Toxicological Sciences 68, 403-419 (2002)
Copyright © 2002 by the Society of Toxicology
ENVIRONMENTAL TOXICOLOGY |
Tissue-Specific Expression of AHR2, ARNT2, and CYP1A in Zebrafish Embryos and Larvae: Effects of Developmental Stage and 2,3,7,8-Tetrachlorodibenzo-p-dioxin Exposure
* Environmental Toxicology Center, University of Wisconsin, Madison, Wisconsin 53705;
Department of Environmental and Molecular Toxicology and Marine/Freshwater Biomedical Sciences Center, Oregon State University, Corvallis, Oregon 97333;
Department of Pharmaceutical Sciences, School of Pharmacy, University of Colorado Health Sciences Center, Denver, Colorado 80262;
Biology Department, Woods Hole Oceanographic Institute, Woods Hole, Massachusetts 02543; and
¶ School of Pharmacy, University of Wisconsin, Madison, Wisconsin 53705
Received February 2, 2002; accepted April 10, 2002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To better understand the role of the aryl hydrocarbon receptor (AHR) signaling pathway in causing tissue-specific signs of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) toxicity in zebrafish, the temporal and spatial expression of the zebrafish aryl hydrocarbon receptor 2 (zfAHR2), aryl hydrocarbon receptor nuclear translocator 2 (zfARNT2), and an AHR regulated gene, cytochrome P4501A (zfCYP1A), were assessed in larvae exposed to vehicle or TCDD (1.55 nM) from 3–4 h postfertilization (hpf). Coexpression of a transcriptionally active AHR pathway was apparent by the expression of zfCYP1A mRNA and protein in certain larval tissues. zfCYP1A protein was first detected in the skin and vasculature of TCDD-exposed larvae at 36 hpf. Vascular-specific zfCYP1A protein expression continued from 36 to120 hpf at which time it was also detected in the heart, kidney, and liver. zfCYP1A mRNA was observed in TCDD treated larvae as early as 24 hpf in the developing vascular system. Vascular specific zfCYP1A mRNA expression in the head, trunk, and tail by 36 hpf in TCDD-exposed larvae, confirmed immunohistochemical localization. The expression of zfAHR2 and zfARNT2 mRNAs was generally similar in control and TCDD-exposed larvae. Coexpression of zfAHR2, zfARNT2, and zfCYP1A mRNAs was evident in TCDD-exposed larvae by 36 hpf and in the vasculature, heart, and trunk kidney by 48 hpf, well before the first signs of overt developmental toxicity are observed. In addition to their function in response to AHR agonists, zfAHR2 and zfARNT2 may be involved in development and function of the nervous system. zfAHR2 and zfARNT2 were detected in the brain, spinal cord, and sensory organs. However, TCDD-induced zfCYP1A expression was not detected in these tissues. Taken together, these results are consistent with the notion that the cardiovascular system is a primary target of TCDD developmental toxicity in zebrafish.
Key Words: AHR; ARNT; CYP1A; TCDD; zebrafish; embryo; larva; development; expression; toxicity.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Larval fish exposed to sufficiently high tissue concentrations of aryl hydrocarbon receptor (AHR) agonists such as the polychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), and biphenyls (PCBs) exhibit similar signs of developmental toxicity. These signs include yolk sac, pericardial and meningeal edema, decreased cardiac output, reduced blood flow to most peripheral vascular beds, anemia, multifocal hemorrhage, impaired swimbladder inflation, craniofacial malformations typified by impaired lower jaw development, stunted growth, and mortality (Belair et al., 2001
; Elonen et al., 1998; Henry et al., 1997; Johnson et al., 1998; Spitsbergen et al., 1990; Teraoka et al., 2002; Walker and Peterson, 1991; Walker et al., 1991; Zabel et al., 1995). Exposure of the most sensitive fish species to such AHR agonists is of ecological concern because of their potential to reduce feral fish populations secondary to mortality at the larval stage of development (Walker and Peterson, 1994; Zabel et al., 1995).
Effort to understand the molecular mechanism of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) developmental toxicity has led to the characterization of the AHR pathway in mammals and other model organisms including zebrafish. In mammals, much of the toxicity of TCDD-like AHR ligands is dependent upon their binding to the cytosolic AHR (Rowlands and Gustafsson, 1997; Schmidt and Bradfield, 1996). Ligand-bound and nuclear-localized AHR dimerizes with the aryl hydrocarbon nuclear translocator (ARNT) and this heterodimeric complex associates with specific DNA sequences termed dioxin response elements (DREs), altering expression of downstream genes such as cytochrome P4501A1 (CYP1A1; Sutter and Greenlee, 1992; Sutter et al., 1991). Both AHR expression and ligand activation of the AHR appear to be essential for the manifestation of much of TCDD developmental toxicity in mammals, since AHR null mice exhibit few signs of TCDD toxicity (Fernandez-Salguero et al., 1996; Mimura et al., 1997; Peters et al., 1999). The zebrafish AHR signal transduction pathway is generally similar to that of mammals with the notable exception that fish possess at least two AHR genes while mammals have only one (Hahn et al., 1997).
Two distinct classes of AHR genes (AHR1 and 2) have been identified in fish (Hahn et al., 1997). Phylogenetic comparisons suggest that the 2 forms of fish AHR arose from a gene duplication event during fish evolution (Hahn et al., 1997). Sequence analysis reveals that zfAHR1 shares the greatest sequence similarity with the mammalian AHRs (Andreasen et al., 2002). The nomenclature of the fish AHRs and ARNTs used here follows the evolutionary conventions suggested in recent reports (Hahn et al., 1997). An ARNT1 has not been identified in fish, however, rtARNTb identified in rainbow trout shares sequence homology with both ARNT and ARNT2 (Pollenz et al., 1996; Powell and Hahn, 2000). Full-length zebrafish AHR2 (zfAHR2) and ARNT2 (zfARNT2) cDNAs have been cloned and their translation products characterized (Tanguay et al., 1999, 2000). Four splice variants of zfARNT2 (zfARNT2a, zfARNT2b, zfARNT2c, and zfARNT2x) have been identified. Only zfARNT2b has been demonstrated to be transcriptionally active with zfAHR2. zfAHR2 and zfARNT2b form a functional heterodimer in vitro. The zfAHR2/zfARNT2b complex specifically recognizes DREs in gel shift experiments and induces DRE driven transcription in response to TCDD exposure in COS-7 cells (Tanguay et al., 1999, 2000). A full-length zfAHR1 corresponding to a partial fragment previously described by Wang and coworkers (Wang et al., 1998) also has been cloned and characterized (Andreasen et al., 2002).
The similar signs of developmental TCDD toxicity observed across fish species suggest there is a common mechanism of toxicity among fish (Walker and Peterson, 1994). In larval fish this mechanism is hypothesized to involve AHR, since the earliest signs of TCDD toxicity correlate with expression and activation of AHR in trout and zebrafish larvae (Guiney et al., 1997; Henry et al., 1997; Tanguay et al., 1999). Furthermore, the ability of various PCDD, PCDF, and PCB congeners to activate the AHR pathway is correlated with their potency as mediators of TCDD-like developmental toxicity in rainbow trout (Walker and Peterson, 1991; Zabel et al., 1995, 1996). If the AHR is required for TCDD toxicity, components of the AHR signaling pathway in zebrafish larvae should be expressed and active in tissues directly responsive to TCDD. Identifying these tissues and the onset of AHR activity could provide insight into the initiating events of TCDD developmental toxicity. In previous work we qualitatively determined the expression of zfAHR2, zfARNT2b, and zfCYP1A mRNAs isolated from extracts of zebrafish larvae (Tanguay et al., 1999, 2000).
Objectives of the present study were to: (1) quantify the expression of zfAHR2, zfARNT2a, zfARNT2b/c, and zfCYP1A mRNAs during larval development in zebrafish, (2) assess the effect of TCDD exposure on abundance of these mRNAs, (3) determine if the transcripts are colocalized particularly in larval tissues involved in TCDD toxicity, (4) determine if TCDD exposure alters the temporal and/or spatial expression of the mRNAs, and (5) identify larval tissue where AHR activation by TCDD has occurred based on positive immunofluorescence for zfCYP1A. Unfortunately tissue specific expression pattern of zfAHR1 in embryos and larvae could not be determined due to low tissue abundance. To address these objectives vehicle (control) and TCDD exposed zebrafish embryos and larvae were assessed at designated times of development from 12 to 120 h postfertilization (hpf). Since the cardiovascular system is a target of TCDD toxicity in zebrafish larvae, expression of the AHR regulated gene zfCYP1A, and zfAHR2, zfARNT2a, and zfARNT2b/c were investigated beginning at 12 hpf. This is prior to initiation of the heartbeat at 24 hpf and before the first signs of TCDD developmental toxicity are manifested at 72 hpf. Previous studies using high exposure concentrations of TCDD found that zfCYP1A mRNA was induced as early as 24 hpf (Tanguay et al., 1999), but the specific tissues in which the expression occurred was not known. Also it was uncertain if the AHR pathway would be activated by TCDD in zebrafish larval tissues in which toxicity was manifested, and if increased zfCYP1A expression in these tissues would precede the earliest occurring signs of toxicity, namely pericardial edema, reduced trunk blood flow, impaired lower jaw development, and reduced growth at 72 hpf (Henry et al., 1997; Teraoka et al., 2002). In the present study we demonstrate that expression of mRNAs for zfAHR2, zfARNT2a, zfARNT2b/c, and zfCYP1A occur during embryonic and larval development, that zfCYP1A protein is highly expressed in the cardiovascular system of larvae following TCDD exposure, that zfAHR2 and zfARNT2 mRNAs have similar, but not completely overlapping expression patterns, and that transcripts of zfAHR2 and zfARNT2 are found in larval tissues that are involved in TCDD toxicity and that respond to TCDD by induction of zfCYP1A.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Static exposure of zebrafish embryos to waterborne TCDD.
For in situ hybridization 1000 zebrafish embryos (AB strain, Eugene, OR) at 3 hpf with intact chorions were statically exposed for 1 h to either acetone vehicle (control) or an LC100 concentration of TCDD (1.55 nM or 0.5 µg/l) dissolved in acetone (0.1 %, v/v). This concentration of TCDD was chosen to ensure that the AHR signaling pathway would be activated in all responsive tissues to such an extent that zfCYP1A mRNA could be detected by in situ hybridization. Following the 1 h TCDD exposure all embryos were washed 3 times with TCDD-free water. Vehicle- and TCDD-exposed embryos were randomly culled to 100 for isolation at different time points and reared in 150-mm culture plates containing TCDD-free water at 28°C. Culture media was changed twice daily. One hundred control and TCDD-exposed larvae were euthanized with tricaine methanesulfonate (MS-222, 40 mg/l) and fixed overnight at 4°C in 4% paraformaldehyde in PBS at 12, 18, 24, 36, 48, 72, 96, and 120 hpf. This time course allowed TCDD-exposed larvae to exhibit signs of toxicity without losing significant numbers to mortality, which began around 144 hpf. Fixed larvae were dehydrated in a graded series of methanol solutions and stored in 100% methanol at –20°C. Fifty control and TCDD-exposed larvae were raised for 14 days to confirm that 100% mortality was caused by TCDD. Embryos were exposed to TCDD in this manner on at least 3 separate occasions for in situ mRNA analysis. For quantification of mRNA abundance embryos were exposed to vehicle or TCDD as described above; however, more replicates with fewer individuals were exposed. For this experiment 5 groups of vehicle (control) and 5 groups of TCDD-exposed zebrafish larvae (50 larvae/group) were euthanized at each of the following times: 24, 36, 48, 72, and 120 hpf, frozen in liquid nitrogen, and stored at –70°C for RNA isolation and real-time PCR quantification of mRNAs. Additionally whole RNA was isolated from adult zebrafish as a reference for relative mRNA abundance. Embryos and larvae used for whole mount immunohistochemical analysis were exposed to vehicle control or TCDD as above with 6 individuals per time point (24, 36, 48, 72, and 120 hpf) and treatment.
Time course of zfAHR2, zfARNT2a, zfARNT2b/c, and zfCYP1A mRNA abundance.
Total RNA was isolated from vehicle control and TCDD treated groups of fish (n = 50/group). Tissues were shredded using glass beads according to the manufacturer and RNA was isolated using a Qiashredder homogenizer (Qiagen, Chatsworth, CA). One µg of total RNA from each group was resolved on a 1% denaturing formaldehyde agarose gel to ensure quality and concentration of the RNA. cDNA was prepared from 2 µg of RNA per group using Superscript II (Life Technologies, Gaithersberg, MD) and oligo dT primers in a 20 µl volume. Quantitative PCR using gene specific primers was conducted using the Light CyclerTM (Roche, Indianapolis, IN). Specifically, 1 µl of each cDNA pool was used for each PCR reaction in the presence of SYBR Green, according to the manufacturer’s instructions. Agarose gel electrophoresis and thermal denaturation (melt curve analysis) were used to confirm specific product formation.
Oligonucleotides.
All oligonucleotide primers were synthesized by the University of Wisconsin Biotechnology Center and are written 5' to 3', F indicates forward primers corresponding to sense strands while antisense reverse primers are designated with an R. Primers for zfAHR2 and zfARNT2 subcloning were: zfAHR2F1, gcttccgcaaccagaattc; zfAHR2R1, cctagaagttgagtttaaagc; zfARNT2F, catttacagaaaccgagcaaaa; zfARNT2R, cacagtgaaatattccttgatc. Oligonucleotides for real-time PCR were: ß-actinF, aagcaggagtacgatgagtc; ß-actinR, tggagtcctcagatgcattg; zfAHR2F2, acggtgaagctctcccata; zfAHR2R2, agtaggtttctctggccac; zfARNT2F2, gactgaattcctttcgcgccac; zfARNT2b/cR, ctggagctgcttgacgttg, zfARNT2aR cacagtgaaatattccttgatc; zfCYPF, tgccgatttcatccctttcc; zfCYPR, agagccgtgctgatagtgtc.
Probes for in situ hybridization.
A fragment of the 3' end of zfAHR2 was amplified using primers zfAHR2F1 and zfAHR2R1 and subcloned into pGEM-T EASY vector (Promega, Madison, WI). A fragment containing the 5' end of zfARNT2 was generated using primers zfARNT2F and zfARNT2R and subcloned into pGEM-T EASY vector. A partial zfCYP1A clone was obtained from Dr. C. H. Hu of National Taiwan Ocean University. The zfCYP1A sequence shared 79% amino acid identity with the rainbow trout CYP1A (gene bank accession number U62796). zfAHR2, zfARNT2, and zfCYP1A specific digoxigenin-UTP riboprobes were generated from linearized vectors. The zfAHR2 riboprobes corresponded to position bp +2236 to 4147 within the C-terminus of the cDNA (Fig. 1). The zfARNT2 riboprobes (bp –108 to + 813) encompass part of the 5' UTR and most of the PAS B domain (Fig. 1). The 1107 bp zfCYP1A digoxigenin-UTP riboprobe (bp +159 to 1266) corresponds to the rainbow trout amino acids 53 to 422 of the 522 amino acid full-length protein (Fig. 1).
|
Whole mount in situ hybridization of zfAHR2, zfARNT2a/b/c, and zfCYP1A.
Whole mount in situ hybridization was conducted with both sense and antisense digoxigenin RNA probes that were visualized with anti-digoxigenin Fab fragments conjugated with alkaline phosphatase (Roche Molecular Biochemicals) according to standard methods (Westerfield, 1995
). All hybridization was conducted at 62°C. The riboprobe hybridization concentration of zfAHR2 and zfARNT2a/b/c was 1.25 µg/ml and zfCYP1A was hybridized at 500 ng/ml. At least 15 individual organisms were examined for each riboprobe within a treatment group. Images are representative of at least 3 separate TCDD or vehicle exposure experiments. Localization of mRNA expression was confirmed by sectioning embryos and larvae after whole mount in situ hybridization. Embryos and larvae were embedded in paraffin and sectioned. Five µm sections were then deparaffinized and counter-stained with nuclear fast red. The scoring system used to evaluate staining intensity was: 0 = not detected, 1 = light stain, 2 = moderate stain, and 3 = intense stain. Other larvae were frozen, sectioned, and stained prior to light microscopic analysis and evaluation of staining intensity.
Whole mount immunolocalization of zfCYP1A.
The tissue distribution of zfCYP1A protein in zebrafish embryos and larvae was determined using the monoclonal antibody MAb 1-12-3 (Park et al., 1986). Stegeman and coworkers have used this antibody to specifically detect CYP1A in several fish species (Guiney et al., 1997; Iwata and Stegeman, 2000; Schlezinger and Stegeman, 2000; Smolowitz et al., 1992; Stegeman et al., 1989, 2001; Van Veld et al., 1997). TCDD-exposed and vehicle control embryos were fixed overnight in 4% paraformaldehyde in PBS and washed in PBS + 0.1% Tween-20 (PBST). Larvae were permeablized with –20°C acetone for 20 min and rinsed in PBST before digestion with collagenase (1 mg/ml) for 30–90 min. Permeabilized larvae were blocked in 10% normal calf serum in PBST before addition of MAb 1-12-3 (0.3 ug/ml). Following several 10 min washes with PBST, larvae were incubated with a secondary antibody, (Alexa-488 conjugated goat anti-mouse, Molecular Probes, Eugene, OR) for 5 h at 22°C. Tissues were then washed 3 times for 10 min in PBST and visualized by epifluorescence microscopy.
Imaging.
Whole mount in situ images were visualized using a Wild M8 stereomicroscope or a Nikon OPTIPHOT-2 compound microscope attached to a Sony CCD color camera. Video images were captured using Image Pro Plus software (Media Cybernetics, Silver Springs, MD). Fluorescent immunohistochemistry images were digitally captured using a MicroMax 5 Mhz camera (Princeton Instruments, Trenton, NJ).
Statistical analyses.
Gene-specific mRNA abundance was determined at designated times of development from 5 groups of zebrafish larvae (50 larvae/group) exposed to either vehicle or TCDD, respectively. Abundance of mRNAs detected by quantitative real-time PCR was expressed as mean ± SE. Significance of stage of development and TCDD exposure on zfAHR2, zfARNT2a, zfARNT2b/c, and zfCYP1A mRNA abundance was assessed by two-way analysis of variance on log10 transformed data. Differences between groups were analyzed by the Tukey method with significance set at p < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Time Course of zfCYP1A, zfAHR2, zfARNT2a, and zfARNT2b/c mRNA Expression in Whole Body Extracts of Larvae Exposed to Vehicle or TCDD
RNA was isolated from vehicle-exposed zebrafish at different stages of embryonic and larval development and the abundances of zfCYP1A, zfAHR2, zfARNT2a, and zfARNT2b/c mRNAs were determined relative to ß-actin mRNA abundance (Fig. 2
, shaded bars, panels A, B, C, and D). In addition, the effect of exposing zebrafish embryos (3–4 hpf) to an LC100 concentration of TCDD on expression of these same messages was assessed (Fig. 2, solid bars, panels A, B, C, and D). The abundance of each mRNA was also assessed in whole adult extracts as a reference for relative abundance in the whole adult. Additionally the detection of each mRNA in the adult served as a positive control for the assay. Quantitative real-time PCR was used to measure mRNA abundance. More specifically, standard curves allowed assignment of relative concentrations to zfCYP1A, zfAHR2, zfARNT2a, and zfARNT2b/c mRNAs and these values were normalized to zebrafish ß-actin mRNA abundance at each developmental time as a control for mRNA recovery and reverse transcription efficiency. While differences in primer hybridization efficiency prevent the measurement of one mRNA species from being compared to another, general trends in relative mRNA abundance are observable. For the genes evaluated in Figure 2 relative mRNA abundance (from highest abundance to lowest abundance) followed the general trend: zfCYP1A (TCDD) > zfAHR2 (TCDD, vehicle), zfARNT2b/c (TCDD, vehicle), zfCYP1A (vehicle) >> zfARNT2a (TCDD, vehicle). Essentially the same trend was observed for the relative abundance of these messages in Northern blot analyses of zebrafish larvae extracts (Tanguay et al., 1999, 2000).
|
In vehicle-exposed zebrafish embryos, low constitutive expression of zfCYP1A mRNA was detected at 24 hpf with elevated expression at 120 hpf (Fig. 2A
). Note in Figure 2A that the left ordinate is for the vehicle-exposed group and that its scale is 1/10th of that for TCDD-exposed group shown on the right ordinate. TCDD exposure caused elevated expression of zfCYP1A, relative to that in the vehicle (control) group, from 24 to 120 hpf (Fig. 2A). zfAHR2 mRNA was also detected in vehicle-exposed zebrafish embryos and larvae from 24 to 120 hpf (Fig. 2B). zfAHR2 relative mRNA abundance in the TCDD treatment group was similar to the control group at 24, 36, 72, and 120 hpf. However, at 48 hpf TCDD exposure caused a significant increase in zfAHR2 message. To quantify the levels of zfARNT2 splice variants, 2 sets of primers were used. One set amplified only zfARNT2a and another amplified both zfARNT2b and zfARNT2c. The mRNA splice variants of zfARNT2, namely zfARNT2a and zfARNT2b/c, were detected at 24 hpf with very little change in expression through 120 hpf (Figs. 2C and 2D). zfARNT2a (Fig. 2C) was expressed at much lower levels than zfARNT2b/c. This indicates that the contribution of zfARNT2a to the in situ localization of zfARNT2 mRNAs (shown later in Fig. 11) was minimal. Finally, the relative abundance of zfARNT2a and zfARNT2b/c mRNA splice variants were not altered significantly by embryonic exposure to TCDD (Figs. 2C and 2D).
|
15). Lateral or dorsal views of larvae hybridized with the riboprobe at: (A) 12, (B) 18, (C) 24, (D) 36, (E) 48, (F) 72, and (G) 96 hpf are shown. BA, branchial arches; C, cerebellum; HA, hypobranchial artery; Int, intestine; SC, spinal cord; OP, olfactory placode; OT, optic tectum; T, telencephalon.Tissue-Specific Expression of zfCYP1A Protein and mRNA in Larvae Exposed to Vehicle or TCDD
The developmental and tissue specific expression of zfCYP1A mRNA and protein in vehicle- and TCDD-exposed embryos and larvae at 24, 36, 48, and 120 hpf is summarized in Table 1
. zfCYP1A protein was detected in whole mount embryos and larvae immunohistochemically and is denoted by gray shaded cells. zfCYP1A mRNA was detected by whole mount hybridization with an antisense zfCYP1A mRNA probe. Localization of zfCYP1A mRNA was verified histologically in sectioned embryos and larvae that were hybridized with the antisense zfCYP1A mRNA. The intensity of riboprobe signal detected in sectioned tissues was scored with 0 being no detection; 1, light; 2, moderate; and 3, very intense expression. When the intensity of zfCYP1A mRNA expression was different between organisms in the same treatment and age group the average intensity was reported. For each riboprobe, the difference between individual organisms never varied by one unit on this scale. In some cases zfCYP1A mRNA was detected in whole mount embryos and larvae but not upon sectioning due to low expression and or size of the tissue. An asterisk indicates positive zfCYP1A mRNA expression when this occurred.
|
In vehicle (control) zebrafish embryos and larvae, constitutive expression of zfCYP1A mRNA and protein was essentially undetected until 120 hpf where it is primarily localized in tissues of meso- and endodermal origin at low levels (Table 1
). In tissues derived from ectoderm there was no expression of zfCYP1A protein detected at any time from 24 to 120 hpf. Message for zfCYP1A was only detected at very low levels in the ear and vent. In tissues derived from mesoderm in the vehicle (control) group, zfCYP1A mRNA and protein were detected at low levels in the vasculature of the head, trunk, and tail and message was detected at a low level in the heart. In tissues derived from endoderm in vehicle-exposed zebrafish, expression of zfCYP1A mRNA and protein was detected in the intestinal tract at 120 hpf and message was expressed at a low level in the oropharyngeal mucosa and vent.
TCDD exposure markedly increased the diversity of tissues expressing zfCYP1A mRNA and protein with the ectoderm-, mesoderm-, and endoderm-derived tissues all expressing zfCYP1A in all zebrafish embryos and larvae examined (Table 1). TCDD exposure also resulted in earlier developmental expression of zfCYP1A message and protein. In tissues derived from ectoderm, zfCYP1A message and protein was induced in the skin of the head, trunk, and tail. Additionally TCDD strongly induced zfCYP1A mRNA in the neuroepithelium of the ear at 48 and 120 hpf, however, protein could not be positively detected in whole mount tissues since larvae were not sectioned after whole mount immunohistochemical analysis. In the mesoderm, zfCYP1A mRNA preceded protein detection in the vasculature being expressed as early as 24 hpf in head, trunk, and yolk sac blood vessels. Expression of zfCYP1A mRNA was also observed following TCDD exposure in the head and trunk kidney at 36, 48, and 120 hpf. The zfCYP1A protein could not be readily detected in the kidney since the tissues were not sectioned, but kidney expression was detected by in situ hybridization. Endodermal expression of zfCYP1A protein and mRNA was detected in the intestinal mucosa and anal vent of the enteric tract at 48 and 120 hpf. Lastly, the liver of TCDD-exposed larvae expressed high levels of zfCYP1A protein and mRNA at 48 and 120 hpf.
Whole Mount Immunolocalization of zfCYP1A in Larvae Exposed to TCDD
The specificity of immunolocalization was ensured by the lack of signal in embryos incubated without the CYP1A antibody (MAb 12-3-1) but with the secondary antibody, Alexa-488 (data not shown). Specific immunolocalization of zfCYP1A was difficult to detect prior to 72 hpf due to autoflourescence of the yolk. zfCYP1A protein was first detected in TCDD-exposed larvae at 36 hpf (Fig. 3A) as fluorescence in the anal and urinary pores (AP, UP; Fig. 3B), developing vasculature of the tail including the caudal artery and vein (CA, CV), and intersegmental vessels (SE; Fig. 3B), and as a punctate pattern of fluorescence in the skin (Fig. 3C). In TCDD-exposed zebrafish larvae at 48 hpf (Fig. 3D) and 72 hpf (Fig. 3F), zfCYP1A immunofluorescence was evident in the vasculature throughout the embryo including the intersegmental vessels (SE), caudal artery and vein (CA, CV), dorsal aorta, pectoral vasculature, gill arches and vessels in the head (BA) as well as in the anal and urinary pores (AP, UP; Figs. 3D, 3E, and 3F). Since the heart overlays the yolk sac, specific zfCYP1A localization of the heart could not be determined until the heart separated from the yolk sac at 72 hpf.
|
Constitutive expression of zfCYP1A protein was detected in vehicle-exposed zebrafish larvae at 120 hpf in all individuals examined (Fig. 4A
). In this representative larva weak zfCYP1A immunofluorescence was observed in the head, gastrointestinal tract, and vasculature throughout the body. These findings were consistent with the time course of zfCYP1A mRNA expression measured by real-time PCR and by whole mount in situ hybridization results both of which demonstrated that zfCYP1A message was only weakly expressed in vehicle-exposed larvae at 120 hpf and was localized in the gastrointestinal tract (Fig. 2A, Figs. 8E–F, Table 1). In sharp contrast, embryonic exposure to TCDD resulted in elevated levels of zfCYP1A protein in zebrafish larva at 120 hpf (Figs. 4B–D). The most predominate tissue in which increased zfCYP1A expression occurred was the vascular endothelium throughout the body (Figs. 4B–D). The lateral view of this TCDD-exposed larva shows vascular specific detection of zfCYP1A in the head, trunk, and tail (Fig. 4B). The lateral and ventral views demonstrate that zfCYP1A was also immunolocalized to the heart, liver, and enteric tract of the TCDD treated larva (Figs. 4B–C).
|
|
Evaluation of the immunofluorescence for zfCYP1A in individual blood vessels in various regions of the head of a representative TCDD-exposed zebrafish larva at 120 hpf is shown in Figure 5
. Vascular endothelium-specific zfCYP1A immunofluorescence was observed in essentially all vessels examined in the head including: branchial arches (BA), posterior cerebral vein (PCeV), primordial midbrain channel (PMBC), and prosencephalic artery (PrA; Fig. 5A, lateral view); branchial arches (BA), hypobranchial artery (HA), opercular artery (ORA), and optic vein (OV; Fig. 5B, ventral view); dorsal longitudinal vein (DLV), mesencephalic vein (MsV), and posterior cerebral vein (PCeV; Fig. 5C, dorsal view); inner optic circle (IOC) and nasal artery (NA) of the eye (Fig. 5D, lateral view); and pectoral artery (PA) and pectoral vein (PV) of the pectoral fin (Fig. 5E, lateral view).
|
Immunolocalization of zfCYP1A to the vascular endothelium in the trunk and tail regions of a representative TCDD-exposed zebrafish larva at 120 hpf is shown in Figure 6
. In the trunk and tail, zfCYP1A was evident in the vascular endothelium of the dorsal aorta (DA), intersegmental vessels (SE), and posterior cardinal vein (Fig. 6A) and in the subintestinal vein (SIV; Fig. 6B). Figure 6 also reveals positive immunofluorescence for zfCYP1A in the kidney, intestine, and anal pore.
|
Immunofluorescent detection of zfCYP1A in the 120 hpf zebrafish larvae exposed to TCDD was also observed in the heart ventricle (V) and atrium (A; Fig. 7A
) and in the liver (L) and hepatic portal vein (HPV; Fig. 7B).
|
Whole Mount in Situ Hybridization: Lack of Staining with Sense RNA Probes for zfCYP1A, zfAHR2, and zfARNT2a/b/c mRNAs
Since high quality antibodies are not yet available to detect zfAHR2 and zfARNT2, whole mount in situ hybridization was used to assess the temporal and spatial expression of zfCYP1A, zfAHR2, and zfARNT2a/b/c mRNAs in zebrafish at 12, 18, 24, 36, 48, 72, 96, and 120 hpf. Antisense hybridization specificity was confirmed for each riboprobe using sense strand specific riboprobes and the tissue specificity and staining intensity for each probe was confirmed by sectioning the whole mount embryos and larvae. None of the control sense probes produced staining at any developmental stage or in any larval tissues except for the swimbladder. This lack of staining for sense strand AHR2 riboprobe is illustrated in zebrafish embryos and larvae at 12, 18, 24, 36, 48, and 96 hpf (Fig. 8
).
zfCYP1A mRNA Localization in Larvae Exposed to Vehicle or TCDD
Constitutive zfCYP1A mRNA expression was not detected by whole mount in situ hybridization in vehicle-exposed zebrafish larvae until 96 hpf when it was observed by light blue staining in endoderm-derived tissues such as the liver and enteric tract (Fig. 9E, top). This finding is consistent with elevated constitutive expression of zfCYP1A mRNA being detected by RT-PCR in vehicle (control) larvae at 120 hpf. In contrast to the weak and limited expression of zfCYP1A mRNA expression in vehicle (control) larvae, expression was readily detected in zebrafish larvae exposed to TCDD (Fig. 9). Expression was first evident at 24 hpf in tissues of mesodermal and ectodermal origin including the dorsal aorta (DA), anal pore (AP), urinary pore (UP), and skin (Fig. 9A, bottom). By 36 hpf, zfCYP1A mRNA expression was elevated in blood vessels throughout the body including the caudal artery (CA), caudal vein (CV), common cardinal vein (CCV), dorsal aorta (DA), gill arches, head vasculature (HV), posterior cardinal vein (PCV), as well as in the hatching gland (HG; Fig. 9B, bottom). TCDD-induced zfCYP1A mRNA expression at 48 hpf was observed in tissues derived from ectoderm, mesoderm, and endoderm. Ectoderm derived tissues expressing zfCYP1A mRNA included the ear and skin associated with the head, trunk, tail, and yolk sac. Expression was also observed in mesoderm-derived tissues represented by the vasculature including the head vasculature (HV), fin vasculature including the pectoral vein (PV) and pectoral artery (PA), posterior cardinal vein (PCV), caudal vein (CV), and caudal artery (CA). Endoderm derived tissues expressing zfCYP1A message included the mucosa associated with the anal pore (AP) and urinary pore (UP; Fig. 9C, bottom). The expression pattern at 48 hpf remained similar at 72 and 96 hpf and was dominated by vascular-specific localization. At 72 hpf this was represented by expression of zfCYP1A message in the branchial arches (BA), head vasculature (HV), and subintestinal veins (SIV; Fig. 9D, bottom) and at 96 hpf by expression in the branchial arches (BA; Fig. 9E, bottom). At 72 hpf expression of zfCYP1A mRNA was evident in the heart (H), anal pore (AP) and urinary pore (UP; Fig. 9D, bottom). By 96 hpf expression is much more evident in tissues of endodermal origin including the liver (L) and intestinal tract (Int) as well as in tissues of ectodermal origin such as the brain (BR) and ear (Fig. 9E, bottom). This pattern of zfCYP1A message expression persists at 120 hpf (Fig. 9F, bottom).
|
Tissue-Specific Expression of zfAHR2 and zfARNT2a/b/c mRNAs in Embryos and Larvae Exposed to Vehicle or TCDD
Tables 2 and 3
summarize the results for tissue-specific expression of zfAHR2 and zfARNT2a/b/c mRNAs from vehicle control and TCDD-exposed embryos and larvae that were hybridized with gene specific riboprobes at 24, 36, 48, and 120 hpf. Abundance and localization of mRNA expression was empirically determined histologically by sectioning embryos and larvae after whole mount in situ hybridization. The intensity of riboprobe signal detected in sectioned tissues was empirically noted with 0 being no detection; 1, light; 2, moderate; and 3, very intense expression. When the intensity of zfAHR2 or zfARNT2a/b/c mRNA expression was different between individual organisms in the same treatment and age group, the average intensity was reported. In some cases either zfAHR2 or zfARNT2a/b/c mRNA was detected in whole mount embryos and larvae but not detected histologically due to low expression and/or size of the tissue. An asterisk (*) in Tables 2 and 3 indicates positive zfAHR2 or zfARNT2a/b/c mRNA expression when this occurred. Shaded cells in Tables 2 and 3 indicate that TCDD-induced zfCYP1A protein and/or mRNA was detected in those particular tissues. Finally according to the in situ analysis there were essentially no differences in tissue-specific expression of either zfAHR2 or zfARNT2a/b/c mRNAs between vehicle and TCDD-exposed larvae at 24, 36, 48, or 120 hpf.2 Therefore, scores for zfAHR2 and zfARNT2a/b/c mRNA expression intensity in the vehicle (control) and TCDD exposure groups were combined at 24, 36, 48, and 120 hpf (Tables 2 and 3).
|
|
zfAHR2 and zfARNT2a/b/c mRNAs were expressed in several tissues derived from ectoderm at 24, 36, 48, and 120 hpf (Table 2
). In the head both transcription factors were expressed in the skin, brain, sensory organs including the eyes and ears and gill. Of the various positively staining structures in the head the highest expression of zfAHR2 and zfARNT2a/b/c messages were in the ears. In tissues derived from ectoderm in the trunk, tail, and yolk sac the expression of these 2 mRNAs was weak and occurred in the skin, spinal cord, and pectoral fins of the trunk, skin and caudal fin of the tail, and skin of the yolk sac. Expression of zfCYP1A protein or mRNA in response to TCDD exposure (shaded cells in Table 2) was more restricted than expression of zfAHR2 and zfARNT2a/b/c mRNAs. This was illustrated by colocalization of zfCYP1A, zfAHR2, and zfARNT2a/b/c occurring in only 2 ectoderm-derived tissues, the ears and skin.
Expression of mRNAs for zfAHR2 and zfARNT2a/b/c in tissues derived from mesoderm and endoderm was also observed in zebrafish larvae (Table 3). In general the expression of these messages in tissues of mesodermal and endodermal origin was weak and widespread. For zfAHR2 mRNA low expression was observed in mesoderm-derived tissues of the head (blood vessels, meninges of the brain, and head kidney), trunk (heart, blood vessels, peritoneum, somites, lateral plate, and trunk kidney), tail (blood vessels, caudal fin, and somites), and yolk sac (blood vessels). For endoderm-derived tissues low expression of zfAHR2 mRNA was observed in the head (pharyngeal mucosa) and trunk (mucosa of the enteric tract and gall bladder) but was conspicuously absent from the liver. Expression of zfARNT2a/b/c generally followed the tissue distribution of zfAHR2 for mesoderm- and endoderm-derived tissues with certain exceptions. Most obvious was the greater expression of zfARNT2a/b/c in blood vessels, peritoneum, and caudal fin mesenchyme. Expression of zfCYP1A protein or mRNA in response to TCDD exposure (shaded cells in Table 3) was more restricted than expression of zfAHR2 and zfARNT2a/b/c mRNAs. Colocalization of zfCYP1A, zfAHR2, and zfARNT2a/b/c occurred in only 3 mesoderm-derived tissues, blood vessels, heart, and trunk kidney. For endoderm-derived tissues only the enteric tract exhibited co-expression of zfCYP1A with zfAHR2 and zfARNT2a/b/c.
zfAHR2 mRNA Localization in Larvae Exposed to Vehicle or TCDD
Since there were essentially no differences between the vehicle (control) and TCDD treatment groups in zfAHR2 mRNA expression, whole mount in situ hybridization images of zebrafish embryos and larvae, representative of both groups, are shown at 12, 18, 24, 36, 48, 72, and 96 hpf, respectively (Fig. 10). At 12 hpf zfAHR2 was ubiquitously expressed in the embryo (Fig. 10A) but by 18 hpf increased expression of zfAHR2 mRNA was observed in the embryo trunk in the general region where hemangioblasts are found (Fig. 10B). By 24 hpf, zfAHR2 mRNA expression was observed in skin of the head, trunk, tail, and yolk sac, regions of the brain including the optic tectum (OT) and cerebellum (C), spinal cord, and sensory organs including the eye and ear (Fig. 10C). Blood vessels throughout the larval body were found to express zfAHR2 at 36 hpf and they included the branchial arches (BA), caudal artery (CA), caudal vein (CV), common cardinal vein (CCV), and dorsal aorta (DA; Fig. 10D). The hatching gland (HG), ear, and anal pore also expressed zfAHR2 message at this time (Fig. 10D). At 48 hpf, zfAHR2 mRNA continued to be expressed in the vasculature and was detected in the heart (H) and in the pectoral artery (PA) and pectoral vein (PV) of the pectoral fin (Fig. 10E). At 72 hpf zfAHR2 was observed in the heart (H), hypobranchial artery (HA), subintestinal veins (SIV), and developing jaw (Fig. 10F). At 96 hpf the pattern of zfAHR2 mRNA expression was generally similar to that observed at 72 hpf (Fig. 10G). Expression of zfAHR2 message in the vasculature continued through 120 hpf, but expression in the brain, eyes, spinal cord, kidney, and trunk and tail somites (seen at earlier stages of development) was not detected at this time (Tables 2 and 3). However, zfAHR2 mRNA was detected in the intestinal mucosa and gall bladder at 120 hpf (Table 3).
|
zfARNT2a/b/c mRNA Localization in Larvae Exposed to Vehicle or TCDD
zfARNT2a/b/c mRNA expression was low at all stages of development and essentially unaffected by TCDD exposure. As with zfAHR2, zfARNT2a/b/c was expressed primarily in tissues of ectodermal and mesodermal origin. zfARNT2a/b/c mRNA expression was ubiquitously expressed in the zebrafish embryo at 12 hpf (Fig. 11A
). At 18 hpf zfARNT2a/b/c expression was localized in the brain and spinal cord and was especially abundant in the forebrain including the olfactory rosette (Fig. 11B). At 24 hpf expression of zfARNT2a/b/c message was confined predominantly to the brain localizing in the olfactory placode (OP), telencephalon (T), optic tectum (OT), cerebellum (C), spinal cord (SC) and sensory organs including the eye and ear (Fig. 11C, lateral and dorsal views). At this time zfARNT2a/b/c mRNA was also observed in the hatching gland and yolk sac skin. By 36 hpf, expression continued in the central nervous system and could also be detected in the cardiovascular system including the pericardium of the heart and in blood vessels in the head, trunk, tail, and yolk sac (Fig. 11D). The expression pattern of zfARNT2a/b/c mRNA remained similar at 48 hpf with additional expression in the pectoral fin (Fig. 11E). Also the expression of zfARNT2a/b/c mRNA in the vasculature throughout the body of the zebrafish larva at 48 hpf was colocalized with zfAHR2 mRNA. At 72 hpf zfARNT2a/b/c message could be detected in the branchial arches (BA), developing jaw, and brain (Fig 11F). At 96 hpf zfARNT2a/b/c mRNA was expressed in vascular structures including the branchial arches (BA) and hypobranchial artery (HA) as well as the enteric tract including the intestine (Int; Fig. 11G). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Determining the expression pattern and abundance of zfAHR2 and zfARNT2 is crucial for understanding normal development in addition to determining a mechanistic understanding of TCDD developmental toxicity in zebrafish. Research in laboratory mammals has demonstrated the necessity of AHR and ARNT2 for normal development. The AHR is required for normal vascular development in mice (Lahvis et al., 2000
; Schmidt et al., 1996). In addition, developmental TCDD toxicity requires a functional AHR. Much of the developmental toxicity of TCDD is presumed to involve the AHR since developing mice lacking a functional AHR are insensitive to acute TCDD exposure (Fernandez-Salguero et al., 1996; Mimura et al., 1997; Peters et al., 1999). ARNT2 is also important for normal development. Mice deficient in ARNT2 die within 24 h of birth and have impaired hypothalamic development (Hosoya et al., 2001). In zebrafish ARNT2 may be necessary for normal development since over expression of a dominant negative, truncated splice variant of zfARNT2 (ARNT2x) alters development and is embryo lethal (Hsu et al., 2001). The known members of the AHR pathway in zebrafish include zfAHR1, zfAHR2, and 4 zfARNT2 splice variants. For this study we chose to investigate the members that are functionally active in response to TCDD. zfAHR1 is not suspected to be involved in TCDD toxicity since it does not bind ligand efficiently, is not transcriptionally active with zfARNT2, and has a very limited tissue distribution (Andreasen et al., 2002). Only 1 splice variant of zfARNT2, zfARNT2b, is known to be functionally active with zfAHR2 (Tanguay et al., 2000). For these reasons it is implied that zfAHR2 and zfARNT2b are the functional members of the AHR pathway in zebrafish. The findings from this study suggest that (1) zfAHR2, zfARNT2a, and zfARNT2b/c and zfCYP1A are all expressed early in development prior to the first signs of TCDD toxicity; (2) TCDD exposure has no effect on zfARNT2 abundance and only a marginal transient inductive effect on zfAHR2; (3) zfAHR2, zfARNT2 and TCDD induced zfCYP1A mRNAs colocalize in several tissues affected by TCDD exposure; (4) TCDD exposure has essentially no effect on the temporal and spatial localization of zfAHR2 and zfARNT2; and (5) immunolocalization of TCDD-induced zfCYP1A identified the cardiovascular and osmoregulatory systems as sites of active AHR signaling.
TCDD-Induced Cardiovascular Toxicity in Zebrafish Larvae
The cardiovascular system is a target of TCDD induced developmental toxicity in zebrafish (Henry et al., 1997). In zebrafish larvae, exposed to TCDD at the embryonic stage of development, circulation is normal from onset of the beating heart at 24 hpf until about 72 hpf when it is first detected as being reduced in the trunk and tail (Henry et al., 1997). Later in development a reduction in blood flow is observed in the head and gills (99 hpf), and is followed by reduced heart rate (125 hpf). TCDD-exposed larvae also show signs of pericardial edema, impaired lower jaw development, and reduced body growth by 72 hpf followed by yolk sac edema and reduced swimbladder inflation by 96 hpf. Mortality begins to occur at 144 hpf. Most of these signs of toxicity, with the notable exception of impaired lower jaw development (Teraoka et al., 2002), are consistent with impaired cardiovascular function. Research in rainbow trout and lake trout larvae has shown that TCDD can impair heart development and increase vascular permeability leading to edema. This appears to involve activation of AHR signaling because the TCDD dose response relationship for increased CYP1A protein expression in the vascular endothelium of lake trout larvae was discovered to be similar to that for larval mortality (Guiney et al., 1997). As with zebrafish, lake trout larvae exposed to TCDD exhibit circulatory failure and extensive edema prior to mortality. Furthermore, CYP1A expression in the vascular endothelium of lake trout larvae occurs prior to the onset of increased vascular permeability and edema (Guiney et al., 2000). Expression of CYP1A in the vasculature of zebrafish larvae also occurs well before the onset of edema. A feature of the developmental cardio-toxicity in rainbow trout larvae is the finding that heart size is greatly reduced by TCDD (Hornung et al., 1999). Thus, in zebrafish and trout larvae it is possible that either increased vascular permeability and/or reduced cardiac output could lead to reduced blood pressure and renal perfusion culminating in edema that is the hallmark sign of TCDD developmental toxicity in fish. A compromised cardiovascular system would result in reduced blood flow and may decrease the exchange of gases at the gills, swimbladder, and skin. Other signs of TCDD developmental toxicity in zebrafish and other fish could be accounted for by insufficient gas exchange promoting tissue necrosis, reduced growth, and inhibition of swimbladder inflation.
Abundance and Tissue Specific Expression of zfAHR2, zfARNT2, and zfCYP1A
Constitutive levels of zfCYP1A mRNA and protein were detected at 24 and 120 hpf respectively. This is earlier than the previously reported initial constitutive zfCYP1A mRNA and protein expression using northern and western analysis (Mattingly and Toscano, 2001; Tanguay et al., 1999). The use of real time PCR and whole mount immunofluorescence allowed for detection of the previously unknown increase in the constitutive expression of zfCYP1A mRNA and protein at 120 hpf due to the sensitivity of the assays used. This correlates with the increase in constitutive aryl hydrocarbon hydroxylase or EROD activity observed after hatching in Fundulus heteroclitus and zebrafish respectively (Binder and Stegeman, 1984; Binder et al., 1985; Mattingly and Toscano, 2001).
TCDD exposure induced the level of zfCYP1A expression in zebrafish embryos substantially at 24 hpf and it remained elevated thereafter. This indicates that TCDD activates the AHR pathway, potentially involving zfAHR2 and zfARNT2a/b/c, before the first signs of developmental toxicity are manifested. Whole mount immunolocalization of zfCYP1A protein was studied to identify potential sites of TCDD toxicity by detecting tissues with an activated AHR pathway. Tissue and developmental stages expressing an active AHR signaling pathway were determined using a monoclonal antibody (MAb1-12-3) to detect zfCYP1A protein in situ at several developmental time points. TCDD induced zfCYP1A protein expression was first localized to the skin and developing vasculature of the zebrafish larva prior to the first signs of toxicity. The cardiovascular system, kidney, and liver of TCDD-exposed zebrafish larvae all expressed zfCYP1A protein by 120 hpf. This larval expression pattern is similar to that seen in F. heteroclitus (Toomey et al., 2001). The tissue distribution of zfCYP1A in TCDD-exposed larvae at 120 hpf also resembles that seen in TCDD-exposed lake trout larvae and the adult pattern observed in several other fish species including zebrafish (Andreasen et al., 2002; Buchmann et al., 1993; Guiney et al., 1997; Husoy et al., 1994; Schlezinger and Stegeman, 2000; Smolowitz et al., 1991, 1992; Stegeman et al., 1989, 2001; Van Veld et al., 1997). The use of zfCYP1A mRNA and protein localization as biomarkers of a TCDD activated AHR pathway has demonstrated, like in the lake trout larva (Guiney et al., 1997), that the zebrafish larva vasculature is the first site of CYP1A expression. Additionally this expression localizes to one of the first sites of TCDD developmental toxicity, reduced blood flow in the trunk at 72 hpf. Besides the vasculature, the AHR pathway also is activated by TCDD in tissues involved in osmoregulation and circulation, including the gills, kidney, heart, and skin. Whether zfCYP1A is directly involved in TCDD toxicity is unknown. Studies in O. latipes and F. heteroclitus suggest that TCDD induced apoptosis occurs in the developing vasculature (Cantrell et al., 1998; Toomey et al., 2001) possibly accounting for vascular related signs of toxicity. However this could not be confirmed in rainbow trout larvae (Hornung et al., 1999).
To determine whether zfAHR2 and zfARNT2a/b/c could play a role in TCDD induced zfCYP1A expression, mRNA abundance and tissue specific localization during development was examined. zfAHR2 and zfARNT2a/b/c were both expressed as early as 12 hpf. This precedes the initiation of a beating heart and the first overt signs of TCDD developmental toxicity. This is consistent with results in F. heteroclitus where expression of AHR2 and ARNT2 preceded the onset of circulation (Powell et al., 2000). The abundance of zfAHR2 and zfARNT2 mRNAs is relatively constant from 24 to 120 hpf.
Previous studies in rainbow trout and zebrafish cell lines and whole trout and zebrafish demonstrated that AHR2 but not ARNT2 mRNAs could be induced by TCDD-exposure (Abnet et al., 1999; Tanguay et al., 1999). However, it was not known if a lower TCDD-exposure concentration could either increase the abundance or alter the localization of zfAHR2 or zfARNT2a/b/c mRNA. Expression of zfARNT2a/b/c mRNA was not altered by TCDD exposure. However, whole embryonic mRNA abundance of zfAHR2 was marginally, but significantly increased by TCDD at 48 hpf. This was also observed in zebrafish embryos exposed to higher TCDD concentrations (3.7–30 nM) than used in the present study (1.55 nM) (Tanguay et al., 1999). Since the lower TCDD concentration used in the current investigation was still sufficient to cause 100% mortality and only transiently altered zfAHR2 mRNA abundance it suggests that TCDD developmental toxicity in the zebrafish is not likely dependent on an elevation of whole larva zfAHR2 mRNA expression.
zfAHR2 and zfARNT2b form functional dimeric complexes in vitro, but it was unknown whether these 2 proteins would be co-expressed in TCDD responsive tissues. If zfAHR2 and zfARNT2 are essential for TCDD responsiveness in vivo, they should be expressed in the same tissues, and should co-localize with TCDD-induced zfCYP1A mRNA and protein expression. zfAHR2, zfARNT2, and TCDD-induced zfCYP1A mRNAs were similarly localized in the vasculature, heart, trunk kidney, ear, and skin by 48 hpf. zfAHR2 and zfARNT2 are presumed to be functional during early development because the TCDD-induced zfCYP1A induction co-localizes with zfAHR2 and zfARNT2 at this time. Confirming the quantitative results, TCDD not only had little effect on whole larval mRNA abundance but localization of zfAHR2 and zfARNT2 were not drastically affected either. Immunohistochemical localization of zfCYP1A suggests that zfAHR2 and zfARNT2a/b/c are responsible for the induction of CYP1A by TCDD as both are co-localized in the vasculature and skin. Identification of the vascular system as a site of zfCYP1A expression is not unique to TCDD-exposed zebrafish larvae. Vascular specific and ligand dependent CYP1A expression has been observed in every fish species studied as well as in the chicken and mouse (Dey et al., 1999; Guiney et al., 1997; Iwata and Stegeman, 2000; Schlezinger and Stegeman, 2000; Smolowitz et al., 1992; Stegeman et al., 1989, 2001; Van Veld et al., 1997; Walker et al., 2000). Expression in the nervous system was similar between zfAHR2 and zfARNT2 in some tissues. zfAHR2 and zfARNT2 were detected in the brain, spinal cord, and sensory organs in the head; however, TCDD-induced zfCYP1A expression was not. The lack of detection of zfCYP1A expression in the head (except for vascular tissues) suggests that zfAHR2 and zfARNT2 most likely have other physiological functions besides altering expression of the TCDD induced battery of genes. A comparison of the localization of AHR, ARNT, and ARNT2 in rat brain suggests that ARNT2 may interact with other PAS proteins (Petersen et al., 2000). The very neuronal specific expression of zfARNT2 as early as