*Manuscript Click here to view linked References

1 2 3 4 5 1 Regulation of Ahr signaling by Nrf2 during development: 6 7 8 2 Effects of Nrf2a deficiency on PCB126 embryotoxicity in 9 10 3 (Danio rerio) 11 12

13 1 1 1 2 2 14 4 Michelle E. Rousseau , Karilyn E. Sant , Linnea R. Borden , Diana G. Franks , Mark E. Hahn , 15 1,2,* 16 5 Alicia R. Timme-Laragy 17 18 6 1. Department of Environmental Health Sciences, School of Public Health and Health 19 20 7 Sciences, University of Massachusetts Amherst, Amherst, MA 01003 21 2. Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543 22 8 23 24 9 * Corresponding author, 686 N. Pleasant St., Goessmann 149C, Amherst, MA 01003, 25 26 10 phone 413-545-7423 27 28 11 29 30 31 12 Michelle E. Rousseau [email protected] 32 33 13 Karilyn E. Sant [email protected] 34 35 36 14 Linnea R. Borden [email protected] 37 38 39 15 Diana G. Franks [email protected] 40 41 16 Mark E. Hahn [email protected] 42 43 44 17 Alicia R. Timme-Laragy [email protected] 45 46 18 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 1 63 64 65 1 2 3 4 19 Abstract 5 6 7 20 The embryotoxicity of co-planar PCBs is regulated by the aryl hydrocarbon (Ahr), and 8 21 has been reported to involve oxidative stress. Ahr participates in crosstalk with another 9 10 22 , Nfe2l2, or Nrf2. Nrf2 binds to response elements to regulate the 11 12 23 adaptive response to oxidative stress. To explore aspects of the crosstalk between Nrf2 and Ahr 13 and its impact on development, we used zebrafish (Danio rerio) with a mutated DNA binding 14 24 15 25 domain in Nrf2a (nrf2afh318/fh318), rendering these embryos more sensitive to oxidative stress. 16 17 26 Embryos were exposed to 2 nM or 5 nM PCB126 at 24 hours post fertilization (prim-5 stage of 18 pharyngula) and examined for expression and morphology at 4 days post fertilization (dpf; 19 27 20 28 protruding –mouth stage). Nrf2a mutant eleutheroembryos were more sensitive to PCB126 21 22 29 toxicity at 4 dpf, and in the absence of treatment also displayed some subtle developmental 23 24 30 differences from wildtype embryos, including delayed inflation of the swim bladder and smaller 25 31 yolk sacs. We used qPCR to measure changes in expression of the nrf gene family, keap1a, 26 27 32 keap1b, the ahr gene family, and known target . cyp1a induction by PCB126 was 28 29 33 enhanced in the Nrf2a mutants (156-fold in wildtypes vs. 228-fold in mutants exposed to 5 nM). 30 34 Decreased expression of heme oxygenase (decycling) 1 (hmox1) in the Nrf2a mutants was 31 32 35 accompanied by increased nrf2b expression. Target genes of Nrf2a and AhR2, 33 34 36 NAD(P)H:quinone oxidoreductase 1 (nqo1) and glutathione S-transferase, alpha-like (gsta1), 35 37 showed a 2-5-fold increase in expression in the Nrf2a mutants as compared to wildtype. This 36 37 38 study elucidates the interaction between two important transcription factor pathways in the 38 39 39 developmental toxicity of co-planar PCBs. 40 40 41 42 43 41 Key Words: transcription factor, dioxin-like compound, gene expression, embryonic 44 45 42 development, swim bladder, yolk 46 47 43 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 1 63 64 65 1 2 3 4 44 Abbreviations: 5 6 7 45 Ahr: aryl hydrocarbon receptor 8 46 Ahrr: aryl hydrocarbon receptor repressor 9 10 47 ARE: antioxidant 11 12 48 Cyp1a: cytochrome P540 1a 13 DLC: dioxin-like compound 14 49 15 50 DMSO: dimethyl sulfoxide 16 17 51 dpf: days post fertilization 18 hpf: hours post fertilization 19 52 20 53 Keap1: Kelch-like ECH-associated 1 21 22 54 Nrf2: nuclear factor erythroid related factor 2-like-2 23 24 55 PAH: polycyclic aromatic hydrocarbon 25 56 PCB126: 3,3’,4,4’,5-pentachlorobiphenyl 26 27 57 ROS: reactive oxygen species 28 29 58 tBOOH: tert butylhydroperoxide 30 59 tBHQ: tert butylhydroquinone 31 32 60 TCDD: 2,3,7,8-tetracholordibenzo-p-dioxin 33 XRE: xenobiotic response element, aka dioxin response element 34 61 35 62 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 2 63 64 65 1 2 3 4 63 1. Introduction 5 6 64 PCBs are carcinogenic and teratogenic persistent organic pollutants. Along with other dioxin- 7 8 65 like-compounds, co-planar PCBs are potent, chlorinated ligands for the aryl hydrocarbon 9 66 receptor (Ahr), which has been shown to mediate toxicity in numerous mammalian and fish 10 11 67 species (e.g. rev. in (Bock, 2013; Denison et al., 2011; Hahn et al., 2006; King-Heiden et al., 12 13 68 2012)). These ligands for the Ahr are also associated with oxidative stress, such as via 14 69 uncoupling of the cytochrome P450 catalytic cycle and the release of superoxide and/or 15 16 70 hydrogen peroxide (Dalton et al., 2002). The adaptive response to oxidative stress is largely 17 18 71 regulated by the redox-sensitive transcription factor nuclear factor erythroid related factor 2-like- 19 72 2 (Nfe2l2 or Nrf2- see footnote). Nrf2 regulates transcription of numerous antioxidant and 20 21 73 cytoprotective genes, and also participates in crosstalk with the Ahr to regulate transcription of 22 23 74 genes, many of which encode Phase I and Phase II enzymes (Wakabayashi et al., 2010). 24 25 75 The Ahr is a member of the basic helix-loop-helix Per-Arnt-Sim (bHLH/PAS) family of 26 27 76 transcription factors. Briefly, it is a constitutively expressed, cytosolic protein. In the canonical 28 29 77 Ahr activation pathway, Ahr translocates to the nucleus upon ligand binding, where it forms a 30 78 heterodimer with the aryl hydrocarbon nuclear translocator (Arnt). This heterodimer binds to the 31 32 79 xenobiotic response element (XRE) in promoter regions and initiates the transcription of a 33 34 80 variety of genes. These genes are mostly comprised of Phase I and Phase II enzymes which 35 81 are often referred to as the “classical Ahr gene battery,” but also include target genes involved 36 37 82 in numerous other functions including liver development, cell cycle progression, cellular 38 39 83 differentiation, inflammation, and (Bock, 2013; Denison, et al., 2011). 40 41 84 Zebrafish have three Ahr genes (ahr1a, ahr1b, ahr2), whereas most mammals have only one 42 43 85 (Ahr). Ahr2 has been shown to mediate the effects of 3,3’,4,4’,5-pentachlorobiphenyl (PCB126), 44 45 86 2,3,7,8-tetracholordibenzo-p-dioxin (TCDD; dioxin), and some high molecular-weight PAHs; the 46 87 roles of Ahr1a and Ahr1b are not yet fully understood (Andreasen et al., 2002; Garner et al., 47 48 88 2013; Goodale et al., 2012; Jonsson et al., 2007; Karchner et al., 2005; Prasch et al., 2003). 49 50 89 PCB126 exposure in zebrafish embryos has been shown to induce expression of cyp1a, 51 90 cyp1b1, cyp1c1, and cyp1c2 in an Ahr2-mediated manner (Billiard et al., 2006; Jonsson, et al., 52 53 91 2007). cyp1a has been shown to be the most responsive of the cyp1 genes, and either 54 55 92 expression of this gene or enzyme activity (ethoxyresorufin O-deethylase; EROD) is often used 56 93 as a biomarker of Ahr ligand activation. Numerous studies of zebrafish embryos exposed to 57 58 94 PCB126 or TCDD during development have demonstrated a suite of morphological deformities 59 60 95 collectively referred to as “blue sac disease,” which includes pericardial edema, craniofacial and 61 62 3 63 64 65 1 2 3 4 96 malformations, and reduced swim bladder inflation; these deformities have been shown to 5 6 97 be primarily regulated by Ahr2 (Garner, et al., 2013; Jonsson, et al., 2007; Jonsson et al., 7 8 98 2012; Lanham et al., 2014; Prasch, et al., 2003), although the importance of other signaling 9 99 pathways has also been demonstrated, including interactions with Wnt, Cox-2, Sox9 (Yoshioka 10 11 100 et al., 2011). 12 13 In addition to activating Ahr2, exposure to co-planar PCBs and dioxins has also resulted in the 14 101 15 102 generation of reactive oxygen species (ROS) and/or oxidative stress in mammals, frogs, birds, 16 17 103 and fish, via several mechanisms including generation of reactive metabolites, an increase in 18 mitochondrial respiration and hydrogen peroxide, uncoupling of the cytochrome P450 catalytic 19 104 20 105 cycle, glutathione depletion, and inflammation (Dalton, et al., 2002; Gillardin et al., 2009; Lu et 21 22 106 al., 2011; Nebert et al., 2000; Schlezinger et al., 2000; Schlezinger et al., 2006; Senft et al., 23 24 107 2002). Oxidative stress is defined as a loss of redox signaling and control (Jones, 2006), and 25 108 the oxidative stress response is defined as the resulting changes in gene expression that serve 26 27 109 to mitigate the oxidative challenge (Hahn et al., 2014). Despite the many studies that 28 29 110 demonstrate an oxidative stress response to these co-planar PCBs and dioxins, there have also 30 111 been studies that have failed to identify an oxidative stress response. In zebrafish for example, 31 32 112 while nrf2a has been shown to be up-regulated by TCDD (Hahn, et al., 2014), zebrafish 33 34 113 embryos sampled immediately after a 6 h exposure to TCDD showed no evidence of altered 35 114 expression of other genes typically found in the oxidative stress response at 4 dpf, and there 36 37 115 was an increase in gstp1 expression only after 48 h of a TCDD exposure that began at 24 hours 38 39 116 post fertilization (hpf) (Hahn, et al., 2014). An oxidative stress response was not observed in 40 117 response to TCDD at least two other studies (Alexeyenko et al., 2010; Wang et al., 2013). This 41 42 118 may be due to many factors including differences in timing, dose, developmental stage, species 43 44 119 sensitivity to Ahr ligand binding, or exposure duration, or it is possible that ROS generation may 45 120 not always be sufficient to alter redox signaling or activate an oxidative stress response. 46 47 121 48 49 122 Nrf2 has been designated the “master regulator” of the adaptive response to oxidative stress 50 123 (Ohtsuji et al., 2008). Nrf2 is a basic-region transcription factor that is 51 52 124 constitutively and ubiquitously expressed and bound to Kelch-like ECH-associated protein 1 53 54 125 (Keap1) within the cytosol, targeting it for ubiquitination (Cullinan et al., 2004). There are several 55 126 ways in which Nrf2 can be activated, such as protein kinase RNA-like 56 57 127 kinase (PERK) signaling from the endoplasmic reticulum, phosphorylation, and electrophilic or 58 59 128 ROS interactions (reviewed in (Bryan et al., 2013; Niture et al., 2014)). After activation, Nrf2 60 61 62 4 63 64 65 1 2 3 4 129 accumulates in the nucleus where it dimerizes with small v- avian musculoaponeurotic 5 6 130 fibrosarcoma oncogene homolog (Maf) to upregulate transcription of numerous 7 8 131 cytoprotective genes (Ma et al., 2004). Nrf2 mediates gene expression through the antioxidant 9 132 response element (ARE). AREs are frequently located in promoter regions of genes involved in 10 11 133 the oxidative stress response and Phase II detoxification, but also have more pleiotropic roles in 12 13 134 functions such as cell cycle regulation and lipid metabolism (Malhotra et al., 2010). Disruption of 14 135 NRF2 in mice has been linked to an increase in pulmonary inflammation, emphysema, 15 16 136 neurodegeneration, and chemical , thus indicating the importance of Nrf2’s 17 18 137 cytoprotective role (Kensler et al., 2007). NRF2 knockout mice have also been shown to have 19 138 defects in liver growth and size (Beyer et al., 2008), reduced bile duct microbranching (Skoko et 20 21 139 al., 2014), and retinal vasculature density (Wei et al., 2013); in some genetic backgrounds, 22 23 140 NRF2-/- mice have a congenital defect resulting in an intrahepatic shunt (Skoko, et al., 2014). 24 25 141 Zebrafish have partitioned the function of Nrf2 between two genes, nrf2a and nrf2b (reviewed in 26 27 142 (Hahn et al., 2015)). We have previously shown that Nrf2a is generally involved in activating 28 29 143 gene expression, while Nrf2b is involved in repressing gene expression (Timme-Laragy et al., 30 144 2012). The nrf2afh318 mutant is a recessive loss-of-function allele generated via the zebrafish 31 32 145 Targeted Induced Local Lesions in Genomes (TILLING) mutagenesis project (Mukaigasa et al., 33 fh318 34 146 2012). The phenotype of nrf2a fish is similar to that of NRF2 -/- mice (e.g (Itoh et al., 1997)) 35 147 in that they display normal development and and enhanced sensitivity to oxidants 36 37 148 (Mukaigasa, et al., 2012). nrf2afh318 larvae are more sensitive to oxidative stress, but no 38 39 149 morphologic or growth differences between Nrf2a mutant vs. wildtype larvae or adults have 40 150 been identified (Mukaigasa, et al., 2012). 41 42 43 151 Crosstalk between Ahr and Nrf2 has been demonstrated in mammalian systems and has been 44 45 152 shown to be important in the transcriptional regulation of genes encoding many Phase I and 46 153 Phase II detoxification enzymes (Anwar-Mohamed et al., 2011; Lo and Matthews, 2013; Miao 47 48 154 et al., 2005; Tijet et al., 2006; Wang, et al., 2013; Yeager et al., 2009). TCDD has been shown 49 50 155 to increase both protein level and nuclear accumulation of NRF2 in mice (Yeager, et al., 2009) 51 156 and in Hepa1c1c7 cells (Wang, et al., 2013), without evidence of oxidative stress; rather, Wang 52 53 157 et al. provided evidence suggesting the AHR upregulated transcription of NRF2 and formed 54 55 158 protein complexes with both NRF2 and KEAP1 that contributed to the stability of NRF2 and 56 159 subsequent ARE-activation (Wang, et al., 2013). However, the interactions of these pathways 57 58 160 have not been thoroughly studied during embryonic development, which is a critical time for 59 60 161 chemical sensitivity that is fundamentally different than adult physiology. 61 62 5 63 64 65 1 2 3 4 162 Zebrafish contain multiple paralogs of both Ahr and Nrf2, each of which demonstrates a 5 6 163 seemingly different function. Here, we tested the hypothesis that Nrf2a plays a protective role in 7 8 164 PCB126 embryotoxicity. We used PCB126 as a model ligand to study the interaction between 9 165 the Ahr and Nrf2 pathways during embryonic development in zebrafish embryos lacking a 10 11 166 functional Nrf2a. We present data on differential sensitivity to PCB126 embryotoxicity between 12 13 167 the Nrf2a mutant and wildtype eleutheroembryos, differential gene expression of ahr and nrf- 14 168 family members and target genes, and also identify subtle developmental differences between 15 16 169 the two genotypes. 17 18 19 170 2. Materials and Methods 20 21 171 2.1 Chemicals 22 23 172 Polychlorinated biphenyl-126 (PCB126) was purchased from Ultra Scientific (N. Kingstown, RI, 24 25 173 USA), and was dissolved in dimethyl sulfoxide (DMSO, Fisher Scientific, Pittsburgh, PA, USA). 26 174 Solutions were vortexed prior to use. Tert-butylhydroquinone (tBHQ) was purchased from Acros 27 28 175 Organics (Thermo Fisher Scientific, Waltham, MA, USA). 29 30 176 2.2 Animals 31 32 177 Zebrafish heterozygous for the nrf2afh318 and nrf2afh319 alleles were generated through the 33 34 178 TILLING mutagenesis Project (R01HD076585) and obtained as embryos from the Moens 35 Laboratory (Fred Hutchinson Cancer Research Center, Seattle, WA, USA). This project aims to 36 179 37 180 generate loss-of-function mutations in zebrafish genes. The nrf2afh318 and nrf2afh319 alleles 38 39 181 encode Nrf2a proteins with point mutations resulting in single changes. Both of 40 41 182 these mutations are located within the DNA binding domain that associates with AREs, and the 42 183 wild type residue at these locations is conserved between zebrafish Nrf2a, human, mouse, and 43 44 184 the chicken orthologs. The fh318 allele point mutation results in a change from R to L at amino 45 46 185 acid 479 and the fh319 allele results in a change from N to I at amino acid 488; these amino 47 186 acid residue numbers are derived from the TILLING project background sequence. Because 48 49 187 these changes are in the DNA-binding domain, nrf2a should still be transcribed and translated, 50 51 188 and still maintain a functional relationship with Keap1a and Keap1b (the zebrafish co-orthologs 52 189 of the mammalian Keap1 (Li et al., 2008)). When we amplified, cloned, and sequenced the full- 53 54 190 length nrf2afh318 and nrf2afh319 cDNAs (see below), we found a discrepancy between the 55 56 191 TILLING nrf2a sequence and our sequences, with the TILLING sequence missing a six amino 57 192 acid segment between residues 186-193. Supplemental Figure 1 shows the protein alignment 58 59 193 encoded by the nrf2a-318-2 clone with other zebrafish Nrf2a sequences including the TILLING 60 61 62 6 63 64 65 1 2 3 4 194 sequence. Our sequence agrees with that identified by (Mukaigasa, et al., 2012), which 5 fh318 fh319 6 195 numbers the changed amino acids as R485L and N494I for the nrf2a and nrf2a alleles, 7 8 196 respectively. 9 10 197 Adult fish were maintained at 28.5°C on a 14 hour (h) light cycle, 10 h dark light cycle in an 11 12 198 Aquatic Habitats zebrafish system (for experiments conducted at the Woods Hole 13 Oceanographic Institution (WHOI)) or an Aquaneering zebrafish system (at the University of 14 199 15 200 Massachusetts Amherst (UMASS)). To obtain homozygous wildtype and mutant fish, the 16 17 201 heterozygous adults were crossed against Tubigen-longfin (TL) wildtype fish for two 18 generations, then in-crossed and genotyped. Adult homozygous zebrafish were crossed to 19 202 20 203 obtain homozygous wildtype (nrf2a +/+) or mutant (nrf2a fh318/fh318) embryos. All embryos were 21 22 204 maintained at low density with daily water changes at 28.5°C on a 14 h light cycle, 10 h dark 23 24 205 light cycle in 0.3x Danieau’s water throughout the duration of the experiment. This study was 25 206 conducted in accordance with the recommendations in the Guide for the Care and Use of 26 27 207 Laboratory Animals of the National Institutes of Health. Animal protocols were approved by the 28 29 208 Institutional Animal Care and Use Committees (IACUCs) of WHOI (Woods Hole, MA, USA, 30 209 Animal Welfare Assurance Number A3630-01) and UMASS (Amherst, MA, USA, Animal 31 32 210 Welfare Assurance Number A3551-01). 33 34 211 2.3 Genotyping 35 36 212 To identify mutant (nrf2afh318) and wildtype alleles (nrf2a+), we isolated DNA from individual fish 37 38 213 fin clips using standard methods. Briefly, the fin was digested with proteinase K overnight, then 39 subjected to a phenol-chloroform DNA extraction. DNA was quantified using a BioDrop μLITE 40 214 41 215 (BioDrop, Cambridge, UK), and only samples that met quality control standards (260/280 ratio 42 43 216 1.8-2) were used in subsequent steps. For the PCR reaction, 200 ng of DNA was amplified 44 45 217 using a proofreading polymerase (Advantage polymerase, Clontech, Mountain View, CA, USA), 46 218 in a 25 µL reaction containing 10 mM dNTPs, 10 µM Reverse primer (5’ - ACC AAC AGG CGG 47 48 219 CGA TAA TGA CTT – 3’), and 10 µM forward primer (5’- AAC GAT GGC TCC CAG ACT CCA 49 50 220 CT – 3’). The PCR reaction was carried out on an Eppendorf Mastercycler Nexus Gradient 51 221 thermocycler (Eppendorf, Hauppauge, NY, USA). The PCR conditions were 94°C for 10 min, 52 53 222 then 35 cycles of 94°C for 10 seconds, 62.9°C for 30 seconds, and 72°C for 60 seconds. This 54 55 223 was followed by a 7 minute incubation at 72°C, then the reaction was cooled and held at 4°C. 56 224 This resulted in one clear band of 835 bp. This product was then cut with the restriction enzyme 57 58 225 Hpy99I and Cut Smart Buffer (New England Biolabs, Ipswich, MA, USA) in a 25 µL reaction 59 60 226 containing 1 µL of PCR product for two h at 37°C. The fh318 mutation in the nrf2a gene results 61 62 7 63 64 65 1 2 3 4 227 in an amino acid change from an R to an L at amino acid 485. The restriction enzyme cut the 5 6 228 wildtype sequence resulting in two bands, at 530 and 305 bp, while the mutant sequence 7 8 229 remained uncut. To visualize the bands, 10 µL of the reaction was mixed with 4 µL of 9 230 MaestroSafe loading buffer (MaestroGen, Las Vegas, NV, USA), run on a 2% agarose 10 11 231 electrophoresis gel, and imaged on a Syngene Pxi (Syngene, Frederick, MD, USA). 12 13 232 2.4 Cloning of Nrf2a alleles 14 15 233 To assess the functional consequences of the nrf2a point mutations in vitro, we amplified the 16 fh318 fh319 17 234 full-length nrf2a and nrf2a cDNAs, cloned them into pcDNA and verified the sequences. 18 To clone these mutant genes, total RNA was isolated from the fin of one nrf2afh318 mutant fish 19 235 20 236 and 20-pooled embryos collected from nrf2afh319 mutant fish using STAT-60 (Tel-Test, Inc., 21 fh319 22 237 Friendswood, TX, USA). Total RNA isolated from embryos collected from nrf2a fish 23 24 238 containing the wild-type allele (as determined through fin-clipping and genotyping) was used as 25 239 the negative control. Total RNA was quantified by Nanodrop (Thermo Fisher Scientific), and 26 27 240 cDNA synthesized using the iScript kit (Biorad, Hercules, CA, USA) from 1 µg of total RNA. 28 29 241 PCR was run to amplify the full-length nrf2a gene using primers that incorporated XbaI and 30 242 HindIII restriction enzyme sites at the 5’ end of each primer, facilitating ligation into the pcDNA 31 32 243 3.2/DEST vector. Using Advantage II Polymerase (Clontech), the PCR program was as follows: 33 34 244 94°C, 1 minute; 5 cycles of 94°C, 30 seconds, 60°C, 15 seconds, 68°C, 2 minutes; 30 cycles of 35 245 94°C, 5 seconds, 68°C, 2 minutes. PCR products were run on a 1% agarose gel, and the 36 37 246 product bands extracted and gene-cleaned (MP Biomedicals, Santa Ana, CA, USA). The gene- 38 39 247 cleaned PCR products, along with 1 µg pcDNA plasmid DNA, were cut with HindIII/XbaI 40 248 (Promega, Madison, WI) for 2 h at 37°C. The cut products were again gene-cleaned to remove 41 42 249 the restriction enzymes, and the nrf2afh318, nrf2afh319, and nrf2a+ non-mutant allele derived from 43 44 250 the fh319 heterozygous line were ligated onto linear pcDNA. These ligations were transformed 45 251 into JM109 competent cells with resulting transformants spread on LB plates supplemented with 46 47 252 ampicillin and incubated overnight at 37°C. Single colonies were picked into 2 ml LB ampicillin 48 49 253 liquid cultures and grown overnight at 37°C with shaking. Plasmid mini-preps were made using 50 254 the Pure Yield Plasmid kit (Promega). To screen for mutants, plasmid DNA from the nrf2afh318 51 52 255 clones was cut with restriction enzyme Hpy99I as described above. Plasmid DNA from the 53 fh319 54 256 nrf2a clones was cut with BglII for 2 h at 37°C. Cut products were run on a 1% agarose gel, 55 257 and clones positive for the insert were sent to Eurofins-MWG (Birmingham, AL, USA) for 56 57 258 sequencing. Large plasmid preps (Qiagen, Valencia, CA, USA) were made from the nrf2a-318-2 58 59 60 61 62 8 63 64 65 1 2 3 4 259 mutant clone, and the nrf2a-319S mutant clone. These preps were used in the luciferase assay 5 6 260 described below. 7 8 261 2.5 Transient transfection of mutant and wildtype alleles and luciferase assay 9 10 262 Three 48-well cell culture plates were seeded with Cos-7 cells (African Green Monkey , 11 5 12 263 from ATCC, Manassas, VA, USA) at a density of 1.6 x10 cells/ml, and 250 µl/well. Plates #1 13 and #2 were identical, and used for transfection, while plate #3 cells were used to check the 14 264 15 265 expression level of the Nrf2a proteins, by western blot with antibodies against Nrf2a. Cos-7 cells 16 17 266 in the plates were grown at 37°C with 5% CO2 for 24 h prior to transfection. X-tremeGENE HP 18 DNA transfection reagent (Roche, Genentech, S. San Francisco, CA, USA) was used at a 3:1 19 267 20 268 ratio of reagent: DNA. Transfection components and their concentrations were as follows: 50 21 22 269 ng/well ARE-luciferase, 3 ng/well Renilla, 25 ng/well wild type nrf2a DNA (pCS2-nrf2a, a 23 24 270 generous gift from Dr. Makoto Kobyashi, University of Tsukuba, Tennodai, Tsukuba, Japan 25 271 (Kobayashi et al., 2002)), 25 ng/well nrf2a-318-2 mutant DNA, 25 ng/well nrf2a-319-S mutant 26 27 272 DNA, 50 ng/well zf-keap-1a DNA, and 50 ng/well zf-keap-1b DNA (both keap1 plasmids were 28 29 273 also a gift from Dr. Kobayashi). Plate #1 cells were dosed with DMSO, while cells in plate #2 30 274 were dosed with 10 µM tBHQ at 19 hours post transfection for 5 h. The luciferase assay was run 31 32 275 using the Dual-Luciferase Reporter Assay System (Promega), and fluorescence was read on a 33 34 276 Turner TD 20/20 Luminometer. Wells on the Western Blot plate were transfected with 50 ng or 35 277 500 ng nrf2a, nrf2a-318-2, or nrf2a-319S DNA only. After 15 h, the cells were rinsed with 1x 36 37 278 PBS, and collected in 40 µl 2x sample treatment buffer/well. Six replicate wells of cell lysates 38 39 279 were pooled into a 1.5 ml Eppendorf tube, boiled for 5 minutes and frozen at -80°C. 40 41 280 Cell lysates were loaded on an 8% acrylamide gel, 30 µl of lysate/well, and run at 100V for 2 h. 42 43 281 The proteins were transferred to a Nytran substrate using the semi-dry transfer method, and 44 45 282 blocked in milk/TBST solution overnight at 4°C. The primary antibody used was Nrf2a rabbit 46 283 #5136 antibody at 1 µg/ml, with goat-anti-rabbit horse-radish peroxidase (GARHRP) secondary 47 48 284 antibody at 1:5000. Blots were exposed to the enhanced chemiluminescence (ECL) Prime 49 50 285 Western Blotting Detection Reagent (Amersham, GE Healthcare Biosciences, Piscataway, NJ, 51 286 USA), and images developed on Hyperfilm ECL (Amersham). The Nrf2a antibody was 52 st 53 287 generated by 21 Century Biochemicals (Marlboro, MA, USA) against the Nrf2a-specific peptide 54 55 288 sequence NMPMQETLDMNAFMKPST, and validated using standard procedures. 56 57 58 59 60 61 62 9 63 64 65 1 2 3 4 289 2.6 PCB Exposure 5 6 290 The model Ahr ligand used in this study was PCB126. At 24 hpf (prim-5 of the pharyngula 7 8 291 stage, hereforth referred to as 24 hpf as per staging by (Kimmel et al., 1995)), embryos were 9 292 exposed for 24 h in triplicate pools of 5 embryos per dose, in 20 mL glass vials. DMSO 10 11 293 concentration was 0.01 % in both control and PCB-treated embryo groups, in a total volume of 12 13 294 10 mL 0.3x Danieau’s water. The concentrations of PCB126 included 2 nM and 5 nM. At 4 dpf 14 295 (protruding mouth stage, hereforth referred to as 4 dpf), embryos were either imaged for 15 16 296 morphology, or fixed in 100 µL of RNA later (Ambion, Life Technologies, Grand Island, NY, 17 18 297 USA) and stored at -80°C until RNA extraction. Morphology experiments were repeated three 19 298 times; two of these were analyzed at 4 dpf, and a third at 5 dpf. 20 21 22 299 2.7 Morphometrics 23 24 300 At 4 dpf, eleutheroembryos were mounted in 3% methylcellulose and imaged on a Zeiss 25 301 dissecting microscope with a Zeiss AxioCam MR Color CCD camera at 50x magnification 26 27 302 (Zeiss, Peabody, MA, USA). Zeiss Axiocam software was used to take measurements of the 28 29 303 pericardial area, an assessment made as to whether the swim bladder was inflated, and a 30 304 deformity score assessing the truncation of Meckel’s cartilage was given using previously 31 32 305 established parameters that incorporate both frequency and severity (Harbeitner et al., 2013). 33 34 306 The ventral-dorsal distance was measured by dropping a straight line from the third somite to 35 307 the bottom of the yolk sac. 36 37 +/+ fh318/fh318 38 308 To measure the yolk area of early stage embryos, 19 nrf2a and 20 nrf2a blastula-stage 39 40 309 embryos (3 hpf), embryos were imaged on an EVOS Auto FL at 100x magnification. To 41 310 standardize measurement of yolk sacs, the diameter was measured twice. The first 42 43 311 measurement began at the edge of the yolk sac closest to the center of the embryonic cell 44 45 312 mass, and this line extended to the furthest point across the yolk sac. The second diameter 46 313 measurement was of the line perpendicular to the first one. The volume for an ellipse was used 47 48 314 to calculate yolk sac area: A = πab, where “a” is the radius of one measurement, and “b” is the 49 2 2 50 315 radius of the other (0.5*diameter). Results are reported in microns (µm ). 51 52 316 2.8 Total RNA extraction and reverse transcription 53 54 317 RNA extractions were carried out according to the GeneJET RNA Purification Kit Total RNA 55 56 318 Purification Protocol (Thermo Fisher Scientific Inc,, Waltham, MA, USA). RNA quality and 57 319 quantity was analyzed spectrophotometrically using a BioDrop μLITE (BioDrop, Cambridge, UK, 58 59 320 USA). The iScript cDNA Synthesis Kit for reverse transcription PCR (Bio-Rad Laboratories, 60 61 62 10 63 64 65 1 2 3 4 321 Hercules, CA, USA) was used according to the manufacturer’s instructions using 360 ng of RNA 5 6 322 and carried out in an Eppendorf Mastercycler Nexus Gradient thermocycler (Eppendorf, 7 8 323 Hauppauge, NY, USA) according the reaction protocol provided by iScript. 9 10 324 2.9 Quantitative real-time PCR 11 12 325 cDNA was diluted to a working concentration of 0.125 ng/μL. A 20 μL qPCR reaction contained 13 5 pM of each primer, 10 μL 2x iQ SYBR Green Supermix (Bio-Rad), 5 μL nuclease free water 14 326 15 327 and 0.5 ng cDNA template (4 μL of 0.125 ng/μL). qRT-PCR was performed in a Bio-Rad CFX 16 o 17 328 Connect Real-Time System under the following thermal cycle conditions: 3 min at 95 C, 18 followed by 40 replicates of 15 s at 95oC, then 60 – 68oC depending on the melt temperature of 19 329 20 330 the primers used for each gene (see Supplemental Table 1). A dissociative curve calculation 21 22 331 step was completed one time at the end of each. All primers have been previously validated for 23 24 332 efficiency and specificity. All samples were run in technical duplicates and each run included a 25 333 no-template control. 26 27 28 334 qPCR data were analyzed using the Bio-Rad CFX Manager Software, Version 3.0 (Bio-Rad). 29 335 The ∆∆C method was used to calculate fold change compared to β-actin. Values are presented 30 T 31 336 as mean ± SEM, and N is defined as the number of pools of embryos. Expression of the 32 33 337 reference genes (β-actin and elongation factor 1-alpha (ef1a)) was not altered by any of the 34 338 treatments. 35 36 37 339 2.10 38 39 340 Genes that were measured by qPCR were also searched for putative AREs in the region from 40 341 10 kb upstream of the transcriptional start site through the end of the 2nd exon. Zebrafish- 41 42 342 specific (TGA(G/C)nnnTC) and mammalian (TGA(G/C)nnnGC) consensus sequences were 43 44 343 identified, as previously described in (Suzuki et al., 2005) and performed in (Timme-Laragy, et 45 344 al., 2012). The Nrf2-Maf complex is able to bind to both of these sequences in the zebrafish 46 47 345 while maintaining function (Suzuki, et al., 2005). The genomic search was done interrogating 48 49 346 sequences identified by UCSC Genome Browser (Zebrafish Sep. 2014 (GRCz10/danRer10) 50 347 Assembly (http://genome.ucsc.edu/)). 51 52 53 348 2.11 Statistical Analysis 54 55 349 Statview for Windows (version 5.0.1; SAS Institute, Cary, NC, USA) was used to determine 56 350 differences among the data. Data were log-normalized and subjected to a two-factor ANOVA for 57 58 351 genotype and treatment, or plasmid and treatment. If an ANOVA yielded significance (p < 0.05), 59 60 61 62 11 63 64 65 1 2 3 4 352 Fisher's protected least significant differences test was used as a post hoc test. For yolk area 5 6 353 measurements, a F-test and T-test were performed in Excel, using a 2-tailed distribution. 7 8 9 354 3. Results 10 11 355 3.1 The Nrf2a mutant allele encodes a protein deficient in ability to activate transcription 12 13 356 via ARE. 14 fh318 fh319 15 357 To assess the functional consequences of nrf2a and nrf2a mutations, we measured the 16 358 ability of these mutants to activate transcription of an antioxidant response element (ARE) 17 fh318 fh319 18 359 reporter gene (luciferase) in comparison to wild-type Nrf2a protein. Both Nrf2a and Nrf2a 19 20 360 proteins were expressed strongly in COS-7 cells, as measured by Western blot (Fig. 1A). In the 21 361 transient transfection assay, wild-type Nrf2a increased luciferase even in the absence of 22 23 362 exogenous activator, and there was no additional enhancement in the presence of the Nrf2 24 25 363 activator tBHQ (Fig. 1B). Wild-type Nrf2a activity was repressed by co-transfection of a plasmid 26 364 expressing zebrafish Keap1a, but only weakly repressed by Keap1b. tBHQ had no effect on the 27 28 365 activity of Nrf2a in the presence of Keap1a repression. 29 30 In contrast to the wild-type Nrf2a, the transcriptional activity of Nrf2afh318 was greatly diminished, 31 366 32 367 regardless of whether tBHQ was present (Fig. 1B). A small amount of residual activity was 33 34 368 evident, and this was eliminated by co-transfection of Keap1a but only slightly affected by 35 Keap1b. The Nrf2afh319 protein retained nearly full transactivation activity that was slightly 36 369 37 370 enhanced by tBHQ (Fig. 1B). Therefore, the Nrf2afh319 mutant zebrafish line was not used in 38 fh318 39 371 subsequent experiments. Together, these results suggest that the Nrf2a allele encodes a 40 41 372 Nrf2a protein that is greatly diminished in activity but retains some residual activity, at least in 42 373 vitro. 43 44 45 374 3.2 Nrf2a mutants are more sensitive to PCB126 teratogenesis than wildtype 46 375 eleutheroembryos. 47 48 376 Pericardial edema and craniofacial malformations are some of the best characterized 49 50 377 deformities associated with PCB126 and DLC embryotoxicity in fish embryos. To determine 51 fh318/fh318 378 whether nrf2a mutants were more or less sensitive than wild type embryos to PCB126 52 53 379 embryotoxicity, we exposed embryos to 2 nM or 5 nM PCB126 for 1 h from 25-26 hpf, and 54 55 380 measured malformations at 4 dpf and 5 dpf (Fig 2 and 3, respectively). Pericardial edema and 56 jaw deformities were more severe in nrf2afh318/fh318 eleutheroembryos than in wildtype controls 57 381 58 382 when assessed at 4 dpf (Fig 2). At 5 nM PCB126, these deformities are present in only a subset 59 60 383 of exposed embryos, and wildtype eleutheroembryos did not have a statistically significant 61 62 12 63 64 65 1 2 3 4 384 increase in pericardial edema area at 4 dpf, whereas the mutants exposed to 5 nM PCB126 had 5 6 385 nearly twice the pericardial area of controls (Fig. 2D). The mean jaw deformity score was a 7 8 386 more sensitive measure of embryotoxicity, with a significant increase in the deformity score at 2 9 387 nM in the mutants. A significant increase in jaw deformities was observed in both genotypes at 5 10 11 388 nM (Fig. 2E). At both exposure concentrations, the jaw deformities in the mutants were more 12 13 389 severe and more frequent than those in the wildtype eleutheroembryos. However, at 5 dpf, both 14 390 the nrf2a mutant and wildtype eleutheroembryos demonstrated similar pericardial and jaw 15 16 391 deformity severity (Fig. 3C, Fig. 3D). 17 18 We identified two notable developmental differences between the nrf2afh318/fh318 and wildtype 19 392 20 393 controls. First, ventral-dorsal distance, which was used to estimate yolk usage, was significantly 21 fh318/fh318 22 394 lower in nrf2a versus wildtype fish (Figs. 2A, 2B). While the ventral-dorsal distance 23 24 395 increased in a PCB dose-dependent manner in both genotypes, reflecting poor yolk sac 25 396 absorption, it was surprising that the nrf2a mutant DMSO control eleutheroembryos had, on 26 27 397 average, a significantly shorter dorsal-ventral distance than age-matched wildtype fish, by 28 29 398 approximately 8% at 4 dpf (Fig. 2B) At 5 dpf, ventral-dorsal distance between nrf2a mutants and 30 399 wildtype eleutheroembryos remained significantly different (Fig. 3A). To determine if the shorter 31 32 400 dorsal-ventral measurements in the mutant fish was due to nrf2afh318/fh318 fish laying smaller 33 34 401 eggs, we measured the diameter of the yolk in fertilized embryos (3 hpf) from Nrf2a 35 402 homozygous wildtype or nrf2afh318/fh318 adults. Surprisingly, mutants had, on average, larger yolk 36 37 403 sac areas than the wildtype embryos (p < 0.0001; Fig. 4). 38 39 The second notable developmental difference was with respect to the timing of swim bladder 40 404 41 405 inflation. In the mutant fish, only 20% of the DMSO controls had inflated swim bladders at the 4- 42 43 406 dpf timepoint, compared to 90% of the wildtype DMSO controls. Treatment with PCB126 44 45 407 effectively prevented swim bladder inflation in a dose-dependent manner among wildtype fish at 46 408 4 dpf (25% of embryos treated with 2 nM, and 0% treated with 5 nM; Fig 2C and as previously 47 48 409 reported (Jonsson, et al., 2012)). This is in contrast to the nrf2a mutant embryos, where none of 49 50 410 the PCB-treated embryos had successfully inflated swim bladders at 4 dpf (Fig. 2C). However, 51 411 interpretation of these data was confounded by a 24-h delay in swim bladder inflation among the 52 fh318/fh318 53 412 nrf2a eleutheroembryos controls. When assessed at 5 dpf, there were no differences in 54 55 413 swim bladder inflation between the control wildtype or mutant nrf2a genotypes (Fig. 3B). At 5 56 414 dpf, 80% of DMSO controls had inflated swim bladders, while only 40% of embryos treated with 57 58 415 2 nM and 0-10% of embryos treated with 5 nM had successfully inflated swim bladders (Fig. 59 60 416 3B). 61 62 13 63 64 65 1 2 3 4 417 3.3 Regulation of Ahr pathway gene expression by Nrf2a 5 6 418 To confirm that the embryos had been effectively dosed with PCB126, we measured induction 7 8 419 of cyp1a, a primary biomarker of ligand activation of the Ahr. There was no significant difference 9 420 in Cyp1a enzyme activity (in vivo EROD activity, data not shown), and no difference in cyp1a 10 11 421 gene expression between genotypes in the DMSO control (Fig. 5A). There was a significant 12 13 422 dose-dependent increase in gene expression of cyp1a in response to 2 and 5 nM PCB126 in 14 423 both genotypes, but the nrf2a mutants had nearly double the induction of cyp1a compared to 15 16 424 wildtype fish. 17 18 To determine whether the enhanced cyp1a induction by PCB126 was due to differential 19 425 20 426 expression of ahr genes, we measured expression of ahr2 (the ortholog that has been 21 22 427 demonstrated to regulate co-planar PCB and TCDD embryotoxicity). Expression of ahr2 was not 23 24 428 significantly different between the two genotypes at 4 dpf. The expression of ahr2 was induced 25 429 by PCB126 exposure in a dose-dependent manner and to a similar extent in both mutant and 26 27 430 wildtype embryos (Fig. 5B). 28 29 431 Because there have been some reports that cyp1a induction can be regulated by the other ahr 30 31 432 genes, ahr1a and ahr1b, we also measured expression of these genes. Expression of ahr1a 32 33 433 was similar in control embryos of the two genotypes, but the magnitude of induction following 34 434 PCB126 treatment differed between genotypes. ahr1a was only inducible in the nrf2a wildtype 35 36 435 fish (2.6 and 4.3 fold for the 2 nM and 5 nM exposures respectively, Fig. 5C); expression of this 37 fh318/fh318 38 436 gene did not change with PCB126 treatment in the nrf2a eleutheroembryos at 4 dpf. 39 Similarly, expression of ahr1b was slightly induced by PCB126 in wild type embryos (2-fold with 40 437 41 438 5 nM) but was not induced in the nrf2a mutants (Fig 5D). Interestingly, the nrf2a mutants had 42 43 439 significantly lower basal expression (70-75% lower) of ahr1b compared to the wildtype fish at 44 45 440 this stage. 46 47 441 Basal expression of the two Ahr repressor genes, ahrra and ahrrb, both tended to be lower in 48 49 442 the nrf2a mutants, but this was not statistically significant. Both of these genes demonstrated 50 443 dose-dependent increases in expression in both genotypes with 2 nM and 5 nM PCB126 51 52 444 treatments, but the induction level was greater in the nrf2a wildtype fish (Figs. 5E, 5F). Wildtype 53 54 445 zebrafish showed a 21.5- and 34.6- fold increase in expression of ahrra at the 2 nM and 5 nM 55 446 PCB126 treatments respectively, compared to 14.5- and 27.5-fold increases in the mutant fish 56 57 447 (Fig. 5E). Inducible expression of ahrrb was not as high as ahrra; wildtype expression increased 58 59 60 61 62 14 63 64 65 1 2 3 4 448 10.4 and 23.9-fold in the wildtypes, and 3.2- and 6.8-fold in the mutants, approximately 10-fold 5 6 449 lower than ahrra expression at these doses of PCB126 (Fig. 5F). 7 8 450 3.4 Regulation of Nrf-family and Keap genes by Nrf2a 9 10 451 The influence of nrf2a function on the expression of other nrf genes was investigated. There 11 12 452 were no statistically significant differences in the basal expression of either nrf1a or nrf1b 13 between the wildtype and nrf2a mutant eleutheroembryos (Figs. 6A, 6B). Both genes were 14 453 15 454 upregulated with PCB126 exposure in the wildtype fish only. Expression of nrf1a was 2-fold 16 17 455 greater in the eleutheroembryos exposed to PCB126 treatment (Fig. 6A). Expression of nrf1b 18 was 4-fold higher in the 2 nM PCB126 exposure group, but not significantly different from 19 456 20 457 controls at the 5 nM concentration (Fig. 6B), suggesting an inverted-U-shaped curve. We found 21 22 458 no difference in basal expression of nrf3 between the wildtype and nrf2a mutant 23 24 459 eleutheroembryos, but there was a significant 2-fold induction of nrf3 with PCB126 only in the 25 460 wildtype embryos (Fig. 6D). 26 27 28 461 To ensure that transcription of nrf2a was indeed comparable between the wildtype and mutants, 29 462 we also measured gene expression of nrf2a. There was no significant difference in basal 30 31 463 expression between the genotypes, but there was a difference in inducibility by PCB126 (Fig. 32 33 464 6C). The expression of nrf2a was significantly up-regulated in response to PCB126 in both 34 465 genotypes, but this up-regulation was greater in the wildtype eleutheroembryos (1.7-fold). While 35 36 466 nrf2a expression in the mutant embryos was increased by PCB126 exposure, this increase did 37 38 467 not exceed the basal levels in the wildtype eleutheroembryos (Fig. 6C). 39 40 468 Expression of both keap1a and keap1b genes, the cytosolic repressors of Nrf2a, tended to be 41 42 469 lower in the nrf2a mutant eleutheroembryos, but this was not statistically significant (Figs. 6E, 43 44 470 6F). Both were inducible by PCB126 in the wildtype but not nrf2a mutant eleutheroembryos. 45 471 Expression of keap1a and keap1b increased 2- fold in the wildtype embryos at the 2 nM 46 47 472 treatment but only 1.5- fold in the 5 nM treatment. 48 49 50 473 We measured expression of nrf2b, a repressive paralog of nrf2a, and found a dose-dependent 51 474 increase only in the mutant embryos exposed to PCB126 (5.2-fold and 6.8-fold at the 2 nM and 52 53 475 5 nM exposures respectively; Fig. 7A). We then measured expression of hmox1 and tp53, two 54 55 476 genes negatively regulated by Nrf2b (Timme-Laragy, et al., 2012). Expression of hmox1 56 477 increased 3-fold only in the nrf2a wildtype embryos at 2 nM PCB126, but not at 5 nM (Fig. 7B). 57 58 478 Expression of tp53 did not change significantly with PCB126 exposure in either genotype, 59 60 479 although there was a significant difference in the expression levels between the two genotypes 61 62 15 63 64 65 1 2 3 4 480 with the 5 nM treatment (1.5-fold increase in wildtype and 0.7-fold decrease in Nrf2a mutants; 5 6 481 Fig. 7C). 7 8 482 3.5 Measures of oxidative stress responsive genes in PCB126 embryotoxicity 9 10 483 To determine whether a transcriptional response to PCB126-induced oxidative stress was 11 12 484 evident in this study, we measured three well-established genes characteristic of an oxidative 13 stress response: heat shock protein 70 (hsp70), glutathione-s-transferase pi 1 (gstp1) and 14 485 15 486 glutamate-cysteine ligase catalytic subunit (gclc). If PCB126 were causing oxidative stress, it 16 17 487 would be expected that these genes would be upregulated in the wildtype embryos exposed to 18 PCB126; however, in these exposures, there was no change in expression of these genes with 19 488 20 489 PCB126 exposure and no difference between the two genotypes (Supplemental Fig. 2). 21 22 23 490 3.6 Gene targets co-regulated by Ahr and Nrf2 24 491 Induction of the genes gsta1 and nqo1 can be regulated by both AHR and NRF2 after treatment 25 26 492 with TCDD in mice (Yeager, et al., 2009). Interestingly, nrf2a mutant eleutheroembryos had 27 28 493 higher basal expression of both of these genes, 4-5- fold more gsta1 and 2-fold higher 29 494 expression of nqo1 than the wildtype embryos at the DMSO, 2 nM and 5 nM PCB 126 30 31 495 treatment; however, there was no significant response to PCB126 (Fig. 8). 32 33 34 496 3.7 Putative AREs can be found across Ahr and Nrf2 target promoters 35 497 To determine the potential for Nrf2 to have a direct effect on the abundance and function of the 36 37 498 genes, a bioinformatic search for putative mammalian (M) or zebrafish-specific (Z) AREs was 38 39 499 performed on all candidate genes. Mammalian and/or zebrafish-specific putative AREs were 40 500 found in each gene (Table 1). Several genes had AREs within 1000 bp upstream or 100 bp 41 42 501 downstream of the transcription start site, often deemed key regulatory zones. These genes 43 44 502 include ahr1b, ahr2, ahrra, ahrrb, cyp1a, gstp1, hmox1, hsp70, keap1a, and nrf1b. A subset of 45 503 these genes have AREs within 100 bp of the beginning of Exon 1, including hsp70, nrf1b, gstp1, 46 47 504 ahrra, and ahr2. 48 49 50 505 4. Discussion 51 506 NRF2 and AHR have been shown to positively regulate the expression of one another in 52 53 507 mammals (Ma, et al., 2004; Miao, et al., 2005; Shin et al., 2007). Nrf2 in both fish and 54 55 508 mammals can also regulate its own expression (Kwak et al., 2002; Timme-Laragy, et al., 2012; 56 Williams et al., 2013). In this study, we examined PCB126 embryotoxicity and gene expression 57 509 58 510 during the sensitive period of embryonic development, with a focus on these two important 59 fh318 60 511 pathways, Ahr and Nrf2. We have also further characterized the nrf2a recessive loss-of- 61 62 16 63 64 65 1 2 3 4 512 function mutant fish with respect to some subtle developmental effects and differences in basal 5 6 513 gene expression. While these zebrafish mutants have been previously characterized 7 8 514 (Mukaigasa, et al., 2012), and demonstrated to be similar to the NRF2 null mouse in some 9 515 respects (Mukaigasa, et al., 2012), our luciferase data from over-expressed Nrf2afh318 mutant 10 11 516 protein suggests that there may be some slight DNA binding activity, at least in vitro. Our 12 319 13 517 assessment of the Nrf2a mutant protein indicated that it maintained effective interactions with 14 518 the ARE, consistent with earlier results (Mukaigasa, et al., 2012). 15 16 17 519 Mammalian models lacking NRF2 and zebrafish with either mutated or knocked-down Nrf2a 18 have demonstrated an increased sensitivity to toxicants including TCDD (Lu, et al., 2011), PAHs 19 520 20 521 (Timme-Laragy et al., 2009), and pro-oxidants (Mukaigasa, et al., 2012; Suzuki, et al., 2005; 21 22 522 Timme-Laragy, et al., 2012). In the present study, we used relatively low concentrations of 23 fh318/fh318 24 523 PCB126 that were not expected to cause severe and pervasive deformities. Yet, nrf2a 25 524 fish were found to be more sensitive to PCB126 embryotoxicity, as measured by pericardial 26 27 525 edema area and craniofacial malformations; a surprising result was the transient nature of this 28 29 526 effect. We show here that at 4 dpf the nrf2a mutant embryos have more severe morphological 30 527 effects of PCB126 exposure (Fig. 2), but by 5 dpf, these effects have largely equalized between 31 32 528 the genotypes (Fig. 3). 33 34 fh318/fh318 529 This study identified subtle developmental differences between the wildtype and nrf2a 35 36 530 embryos and eleutheroembryos. There was a one day delay in swim bladder inflation, which is 37 38 531 one of the hallmarks of embryotoxicity caused by exposure to PCB126 and other dioxin-like- 39 compounds (DLCs), and a shortened dorsal-ventral measure, reflecting a slightly smaller yolk 40 532 41 533 sac (Figs. 2-4). Since the mutant embryos start out with larger yolks (at 3 hpf) than wildtype 42 43 534 embryos, we suggest that this smaller yolk may be due to an increased rate of yolk utilization by 44 fh318/fh318 45 535 the nrf2a embryos, and are conducting a follow-up study to better understand this. 46 536 While previous studies have demonstrated that NRF2 is not essential for normal development 47 48 537 and reproduction in both zebrafish and mice (Chan et al., 1996; Itoh, et al., 1997; Mukaigasa, 49 50 538 et al., 2012) some subtle effects in NRF2 mutant mice have been identified. NRF2 -\- mice have 51 539 defects in liver growth and size (Beyer, et al., 2008), reduced bile duct microbranching (Skoko, 52 53 540 et al., 2014), reduced retinal vasculature density (Wei, et al., 2013), and in some genetic 54 55 541 backgrounds, a congenital defect resulting in an intrahepatic shunt (Skoko, et al., 2014). 56 57 542 We identified putative AREs throughout the promoter regions of genes involved in the Ahr, Nrf2, 58 59 543 and general stress responses (Table 1). Several Ahr targets and pathway genes have putative 60 61 62 17 63 64 65 1 2 3 4 544 AREs in their promoters within 1000 bp upstream or 100 bp downstream of the beginning of 5 6 545 Exon 1. These genes include ahr1b, ahr2, ahrra, ahrrb, and cyp1a. Because of the close 7 8 546 proximity of these loci to the TATAA box, it is highly possible that Nrf2 binding in these regions 9 547 may in some way affect transcriptional activation. Deletion of this region of the gstp1 gene in 10 11 548 zebrafish has been demonstrated to completely eliminate induction by Nrf2a (Suzuki, et al., 12 13 549 2005). Though indirect crosstalk between the Ahr and Nrf2 signaling pathways is likely, it is also 14 550 possible that Nrf2 may directly activate these genes. However, these are putative AREs and 15 16 551 thus confirmation of Nrf2 transcription factor binding and activation of these targets would be 17 18 552 required in order to determine whether Nrf2 does activate or repress transcription of these 19 553 genes. Several other groups have performed ChIP-seq in various tissues and cell lines in order 20 21 554 to discern whether basal and/or induced NRF2 has any novel targets in various tissues 22 23 555 (Campbell et al., 2013; Chorley et al., 2012; Malhotra, et al., 2010). Though these studies 24 556 confirmed NRF2 binding to a similar core of genes associated with the antioxidant response, 25 26 557 such as those included in glutathione synthesis and phase II metabolism, these studies also 27 28 558 identified novel targets for each cell line or tissue. 29 30 559 We identified significant changes in gene expression during development among the ahr target 31 32 560 genes that were altered in the absence of a fully-functional Nrf2a, with respect to both basal and 33 34 561 PCB126-inducible expression (summarized in Table 2). For example, nrf2a mutants had nearly 35 562 double the induction of cyp1a compared to wildtype fish in response to PCB126 (Fig.5A). As Ahr 36 37 563 activation is directly related to the severity of deformities after exposure to DLCs, this finding 38 39 564 concurs with the enhanced severity of deformities observed in these mutants at the 4 dpf 40 565 timepoint (Fig. 2). However, there are conflicting findings in the literature with regards to the role 41 42 566 of Nrf2 in expression of cyp1a. In some studies of adult NRF2 knockout mice treated with 43 44 567 TCDD, expression of Cyp1a1 was lower than in wildtype livers (Anwar-Mohamed, et al., 2011), 45 568 as was expression of Ahr, Cyp1a1, and Cyp1b1 in mouse embryo fibroblasts derived from 46 47 569 NRF2 knockout mice (Shin, et al., 2007). Still others found no change in the basal or inducible 48 49 570 expression of Cyp1a1 or Cyp1a2 in livers of adult NRF2 knockout mice (Aleksunes and 50 571 Klaassen, 2012; Noda et al., 2003; Yeager, et al., 2009). There may be several reasons for 51 52 572 this discrepancy in results pertaining to the basal and inducible expression of cyp1a in our study 53 54 573 and in the literature overall. There may be differences related to developmental events or the 55 574 inclusion of whole embryo gene expression compared to isolated adult liver tissues measured in 56 57 575 other studies. There may also be differences related to the dose and the inducer used. 58 59 576 60 61 62 18 63 64 65 1 2 3 4 577 The greater induction in cyp1a expression in Nrf2a mutants in our study may also be due to 5 6 578 concurrent changes in expression of the aryl hydrocarbon receptor repressor (ahrr) genes. It is 7 8 579 conceivable that the diminished induction in expression of the ahrr genes could result in 9 580 diminished repression of Ahr2 transcriptional activity and subsequently contribute to the 10 11 581 enhanced expression of cyp1a in the mutants. The molecular mechanism by which Nrf2a affects 12 13 582 expression of ahrra and ahrrb is also unknown and requires further investigation. However, both 14 583 ahrr genes contain putative AREs in their promoters (7 in ahrra, and 6 in ahrrb; Table 1). For 15 16 584 both ahrra and ahrrb, there is an ARE proximal to the transcription start site, suggesting that 17 18 585 they may be directly induced by Nrf2 activation. Increasing PCB concentrations increased ahrra 19 586 and ahrrb expression in nrf2a mutants, albeit to a significantly lesser degree for ahrrb (Fig. 5). 20 21 587 To our knowledge, the function of AHRR has not yet been examined in NRF2 knockout mice. 22 23 588 24 589 The induction of cyp1a in zebrafish is largely regulated by Ahr2, although some studies suggest 25 26 590 that Ahr1a, which is not capable of binding to TCDD (Karchner, et al., 2005), may have some 27 28 591 role in the endogenous regulation of cyp1a (Garner, et al., 2013; Goodale, et al., 2012; 29 592 Incardona et al., 2006). It has been shown that a knockdown of Ahr1a via morpholino 30 31 593 exacerbated cardiac toxicity of PCB126 and increased Cyp1a activity (EROD activity) in 32 33 594 exposed embryos (Garner, et al., 2013). A knockdown of Ahr1b showed no such visible or 34 595 quantitative effect on PCB126 toxicity (Garner, et al., 2013). In another study, TCDD was unable 35 36 596 to induce ahr1a expression in zebrafish larvae at 48 or 72 hpf (Karchner, et al., 2005). In the 37 38 597 present study at 4 dpf, ahr1a and ahr1b were expressed more highly in wild-type embryos than 39 598 mutants and there was a dose-dependent increase in expression with increasing concentration 40 41 599 of PCBs in wild-type embryos (Fig. 5). Interestingly, there was no induction in the mutant 42 43 600 embryos. This would suggest that Nrf2 activation may be necessary for upregulation of ahr1a 44 601 and ahr1b; the presence of putative AREs in their promoters provides additional support for this 45 46 602 mechanism. 47 48 49 603 It is important to note here that not all Ahr target genes responded in a similar manner in this 50 604 study (Table 2). For example, there are conflicting patterns with cyp1a being more inducible in 51 52 605 the nrf2a mutants, while the ahrr genes were less inducible. This has also been reported by 53 54 606 others (Yeager, et al., 2009) (Nukaya et al., 2010), and reflects the complex nature of gene 55 607 network interactions that will require further study to dissect the contribution of Nrf2. 56 57 608 58 59 60 61 62 19 63 64 65 1 2 3 4 609 Nqo1 has been categorized as part of the mammalian “AhR battery of genes.” However, Ahr 5 6 610 crosstalk with Nrf2 may be necessary for proper induction of Nqo1 (Yeager et al, 2009, Wang et 7 8 611 al, 2013). NRF2-null mice have shown a decrease in NQO1 protein when exposed to 3-MC, an 9 612 inducer of Phase 1 genes, and BHA, an inducer of phase 2 genes (Jin et al., 2014). Likewise, 10 11 613 Gstp, a gene typically associated with the “Nrf2 battery of genes”, experienced a decrease in 12 13 614 protein expression when treated with BHA in AHR-null mutant mice, thus suggesting a 14 615 relationship between Nrf2 and Ahr in the regulation of these genes (Noda, et al., 2003). Our 15 16 616 results show an unexpected increase in basal expression and induction of nqo1, and a lack of 17 18 617 induction of gstp1 when treated with low concentrations of the Ahr agonist, PCB126, in the nrf2a 19 618 mutants (Fig 8). Though our nqo1 expression pattern is different from mammalian studies, it is 20 21 619 consistent with what has been found in the zebrafish model using the Nrf2 activator tBHQ, the 22 23 620 Ahr activator BaP, or water soluble oil fractions (Corrales et al., 2014; dos Anjos et al., 2011; 24 621 Hahn, et al., 2014). 25 26 27 622 In the current study, we found that some Phase II genes we measured (nqo1, gsta1) were 28 29 623 expressed at higher levels in the nrf2a mutant fish compared to wildtype controls, with no 30 624 significant change in expression associated with PCB126 exposure in either genotype (Fig. 8). 31 32 625 This is contrary to what has been found for most other Phase II genes, whose expression in 33 34 626 adult mice has been shown to increase with NRF2 activation, and decrease in the NRF2 null 35 627 mouse model, the exception being Sult1a1, which decreased with Nrf2 activation (Wu et al., 36 37 628 2012). Although both NQO1 and GSTa1 are NRF2 targets in mammalian systems, there are 38 39 629 contradictory data concerning whether this is true in zebrafish (Hahn, et al., 2014; Kobayashi, 40 630 et al., 2002; Nakajima et al., 2011). Other genes have been found to be over-expressed in 41 42 631 Nrf2afh318/fh318 larvae, including gpx1b (Mukaigasa, et al., 2012). This upregulation may be 43 44 632 compensatory to counteract the reduced antioxidant response capacity in the mutant embryos. 45 It is important in interpreting our results to keep in mind the nature of this Nrf2a mutation; all 46 633 47 634 other functional elements of the Nrf2a protein remain intact, enabling full interactions with Keap1 48 49 635 and other proteins. This is inherently different from the NRF2 null mouse, in which the much of 50 the gene was removed and no functional NRF2 protein synthesized (Chan, et al., 1996). 51 636 52 53 637 Other members of the nrf gene family are capable of binding to AREs and regulating gene 54 55 638 expression, and mammalian NRF1 and NRF2 have been shown to have some overlap in the 56 639 sets of genes that they regulate (Ohtsuji, et al., 2008). We therefore considered the possibility 57 58 640 that in the nrf2a mutant embryos, other nrf family members may be compensating for the 59 60 641 deficient transcriptional activity of Nrf2a. Previous research demonstrated that the inducible, but 61 62 20 63 64 65 1 2 3 4 642 not basal, expression of nrf1a and nrf1b following exposure to tert butylhydroperoxide was 5 6 643 dependent on the presence of Nrf2a (Williams, et al., 2013). However, we found no differences 7 8 644 in the basal expression of these genes, although there was a trend towards a reduced basal 9 645 expression in nrf1a, and nrf1b (Fig. 6). Williams et al. identified ARE and XREs in the promoters 10 11 646 of the Nrf family members (nrf1a, nrf1b, nrf2b, nrf3, ), and with the exception of nrf3, 12 13 647 expression each of these can be upregulated by tert butylhydroperoxide (Williams, et al., 2013). 14 648 In Nrf2a-morpholino embryos, the upregulation of nfe2, nrf1a, and nrf1b was either lost or 15 16 649 greatly diminished, but the upregulation of nrf3 was enhanced by Nrf2a knockdown. Here we 17 18 650 found that these nrf family genes, as well as the cytosolic repressor proteins keap1a and 19 651 keap1b were inducible by PCB126, but only in the wildtype eleutheroembryos (Fig. 6), indicating 20 21 652 that Nrf2a is an important contributor to this response. 22 23 24 653 Our previous studies of the novel nrf2a gene paralog, nrf2b, identified a role for this gene in the 25 654 negative transcriptional regulation of tp53 and hmox1, and also showed that inducible 26 27 655 expression of both nrf2a and nrf2b was dependent on Ahr2 (Timme-Laragy, et al., 2012). We 28 29 656 expand upon that finding here, and have shown that inducible expression of nrf2b with PCB126 30 657 is also dependent on Nrf2a (Fig. 7). Our data also provides support for the previously identified 31 32 658 negative transcriptional regulation of hmox1 by Nrf2b: the induction of hmox1 by PCB126 in the 33 34 659 wildtype embryos was not observed in the nrf2a mutant embryos, and this was correlated with 35 660 the increased expression of nrf2b. 36 37 38 661 It is well-established that PCB126 is capable of causing oxidative stress in mammals, cell 39 culture, and fish models (Arzuaga et al., 2006; Chen, 2010; Hassoun and Periandri-Steinberg, 40 662 41 663 2010; Park et al., 2010). For example, Schlezinger et al. demonstrated that ROS can be 42 43 664 generated from PCB126 exposure via the uncoupling of the catalytic cycle of Cyp1a 44 45 665 (Schlezinger, et al., 2006). Other groups have demonstrated oxidative stress resulting from 46 666 PCB126 exposure in zebrafish embryos, including a transcriptional oxidative stress response; 47 48 667 however, the concentrations used in those studies were 100 nM (Na et al., 2009) and 64 and 49 50 668 128 µg/L (Liu et al., 2014) which are drastically greater than the highest dose used in this study 51 669 (5 nM or 7 µg/L). ROS generated at low PCB126 concentrations may be difficult to detect, and 52 53 670 may not have been captured in our experimental design. Or perhaps oxidative stress may not 54 55 671 be a primary factor contributing to the deformities caused by low concentrations of PCB126. 56 672 Another point of consideration is that nrf2a fh318/fh318 fish have been shown to be more sensitive 57 58 673 to some, but not all, sources of ROS and electrophiles. For example, while nrf2a mutants were 59 60 674 more sensitive to peroxides and acetaminophen, they were only moderately more sensitive to 61 62 21 63 64 65 1 2 3 4 675 paraquat, and not more sensitive to menadione (Mukaigasa, et al., 2012). The only indication of 5 6 676 an oxidative stress response in our data was the upregulation of hmox1 by PCB126; otherwise 7 8 677 we did not identify changes in gene expression that would be typical of an oxidative stress 9 678 response (Hahn, et al., 2014). This study used low concentrations of PCB126, yet still 10 11 679 demonstrated an important role for Nrf2a in the transcriptional response and embryotoxicity. 12 13 680 The mechanistic function of Nrf2a in these processes is not entirely clear, but these studies 14 681 suggest a noncanonical role of Nrf2a, or that interaction with the Ahr itself may be important. 15 16 682 Wang et al. provided evidence suggesting that AHR upregulated the transcription of Nrf2 and 17 18 683 formed protein complexes with both NRF2 and KEAP1 that contributed to the stability of NRF2 19 684 and subsequent ARE-activation (Wang, et al., 2013). Thus, regulation of Nrf2 by Ahr could 20 21 685 occur via transcriptional regulation as well as direct protein-protein interactions. 22 23 24 686 In the age of transcriptomics, it is becoming increasingly important to understand the role of 25 687 crosstalk between transcriptional pathways. It is no longer effective to think of pathways as 26 27 688 simply linear models; rather, they need to be examined as networks that interact with each 28 29 689 other. Studying the transcriptional regulators of such gene networks, and their interactions, will 30 690 be important to the field of toxicology in understanding toxicant responses and predicting 31 32 691 toxicity. 33 34 35 692 5. Conclusions 36 693 Here we have taken a targeted candidate gene approach to focus on important genes in the 37 38 694 Nrf2a and Ahr2 pathways. We have shown complex interactions between these regulatory 39 40 695 pathways, and resulting consequences for PCB126 embryotoxicity. We have suggested that the 41 696 Nrf2a fh318 mutant allelic variant may be slightly functional at the transcriptional level, at least in 42 43 697 vitro. We identified some important developmental differences between the mutants and the 44 45 698 wildtype embryos, and a transient increase in sensitivity to PCB126 embryotoxicity in the nrf2a 46 699 mutants. This is correlated with a higher induction of cyp1a in these mutants. Finally, we 47 48 700 demonstrated that expression of members of the ahr and nrf2 gene families are affected by 49 50 701 Nrf2a. These findings will have important consequences for the use of these mutant fish by 51 702 others, and also contribute to our understanding of crosstalk between two pathways that are 52 53 703 important to the study of toxicology. 54 55 56 704 57 58 59 60 61 62 22 63 64 65 1 2 3 4 705 Supplemental Data 5 6 706 Supplemental Table 1. List of qPCR primers, melting temperature, and sources. 7 8 fh318 9 707 Supplemental Figure 1. Protein alignment of the translated nrf2a allele with other Nrf2a 10 708 protein sequences. 11 12 13 709 Supplemental Fig.2. . Quantitative real-time PCR measures for oxidative stress response genes 14 15 710 in nrf2a wildtype and fh318 mutant eleutheroembryos at 4 dpf. 16 17 711 [These figure legends are shown at the end of the paper 18 19 712 ] 20 21 713 Funding 22 23 714 University of Massachusetts Amherst Commonwealth Honors College Research grant (to M.R.), 24 25 715 National Institutes of Health grants R01ES016366 (MEH), R01ES006272 (MEH), and 26 F32ES017585 (ART-L). 27 716 28 29 717 30 31 32 718 Acknowledgements 33 719 We would like to acknowledge the excellent fish care provided by Brandy Joyce and Gale Clark 34 35 720 at WHOI, and by Karen Melendez, Derek Luthi, Sonia Filipczak, Marjorie Marin, Jiali Xu, 36 37 721 Christopher Sparages, Shana Fleishman, Kaylee-Anna Williams, Katrina Borofski, and Gabriella 38 722 McClellen at UMASS. 39 40 41 723 42 43 44 724 References 45 725 Aleksunes, L. M., and Klaassen, C. D. (2012). Coordinated regulation of hepatic phase I and II 46 47 726 drug-metabolizing genes and transporters using AhR-, CAR-, PXR-, PPARalpha-, and 48 727 Nrf2-null mice. Drug metabolism and disposition: the biological fate of chemicals 40(7), 49 728 1366-79, 10.1124/dmd.112.045112. 50 729 Alexeyenko, A., Wassenberg, D. M., Lobenhofer, E. K., Yen, J., Linney, E., Sonnhammer, E. L., 51 730 and Meyer, J. N. (2010). Dynamic zebrafish interactome reveals transcriptional 52 731 mechanisms of dioxin toxicity. PloS one 5(5), e10465, 10.1371/journal.pone.0010465. 53 732 Andreasen, E. A., Spitsbergen, J. M., Tanguay, R. L., Stegeman, J. J., Heideman, W., and 54 733 Peterson, R. E. (2002). Tissue-specific expression of AHR2, ARNT2, and CYP1A in 55 734 zebrafish embryos and larvae: effects of developmental stage and 2,3,7,8- 56 57 735 tetrachlorodibenzo-p-dioxin exposure. Toxicological sciences : an official journal of the 58 736 Society of Toxicology 68(2), 403-19. 59 737 Anwar-Mohamed, A., Degenhardt, O. S., El Gendy, M. A., Seubert, J. M., Kleeberger, S. R., and 60 738 El-Kadi, A. O. (2011). The effect of Nrf2 knockout on the constitutive expression of drug 61 62 23 63 64 65 1 2 3 4 739 metabolizing enzymes and transporters in C57Bl/6 mice livers. Toxicology in vitro : an 5 740 international journal published in association with BIBRA 25(4), 785-95, 6 741 10.1016/j.tiv.2011.01.014. 7 8 742 Arzuaga, X., Wassenberg, D., Di Giulio, R., and Elskus, A. (2006). The chlorinated AHR ligand 9 743 3,3',4,4',5-pentachlorobiphenyl (PCB126) promotes reactive oxygen species (ROS) 10 744 production during embryonic development in the killifish (Fundulus heteroclitus). Aquatic 11 745 toxicology (Amsterdam, Netherlands) 76(1), 13-23, 10.1016/j.aquatox.2005.07.013. 12 746 Beyer, T. A., Xu, W., Teupser, D., auf dem Keller, U., Bugnon, P., Hildt, E., Thiery, J., Kan, Y. 13 747 W., and Werner, S. (2008). Impaired liver regeneration in Nrf2 knockout mice: role of 14 748 ROS-mediated insulin/IGF-1 resistance. The EMBO journal 27(1), 212-23, 15 749 10.1038/sj.emboj.7601950. 16 Billiard, S. M., Timme-Laragy, A. R., Wassenberg, D. M., Cockman, C., and Di Giulio, R. T. 17 750 18 751 (2006). The role of the aryl hydrocarbon receptor pathway in mediating synergistic 19 752 developmental toxicity of polycyclic aromatic hydrocarbons to zebrafish. Toxicological 20 753 sciences : an official journal of the Society of Toxicology 92(2), 526-36, 21 754 10.1093/toxsci/kfl011. 22 755 Bock, K. W. (2013). The human Ah receptor: hints from dioxin toxicities to deregulated target 23 756 genes and physiological functions. Biological chemistry 394(6), 729-39, 10.1515/hsz- 24 757 2012-0340. 25 758 Bryan, H. K., Olayanju, A., Goldring, C. E., and Park, B. K. (2013). The Nrf2 cell defence 26 27 759 pathway: Keap1-dependent and -independent mechanisms of regulation. Biochemical 28 760 pharmacology 85(6), 705-17, 10.1016/j.bcp.2012.11.016. 29 761 Campbell, M. R., Karaca, M., Adamski, K. N., Chorley, B. N., Wang, X., and Bell, D. A. (2013). 30 762 Novel hematopoietic target genes in the NRF2-mediated transcriptional pathway. 31 763 Oxidative medicine and cellular longevity 2013, 120305, 10.1155/2013/120305. 32 764 Chan, K., Lu, R., Chang, J. C., and Kan, Y. W. (1996). NRF2, a member of the NFE2 family of 33 765 transcription factors, is not essential for murine erythropoiesis, growth, and development. 34 766 Proceedings of the National Academy of Sciences of the United States of America 35 93(24), 13943-8. 36 767 37 768 Chen, F. (2010). Induction of oxidative stress and cytotoxicity by PCB126 in JEG-3 human 38 769 cells. Journal of environmental science and health. Part A, 39 770 Toxic/hazardous substances & environmental engineering 45(8), 932-7, 40 771 10.1080/10934521003772311. 41 772 Chorley, B. N., Campbell, M. R., Wang, X., Karaca, M., Sambandan, D., Bangura, F., Xue, P., 42 773 Pi, J., Kleeberger, S. R., and Bell, D. A. (2012). Identification of novel NRF2-regulated 43 774 genes by ChIP-Seq: influence on alpha. Nucleic acids research 44 775 40(15), 7416-7429, 10.1093/nar/gks409. 45 46 776 Corrales, J., Fang, X., Thornton, C., Mei, W., Barbazuk, W. B., Duke, M., Scheffler, B. E., and 47 777 Willett, K. L. (2014). Effects on specific promoter DNA methylation in zebrafish embryos 48 778 and larvae following benzo[a]pyrene exposure. Comparative and 49 779 physiology. Toxicology & pharmacology : CBP 163, 37-46, 10.1016/j.cbpc.2014.02.005. 50 780 Cullinan, S. B., Gordan, J. D., Jin, J., Harper, J. W., and Diehl, J. A. (2004). The Keap1-BTB 51 781 protein is an adaptor that bridges Nrf2 to a Cul3-based E3 ligase: oxidative stress 52 782 sensing by a Cul3-Keap1 ligase. Molecular and cellular biology 24(19), 8477-86, 53 783 10.1128/mcb.24.19.8477-8486.2004. 54 Dalton, T. P., Puga, A., and Shertzer, H. G. (2002). Induction of cellular oxidative stress by aryl 55 784 56 785 hydrocarbon receptor activation. Chemico-biological interactions 141(1-2), 77-95. 57 786 Denison, M. S., Soshilov, A. A., He, G., DeGroot, D. E., and Zhao, B. (2011). Exactly the same 58 787 but different: promiscuity and diversity in the molecular mechanisms of action of the aryl 59 788 hydrocarbon (dioxin) receptor. Toxicological sciences : an official journal of the Society 60 789 of Toxicology 124(1), 1-22, 10.1093/toxsci/kfr218. 61 62 24 63 64 65 1 2 3 4 790 dos Anjos, N. A., Schulze, T., Brack, W., Val, A. L., Schirmer, K., and Scholz, S. (2011). 5 791 Identification and evaluation of cyp1a transcript expression in fish as molecular 6 792 biomarker for petroleum contamination in tropical fresh water ecosystems. Aquatic 7 8 793 toxicology (Amsterdam, Netherlands) 103(1-2), 46-52, 10.1016/j.aquatox.2011.02.004. 9 794 Garner, L. V., Brown, D. R., and Di Giulio, R. T. (2013). Knockdown of AHR1A but not AHR1B 10 795 exacerbates PAH and PCB-126 toxicity in zebrafish (Danio rerio) embryos. Aquatic 11 796 toxicology (Amsterdam, Netherlands) 142-143, 336-46, 10.1016/j.aquatox.2013.09.007. 12 797 Gillardin, V., Silvestre, F., Dieu, M., Delaive, E., Raes, M., Thome, J. P., and Kestemont, P. 13 798 (2009). Protein expression profiling in the African clawed frog Xenopus laevis tadpoles 14 799 exposed to the polychlorinated biphenyl mixture aroclor 1254. Molecular & cellular 15 800 proteomics : MCP 8(4), 596-611, 10.1074/mcp.M800323-MCP200. 16 Goodale, B. C., La Du, J. K., Bisson, W. H., Janszen, D. B., Waters, K. M., and Tanguay, R. L. 17 801 18 802 (2012). AHR2 mutant reveals functional diversity of aryl hydrocarbon receptors in 19 803 zebrafish. PloS one 7(1), e29346, 10.1371/journal.pone.0029346. 20 804 Hahn, M. E., Karchner, S. I., Evans, B. R., Franks, D. G., Merson, R. R., and Lapseritis, J. M. 21 805 (2006). Unexpected diversity of aryl hydrocarbon receptors in non-mammalian 22 806 vertebrates: insights from comparative genomics. Journal of experimental zoology. Part 23 807 A, Comparative experimental biology 305(9), 693-706, 10.1002/jez.a.323. 24 808 Hahn, M. E., McArthur, A. G., Karchner, S. I., Franks, D. G., Jenny, M. J., Timme-Laragy, A. R., 25 809 Stegeman, J. J., Woodin, B. R., Cipriano, M. J., and Linney, E. (2014). The 26 27 810 transcriptional response to oxidative stress during vertebrate development: effects of 28 811 tert-butylhydroquinone and 2,3,7,8-tetrachlorodibenzo-p-dioxin. PloS one 9(11), 29 812 e113158, 10.1371/journal.pone.0113158. 30 813 Hahn, M. E., Timme-Laragy, A. R., Karchner, S. I., and Stegeman, J. J. (2015). Nrf2 and Nrf2- 31 814 related proteins in development and developmental toxicity:Insights from studies in 32 815 zebrafish (danio rerio). Free radical biology & medicine doi: 33 816 10.1016/j.freeradbiomed.2015.06.022, 10.1016/j.freeradbiomed.2015.06.022. 34 817 Harbeitner, R. C., Hahn, M. E., and Timme-Laragy, A. R. (2013). Differential sensitivity to pro- 35 oxidant exposure in two populations of killifish (Fundulus heteroclitus). Ecotoxicology 36 818 37 819 22(2), 387-401, 10.1007/s10646-012-1033-x. 38 820 Hassoun, E. A., and Periandri-Steinberg, S. (2010). Assessment of the roles of antioxidant 39 821 enzymes and glutathione in 3,3',4,4',5-Pentachlorobiphenyl (PCB 126)-induced oxidative 40 822 stress in the tissues of rats after subchronic exposure. Toxicological and 41 823 environmental chemistry 92(2), 301, 10.1080/02772240902846660. 42 824 Incardona, J. P., Day, H. L., Collier, T. K., and Scholz, N. L. (2006). Developmental toxicity of 4- 43 825 ring polycyclic aromatic hydrocarbons in zebrafish is differentially dependent on AH 44 826 receptor isoforms and hepatic cytochrome P4501A metabolism. Toxicology and applied 45 46 827 pharmacology 217(3), 308-21, 10.1016/j.taap.2006.09.018. 47 828 Itoh, K., Chiba, T., Takahashi, S., Ishii, T., Igarashi, K., Katoh, Y., Oyake, T., Hayashi, N., 48 829 Satoh, K., Hatayama, I., Yamamoto, M., and Nabeshima, Y. (1997). An Nrf2/ 49 830 heterodimer mediates the induction of phase II detoxifying enzyme genes through 50 831 antioxidant response elements. Biochemical and biophysical research communications 51 832 236(2), 313-22. 52 833 Jin, Y., Miao, W., Lin, X., Pan, X., Ye, Y., Xu, M., and Fu, Z. (2014). Acute exposure to 3- 53 834 methylcholanthrene induces hepatic oxidative stress via activation of the Nrf2/ARE 54 835 signaling pathway in mice. Environmental toxicology 29(12), 1399-408, 55 56 836 10.1002/.21870. 57 837 Jones, D. P. (2006). Redefining oxidative stress. & redox signaling 8(9-10), 1865- 58 838 79, 10.1089/ars.2006.8.1865. 59 839 Jonsson, M. E., Jenny, M. J., Woodin, B. R., Hahn, M. E., and Stegeman, J. J. (2007). Role of 60 840 AHR2 in the expression of novel cytochrome P450 1 family genes, cell cycle genes, and 61 62 25 63 64 65 1 2 3 4 841 morphological defects in developing zebra fish exposed to 3,3',4,4',5- 5 842 pentachlorobiphenyl or 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicological sciences : an 6 843 official journal of the Society of Toxicology 100(1), 180-93, 10.1093/toxsci/kfm207. 7 8 844 Jonsson, M. E., Kubota, A., Timme-Laragy, A. R., Woodin, B., and Stegeman, J. J. (2012). 9 845 Ahr2-dependence of PCB126 effects on the swim bladder in relation to expression of 10 846 CYP1 and cox-2 genes in developing zebrafish. Toxicology and applied pharmacology 11 847 265(2), 166-74, 10.1016/j.taap.2012.09.023. 12 848 Karchner, S. I., Franks, D. G., and Hahn, M. E. (2005). AHR1B, a new functional aryl 13 849 hydrocarbon receptor in zebrafish: tandem arrangement of ahr1b and ahr2 genes. The 14 850 Biochemical journal 392(Pt 1), 153-61, 10.1042/BJ20050713. 15 851 Kensler, T. W., Wakabayashi, N., and Biswal, S. (2007). Cell survival responses to 16 environmental stresses via the Keap1-Nrf2-ARE pathway. Annual review of 17 852 18 853 pharmacology and toxicology 47, 89-116, 19 854 10.1146/annurev.pharmtox.46.120604.141046. 20 855 Kimmel, C. B., Ballard, W. W., Kimmel, S. R., Ullmann, B., and Schilling, T. F. (1995). Stages of 21 856 embryonic development of the zebrafish. Developmental dynamics : an official 22 857 publication of the American Association of Anatomists 203(3), 253-310, 23 858 10.1002/aja.1002030302. 24 859 King-Heiden, T. C., Mehta, V., Xiong, K. M., Lanham, K. A., Antkiewicz, D. S., Ganser, A., 25 860 Heideman, W., and Peterson, R. E. (2012). Reproductive and developmental toxicity of 26 27 861 dioxin in fish. Molecular and cellular endocrinology 354(1-2), 121-38, 28 862 10.1016/j.mce.2011.09.027. 29 863 Kobayashi, M., Itoh, K., Suzuki, T., Osanai, H., Nishikawa, K., Katoh, Y., Takagi, Y., and 30 864 Yamamoto, M. (2002). Identification of the interactive interface and phylogenic 31 865 conservation of the Nrf2-Keap1 system. Genes to cells : devoted to molecular & cellular 32 866 mechanisms 7(8), 807-20. 33 867 Kwak, M. K., Itoh, K., Yamamoto, M., and Kensler, T. W. (2002). Enhanced expression of the 34 868 transcription factor Nrf2 by cancer chemopreventive agents: role of antioxidant response 35 element-like sequences in the nrf2 promoter. Molecular and cellular biology 22(9), 2883- 36 869 37 870 92. 38 871 Lanham, K. A., Plavicki, J., Peterson, R. E., and Heideman, W. (2014). Cardiac myocyte- 39 872 specific AHR activation phenocopies TCDD-induced toxicity in zebrafish. Toxicological 40 873 sciences : an official journal of the Society of Toxicology 141(1), 141-54, 41 874 10.1093/toxsci/kfu111. 42 875 Li, L., Kobayashi, M., Kaneko, H., Nakajima-Takagi, Y., Nakayama, Y., and Yamamoto, M. 43 876 (2008). Molecular of Keap1. Two Keap1 molecules with distinctive intervening 44 877 region structures are conserved among fish. The Journal of biological chemistry 283(6), 45 46 878 3248-55, 10.1074/jbc.M708702200. 47 879 Liu, H., Nie, F. H., Lin, H. Y., Ma, Y., Ju, X. H., Chen, J. J., and Gooneratne, R. (2014). 48 880 Developmental toxicity, oxidative stress, and related gene expression induced by dioxin- 49 881 like PCB 126 in zebrafish (Danio rerio). Environmental toxicology doi: 50 882 10.1002/tox.22044, 10.1002/tox.22044. 51 883 Lo, R., and Matthews, J. (2013). The aryl hydrocarbon receptor and alpha 52 884 differentially modulate nuclear factor erythroid-2-related factor 2 transactivation in MCF-7 53 885 cells. Toxicology and applied pharmacology 270(2), 139-48, 54 10.1016/j.taap.2013.03.029. 55 886 56 887 Lu, H., Cui, W., and Klaassen, C. D. (2011). Nrf2 protects against 2,3,7,8-tetrachlorodibenzo-p- 57 888 dioxin (TCDD)-induced oxidative injury and steatohepatitis. Toxicology and applied 58 889 pharmacology 256(2), 122-35, 10.1016/j.taap.2011.07.019. 59 890 Ma, Q., Kinneer, K., Bi, Y., Chan, J. Y., and Kan, Y. W. (2004). Induction of murine 60 891 NAD(P)H:quinone oxidoreductase by 2,3,7,8-tetrachlorodibenzo-p-dioxin requires the 61 62 26 63 64 65 1 2 3 4 892 CNC (cap 'n' collar) basic leucine zipper transcription factor Nrf2 (nuclear factor erythroid 5 893 2-related factor 2): cross-interaction between AhR (aryl hydrocarbon receptor) and Nrf2 6 894 signal transduction. The Biochemical journal 377(Pt 1), 205-13, 10.1042/BJ20031123. 7 8 895 Malhotra, D., Portales-Casamar, E., Singh, A., Srivastava, S., Arenillas, D., Happel, C., Shyr, 9 896 C., Wakabayashi, N., Kensler, T. W., Wasserman, W. W., and Biswal, S. (2010). Global 10 897 mapping of binding sites for Nrf2 identifies novel targets in cell survival response through 11 898 ChIP-Seq profiling and network analysis. Nucleic acids research 38(17), 5718-34, 12 899 10.1093/nar/gkq212. 13 900 Miao, W., Hu, L., Scrivens, P. J., and Batist, G. (2005). Transcriptional regulation of NF-E2 p45- 14 901 related factor (NRF2) expression by the aryl hydrocarbon receptor-xenobiotic response 15 902 element signaling pathway: direct cross-talk between phase I and II drug-metabolizing 16 enzymes. The Journal of biological chemistry 280(21), 20340-8, 17 903 18 904 10.1074/jbc.M412081200. 19 905 Mukaigasa, K., Nguyen, L. T., Li, L., Nakajima, H., Yamamoto, M., and Kobayashi, M. (2012). 20 906 Genetic evidence of an evolutionarily conserved role for Nrf2 in the protection against 21 907 oxidative stress. Molecular and cellular biology 32(21), 4455-61, 10.1128/mcb.00481-12. 22 908 Na, Y. R., Seok, S. H., Baek, M. W., Lee, H. Y., Kim, D. J., Park, S. H., Lee, H. K., and Park, J. 23 909 H. (2009). Protective effects of vitamin E against 3,3',4,4',5-pentachlorobiphenyl 24 910 (PCB126) induced toxicity in zebrafish embryos. Ecotoxicology and environmental safety 25 911 72(3), 714-9, 10.1016/j.ecoenv.2008.09.015. 26 27 912 Nakajima, H., Nakajima-Takagi, Y., Tsujita, T., Akiyama, S., Wakasa, T., Mukaigasa, K., 28 913 Kaneko, H., Tamaru, Y., Yamamoto, M., and Kobayashi, M. (2011). Tissue-restricted 29 914 expression of Nrf2 and its target genes in zebrafish with gene-specific variations in the 30 915 induction profiles. PloS one 6(10), e26884, 10.1371/journal.pone.0026884. 31 916 Nebert, D. W., Roe, A. L., Dieter, M. Z., Solis, W. A., Yang, Y., and Dalton, T. P. (2000). Role of 32 917 the aromatic hydrocarbon receptor and [Ah] gene battery in the oxidative stress 33 918 response, cell cycle control, and apoptosis. Biochemical pharmacology 59(1), 65-85. 34 919 Niture, S. K., Khatri, R., and Jaiswal, A. K. (2014). Regulation of Nrf2-an update. Free radical 35 biology & medicine 66, 36-44, 10.1016/j.freeradbiomed.2013.02.008. 36 920 37 921 Noda, S., Harada, N., Hida, A., Fujii-Kuriyama, Y., Motohashi, H., and Yamamoto, M. (2003). 38 922 Gene expression of detoxifying enzymes in AhR and Nrf2 compound null mutant mouse. 39 923 Biochemical and biophysical research communications 303(1), 105-11. 40 924 Nukaya, M., Lin, B. C., Glover, E., Moran, S. M., Kennedy, G. D., and Bradfield, C. A. (2010). 41 925 The aryl hydrocarbon receptor-interacting protein (AIP) is required for dioxin-induced 42 926 hepatotoxicity but not for the induction of the Cyp1a1 and Cyp1a2 genes. The Journal of 43 927 biological chemistry 285(46), 35599-605, 10.1074/jbc.M110.132043. 44 928 Ohtsuji, M., Katsuoka, F., Kobayashi, A., Aburatani, H., Hayes, J. D., and Yamamoto, M. (2008). 45 46 929 Nrf1 and Nrf2 play distinct roles in activation of antioxidant response element-dependent 47 930 genes. The Journal of biological chemistry 283(48), 33554-62, 48 931 10.1074/jbc.M804597200. 49 932 Park, S. H., Jang, J. H., Chen, C. Y., Na, H. K., and Surh, Y. J. (2010). A formulated red 50 933 ginseng extract rescues PC12 cells from PCB-induced oxidative cell death through Nrf2- 51 934 mediated upregulation of heme oxygenase-1 and glutamate cysteine ligase. Toxicology 52 935 278(1), 131-9, 10.1016/j.tox.2010.04.003. 53 936 Prasch, A. L., Teraoka, H., Carney, S. A., Dong, W., Hiraga, T., Stegeman, J. J., Heideman, W., 54 and Peterson, R. E. (2003). Aryl hydrocarbon receptor 2 mediates 2,3,7,8- 55 937 56 938 tetrachlorodibenzo-p-dioxin developmental toxicity in zebrafish. Toxicological sciences : 57 939 an official journal of the Society of Toxicology 76(1), 138-50, 10.1093/toxsci/kfg202. 58 940 Schlezinger, J. J., Keller, J., Verbrugge, L. A., and Stegeman, J. J. (2000). 3,3',4,4'- 59 941 Tetrachlorobiphenyl oxidation in fish, bird and reptile species: relationship to cytochrome 60 61 62 27 63 64 65 1 2 3 4 942 P450 1A inactivation and reactive oxygen production. Comparative biochemistry and 5 943 physiology. Toxicology & pharmacology : CBP 125(3), 273-86. 6 944 Schlezinger, J. J., Struntz, W. D., Goldstone, J. V., and Stegeman, J. J. (2006). Uncoupling of 7 8 945 cytochrome P450 1A and stimulation of reactive oxygen species production by co-planar 9 946 polychlorinated biphenyl congeners. Aquatic toxicology (Amsterdam, Netherlands) 77(4), 10 947 422-32, 10.1016/j.aquatox.2006.01.012. 11 948 Senft, A. P., Dalton, T. P., Nebert, D. W., Genter, M. B., Puga, A., Hutchinson, R. J., Kerzee, J. 12 949 K., Uno, S., and Shertzer, H. G. (2002). Mitochondrial reactive oxygen production is 13 950 dependent on the aromatic hydrocarbon receptor. Free radical biology & medicine 33(9), 14 951 1268-78. 15 952 Shin, S., Wakabayashi, N., Misra, V., Biswal, S., Lee, G. H., Agoston, E. S., Yamamoto, M., and 16 Kensler, T. W. (2007). NRF2 modulates aryl hydrocarbon receptor signaling: influence 17 953 18 954 on adipogenesis. Molecular and cellular biology 27(20), 7188-97, 10.1128/MCB.00915- 19 955 07. 20 956 Skoko, J. J., Wakabayashi, N., Noda, K., Kimura, S., Tobita, K., Shigemura, N., Tsujita, T., 21 957 Yamamoto, M., and Kensler, T. W. (2014). Loss of Nrf2 in mice evokes a congenital 22 958 intrahepatic shunt that alters hepatic oxygen and protein expression gradients and 23 959 toxicity. Toxicological sciences : an official journal of the Society of Toxicology 141(1), 24 960 112-9, 10.1093/toxsci/kfu109. 25 961 Suzuki, T., Takagi, Y., Osanai, H., Li, L., Takeuchi, M., Katoh, Y., Kobayashi, M., and 26 27 962 Yamamoto, M. (2005). Pi class glutathione S-transferase genes are regulated by Nrf 2 28 963 through an evolutionarily conserved regulatory element in zebrafish. The Biochemical 29 964 journal 388(Pt 1), 65-73, 10.1042/bj20041860. 30 965 Tijet, N., Boutros, P. C., Moffat, I. D., Okey, A. B., Tuomisto, J., and Pohjanvirta, R. (2006). Aryl 31 966 hydrocarbon receptor regulates distinct dioxin-dependent and dioxin-independent gene 32 967 batteries. Molecular pharmacology 69(1), 140-53, 10.1124/mol.105.018705. 33 968 Timme-Laragy, A. R., Karchner, S. I., Franks, D. G., Jenny, M. J., Harbeitner, R. C., Goldstone, 34 969 J. V., McArthur, A. G., and Hahn, M. E. (2012). Nrf2b, novel zebrafish paralog of oxidant- 35 responsive transcription factor NF-E2-related factor 2 (NRF2). The Journal of biological 36 970 37 971 chemistry 287(7), 4609-27, 10.1074/jbc.M111.260125. 38 972 Timme-Laragy, A. R., Van Tiem, L. A., Linney, E. A., and Di Giulio, R. T. (2009). Antioxidant 39 973 responses and NRF2 in synergistic developmental toxicity of PAHs in zebrafish. 40 974 Toxicological sciences : an official journal of the Society of Toxicology 109(2), 217-27, 41 975 10.1093/toxsci/kfp038. 42 976 Wakabayashi, N., Slocum, S. L., Skoko, J. J., Shin, S., and Kensler, T. W. (2010). When NRF2 43 977 talks, who's listening? Antioxidants & redox signaling 13(11), 1649-63, 44 978 10.1089/ars.2010.3216. 45 46 979 Wang, L., He, X., Szklarz, G. D., Bi, Y., Rojanasakul, Y., and Ma, Q. (2013). The aryl 47 980 hydrocarbon receptor interacts with nuclear factor erythroid 2-related factor 2 to mediate 48 981 induction of NAD(P)H:quinoneoxidoreductase 1 by 2,3,7,8-tetrachlorodibenzo-p-dioxin. 49 982 Archives of biochemistry and biophysics 537(1), 31-8, 10.1016/j.abb.2013.06.001. 50 983 Wei, Y., Gong, J., Thimmulappa, R. K., Kosmider, B., Biswal, S., and Duh, E. J. (2013). Nrf2 51 984 acts cell-autonomously in endothelium to regulate tip cell formation and vascular 52 985 branching. Proceedings of the National Academy of Sciences of the United States of 53 986 America 110(41), E3910-8, 10.1073/pnas.1309276110. 54 Williams, L. M., Timme-Laragy, A. R., Goldstone, J. V., McArthur, A. G., Stegeman, J. J., 55 987 56 988 Smolowitz, R. M., and Hahn, M. E. (2013). Developmental expression of the Nfe2- 57 989 related factor (Nrf) transcription factor family in the zebrafish, Danio rerio. PloS one 58 990 8(10), e79574, 10.1371/journal.pone.0079574. 59 60 61 62 28 63 64 65 1 2 3 4 991 Wu, K. C., Cui, J. Y., and Klaassen, C. D. (2012). Effect of graded Nrf2 activation on phase-I 5 992 and -II drug metabolizing enzymes and transporters in mouse liver. PloS one 7(7), 6 993 e39006, 10.1371/journal.pone.0039006. 7 8 994 Yeager, R. L., Reisman, S. A., Aleksunes, L. M., and Klaassen, C. D. (2009). Introducing the 9 995 "TCDD-inducible AhR-Nrf2 gene battery". Toxicological sciences : an official journal of 10 996 the Society of Toxicology 111(2), 238-46, 10.1093/toxsci/kfp115. 11 997 Yoshioka, W., Peterson, R. E., and Tohyama, C. (2011). Molecular targets that link dioxin 12 998 exposure to toxicity phenotypes. The Journal of steroid biochemistry and molecular 13 999 biology 127(1-2), 96-101, 10.1016/j.jsbmb.2010.12.005. 14 1000 15 1001 Footnotes 16 17 1002 Nomenclature: nrf is a commonly used notation for NF-E2-related factor genes, which are 18 1003 officially designated as nfe2l (NF-E2-like). For example, nrf2a is officially designated as nfe2l2a. 19 20 1004 Throughout the paper, we use the more common nrf designation. Otherwise, we utilize the 21 22 1005 approved format for designating genes and proteins (see the ZFIN Zebrafish Nomenclature 23 1006 Web site). Human genes and proteins are designated using all capitals (NRF2 and NRF2, 24 25 1007 respectively), rodent genes and proteins as Nrf2 and NRF2 respectively, whereas zebrafish 26 27 1008 genes are designated nrf2 and Nrf2 for genes and proteins, respectively. When not referring to 28 1009 a specific species, we have used the zebrafish notation as a default format. 29 30 31 1010 32 33 34 1011 Figure legends 35 1012 Fig. 1. Activation of ARE-driven luciferase reporter by Nrf2a mutant alleles in vitro. COS-7 cells 36 37 1013 were transfected with plasmids encoding for full-length sequences of a wildtype Nrf2a, the 38 fh318 fh319 39 1014 Nrf2a and Nrf2a mutant alleles, and plasmids for Keap1a and Keap1b. A) Western blot 40 1015 of COS-7 cell lysates showing expression of Nrf2a proteins from the respective plasmids, 41 42 1016 transfected at 50 or 500 ng plasmid per well. Location of MW standards is shown. The predicted 43 44 1017 size of the Nrf2a protein is ~66 kDa. B) Results of transient transfection assays to evaluate the 45 1018 function of the Nrf2a proteins including their ability to be repressed by Keap1 and induced by 46 47 1019 the model Nrf2 activator tBHQ. Results show that the Nrf2afh318 allele is capable of activating the 48 49 1020 ARE reporter at a greatly diminished capacity from the wildtype allele, but activity is above 50 background levels. The Nrf2afh319 allele has slightly less ARE activity levels than the wildtype, 51 1021 52 1022 but is more inducible by 10 µM tBHQ. All Nrf2a proteins were repressed by Keap1a and 53 54 1023 Keap1b, except for repression of the 318-2 plasmid by Keap1b, which was not statistically 55 significant. Two-factor ANOVA for plasmid and tBHQ exposure p < 0.001 followed by Fisher’s 56 1024 57 1025 PLSD. # indicates a significant difference between DMSO and tBHQ; “a” indicates a significant 58 59 1026 difference from the ARE plasmid control; “b” indicates a significant difference of the 318 or 319 60 61 62 29 63 64 65 1 2 3 4 1027 plasmids from the Nrf2a wildtype plasmid activity, and “c” indicates a significant repression by 5 6 1028 Keap1 of Nrf2a-luciferase activation (p < 0.05). N = 3 replicates per group 7 8 fh318 1029 Fig. 2. nrf2a mutant eleutheroembryos are more sensitive to PCB126 embryotoxicity and 9 10 1030 have subtle developmental abnormalities. Homozygous wildtype (WT/WT) and nrf2a mutant 11 12 1031 (M/M) embryos were exposed to 0, 2, or 5 nM PCB126 (with DMSO, 0.01%) at 24 hpf, and 13 assessed for morphology at 4 days post fertilization. A) Representative eleutheroembryos from 14 1032 15 1033 each genotype and treatment group. Quantified measures include B) ventral-dorsal length, C) 16 17 1034 swim bladder inflation, D) pericardial area, and E) severity of the craniofacial malformations. 18 Data are from two independent experiments, which each contained 3 biological replicate pools 19 1035 20 1036 of 5-10 embryos per dose. Statistical significance from a two-factor ANOVA and Fisher’s PLSD 21 22 1037 are designated by “#” for differences between genotypes, and “*” for PCB126 effects within a 23 24 1038 genotype; p < 0.05. 25 26 1039 Fig. 3. nrf2afh318 mutant eleutheroembryos treated with PCB126 at 5 dpf. Homozygous wildtype 27 28 1040 (WT/WT) and nrf2a mutant (M/M) embryos were exposed to 0, 2, or 5 nM PCB126 (with DMSO, 29 1041 0.01%) at 24 hpf, and assessed for morphology at 5 days post fertilization. Quantified measures 30 31 1042 include A) ventral-dorsal distance, B) swim bladder inflation, C) pericardial area, and D) severity 32 33 1043 of the craniofacial malformations. Data are from 3-4 biological replicate pools of 5-10 embryos 34 1044 per dose. Statistical significance from a two-factor ANOVA and Fisher’s PLSD are designated 35 36 1045 by “#” for differences between genotypes, and “*” for PCB126 effects within a genotype; p < 37 38 1046 0.05. 39 40 1047 Fig. 4. nrf2afh318 mutant embryos have larger yolk sacs at 3 hpf than wildtype embryos. At 3 hpf, 41 42 1048 19-20 embryos per genotype were imaged and the yolk area was quantified. Error bar = 1000 43 44 1049 µm. A F-test and T-test were performed in Excel using a 2-tailed distribution, and the genotypes 45 1050 are statistically different (p < 0.05). 46 47 48 1051 Fig. 5. Quantitative real-time PCR measures for ahr-related genes in nrf2a wildtype and fh318 49 50 1052 mutant eleutheroembryos at 4 dpf. Genes measured include A) cyp1a, B) ahr2, C) ahr1a, D) 51 1053 ahr1b, E) ahrra, and F) ahrrb. Data were analyzed by two-factor ANOVA for PCB effect, 52 53 1054 genotype effect, and interactions, followed by a one-factor ANOVA and Fisher’s PLSD (p < 54 55 1055 0.05). Differences significant for PCB treatment within genotypes are noted with a “*” and 56 1056 differences between genotypes exposed to the same treatment are noted with a “#”. N = 3-4 57 58 1057 biological replicates per group, with each replicate containing a pool of 5 fish. 59 60 61 62 30 63 64 65 1 2 3 4 1058 Fig. 6. Quantitative real-time PCR measures for nrf-related genes in nrf2a wildtype and fh318 5 6 1059 mutant eleutheroembryos at 4 dpf. Genes measured include A) nrf1a, B) nrf1b, C) nrf2a, D) 7 8 1060 nrf3, E) keap1a, and F) keap1b. Data were analyzed by two-factor ANOVA for PCB effect, 9 1061 genotype effect, and interactions, followed by a one-factor ANOVA and Fisher’s PLSD (p < 10 11 1062 0.05). Differences significant for PCB treatment within genotypes are noted with a “*” and 12 13 1063 differences between genotypes exposed to the same treatment are noted with a “#”. N = 3-4 14 1064 biological replicates per group, with each replicate containing a pool of 5 fish. 15 16 17 1065 Fig. 7. Quantitative real-time PCR measures for nrf2b and regulatory targets in nrf2a wildtype 18 and fh318 mutant eleutheroembryos at 4 dpf. Genes measured include A) nrf2b, B) hmox1, and 19 1066 20 1067 C) . Data were analyzed by two-factor ANOVA for PCB effect, genotype effect, and 21 22 1068 interactions, followed by a one-factor ANOVA and Fisher’s PLSD (p < 0.05). Differences 23 24 1069 significant for PCB treatment within genotypes are noted with a “*” and differences between 25 1070 genotypes exposed to the same treatment are noted with a “#”. N = 3-4 biological replicates per 26 27 1071 group, with each replicate containing a pool of 5 fish. 28 29 1072 Fig. 8. Quantitative real-time PCR measures for Phase II genes in nrf2a wildtype and fh318 30 31 1073 mutant eleutheroembryos at 4 dpf. Genes measured include A) nqo1 and B) gsta1. Data were 32 33 1074 analyzed by two-factor ANOVA for PCB effect, genotype effect, and interactions, followed by a 34 1075 one-factor ANOVA and Fisher’s PLSD (p < 0.05). Differences significant for PCB treatment 35 36 1076 within genotypes are noted with a “*” and differences between genotypes exposed to the same 37 38 1077 treatment are noted with a “#”. N = 3-4 biological replicates per group, with each replicate 39 containing a pool of 5 fish. 40 1078 41 42 1079 43 44 45 1080 Tables 46 1081 Table 1. Identification of putative Antioxidant Response Elements (AREs) in the region starting 47 48 1082 10 kb upstream from the transcription start site through the end of the 2nd exon. The zebrafish- 49 50 1083 specific (TGA(G/C)nnnTC) and mammalian (TGA(G/C)nnnGC) ARE sequences were manually 51 1084 identified in each genomic sequence. 52 53 Gene Accession # # Zebrafish ARE Locations # Mammalian ARE Locations 54 name Zebrafish [TGA(G/C)nnnTC] Mammalia [TGA(G/C)nnnGC] 55 AREs n AREs 56 57 ahr1a NM_131028 4 -9660, -9476, -7930, - 5 -7774, -7691, -6079, -5434, -4814 58 2323 59 60 61 62 31 63 64 65 1 2 3 4 5 ahr1b NM_001024 4 -7512, -3509, -2455, -194 3 -9819, -9262, -5418 816 6 7 ahr2 NM_131264 9 -8506, -8214, -6746, - 10 -9909, -7982, -5255, -3903, -261, 8 2670, -74, +8889, +7989, +8822, +11120, +12115, 9 +17669, +18540, +19184 +15528 10 11 ahrra NM_001035 3 -1864, -1383, -21 4 -6279, -5659, -5036, -4907 12 265 13 14 ahrrb NM_001033 5 -8109, -8021, -3768, - 1 -6395 15 920 1192, -161 16 cyp1a NM_131879 4 -8567, -5393, +3202, 4 -9968, -2472, -1965, -331 17 +3440 18 19 gclc NM_199277 6 -3115, +4635, +6218, 11 -9742, -6187, -5525, -4047, - 20 +6771, +7235, +7358 3296, +178, +1741, +3460, 21 +3772, +6043, +7223 22 23 gsta1 NM_213394 9 -9661, -5460, +3443, 10 -9387, -1317, +2080, +6612, 24 +4414, +4488, +4851, +7289, +9816, +10774, +11968, 25 +9990, +11900, +15085 +12652, +12944 26 27 gstp1 NM_131734 4 -5560, -4371, -4061, -33 2 -5187, -1043 28 hmox1 NM_001127 8 -7983, -7974, -4294, - 2 +18, +272 29 516 2280, -2060, -1087, +29, 30 +433 31 32 hsp70 NM_131397 7 -8921, -5727, -5518, - 5 -7418, -6640, -33, +1441, +1871 33 5466, -5133, -647, +218 34 35 keap1 NM_182864 3 -4726, -1743, +6666 8 -817, +1389, +1627, +2425, 36 a +3807, +4089, +7108, +7669 37 38 keap1 NM_001113 6 -9831, -6073, -5456, - 3 -9798, -8904, -7486 b 477 3391, +536, +1234 39 40 NM_212855 7 -7436, -6034, -5484, - 3 -6657, +1805, +9766 41 a 1461, +3531, +8657, 42 +9025 43 44 nfe2l1 NM_001278 10 -8942, -7257, -6739, - 4 -3736, -2234, +1803, +2683 45 b 842 5758, -5670, -1929, -748, 46 -31, +1859, +3480 47 48 NM_182889 4 -5710, -5431, -5355, 3 -2812, +1685, +13484 49 a +3418 50 nfe2l2 NM_001257 5 -9572, -1293, -1116, 5 -6837, -4311, +2366, +4102, 51 b 183 +2636, +3853 +5007 52 53 nfe2l3 NM_213231 2 -5360, -3675 6 -9497, -9121, -7259, -6205, +857, 54 +1437 55 56 nqo1 NM_001204 0 3 -9195, -5399, +3781 57 272 58 1085 59 60 61 62 32 63 64 65 1 2 3 4 1086 Table 2. Comparison of changes in gene expression between Nrf2a wildtype (wt) and 5 6 1087 Nrf2afh318/fh318 (m) embryos at 96 hpf. Expression levels that were not different between wt 7 8 1088 and m are indicated with “=” while < and > indicate that the wt fish had expression either 9 1089 significantly less than or greater than the mutants. 10 11 Basal PCB126 12 13 Gene names expression inducible 14 15 ahr2 wt = m wt = m 16 17 cyp1a wt = m wt < m 18 19 ahrra, ahrrb, nrf2a wt = m wt > m 20 21 ahr1a, hmox1, wt = m wt 22 keap1a, keap1b, 23 nrf1a, nrf2b, nrf3 24 25 ahr1b wt > m wt 26 27 nrf2b wt = m m 28 29 gclc, gstp1, hsp70, wt = m - 30 nfe2, tp53 31 32 gsta1, nqo1 wt < m - 33 34 35 1090 36 37 38 1091 39 40 41 1092 42 43 44 45 1093 46 47 48 1094 49 50 51 1095 52 53 54 1096 55 56 57 58 1097 Figures 59 60 1098 61 62 33 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 1099 40 41 42 1100 Fig. 1. Activation of ARE-driven luciferase reporter by Nrf2a mutant alleles in vitro. COS-7 cells 43 1101 were transfected with plasmids encoding for full-length sequences of a wildtype Nrf2a, the 44 45 1102 Nrf2afh318 and Nrf2afh319 mutant alleles, and plasmids for Keap1a and Keap1b. A) Western blot 46 47 1103 of COS-7 cell lysates showing expression of Nrf2a proteins from the respective plasmids, 48 1104 transfected at 50 or 500 ng plasmid per well. Location of MW standards is shown. The predicted 49 50 1105 size of the Nrf2a protein is ~66 kDa. B) Results of transient transfection assays to evaluate the 51 52 1106 function of the Nrf2a proteins including their ability to be repressed by Keap1 and induced by 53 1107 the model Nrf2 activator tBHQ. Results show that the Nrf2afh318 allele is capable of activating the 54 55 1108 ARE reporter at a greatly diminished capacity from the wildtype allele, but activity is above 56 fh319 57 1109 background levels. The Nrf2a allele has slightly less ARE activity levels than the wildtype, 58 1110 but is more inducible by 10 µM tBHQ. All Nrf2a proteins were repressed by Keap1a and 59 60 1111 Keap1b, except for repression of the 318-2 plasmid by Keap1b, which was not statistically 61 62 34 63 64 65 1 2 3 4 1112 significant. Two-factor ANOVA for plasmid and tBHQ exposure p < 0.001 followed by Fisher’s 5 6 1113 PLSD. # indicates a significant difference between DMSO and tBHQ; “a” indicates a significant 7 8 1114 difference from the ARE plasmid control; “b” indicates a significant difference of the 318 or 319 9 1115 plasmids from the Nrf2a wildtype plasmid activity, and “c” indicates a significant repression by 10 11 1116 Keap1 of Nrf2a-luciferase activation (p < 0.05). N = 3 replicates per group. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 35 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 1117 42 43 Fig. 2. nrf2afh318 mutant eleutheroembryos are more sensitive to PCB126 embryotoxicity and 44 1118 45 1119 have subtle developmental abnormalities. Homozygous wildtype (WT/WT) and nrf2a mutant 46 47 1120 (M/M) embryos were exposed to 0, 2, or 5 nM PCB126 at 24 hpf, and assessed for morphology 48 at 4 days post fertilization. A) Representative eleutheroembryos from each genotype and 49 1121 50 1122 treatment group. Quantified measures include B) ventral-dorsal length, C) swim bladder 51 52 1123 inflation, D) pericardial area, and E) severity of the craniofacial malformations. Data are from 53 54 1124 two independent experiments, which each contained 3 biological replicate pools of 5-10 55 1125 embryos per dose. Statistical significance from a two-factor ANOVA and Fisher’s PLSD are 56 57 1126 designated by “#” for differences between genotypes, and “*” for PCB126 effects within a 58 59 1127 genotype; p < 0.05. 60 61 62 36 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 1128 29

30 fh318 31 1129 Fig. 3. nrf2a mutant eleutheroembryos treated with PCB126 at 5 dpf. Homozygous wildtype 32 1130 (WT/WT) and nrf2a mutant (M/M) embryos were exposed to 0, 2, or 5 nM PCB126 at 24 hpf, 33 34 1131 and assessed for morphology at 5 days post fertilization. Quantified measures include A) 35 36 1132 ventral-dorsal distance, B) swim bladder inflation, C) pericardial area, and D) severity of the 37 1133 craniofacial malformations. Data are from 3-4 biological replicate pools of 5-10 embryos per 38 39 1134 dose. Statistical significance from a two-factor ANOVA and Fisher’s PLSD are designated by “#” 40 41 1135 for differences between genotypes, and “*” for PCB126 effects within a genotype; p < 0.05. 42 43 1136 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 37 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 1137 31 32 Fig. 4. nrf2a mutant embryos have larger yolk sacs at 3 hpf than wildtype embryos. At 3 hpf, 33 1138 34 1139 19-20 embryos per genotype were imaged and the yolk area was quantified. A) Representative 35 36 1140 embryos from the wildtype and fh318 mutants; error bar = 1000 µm. B) Quantification of the yolk 37 sacs. A F-test and T-test were performed in Excel using a 2-tailed distribution, and the 38 1141 39 1142 genotypes are statistically different (p < 0.05). 40 41 42 1143 43 44 45 1144 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 38 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 1145 35 36 37 1146 Fig. 5. Quantitative real-time PCR measures for ahr-related genes in nrf2a wildtype and mutant 38 1147 eleutheroembryos at 4 dpf. Genes measured include A) cyp1a, B) ahr2, C) ahr1a, D) ahr1b, E) 39 40 1148 ahrra, and F) ahrrb. Data were analyzed by two-factor ANOVA for PCB effect, genotype effect, 41 42 1149 and interactions, followed by a one-factor ANOVA and Fisher’s PLSD (p < 0.05). Differences 43 1150 significant for PCB treatment within genotypes are noted with a “*” and differences between 44 45 1151 genotypes exposed to the same treatment are noted with a “#”. N = 3-4 biological replicates per 46 47 1152 group, with each replicate containing a pool of 5 fish. 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 39 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 1153 35 36 37 1154 Fig. 6. Quantitative real-time PCR measures for nrf-related genes in nrf2a wildtype and mutant 38 39 1155 eleutheroembryos at 4 dpf. Genes measured include A) nrf1a, B) nrf1b, C) nrf2a, D) nrf3, E) 40 41 1156 keap1a, and F) keap1b. Data were analyzed by two-factor ANOVA for PCB effect, genotype 42 1157 effect, and interactions, followed by a one-factor ANOVA and Fisher’s PLSD (p < 0.05). 43 44 1158 Differences significant for PCB treatment within genotypes are noted with a “*” and differences 45 46 1159 between genotypes exposed to the same treatment are noted with a “#”. N = 3-4 biological 47 1160 replicates per group, with each replicate containing a pool of 5 fish. 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 40 63 64 65 1 2 3 4 1161 5 6 7 8 9 10 11 12 13 14 15 16 17 1162 18 19 20 1163 Fig. 7. Quantitative real-time PCR measures for nrf2b and regulatory targets in nrf2a wildtype 21 22 1164 and mutant eleutheroembryos at 4 dpf. Genes measured include A) nrf2b, B) hmox1, and C) 23 1165 p53. Data were analyzed by two-factor ANOVA for PCB effect, genotype effect, and 24 25 1166 interactions, followed by a one-factor ANOVA and Fisher’s PLSD (p < 0.05). Differences 26 27 1167 significant for PCB treatment within genotypes are noted with a “*” and differences between 28 1168 genotypes exposed to the same treatment are noted with a “#”. N = 3-4 biological replicates per 29 30 1169 group, with each replicate containing a pool of 5 fish. 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 41 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1170 16 17 18 1171 Fig. 8. Quantitative real-time PCR measures for Phase II genes in nrf2a wildtype and mutant 19 20 1172 eleutheroembryos at 4 dpf. Genes measured include A) nqo1 and B) gsta1. Data were analyzed 21 1173 by two-factor ANOVA for PCB effect, genotype effect, and interactions, followed by a one-factor 22 23 1174 ANOVA and Fisher’s PLSD (p < 0.05). Differences significant for PCB treatment within 24 25 1175 genotypes are noted with a “*” and differences between genotypes exposed to the same 26 1176 treatment are noted with a “#”. N = 3-4 biological replicates per group, with each replicate 27 28 1177 containing a pool of 5 fish. 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 42 63 64 65 1 2 3 4 1178 Supplemental Table 1. List of qPCR primers, melting temperature, and sources. 5 6 7 8 Gene Forward primer (5’-3’) Tm Reference 9 Reverse primer (5’-3’) 10 11 ahr1a TTGTTCTTGGCTACACAGAGGCAGAGC 66 Karchner et al., 2005 12 CCCAGTTTCACCGGTCCTCATCATCC 13 14 ahr1b GGTTTGTCGTCAAACAACAGTAACCACG 65 Karchner et al., 2005 15 CCACCAACACAAAGCCATTAAGAGCCTG 16 17 ahr2 CTACTTGGGCTTCCATCAGTCG 65 Evans et al., 2005 18 GTCACTTGAGGGATTGAGAGCG 19 20 ahrra GCGCATCAAGAGCTTCTGCAGCGTGTT 65 Evans et al., 2005 21 CCACTGACGACCAGCGCAAACCCT 22 23 ahrrb AAACAGAAGCTCTGGCCGCAGTGA 65 Evans et al., 2005 24 CTGTATGCCCATGAAGCGTCCTGAG 25 26 b-actin CAACAGAGAGAAGATGACACAGATCA 65 Evans et al., 2005 27 GTCACACCATCACCAGAGTCCATCAC 28 29 cyp1a GCATTACGATACGTTCGATAAGGAC 65 Evans et al., 2005 30 GCTCCGAATAGGTCATTGACGAT 31 32 ef1a CTTCTCAGGCTGACTGTGC 60 McCurley & Callard, 2008 33 CCGCTAGGATTACCCTCC 34 35 gclc AACCGACACCCAAGATTCAGCACT 60 Timme-Laragy et al., 2009 36 CCATCATCCTCTGGAAACACCTCC 37 38 gsta1 TCACACCTGCCGAAAACAAAG 65 Richter et al., 2011 39 CCACGAGGAAAGAAGAGTTTGC 40 41 gstp1 CGACTTGAAAGCCACCTGTGTC 67 Timme-Laragy et al., 2012 42 CTGTCGTTTTTGCCATATGCAGC 43 hmox1 et al 44 GCTTCTGCTGTGCTCTCTATACG 62 Timme-Laragy ., 2012 45 CCAATCTCTCTCAGTCTCTGTGC 46 hsp70 CAAAGGCAAATCCTCAGAGC 64 Timme-Laragy et al., 2012 47 48 CACAAAGTGGTTCACCATGC 49 keap1a CCGGTTGTCCCCGTCAAACC 65 - 50 CCGGTTGTCCCCGTCAAACC 51 52 keap1b CCAGATCGACAGCGTGGTTC 65 - 53 CCTGGCTGAAGTTCATGTAG 54 55 nfe2 AATTTGGAGATGGACTTGGCCTGG 60 Timme-Laragy et al., 2012 56 GCATCACAAGTGGCTGGAATGGAT 57 58 nqo1 TTCAGTACCCACTCTACTGG 63 Hahn et al., 2014 59 GCATGGCCCTCTTATTCTTG 60 61 62 43 63 64 65 1 2 3 4 5 nrf1a CCAGAGTTGACAGGTCCTGG 64 Timme-Laragy et al., 2012 6 CATAACCTGTGATTCCATGATAGAC 7 8 nrf1b GCAGGACATGGAGGTGAACAATACG 66 Timme-Laragy et al., 2012 9 GGATCGTGGGAGCCCAAAATTTCC 10 11 nrf2a GAGCGGGAGAAATCACACAGAATG 65 Timme-Laragy et al., 2012 12 CAGGAGCTGCATGCACTCATCG 13 14 nrf2b GGCAGAGGGAGGAGGAGACCAT 68 Timme-Laragy et al., 2012 15 AAACAGCAGGGCAGACAACAAGG 16 17 nrf3 GCATGAGGATTTAGTGGTTAGTGG 64 Timme-Laragy et al., 2012 18 GGAGTCAAAATCATCAAAGTCAG 19 tp53 et al., 20 CCCGGATGGAGATAACTTG 60 Duan 2011 21 CACAGTTGTCCATTCAGCAC 22 1179 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 44 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Supplemental Figure 1. Protein alignment of the translated nrf2a fh318 allele with other Nrf2a 43 1180 44 1181 protein sequences. 45 46 1182 47 48 1183 Source of sequences: drNrf2a (consensus sequence); “tilling Nrf2a” 49 1184 (http://research.fhcrc.org/moens/en/mutants.html); BAC10573.1 (Kobayashi et al 2002); 50 51 1185 NP_878309.1 (GenBank reference sequence, derived from BAC10573.1); AAH45852.1 52 53 1186 (GenBank direct submission); AAI52660.1 (GenBank direct submission); Nrf2a clone 318-M2 54 1187 (our sequence of mutant fh318). 55 56 57 58 59 60 61 62 45 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 1188 27 28 29 1189 Supplemental Fig. 2. . Quantitative real-time PCR measures for oxidative stress response 30 1190 genes in nrf2a wildtype and fh318 mutant eleutheroembryos at 4 dpf. Genes measured include 31 32 1191 A) hsp70, B) gclc, and C) gstp1. Data were analyzed by two-factor ANOVA for PCB effect, 33 34 1192 genotype effect, and interactions, followed by a one-factor ANOVA and Fisher’s PLSD (p < 35 0.05). N = 3-4 biological replicates per group, with each replicate containing a pool of 5 fish. 36 1193 37 38 1194 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 46 63 64 65 1 2 3 4 1195 Supplemental References: 5 6 1196 Duan, J., Ba, Q., Wang, Z., Hao, M., Li, X., Hu, P., Zhang, D., Zhang, R., and Wang, H. (2011). 7 1197 Knockdown of ribosomal protein S7 causes developmental abnormalities via p53 dependent 8 1198 and independent pathways in zebrafish. Int J Biochem Cell Biol 43(8), 1218-27, 9 10 1199 10.1016/j.biocel.2011.04.015. 11 12 1200 Evans, B. R., Karchner, S. I., Franks, D. G., and Hahn, M. E. (2005). Duplicate aryl hydrocarbon 13 1201 receptor repressor genes (ahrr1 and ahrr2) in the zebrafish Danio rerio: structure, function, 14 1202 evolution, and AHR-dependent regulation in vivo. Archives of biochemistry and biophysics 15 1203 441(2), 151-67, 10.1016/j.abb.2005.07.008. 16 17 1204 Hahn, M. E., McArthur, A. G., Karchner, S. I., Franks, D. G., Jenny, M. J., Timme-Laragy, A. R., 18 1205 Stegeman, J. J., Woodin, B. R., Cipriano, M. J., and Linney, E. (2014). The transcriptional 19 1206 response to oxidative stress during vertebrate development: effects of tert-butylhydroquinone 20 1207 and 2,3,7,8-tetrachlorodibenzo-p-dioxin. PloS one 9(11), e113158, 21 1208 10.1371/journal.pone.0113158. 22 23 1209 Karchner, S. I., Franks, D. G., and Hahn, M. E. (2005). AHR1B, a new functional aryl 24 hydrocarbon receptor in zebrafish: tandem arrangement of ahr1b and ahr2 genes. The 25 1210 26 1211 Biochemical journal 392(Pt 1), 153-61, 10.1042/bj20050713. 27 28 1212 Kobayashi, M., Itoh, K., Suzuki, T., Osanai, H., Nishikawa, K., Katoh, Y., Takagi, Y., and 29 1213 Yamamoto, M. (2002). Identification of the interactive interface and phylogenic conservation of 30 1214 the Nrf2-Keap1 system. Genes to cells : devoted to molecular & cellular mechanisms 7(8), 807- 31 1215 20. 32 33 1216 McCurley, A. T., and Callard, G. V. (2008). Characterization of housekeeping genes in 34 1217 zebrafish: male-female differences and effects of tissue type, developmental stage and 35 1218 chemical treatment. BMC Mol Biol 9, 102, 10.1186/1471-2199-9-102. 36 37 1219 Richter, C. A., Garcia-Reyero, N., Martyniuk, C., Knoebl, I., Pope, M., Wright-Osment, M. K., 38 Denslow, N. D., and Tillitt, D. E. (2011). Gene expression changes in female zebrafish (Danio 39 1220 40 1221 rerio) brain in response to acute exposure to methylmercury. Environmental toxicology and 41 1222 chemistry / SETAC 30(2), 301-8, 10.1002/etc.409. 42 43 1223 Timme-Laragy, A. R., Karchner, S. I., Franks, D. G., Jenny, M. J., Harbeitner, R. C., Goldstone, 44 1224 J. V., McArthur, A. G., and Hahn, M. E. (2012). Nrf2b, novel zebrafish paralog of oxidant- 45 1225 responsive transcription factor NF-E2-related factor 2 (NRF2). The Journal of biological 46 1226 chemistry 287(7), 4609-27, 10.1074/jbc.M111.260125. 47 48 1227 Timme-Laragy, A. R., Van Tiem, L. A., Linney, E. A., and Di Giulio, R. T. (2009). Antioxidant 49 1228 responses and NRF2 in synergistic developmental toxicity of PAHs in zebrafish. Toxicological 50 1229 sciences : an official journal of the Society of Toxicology 109(2), 217-27, 10.1093/toxsci/kfp038. 51 52 53 54 55 56 57 58 59 60 61 62 47 63 64 65