* Integrated Toxicology Program and Nicholas School of the Environment and Earth Sciences, Duke University, Durham, North Carolina 27708; and Department of Population Health and Pathobiology, College of Veterinary Medicine, North Carolina State University, Raleigh, North Carolina 27606
2 To whom correspondence should be addressed at Division of Environmental Sciences and Policy, Nicholas School of the Environment and Earth Sciences, Duke University, Durham, NC 277080328. Fax: 9196848741. E-mail: swkull{at}duke.edu.
Received October 28, 2004; accepted February 4, 2005
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: Japanese medaka; TCDD; gene expression; brain; liver; testis.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In recent years, genome-wide expression analysis has been exploited in vitro (Adachi et al., 2004; Frueh et al., 2001
; Puga et al., 2000
) and in vivo (Kurachi et al., 2002
) to uncover additional AhR-dependent or -independent TCDD-responsive genes. However, the majority of these studies have centered on gene expression analysis in TCDD-exposed hepatoma cells or liver. To our knowledge, wide-scale gene expression analysis has not been thoroughly investigated with extra-hepatic organs affected by TCDD. Moreover, although much research has focused on acute and chronic dioxin toxicity in fish model systems (Hahn, 2001
), and several research groups have significantly advanced gene microarray technology for ecotoxicological fish models (Denslow et al., 2004
; Larkin et al. 2003
; Oleksiak et al., 2002
; Williams et al., 2003
), to date there are no published studies investigating organ-specific gene expression responses in fish following TCDD exposure.
To this end, we used suppression subtractive hybridization (SSH) as a screening tool to initially evaluate qualitative gene expression changes in male Japanese medaka (Oryzias latipes) organs (whole brain, liver, and testis) after intraperitoneal TCDD injection and exposure for 48 h. At present, oligonucleotide- or cDNA-based microarrays are not commercially available for Japanese medaka. Like microarray-based gene expression analysis, SSH analysis provides qualitative evaluation of mRNA-level differences between control and toxicant-exposed animal tissues. Suppression subtractive hybridization relies on hybridization-dependent subtraction of equally abundant transcripts and selective PCR-amplification and enrichment of differentially expressed transcripts (Diatchenko et al., 1996). However, SSH is biased toward enrichment of transcripts that are highly abundant in one sample but not the other, and low abundant transcripts that are differentially expressed are not as likely to be identified (Ji et al., 2002
). Therefore, in this study, after SSH-based identification of differentially expressed transcripts, expression of genes suspected to be strongly responsive to TCDD was tested with organ-specific replicate nylon membrane arrays, and relative CYP1A and AHR1 transcript levels were measured by real-time RT-PCR. Moreover, qualitative histopathologic evaluation was used to relate organ pathology with gene expression patterns. Overall, we demonstrate that TCDD induces organ-specific qualitative and semi-quantitative gene expression differences in male medaka, and that these differences are associated with histopathological changes.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Test animals.
Oryzias latipes (Japanese medaka) is a small (34 cm) egg-laying freshwater fish native to Japan, Korea, and eastern China, and it is a well-established developmental, genetic, and toxicological animal model (Law et al., 2003; Wittbrodt et al., 2002
). Fish used for this study were collected from an orange-red line under standard recirculating aquaculture conditions. All fish were handled and treated according to protocols approved by the Duke Institutional Animal Care and Use Committee (IACUC).
TCDD exposures.
Adult male medaka (67 months old) (350-mg body weight) were isolated to a separate aquarium tank and acclimated under recirculating culture conditions for 1 week prior to injection. Fish were intraperitoneal (i.p.)-injected in the abdominal region with a sterile, glass 25-µl Hamilton syringe equipped with an ultra-fine needle with 1 µl HPLC-grade DMSO (vehicle) or 1 µl TCDD stock (3.5 ng/µl;
10 µg-TCDD/kg-body weight nominal). For cDNA subtraction library generation, six fish per treatment were incubated static at 25°C under 16 h:8 h light:dark conditions for 48 h in ethanol-rinsed 2 l glass beakers containing sterile, embryo rearing medium (ERM) (17.1 mM NaCl, 272 µM CaCl2 2H2O, 402 µM KCl, 661 µM MgSO4 7H2O; pH 7.2) (Kirchen and West, 1976
). For cDNA macroarray and histopathological analysis, 12 fish per treatment were incubated static in two replicate, ethanol-rinsed 2 l glass beakers (6 fish per beaker) containing sterile ERM. For both experiments, fish were not fed, and no mortalities or abnormal behavior were observed throughout the 48-h exposure period. At test termination, medaka were removed and anesthetized in ice-cold ERM. Using scissors and forceps cleaned with RNaseZAP (Sigma) and 100% ethanol to reduce RNase contamination, organs (whole brain, liver, and testis) were removed and immediately frozen in liquid nitrogen. Individual organs from six DMSO- or TCDD-exposed fish were pooled and used for generation of cDNA subtraction libraries (1 pool/treatment/organ); individual organs from four fish within each treatment replicate were pooled for cDNA macroarray analysis (2 pools/treatment/organ). For the second experiment, the remaining two fish from each treatment replicate (4 fish total per treatment) were sacrificed for histopathological analysis.
Histopathological analysis.
At 48 h post-injection, DMSO-exposed and TCDD-exposed male medaka were anesthetized in ice-cold ERM and peduncle transected caudal to the anus using clean forceps. Clean scissors were then used to open the abdominal cavity from rostral to the anus. This permitted fixative to make contact will all internal viscera. Fish were fixed in 2% paraformaldehyde/phosphate buffered saline (PBS; pH 7.4) for 72 h at 4°C, and stored in 6% sucrose/PBS (pH 7.4) at 4°C until mounting and sectioning. Fish were oriented in lateral recumbency, paraffin-embedded, and 68µm thick-step sections through the whole body were mounted on glass slides and stained with hematoxylin and eosin. All animals were surveyed and imaged (40x and 100x) with a Nikon Eclipse E600 light microscope, a Nikon DXM 1200 digital camera, and EclipseNet imaging software (Nikon).
Total RNA isolation.
For each treatment-specific RNA extraction, pooled organs (whole brain, liver, or testis) were homogenized with 1 ml RNA Bee (TelTest) in a stainless-steel Polytron homogenizer (Kinematica) cleaned with RNaseZAP (Sigma), DEPC-treated water, and sterile de-ionized water. After homogenization, 200 µl chloroform was added and the homogenate was mixed vigorously for 30 s. The mixture was then chilled on ice for 5 min and centrifuged at 12,000 x g at 4°C for 15 min. The upper aqueous phase was transferred to a new 2-ml RNase-free tube containing 500 µl isopropanol and mixed well by inverting the tubes. Samples were centrifuged at 12,000 x g at 4°C for 5 min, the supernatant was discarded, and the RNA pellet was washed once by vortexing with 1 ml 75% ethanol. After centrifugation at 7500 x g at 4°C for 5 min, the ethanol wash was removed and the sample air-dried for 10 min. The resultant pellet was re-suspended in 100 µl RNase-free water warmed to 52°C. To eliminate DNA contamination, each sample was on-column-digested with DNase using an RNase-free DNase Set according to the manufacturer's instructions (Qiagen), and eluted with 30 µl warmed (52°C) RNase-free water. RNA quantity and quality were verified with a NanoDrop ND-1000 spectrophotometer. Total RNA concentrations averaged 4004500 ng/µl with 260/280 ratios 2.0.
Generation of subtraction libraries.
Suppression subtractive hybridization was performed based on the original procedures of Diatchenko et al. (1996). Following the manufacturer's instructions, dsDNA was synthesized from 2 µg total RNA using a BD SMART PCR cDNA Synthesis Kit (BD Biosciences); common-level cDNAs were then subtracted, and differentially expressed sequences were enriched using a Clontech PCR-Select cDNA Subtraction Kit (BD Biosciences). For each organ, subtracted cDNA libraries were generated in forward and reverse directions. cDNAs from the resultant libraries (2 libraries per organ) were inserted into pCR 2.1 vector (Invitrogen) and cloned using INV
F' bacterial cells (Invitrogen). After overnight growth on agar plates, 384 individual bacterial colonies each for forward- and reverse-subtracted libraries (768 total colonies per organ) were individually transferred with sterile toothpicks to 96-well microplates containing 100 µl LB media with ampicillin (100 µg/ml) per well. Microplates were covered with breathable membranes (USA Scientific), and colonies were grown overnight at 37°C with constant agitation at 150 rpm. After overnight growth, all isolated colonies were stored in 60% glycerol at 80°C until differential screening.
Clones containing subtracted cDNAs were screened for false-positives with a PCR-Select Differential Screening Kit (BD Biosciences). Briefly, individual cDNA inserts were PCR-amplified with 1 µl bacterial culture template and nested primers (NP-1 and NP-2R) provided by the manufacturer. Polymerase chain reaction products (5 µl) were denatured at 65°C for 5 min with 0.6 M NaOH, chilled on ice for 23 min, and spotted with a vacuum dot blotter in 96-well format on duplicate positively charged nylon membranes (75 mm x 115 mm) (Roche Diagnostics). Spotted membranes were neutralized in 0.5 M Tris/1.5 M NaCl for 4 min, rinsed in water for 2 min, and UV-cross-linked prior to drying and storage.
Each duplicate set of membranes was screened with probes from the appropriate organ-specific forward- and reverse-subtracted cDNA library. Forward- and reverse-subtracted cDNA probes were generated at 37°C for 60 min with [32P] dATP (or dCTP) and purified using Microspin G-25 columns (Amersham Biosciences). Immediately following probe purification, probe radioactivity was quantified in a scintillation counter. 32P-labeled library probes were then hybridized to subtracted clones arrayed on membranes according to the manufacturer's instructions (BD Biosciences). After washes, hybridized membranes were exposed onto phosphor screens for 72 h and scanned with a Storm 860 Phosphoimager (Molecular Dynamics). Positive clones were identified based on visual inspection of each set of membranes spotted with either forward- or reverse-subtracted libraries. Forward library probes were expected to hybridize to forward library clones but not to reverse library clones. Alternatively, reverse library probes were expected to hybridize to reverse library clones but not to forward library clones. In general, these predictions were true. However, some forward and reverse library probes hybridized equally to both library clones; these represented false-positive transcripts that were not differentially expressed between DMSO-treated and TCDD-treated organs. Based on visual inspection of scanned membranes, only positive clones from both libraries for each organ showing a clear difference in hybridization were individually transferred with sterile toothpicks to new 96-well master microplates containing 100-µl LB medium with ampicillin (100 µg/ml) per well. These colonies were re-grown overnight at 37°C with constant agitation at 150 rpm. After overnight growth, all isolated stock colonies were stored in 60% glycerol at 80°C.
cDNA inserts from re-grown positive clones were PCR-amplified with reagents provided in the PCR-Select Differential Screening Kit (BD Biosciences). After PCR amplification, products were resolved on 1.2% agarose gels to ensure relative cDNA quality and quantity. Polymerase chain reaction products were then treated with Exo1 and SAP and stored at 20°C until sequencing. A total of 780 subtracted cDNAs were sequenced by the Duke University Center for Genome Technology (Table 1). Individual sequence reads were compared to National Center for Biotechnology Information (NCBI) nucleotide and protein databases using the BLASTn and BLASTx programs. Sequence reads with E < 0.05 were assigned putative identities; all others were assumed to have no significant match with known sequences in NCBI databases. An E-value threshold of 0.05 was chosen because the probability of detecting a significant high-scoring sequence pair at the 95% confidence level is P = 1 eE (Karlin and Altschul, 1990). Thus, p = 0.049 when E = 0.05, and p < 0.05 denotes a statistically significant match to known sequences in NCBI databases.
|
|
Statistical analysis of cDNA array data.
Raw spot-density data for each membrane were individually quantified by ArrayGauge v2.1 (FugiFilm), exported as a spreadsheet into Microsoft Excel, and coded by gene, treatment, treatment replicate, membrane replicate, and spot replicate. These hierarchical levels of identification were critical for appropriate statistical analyses, as outlined below. Spot-intensity values on each membrane were normalized by dividing mean intensity values for salmon sperm DNA and Arabidopsis Cab1 on the same membrane. As such, for each membrane, raw intensity values were individually adjusted for nonspecific binding caused by the blocking agent (salmon sperm DNA) and differences in hybridization among membranes (Arabidopsis Cab1). These coded and normalized data were then imported into SAS v9.1 (SAS Institute Inc., Cary, NC), and outliers were detected by means of a ROBUSTREG procedure ( = 0.05). Outliers represented approximately 410% of the data, depending on the array, and were generally the result of random hybridization errors. After outliers were eliminated from the entire data set, normalized values were tested for treatment-specific differences on a gene-by-gene basis with a mixed linear model (
= 0.05) (a MIXED procedure in SAS) and multiple comparison of least-squares means (Wolfinger et al., 2001
). As these data were generated using a nested experimental design, we used a mixed model to account for variability linked with each level of hierarchy. This model handles unbalanced data (resulting from outlier elimination) and considers both fixed- and random-effects parameters associated with known a prioriassigned variables (gene and treatment) and unknown random variables (treatment replicate, membrane replicate, and spot replicate), respectively, assumed to affect the variability of the data. Because this statistical model is conservative, significant treatment differences in array data were reported at the 90% confidence level (p < 0.10).
Real-time RT-PCR.
First-strand cDNAs were generated from the same mRNAs used for cDNA macroarrays. For each 20-µl reaction, total RNA (2 µg) was diluted with RNase-free water to a final volume of 10 µl, and 1 µl oligo(dT)15 (500 µg/ml; Promega) and 1 µl 10 mM dNTPs were mixed with diluted RNA. The mix was heated to 65°C for 5 min and chilled on ice for 2 min. After centrifugation, 4 µl 5x first-strand buffer (Invitrogen), 2 µl 0.1 M DDT, and 1 µl RNase OUT inhibitor (40 U/µl; Invitrogen) were added to each reaction and heated to 37°C. After 2 min incubation, 1 µl Superscript Reverse Transcriptase (200 U/µl; Invitrogen) was added to each reaction and mRNA reverse transcribed at 37°C for 1 h. All reactions were inactivated by incubating at 70°C for 15 min. First-strand cDNAs were stored at 20°C until real-time PCR.
Relative levels of CYP1A, AHR1, and ß-actin transcripts were measured with real-time PCR. The following medaka-specific real-time PCR primers were designed with PrimerQuest (Integrated DNA Technologies): CYP1A (Accession ID: AY297923), forward primer 5'-ACATCGGCCTGAACCGAAATCCTA-3', reverse primer 5'-TGCTTCATTGTGAGCCCGTACTCT-3'; AHR1 (Accession ID: AB065092), forward primer 5'-GTTGTCGGTGAAATTCCCGGCTTT-3', reverse primer 5'- ACTTGAGTCTGCCCTGGATGTTCA-3'; and ß-actin (Accession ID: D89627), forward primer 5'- ACAACGGATCTGGCATGTGCAAAG-3', reverse primer 5'- AGGGCTGTGATCTCCTTCTGCATT-3'. Cytochrome P450 1A, AHR1, and ß-actin cDNAs were PCR-amplified separately in duplicate in a 96-well PCR plate with an ABI PRISM 7000 Sequence Detection System (Applied Biosystems). For each 25-µl real-time PCR reaction, first-strand cDNAs were amplified with 2 µl (200 µg) first-strand cDNA, 9 µl RNase-free water, 0.75 µl 10 µM forward primer (0.3 µM), 0.75 µl 10 µM reverse primer (0.3 µM), and 12.5 µl 2x QuantiTect SYBR Green PCR Master Mix (Qiagen). Real-time PCR reaction conditions were: 95°C for 15 min followed by 44 cycles of 94°C for 15 s, 55°C for 30 s, and 72°C for 1 min. Relative quantitation of gene expression within each reaction was calculated according to the manufacturer's instructions (User Bulletin No. 2, ABI PRISM 7700 Sequence Detection System, Applied Biosystems). For each sample, the threshold cycle for reference (ß-actin) amplification (Ct,ß-actin) was subtracted from the threshold cycle for target amplification (Ct,CYP1A or Ct,AHR1) to yield a Ct. The threshold cycle represents the cycle number at which the fluorescence signal is significantly above the baseline. For each organ, the mean or standard deviation of
Ct for DMSO-treated samples was subtracted from the mean or standard deviation of
Ct for TCDD-treated samples to yield a mean and standard deviation of
Ct for each target and organ. Fold induction relative to DMSO-treated organs and 95% confidence intervals were calculated with 2
Ct.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
For each organ-specific library, redundant clones (those matching the same gene) likely represented higher abundance transcripts. In this study, a total of 98, 258, and 214 sequences from the brain, liver, and testis, respectively, had a significant match (E < 0.05) with known genes based on BLASTn and BLASTx analysis. These 570 sequences from 3 forward and 3 reverse libraries represented 335 total genes (Tables 1, 2, and 3 in the Supplementary Material online). Although we originally picked the same number of colonies (768) for each organ, the total number of genes was different from brain-, liver-, and testis-based libraries: 58, 112, and 165, respectively (Table 1). In part, the frequency of one or more genes within both libraries affected the total gene diversity and number of distinct genes identified within each organ. For example, cytochrome P450 1A (CYP1A) was identified 28 times in the forward library generated from the brain, suggesting that CYP1A transcripts dominated the original RNA pool from TCDD-treated brain (see Table 1 in the Supplementary Material online). Likewise, complement C3-1, transferrin, and UDP-glucoronosyl transferase were identified 23, 20, and 19 times, respectively, in liver-based libraries, thus decreasing the overall gene diversity and number of distinct genes. However, in testis-based libraries, no single transcript dominated the RNA pool, resulting in enhanced detection of many different transcripts that were, in general, similarly responsive to TCDD (see Table 3 in the Supplementary Material online).
Of the 335 total genes identified among all three organs, only 23% of TCDD-responsive genes in all organs were shared between any two organs (Fig. 1). Moreover, just 4 (1.2%) genesCYP1A, ferritin, glyceraldehyde 3-phosphate dehydrogenase, and mitochondrial DNAwere identified in all three organs, suggesting a high degree of organ specificity at the level of gene expression in response to TCDD exposure (Fig. 1). After BLAST identification, genes identified within each organ were annotated by biological process based on Gene Ontology (GO) using LocusLink (http://www.ncbi.nlm.nih.gov/LocusLink/) (Pruitt and Maglott, 2001). Excluding genes with unknown processes, brain, liver, and testis libraries consisted of 16, 16, and 24 different GO terms, respectively (Tables 1, 2, and 3 in the Supplementary Material online). Of the top six GO terms, genes involved in metabolism represented 31% (13 genes) and 32% (26 genes) of genes identified in the brain and liver (Fig. 2a and 2b). Beyond metabolism, GO terms that dominated the brain included organogenesis (8 genes), protein transport (6 genes), unknown processes (6 genes), signal transduction (5 genes), and energy homeostasis (3 genes) (Fig. 2a); and GO terms that dominated the liver included unknown processes (15 genes), signal transduction (14 genes), immune response (12 genes), blood coagulation (9 genes), and proteolysis (6 genes) (Fig. 2b). In the testis, however, 41% (50 genes) of identified genes had no known function (Fig. 2c), with remaining dominant GO terms including proteolysis (16 genes), signal transduction (15 genes), metabolism (14 genes), cell proliferation (13 genes), and cell motility (12 genes).
|
|
Of 42 genes evaluated, liver-derived CYP1A mRNA was the only transcript significantly higher (1.6-fold; p < 0.06) in male medaka brain after a 48-h TCDD exposure (Table 2). Interestingly, although brain-derived CYP1A was also spotted on this array, no significant treatment differences were detected. After sequence alignment of these two cDNAs to full-length medaka CYP1A cDNA (total length = 2349 bp), brain-derived CYP1A cDNA spanned 194735 bp, whereas liver-derived CYP1A cDNA spanned 16622206 bp (data not shown). As the liver-derived CYP1A cDNA sequence was 3'-flanked, and reverse transcriptase is 3'-biased when generating 32P-labeled cDNA, the cDNA pool likely consisted of a higher proportion of 3'-end-radiolabeled CYP1A. Accordingly, this may explain greater hybridization to liver CYP1A spotted on the array and consequent TCDD-induced differences detected with liver-derived CYP1A but not brain-derived CYP1A.
In TCDD-exposed male medaka liver, 12 (29%) of 42 screened transcripts were significantly higher (p < 0.10), and 5 of these transcripts were significant at = 0.05 (Table 2); no significantly responsive transcripts screened were lower in TCDD-exposed liver. As brain-derived and liver-derived CYP1A mRNAs were 1.6- and 2.5-fold higher, respectively (Table 2), differences between these two CYP1A are likely attributable to 3'-biased reverse transcriptase as mentioned above. Interestingly, all statistically significant TCDD-responsive transcripts exhibited treatment:control ratios
1.19-fold, suggesting that conventional 1.5- or 2-fold threshold designations are not appropriate for array data in this study. Indeed, an arbitrary 2-fold threshold would have excluded 92% (11/12) of the significant responses in the liver. Five of 12 differentially expressed genes were originally identified in either brain or testis, and 15 of 22 liver-derived genes spotted onto this array were not significantly responsive (Table 2). This suggests that (1) suppressive subtraction hybridization failed to enrich for other TCDD-responsive liver genes identified in the brain or testis, and/or (2) signal detection on these arrays is limited.
In TCDD-exposed medaka testis, 34 (81%) of 42 screened transcripts were significantly lower (p < 0.10); 17 of 34 were significant at = 0.05, and 11 of 34 were significant at
= 0.01 (Table 2). Significantly higher levels of transcripts were not detected in TCDD-exposed testis, and no significant differences were detected with brain- or liver-derived CYP1A. In addition, 24 of 34 differentially expressed genes were originally identified in either brain or liver (Table 2), suggesting, as in the liver, that our subtracted testis cDNA libraries may not include additional genes potentially responsive to TCDD. Similar to liver, all significant TCDD-responsive transcripts exhibited treatment:control ratios
0.843, likewise suggesting that conventional 1.5- or 2-fold threshold designations are not appropriate for array data in this study.
Real-Time RT-PCR Analysis of CYP1A and AHR1 in TCDD-Exposed Medaka Brain, Liver, and Testis
Real-time RT-PCR was used for relative quantitation of CYP1A and AHR1 transcripts in the same RNA used for cDNA macroarray hybridizations. Relative to DMSO-treated fish, mean CYP1A transcript levels were roughly 198-, 15-, and 1.5-fold higher in brain, liver, and testis, respectively, at 48 h after TCDD exposure (Table 3). Thus, there were significant organ-dependent differences in CYP1A transcript levels at this sampling time and dose. Alternatively, compared to CYP1A, AHR1 was not significantly induced in TCDD-treated organs at 48 h, and it did not exceed levels >2.0-fold in any of the organs (Table 3).
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
For array-based analysis, a separate cohort of 67-month-old male medaka was exposed for 48 h to DMSO or TCDD, as was initially done for subtraction library generation. Array results showed that expression of some of the SSH-identified transcripts was significantly different between DMSO-treated and TCDD-treated brain, liver, or testis; however, within each organ, not all transcripts originally isolated from the respective organ were significantly responsive, suggesting that these arrays exhibited low sensitivity. Moreover, some significant differentially expressed transcripts within the liver and testis were not originally identified in that same organ, suggesting that SSH initially failed to enrich for some differentially expressed mRNAs. Thus, given these limitations of SSH, larger subtraction libraries need to be screened in order to maximize identification of TCDD-responsive genes.
Analysis of overall array data trends demonstrates that the number and significance of TCDD-responsive genes increased from brain to liver to testis, a finding that corroborates qualitative trends in mRNA library diversity (an increase in the number of unique genes identified) and histopathological changes (an increase in the degree of histological changes observed). In addition, relative to DMSO-treated fish, all significantly different transcript responses screened on these arrays were higher (upregulated) in TCDD-treated brain and liver, whereas all significantly different responses were lower (downregulated) in TCDD-treated testis. Therefore, TCDD-exposed male medaka brain, liver, and testis have significant qualitative and semi-quantitative differences in mRNA-level responsesincluding the direction of response (upregulated or downregulated)with putative associations with histopathological differences. However, as with many current toxicogenomic studies, accurate correlations of TCDD-induced gene expression changes with observed pathologies were not conclusive in this study. Indeed, because RNA was only isolated 48 h after TCDD treatment, gene expression responses at this time point may be primary or secondary to histopathological changes.
In this study, we detected significantly higher CYP1A mRNA levels in TCDD-treated adult male medaka brain and liver based on cDNA array analysis and real-time PCR. Based on array data, liver CYP1A mRNAs were roughly 1.5-times higher than in brain, as was similarly observed in CYPLucR+/ mice treated with 3.2 µg TCDD/kg tissue for 24 h (Galijatovic et al., 2004) and Sprague-Dawley rats treated with 10 µg-TCDD/kg tissue for 28 days (Huang et al., 2000
). However, based on real-time RT-PCR, relative brain CYP1A mRNAs were over 13 times higher than in liver, indicating that real-time RT-PCR was significantly more sensitive than arrays for detecting differences in CYP1A mRNA among organs. Although we did not localize CYP1A mRNA or protein on tissue sections, high CYP1A transcript levels in the brain are likely due to induction within the vascular endothelium rather than the parenchyma, as was reported in TCDD-exposed zebrafish embryos (Dong et al., 2002
). Interestingly, while CYP1A transcripts were induced 197-fold in the brain based on real-time RT-PCR, we observed no gross histopathological changes at 48 h. Likewise, while CYP1A mRNA and protein were strongly induced in all major brain regions (telencephlalon, diencephalon, mesencephalon, rhombencephalon, cerebellum, and pituitary) of gilthead seabream (Sparus aurata) after a 2-day static exposure to 6 pg TCDD/l seawater (Ortiz-Delgado et al. 2002
), no gross histopathological changes were reported.
Because we did not identify AHR or ARNT transcripts in brain, liver, or testis subtraction libraries, AHR and ARNT mRNA were not measured by array analysis in this study. However, we measured AHR1 expression by real-time RT-PCR in the same samples used for array analysis, and we found that, compared to CYP1A, AHR1 was not significantly induced in any organ. Likewise, Huang et al. (2000) reported that AHR and ARNT mRNA levels in TCDD-treated rat liver were not significantly different from vehicle controls, and Tanguay et al. (2000)
reported that TCDD did not induce dioxin responsive reporter gene expression in COS-7 cells expressing zebrafish ARNT2b, or ARNT2c. Thus, because of minimal induction after TCDD exposure, SSH was not sensitive enough to enrich for AHR or ARNT isoforms in this study.
Based on cDNA array analysis of TCDD-exposed liver mRNA, we detected significantly higher transcripts associated with drug/toxicant metabolism (CYP1A, UDP-glucuronosyl transferase (UDPGT)), immune response (hepcidin precursor, MHC class I), lipid transport (apolipoprotein C-I), complement activation (complement C2), carbohydrate metabolism (-amylase), mitosis (c-myc binding protein), microtubule movement (
-tubulin), and sex differentiation (male sex-determining protein). Cytochrome P450 1A and UDPGT are well-established target genes for AHR-mediated transcription (Nebert, 2000). Thus, significantly higher levels of these transcripts detected in this study supported liver exposure to TCDD, whereas remaining differences (excluding male sex-determining protein) were likely related to acute-phase responses such as host defense and inflammation. Even at 48 h, we observed glycogen depletion in the liver. Similar histopathological observations were reported in adult female rainbow trout (Oncorhynchus mykiss) liver after exposure to TCDD-spiked food (90 ng-TCDD/kg-food) for 100 days (Walter et al., 2000
) and in adult male zebrafish (Danio rerio) liver after exposure to 70 µg TCDD/g tissue for 5 days (Zodrow et al., 2004
). Glycogen serves as a major glucose reserve in fish liver, and glycogen mobilization to the plasma occurs during high-energy demand situations such as starvation or stress (Hinton et al., 2001
). In this study, mRNA levels of
-amylasean enzyme involved in glycogen metabolism in fish liver (Murat, 1976
)were significantly higher in TCDD-exposed livers, suggesting that
-amylase played a role in glycogen mobilization and depletion in response to increased xenobiotic-related metabolic demand. Contrary to the observations in TCDD-exposed zebrafish (Zodrow et al., 2004
), we did not observe histopathological signs of lipidosis in the liver; however, the nominal dose used in the present study was 7 times less (70 µg/kg vs. 10 µg/kg) and exposure time was less than half the duration (120 h vs. 48 h). Thus, a 10 µg/kg i.p.-injected exposure for 48 h was not sufficient to induce the degree of lipid accumulation commonly observed in xenobiotic-exposed fish livers (Hinton, 2001).
To our knowledge, this is the first report of hepcidin and TBT-binding protein mRNA in medaka liver, and the only published demonstration of differential expression of these two genes after TCDD exposure. Hepcidin is a novel liver-derived antimicrobial peptide originally discovered in urine from human donors (Park et al., 2001). This 2025-amino acid peptide has since been identified in fish including hybrid striped bass (Morone saxatilis x M. chrysops) (Shike et al., 2002
), winter flounder (Pseudopleuronectes americanus) (Douglas et al., 2003
), Atlantic salmon (Salmo salar) (Douglas et al., 2003
), and zebrafish (Danio rerio) (Shike et al., 2004
), and it is highly expressed in white bass (M. chrysops) liver after bacterial challenge (Shike et al., 2002
). Although the precise role of hepcidin in chemically induced liver toxicity is not established, hepcidin is known to be a key regulator of iron absorption and a mediator of inflammation (Ganz, 2003
). In fish and mammals, iron metabolism and transport in liver is a well-known component of an acute phase response to toxicant exposure. In this study, hepcidin was likely involved in immune- and/or inflammation-related pathways in TCDD-exposed medaka liver. Significantly higher mRNA levels of tributyltin (TBT)-binding protein were also detected in TCDD-exposed liver, but the precise function of this protein is also unknown. Tributyltin-binding protein is a serum-derived protein identified from TBT-exposed Japanese flounder (Paralichthys olivaceus), and it is thought to play a role in xenobiotic transport throughout the fish vascular system (Shimasaki et al., 2002
).
Based on qualitative evaluation of histopathology and semi-quantitative cDNA array analysis, the most significant differences observed at 48 h were in TCDD-exposed male medaka testis. Morphological differences included disorganization of spermatogenesis at the testis periphery, disruption of the interstitium, Leydig cell swelling, and Sertoli cell vacuolation. Similar adverse effects in the germinal epithelium have also been observed in rat testes 7 days post-exposure to a single 3 µg/kg dose of TCDD (Rune et al., 1991). Based on array data, significantly affected transcripts isolated from TCDD-treated whole medaka testes were decreased more than 1.2-fold and were associated with 21 known biological processes, suggesting that overall gene transcription in the testis was suppressed at 48 h after a single TCDD injection; however, significant differences in CYP1A and AHR1 mRNA were not detected between control and TCDD-treated male medaka testis based on arrays (CYP1A) and real-time PCR (CYP1A and AHR1), findings similar to those observed in adult male rats (Roman et al., 1998
). Thus, contrary to the liver findings, these data suggest that TCDD-induced gene expression responses and gross histopathological changes are putatively CYP1A-independent; however, based on conventional RT-PCR, AHR1 mRNA is strongly expressed in DMSO- and TCDD-treated medaka testis (data not shown). Rather, acute testicular toxicity at 48 h may be due to induction of reactive oxygen species (ROS) and inhibition of antioxidant enzymes. For example, administration of 10 µg TCDD/kg body weight to adult rats for 4 days significantly increased hydrogen peroxide levels and lipid peroxidation, and it significantly decreased superoxide dismutase, catalase, glutathione reductase, and glutathione peroxidase in epididymal sperm (Latchoumycandane et al., 2003
). Although our study focused only on acute effects, and although medaka testes were likely able to recover after a single dose of TCDD, long-term TCDD-induced ROS generation may lead to apoptosis in a variety of cell types within the testis. Based on recent studies using terminal dideoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assays, adult male medaka chronically exposed to 100 µg/l nonylphenol or 0.01 µg/l 17
-ethinylestradiol resulted in increased apoptosis of spermatocytes, Leydig cells, and Sertoli cells (Weber et al., 2002
, 2004
). Likewise, long-term TCDD exposure may lead to cell death in Sertoli cells, Leydig cells, and/or immature germ cells of adult medaka. Moreover, persistent TCDD-induced toxicity to these cell types (especially during development) may lead to decreased mature spermatozoa numbers and male infertility. For instance, low TCDD doses (
1.0 µg TCDD/kg tissue) during various stages of male rat development significantly decreased sperm production by more than half once sexual maturity was reached (Gray et al., 1995
; Mably et al., 1992
).
Of those genes significantly responsive to TCDD, two genesmale sex-determining protein and protamineare critical for testicular development and spermatogenesis in male medaka. In medaka, male sex-determining protein is a Y-linked DM-domaincontaining gene required for testicular development (Matsuda et al., 2002), and transcripts are normally expressed in Sertoli cells of adult male testis (Nanda et al., 2002
). Because heritable mutations of male sex-determining protein in medaka result in XY female offspring (Matsuda et al., 2002
), TCDD-induced decreases of this protein in the testis may have the potential to disrupt spermatogenesis or induce long-term feminizing effects. In addition, male sex-determining protein transcripts were significantly higher in TCDD-exposed medaka liver, as has been reported with the mammalian ortholog SRY in TCDD-exposed HepG2 cells (Adachi et al., 2004
); however, at present the role of liver-derived sex-determining protein in medaka is unknown. Protamine is an arginine-rich, DNA-binding protein that replaces histones in maturing spermatids during nuclear condensation (Saiki et al., 1997
), and heritable mutations in the protamine gene disrupt nuclear formation, spermatogenesis, and infertility in male mice (Cho et al., 2001
). Thus, TCDD-induced decreases in protamine may have long-term effects on normal sperm maturation in male medaka.
Taken together, these findings demonstrate that TCDD induces organ-specific differential gene expression in adult male medaka, and these responses are generally associated with histopathological changes. After a 48-h 10 µg TCDD/kg tissue exposure, adverse effects at the mRNA and histological levels increased from brain to liver to testis. In addition, we identified unique mRNA-level biomarkers of TCDD-induced hepatic and testicular toxicity in medaka that may be further pursued by applying focused monitoring approaches such as real-time PCR or in situ hybridization, as well as functional approaches such as drug-based inhibition or RNAi.
![]() |
NOTES |
---|
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Cantrell, S. M., Joy-Schlezinger, J., Stegeman, J. J., Tillitt, D. E., and Hannink, M. (1998). Correlation of 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced apoptotic cell death in the embryonic vasculature with embryotoxicity. Toxicol. Appl. Pharmacol. 148, 2434.[CrossRef][ISI][Medline]
Cho, C., Willis, W. D., Goulding, E. H., Jung-Ha, H., Choi, Y. C., Hecht, N. B., and Eddy, E. M. (2001). Haploinsufficiency of protamine-1 or -2 causes infertility in mice. Nat. Genet. 28, 8286.[CrossRef][ISI][Medline]
Denslow, N. D., Kocerha, J., Sepulveda, M. S., Gross, T., and Holm, S. E. (2004). Gene expression fingerprints of largemouth bass (Micropterus salmoides) exposed to pulp and paper mill effluents. Mutat. Res. 552, 1934.[ISI][Medline]
DeVito, M. J., and Birnbaum, L. S. (1994). Toxicology of dioxins and related chemicals. In Dioxins and Health (A. Schecter, Ed.), pp. 139162. Plenum Press, New York.
Diatchenko, L., Lau, Y. F., Campbell, A. P., Chenchik, A., Moqadam, F., Huang, B., Lukyanov, S., Lukyanov, K., Gurskaya, N, Sverdlov, E. D., and Siebert, P. D. (1996). Suppression subtractive hybridization: A method for generating differentially regulated or tissue-specific cDNA probes and libraries. Proc. Natl. Acad. Sci. U.S.A. 93, 60256030.
Dong, W., Teraoka, H., Yamazaki, K., Tsukiyama, S., Imani, S., Imagawa, T., Stegeman, J. J., Peterson, R. E., and Hiraga, T. (2002). 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity in the zebrafish embryo: Local circulation failure in the dorsal midbrain is associated with increased apoptosis. Toxicol. Sci. 69, 191201.
Dong, W., Teraoka, H., Tsujimoto, Y., Stegeman, J. J., and Hiraga, T. (2004). Role of aryl hydrocarbon receptor in mesencephalic circulation failure and apoptosis in zebrafish embryos exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol. Sci. 77, 109116.
Douglas, S. E., Gallant, J. W., Liebscher, R. S., Dacanay, A., and Tsoi, S. C. M. (2003). Identification and expression analysis of hepcidin-like antimicrobial peptides in bony fish. Dev. Comp. Immunol. 27, 589601.[CrossRef][ISI][Medline]
Frueh, F. W., Hayashibara, K. C., Brown, P. O., and Whitlock, J. P., Jr. (2001). Use of cDNA microarrays to analyze dioxin-induced changes in human liver gene expression. Toxicol. Lett. 122, 189203.[CrossRef][ISI][Medline]
Galijatovic, A., Beaton, D., Nguyen, N., Chen, S., Bonzo, J., Johnson, R., Meada, S., Karin, M., Guengerich, F. P., and Tukey, R. H. (2004). The human CYP1A1 gene is regulated in a developmental and tissue-specific fashion in transgenic mice. J. Biol. Chem. 279, 2396923976.
Ganz, T. (2003). Hepcidin, a key regulator of iron metabolism and mediator of anemia of inflammation. Blood 102, 783788.
Gray, L. E., Jr., Kelce, W. R., Monosson, E., Ostby, J. S., and Birnbaum, L. S. (1995). Exposure to TCDD during development permanently alters reproductive function in male Long Evans rats and hamsters: Reduced ejaculated and epididymal sperm numbers and sex accessory gland weights in offspring with normal androgenic status. Toxicol. Appl. Pharmacol. 131, 108118.[CrossRef][ISI][Medline]
Hahn, M. E. (2001). Dioxin toxicology and the aryl hydrocarbon receptor: Insights from fish and other non-traditional models. Mar. Biotechnol. 3, S224S238.[CrossRef][ISI][Medline]
Hinton, D. E., Segner, H., and Braunbeck, T. (2001). Toxic responses of the liver. In Target Organ Toxicity in Marine and Freshwater Teleosts (D. Schlenk and W.H. Benson, Eds.), Vol. 1Organs, pp. 224268. Taylor and Francis, London.
Hoffer, A., Chang, C.-Y., and Puga, A. (1996). Dioxin induces transcription of fos and jun genes by Ah receptor-dependent and -independent pathways. Toxicol. Appl. Pharmacol. 141, 238247.[CrossRef][ISI][Medline]
Huang, P., Rannug, A., Ahlbom, E., Håkansson, and Ceccatelli, S. (2000). Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on the expression of cytochrome P450 1A1, the aryl hydrocarbon receptor, and the aryl hydrocarbon receptor nuclear translocator in rat brain and pituitary. Toxicol. Appl. Pharmacol. 169, 159167.[CrossRef][ISI][Medline]
Ji, W., Wright, M. B., Cai, L., Flament, A., and Lindpaintner, K. (2002). Efficacy of SSH PCR in isolating differentially expressed genes. BMC Genomics 3, 1218.[CrossRef][Medline]
Jokinen, M. P., Walker, N. J., Brix, A. E., Sells, D. M., Haseman, J. K., and Nyska, A. (2003). Increase in cardiovascular pathology in female Sprague-Dawley rats following chronic treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin and 3,3',4,4',5-pentachlorobiphenyl. Cardiovasc. Toxicol. 3, 299310.[CrossRef][Medline]
Kamath, A. B., Xu, H., Nagarkatti, P. S., and Nagarkatti, M. (1997). Evidence for the induction of apoptosis in thymocytes by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol. Appl. Pharmacol. 142, 367377.[CrossRef][ISI][Medline]
Karlin, S., and Altschul, S. F. (1990). Methods for assessing the statistical significance of molecular sequence features by using general scoring schemes. Proc. Natl. Acad. Sci. U.S.A. 87, 22642268.
Kirchen, R. V., and West, W. R. (1976). The Japanese medaka. Its care and development. Carolina Biological Supply, Burlington, NC, 39 pp.
Kurachi, M., Hashimoto, S., Obata, A., Nagai, S., Nagahata, T., Inadera, H., Sone, H., Tohyama, C., Kaneko, S., Kobayashi, K., and Matsushima, K. (2002). Identification of 2,3,7,8- tetrachlorodibenzo-p-dioxin in mouse liver by serial analysis of gene expression. Biochem. Biophys. Res. Commun. 292, 368377.[CrossRef][ISI][Medline]
Larkin, P., Folmar, L. C., Hemmer, M. J., Poston, A. J., and Denslow, N. D. (2003). Expression profiling of estrogenic compounds using a sheepshead minnow cDNA macroarray. EHP Toxicogenomics 111, 2936.[Medline]
Law, J. M. (2003). Issues related to the use of fish models in toxicologic pathology: Session introduction. Toxicol. Pathol. 31, 4952.[Medline]
Latchoumycandane, C., Chitra, K. C., and Mathur, P. P. (2003). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) induces oxidative stress in the epididymis and epididymal sperm of adult rats. Arch. Toxicol. 77, 280284.[ISI][Medline]
Mably, T. A., Bjerke, D. L., Moore, R. W., Gendron-Fitzpatrick, A., and Peterson, R. E. (1992). In utero and lactational exposure of male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin. 3. Effects on spermatogenesis and reproductive capability. Toxicol. Appl. Pharmacol. 114, 118126.[CrossRef][ISI][Medline]
Marlowe, J. L., Knudsen, E. S., Schwemberger, S., and Puga, A. (2004). The aryl hydrocarbon receptor displaces p300 from E2F-dependent promoters and represses S phase-specific gene expression. J. Biol. Chem. 279, 2901329022.
Matsuda, M., Nagahama, Y., Shinomiya, A., Sato, T., Matsuda, C., Kobayashi, T., Morrey, C. E., Shibata, N., Asakawa, S., Shimizu, N., Hori, H., Hamaguchi, S., and Sakaizumi, M. (2002). DMY is a Y-specific DM-domain gene required for male development in the medaka fish. Nature 417, 559563.[CrossRef][ISI][Medline]
Matsumura, F. (2003). On the significance of the role of cellular stress response reactions in the toxic actions of dioxin. Biochem. Pharmacol. 66, 527540.[CrossRef][ISI][Medline]
Murat, J. C. (1976). Studies on glycogenolysis in carp liver: evidence for an amylase pathway for glycogen breakdown. Comp. Biochem. Physiol. B 55, 461465.[CrossRef][ISI][Medline]
Nanda, I., Kondo, M., Hornung, U., Asakawa, S., Winkler, C., Shimizu, A., Shan, Z., Haaf, T., Shimizu, N., Shima, A., Schmid, M., and Schartl, M. (2002). A duplicated copy of DMRT1 in the sex-determining region of the Y chromosome of the medaka, Oryzias latipes. Proc. Natl. Acad. Sci. U.S.A. 99, 1177811783.
Nebert, D. W., Roe, A. L., Dieter, M. Z., Solis, W. A., Yang, Y., and Dalton, T. P. (2000). Role of the aromatic hydrocarbon receptor and [Ah] gene battery in the oxidative stress response, cell cycle control, and apoptosis. Biochem. Pharmacol. 59, 6585.[CrossRef][ISI][Medline]
Oleksiak, M. F., Churchill, G. A., and Crawford, D. L. (2002). Variation in gene expression within and among natural populations. Nat. Genet. 32, 261266.[CrossRef][ISI][Medline]
Ortiz-Delgado, J. B., Sarasquete, C., Behrens, A., González de Canales, M. L., and Segner, H. (2002). Expression, cellular distribution and induction of cytochrome P4501A (CYP1A) in gilthead seabream, Sparus aurata, brain. Aquat. Toxicol. 60, 269283.[CrossRef][ISI][Medline]
Park, C. H., Valore, E. V., Waring, A. J., and Ganz, T. (2001). Hepcidin, a urinary antimicrobial peptide synthesized in the liver. J. Biol. Chem. 276, 78067810.
Puga, A., Maier, A., and Medvedovic, M. (2000). The transcriptional signature of dioxin in human hepatoma HepG2 cells. Biochem. Pharmacol. 60, 11291142.[CrossRef][ISI][Medline]
Pruitt, K. D., and Maglott, D. R. (2001). RefSeq and LocusLink: NCBI gene-centered resources. Nucleic Acids Res. 29, 137140.
Roman, B. L., Pollenz, R. S., and Peterson, R. E. (1998). Responsiveness of the adult male rat reproductive tract to 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure: Ah receptor and ARNT expression, CYP1A1 induction, and Ah receptor down-regulation. Toxicol. Appl. Pharmacol. 150, 228239.[CrossRef][ISI][Medline]
Rune, G. M., de Souza, P., Krowke, R., Merker, H. J., and Neubert, D. (1991). Morphological and histochemical pattern of response in rat testes after administration of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Histol. Histopathol. 6, 459476.[ISI][Medline]
Saiki, A., Tamura, M., Matsumoto, M., Katowgi, J., Watanabe, A., and Onitake, K. (1997). Establishment of in vitro spermatogenesis from spermatocytes in the medaka, Oryzias latipes. Dev. Growth. Differ. 39, 33744.[CrossRef][ISI][Medline]
Senft, A. P., Dalton, T. P., Nebert, D. W., Genter, M.D., Hutchinson, R. J., and Shertzer, H. G. (2002). Dioxin increases reactive oxygen production in mouse liver mitochondria. Toxicol. Appl. Pharmacol. 178, 1521.[CrossRef][ISI][Medline]
Shike, H., Lauth, X., Westerman, M. E., Ostland, V. E., Carlberg, J. M., Van Olst, J. C., Shimizu, C., Bulet, P., and Burns, J. C. (2002). Bass hepcidin is a novel antimicrobial peptide induced by bacterial challenge. Eur. J. Biochem. 269, 22322237.
Shike, H., Shimizu, C., Lauth, X., and Burns, J. C. (2004). Organization and expression analysis of the zebrafish hepcidin gene, an antimicrobial peptide gene conserved among vertebrates. Dev. Comp. Immunol. 28, 747754.[CrossRef][ISI][Medline]
Shimasaki, Y., Oshima, Y., Yokota, Y., Kitano, T., Nakao, M., Kawabata, S., Imada, N., and Honjo, T. (2002). Purification and identification of a tributyltin-binding protein from serum of Japanese flounder, Paralichthys olivaceus. Environ. Toxicol. Chem. 21, 12291235.[CrossRef][ISI][Medline]
Tanguay, R. L., Andreasen, E., Heideman, W., and Peterson, R. E. (2000). Identification and expression of alternatively spliced aryl hydrocarbon nuclear translocator 2 (ARNT2) cDNAs from zebrafish with distinct functions. Biochim. Biophys. Acta 1494, 117128.[ISI][Medline]
Toomey, B. H., Bello, S., Hahn, M. E., Cantrell, S., Wright, P., Tillitt, D. E., and Di Giulio, R. T. (2001). 2,3,7,8-Tetrachlorodibenzo-p-dioxin induces apoptotic cell death and cytochrome P4501A expression in developing Fundulus heteroclitus embryos. Aquat. Toxicol. 53, 127138.[CrossRef][ISI][Medline]
Walter, G. L., Jones, P. D., and Giesy, J. P. (2000). Pathologic alterations in adult rainbow trout, Oncorhychus mykiss, exposed to dietary 2,3,7,8- tetrachlorodibenzo-p-dioxin. Aquat. Toxicol. 50, 287299.[CrossRef][ISI][Medline]
Weber, L. P., Kiparissi, Y., Hwang, G. S., Niimi, A. J., Janz, D. M., and Metcalfe, C. D. (2002). Increased cellular apoptosis after chronic aqueous exposure to nonylphenol and quercetin in adult medaka (Oryzias latipes). Comp. Biochem. Physiol. Part C 131, 5159.[CrossRef]
Weber, L. P., Balch, G. C., Metcalfe, C. D., and Janz, D. M. (2004). Increased kidney, liver, and testicular cell death after chronic exposure to 17-ethinylestradiol in medaka (Oryzias latipes). Environ. Toxicol. Chem. 23, 792797.[CrossRef][ISI][Medline]
Wittbrodt, J., Shima, A., and Schartl, M. (2002). MedakaA model organism from the Far East. Nat. Rev. Genet. 3, 5364.[CrossRef][ISI][Medline]
Williams, T. D., Gensberg, K., Minchin, S. D., and Chipman, J. K. (2003). A DNA expression array to detect toxic stress response in European flounder (Platichthys flesus). Aquat. Toxicol. 65, 141157.[CrossRef][ISI][Medline]
Wolfinger, R. D., Gibson, G., Wolfinger, E. D., Bennett, L., Hamadeh, H., Bushel, P., Afshari, C., and Paules, R. S. (2001). Assessing gene significance from cDNA microarray expression data via mixed models. J. Comput. Biol. 8, 625637.[CrossRef][ISI][Medline]
Zodrow, J. M., Stegeman, J. J., and Tanguay, R. L. (2004). Histological analysis of acute toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in zebrafish. Aquat. Toxicol. 66, 2538.[CrossRef][ISI][Medline]
Zook, D. R., and Rappe, C. (1994). Environmental sources, distribution, and fate of polychlorinated dibenzodioxins, dibenzofurans, and related organochlorines. In Dioxins and Health (A. Schecter, Ed.), pp. 79113. Plenum Press, New York.