Assessment of the Effects of Chemicals on the Expression of Ten Steroidogenic Genes in the H295R Cell Line Using Real-Time PCR
Klara Hilscherova,* Paul D. Jones,*1 Tannia Gracia,* John L. Newsted,; Xiaowei Zhang, *¿ J. T. Sanderson,§ Richard M. K. Yu,¿ Rudolf S. S. Wu,¿ and John P. Giesy ***
*Department of Zoology, National Food Safety and Toxicology Center, Center for Integrative Toxicology, Michigan State University, East Lansing, Michigan 48824; +ENTRIX Inc., East Lansing, Michigan 48864; ¿ Center for Coastal Pollution and Conservation City University of Hong Kong, Kowloon, Hong Kong, SAR China; and §Institute for Risk Assessment Sciences (IRAS), University of Utrecht, 3508 TD Utrecht, Netherlands
Received May 7, 2004; accepted June 3, 2004
The potential for a variety of environmental contaminants to disturb endocrine function in wildlife and humans has been of recent concern. While much effort is being focused on the assessment of effects mediated through steroid hormone receptor-based mechan- isms, there are potentially several other mechanisms that could lead to endocrine disruption. Recent studies have demonstrated that a variety of xenobiotics can alter the gene expression or activity of enzymes involved in steroidogenesis. By altering the production or catalytic activity of steroidogenic or steroid-catabolizing enzymes, these chemicals have the potential to alter the steroid balance in organisms. To assess the potential of chemicals to alter steroidogen- esis, an assay system was developed using a human adrenocortical carcinoma cell line, the H295R cell line, which retains the ability to synthesize most of the important steroidogenic enzymes. Methods were developed, optimized, and validated to measure the expression of 10 genes involved in steroidogenesis by the use of real-time quan- titative reverse transcriptase PCR. The effects of several model chemicals known to alter steroid metabolism, both inducers and inhibitors, were assessed. Similar expression patterns were observed for chemicals acting through common mechanisms of action. Time- course studies demonstrated distinct time-dependent expression profiles for chemicals able to modulate steroid metabolism. The assay, which allows simultaneous analysis of the expression of numerous steroidogenic enzymes, would be useful as a sensitive and integrative screen for the many effects of chemicals on steroidogenesis.
Key Words: steroidogenesis; bioassay; xenoestrogens; screening.
Recently there has been much interest in the effects of endo- crine disrupters on wildlife (Ankley et al., 1998) and humans (Kavlock et al., 1996). The Safe Drinking Water Act Amend- ments of 1995 and the Food Quality Protection Act of 1996 mandate screening for endocrine-disrupting properties of chemi- cals in drinking water or pesticides used in food production. In
response to this legislation, the federal Endocrine Disrupter Screening and Testing Advisory Committee (EDSTAC) recom- mended that chemicals be screened as agonists or antagonists of estrogen (ER), androgen (AR), and thyroid (ThR) hormone receptors (EDSTAC Final Report, 1998).
One type of endocrine disruption takes place when xenobio- tics mimic steroid hormones. Of particular concern have been those compounds that mimic endogenous estrogens, sometimes called xenoestrogens. While some reports indicated that endo- crine disruption functioned through this mechanism of action, subsequent studies have found that some compounds have more complex mechanisms of action. It has been observed that some compounds can bind to the androgen receptor and function as either androgen agonists or antagonists. Although the effects of endocrine-disrupting chemicals (EDCs) and methods to screen for them have focused on direct interactions with steroid hor- mone receptors such as ER, AR, and ThR, EDCs can operate several different ways. Firstly, there are several other receptor- mediated processes that control sexual development and homeo- stasis. Secondly, there are also some nonreceptor-mediated mechanisms. Finally, there are compounds that can modulate steroid hormone production or breakdown and cause endocrine disruption without acting as hormone mimics. These effects are often exerted indirectly via various effects on common signal transduction pathways or by acting on steroid metabolism pathways.
One such example is the effect of the herbicide atrazine. Atrazine has been observed to cause estrogenic effects both in vitro and in vivo but does not bind to the estrogen receptor (Connor et al., 1996; Sanderson et al., 1999, 2000, 2001). While the effects observed in vitro occurred at relatively great concen- trations, these results serve as an example of the types of effects that can be observed with in vitro tests. The family of 2-chloro- s-triazine herbicides had a common ability to induce the catalytic activity and mRNA levels of CYP19 using the H295R cell line as a steroidogenic model system (Sanderson et al., 2000, 2001). The H295R (a subpopulation of H295 that forms a monolayer in culture) human adrenocortical carcinoma cell line has been
1 To whom correspondence should be addressed at Michigan State University, 224 National Food Safety and Toxicology Center, East Lansing, MI 48824-1311. Fax: (517) 432-2310. E-mail: jonespa7@msu.edu.
Downloaded from http://toxsci.oxfordjournals.org/ at UNIVERSITY OF AUCKLAND LIBRARY on May 2, 2015
characterized in detail and shown to express most of the key enzymes involved in steroidogenesis (Gazdar et al., 1990; Rainey et al., 1993; Staels et al., 1993). Sanderson and coworkers suggested that the effects they observed in the H295R cells occurred by the inhibition of phosphodiesterase with a concomitant increase in cyclic-AMP. The model com- pound 8-bromo-c-AMP also resulted in the upregulation of CYP19 (aromatase) mRNA.
While this mechanism may not be operating in vivo at all times in all tissues of all species or at relevant environmental concen- trations, it is a plausible explanation for the observation that atrazine induced luciferase activity under the control of the ER in MVLN cells (MCF-7-luc, MVLN; Villeneuve et al., 1998). However, experiments demonstrating the expression of aromatase in this cell line have yielded equivocal results. Thus, in addition to other indirect mechanisms of action, it is possible that natural and synthetic chemicals can modulate the endocrine system by acting as direct or indirect stimulators or inhibitors of the enzymes involved in the production, transfor- mation, and or elimination of steroid hormones. Here we present a procedure for screening for the effects of chemicals on the profile of expression of steroidogenic genes. Specifically, we report methods to simultaneously measure mRNA concentra- tions for 10 steroidogenic enzymes and two housekeeping genes in cultured H295R cells.
The key genes measured in the current study include CYP11A (cholesterol side-chain cleavage); CYP11B1 (steroid 11ß- hydroxylase); CYP11B2 (aldosterone synthetase); CYP17 (steroid 17a-hydroxylase and/or 17,20 lyase); CYP19 (aroma- tase); 17฿HSD1,17฿HSD4, CYP21B2(steroid 21-hydroxylase), and 3฿HSD2 (3ß-hydroxysteroid dehydrogenase); HMGR (hydroxymethylgutaryl CoA reductase); and the cholesterol transfer protein StAR (steroid acute regulatory protein). The H295R cells used have the physiological characteristics of zon- ally undifferentiated human fetal adrenal cells, with the ability to produce the steroid hormones of each of the three phenotypically distinct zones found in the adult adrenal cortex (Fig. 1; Gazdar et al., 1990; Staels et al., 1993). Since the cells maintain the ability to express these genes and produce these enzymes, which might otherwise only be expressed in certain tissues or periods of ontogeny, they are a useful model system for potential effects on steroidogenesis.
MATERIALS AND METHODS
Forskolin, 8BrcAMP, Phorbol-12-myristate-13-acetate (PMA), lovastatin, ketoconazole, aminoglutethimide, androstenedione, and spironolactone were obtained from Sigma Chemical Co. (St. Louis, MO). Metyrapone was from Aldrich (St. Louis, MO), and daidzein was from ICN Biochemicals Inc. (Aurora, OH). The chemicals used in this study were chosen based on their variety of known effects on steroid metabolism. That is, aminoglutethimide is an aromatase inhibitor; lovastatin is metabolized to produce a specific hydroxymethylglutaryl- CoA reductase (HMGR) inhibitor; 8BrcAMP and forskolin increase cellular cAMP concentrations; PMA is a diacylglycerol analogue that activates protein
Cholesterol
CYP11A
CYP17
Pregnenolone
17a-OH-
CYP17
Pregnenolone
DHEA
3B-HSD
3B-HSD
3B-HSD
Progesterone
CYP17
17a-OH-
CYP17
Androstene
CYP21
Progesterone
-dione
11-Deoxy-
CYP21
corticosterone
17B-HSD
CYP11B1
11-Deoxycortisol
Testosterone
Corticosterone
CYP11B1
CYP19
CYP11B2
Cortisol
17 B-Estradiol
Aldosterone
Zona glomerulosa
Zona reticularis
Zona
fasciculata
kinase C; ketoconazole works principally by the inhibition of cytochrome P450 14a-demethylase (P45014DM); and daidzein is a weak estrogen receptor agonist.
The H295R human adrenocortical carcinoma cell lines were obtained from the American Type Culture Collection (ATCC CRL-2128; ATCC, Manassas, VA) and were grown in 75 cm2 flasks with 12.5 ml of supplemented medium at 37°℃ with a 5% CO2 atmosphere. Supplemented medium was a 1:1 mixture of Dulbecco’s modified Eagle’s medium with Ham’s F-12 Nutrient mixture with 15 mM HEPES buffer. The medium was supplemented with 1.2 g/l Na2CO3, ITS + Premix (1 ml Premix/100 ml medium), and 12.5 ml/ 500 ml NuSerum (BD Bioscience, San Jose, CA). Final component concentra- tions in the medium were as follows: 15 mM HEPES, 6.25 µg/ml insulin, 6.25 µg/ml transferrin, 6.25 ng/ml selenium, 1.25 mg/ml bovine serum albumin, 5.35 µg/ml linoleic acid, and 2.5% NuSerum. The medium was changed two to three times per week and cells were detached from flasks for subculturing by use of trypsin/EDTA (Sterile 1X Trypsin-EDTA; Life Technologies Inc., Grand Island, NY). Cells were exposed to chemicals of interest in 6-well Tissue Culture Plates (Nalgene Nunc Inc., Rochester, NY). Cells were dosed with chemicals dissolved in DMSO for 48-72 h after plating.
RNA isolation. Before nucleic acid isolation and analysis, cell viability was determined. Cells were visually inspected under a microscope to evaluate via- bility and cell numbers. Also, cell viability was determined with the Live/Dead cell viability kit (Molecular Probes, Eugene, OR). Cell death was only observed for 17a-Ethynylestradiol and lovastatin at concentrations greater than 30 uM; ketoconazole and cyproterone acetate inhibited cell growth at concentrations greater than 30 µM. No adverse effects on cell growth or viability were observed for any of the tested chemicals at maximum concentrations ranging from 30 to 100 uM. Exposures in which either cell death or decreased viability was observed were not used for gene expression analysis.
After removal of the medium, cells were lysed in the culture plate by the addition of 580 ul/well of Lysis Buffer-ß-ME mixture (Stratagene, La Jolla, CA). Cells were mixed and collected by repeated pipetting and transferred to a micro- centrifuge tube that was mixed to homogenize and ensure low viscosity of the lysate. After mixing, the homogenate was transferred to a prefilter spin cup seated in a 2-ml tube and was centrifuged in a microcentrifuge for 5 min. The spin cup was removed from the receptacle tube and discarded. For RNA isolation, 700 ul of 70% ethanol was added to the filtrate and the tube was vortexed to mix thoroughly. Half of the mixture was transferred to an RNA binding spin cup seated in a fresh 2-ml tube and this was then centrifuged for 1 min. The spin cup was removed and retained and the filtrate was discarded. This procedure was repeated with the same spin cup using the second half of the sample.
To remove residual DNA prior to reverse transcription, DNase treatment was used; 600 ul of 1 X low-salt wash buffer were added to the spin cup containing the RNA, this was centrifuged for 1 min, and the filtrate was discarded. Next, 55 ul of RNase Free-DNase I solution (Stratagene) were added to the fiber matrix inside the spin cup. The sample was incubated at 37℃ for 15 min. The sample was then washed with 600 ul of 1X high-salt wash buffer and 600 ul of 1X low-salt wash buffer, centrifuged at maximum speed for 30-60 s and discarding the filtrate after each wash. A final wash was done by adding 300 ul of 1X low-salt wash buffer to the spin cup, and the tube was centrifuged for 2 min to dry the fiber matrix. The spin cup was transferred to a fresh 1.5-ml microcentrifuge tube and 80 ul of nuclease-free water was added directly onto the center of the fiber matrix inside the spin cup. The tube was incubated for 2 min at room temperature before centrifugation for 1 min. This elution step was repeated to maximize the yield of RNA. The purified RNA was used immediately for RT-PCR or was stored at -80°℃ until analysis.
An appropriate dilution of the RNA sample (1:50) was prepared for RNA quantitation. The absorbance of the RNA solution was measured at 260 and 280 nm and the 260/280 ratio was calculated. The concentration of total RNA was estimated using the A260 value and a standard with an A260 of 1 that was equivalent to 40 µg RNA/ml.
cDNA preparation. Total RNA (1-5 µg) was combined with 50 uM oligo-(dT)20 and 10 mM dNTPs diethylpyrocarbamate- (DEPC-) treated water to a final volume of 12 ul. RNA and primers were denatured at 65℃ for 5 min and then incubated on ice for 5 min. Reverse transcription was performed using 8 ul of a master mix containing the following: 5X cDNA synthesis buffer, 0.1 M DTT, RNase OUT 40 U/ul, Cloned AMV Reverse Tran- scriptase (Invitrogen, Carlsbad, CA), and DEPC-treated water. Reactions were incubated at 50℃ for 45 min and were terminated by incubation at 85℃ for 5 min. Samples were either used directly for PCR or were stored at -20℃ until analysis.
Real-time PCR. Real-time PCR (quantitative PCR) was performed by using a Smart Cycler System (Cepheid, Sunnyvale, CA) in 25-ul sterile tubes using a master mix containing the following: 25 mM MgCl2, 1 U/ul AmpErase (Applied Biosystems, Foster City, CA), 5 U/ul Taq DNA polymerase AmpliTaq Gold, 10X SYBR Green (PE Biosystems, Warrington, UK), nuclease-free water, and between 10 pg and 1 µg of cDNA. The Thermal Cycling program was 94℃ for 10 min as follows: 50-60℃ for 30 s to 1 min; 68-72℃ for 1 min/kb followed by 35-40 cycles of 94℃ for 15-40 s; 50-60℃ for 30 s to 1 min; 68-72℃ for 1 min/kb; and a final cycle of 94℃ for 15-40 s, 50-60°℃ for 30 s to 1 min, and 72℃ for 5-10 min. Melting curve analyses were performed immediately following the final PCR cycle to differentiate between the desired amplicons and any primer- dimers or DNA contaminants.
For quantification of PCR results, Ct (the cycle at which the fluorescence signal is first significantly different from background) was determined for each reaction. Ct values for each gene of interest were normalized by division by the Ct for the endogenous control gene to produce ACt. Therefore, the difference between AC, values for a control and a chemically exposed culture (designated AAC) represent the degree of induction or inhibition of the gene of interest. Moreover, the degree of induction or inhibition can be calculated as a fold difference using the following relationship:
Xexp/Xcon =2-44Ct
where Xexp and Xcon represent the degree of expression in the exposed and control samples, respectively, and Xexp/Xcon, therefore, represents the fold induction. All data are reported and were statistically analyzed as fold induction between exposed and control cultures. Gene expression was measured at least in triplicate for each control or exposed cell culture and each exposure was repeated at least three times.
Statistical analysis. Statistical analyses of gene expression profiles were conducted using SYSTAT 10 (SPSS Inc., Chicago, IL). Differences in gene expression were evaluated by ANOVA followed by Tukey’s test. Differences with p ≤0.05 were considered significant. Statistical analysis of sequence homo- logies between amplicons and the GenBank database were conducted using the BLAST algorithm on the National Center for Biotechnology Information website (http://www.ncbi.nlm.nih.gov/).
RESULTS
PCR Assay Procedures
Quantitative PCR (Q-RT-PCR) conditions, including sense and antisense primers, temperatures, times, and reagent concen- trations were optimized for all the steroidogenic genes (Table 1). Each amplicon yielded a single peak when the melting tempera- ture curve was analyzed at the conclusion of the PCR reaction (Fig. 2). To further confirm the identities of the amplified sequences, the PCR products were analyzed by agarose gel electrophoresis (Fig. 3). After optimization, each PCR reaction produced a single amplicon of the expected size. No additional bands or excessive levels of primer-dimer products were
| Gene | Product length | Annealing ºC (s)* | Primer concentration (µM) | Sense primer | Antisense primer |
|---|---|---|---|---|---|
| 18S rRNA | 124 | 62 (60) | 0.4 | CGTCTGCCCTATCAACTTTCG | TGCCTTCCTTGGATGTGGTAG |
| B-actin | 100 | 64 (60) | 0.2 | CACTCTTCCAGCCTTCCTTCC | AGGTCTTTGCGGATGTCCAC |
| CYP11A | 137 | 62 (50) | 0.4 | GAGATGGCACGCAACCTGAAG | CTTAGTGTCTCCTTGATGCTGGC |
| CYP11B2 | 146 | 62 (50) | 0.2 | TCCAGGTGTGTTCAGTAGTTCC | GAAGCCATCTCTGAGGTCTGTG |
| CYP17 | 134 | 64 (60) | 0.6 | AGCCGCACACCAACTATCAG | TCACCGATGCTGGAGTCAAC |
| CYP19 | 128 | 64 (50) | 0.4 | AGGTGCTATTGGTCATCTGCTC | TGGTGGAATCGGGTCTTTATGG |
| CYP21 | 108 | 64 (50) | 0.4 | CGTGGTGCTGACCCGACTG | GGCTGCATCTTGAGGATGACAC |
| 3฿HSD2 | 95 | 60 (50) | 0.4 | TGCCAGTCTTCATCTACACCAG | TTCCAGAGGCTCTTCTTCGTG |
| 17₿HSD1 | 136 | 64 (60) | 0.4 | CTCCCTCTGACCAGCAACC | TGTGTCTCCCACGCAATCTC |
| 17฿HSD4 | 121 | 62 (50) | 0.4 | TGCGGGATCACGGATGACTC | GCCACCATTCTCCTCACAACTC |
| StAR | 168 | 64 (40) | 0.4 | GTCCCACCCTGCCTCTGAAG | CATACTCTAAACACGAACCCCACC |
| HMGR | 152 | 60 (50) | 0.4 | TGCTTGCCGAGCCTAATGAAAG | AGAGCGTTCGTGGGTCCATC |
*All PCR reactions were extended at 72℃ for 30 s and denatured at 95℃ for 15 s.
detected in any of the amplified DNA samples. To definitively confirm the identity of the amplicons, the DNA sequence of each band was determined (Table 2). During amplicon sequencing, the initial sequence determination (i.e., the first 20-30 base pairs) can be unclear and this low-quality sequence was identi- fied by the sequencing facility (Michigan State University, Macromolecular Structure Facility, personal communication). Thus, only the middle portion of the sequence is of sufficient
250
Fluoresence (1st derivative)
200
150
100
50
0
55
65
75
85
95
Temperature (C)
quality to match. This is why the sequences determined are not the full length of the amplicon.
The sequence of the amplicons showed a minimum of 89% homology to the desired target sequence and a minimum significance value (E-value) of 9 × 10-8. The Expect value (E) is a parameter that describes the number of hits one can expect to observe by chance when searching a database of a particular size. The value decreases exponentially with the Score (S) that is assigned to a match between two sequences. The E value describes the random background noise that exists for matches between sequences. For example, an E value of 1 assigned to a “hit” can be interpreted as meaning that, in a database of the current size, one might expect to see one match with a similar score simply by chance. Hence, the smaller the E-value or the closer it is to 0, the more significant the match. The BLAST programs report E values rather than p values because it is easier to interpret the difference between, for example, E values of 5 and 10 than p values of 0.993 and 0.99995. However, when E < 0.01, p values and E values are nearly identical.
Due to the relatively short length of the amplified DNA, some bands could not be sequenced. While some differences were detected from the published sequences, these differ- ences are not likely to be significant given that only a single sequence determination was conducted and the possibility that genetic variants different from the published sequences could occur.
| Gene | MW marker | 18sRNA | CYP11B2 | 17ßHSD4 | CYP11A | 3BHSD2 | CYP19 | HMGR | CYP21 | 178HSD1 | STAR | CYP11B1 | 17ßHSD2 | 17ßHSD4 | 17ßHSD1 | B-actin | 178HSD2 | MW marker | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Size (bases) | 124 | 146 | 121 | 137 | 95 | 128 | 152 | 108 | 136 | 168 | 147 | 135 | 121 | 136 | 100 | 135 |
| Target | Amplicon sequence* | Identities | E value |
|---|---|---|---|
| 18S | CGGGGAATCAGGGTTCGATTCCGGATCGGGAGCCTGAGAAACGGCTACCACATCCAAGGAAGG | 61/63 | 2e-22 |
| B-actin | TCACTCCATCATGAAGTGTGACGTGTACATCCGCAAAGA | 36/37 | 3e-9 |
| CYP11A | CCGATGCTACAGCTGGTCCCCCTCCTCAAAGCCAGCATCAAGGAGACACTA | 50/50 | 3e-19 |
| CYP11B2 | GCAGTGCAGCATGGGAAAGGAATAAGGGGGCAACAAGGTGCACAGACCTCAGAGATGGCT | 60/60 | 4e-25 |
| CYP17 | CACAAGGCCAACGTTGACTCCAGCATCGGT | 30/30 | 9e-8 |
| CYP19 | AGAGTTTGAGGGAGATCCAGTCGGTGAAGAAACCGTATCCATAAAGACCCGATTCCA | 54/57 | 4e-16 |
| CYP21 | CGCCCTCCCTGCAGCCCCTGCCCCACTGCCGTGTCATCCTC | 39/40 | 5e-11 |
| 17₿HSD1 | AAAGGAAGGCTTATCCTTGAGATTGCGTGGGAGACACAA | 37/37 | 1e-11 |
| 17฿HSD4 | AGCCAGAGTATGTGGCACCTCTTGTCCTCTGGCTTTGTCACGAGAGTTGTGAGGAGAATGGTG | 62/63 | 2e-24 |
| HMGR | TCCTGTGGCCAGGAGGTTTGACTGAAACATTCACACAGGGCTCTTTGATGGACCCACGAACGC | 63/65 | 2e-21 |
| StAR | CCAGGAGAATCCCTACTGGAAGCCTGCAAGTCTAAGATCTCCATCTGGTGACAGTGGCATGGGT- GGGGTTCGTGTT | 74/75 | 2e-31 |
| CYP11B1 | CTTGTCCCCAGCCCTACCTGGCCACTTTCTCCAGCAAGCACTGTCCTCTGGGCAGTTTGCACCCA- TCCCTCCCAGT | 73/76 | 6e-25 |
| 3฿HSD2 | TGCTTTGTGCAGTATCTGGATGCGNTGGGCTTGATGTATTTGCCGGAGTCTTGAATGAAAAGGG- ACCAGGAGCTGAGGAATTGCNAANAACCTGCTCTCCGC | 93/104 | 1e-21 |
Note. See text for discussion of the statistical significance of the E value. * All sequences are listed s 5’-3’.
The PCR methods were also optimized to ensure optimum efficiency (100%) over a range of tested RNA concentrations. Relative efficiencies were also determined to ensure that quan- tification of sequences of interest relative to housekeeping genes would remain constant even at a wide range of relative message concentrations (Fig. 4). The determination of all sequences of interest could be achieved quantitatively with 100% efficiency over a range of at least four orders of magnitude.
Chemical Exposure Results
In Exposure 1, H295R cells were exposed to several model inducers for 24 h. At the end of 24 h, relative responses of 10 genes involved in the steroidogenic pathway were evaluated and compared to negative controls that were analyzed along with the treated cells (Fig. 5). The levels of gene expression in blank and solvent control cell cultures were remarkably consistent when normalized to the housekeeping genes ß-actin (Table 3) or 18S ribosomal RNA (Table 4). However, an evaluation of the coeffi- cients of variation (CV) for blanks indicated that the variability in gene expression associated with 18S RNA-normalized data were greater than those associated with data normalized to B-actin. To evaluate the amount of variability associated with the housekeeping genes, separate from that associated with the measurement steroidogenic genes, a comparison of Ct values for 18S RNA and ß-actin among all treatments was conducted with blank data from Exposure 1. Results of the data analysis indi- cated that the variability associated with 18S RNA (average CV of 26%) was greater than the variability associated with ß-actin (average CV of 2.1%). Also, the coefficients of variation for 18S RNA ranged from 0.52 to 115% while for ß-actin the range was
0.95 to 3.94%. This result demonstrates that gene expression data normalized to 18S RNA would incorporate additional sources of variability not associated with the measurement of specific genes.
Treatment of H295R cells with model inducers resulted in significant changes in gene expression (Fig. 5). Treatment of H295R cells with forskolin and 8BrcAMP resulted in significant increases in expression of CYP17, CYP21, CYP11A, 3-HSD2, StAR, and CYP11B2 as normalized by ß-actin. Also, treatment with 8BrcAMP resulted in a significant increase in CYP19 gene expression. Of the genes that were significantly altered, CYP11B2 was induced to the greatest extent (>15-fold increase in cells treated with forskolin or 8BrcAMP). PMA treatment of H295R cells resulted in statistically significant increases in CYP21 and CYP19 gene expression, while lovastatin did not significantly alter the expression of any steroidogenic genes. However, while CYP11B2 expression was altered by PMA, the alteration in gene expression was not significantly different from that of the solvent control. Treatment of the cells with PMA also resulted in a decrease in the expression of CYP11A (3.3-fold), CYP17(10.9-fold), and HMGR (2.9-fold), but none of these reductions were statistically significant. In contrast to the differences in gene expression observed with data normalized to B-actin, no statistically significant differences were noted for treatments where gene expression was normalized to 18S RNA. The relatively great variability in measured 18S RNA activity in both the controls and treated cells masked any alterations in gene expression due to chemical treatment (Table 4). Thus, while there was a 47-fold increase in 30HSD2 in 8BrcAMP- treated cells, compared to a 12-fold increase noted in ß- actin-normalized data, normalization of expression to 18S RNA introduced variability into the data and resulted in no significant differences.
Relative efficiency - Difference CYP17 ß-actin (calc. from means)
5
4
y = - 0.045x + 1.9549
R2= 0.2111
Ct
3
2
T
2
1
0
-4
-3
-2
-1
0
1
2
3
log ng RNA
Relative efficiency - Difference STAR ß-actin (calc. from means)
6
5
y = - 0.0379x + 2.1567
4
R2 = 0.0325
0
∆
3
2
2
E
1
0
-4
-3
-2
-1
0
1
2
3
log ng RNA
To further elucidate the effects of forskolin and PMA on gene expression, cells were exposed to different concentrations of these compounds over time periods up to 48 h (Fig. 6). In general, treatment with PMA resulted in greater alteration in gene expres- sion at 12 than at 24 h for both 10 and 40 nM PMA. As was observed in Exposure 1, PMA reduced the expression of CYP11A and CYP17; the inhibition of CYP17 was not apparent until 24 h, whereas the reduction of CYP11A was initially evident at 12 h and continued on to 24 h. In cells treated with 10 nM PMA, the expression of CYP11A was somewhat greater at 24 than at 12 h. Furthermore, when CYP11A levels at 24 h in the 10 nM-PMA group were compared to levels at 12 and 24 h in the 40-nM PMA treatment group, no significant differences were observed. These results suggest some recovery for this gene may have occurred, but the exact mechanism of this recovery is unknown at this time. The most significant effect of exposure to PMA was the large increase in CYP19 and 3HSD2 gene expression at 12 h for both the tested concentrations. The con- centration of CYP19 mRNA was increased 240- and 274-fold by 10 and 40 nM PMA, respectively. Also, 13ßHSD2 gene expres- sion was increased 43.2- and 23-fold by 10 and 40 uM PMA, respectively. The expression of these genes was approximately 10-fold less at 24 h than it was at 12 h, with the expression levels of both genes being less than 1.5-fold different between con- centrations. This general pattern of greater gene expression at
12 compared to 24 h occurred for most of the genes analyzed. Furthermore, at 24 h there was little difference in gene expres- sion between cells treated with 10 or 40 nM PMA for genes monitored in the exposure. The consistency of this result among genes and between concentrations as well as to the results of the previous exposures adds to the validity of the great levels of mRNA induction observed.
Time- and concentration-dependent changes in gene expres- sion were observed in cells treated with forskolin (Fig. 6). While gene expression at both doses tended to be greater at 12 h than at 24 h, like that observed with PMA, expression at 48 h for some of the genes returned to the levels measured at 12 h. Thus, for genes such as CYP17, CYP11A, and StAR, this resulted in measured gene expression that resembled an inverted time-response curve.
As was observed in Exposure 1, 3฿HSD2, CYP11B2, and CYP19 were the genes for which expression was increased to the greatest extent over the three time periods. However, several specific time- and concentration-related differences in gene expression were noted among these three genes. For instance, a 40-fold induction in CYP19 gene expression was observed at 12 h in cells exposed to 10 or 50 uM forskolin. This was followed by a reduction in gene expression to approximately 20-fold induction at 24 and 48 h sampling time in both treat- ment groups. For CYP11B2 at 12 h, there was a 68- or 34-fold increase in gene expression in cell treated with 10 or 50 µM,
40
35
30
8BrcAMP
25
FOLD
20
15
10
ß-Actin normalized
HILSCHEROVA ET AL.
5
40
35
17bHSD1
30
17bHSD4
25
20
TPA
3bHSD2
CYP11A
FOLD
CYP11B2
15
40
CYP17
10
35
CYP19
30
25
FOLD
Forskolin
CYP21
HMGR
5
20
STAR
15
40
10
35
30
5
17bHSD1
17bHSD4
25
Blank
3bHSD2
FOLD
20
CYP11A
15
CYP11B2
40
CYP17
10
35
CYP19
17bHSD1
30
Lovastatin
CYP21
17bHSD4
HMGR
5
25
STAR
3bHSD2
CYP11A
FOLD
20
CYP11B2
15
CYP
10
CYP19
CYP21
17bHSD1
HMGR
5
17bHSD4
35
STAR
3bHSD2
CYP11A
30
CYP11B2
25
8BrcAMP
40
CYP17
20
35
CYP19
FOLD
30
CYP21
15
17bHSD1
25
HMGR
17bHSD4
10
3bHSD2
20
18S normalized
FOLD
SC
STAR
CYP11A
15
5
CYP11B2
CYP
10
CYP19
CYP21
5
HMGR
35
STAR
30
17bHSD1
25
17bHSD4
20
3bHSD2 CYP11A
FOLD
TPA
15
17bHSD1
17bHSD4
Downloaded from http://toxsci.oxfordjournals.org/ at UNIVERSITY OF AUCKLAND LIBRARY on May 2, 2015
35
CYP11B2
CYP
10
3bHSD2
30
CYP19
CYP11A
CYP11B2
25
Forskolin
CYP21
5
20
HMGR
CYP17
FOLD
STAR
CYP19
15
CYP21 HMGR STAR
10
35
30
5
17bHSD
25
17bHSD
20
control; blank, unexposed cells.
solvent control. Values presented are the means of three determinations on each of three replicate exposures. PMA, phorbol-12-myristate 13-acetate; sc, solvent
FIG. 5. The effects of different chemicals on expression of steroidogenic enzyme genes in H2195R cells in culture. Expression of steroidogenic genes was normalized to the expression of either ß-actin or 18S ribosomal RNA as indicated. Fold induction represents the increase in expression compared to the relevant
CYP1 YP11A
FOLD
Blank
3bHSD2
15
35
CYP11B2
10
CYP
30
CYP19
17bHSD1
25
Lovastatin
CYP
5
17bHSD4
20
HMGR
3bHSD2
CYP11A
FOLD
STAR
15
CYP11B2
CYP
10
CYP19
CYP21 HMGR
5
17bHSD1
17bHSD4
STAR
3bHSD2 CYP11A
35
CYP11B2
CYP
30
CYP19
17bHSD1
25
CYP21
17bHSD4
FOLD 15
20
HMGR
SC
STAR
3bHSD2
CYP11A
CYP11B2
10
CYP1
CYP19
CYP21
5
HMGR STAR
17bHSD1
17bHSD4
3bHSD2
CYP11A
CYP11B2
CYP
CYP19
CYP21
HMGR STAR
| Gene | Treatment | |||||
|---|---|---|---|---|---|---|
| Blank | Solvent | Forskolin | 8BrcAMP | PMA | Lovastatin | |
| CYP17 | 1.06 ± 0.18 | 1.05 ± 0.021 | 4.58 ± 0.59* | 3.68 ± 0.98* | 0.34 ± 0.40 | 0.44 ± 0.37 |
| CYP21 | 1.11 ± 0.08 | 1.00 ± 0.11 | 3.22 ± 0.93* | 2.51 ± 1.32* | 2.40 ± 0.23* | 1.87 ± 0.48 |
| CYP11A | 1.06 ± 0.24 | 1.00 ± 0.10 | 2.98 ± 0.50* | 2.69 ± 0.47* | 0.57 ± 0.40 | 1.18 ± 0.82 |
| CYP19 | 1.01 ± 0.19 | 1.00 ± 0.08 | 5.93 ± 0.94 | 6.53 ± 1.05* | 7.42 ± 4.94* | 3.23 ± 3.64 |
| StAR | 0.94 ± 0.31 | 1.01 ± 0.17 | 3.71 ± 0.62* | 3.86 ± 0.71* | 1.45 ± 0.30 | 1.10 ± 0.31 |
| 3฿HSD2 | 1.17 ± 0.26 | 1.07 ± 0.43 | 8.57 ± 0.49* | 12.1 ± 3.92* | 1.92 ± 1.02 | 2.36 ± 1.63 |
| HMGR | 1.19 ± 0.57 | 1.12 ± 0.55 | 1.45 ± 0.58 | 2.09 ± 0.97 | 1.41 ± 1.21 | 1.42 ± 0.94 |
| 17฿HSD1 | 0.96 ± 0.17 | 1.00 ± 0.08 | 0.98 ± 0.20 | 0.85 ± 0.29 | 0.88 ± 0.06 | 0.92 ± 0.41 |
| 17฿HSD4 | 1.09 ± 0.10 | 1.02 ± 0.23 | 1.06 ± 0.28 | 1.00 ± 0.35 | 0.89 ± 0.24 | 1.11 ± 0.30 |
| CYP11B2 | 1.35 ± 0.29 | 1.08 ± 0.15 | 17.8 ± 2.76* | 25.7 ± 8.34* | 3.68 ± 2.78 | 1.72 ± 0.34 |
Note. Cell exposures were for 24 h; relative gene activity expressed as means and standard deviations. *Statistically different from solvent control (p < 0.05).
| Gene | Treatment | |||||
|---|---|---|---|---|---|---|
| Blank | Solvent | Forskolin | 8BrcAMP | PMA | Lovastatin | |
| CYP17 | 0.86 ± 0.03 | 1.01 ± 0.15 | 4.58 ± 1.20* | 20.5 ± 28.6 | 0.29 ± 0.27 | 0.25 ± 0.18 |
| CYP21 | 0.92 ± 0.07 | 1.02 ± 0.23 | 3.24 ± 1.18* | 8.82 ± 9.46 | 2.74 ± 0.96 | 1.30 ± 0.43 |
| CYP11A | 0.86 ± 0.10 | 1.01 ± 0.13 | 2.99 ± 0.91* | 12.0 ± 14.8 | 0.56 ± 0.19 | 0.69 ± 0.38 |
| CYP19 | 0.83 ± 0.10 | 1.02 ± 0.22 | 5.87 ± 1.14* | 35.3 ± 48.2 | 9.45 ± 7.13 | 3.03 ± 4.17 |
| StAR | 0.80 ± 0.33 | 1.00 ± 0.04 | 3.70 ± 0.96* | 17.5 ±21.1 | 1.59 ± 0.47 | 0.74 ± 0.13 |
| 3฿HSD2 | 0.99 ± 0.32 | 1.05 ± 0.37 | 8.51 ± 1.47* | 47.3 ± 53.4 | 1.93 ± 0.33 | 1.46 ± 0.73 |
| HMGR | 1.03 ± 0.59 | 1.08 ± 0.48 | 1.44 ± 0.60 | 11.6 ± 16.5 | 1.32 ± 0.68 | 0.85 ± 0.40 |
| 17₿HSD1 | 0.79 ± 0.09 | 1.01 ± 0.19 | 0.89 ± 0.28 | 4.20 ± 5.61 | 0.97 ± 0.24 | 0.61 ± 0.19 |
| 17฿HSD4 | 0.91 ± 0.20 | 1.01 ± 0.21 | 1.07 ± 0.37 | 4.17 ± 5.00 | 0.95 ± 0.07 | 0.74 ± 0.06 |
| CYP11B2 | 1.11 ± 0.21 | 1.00 ± 0.06 | 17.4 ± 0.63* | 159 ± 230 | 3.54 ± 1.42 | 1.17 ± 0.21 |
Note. Cell exposures were for 24 h; relative gene activity expressed as means and standard deviations. *Statistically different from solvent control (p < 0.05).
respectively. Thereafter, there was a decrease in CYP11B2 expression that resulted in expression levels that were similar among dose groups for the 24 and 48 h time points. In contrast to CYP11B2, there was no concentration-related difference in expression of 3ßHSD2 with time but there was a general trend of reduced gene expression (11-fold reduction) over the experimental time period. Interestingly, while all previous expo- sures caused little effect on HMGR expression at 12 or 24 h of exposure to either 10 or 50 uM, an exposure to forskolin for 48 h resulted in a considerable and similar increase in the expression of this gene.
The reproducibility of the gene expression in the H295R bioassay was evaluated with forskolin- and PMA-treated cells from Exposures 1 and 3 (Table 5). In cells exposed to forskolin, the only significant differences in gene expression between the two experiments were for CYP21, CYP19, StAR, and CYP11B2.
Of these genes, only CYP21 and CYP19 had interassay differences that were greater than 2-fold. In general, the expres- sion of these genes was greater in Exposure 3 than in Exposure 1, and overall significances of these gene activities relative to sol- vent control values were consistent between assays. That is, if there was a significant alteration in gene expression in one assay it was also significant in the second assay. In the PMA exposures, only CYP17, CYP21, CYP19, 3ßHSD2, and CYP11B2 signifi- cantly differed between assays. As was observed with forskolin, the level of gene expression in Exposure 3 was generally greater than that measured in Exposure 1, with all fold differences being greater than 2.5. Again, while the magnitude of gene expression activity differed between the two assays, the significances of gene expression as a consequence of PMA exposure were simi- lar, indicating that the gene expression profile remained the same between assays. Overall, this analysis indicates that while there
Downloaded from http://toxsci.oxfordjournals.org/ at UNIVERSITY OF AUCKLAND LIBRARY on May 2, 2015
Forskolin 10 µM
Forskolin 50 µM
100
50
40
30
20
10
FOLD
10
FOLD
HOURS
1
12
24
17bHSD1
17bHSD4
3bHSD2
CYP11A
CYP11B2
CYP17
CYP19
CYP21
HMGR
STAR
48
17bHSD1
17bHSD4
3bHSD2
CYP11A
CYP11B2
CYP17
CYP19
CYP21
HMGR
STAR
PMA 10 nM
PMA 40 nM
100
100
FOLD
10
FOLD
10
1
1
HOURS
12
0.
24
0
17bHSD1
17bHSD4
3bHSD2
CYP11A
CYP11B2
CYP17
CYP19
CYP21
HMGR
STAR
17bHSD1
17bHSD4
3bHSD2
CYP11A
CYP11B2
CYP17
CYP19
CYP21
HMGR
STAR
is some interassay variability in absolute expression, the overall conclusion that can be drawn relative to gene expression profiles is consistent between assays.
In another set of exposures (Exposure 2), the effects of a variety of inhibitors of steroidogenic genes were examined (Table 6). The most wide-ranging effects were observed after exposure to ketoconazole and spironolactone. Spironolactone decreased the expression of StAR by greater than 90% yet caused a 7-fold induction of CYP19. Spironolactone also significantly decreased the expression of 17ßHSD4 (but not 17BHSD1), 3ßHSD2, and CYP11A. Ketoconazole decreased the expression of CYP11A and 30HSD2, while it increased the expression of CYP21, CYP19, HMGR, 17BHSD1, and CYP11B2. Of particular interest is the fact that ketoconazole was the only inhibitor tested that resulted in a significant induction of CYP21 and CYP11B2. No other inhibitor tested
significantly altered the expression of these two genes. None of the model inhibitors significantly induced the expression of StAR, but expression of this gene was significantly decreased by exposure to spirolactone, aminoglutethimide, or daidzein. How- ever, the reduction in StAR was not correlated to any general decrease in the expression of the other steroidogenic genes mon- itored in the study. In contrast to the other genes monitored in this experiment, the expression of CYP17 was not affected by any of the inhibitor chemicals.
DISCUSSION
An analytical procedure was developed that is capable of measuring gene expression of a range of steroidogenic enzymes as a result of exposure to chemicals. The Q-RT-PCR procedure was chosen over traditional enzyme assay techniques because it
offers the opportunity to screen expression of a wide range of genes using a single technique and a limited amount of sample. This latter criterion was essential to the development of a cell culture-based bioassay approach.
The production of steroids is a complex process with multiple sensitive control points. Given the complexity of the system and the number of enzymes and substrates involved, the potential for xenobiotic chemicals to interfere with this process is relatively great. Indeed the presence of genetic deficiencies in these ster- oidogenic enzymes leads to a condition known as congenital adrenal hyperplasia (CAH), which is often fatal (Richmond et al., 2001). While this condition is most frequently caused by a deficiency in CYP21B (Chiou et al., 1990), deficiencies in StAR
| Gene | Forskolin (50 µM) | PMA (40 nM) | ||
|---|---|---|---|---|
| Fold change | p value | Fold change | p value | |
| CYP17 | -1.2 | 0.138 | +2.51* | 0.001 |
| CYP21 | +3.1* | 0.001 | +5.30* | 0.009 |
| CYP11A | -1.2 | 0.230 | -1.01 | 0.953 |
| CYP19 | +3.1* | 0.018 | +3.6* | 0.001 |
| StAR | -1.05* | 0.05 | +1.09 | 0.707 |
| 3฿HSD2 | -1.05 | 0.634 | +4.3* | 0.001 |
| HMGR | -1.73 | 0.216 | +1.18 | 0.430 |
| 17₿HSD1 | -1.37 | 0.176 | +1.17 | 0.231 |
| 17฿HSD4 | -1.10 | 0.667 | +1.50 | 0.075 |
| CYP11B2 | +1.81* | 0.011 | +2.95* | 0.001 |
Note. Fold change indicates direction and difference between Exposures 1 and 3; p value is based on the results of t-test between Exposures 1 and 3. *Statistically significant difference between the two experiments.
and other steroidogenic enzymes are also capable of causing CAH (Richmond et al., 2001).
Mobilization of cholesterol to CYP11A, also known as CYP450scc, and its conversion to pregnenolone are the first and rate-limiting steps in the conversion of cholesterol to steroid hormones and is a point of both acute and chronic control (Hu et al., 2001; 2002). In our study, only 8BrcAMP and forskolin resulted in significant increases in CYP11A1 expression; PMA significantly decreased CYP11A expression, while lovastatin appeared to have little effect on this enzyme. Alterations in CYP11A are also noteworthy since some studies indicate that the expression of other steroidogenic enzymes is coordinated with CYP11A. For example, it has been demonstrated that CYP11A activity may be coordinated with CYP11B1 activity by the physical proximity of the two enzymes (Cauet et al., 2001). Such physical interrelationships between enzymes may be of greater significance in vivo than in vitro due to the tissue-specific expression of some enzymes. In fact, some tissues, particularly the adrenal gland, exhibit differential enzyme expression within different regions of the tissue (Gazdar et al., 1990; Sanderson et al., 2000; Staels et al., 1993).
Some of the chemicals tested resulted in some increase in CYP21 gene expression. The CYP21 gene product is required for the synthesis of both aldosterone and corticosteroids. Defi- ciency of this enzyme in CAH results in deficiencies in both cortisol and aldosterone that are also accompanied by overpro- duction of androgens (Chiou et al., 1990, Richmond et al., 2001). The overproduction of androgens is due to a combination of the general adrenal hyperplasia and substrate accumulation related to inhibition of the gluco- and mineralocorticoid pathways. In our experiments, increases in CYP21 would be expected to lead to increased synthesis of cortisol and aldosterone and may result in decreased substrate availability for androgen and estrogen production.
CYP17 catalyzes the conversion of aldosterone to cortico- steroid substrates and ultimately to sex steroid substrates.
| Gene | Treatment | |||||||
|---|---|---|---|---|---|---|---|---|
| Blank | Solvent | Metyrapone | Daidzein | Ketoconazole | AMG | Androstedione | Spironolactone | |
| CYP17 | 1.16 (0.26) | 1.00 (0.06) | 0.81 (0.01) | 0.75 (0.05) | 0.93 (0.15) | 0.78 (0.08) | 1.29 (0.03) | 0.77 (0.46) |
| CYP21 | 1.17 (0.15) | 1.02 (0.21) | 1.11 (0.35) | 1.14 (0.14) | 2.18* (0.27) | 0.89 (0.19) | 0.68 (0.26) | 1.36 (0.32) |
| CYP11A | 1.06 (0.07) | 1.00 (0.04) | 0.83* (0.03) | 0.88 (0.07) | 0.73* (0.06) | 0.84* (0.10) | 1.07 (0.14) | 0.37* (0.07) |
| CYP19 | 1.11 (0.41) | 1.00 (0.10) | 1.62 (0.35) | 1.30 (0.18) | 4.45* (0.95) | 1.21 (0.12) | 1.41 (0.20) | 7.13* (2.72) |
| StAR | 1.07 (0.31) | 1.01 (0.18) | 1.05 (0.42) | 0.31* (0.18) | 1.61 (0.65) | 0.29* (0.07) | 0.74* (0.45) | 0.13* (0.13) |
| 3฿HSD2 | 1.15 (0.08) | 1.03 (0.31) | 0.48* (0.18) | 0.50* (0.12) | 0.43* (0.09) | 0.92 (0.30) | 0.76 (0.38) | 0.26* (0.12) |
| HMGR | 0.98 (0.27) | 1.01 (0.48) | 1.01 (0.06) | 0.89 (0.14) | 1.66* (0.24) | 0.97 (0.26) | 1.22* (0.29) | 0.73 (0.29) |
| 17₿HSD1 | 1.18 (0.29) | 1.01 (0.12) | 1.70 (0.12) | 1.60 (0.42) | 2.00* (0.79) | 1.41 (0.23) | 1.70 (0.31) | 1.38 (0.65) |
| 17฿HSD4 | 1.00 (0.13) | 1.04 (0.35) | 1.15 (0.10) | 1.16 (0.19) | 0.99 (0.69) | 0.41* (0.44) | 0.21* (0.03) | 0.34* (0.28) |
| CYP11B2 | 0.92 (0.16) | 1.02 (0.24) | 1.65 (0.42) | 1.13 (0.03) | 5.89* (1.82) | 0.87 (0.19) | 1.25 (0.120 | 0.76 (0.16) |
Note. Cell exposures were for 24 h; relative gene activity expressed as means with standard deviations in parentheses. *Statistically different from solvent control (p < 0.05).
Downloaded from http://toxsci.oxfordjournals.org/ at UNIVERSITY OF AUCKLAND LIBRARY on May 2, 2015
| Chemical | CYP11A | CYP11B2 | CYP17 | CYP19 | CYP21 | 17₿HSD1 | 17₿HSD4 | 3฿HSD2 | HMGR | StAR |
|---|---|---|---|---|---|---|---|---|---|---|
| Inducers | ||||||||||
| 8BrcAMP | " | 1111 | " | 11 | " | – | – | 111 | – | " |
| PMA | # | – | 111 | 11 | " | # | – | |||
| Forskolin | " | 1111 | " | 11 | " | – | – | 11 | – | " |
| Lovastatin | – | " | – | – | " | – | – | " | " | – |
| Inhibitors | ||||||||||
| Aminogluteth-imide | – | – | – | – | – | – | 11 | – | – | # |
| Androstedione | – | – | – | – | – | 11 | – | |||
| Spironolactone | # | – | – | 11 | – | – | 11 | 11 | – | |
| Daidzein | – | – | – | – | – | – | – | # | – | # |
| Ketoconazole | – | 11 | – | 1 | 1 | 1 | – | # | – | – |
| Metyrapone | – | – | – | – | – | – | – | # | – | – |
Note. Symbols indicate difference relative to control. 1, 2-fold or more; 11, 5-fold or more; 111, 10-fold or more; 1111, 15-fold or more. All other differences less than 2-fold.
Therefore, it is possible that this enzyme could redirect steroid output from mineralocorticoids to glucocorticoids or weak androgens. Inhibition of CYP17 would have the opposite effect. Supporting this hypothesis is the observation that treatment with PMA, which results in an almost complete inhibition of CYP17 expression, results in the greatest increase in CYP19 expression (p < 0.01). CYP19 is responsible for the final conversion of androgens to estrogens.
While only a limited number of chemicals were tested in this study, distinct gene expression profiles are apparent (Table 7). In particular, similar expression patterns were observed for 8BrcAMP and forskolin. These patterns included relatively great increases in the expression of 3฿HSD2 and CYP11B2 and moderate increases in expression of CYP11A, CYP17, CYP19, CYP21, HMGR, and StAR. In contrast, PMA resulted in decreases in CYP11A and CYP17, moderate increases in CYP11B2 and CYP21, and a greater increase in CYP19. This variety of responses demonstrates the utility of the H295R cell line for the detection of both induction and down- regulation of gene expression for steroidogenic enzymes (Heneweer et al., 2004). Lovastatin resulted in only moderate increases of 3ßHSD2, CYP11B2, CYP21, and HMGR expres- sion. We hypothesize that the expression profiles observed for forskolin and 8BrcAMP, which were similar, resulted from increased signaling through the cAMP pathway and that other chemicals causing a similar alteration would result in a similar expression profile, as reported previously (Sanderson et al., 2002). It has been shown that forskolin is able to increase cellular cAMP concentrations in H295R cell line (Sanderson et al., 2002). In contrast, the expression profiles observed for PMA and lovastatin appear to have been produced by a signal- ing pathway other than the cAMP pathway and were distinct from each other. PMA exerts effects on steroidogenesis pri- marily through the MAPKC pathway and so would be expected to have an expression profile distinct from the cAMP- dependent pathways. Lovastatin is known to specifically inhibit
HMG-CoA reductase activity and, as expected, treatment with this chemical increased the expression of HMG-CoA reductase in H295R cells.
It has been hypothesized in recent studies that the ability of chemicals to alter activity of CYP19 (aromatase) represents a potential mechanism of endocrine disruption (Hayes et al., 2002; Heneweer et al., 2004; Sanderson et al., 2002). While several of the chemicals tested in this study altered the expression of CYP19, this gene was in no case the only gene whose expression was altered. Indeed, in no case was the altera- tion in the expression of CYP19 the most significant alteration in gene expression (Table 7). These observations clearly demon- strate the need to examine alterations in steroid metabolic pro- cesses in a far more holistic fashion, evaluating many different end points including but not limited to gene expression. While gene expression profiling offers detailed information on alterations in gene regulation, other procedures such as the measurement of enzymes activities and amounts of steroids produced offer more proximal measures of the effects of che- micals on steroidogenesis.
The ability to assess all of the key enzymes involved in steroidogenesis in a single assay procedure will clearly be of great interest to those studying the effects of xenobiotics on steroidogenesis. While initial work has focused on specific enzymes such as aromatase, the assay we have presented allows for more general assessment of steroidogenesis by evaluating both enzymes that determine the overall rate of steroidogenesis as well as those specific enzymes that can influence the overall fate or balance of steroid production. The H295R cell line has been previously used in such a bioassay approach, but the end points in those studies were either mRNA species and one or two specific enzymes (Sanderson et al., 2001, 2002) or were a variety of enzyme activities (Ohno et al., 2002).
Our findings demonstrate that the genes within the steroido- genesis pathway are not expressed to the same extent and that
different chemicals result in different relative changes in the expression of various genes. Chemical agents have the potential to alter gene expression profiles and, potentially, the steroids produced by this pathway. The changes in patterns of relative expression can be used to classify chemicals of unknown mechanisms of action on the steroidogenic pathways. In this way, chemicals can be grouped for further testing of a reduced set of model chemicals and for risk assessments.
ACKNOWLEDGMENTS
This study was funded by U.S. EPA, ORD Service Center/NHEERL, Contract GS-10F-0041L. We acknowledge many helpful discussions and manuscript review by Dr. Ralph Cooper, Dr. Jerome Goldman, and Dr. Robert Kavlock, Endocrinology Branch, NHEERL, U.S. EPA, Research Triangle Park, North Carolina. Support was also given by ENTRIX Inc. and U.S. EPA; no conflicts of interest exist.
REFERENCES
Ankley, G., Mihaich, E., Stahl, R., Tillitt, D., Colborn, T., McMaster, S., Miller, R., Bantle, J., Campbell, P., Denslow, N., et al. (1998). Overview of a workshop on screening methods for detecting potential (anti-) estro- genic/androgenic chemicals in wildlife. Environ. Toxicol. Chem. 17, 68-87. Cauet, G., Balbuena, D., Achstetter, T., and Dumas, B. (2001). CYP11A1 sti- mulates the hydroxylase activity of CYP11B1 in mitochondria of recombinant yeast in vivo and in vitro. Eur. J. Biochem. 268, 4054-4062.
Chiou, S. H., Hu, M. C., and Chung, B. C. (1990). A missense mutation at Ile172-Asn or Arg356-Trp causes steroid 21-hydroxylase deficiency. J. Biol. Chem. 265, 3549-3552.
Connor, K., Howell, J., Chen, I., Liu, H., Berhane, K.,Sciarretta, C., Safe, S., and Zacherewski, T. (1996). Failure of chloro-s-triazine-derived compounds to induce estrogen receptor-mediated responses in vivo and in vitro. Fundam. Appl. Toxicol. 30, 93-101.
EDSTAC Final Report (1998). Endocrine Disruptor Screening and Testing Advisory Committee Final Report. U.S. Environmental Protection Agency. www.epa.gov/opptintr/opptendo/finalrpt.htm.
Gazdar, A. F., Oie, H. K., Shackleton, C. H., Chen, T. R., Triche, T. J., Myers, C. E., Chrousos, G. P., Brennan, M. F., Stein, C. A., and La Rocca, R. V. (1990). Estab- lishment and characterization of a human adrenocortical carcinoma cell line that expresses multiple pathways of steroid biosynthesis. Cancer Res. 50, 5488-5496.
Hayes, T. B. Collins, A., Lee, M., Mendoza, M., Noriega, N., Stuart, A. A., and Vonk, A. (2002). Hermaphroditic, demasculinized frogs after exposure to the herbicide atrazine at low ecologically relevant doses. Proc. Nat. Acad. Sci. USA 99, 5476-5480.
Heneweer, M., Van den Berg, M., and Sanderson, J. (2004). A comparison of human H295R and rat R2C cell lines as in vitro screening tools for effects on aromatase. Toxicol. Lett. 146, 183-194.
Hu, M. C., Chiang, E.F .- L., Tong, S. K., Lai, W., Hsu, N.C., Wang,L.C .- K., and Chung, B. C. (2001). Regulation of steroidogenesis in transgenic mice and zebrafish. Mol. Cell. Endocrinol. 171, 9-14.
Hu, M. C., Hsu, N. C., El Hadj, N. B., Pai, C. I., Chi, H. P. C, Wang, K. L., and Chung, B. C. (2002). Steroid deficiency syndromes in mice with targeted disruption of CYP11A1. Mol. Endocrinol. 16, 1943-1950.
Kavlock, R. T., Daston, G. P., De Rosa, C., Fenner-Crisp, P., Gray, L. E., Kaattari, S., Lucier, G., Luster, M., Mac, M. J., Maczka, C., et al. (1996). Research needs for the risk assessment of health and environmental effects of endocrine disruptors: A report of the U.S. EPA sponsored workshop. Environ. Health Perspect. 104, 715-740.
Ohno, S., Shinoda, S., Toyoshima, S., Nakazawa, H., Makino, T., and Nakajin, S. (2002). Effects of flavonoid phytochemicals on cortisol production and on activities of steroidogenic enzymes in human adrenocortical H295R cells. J. Steroid Biochem. Mol. Biol. 80, 355-363.
Rainey, W. E., Bird, I. M., Sawetawan, C., Hanley, N. A., McCarthy, J. L., McGee, E. A., Wester, R., and Mason, J. I. (1993). Regulation of human adrenal carcinoma cell (NCI-H295) production of C19 steroids. J. Clin. Endocrinol. Metab. 77, 731-737.
Richmond, E. J., Flickinger, C. J., McDonald, J. A., Lovell, M. A., and Rogol, A. D. (2001) Lipoid congenital adrenal hyperplasia (CAH): Patient report and a mini-review. Clin. Pediatr. (Phila.) 40, 403-407.
Sanderson, J. T., Boerma, J., Lansbergen, G. W., and Van den Berg, M. (2002). Induction and inhibition of aromatase (CYP19) activity by various classes of pesticides in H295R human adrenocortical carcinoma cells. Toxicol. Appl. Pharmacol. 182, 44-54.
Sanderson, J. T., Heneweer, M., Seinen, W., Giesy, J. P., and Van den Berg, M. (1999). Chloro-s-triazine herbicides and certain metabolites induce aro- matase (CYP19) activity in H295R human adrenocortical carcinoma cells. Organohalogen Compounds 42, 5-8.
Sanderson, J. T., Seinen, W., Giesy, J. P., and Van den Berg, M. (2000).2-Chloro- S-triazine herbicides induce aromatase (CYP19) activity in H295R human adrenocortical carcinoma cells: A novel mechanism for estrogenicity. Toxicol. Sci. 54, 121-127.
Sanderson, J., Thomas, R. J., Letcher, M., Heneweer, Giesy, J. P., and Van den Berg, M. (2001). Effects of chloro-S-triazine herbicides and metabolites on aromatase (CYP19) activity in various human cell lines and on vitellogenin production in male carp hepatocytes. Environ. Health Perspect. 109, 1027-1031.
Staels, B., Hum, D. W., and Miller, W.L. (1993). Regulation of steroidogenesis in NCI-H295R cells: A cellular model of the human fetal adrenal. Mol. Endocrinol. 7, 423-433.
Villeneuve, D. L., Blankenship, A. L., and Giesy, J. P. (1998). Estrogen receptors environmental xenobiotics. In Toxicant-Receptor Interactions and Modula- tion of Gene Expression. (M. S. Denison and W. G. Helferich, Eds.), pp. 69-99. Lippincott-Raven Publishers, Philadelphia.
Downloaded from http://toxsci.oxfordjournals.org/ at UNIVERSITY OF AUCKLAND LIBRARY on May 2, 2015