Physiol. Genomics Add DOIs to your references at manuscript stage!
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 33: 148-158, 2008. First published January 29, 2008; doi:10.1152/physiolgenomics.00214.2007
1094-8341/08 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow All Versions of this Article:
33/2/148    most recent
00214.2007v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, J.
Right arrow Articles by Heideman, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, J.
Right arrow Articles by Heideman, W.
Received 13 September 2007; accepted in final form 25 January 2008.
Physiological Genomics 33:148-158 (2008)
1094-8341/08 $8.00 © 2008 American Physiological Society

Comparative genomics identifies genes mediating cardiotoxicity in the embryonic zebrafish heart

Jing Chen1, Sara A. Carney2, Richard E. Peterson1,2 and Warren Heideman1,2

1 Pharmaceutical Sciences Division, School of Pharmacy
2 Molecular and Environmental Toxicology Center, University of Wisconsin, Madison, Wisconsin


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Retinoic acid (RA) and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) activate distinct ligand-dependent transcription factors, and both cause cardiac malformation and heart failure in zebrafish embryos. We hypothesized that they cause this response by hyperactivating a common set of genes critical for heart development. To test this, we used microarrays to measure transcript changes in hearts isolated from zebrafish embryos 1, 2, 4, and 12 h after exposure to 1 µM RA. We used hierarchical clustering to compare the transcriptional responses produced in the embryonic heart by RA and TCDD. We could identify no early responses in common between the two agents. However, at 12 h both treatments produced a dramatic downregulation of a common cluster of cell cycle progression genes, which we term the cell cycle gene cluster. This was associated with a halt in heart growth. These results suggest that RA and TCDD ultimately trigger a common transcriptional response associated with heart failure, but not through the direct activation of a common set of genes. Among the genes rapidly induced by RA was Nr2F5, a member of the COUP-TF family of transcriptional repressors. We found that induction of Nr2F5 was both necessary and sufficient for the cardiotoxic response to RA.

retinoic acid; 2,3,7,8-tetrachlorodibenzo-p-dioxin; microarray


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
ALL-TRANS RETINOIC ACID (RA) is an active derivative of vitamin A that plays an important role in cardiac development. Abnormal levels of RA activity, in either excess or deficiency, cause cardiac malformation in developing vertebrates.

Excess RA alters heart formation in mouse (8), Xenopus (12), and zebrafish (30) embryos. Depletion of RA causes altered heart development in the mouse (23, 38), while BMS493, an RA receptor antagonist, also compromises mouse heart development (8). The nof mutation, blocking RA synthesis, causes severe pericardial edema in zebrafish (16). These results and others indicate that RA plays an important role in vertebrate heart development.

RA is an agonist for the retinoic acid receptor (RAR), a DNA binding nuclear hormone receptor with three isoforms, {alpha}, β, and {gamma} (4). Unliganded RAR forms a heterodimer with retinoid X receptor (RXR) and binds to DNA sequence elements referred to as retinoic acid response elements (RAREs) where the heterodimer inhibits transcription. Agonist binding to the RAR alters the complex, leading to recruitment of coactivators and active transcription of target genes (3).

We have found (1, 34) that 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) also causes cardiovascular abnormalities in developing zebrafish embryos. TCDD produces pericardial edema, reduced heart chamber size, and heart failure in the zebrafish embryo (1).

As with RA, the effects of TCDD are mediated by a ligand-activated transcription factor, the aryl hydrocarbon receptor (AHR). Agonists such as TCDD bind to AHR, causing it to form a heterodimer with the related basic helix-loop-helix PAS protein, ARNT. The heterodimer then binds to DNA sequence motifs known as dioxin response elements to initiate transcription of target genes (11, 29).

One of the most important obstacles in understanding the mechanism by which AHR agonists such as TCDD affect the development of the heart has been identification of the genes targeted by the activated AHR/ARNT complex. In recent work we (6) developed methods for measuring changes in mRNA levels in embryonic hearts exposed to TCDD with time course microarray experiments. These experiments revealed a very rapid induction of putative AHR target genes within 1–4 h after TCDD exposure, followed by a striking downregulation of a large set of genes involved in cell cycle progression. This latter response coincides with heart failure and a halt in cardiomyocyte proliferation.

TCDD and RA exposure both produce malformations in embryonic hearts, presumably through the activation of their respective receptors. In both cases the receptors directly activate gene expression, but these receptors bind to distinct DNA sequence motifs. Whether or not there are important genes containing binding sites for both receptors, allowing regulation by both RA and TCDD, is not known. This raises a question: Do RA and TCDD activate a common set of genes that lead to heart malformation and failure? The recent development of techniques for global analysis of gene expression changes in hundreds of intact hearts isolated from synchronously developing embryos (5, 6) allows us now to address the hypothesis that TCDD and RA affect the developing heart through a common set of target genes.

We were also interested in the possibility that the hypothesis would be incorrect, that RA and TCDD would produce similar signs of cardiotoxicity without activating a common set of genes. In this case, we hypothesized that the progressive heart failure produced by the two agents would nonetheless be associated with a characteristic set of gene expression changes, diagnostic for the condition of the damaged hearts.

Finally, while the effects of RA on heart development have been widely studied, we do not yet know which gene targets lead to cardiotoxic responses in the embryonic heart. The use of microarray experiments offers the chance of identifying such gene targets activated by RAR/RXR.

Here we report the pattern of gene expression changes in embryonic zebrafish hearts following exposure to excess RA compared with the previously identified changes produced by TCDD. We found little overlap in early gene expression changes induced by RA and TCDD, indicating that RA and TCDD initiate cardiotoxicity through distinct mechanisms. However, we found that, as time progressed, the heart failure produced by either agent was preceded by a common set of gene expression changes. This was characterized by the marked downregulation of a set of genes associated with DNA replication and cell division. We found that the transcript with the strongest induction in response to RA exposure was Nr2F5, a member of the chicken ovalbumin upstream promoter-transcription factor (COUP-TF) family of nuclear receptors. Overexpression of Nr2F5 phenocopied cardiac defects produced by RA exposure, while blockade of Nr2F5 expression was sufficient to block RA-induced cardiotoxicity.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Zebrafish embryos.
All zebrafish embryos were kept at 27°C in egg water (60 µg/ml Instant Ocean Salts; Aquarium Systems, Mentor, OH) with a 14:10-h light-dark cycle. Cmcl2::GFP embryos were used for heart extraction for microarray analysis (5). Albino embryos were used for imaging in the experiments to assay cardiovascular function. Cmcl2::dsRed2-nuc embryos raised in 0.003% phenylthiourea egg water were used for imaging in the experiment to assay cardiac myocyte number (1, 21).

All-trans retinoic acid and vehicle exposure.
All-trans RA (≥98% purity, Sigma-Aldrich, St. Louis, MO) was kept at –80°C as a stock solution at 0.01 M in 100% DMSO. Before each experiment, RA was diluted to 1 µM/0.1% DMSO in egg water for embryo exposure; embryos were exposed to 0.1% DMSO as vehicle controls. RA solution was renewed every 24 h. For morpholino (MO) experiments, RA stock solution was diluted to 0.1 µM in 0.1% DMSO. All manipulations of RA were performed in subdued light.

Experimental design.
For assays of stroke volume, cardiac output, and heart rate, six replicate experiments of 5 zebrafish embryos each were used per treatment group. For assays of peripheral blood flow, pericardial edema, and cardiac myocyte number, five replicate experiments of 5 zebrafish embryos each were used. For microarray analysis, three arrays were used for each sample time point, with each array using RNA samples collected from independently conducted experiments. Each array replicate used RNA collected from ~500 hearts, collected in six or seven blocks. For each block, 200 exposed embryos were used for heart extraction. The extraction period, including the time spent collecting hearts under the dissecting microscope, was limited to 30 min, yielding 80–100 hearts per extraction session. Real-time quantitative PCR analysis used triplicate RNA samples collected from independent heart extractions.

Heart rate, stroke volume, and cardiac output.
Embryos were mounted in 5% methylcellulose, and video images were captured with a MotionScope camera mounted on a Nikon TE300 inverted scope with 250 frames/s. Frames containing ventricular end-diastolic and end-systolic images were identified, and end-diastolic and end-systolic volumes (EDV and ESV) of the ventricles were estimated according to Simpson's method (see Refs. 6, 10, 28); stroke volume = EDV – ESV. Heart rate of each embryo was measured by counting beats per minute on the live video, and cardiac output was calculated as the product of stroke volume x heart rate.

Cardiac myocyte number.
Cmlc2::dsRed2-nuc embryos, expressing Discosoma red fluorescent protein (RFP) in cardiomyocyte nuclei, were raised in 0.003% phenylthiourea egg water to inhibit pigmentation, anesthetized, mounted in Lebovitz's L15 medium-10% fetal bovine serum (FBS) on a glass slide, and covered with a coverslip. Images of the hearts were taken with a Princeton Instruments Micromax charge-coupled device camera mounted on an inverted microscope, and the number of red fluorescent nuclei was counted for each heart image (1, 6, 21).

Peripheral blood flow.
Video images of blood flow in the posterior intersegmental vein in the trunk of either RA- or DMSO-exposed embryos were captured with a Princeton Instruments Micromax charge-coupled device camera mounted on an inverted microscope. Time-lapse frames lasting 10 s were identified, and the number of red blood cells passing a chosen landmark in the posterior intersegmental vein during 10 s was counted (6, 25).

Pericardial sac area.
Images of embryos were captured on a Leica MZ16 stereomicroscope, and their pericardial areas were outlined and measured with the program Scion Image (Scion, Frederick, MD) in order to quantify the severity of pericardial edema (6, 25).

Heart extraction.
Before heart extraction, ~200 embryos were anesthetized with Tricaine-S (Aquatic EcoSystems, Apopka, FL), collected in a microcentrifuge tube, and immersed in 1 ml of Lebovitz's L-15 medium-10% FBS. Hearts of zebrafish embryos were separated from the bodies by sheer force generated by drawing and repelling the embryos through a 19-gauge syringe needle, followed by size fraction of the disrupted embryos and manual retrieval of the individual green fluorescent protein (GFP)-expressing hearts with a pipette under epifluorescence. The hearts were snap frozen in liquid nitrogen and stored at –80°C for further analysis (8).

Microarray analysis and data processing.
Total RNA (1 µg/500 hearts) was extracted with a Qiagen RNeasy mini kit according to the manufacturer's protocol (Qiagen, Valencia, CA). Biotinylated cRNA was generated with Affymetrix One-cycle Target Labeling and Control Reagents according to the manufacturer's protocol (Affymetrix, Santa Clara, CA), followed by fragmentation and hybridization to Affymetrix Genechip Zebrafish Arrays. Hybridized arrays were stained on an Affymetrix Fluidics Station 400 and scanned in an Agilent Gene Array Scanner. All data have been deposited in the GEO database with accession number GSE9020. All .cel images were preprocessed with the GC-RMA function in ArrayAssist Express version 4.0 software (Stratagene, La Jolla, CA). Briefly, GC-RMA (39) is an array preprocessing function similar to robust multiarray analysis (RMA) (17) but incorporating mismatch probe sets by using a model based on GC content of the probe. The method is composed of three steps: background correction based on the theory that nonspecific binding tends to be directly related to GC content of the probe; quantile normalization; and summarization by the mean polish method. All summarized GC-RMA output signals were then log2 transformed. Transcripts that were altered by twofold or more compared with the control were selected for analysis. The raw .cel data were also preprocessed by Affymetrix Microarray Suite (MAS) 5.0 software in order to obtain information about whether the signal was present (P) or absent (A) for a particular probe set in a given sample. A probe set was determined as "A" if its MAS 5.0 signal was "A" in at least three of six samples from a given time point, unless all three "A"s were confined solely to either the DMSO or the RA group. This decision was made as an attempt to include samples in which one of the treatments brought the transcript levels above background, while excluding samples with questionable expression under both conditions. After removal of absent signals, the transformed GC-RMA signals of all probe sets with twofold changes were input into the TIGR MultiExperiment Viewer (TMEV) software (27) from the Institute for Genomic Research (TIGR), and the probe sets with significant changes were selected by two-class unpaired significant analysis of microarray (SAM) (33) with a ≤10% false discovery rate (FDR). Data from the RA experiment shown in Fig. 5 were organized with the self-organizing map (SOM) program from ArrayAssist Expression software, using transcripts altered significantly by twofold or more by RA exposure (31). Hierarchical clustering (HCL) (14) with euclidean distances and average linkage was used in Fig. 8 to organize the combined results from the RA exposure experiment described in this work and the TCDD experiment from Carney et al. (6). Again, the analysis was limited to transcripts altered significantly by twofold or more.

Real-time PCR.
First-strand cDNA synthesis was performed with total RNA isolated from 100 hearts per replicate with a SuperScript III First-Strand Synthesis System for RT-PCR kit from Stratagene according to the manufacturer's protocol. Real-time PCR was performed with 1 µl of cDNA from each replicate and gene-specific primers with a LightCycler FastStart DNA Master SYBR Green I kit (Roche, Indianapolis, IN) in the Light Cycler (Roche Applied Science, Indianapolis, IN) according to the manufacturer's protocol. The signal output for each gene was normalized to the level of β-actin to generate a relative expression ratio. Three replicate experiments were repeated at each time point for each gene, and the fold changes were averaged for a given sample. The change in gene expression was expressed as log2[fold change (RA/DMSO)]. Primer sequences were β-actin: forward (F): 5'-aagcaggagtacgatgagtc-3', reverse (R): 5'-tggagtcctcagatgcattg-3'; cyp26b1: F: 5'-atgaggctctggaaagctacc-3', R: 5'-ggagtgagtctcttgctcgaa-3'; Nr0B2: F: 5'-ataaccttgccgcggact-3', R: 5'-gcattacaacgctgcatcag-3', Nr2F5: F: 5'-gacagaatgttgccatgcc-3'; R: 5'-tcctgggccaaattagca-3'; Dehydrogenase/reductase (SDR family) member 3: F: 5'-gagatggcttgttgtccca-3', R: 5'-attgcattgacctgtgctaca-3'; Growth arrest-specific 6: F: 5'-cactgacgcttccattagacc-3'; R: 5'-gcgtttcagtctcacacacag-3'; BCL2-antagonist of cell death: F: 5'-gcgcgtcagatgagtcaat-3', R: 5'-tgttcaaagcagagcaccc-3'; Myeloblastosis oncogene-like 2: F: 5'-gatcagtcaggatcagtgctca-3', R: 5'-ccactggtggacagtagaaatg-3'; HoxB5a: F: 5'-tgtatcccaagacctgctacg-3', R: 5'-atccataacatgccgaaagg-3'; HoxB5b: F: 5'-gcgcaggaggagaatagaga-3', R: 5'-ggtgagaacaagctgaaggtg-3'.

Nr2F5 mRNA overexpression and morpholino injection.
Zebrafish nr2f5 (GenBank accession no. NM_131186) was cloned from a heart-specific first-strand cDNA pool with a pfuUltra II Fusion HS DNA Polymerase PCR kit (Stratagene) and the following primers: F (engineered with 5' HindIII site): 5'-acccaagcttatggcaatggtagtaaatcagtgg-3', R (engineered with 3' BamHI site): 5'-acgcggatcctcagggcccgttctcatt-3'. The nr2f5 PCR product was cloned into pSP64 Poly(A) vector between HindIII and BamHI sites in the multicloning site (MCS) region to generate pSP64-Nr2F5. Clones with correct sequence verified by DNA sequencing were selected. A second set of primers, F (engineered with 5' SP6 promoter region followed by Kozak sequence): 5'-atttaggtgacactatagaagcgccgccaccatggcaatggtagtaaatcagtgg-3' and R (oligo 30T): 5'-ttttttttttttttttttttttttttttttttt-3', was used to amplify nr2f5 PCR product from pSP64-Nr2F5. 7-Methylguanosine-capped Nr2F5 mRNA and control mRNA were generated with a mMESSAGE mMACHINE SP6 Kit (Ambion, Austin, TX) from Nr2F5 PCR product and linearized MCS region in the pSP64 vector backbone, respectively. One hundred picograms of either Nr2F5 mRNA or control mRNA was injected into one-cell stage AB embryos with a Narishige IM300 Microinjector (Tokyo, Japan), and images of the hearts were taken at 72 hours post fertilization (hpf). Fluorescein-tagged zebrafish Nr2F5 morpholino (MO) (Gene Tool, Philomath, OR) was designed to block the translation initiation site with sequence as follows: 5'-cactgatttactaccattgccatgc-3'. Three nanoliters of 0.6 mM Nr2F5 MO or control MO (Gene Tools' standard control morpholino: 5'-ctcttacctcagttacaatttata-3') in Danieau's solution [in mM: 58 NaCl, 0.7 KCl, 0.4 MgSO4, 0.6 Ca(NO3)2, 5 HEPES, pH 7.6] was injected with the same device into cmlc2::GFP one-cell-stage embryos. Injected embryos were then assessed 4 h later to be certain that strong fluorescence was distributed evenly inside the dividing cells as an indication of successful injection.

Statistical analysis.
For measurements of cardiac function, peripheral blood flow, pericardial sac area, and cardiac myocyte number, the parametric distribution of the variances of the data was confirmed by Levene's test or the Cochran C, Hartley, Bartlett test for homogeneity. The significance of difference in mean values between RA and control treatment groups was determined by two-way ANOVA followed by the Fisher least significant difference test of Statistica 7.0 (StatSoft, Tulsa, OK). Results are presented as means ± SE, with level of significance at P ≤ 0.05.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Experimental design.
Our immediate goal was to test the hypothesis that RA and TCDD both cause cardiotoxicity through the activation of a common set of target genes in the heart. Our intent was to take advantage of the previously published TCDD array data and collect a matching set of data from RA-exposed hearts. This goal set several constraints on the design of the experiment, which is outlined in Fig. 1.


Figure 1
View larger version (7K):
[in this window]
[in a new window]

  		Fig. 1. Schematic of experimental design. The timing of microarray analysis and assessments of cardiac function and heart morphology in zebrafish embryos is shown. RA, retinoic acid.

 
Each embryonic heart is composed of ~300 cells, and we have found that successful hybridization of Affymetrix zebrafish arrays requires the RNA from ~400 hearts. Heart extraction is optimal at ~72 hpf, a factor that had previously led us to start our TCDD exposure experiment at 72 hpf (6). RA exposure was therefore initiated at 72 hpf, and hearts were collected at 1, 2, 4, and 12 h after exposure. This allows identification of the early changes produced by direct activation via RAR, as well as later changes associated with cardiotoxicity. This time course also corresponds to the time points previously used in the TCDD exposure experiment.

We also needed to demonstrate that RA still produced a cardiotoxic response in 72 hpf zebrafish. Therefore, Fig. 1 also outlines gathering of cardiotoxicity data to confirm this. It should be emphasized that the effects of RA on heart development have been well characterized, and the purpose of these experiments was not to replicate work characterizing the effects of RA on the developing heart.

We expected that RA activation of RAR would produce transcriptional responses relatively rapidly. To follow this, we collected samples for microarray analysis at 1, 2, 4, and 12 h after initial exposure. We also wanted to relate changes in gene expression to cardiotoxic effects produced by RA. Since this takes time to be manifested, these data were gathered at 4, 8, 12, 24, 36, and 48 h after initial RA exposure. Finally, to collect data about heart cell number, cardiac myocytes were counted at 12 and 36 h.

Retinoic acid causes cardiac toxicity in embryonic zebrafish.
We found that RA exposure beginning at 72 hpf caused numerous cardiovascular defects, including pericardial edema, heart elongation, and reduced heart size (Fig. 2). In the lateral view, pericardial sac edema is clearly evident, along with an elongated atrium and a compacted ventricle. In the ventral view, pericardial edema is also evident, along with the reversal of the normal looping and elongated atrium. These defects, indicated by arrows in Fig. 2, were first evident at 24 h after exposure and became progressively more severe with time. This severe pericardial edema and decreased heart size was similar to the cardiac abnormalities observed in zebrafish embryos exposed to RA at the nine-somite stage (30).


Figure 2
View larger version (113K):
[in this window]
[in a new window]

  		Fig. 2. Differential interference contrast (DIC) images of embryos 4, 8, 12, 24, 30, and 48 h after initiation of RA exposure. Zebrafish embryos were exposed continuously to either 1 µM RA or 0.1% DMSO as a vehicle control starting at 72 hours post fertilization (hpf). DIC images of the hearts were taken at the indicated time points. Arrows indicate significant pericardial edema and heart malformation.

 
We used a computer program to outline and calculate the change in pericardial area caused by RA exposure (Fig. 3). This confirmed an increase in pericardial edema by 30 h after initial RA exposure.


Figure 3
View larger version (21K):
[in this window]
[in a new window]

  		Fig. 3. Severity of pericardial edema in response to RA exposure. Embryos were continuously exposed to either 1 µM RA or 0.1% DMSO as a vehicle control starting at 72 hpf. Pericardial sac area of the embryos was outlined and measured at the indicated time points as described in MATERIALS AND METHODS. Values are means ± SE of 5 replicates for each group. *Significant difference between RA and control for a given time point (P ≤ 0.05).

 
RA exposure also altered heart function, reducing ventricular stroke volume to nearly zero by 48 h after dosing (Fig. 4A). Since RA had little if any effect on heart rate (Fig. 4B), the altered stroke volume produced a parallel decrease in cardiac output (Fig. 4C). Reduced cardiac output was clearly evident by subjective observation 30 h after exposure, with halted blood flow throughout the embryo. This was confirmed by measurements of blood blow in the posterior intersegmental vein (Fig. 4D).


Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 4. Assessment of cardiac functions at designated times after dosing in RA- and DMSO control-exposed embryos. Embryos were exposed to either 1 µM RA or 0.1% DMSO continuously starting from 72 hpf. A: stroke volume calculated from end-diastolic and end-systolic volume values as described in MATERIALS AND METHODS. B: heart rate expressed as beats per minute. C: cardiac output calculated from stroke volume and heart rate as described in MATERIALS AND METHODS. D: peripheral blood flow measured as described in MATERIALS AND METHODS. All values are means ± SE of 6 replicates for stroke volume, heart rate, and cardiac output and 5 replicates for peripheral blood flow. *Significant difference between RA group and its respective vehicle control for a given time point (P ≤ 0.05).

 
Changes in transcript abundance in hearts of RA-exposed zebrafish embryos.
We used DNA microarrays to identify transcripts that were altered in embryonic hearts exposed to RA. The number of affected transcripts increased from 30 at the first hour to a total of 278 affected by 12 h. Altogether, a total of 309 transcripts were altered by at least twofold for at least one time point. In general, the number of genes that were upregulated was similar to the number of genes downregulated by RA exposure.

We used a SOM algorithm to divide the transcripts into sets of genes with similar temporal patterns of expression (Fig. 5, B and C, and Supplemental Table S1).1 Cluster 1 was composed of genes that were immediately upregulated and remained up at all time points. This cluster consisted primarily of transcripts encoding metabolism-related enzymes (5 genes) and transcription factors (3 genes) and cell signaling transcripts (4 genes). Cluster 2 was composed of transcripts that increased in abundance steadily over time. This cluster was enriched in cell signaling genes (9) and transcription factors (15). Cluster 3 was composed of transcripts that in general were not upregulated until the latest time point at 12 h. This cluster included transcripts involved in cell signaling (8), metabolism (12), and cytokine and immune functions (7) as well as encoding structural proteins (6) and transcription factors (5). Transcripts in cluster 4 had no clear pattern of regulation across the time course but were in many cases downregulated. These transcripts tended to return to normal levels by the 12 h time point and encoded cell cycle/DNA replication components (6), structural proteins (5) and metabolism-related enzymes (6). Cluster 5 was composed of genes that were sharply downregulated at the 12 h time point. This was the largest single cluster, containing primarily transcripts related to cell cycle progression/DNA replication (33), transcripts encoding metabolism enzymes (33), and cell signaling transcripts (12).


Figure 5
View larger version (44K):
[in this window]
[in a new window]

  		Fig. 5. RA-induced changes in mRNA expression levels in the embryonic heart. Embryos were continuously exposed to either 1 µM RA or 0.1% DMSO control starting from 72 hpf. A: representative view of an isolated heart from a 72 hpf cmlc2::GFP transgenic embryo collected for microarray analysis. Top: DIC image. Bottom: specific green fluorescent protein (GFP) expression in the heart under epifluorescence. B: self-organizing map (SOM) clustering of RA-induced changes in cardiac transcripts, producing 5 clusters. Genes with expression levels significantly changed by RA treatment by 2-fold or greater for at least 1 of the time points were selected for SOM clustering with euclidean distance and average linkage. Colors in the heat map indicate the extent of change compared with control displayed as log2-transformed values: green, downregulation; black, no change; red, upregulation. Log2-transformed fold changes of Nr2F5 mRNA expression level on RA treatment were 1.14, 4.23, 5.80, and 4.70 at 1, 2, 4, 12 h after dosing, respectively. The heat map row showing data for Nr2F5 mRNA is indicated by an arrow. C: line graph depiction of the data for each of the 5 clusters, in which each transcript is displayed as an individual line.

 
Cluster 1 is notable because it contains the most rapidly induced transcripts (Supplemental Table S1). As might be expected, many of these transcripts encode proteins involved in negative feedback control of RA levels, thus affording a homeostatic response to the sudden increase in RA concentration. Cluster 1 includes cyp26a1 and cyp26b1, encoding cytochrome P-450s that inactivate RA by converting it to 4-oxo-RA and 4-hydroxy-RA (35, 36). This group also contained Dehydrogenase/reductase (SDR family) member 3, which belongs to the short-chain dehydrogenase family that catalyze the reduction of all-trans retinal to all-trans retinol (13). The most prominent member of cluster 1 was Nr2F5, a member of the COUP-TF family that functions to repress transcription of steroid-thyroid hormone receptor target genes. Thus Nr2F5 acts antagonistically to RAR (9, 19, 32). Nr2F5 was among the most prominently altered transcripts in the experiment; it was upregulated at the earliest time point, was increased >55-fold by 4 h after RA exposure, and was persistently upregulated throughout the time course.

Cluster 5 was remarkable in that a large fraction of this group was made up of transcripts involved with promoting cell division and growth. These included transcripts that function in DNA replication, DNA repair, cell cycle progression, cellular proliferation, cell division, transcription, and chromosome assembly and maintenance (Supplemental Table S1). The marked downregulation of these messages occurred at a time when RA was causing the normal increase in cardiomyocyte numbers to cease (see Fig. 7).

To validate the results obtained with the microarrays we collected hearts for quantitative PCR (Fig. 6). Triplicate independent replicates showed very close matches to the data obtained with the microarray experiments. This lends confidence to the array results.


Figure 6
View larger version (9K):
[in this window]
[in a new window]

  		Fig. 6. Quantitative PCR (qRT-PCR) validation ofthe microarray results. Changes in the expression for 9 representative genes measured in the microarray experiment were validated with real-time PCR. Embryos were exposed to either 1 µM RA or 0.1% DMSO as described in the text, and hearts were isolated for RNA preparation in 3 replicate experiments. Real-time PCR was performed as described in MATERIALS AND METHODS. Values are mean ± SE log2-transformed fold changes (RA/DMSO) for 3 replicates.

 
The apparent reduction in heart size coupled with the downregulation of cell cycle genes contained in cluster 5 suggested the possibility that RA was producing a response in which heart cells were no longer increasing in number. To test this, we used RFP-labeled cardiomyocytes to count the number of myocytes in the RA and control hearts. While in the control hearts the number of cells increased over time, growth slowed in the RA-exposed hearts, and at 30 h after initial RA exposure we observed a significant reduction in the number of cardiac myocytes compared with the controls (Fig. 7).


Figure 7
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 7. RA exposure causes a reduction in cardiomyocyte numbers. Cmlc2::dsRed-nuc embryos were exposed to 1 µM RA or 0.1% DMSO control starting from 72 hpf. Embryos were flat mounted, and the number of cardiomyocytes expressing the nuclear marker was counted under epifluorescence at 12 and 30 h after dosing. Values are means ± SE of 6 replicates for each group. *Significant difference between an RA point and its respective vehicle control (P ≤ 0.05).

 
Comparison of RA and TCDD transcript changes.
Overall, RA exposure produced cardiotoxicity similar to that induced by TCDD (Fig. 8B). Both compounds caused severe pericardial edema, heart elongation, reduced heart size, decreased cardiomyocyte number, and circulatory collapse (23). In addition, both RA and TCDD are thought to produce their effects through activation of transcriptional regulators. Therefore, we reasoned that the similarities in cardiac toxicity produced by TCDD and RA might be due to a common set of transcriptional events. To test this, we combined the microarray data from this report with previously published microarray data examining the effects of TCDD on heart transcripts (6). We selected and combined data for transcripts that were altered by twofold or more in either of the experiments and used HCL to look for similarities between the patterns of transcriptional changes produced by the two agents (Fig. 8, A and C).


Figure 8
View larger version (55K):
[in this window]
[in a new window]

  		Fig. 8. Comparison of gene expression changes produced in the zebrafish embryonic heart by RA and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure. For RA treatment, embryos were exposed to 1 µM RA or DMSO as described in text; for TCDD treatment, embryos were exposed to TCDD or DMSO starting at 72 hpf (9). Both microarray analyses were conducted on hearts isolated at 1, 2, 4, and 12 h after dosing. A: hierarchical clustering (HCL) of all genes with significant changes of ≥2-fold in either of the RA and TCDD microarray analyses. HCL was performed with euclidean distance and average linkage. Colors indicate the extent of regulations that correspond to the values of log2-transformed fold changes: green, downregulation; black, no change; red, upregulation. Box outlines the set of transcripts with a similar pattern of expression in both RA and TCDD profiles. B: comparison of the heart morphology at 120 hpf induced by RA and TCDD treatments. Top left, DMSO control to RA treatment; top right, DMSO control to TCDD treatment; bottom left, RA treatment; bottom right, TCDD treatment. C: Gene Ontology descriptions for the boxed cell cycle gene cluster gene set.

 
We could find little if any overlap between the initial responses produced by RA and TCDD at 1, 2, or 4 h. This is revealed in the heat map in Fig. 8A, where transcripts shown as red or green are strongly regulated and unchanged transcripts are indicated by shades of black. In most cases at the early time points, the transcripts showing strong induction or repression in response to one treatment showed little response to the other treatment. However, we did find a cluster of overlapping transcripts at the 12 h time point, a point at which the hearts from both treatment types showed distinct signs of failure. This relatively large group of downregulated genes consists of transcripts involved in cell cycle progression (Fig. 8C, Supplemental Table S2). This overlapping group of transcripts corresponds to cluster 5 from Fig. 5 and cluster 9 from Carney et al. (6).

Overall, our results are consistent with a model in which the divergent RA and TCDD signaling pathways do not initially produce a common transcriptional response. However, both agents produce effects that ultimately trigger a heart failure response that appears to be similar for both compounds. Once initiated, this response appears to proceed in parallel with both agents, producing a downregulation of a cell cycle gene cluster (CCGC) and a halt in myocyte growth.

Nr2F5 mediates RA-induced cardiac toxicity in 72 hpf zebrafish embryos.
As mentioned above, Nr2F5 was massively upregulated after RA exposure. Nr2F5 is a member of the COUP-TF family, a group of orphan steroid hormone receptors. These proteins function as transcriptional repressors, countering the activation produced by steroid-thyroid hormone receptors such as RAR.

We hypothesized that the rapid increase in Nr2F5 expression produced by RA exposure plays an important role in causing heart malformations. An alternative hypothesis is that changes in Nr2F5 levels are just one part of the overall RA response, and do nothing to cause the cardiotoxic effects of RA. If the first hypothesis is true, then artificially increasing Nr2F5 expression should phenocopy the heart malformations produced by RA exposure. To test this we microinjected Nr2F5 mRNA into embryos at the one-cell stage to increase Nr2F5 expression. This recapitulated the cardiac response to RA, producing a small, stretched heart with pericardial edema (Fig. 9). Because of the mosaic expression of injected mRNA in zebrafish embryos, we expected that a limited percentage of the injected embryos would have sufficient expression levels in the right location in the fish to produce the full response. Accordingly, we found that 17% of embryos injected with Nr2F5 mRNA (11 of 63) exhibited clearly apparent signs matching RA-induced cardiac malformations by 72 hpf. In sharp contrast, 0% of the embryos microinjected with control mRNA exhibited cardiac malformations.


Figure 9
View larger version (63K):
[in this window]
[in a new window]

  		Fig. 9. Nr2F5 as a critical target in RA-induced cardiac toxicity in zebrafish embryos. One-cell-stage embryos were injected with either control or Nr2F5 mRNA as described in MATERIALS AND METHODS, and images of the hearts were taken at 72 hpf. Top: typical appearance of embryo after injection of 100 pg of control RNA in 1-cell-stage embryo showing no effect on heart development. Middle and bottom: range of appearance of embryos injected with 100 pg Nr2F5 mRNA, showing small and elongated hearts with mild to severe pericardial edema. Arrows point to the hearts.

 
If the induction of Nr2F5 mRNA by RA exposure plays an important part in producing cardiac malformation, then we should be able to protect embryonic hearts from RA toxicity by blocking the induction of Nr2F5. We tested this using morpholino oligos (MOs) that specifically block Nr2F5 translation by hybridizing to the mRNA. This allowed us to determine whether knocking down Nr2F5 translation could rescue RA-treated embryos from cardiotoxicity. Zebrafish embryos at the 1- to 2-cell stage were injected with either a control MO or a Nr2F5 MO and then starting at 72 hpf were continuously exposed to either RA or vehicle. We found that the specific Nr2F5 MO targeted to the Nr2F5 translational start site blocked the cardiotoxicity produced by RA exposure (Fig. 10A). The hearts of the RA-exposed embryos treated with the Nr2F5 MO appeared normal and were indistinguishable from the hearts of control embryos not exposed to RA. In sharp contrast, control MO had no ability to protect the heart from RA exposure.


Figure 10
View larger version (34K):
[in this window]
[in a new window]

  		Fig. 10. Nr2F5 morpholino (MO) blocks RA-induced cardiac toxicity in zebrafish. A: 1-cell-stage embryos were injected with either Nr2F5 MO or control MO as described in MATERIALS AND METHODS. At 72 hpf the embryos were exposed to 0.1 µM RA or 0.1% DMSO as the vehicle control, and images of the hearts were recorded at 120 hpf. Top: Nr2F5 MO- and control MO-injected embryos exposed to DMSO, showing normal heart development; bottom, Nr2F5 MO- and control MO-injected embryos exposed to RA, showing rescue from the effect of RA for the Nr2F5 morphants, but not for the embryos receiving the control MO. Arrows point to the hearts. B: lateral views of embryos from the experiment described in A were photographed and the area of the pericardial sac was measured for each embryo with NIH Image. Gray bars, embryos treated with RA; open bars, control embryos exposed to DMSO. Values for edema are means ± SE from 4 separate experiments with n = 3–5 embryos/group in each experiment. *Significant difference between Nr2F5 MO and control MO (P ≤ 0.001).

 
We used lateral-view digital images to measure the protection afforded by the Nr2F5 MO against pericardial edema in fish exposed to RA (Fig. 10B). This clearly showed that while control MO provided no protection, the Nr2F5 MO was effective in rescuing the developing zebrafish from signs of pericardial edema. Taken together, our results indicate that induction of Nr2F5 by RA exposure plays a critical part in producing cardiac toxicity.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
RA cardiotoxicity at 72 hpf.
Many studies of RA cardiotoxicity have concentrated on early events such as precardiac field formation, chamber morphogenesis, looping, and cardiac myocyte differentiation (8, 12, 18, 30). However, our experiments examined developmental stages that occur after these basic pattern formation events are completed. By 72 hpf the zebrafish heart has fully formed, functional, well-looped chambers. This did not prevent RA from producing a cardiotoxic effect. Because the cardiotoxicity produced by RA in our study occurred after basic heart formation was completed, the cardiotoxic response that we observed could not be due to the blockade of events occurring before the RA exposure and must instead be due to interference with processes ongoing at 72 hpf. This raises the question of whether heart patterning events are irreversible after completion, or whether some aspects of heart patterning or organogenesis must be actively maintained even after completion. In our experiments, RA exposure appeared to elongate the heart, reducing looping and leaving a heart in which the atrium lies almost directly posterior to the ventricle. This suggests that an active process maintains the heart pattern after formation, at least at 72 hpf, and that this process involves RA signaling. Recent work examining the ability of adult heart cells to proliferate in response to wounding or growth signals indicates that proliferative signals are involved in maintaining the structure of the heart throughout zebrafish life (37).

Transcriptional response to RA in embryonic zebrafish heart.
By rapidly isolating hearts from RA-exposed zebrafish embryos we were able to follow the transcriptional response to RA as it progressed to cardiotoxicity. This yielded a group of genes that could be divided in terms of time course and known function. Our expectation was that the direct targets for the activated RAR complex would be found among the genes affected soon after RA addition. Not surprisingly, the transcripts affected earliest included mRNAs encoding proteins known to antagonize RA action, indicating that RAR activation is linked to a negative feedback loop maintaining RA homeostasis.

While it might be expected that direct targets of RA-activated RAR/RXR would be revealed as rapidly upregulated transcripts, we observed an almost equal number of downregulated transcripts at 1 h after RA exposure. This suggests either that RA-induced transcripts are rapidly translated to produce transcriptional inhibitors that act within this first hour or that RA-activated RAR/RXR can in some manner cause the repression of specific target genes.

There is no doubt that RA treatment produced effects in the developing zebrafish that were not limited to altered mRNAs in the heart. These effects would be expected to contribute to the overall response to RA, but an intrinsic limitation of our focus on the heart is that these effects would not be identified with our approach.

Comparison of transcriptional responses to RA and TCDD.
Both RA and TCDD exposure rapidly altered transcript levels in the zebrafish heart. While the compounds produced similar types of cardiotoxicity, we found little evidence that the two different compounds cause cardiotoxicity by induction of the same set of target genes. The transcriptional changes produced by the two agents were almost entirely nonoverlapping during the first three time points, a period during which we expected the receptors to be directly regulating transcription.

Nonetheless, as toxicity progressed into the early stages of heart failure at the 12 h point, we found a cluster of downregulated transcripts that was common to both treatments. This corresponded to a CCGC that had been earlier identified as a late response to TCDD treatment (6). This cluster is made up of genes known to play important roles in cell cycle progression. Furthermore, many of these genes were downregulated substantially after RA exposure, suggesting that this produces a halt in cardiomyocyte proliferation. However, the degree to which these genes must be repressed in order to halt heart cell growth in the developing zebrafish has not been determined, so we cannot say with certainty that the drop in CCGC expression causes a halt in heart growth. Nevertheless, our results suggest a model in which different cardiotoxic agents can trigger a response that represses cell proliferation genes and halts the growth of the heart in the developing zebrafish. We have not identified any common pathway that could lead to this; however, it is possible that the response is triggered by a common hemodynamic failure caused by divergent RA and TCDD signaling events.

The COUP-TF protein Nr2F5.
An obvious goal for microarray studies is to identify the important downstream transcripts that control biological responses. Among the more interesting potential RA targets found in cluster 1 was Nr2F5, a member of the COUP-TF family. This transcript stands out because of the rapidity and degree of induction by RA. The sequence of a canonical RARE is described as two direct PuG(G/T)TCA repeats separated by n number of spacers, PuG(G/T)TCA (X)n PuG(G/T)TCA, summarized by Chambon (7). The most frequent number of spacers is 5, 2, or 1; however, wider spacers and more complex organizations of the motifs have been observed (3, 7). For Nr2F5 we found two TGGTCA motifs with a spacer of n = 13 at the –1410 to –1433 position, indicating that the rapid induction by RA could reflect a direct interaction between RAR and the Nr2F5 promoter.

Previous reports have shown that COUP-TF proteins are essential for the function of RA in growth inhibition and apoptosis in cancer cell lines (20). Targeted deletion of COUP-TF II in mouse embryos led to defects in heart development and angiogenesis (24). This known association with RA and the heart, taken with the prominent induction by RA in our experiments, made Nr2F5 an attractive candidate for a mediator of RA-induced cardiotoxicity. We found that overexpression of Nr2F5 could phenocopy and thus explain the cardiotoxicity produced by RA. While it can be argued that the production of heart failure and pericardial edema could be due to nonspecific effects of mRNA injection or protein overexpression, we also found that Nr2F5 MOs blocked the cardiotoxicity produced by RA exposure. It is difficult to imagine how nonspecific mechanisms could lead to such a dramatic rescue. The simplest explanation for our results is that Nr2F5 plays a role in mediating the cardiotoxic response to RA in the zebrafish heart.

Nr2F5 is one of three COUP-TF family members found in zebrafish (15). COUP-TF proteins act as repressors at hormone-responsive elements (9, 19, 32) and are needed for RA-induced growth inhibition and apoptosis (20). It is interesting that the level of nuclear COUP-TF has been found to be elevated in heart cells of mice with cardiac hypertrophy (26), while loss of COUP-TF II in knockout mouse embryos produces underdevelopment of the atria and sinus venosa (24). COUP-TF II has been reported to interact with MyoD, the master transcription factor required for skeletal muscle myogenesis. COUP-TF II tethers MyoD away from the activation complex, suppressing MyoD-mediated myogenesis in pluripotential C3H10T1/2 cell lines (2, 22). Although the master transcriptional regulator of cardiac myogenesis has not been found, we speculate that in the developing zebrafish Nr2F5 might inhibit cardiac myogenesis through a mechanism analogous to the inhibition of MyoD in skeletal muscle cells.

RA-induced cardiac toxicity and teratogenic effects on the heart have been widely reported at the anatomic, pathological, and pathophysiological levels. Our findings provide a step forward in understanding the effects of RA on the developing heart at the molecular level.


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by National Institutes of Health (NIH) grant R01-ES-012716 from the National Institute of Environmental Health Sciences (NIEHS) (W. Heideman and R. E. Peterson) and the University of Wisconsin Sea Grant Institute, National Sea Grant College Program, National Oceanic and Atmospheric Administration, US Department of Commerce Grant NA-16RG2257, Sea Grant Project Numbers R/BT-17, R/BT-20, and R/BT-22 (W. Heideman and R. E. Peterson), NIH Grant T32-ES-07015 from the NIEHS (S. A. Carney), and the Pharmaceutical Sciences and Molecular and Environmental Toxicology Graduate Programs (University of Wisconsin-Madison). The contents are solely the responsibility of the authors and do not necessarily represent the official view of the NIEHS, NIH.


   ACKNOWLEDGMENTS
 
We thank Dorothy Nesbit for technical assistance with the real-time PCR analysis, Dagmara Antkiewicz for expertise with microscopy imaging, and Dr. Craig Struble, Department of Mathematics, Statistics, and Computer Science, Marquette University (Milwaukee, WI), for assistance in computational RARE analysis.

Present address for S. A. Carney: Skin Science Research, P&TD Technology, Kimberly Clark Corporation, Neenah, WI 54956.


   FOOTNOTES
 
Address for reprint requests and other correspondence: W. Heideman, School of Pharmacy, Univ. of Wisconsin, 777 Highland Ave., Madison, WI 53705 (e-mail: wheidema{at}wisc.edu).

Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

1 The online version of this article contains supplemental material. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Antkiewicz DS, Burns CG, Carney SA, Peterson RE, Heideman W. Heart malformation is an early response to TCDD in embryonic zebrafish. Toxicol Sci 84: 368–377, 2005.[Abstract/Free Full Text]
  2. Bailey P, Sartorelli V, Hamamori Y, Muscat GE. The orphan nuclear receptor, COUP-TF II, inhibits myogenesis by post-transcriptional regulation of MyoD function: COUP-TF II directly interacts with p300 and myoD. Nucleic Acids Res 26: 5501–5510, 1998.[Abstract/Free Full Text]
  3. Bastien J, Rochette-Egly C. Nuclear retinoid receptors and the transcription of retinoid-target genes. Gene 328: 1–16, 2004.[CrossRef][Web of Science][Medline]
  4. Brand N, Petkovich M, Krust A, Chambon P, de The H, Marchio A, Tiollais P, Dejean A. Identification of a second human retinoic acid receptor. Nature 332: 850–853, 1988.[CrossRef][Medline]
  5. Burns CG, MacRae CA. Purification of hearts from zebrafish embryos. Biotechniques 40: 274, 276, 278, 2006.
  6. Carney SA, Chen J, Burns CG, Xiong KM, Peterson RE, Heideman W. Aryl hydrocarbon receptor activation produces heart-specific transcriptional and toxic responses in developing zebrafish. Mol Pharmacol 70: 549–561, 2006.[Abstract/Free Full Text]
  7. Chambon P. A decade of molecular biology of retinoic acid receptors. FASEB J 10: 940–954, 1996.[Abstract]
  8. Chazaud C, Chambon P, Dolle P. Retinoic acid is required in the mouse embryo for left-right asymmetry determination and heart morphogenesis. Development 126: 2589–2596, 1999.[Abstract]
  9. Cooney AJ, Tsai SY, O'Malley BW, Tsai MJ. Chicken ovalbumin upstream promoter transcription factor (COUP-TF) dimers bind to different GGTCA response elements, allowing COUP-TF to repress hormonal induction of the vitamin D3, thyroid hormone, and retinoic acid receptors. Mol Cell Biol 12: 4153–4163, 1992.[Abstract/Free Full Text]
  10. Coucelo J, Joaquim N, Coucelo J. Calculation of volumes and systolic indices of heart ventricle from Halobatrachus didactylus: echocardiographic noninvasive method. J Exp Zool 286: 585–595, 2000.[CrossRef][Web of Science][Medline]
  11. Denison MS, Fisher JM, Whitlock JP Jr. The DNA recognition site for the dioxin-Ah receptor complex. Nucleotide sequence and functional analysis. J Biol Chem 263: 17221–17224, 1988.[Abstract/Free Full Text]
  12. Drysdale TA, Patterson KD, Saha M, Krieg PA. Retinoic acid can block differentiation of the myocardium after heart specification. Dev Biol 188: 205–215, 1997.[CrossRef][Web of Science][Medline]
  13. Duester G. Families of retinoid dehydrogenases regulating vitamin A function: production of visual pigment and retinoic acid. Eur J Biochem 267: 4315–4324, 2000.[Web of Science][Medline]
  14. Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 14863–14868, 1998.[Abstract/Free Full Text]
  15. Fjose A, Nornes S, Weber U, Mlodzik M. Functional conservation of vertebrate seven-up related genes in neurogenesis and eye development. EMBO J 12: 1403–1414, 1993.[Web of Science][Medline]
  16. Grandel H, Lun K, Rauch GJ, Rhinn M, Piotrowski T, Houart C, Sordino P, Kuchler AM, Schulte-Merker S, Geisler R, Holder N, Wilson SW, Brand M. Retinoic acid signalling in the zebrafish embryo is necessary during pre-segmentation stages to pattern the anterior-posterior axis of the CNS and to induce a pectoral fin bud. Development 129: 2851–2865, 2002.[Web of Science][Medline]
  17. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 31: e15, 2003.
  18. Keegan BR, Feldman JL, Begemann G, Ingham PW, Yelon D. Retinoic acid signaling restricts the cardiac progenitor pool. Science 307: 247–249, 2005.[Abstract/Free Full Text]
  19. Kliewer SA, Umesono K, Heyman RA, Mangelsdorf DJ, Dyck JA, Evans RM. Retinoid X receptor-COUP-TF interactions modulate retinoic acid signaling. Proc Natl Acad Sci USA 89: 1448–1452, 1992.[Abstract/Free Full Text]
  20. Lin B, Chen GQ, Xiao D, Kolluri SK, Cao X, Su H, Zhang XK. Orphan receptor COUP-TF is required for induction of retinoic acid receptor beta, growth inhibition, and apoptosis by retinoic acid in cancer cells. Mol Cell Biol 20: 957–970, 2000.[Abstract/Free Full Text]
  21. Mably JD, Mohideen MA, Burns CG, Chen JN, Fishman MC. heart of glass regulates the concentric growth of the heart in zebrafish. Curr Biol 13: 2138–2147, 2003.[CrossRef][Web of Science][Medline]
  22. Muscat GE, Rea S, Downes M. Identification of a regulatory function for an orphan receptor in muscle: COUP-TF II affects the expression of the myoD gene family during myogenesis. Nucleic Acids Res 23: 1311–1318, 1995.[Abstract/Free Full Text]
  23. Niederreither K, Vermot J, Messaddeq N, Schuhbaur B, Chambon P, Dolle P. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128: 1019–1031, 2001.[Abstract]
  24. Pereira FA, Qiu Y, Zhou G, Tsai MJ, Tsai SY. The orphan nuclear receptor COUP-TFII is required for angiogenesis and heart development. Genes Dev 13: 1037–1049, 1999.[Abstract/Free Full Text]
  25. Prasch AL, Teraoka H, Carney SA, Dong W, Hiraga T, Stegeman JJ, Heideman W, Peterson RE. Aryl hydrocarbon receptor 2 mediates 2,3,7,8-tetrachlorodibenzo-p-dioxin developmental toxicity in zebrafish. Toxicol Sci 76: 138–150, 2003.[Abstract/Free Full Text]
  26. Sack MN, Disch DL, Rockman HA, Kelly DP. A role for Sp and nuclear receptor transcription factors in a cardiac hypertrophic growth program. Proc Natl Acad Sci USA 94: 6438–6443, 1997.[Abstract/Free Full Text]
  27. Saeed AI, Sharov V, White J, Li J, Liang W, Bhagabati N, Braisted J, Klapa M, Currier T, Thiagarajan M, Sturn A, Snuffin M, Rezantsev A, Popov D, Ryltsov A, Kostukovich E, Borisovsky I, Liu Z, Vinsavich A, Trush V, Quackenbush J. TM4: a free, open-source system for microarray data management and analysis. Biotechniques 34: 374–378, 2003.[Web of Science][Medline]
  28. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr 2: 358–367, 1989.[Medline]
  29. Schmidt JV, Bradfield CA. Ah receptor signaling pathways. Annu Rev Cell Dev Biol 12: 55–89, 1996.[CrossRef][Web of Science][Medline]
  30. Stainier DY, Fishman MC. Patterning the zebrafish heart tube: acquisition of anteroposterior polarity. Dev Biol 153: 91–101, 1992.[CrossRef][Web of Science][Medline]
  31. Toronen P, Kolehmainen M, Wong G, Castren E. Analysis of gene expression data using self-organizing maps. FEBS Lett 451: 142–146, 1999.[CrossRef][Web of Science][Medline]
  32. Tsai SY, Tsai MJ. Chick ovalbumin upstream promoter-transcription factors (COUP-TFs): coming of age. Endocr Rev 18: 229–240, 1997.[Abstract/Free Full Text]
  33. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 98: 5116–5121, 2001.[Abstract/Free Full Text]
  34. Walker MK, Peterson RE. Aquatic toxicity of dioxins and related chemicals. In: Dioxins and Health, edited by Schecter A. New York: Plenum, 1994, p. 347–387.
  35. White JA, Guo YD, Baetz K, Beckett-Jones B, Bonasoro J, Hsu KE, Dilworth FJ, Jones G, Petkovich M. Identification of the retinoic acid-inducible all-trans-retinoic acid 4-hydroxylase. J Biol Chem 271: 29922–29927, 1996.[Abstract/Free Full Text]
  36. White JA, Ramshaw H, Taimi M, Stangle W, Zhang A, Everingham S, Creighton S, Tam SP, Jones G, Petkovich M. Identification of the human cytochrome P450, P450RAI-2, which is predominantly expressed in the adult cerebellum and is responsible for all-trans-retinoic acid metabolism. Proc Natl Acad Sci USA 97: 6403–6408, 2000.[Abstract/Free Full Text]
  37. Wills AA, Holdway JE, Major RJ, Poss KD. Regulated addition of new myocardial and epicardial cells fosters homeostatic cardiac growth and maintenance in adult zebrafish. Development 135: 183–192, 2008.[Abstract/Free Full Text]
  38. Wilson JG, Warkany J. Aortic-arch and cardiac anomalies in the offspring of vitamin A deficient rats. Am J Anat 85: 113–155, 1949.[CrossRef][Web of Science][Medline]
  39. Wu Z, Irizarry RA, Gentleman R, Martinez-Murillo F, Spencer F. A Model-Based Background Adjustment for Oligonucleotide Expression Arrays. Johns Hopkins University, Department of Biostatistics Working Papers, Working Paper 1, 2004.




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow All Versions of this Article:
33/2/148    most recent
00214.2007v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, J.
Right arrow Articles by Heideman, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, J.
Right arrow Articles by Heideman, W.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2008 by the American Physiological Society.