ToxSci Advance Access originally published online on September 28, 2005
Toxicological Sciences 2006 90(1):61-72; doi:10.1093/toxsci/kfi341
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
HIGHLIGHTED ARTICLE |
Toxicogenomic Profiling of the Hepatic Tumor Promoters Indole-3-Carbinol, 17ß-Estradiol and ß-Naphthoflavone in Rainbow Trout
,
,1
* Department of Environmental and Molecular Toxicology, Marine and Freshwater Biomedical Sciences Center and Linus Pauling Institute, Center for Gene Research and Biotechnology, Environmental Health Sciences Center, and Department of Statistics, Oregon State University, Corvallis, Oregon 97331
1 To whom correspondence should be addressed at Marine and Freshwater Biomedical Sciences Center, 435 Weniger Hall, Oregon State University, Corvallis, OR 97331. E-mail: david.williams{at}oregonstate.edu.
Received August 15, 2005; accepted September 19, 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Indole-3-carbinol (I3C), from cruciferous vegetables, has been found to suppress or enhance tumors in several animal models. We previously reported that dietary I3C promotes hepatocarcinogenesis in rainbow trout (Oncorhynchus mykiss) at concentrations that differentially activated estrogen receptor (ER) or aryl hydrocarbon receptor (AhR)-mediated responses based on individual protein biomarkers. In this study, we evaluated the relative importance of these pathways as potential mechanisms for I3C on a global scale. Hepatic gene expression profiles were examined in trout after dietary exposure to 500 and 1500 ppm I3C and 3,3'-diindolylmethane (DIM), a major in vivo component of I3C, and were compared to the transcriptional signatures of two model hepatic tumor promoters: 17ß-estradiol (E2), an ER agonist, and ß-naphthoflavone, an AhR agonist. We demonstrate that I3C and DIM acted similar to E2 at the transcriptional level based on correlation analysis of expression profiles and clustering of gene responses. Of the genes regulated by E2 (fold change
2.0 or 
0.50), most genes were regulated similarly by DIM (87–92%) and I3C (71%), suggesting a common mechanism of action. Of interest were upregulated genes associated with signaling pathways for cell growth and proliferation, vitellogenesis, and protein folding, stability, and transport. Other genes downregulated by E2, including those involved in acute-phase immune response, were also downregulated by DIM and I3C. Gene regulation was confirmed by qRT-PCR and Western blot. These data indicate I3C promotes hepatocarcinogenesis through estrogenic mechanisms in trout liver and suggest DIM may be an even more potent hepatic tumor promoter in this model. Key Words: indole-3-carbinol; 3,3'-diindolylmethane; 178-estradiol; gene expression; rainbow trout.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Indole-3-carbinol (I3C) is a naturally occurring glucosinolate hydrolysis product found in significant concentrations in Brassica vegetables such as broccoli, cauliflower and cabbage (Fig. 1; McDanell et al., 1988
). 3,3'-Diindolylmethane (DIM) is the major I3C acid condensation product formed after oral administration and measured in the liver after absorption and distribution in trout and rodent models (Anderton et al., 2004; Dashwood et al., 1989; Stresser et al., 1995b). Both indole phytochemicals are also available as dietary supplements and are promoted for their well-established chemoprotective effects. I3C is chemoprotective in a number of animal models, particularly when administered in the diet concurrent with, or prior to, the carcinogen, effectively blocking initiation (Grubbs et al., 1995; Kojima et al., 1994). In some studies, decreased tumor incidence was correlated with inhibition of carcinogen–DNA adducts, indicating that the ability of indoles to induce cytochrome P450s (CYPs) involved in detoxication of procarcinogens through the aryl hydrocarbon receptor (AhR) or inhibition of CYPs capable of bioactivation are possible mechanisms for chemoprevention (Dashwood et al., 1994; Stresser et al., 1995a). Other mechanisms determined in vitro, including the ability of both I3C and DIM to alter cell cycle progression, proliferation, and apoptosis, suggest indoles may also target other stages of carcinogenesis (Kim and Milner, 2005). Of particular interest is the ability of indoles to act as anti-estrogens in certain systems by antagonizing the estrogen receptor (ER)-mediated actions of endogenous 17ß-estradiol (E2) or by metabolizing E2 to less estrogenic forms through induction of CYPs (Lord et al., 2002; Meng et al., 2000).
|
Despite clear evidence for chemoprotective effects, I3C has also been found to promote tumor formation in multiple organs in rodent and trout models after dietary exposure post-initiation (Oganesian et al., 1999
; Stoner et al., 2002; Yoshida et al., 2004). Some studies suggest that the promotional potency of I3C is at least as great as its potency as an anti-initiating agent (Bailey et al., 1991; Oganesian et al., 1999; Stoner et al., 2002). Although the mechanisms for promotion are not well understood, it is possible that ER and/or AhR-mediated processes similar to those described for chemoprevention may be important. We have previously reported that I3C promotes aflatoxin B1 (AFB1)-induced hepatocarcinogenesis in trout at concentrations that differentially induced vitellogenin (VTG) and not CYP1A, while higher concentrations induced both proteins (Oganesian et al., 1999). The relative induction of VTG and CYP1A, which are frequently used as markers for activation of ER and AhR-mediated pathways, respectively, suggest that ER-mediated responses may be important for promotion by I3C in trout. Further, in vitro studies with DIM have shown it to have estrogenic activity in certain cancer cells by ligand-independent activation of ER (Leong et al., 2004; Riby et al., 2000) and we have also found DIM to induce VTG in trout (Shilling and Williams, 2001). However, promotion of endometrial adenocarcinoma by I3C in rats was recently correlated with the induction of CYP1A and CYP1B enzymes and sequential formation of toxic E2 catechol metabolites, suggesting AhR-mediated pathways may be more important (Yoshida et al., 2004). Therefore, while the mechanisms of action for I3C and DIM have been found to involve both ER and AhR-mediated pathways, the relative importance of either in trout liver has not been evaluated by a comprehensive toxicogenomics approach.
In this study, we determined the relative importance of ER and AhR-mediated pathways in the mechanism of indole phytochemicals by microarray analysis. One of the inherent strengths of microarray technology is the ability to perform correlation analyses on compounds of interest to reveal commonality in global gene networks and provide insight into potential mechanisms of action. Hepatic gene expression profiles were examined in trout after dietary exposure to I3C and DIM at concentrations mimicking those from the tumor promotion study. Indole profiles were compared to the transcriptional signatures of two model hepatic tumor promoters: E2, an ER-agonist, and ß-naphthoflavone (ßNF), an AhR-agonist (Bailey et al., 1989; Nunez et al., 1989). We demonstrate that transcriptional profiles of I3C and DIM strongly overlap with E2 based on correlation analyses. These data indicate that I3C acts similar to E2 in trout liver in vivo and likely promotes hepatocarcinogenesis through estrogenic mechanisms. Interestingly, these data also suggest DIM may have a greater promotional potency than I3C in the trout tumor model based on this mechanism.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Materials.
Analytical grade I3C, ßNF, and E2 were purchased from Sigma Chemical (St. Louis, MO). DIM was kindly donated by BioResponse (Boulder, CO), and the purity was confirmed by HPLC. All other compounds were purchased from Sigma unless otherwise stated.
Experimental animals and treatments.
Mt. Shasta strain rainbow trout were hatched and reared at the Oregon State University Sinnhuber Aquatic Research Laboratory in 14°C carbon-filtered flowing well water on a 12:12 h light:dark cycle. All animal protocols were performed in accordance with Oregon State University Institutional Animal Care and Use Committee guidelines. Juvenile trout, 12–18 months old, were maintained in separate 375-l tanks (n = 2 tanks) for each treatment with six fish per tank. Animals were fed a maintenance ration (2.8% w/w) of Oregon test diet, a semipurified casein-based diet (Lee et al., 1991). Administration of 500 or 1500 ppm I3C or DIM, 5 ppm E2, 500 ppm ßNF, or 0.15% dimethyl sulfoxide vehicle control in the diet was carried out for 12 days. The indole concentrations in the diet for 500 and 1500 ppm are equivalent to 25 and 76 mg/kg/day, respectively, and were chosen to mimic those used in a trout tumor promotion study with I3C in which 1500 ppm I3C maximally induced both VTG and CYP1A protein biomarkers (Oganesian et al., 1999). Concentrations of E2 and ßNF were also chosen based on their ability to maximally induce VTG and CYP1A, respectively, and act as hepatic tumor promoters in trout (Bailey et al., 1989; Nunez et al., 1989). On day 13, fish were euthanized by deep anesthesia with 250 ppm tricaine methanesulfonate. Approximately 100 mg liver tissue from individual fish was minced, stored in TRIzol Reagent (Invitrogen, Carlsbad, CA), and quick frozen in liquid nitrogen for gene expression analysis. The rest of the liver was quick frozen in liquid nitrogen for protein analysis. All tissues were taken within 1 h of the scheduled time period.
RNA isolation.
Total hepatic RNA was isolated from individual trout liver using TRIzol Reagent followed by cleanup with RNeasy Mini Kits (Qiagen, Valencia, CA) according to manufacturer instructions. Equal amounts of RNA (µg) were pooled from each of the six fish per tank for every treatment (n = 2), except vehicle control in which RNA was pooled for use as a reference sample from 12 fish in both tanks. RNA quality and quantity were assessed by agarose gel electrophoresis, spectrophotometric absorbency at 260/280 nm, and bioanalyzer trace (Bioanalyzer 2100, Agilent, Palo Alto, CA).
Microarray hybridization and analysis.
Salmonid cDNA microarrays (GRASP3.7k v.1) were purchased from B. F. Koop and W. Davidson (Genome Research on Atlantic Salmon Project, University of Victoria, BC, Canada; http://web.uvic.ca/cbr/grasp). Microarray fabrication and quality control have been described previously (Rise et al., 2004). The array contains 3,119 unique Atlantic salmon cDNAs and 438 unique rainbow trout cDNAs (printed in duplicate), which have been found to have high cross-reactivity with rainbow trout targets, >73% and >61%, respectively, similar to that for Atlantic salmon targets. Hybridizations were performed with the Genisphere Array350 kit and instructions (Hatfield, PA) using standard reference design with dye-swapping. Briefly, 10 µg total RNA was reverse-transcribed with Superscript II (Invitrogen) using the Genisphere oligo d(T) primer containing a capture sequence for the Cy3 or Cy5 labelling reagents. Each reaction was spiked with increasing concentrations of three of the Arabidopsis thaliana cDNA controls included on the array: PSII oxygen-evolving complex protein 2 (clone ID 175B23T7), ferrodoxin (clone ID 249A17T7), and protochlorophyllide reductase precursor (clone ID 166N16T7) provided as a gift from Dr. Ed Allen, Oregon State University. Test arrays were hybridized using cDNAs without spiking controls and resulted in no cross-reactivity of trout samples to Arabidopsis control spots on the array (data not shown). Each cDNA sample containing the capture sequence for the Cy3 or Cy5 label was combined with equal amounts reference cDNA (pooled from vehicle control) containing the sequence for the opposite label. Every cDNA sample was dye-swapped and hybridized to two slides as technical replicates. Prior to hybridization, microarrays were processed post-printing by washing twice in 0.1% SDS for 5 min, twice in Milli-Q water for 5 min, and immersing in boiling water for 3 min and then dried by centrifugation. Arrays were then washed in 2x SSC, 0.1% SDS at 49°C for 20 min, 0.1x SSC for 5 min, and Milli-Q water for 3 min prior to drying by centrifugation. The cDNAs (35 µl) were hybridized to arrays in formamide buffer (50% formamide, 8x SSC, 1% SDS, 4x Denhardt's solution) for 16 h at 49°C with 22 x 60 mm Lifterslips (Erie Scientific, Portsmouth, NH). Arrays were then washed once in 2x SSC, 0.1% SDS at 49°C for 10 min, twice in 2x SSC, 0.1% SDS for 5 min, twice in 1x SSC for 5 min, and twice in 0.1x SSC for 5 min and dried by centrifugation. Shaded from light, the Cy3 and Cy5 fluorescent molecules (3DNA capture reagent, Genisphere) were hybridized in formamide buffer for 3 h at 49°C to corresponding capture sequences on cDNAs bound to the arrays. Arrays were washed in the dark with SSC containing 0.1 M DTT and dried as described earlier.
Scanned images (5 µm) were acquired with ScanArray Express (PerkinElmer, Boston, MA) at an excitation of 543 nm for Cy3 and 633 for Cy5 and at 90% power. The photomultiplier tube (PMT) settings for each fluor were set based on intensity of spiked internal Arabidopsis controls to normalize among all slides in the experiment. Image files were quantified in QuantArray (PerkinElmer) and raw mean signal and background values were exported to BioArray Software Environment (BASE) for analysis. Data were background subtracted and normalized by LOWESS, which is recommended for two-color experiments to eliminate dye-related artifacts and produce ratios that are not affected by signal intensity values (Supplementary Table 1). Stringent criteria were used to filter for genes that were regulated at least two-fold compared to vehicle controls consistently in all features (n = 8 per treatment) from biological replicates, dye-swapped technical replicates and duplicate spots printed on arrays. The genes that met these criteria were minimally categorized based on function using Gene Ontology and OMIM databases for putative homolog descriptions. Hierarchical clustering of gene expression profiles was performed with the agglomerative hierarchical clustering method provided in BASE using weighted (center of mass) averaging. Pearson correlation coefficients were calculated in GraphPad Prism (GraphPad Software, San Diego, CA), and Venn diagrams were created with Array File Maker 4.0.
|
Real time qRT-PCR.
To confirm results from microarray analysis, the expression of some genes was also analyzed by real time qRT-PCR. Total RNA was isolated as described previously and was treated with DNase (Invitrogen) according to manufacturer's protocol. cDNA was synthesized from 2 µg RNA with an oligo (dT)18 primer using SuperScript II (Invitrogen) following manufacturer's instructions with a final volume of 100 µl. Synthesized cDNAs (1 µl) were used as templates for amplification of specific gene products in total volumes of 20 µl containing 1x SYBR Green master mix (DyNAmo qPCR kit, Finnzymes, Finland) and 0.3 µM of each primer. Primer sequences are listed in Table 1. Primer sequences were chosen so that the product was contained in the array cDNA sequence to ensure validation of the microarray experiment. PCR was performed using a DNA Engine Cycler and Opticon 2 Detector (MJ Research, Waltham, MA). PCR was carried out for 40 cycles with denaturation at 94°C for 10 s, annealing at optimum temperature for primers (56–58°C) for 20 s, and extension at 72°C for 12 s. DNA amplification was quantified (pg) from the C(T) value based on standard curves to ensure quantification was within a linear range. Standards were created from gel-purified PCR products (QIAX II, Qiagen, Valencia, CA) for each primer set after quantification with PicoGreen dsDNA Quantification Kit (Molecular Probes, Eugene, OR) and serial dilutions ranging from 0.25 to 100 ng DNA. All signals were normalized against ß-actin, and ratios were calculated for treated samples compared to vehicle control as for the microarray analysis. Expression of ß-actin was not altered by treatment based on either microarray analysis or RT-PCR and so was found to be an appropriate housekeeping gene for normalization in this study.
Subcellular fractionation and immunoblot analysis.
Microsomal and cytosolic fractions were prepared from individual livers as described previously (Shilling and Williams, 2001). Protein concentrations were determined by the BioRad protein assay (Hercules, CA). CYP1A and zona radiata (ZR) were detected in liver microsomes and cytosol, respectively. Each sample (10 µg protein) was separated on NuPAGE 3–8% Tris-acetate polyacrylamide gels (Invitrogen) by electrophoresis and transferred to PVDF membranes. Membranes were incubated in BSA block buffer (2% BSA in PBS, pH 7.4) for 1 h at room temperature. Blots were probed with CYP1A mouse anti-trout monoclonal clone C10-7 (1:500 dilution) and ZR rabbit anti-salmon polyclonal clone O-146 (1:1000 dilution; Biosense, Bergen, Norway) for 1 h at room temperature. Membranes were washed four times for 5 min in Tween buffer (0.05% Tween-20 in PBS, pH 7.4). Membranes were incubated in the appropriate anti-mouse or anti-rabbit secondary horseradish peroxidase-conjugated antibodies (1:500; BioRad) for 1 h at room temperature and washed again in Tween buffer. Peroxidase activity was detected using Western Lighting Chemiluminescence Reagent (PerkinElmer) according to the manufacturer's instructions. Bands were visualized using an Alpha Image 1220 Documentation and Analysis System (Alpha Innotech, San Leandro, CA) and quantified as percent above control with Scion Image software (Frederick, MD).
Quantification of VTG by ELISA.
Trout liver cytosol was prepared as described above and quantification of VTG was based on an ELISA previously described (Donohoe and Curtis, 1996; Shilling and Williams, 2001). Briefly, cytosol samples were incubated in 96-well plates at 4°C for 24 h with rabbit anti-chum salmon VTG (1:1500), which was graciously provided by A. Hara at Hokkaido University. Samples were transferred to plates coated with 25 ng/well purified rainbow trout VTG (pre-blocked with 1% BSA) and incubated for 24 h at 4°C. Plates were then incubated with biotin-linked donkey anti-rabbit IgG and streptavidin horseradish peroxidase conjugate (Amersham, Buckinghamshire, England) for 2 h at 37°C and developed with 0.01% 3,3'5,5'-tetramethylbenzidine and 0.01% hydrogen peroxide in 0.5 M sodium acetate, pH 6.0. Colorimetric reactions were stopped after 10 min with 2 M sulfuric acid, and optical density was measured on a SpectraMax 190 plate reader with SoftMax Pro 4.0 software (Molecular Devices, Sunnyvale, CA). VTG concentrations were determined based on comparison to a trout VTG standard curve with a detection limit for this assay of 6.25 ng/ml. VTG was normalized to protein concentration for each sample, and ratios were calculated for treated samples compared to vehicle control similar to microarray analysis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Gene Expression Profiles by I3C, DIM, E2, and BNF
In this study, we determined the relative importance of ER- and AhR-mediated pathways in the mechanism of action of indole phytochemicals in trout by examining hepatic gene expression profiles after dietary exposure to I3C and DIM. Changes in gene expression were analyzed using salmonid cDNA microarrays (GRASP3.7kv.1) to characterize the effects of I3C and DIM in comparison to E2 and ßNF. As described in Material and Methods, two replicates of pooled RNA from six treated animals were hybridized to arrays with dye-swapping. The relationship of gene expression profiles among the different treatments were examined in scatter-plot graphs in which a correlation coefficient (r value) was calculated for each graph based on the linear regression between two profiles (Fig. 2). Pairwise analysis of all 8,736 features on the array indicated high correlations between E2 and 500 ppm DIM, 1500 ppm DIM, and 1500 ppm I3C, r = 0.77, 0.73, and 0.73, respectively (Fig. 2, panels A–C). Comparison of the DIM and I3C treatments resulted in the highest correlation coefficient of r = 0.84 (Fig. 2E), which would be expected since DIM is the primary gastric oligomerization product of I3C after dietary consumption (Dashwood et al., 1989
). In contrast, pairwise analysis suggested a low degree of similarity in gene expression patterns between ßNF and the dietary indoles, r = 0.07–0.09 (Fig. 2, panels G and H), which were similar to that for comparison of ßNF to E2, r = 0.07 (Fig. 2F).
|
Genes were considered differentially expressed if they were
2.0- or 0.5-fold changed compared to vehicle control in all intra- and interarray technical replicates and in both biological replicates for a treatment. Genes that passed the stringency filter are listed in Table 2. Gene descriptions are provided based on sequence homology using the most significant (E < 10–6) BLASTX hit against the current GenBank databases. Supplementary Table 2 lists the E-values and degree of similarity (length and percent identity over aligned region) between salmonid cDNA expressed sequence tags (EST) and the top BLASTX hit. If a salmonid EST had no significant BLASTX hit, then the top BLASTN is listed. Many of the genes listed in Table 2 are known trout genes, however others are only putative homologs based on sequence identity.
|
Hierarchical clustering was used as a visualization tool to identify similarities among biological replicates within a treatment and differences in gene expression between treatments (Fig. 3). Bidirectional hierarchical clustering of genes differentially regulated in at least one treatment group also indicated that there was a high degree of similarity in gene expression patterns among I3C, DIM, and E2 treatments (Fig. 3B). Treatments that clustered together on node II included 500 ppm DIM, 1500 ppm DIM and I3C, and 5 ppm E2. This supports the similarities observed between indoles and E2 by pairwise correlation analysis. Interestingly, ßNF clustered with 500 ppm I3C in node I separately from most other indole treatments; however, this is more likely due to the low number of genes differentially regulated overall by these two treatments than to similarity between ßNF and 500 ppm I3C. The 500 ppm I3C treatment did not have a strong correlation with E2 by pairwise analysis (r = 0.49; Fig. 2D), however, all genes differentially regulated by 500 ppm I3C were also regulated by E2, whereas none were regulated similarly to ßNF (Table 2 and Fig. 4). The only two array features differentially regulated by ßNF were for different cDNAs of CYP1A, which were also similarly regulated by the high dose of I3C.
|
|
Of the 38 cDNAs differentially regulated at least 2-fold by E2, 87% or 92% were similarly regulated by DIM, depending on dose, and 71% were regulated by I3C (Fig. 4). Further, all cDNAs regulated 2-fold by E2, except for thioredoxin, were also regulated at least 1.5-fold by either DIM or I3C, suggesting a common mechanism of action in trout liver. Transcripts encoding vitellogenic liver proteins were the most sensitive markers for the estrogenic response in trout, with expression profiles for VTG 13 to 23-fold above controls by microarray analysis and 250 to 1000-fold by qRT-PCR (Table 2). Other upregulated genes include those involved in cell proliferation, protein stability, and transport. Genes commonly downregulated by these treatments include those important for lipid, glucose, and retinol metabolism, immune regulation, and angiogenesis. While most cDNAs altered by E2 were also altered by treatment with DIM and I3C, there were some treatment-specific effects by the dietary indoles in trout liver (Fig. 4). The majority of these include genes involved in immune function and acute phase response that were downregulated by DIM and I3C and were not differentially regulated by E2, many of which were represented by multiple cDNAs on the array. In cases where there were multiple entries for the same gene, the gene was only entered once in Table 2 unless there were differences in treatment-related responses. It is interesting that most nonvitellogenic genes were almost always more strongly regulated by DIM than E2 based on fold-change values by microarray and qRT-PCR (Table 2, Fig. 5). For purposes of comparison, the concentrations of E2 and ßNF were chosen based on their ability to promote tumors and maximally induce VTG and CYP1A, respectively, which was confirmed in this study. It is apparent, however, that not all estrogen-responsive genes were equally regulated by concentrations that maximally expressed VTG, because it was such a sensitive marker supporting the conservative nature of the comparison.
|
), ßNF (•), DIM (
), and I3C (
) measured by microarray and real-time RT-PCR. Values are expressed as fold change (log2) compared to vehicle-treated control for select genes including cytochrome P4501A (CYP1A), C-type lectin 2–2 (CTL2–2), vitellogenin (VTG), serine-threonine kinase (STK), cathepsin D (CTSD), thioredoxin (TRX), and apolipoprotein B (APOB).Microarray Confirmation by qRT-PCR and Immunoassay
The expression profiles of select genes that were found to be differentially regulated by some treatments, including CYP1A, VTG, C-type lectin 2–2 (CTL2–2), serine/threonine kinase (STK), cathepsin D (CTSD), thioredoxin (TRX), and apolipoprotein B (APOB), were confirmed for all treatments by qRT-PCR using SYBR Green (Fig. 5). Overall, gene expression profiles measured by qRT-PCR confirmed those measured by cDNA array analysis. However, qRT-PCR was more sensitive in several cases than microarray analysis and detected greater changes. In some instances, genes that were not differentially regulated by certain treatments as measured by microarray analysis were found to be differentially regulated at least two-fold by qRT-PCR. For example, CYP1A was only upregulated by 1500 ppm I3C, and ßNF by microarray analysis. However, qRT-PCR analysis of cDNAs did result in greater than two-fold detection of CYP1A for both concentrations of I3C and DIM (Fig. 5). Similarly, E2 treatment even caused an unexpected two-fold upregulation of CYP1A as determined by qRT-PCR. This indicates there were some sensitivity differences between the two methods, and microarray analysis is likely much more conservative at detecting changes than qRT-PCR. Some genes were also confirmed by examining corresponding protein induction for CYP1A, VTG, and zona radiata (ZR), also known as vitelline envelope, by immunoassay (Fig. 6). These proteins were found to correlate well with transcript profiles measured by microarray analysis.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
We have previously reported that I3C promotes AFB1-induced hepatocarcinogenesis post-initiation at concentrations in the diet that were able to induce VTG, but not CYP1A (Oganesian et al., 1999
). These data suggest that estrogenic mechanisms may be important for promotion by indoles, particularly at lower dietary levels. VTG and CYP1A are frequently used as markers for activation of ER and AhR-mediated pathways, respectively, in fish and other models. The mechanisms of action for I3C and DIM have been found to involve both pathways; however, the relative importance of either in trout liver has not been evaluated on a global scale. The purpose of this study was to examine hepatic gene expression profiles after dietary exposure to two indole supplements, I3C and DIM, compared to E2, an ER agonist, and ßNF, an AhR agonist. We demonstrate that I3C and DIM acted similar to E2 at the transcriptional level based on correlation analysis of expression profiles and on clustering of gene responses. Of all the genes two-fold differentially regulated by E2, approximately 87–92% were also similarly regulated by DIM and 71% by I3C. The correlations are likely conservative based on the stringent criteria used to determine differential regulation by array analysis and the lower sensitivity of microarray results observed in comparison to qRT-PCR. These data highlight the strong overlap in transcriptional signatures of dietary indoles with endogenous E2 and suggest that the promotional ability of I3C in trout is through estrogenic mechanisms.
Overall, transcripts encoding vitellogenic liver proteins were the most sensitive markers for the estrogenic response in trout. This is similar to other teleost microarray studies in which VTG and egg envelope proteins were the most responsive hepatic genes regulated after in vivo exposure to estrogenic compounds (Larkin et al., 2002, 2003). The VTG response in trout is also confirmed by prior studies that found DIM induced VTG protein with similar efficacy as E2, although with approximately 200-fold less potency than E2 and 5-fold greater potency than I3C (Shilling and Williams, 2001). Our data show that DIM and I3C were able to induce an estrogenic response at the transcriptional level with similar efficacy to E2, and that DIM was more potent than I3C in vivo, also supporting its role as the active in vivo component of I3C (Anderton et al., 2004; Dashwood et al., 1989; Stresser et al., 1995b).
It was interesting that treatment with ßNF resulted in upregulation of only CYP1A and that this transcript was also upregulated above control levels by all treatments in this study, including E2, as determined by qRT-PCR. Therefore, it is possible that AhR-mediated pathways may also be relevant at lower concentrations of dietary indoles and that cross-talk between AhR and ER-mediated mechanisms is involved. Cross-talk between these pathways has been observed previously and suggests AhR agonists inhibit ER-mediated signaling (Safe et al., 1998). Antagonism was observed with ßNF, which upregulated CYP1A and downregulated VTG as measured by qRT-PCR. The significance of cross-talk in this study is in need of further research. However, the fact that dietary indoles so strongly mimicked E2 and did not result in antagonism of ER-mediated transcripts with upregulation of CYP1A further supports their similarities to E2 compared to ßNF.
We observed consistent downregulation of genes involved in redox regulation and lipid, glucose, and retinol homeostasis and metabolism by estrogenic treatments. Similar downregulation of these genes and gene classes in liver have been reported for rats treated with the potent estrogen and tumor promoter 17
-ethinylestradiol (Stahlberg et al., 2005). Also, an unexpected and consistent downregulation of genes important for angiogenesis, formation of the extracellular matrix, and immune response was measured after estrogenic treatment. Estrogens have been found to activate pro-angiogenic and matrix-membrane factors in certain cell types; however, the inhibitory effect of estrogens on inflammation in vascular endothelia and subsequent inhibition of chemoattractant and acute phase proteins is well documented (Koh, 2002). These signals directly regulate cell adhesion molecules, such as tissue factor, and other vascularization components and may be important for regulation in trout liver endothelial cells. Downregulation of other angiogenic and matrix-related factors in liver after in vivo exposure to estrogens has been documented previously (Larkin et al., 2003; Stahlberg et al., 2005). These data indicate evolutionary conservation of a dual role for estrogens in which they can simultaneously stimulate tumor growth, but may also inhibit tumor invasion and motility (Platet et al., 2004).
The liver is a major metabolic organ responsive to estrogens, which are known hepatic tumor promoters in a number of animal models (Nunez et al., 1989; Yager and Liehr, 1996). The transcriptional profiles in this study for E2, DIM, and I3C provide insight into the estrogenic mechanisms that may be important for promotion in trout liver including cell proliferation, signaling pathways, and protein stability. Many genes involved in protein folding, stability, and transport were upregulated by estrogens in this study including protein disulfide isomerase (PDI), heat shock proteins, peptidylpropyl cis-trans isomerase (PPI), cathepsin D, and translocation proteins. Some genes, such as cathepsin D, are known estrogen-responsive targets. Cathepsin D is a lysosomal protease important for yolk formation in oviparous animals, but the human form also contains an estrogen-responsive element in its promoter and is upregulated by E2 (Cavailles et al., 1991; Kwon et al., 2001). Other genes regulated in this study may be involved in the stability or formation of the ER complex for interaction with the DNA binding domains of target genes. PDI encodes a protein that is involved in enhancement of ER transcriptional activity by stabilizing DNA binding and has also been observed to be transcriptionally upregulated by E2 exposure in human, rat, and fish models (Bowman et al., 2002; Yoshikawa et al., 2000). Hsp108 was previously found in oviparous animals and has a high homology with chaperone hsp90, which is part of a large molecular complex that stabilizes unliganded ER in the cytoplasm and protects it from proteosomal degradation (Kulomaa et al., 1986). Other members of this ER multi-complex include PPI immunophilins and accessory chaperone hsp40 (Carrello et al., 2004). Transcriptional upregulation of both ER chaperone proteins and downstream targets indicate the responses to E2 and dietary indoles are ER mediated.
DIM has previously been found to be estrogenic in vitro through strong ligand-independent activation of ER that is mediated by PKA and MAPK signaling pathways (Leong et al., 2004; Riby et al., 2000). Recent evidence suggests that MAPK and other serine/threonine kinases also mediate estrogenic signaling in trout liver, which may result in ligand-independent activation of ER and subsequent gene transcription (Kullman et al., 2003). We observed upregulation of several kinases by E2, I3C, and DIM treatments in this study, including serine/threonine protein kinase (similar to PAS kinase) and NM23-H2/nucleoside diphosphate kinase B (NDPK-B), indicating that it is possible other intracellular signaling cascades may be activated by estrogens in trout liver. PAS kinase is a novel PAS domain-containing serine/threonine kinase in eukaryotes; however, it is highly conserved throughout evolution (Rutter et al., 2001). The signaling mechanisms of this recently identified kinase are not well understood, but PAS domains regulate the function of many intracellular signaling pathways in response to intrinsic and extrinsic stimuli including redox and nutrient status (Lindsley and Rutter, 2004). Also, NDPK-B, which is part of a class of NDPKs originally identified as housekeeping enzymes for synthesis of nucleoside triphosphates, has more recently been found to function in signal transduction and gene expression, including mediating G-protein activation of cell surface receptors and transcriptionally activating c-myc proto-oncogenes (Otero, 2000). NDPK expression is inversely correlated with tumor metastatic potential in a number of cancers including hepatocellular carcinoma (Bei et al., 1998), but can be upregulated by estrogen via ER-alpha activation in vitro and by other tumor promoters during neoplastic transformation of epidermal cells (Shah et al., 2004; Wei et al., 2004). Taken together, these data suggest that kinase signaling pathways more typically associated with growth factors may be relevant for estrogenic responses, including dietary indoles, in trout liver. We are currently working to determine the relative importance of these mechanisms in trout by examining post-translational effects of E2 and indoles on kinase signaling pathways in liver.
In summary, we demonstrate that indole phytochemicals, I3C and DIM, acted similar to E2 at the transcriptional level based on correlation analysis of expression profiles and on clustering of gene responses. The transcriptional profiles in this study for E2, DIM, and I3C provide insight into the mechanisms that may be important for promotion by estrogens in trout liver, including genes involved in cell proliferation, signaling pathways, and protein stability. Downregulation of transcripts for immune regulation, angiogenesis, and cell adhesion indicate the possibility of estrogens and indoles having some protective effects against tumor invasion and metastasis. This data suggests that DIM may also promote hepatocarcinogenesis in trout by estrogenic mechanisms.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Supplementary data are available online at www.toxsci.oxfordjournals.org. These files are also available through Gene Expression omnibus accession #GSE3324.
|
ACKNOWLEDGMENTS |
|---|
The authors wish to thank Eric Johnson and Greg Gonnerman for care and maintenance of fish and Caprice Rosato for array technical assistance. This work was supported by NIH grants ES07060, ES03850, ES00210, ES11267, and CA90890.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Anderton, M. J., Manson, M. M., Vershoyle, R. D., Gescher, A., Lamb, J. H., Farmer, P. B., Steward, W. P., and Williams, M. L. (2004). Pharmacokinetics and tissue disposition of indole-3-carbinol and its acid condensation products after oral administration to mice. Clin. Cancer Res. 10, 5233–5241.
Bailey, G. S., Dashwood, R. H., Fong, A. T., Williams, D. E., Scanlan, R. A., and Hendricks, J. D. (1991). Modulation of mycotoxin and nitrosamine carcinogensis by indole-3-carbinol: Quantitative analysis of inhibition versus promotion. In Relevance to Human Cancer of N-Nitroso Compounds. (I. K. O'Neill, J. Chen, and H. Bartsch, Eds.), pp. 275–280. IARC Press, Lyon, France.
Bailey, G. S., Goeger, D. E., and Hendricks, J. D. (1989). Factors influencing experimental carcinogenesis in laboratory fish models. In Metabolism of Polynuclear Hydrocarbons in the Aquatic Environment (U. Varanasi, Ed.), pp. 253–268. CRC Press, Boca Raton, FL.
Bei, H., Zhong-Bi, W., and You-Bing, R. (1998). Expression of nm23 gene expression in hepatocellular carcinoma tissue and its relation with metastasis. World J. Gastroenterol. 4, 266–267.[Medline]
Bowman, C. J., Kroll, K. J., Gross, T. G., and Denslow, N. D. (2002). Estradiol-induced gene expression in largemouth bass (Micropterus salmoides). Mol. Cell. Endocrinol. 196, 67–77.[CrossRef][ISI][Medline]
Carrello, A., Allan, R. K., Morgan, S. L., Owen, B. A., Mok, D., Ward, B. K., Minchin, R. F., Toft, D. O., and Ratajczak, T. (2004). Interaction of the Hsp90 cochaperone cyclophylin 40 with Hsc70. Cell Stress Chaperones. 9, 167–181.[CrossRef][ISI][Medline]
Cavailles, V., Augereau, P., and Rochefort, H. (1991). Cathepsin D gene of human MCF7 cells contains estrogen-responsive sequences in its 5' proximal flanking region. Biochem. Biophys. Res. Commun. 174, 816–824.[CrossRef][ISI][Medline]
Dashwood, R. H., Fong, A. T., Arbogast, D. N., Bjeldanes, L. F., Hendricks, J. D., and Bailey, G. S. (1994). Anticarcinogenic activity of indole-3-carbinol acid products: Ultrasensitive bioassay by trout embryo microinjection. Cancer Res. 54, 3617–3619.
Dashwood, R. H., Uyetake, L., Fong, A. T., Hendricks, J. D., and Bailey, G. S. (1989). In vivo disposition of the natural anti-carcinogen indole-3-carbinol after po administration to rainbow trout. Food. Chem. Toxicol. 27, 385–392.[CrossRef][ISI][Medline]
Donohoe, R. M., and Curtis, L. R. (1996). Estrogenic activity of chlordecone, o,p‘-DDT and o,p’-DDE in juvenile rainbow trout: Induction of vitellogenesis and interaction with hepatic estrogen binding sites. Aquat. Toxicol. 36, 31–52.
Grubbs, C. J., Steele, V. E., Casebolt, T., Juliana, M. M., Eto, I., Whitaker, L. M., Dragnev, K. H., Kelloff, G. J., and Lubert, R. L. (1995). Chemoprevention of chemically-induced mammary carcinogenesis by indole-3-carbinol. Anticancer Res. 15, 709–716.[ISI][Medline]
Kim, Y. S., and Milner, J. A. (2005). Targets for indole-3-carbinol in cancer prevention. J. Nutr. Biochem. 16, 65–73.[CrossRef][ISI][Medline]
Koh, K. K. (2002). Effects of estrogen on the vascular well: Vasomotor function and inflammation. Cardiovasc. Res. 55, 714–726.
Kojima, T., Tanaka, T., and Mori, H. (1994). Chemoprevention of spontaneous endometrial cancer in female Donryu rats by dietary indole-3-carbinol. Cancer Res. 54, 1446–1449.
Kullman, S. W., Matsamura, F., and Hinton, D. E. (2003). Estrogen signaling in trout liver: Estrogen receptor mitogenic mediated cellular responses. Environ. Sci. 10, 223–237.
Kulomaa, M. S., Weigel, N. L., Kleinsek, D. A., Beattie, W. G., Conneely, O. M., March, C., Zarucki-Schulz, T., Schrader, W. T., and O'Malley, B. W. (1986). Amino acid sequence of a chicken heat shock protein derived from the complementary DNA nucleotide sequence. Biochemistry 25, 6244–6251.[CrossRef][Medline]
Kwon, J. Y., Prat, F., Randall, C., and Tyler, C. R. (2001). Molecular characterization of putative yolk processing enzymes and their expression during oogenesis and embyogenesis in rainbow trout (Oncorhynchus mykiss). Biol. Reprod. 65, 1701–1709.
Larkin, P., Folmar, L. C., Hemmer, M. J., Poston, A. J., and Denslow, N. D. (2003). Expression profiling of estrogenic compounds using sheepshead minnow cDNA macroarray. Environ. Health Perspect. 111, 839–846.
Larkin, P., Sabo-Attwood, T., Kelso, J., and Denslow, N. D. (2002). Gene expression analysis of largemouth bass exposed to estradiol, nonylphenol, and p,p'-DDE. Comp. Biochem. Physiol. B. 133, 543–557.[CrossRef][Medline]
Lee, B. C., Hendricks, J. D., and Bailey, G. S. (1991). Toxicity of microtoxins in the feed of fish. In Mycotoxins and Animal Feedstuff: Natural Occurrence, Toxicity and Control (J. E. Smith, Ed.), pp. 607–626. CRC Press, Boca Raton, FL.
Leong, H., Riby, J. E., Firestone, G. L., and Bjeldanes, L. F. (2004). Potent ligand-independent estrogen receptor activation by 3,3'-diindolylmethane is mediated by cross-talk between protein kinase A and mitogen-activated protein kinase signaling pathways. Mol. Endocrinol. 18, 291–302.
Lindsley, J. E., and Rutter, J. (2004). Nutrient sensing and metabolic decisions. Comp. Biochem. Physiol. B. 139, 543–559.[CrossRef][Medline]
Lord, R. S., Bongiovanni, B., and Bralley, J. A. (2002). Estrogen metabolism and the diet-cancer connection: Rationale for assessing the ratio of urinary hydroxylated estrogen metabolites. Altern. Med. Rev. 7, 112–129.[Medline]
McDanell, R., McLean, A. E., Hanley, A. B., Heaney, R. K., and Fenwick, G. R. (1988). Chemical and biological properties of indole glucosinolates (glucobrassicin): A review. Food. Chem. Toxicol. 26, 59–70.[CrossRef][ISI][Medline]
Meng, Q., Yuan, F., Goldberg, I. D., Rosen, E. M., Auborn, K., and Fan, S. (2000). Indole-3-carbinol is a negative regulator of estrogen receptor- signaling in human tumor cells. J. Nutr. 130, 2927–2931.
Nunez, O., Hendricks, J. D., Arbogast, D. N., Fong, A. T., Lee, B. C., and Bailey, G. S. (1989). Promotion of aflatoxin B1 hepatocarcinogenesis in rainbow trout by 17ß-estradiol. Aquat. Toxicol. 15, 289–302.
Oganesian, A., Hendricks, J. D., Pereira, C. B., Orner, G. A., Bailey, G. S., and Williams, D. E. (1999). Potency of dietary indole-3-carbinol as a promoter of aflatoxin B1-induced hepatocarcinogenesis: Results from a 9000 animal tumor study. Carcinogenesis 20, 453–458.
Otero, A. S. (2000). NM23/Nucleoside diphosphate kinase and signal transduction. J. Bioenerget. Biomembranes 32, 269–275.[CrossRef][ISI][Medline]
Platet, N., Cathiard, A. M., Gleizes, M., and Garcia, M. (2004). Estrogens and their receptors in breast cancer progression: A dual role in cancer proliferation and invasion. Crit. Rev. Oncol. Hematol. 51, 55–67.[ISI][Medline]
Rees, C. B., and Li, G. (2004). Development and application of a real-time quantitative PCR assay for determining CYP1A transcripts in three genera of salmonids. Aquat. Toxicol. 66, 357–368.[CrossRef][ISI][Medline]
Riby, J. E., Chang, G. H. F., Firestone, G. L., and Bjeldanes, L. F. (2000). Ligand-independent activation of estrogen receptor function by 3,3'-diindolylmethane in human breast cancer cells. Biochem. Pharmacol. 60, 167–177.[CrossRef][ISI][Medline]
Rise, M. L., von Schalburg, K. R., Brown, G. D., Mawer, M. A., Devlin, R. H., Kuipers, N., Busby, M., Beetz-Sargent, M., Alberto, R., Gibbs, A. R., et al. (2004). Development and application of a salmonid EST database and cDNA microarray: Data mining and interspecific hybridization characteristics. Genome. Res. 14, 478–490.
Rutter, J., Michnoff, C. H., Harper, S. M., Gardner, K. H., and McKnight, S. L. (2001). PAS kinase: An evolutionarily conserved PAS domain-regulated serine/threonine kinase. Proc. Natl. Acad. Sci. U.S.A. 98, 8991–8996.
Safe, S., Wang, F., Porter, W., Duan, R., and McDougal, A. (1998). Ah receptor agonists are endocrine disruptors: Antiestrogenic activity and mechanisms. Toxicol. Lett. 102–103, 343–347.
Shah, Y. M., Basrur, V., and Rowan, B. G. (2004). Selective estrogen receptor modulator regulated proteins in endometrial cancer cells. Mol. Cell. Endocrinol. 219, 127–139.[CrossRef][ISI][Medline]
Shilling, A. D., and Williams, D. E. (2001). 3,3'-Diindolylmethane, a major condensation product of indole-3-carbinol, is a potent estrogen in rainbow trout. Toxicol. Appl. Pharmacol. 170, 191–200.[CrossRef][ISI][Medline]
Stahlberg, N., Merino, R., Hernandez, L. H., Fernandez-Perez, L., Sandelin, A., Engstrom, P., Tollet-Egnell, P., Lenhard, B., and Flores-Morales, A. (2005). Exploring hepatic hormone actions using a compilation of gene expression profiles. BMC Physiol. 5, 1472–6793.
Stoner, G., Casto, B., Ralston, S., Roebuck, B., Pereira, C., and Bailey, G. (2002). Development of a multi-organ rat model for evaluating chemopreventive agents: Efficacy of indole-3-carbinol. Carcinogenesis 23, 265–272.
Stresser, D. M., Bjeldanes, L. F., Bailey, G. S., and Williams, D. E. (1995a). The anticarcinogen 3,3-diindolylmethane is an inhibitor of cytochrome P-450. J. Biochem. Toxicol. 10, 191–201.[CrossRef][Medline]
Stresser, D. M., Williams, D. E., Griffin, D. A., and Bailey, G. S. (1995b). Mechanisms of tumor modulation by indole-3-carbinol: Disposition and excretion in male Fischer 344 rats. Drug Metab. Dispos. 23, 965–975.[Abstract]
Wei, S. J., Trempus, C. S., Ali, R. C., Hansen, L. A., and Tennant, R. W. (2004). 12-O-Tetradecanoylphorbol-13-acetate and UV radiation-induced nucleoside diphosphate kinase protein kinase B mediates neoplastic transformation of epidermal cells. J. Biol. Chem. 279, 5993–6004.
Yager, J. D., and Liehr, J. G. (1996). Molecular mechanisms of estrogen carcinogenesis. Annu. Rev. Pharmacol. Toxicol. 36, 203–232.[CrossRef][ISI][Medline]
Yoshida, M., Katashima, S., Ando, J., Tanaka, T., Uematsu, F., Nakae, D., and Maekawa, A. (2004). Dietary indole-3-carbinol promotes endometrial adenocarcinoma development in rats initiated with N-ethyl-N'-nitro-N-nitrosoguanidine, with induction of cytochrome P450s in the liver and consequent modulation of estrogen metabolism. Carcinogenesis 25, 2257–2264.
Yoshikawa, S., Kamada, M., Maegawa, M., Yamamoto, S., Irahara, M., Yamano, S., Aono, T., Kido, H., and Koide, S. S. (2000). Hormonal control of mRNA expression of immunoglobulin binding factor in uterine cervix. Biochem. Biophys. Res. Commun. 279, 898–903.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. D. Benninghoff and D. E. Williams Identification of a Transcriptional Fingerprint of Estrogen Exposure in Rainbow Trout Liver Toxicol. Sci., January 1, 2008; 101(1): 65 - 80. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Kassie, L. B. Anderson, R. Scherber, N. Yu, D. Lahti, P. Upadhyaya, and S. S. Hecht Indole-3-carbinol Inhibits 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone Plus Benzo(a)pyrene-Induced Lung Tumorigenesis in A/J Mice and Modulates Carcinogen-Induced Alterations in Protein Levels Cancer Res., July 1, 2007; 67(13): 6502 - 6511. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Tilton, J. D. Hendricks, G. A. Orner, C. B. Pereira, G. S. Bailey, and D. E. Williams Gene expression analysis during tumor enhancement by the dietary phytochemical, 3,3'-diindolylmethane, in rainbow trout Carcinogenesis, July 1, 2007; 28(7): 1589 - 1598. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||