Perdew, G. H. Evidence that ligandbinding is a key determinant of Ah receptor-mediated transcriptionalactivity. Arch. Biochem. Biophys. 2005, 442 (1), 59−71.(20) Pandini, A.; Denison, M. S.; Song, Y.; Soshilov, A. A.; Bonati, L.Structural and functional characterization of the aryl hydrocarbonreceptor ligand binding domain by homology modeling andmutational analysis. Biochemistry 2007, 46 (3), 696−708.(21) Pandini, A.; Soshilov, A. A.; Song, Y.; Zhao, J.; Bonati, L.;Denison, M. S. Detection of the TCDD binding-fingerprint within theAh receptor ligand binding domain by structurally driven mutagenesisand functional analysis. Biochemistry 2009, 48 (25), 5972−5983.(22) Backlund, M.; Ingelman-Sundberg, M. Different structuralrequirements of the ligand binding domain of the aryl hydrocarbonreceptor for high- and low-affinity ligand binding and receptoractivation. Mol. Pharmacol. 2004, 65 (2), 416−425.(23) Van den Berg, M.; Birnbaum, L.; Bosveld, A. T.; Brunstrom, B.;Cook, P.; Feeley, M.; Giesy, J. P.; Hanberg, A.; Hasegawa, R.;Kennedy, S. W.; Kubiak, T.; Larsen, J. C.; van Leeuwen, F. X.; Liem, A.K.; Nolt, C.; Peterson, R. E.; Poellinger, L.; Safe, S.; Schrenk, D.;Tillitt, D.; Tysklind, M.; Younes, M.; Waern, F.; Zacharewski, T. Toxicequivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans andwildlife. Environ. Health Perspect. 1998, 106 (12), 775−792.(24) Herve, J. C.; Crump, D. L.; McLaren, K. K.; Giesy, J. P.;Zwiernik, M. J.; Bursian, S. J.; Kennedy, S. W. 2,3,4,7,8-pentachlorodibenzofuran is a more potent cytochrome P4501Ainducer than 2,3,7,8-tetrachlorodibenzo-p-dioxin in herring gullhepatocyte cultures. Environ. Toxicol. Chem. 2010, 29 (9), 2088−2095.(25) Flaveny, C. A.; Murray, I. A.; Chiaro, C. R.; Perdew, G. H.Ligand selectivity and gene regulation by the human aryl hydrocarbonreceptor in transgenic mice. Mol. Pharmacol. 2009, 75 (6), 1412−1420.(26) Fredricks, T. B.; Giesy, J. P.; Coefield, S. J.; Seston, R. M.;Tazelaar, D. L.; Roark, S. A.; Kay, D. P.; Newsted, J. L.; Zwiernik, M. J.Multiple lines of evidence risk assessment of terrestrial passerinesexposed to PCDFs and PCDDs in the Tittabawassee River floodplain,Midland, Michigan, USA. Hum. Ecol. Risk Assess. 2011, 17 (1), 159−186.(27) Fredricks, T. B.; Zwiernik, M. J.; Seston, R. M.; Coefield, S. J.;Tazelaar, D. L.; Roark, S. A.; Kay, D. P.; Newsted, J. L.; Giesy, J. P.Effects on tree swallows exposed to dioxin-like compounds associatedwith the Tittabawassee River and floodplain near Midland, Michigan,USA. Environ. Toxicol. Chem. 2011, 30 (6), 1354−1365.(28) Kim, E. Y.; Suda, T.; Tanabe, S.; Batoev, V. B.; Petrov, E. A.;Iwata, H. Evaluation of relative potencies for in vitro transactivation ofthe Baikal seal aryl hydrocarbon receptor by dioxin-like compounds.Environ. Sci. Technol. 2011, 45 (4), 1652−1658.

Environmental Science & Technology Article

dx.doi.org/10.1021/es2043992 | Environ. Sci. Technol. 2012, 46, 2967−29752975