| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
and ß receptors
* IZMB, AG Molekulare Bioenergetik, Universität Bonn, Bonn, Germany; and
Klinik für gynäkologische Endokrinologie und Sterilität, Universität Innsbruck, Innsbruck, Austria
1Correspondence: IZMB, Universität Bonn, Kirschallee 1, Bonn 53115, Germany. E-mail: h.lichtenberg{at}uni-bonn.de
ABSTRACT
The human estrogen receptors (hER)
and hERß, differentially expressed and localized in various tissues and cell types, mediate transcriptional activation of target genes. These encode a variety of physiological reproductive and nonreproductive functions involved in energy metabolism, salt balance, immune system, development, and differentiation. As a step toward developing a screening assay for the use in applications where significant numbers of compounds or complex matrices need to be tested for (anti) estrogenic bioactivity, hER and hERß were expressed in a genetically modified Saccharomyces cerevisiae strain, devoid of three endogenous xenobiotic transporters (PDR5, SNQ2, and YOR1). By using receptor-mediated transcriptional activation of the green fluorescent protein optimized for expression in yeast (yEGFP) as reporter 17 natural, comprising estrogens and phytoestrogens or synthetic compounds among which tibolone with its metabolites, gestagens, and antiestrogens were investigated. The reporter assay deployed a simple and robust protocol for the rapid detection of estrogenic effects within a 96-well microplate format. Results were expressed as effective concentrations (EC50) and correlated to other yeast based and cell line assays. Tibolone and its metabolites exerted clear estrogenic effects, though considerably less potent than all other natural and synthetic compounds. For the blood serum of two volunteers, considerable higher total estrogenic bioactivity than single estradiol concentrations as determined by immunoassay was found. Visualization of a hER/GFP fusion protein in yeast revealed a sub cellular cytosolic localization. This study demonstrates the versatility of (anti) estrogenic bioactivity determination using sensitized S. cerevisiae cells to assess estrogenic exposure and effects.—Hasenbrink, G., Sievernich, A., Wildt, L., Ludwig, J., and Lichtenberg-Fraté, H. Estrogenic effects of natural and synthetic compunds including tibolone assessed in Saccharomyces cerevisiae expressing the human estrogen and ß receptors.
Key Words: S. cerevisiae in vitro assay • hER • hERß
ESTROGENS are an important class of signaling molecules, regulating a diverse range of physiological processes in animals and humans. The actions of these hormones are mediated by intracellular receptor proteins (ERs) that on dimerization act as ligand-activated transcription factors by binding of the DNA-binding domain (DBD) to specific DNA sequences [estrogen responsive elements (ERE)] in the regulatory regions of target gene promoters. Both quality (which genes) and quantity (degree) of transcriptional stimulation in a certain cell depend on the kind and concentration of the activated estrogen receptor(s) and several complex regulating factors, e.g., coactivators. The human estrogen receptors (hER) and ß (hERß) belong to the steroid-thyroid-retinoic acid super family of nuclear receptors. The hERß receptor was identified a decade ago (1)
and found to be expressed in the same tissues as hER except in the liver (hER) and the gastrointestinal tract (hERß). However, the two receptors are often localized to different cell types though heterodimers of ER and ERß have been found in vitro (2)
. Receptor-mediated gene regulation also involves ligand dependent interaction with other nuclear factors such as activating protein-1 with, however, differential activation of the two receptors (3)
. Together with structural, such as N-terminal and ligand-binding pocket domain, differences between hER and hERß, these functional differences suggested a potential significant role for ERß in brain (4)
and cardiovascular functions among other important physiological control systems.
Saccharomyces cerevisiae cells do not contain members of the steroid-thyroid-retinoic acid receptor super family but have been used to study respective receptor expression (5)
and structure-function relationships (6
7
8)
and for the identification of signaling pathway components (9
10
11)
. Such approaches used that by expression of the receptor proteins and on ligand exposure the receptor-mediated signaling can be reconstituted and measured using a reporter gene driven by DNA responsive elements. In the absence of hormone, steroid receptors are found in a complex with molecular chaperones, such as Hsp90, in the cytoplasm or nucleus (12)
. Yeast Hsp82 and YDF1 as counterparts of mammalian molecular chaperones function to maintain steroid receptors in a hormone-binding conformation and to assist activation of the downstream pathway (10)
. The introduction of S. cerevisiae based screens for estrogenic activity (13
14
15
16
17
18)
has also demonstrated the applicability of this type of assay in such diverse disciplines as environmental monitoring and analysis of food components. Besides classical hormone testing such assays served to detect estrogenic activity of natural (e.g., phytoestrogens and synthetic compounds, the latter widely referred as to xenoestrogens among the endocrine disruptors) (19
20
21
22
23)
. Many of the yeast assays deploy the Escherichia coli lacZ reporter connected to two or three copies of the DNA estrogen responsive element to determine estrogenic activity. More recently, utilization of a yeast-optimized green fluorescent protein (yEGFP; 24
) fused to the cis-acting ERE took advantage of the direct fluorescence read out opportunity (17
, 18)
and thus avoided cell lysis procedures.
Previously, we (18)
have described the construction and preliminary characterization of the hER-receptor ERE-GFP-reporter test system in S. cerevisiae. Here, we report on the extension of the transactivation assay toward hERß-receptor expression coupled to the ERE-GFP-reporter in a genetically modified yeast strain, devoid of three endogenous xenobiotic transporters (PDR5, SNQ2, and YOR1), and comparative analysis of natural and synthetic (anti) estrogenic compounds. Among the natural compounds were the endogenous estrogens estradiol-17ß (E2) and estriol (E3); the phyto- or mykoestrogens genistein, coumestrol, and zearalenone; and the synthetic compounds with clinical application ethinylestradiol, 17-estradiol, mestranol, estradiol-3-benzoate, 17-hydroxyprogesterone, norethisteron, tamoxifen, 3-hydroxytamoxifen, and tibolone with its metabolites, or without (current) clinical application as the nonsteroidal estrogen diethylstilbestrol. Transcriptionally activated ERE-GFP fluorescence emissions were normalized to cell growth determined by absorption and correlated to internal reference standards. Obtained dose-response curves served for EC50 value calculation. Assay protocol optimization comprised conditions like agitation, temperature, and pH value to determine the most practical test handling conditions. Finally, the blood serum of two volunteers was also investigated under the optimized test conditions revealing significant differences between the estrogenic bioactivity and E2 concentrations as determined by standard methods.
MATERIALS AND METHODS
Plasmids and yeast strains
The plasmid pcDNA3-hERß, harboring the human beta receptor, was kindly provided by Dr. K. Korach (National Institutes of Health). For expression in S. cerevisiae, the hERß receptor was excised as 1508 bp BamHI-XhoI fragment with the protruding XhoI end filled with T4-DNA polymerase (MBI Fermentas) and gel purified. The episomal vector backbone was the high copy Escherichia coli/yeast shuttle vector pYEX-BX (Clontech, Palo Alto, CA) in which the URA cassette was replaced with the TRP selection marker. After EcoRI linearization, filling in of the protruding end and subsequent BamHI digest both fragments were ligated to yield the 8.8 kb pYEX-hERß plasmid. Plasmid pERE-CYC1-GFP was described before (18)
. Plasmid YEpE12 (hER) was a gift from Dr. A. Jungbauer (University of Vienna). For microscopic visualization, the complete 2012 bp ub-ER fragment was amplified from the YEpE12 DNA template by polymerase chain reaction (PCR) with Pwo polymerase (Novagen) using primers corresponding to the nucleotides 5'-gagaggatccgaattcattatgcagatcttcg-3' and 5'-gagaactagttcagactgtggcagggaaacc-3' containing primer-encoded BamHI and SpeI restriction sites. The fragment was then ligated to the BglII and SpeI restriction sites of the modified pYEX-GFP to yield the fusion construct pYEX-GFP/hER.
For the SNQ2 gene deletion cassette, the pUG6 plasmid with the kanamycin marker gene (25)
was used as template. For the deletion cassettes of the S. cerevisiae YOR1 and PDR5 genes, the pUG6 plasmid (25)
was modified by exchange of the kanamycin with the LEU2 marker. The isopropyl-malate dehydogenase gene was amplified with S. cerevisiae wild-type (S288C) genomic DNA as template by PCR using the oligonucleotides 5'-gagaagatctgagttcgaatctcttagcaacc-3' and 5'-gagagagctccaaattaggaatcgtagtttcatg-3' with primer encoded BglII and SacI restriction sites. The PCR fragment was cleaved with BglII/SacI and ligated with the BglII/SacI-digested pUG6 to yield plasmid pUG6(LEU). PDR5 and YOR1 gene replacement cassettes were PCR generated comprising the LEU2 (PDR5 and YOR1) or kanamycin resistance (SNQ2) gene flanked by 
500 bp gene specific homologous 5'- and 3'-targeting regions. The oligonucleotides used for the SNQ2, YOR1, and PDR5 gene replacement cassettes are listed in Table 1
. All gene replacement cassettes were subcloned to pBSK (SmaI). All recombinant plasmids recovered from transformed E. coli XL1-blue cells were mapped by restriction analysis and confirmed by sequencing (GeneART). Computer analysis of nucleotide and amino acid sequences was performed using the Vnti software (Informax).
|
The yeast strains used throughout this study are summarized in Table 2
. The S. cerevisiae pdr5 snq2 yor1 disruption strain was derived from PLY232 (26)
, using the Cre-loxP-recombination system described by Güldener et al. (25)
, except that for PDR5 and YOR1 the LEU2 marker was used for replacement. After yeast transformation, in LEU2 positive or G-418 (200 mg/l) resistant single cell derived colonies the ORF::loxP-kanMX/LEU2-loxP introduced markers were rescued on expression of the Cre recombinase with plasmid pSH47. Final pSH47 plasmid loss was achieved by growth on YPD medium for 3 days. The eventually obtained triple S. cerevisiae pdr5 snq2 yor 1 mutant strain served as host for expression of the hER-ERE-GFP, hERß-ERE-GFP, and pYEX-GFP/hER plasmids. The plasmids were obtained by standard DNA manipulations according to Sambrook et al. (27)
and used to transform S. cerevisiae pdr5 snq2 yor1 cells to tryptophan, uracil, and leucine prototrophy by standard methods (28)
.
|
Media and growth conditions
All yeast cells were grown aerobically at 30°C. Nutritional requirements appropriate for selection and maintenance of mutants and plasmids in the transformed strain were scored on either liquid complete synthetic SDAP medium (29)
plus 0.5% D-glucose or minimal YNB media consisting of 0.67% yeast nitrogen base (YNB) with (NH4)2SO4, amino acids, and 0.5% D-glucose without uracil, tryptophan, and leucine (the latter to use the LEU2-d function of the pYEX-BX vector to increase for plasmid copy numbers) adjusted to pH 6.4 or 4.5 with NaOH and, where indicated, buffered with 50 mM citric acid monohydrate.
Assay conditions and fluorescence monitoring
For quantitative assessment of growth phenotypes and fluorescence development, logarithmic growing cells (70% budding) were diluted to a start optical density (OD)600 of 0.4 (Pharmacia Ultrospec 2000 Spectrophotometer) corresponding to 3.25 x 106 cells/ml. For each tested compound, at least three tests were carried out on different days. Each experiment consisted of six replica test cultures with respective controls and minimum four test concentrations. Test compounds were dissolved in DMSO and added to the test cultures of yeast strains transformed with the respective expression-reporter system in a total volume of 200 µl. The final concentration of the DMSO solvent did not exceed 0.5%. For serum samples, 20% charcoal-stripped serum (CSS; provided by Dr. Daxenbichler, University of Innsbruck) was used to dissolve E2 as positive control. The growth was estimated by end point OD600 measurements after 16.5 h incubation in transparent 96-well microtitre plates using a microplate reader (Tecan, Spectrofluor Plus). Tests were considered as valid when the turbidity of the control cultures increased at least fivefold during the incubation period. For fluorescence development read-outs, the excitation wavelength was adjusted to 485 nM and emission was observed at 535 (25 nM bandwith).
Data capture and evaluation
After 16.5 h incubation, obtained end point fluorescence (FL) values (corrected for blanks) were divided by growth determined as OD 600 (corrected for blanks, OD) for each replica well to normalize fluorescence for cell number (FL/OD). To increase the reproducibility of results, the FL/OD values obtained for a test compound at a given concentration were expressed as fractional values of the maximal response of a saturating concentration of the reference compound E2, applied in a concentration range from 0–150 pg/ml. The fractional values (response relative to the maximal E2 (RrelßÖ) at a given concentration of test compound (ct) were calculated according to RrelßÖ(ct) = (FL/OD(ct)–bottom)/(top–bottom), with "top" corresponding to the fitted FL/OD at saturating E2 concentration and "bottom" to the fitted FL/OD for the negative (solvent) control. The top and bottom values were obtained by Hill equation fit: 
with y(x) = FL/OD at the actual compound concentration, x = the decadical logarithm of compound concentration, LEC50 = decadical logarithm of EC50, top = fitted maximal FL/OD at saturating concentrations, bottom = fitted maximal FL/OD of negative control and hill_slope as the hill steepness parameter to the FL/OD values for each E2concentration using the R functionnls(The R Foundation for Statistical Computing, http://www.r-project.org/). Dose-response data for test compounds were obtained from the RrelßÖ(ct) for each test compound using the analogous fitting algorithm. EC50, EC20, and EC90values were calculated from the fitted dose-response curves, confidence intervals were determined using the R functionconfint. Relative potencies (E2=1) were determined as ratio of EC50of E2and the compound EC50.
Fluorescence microscopy
Microscopic images were obtained with a Leica TCS 4D confocal microscope (Leica, Wetzlar, Germany) with an 40 x 1.4 oil objective. The excitation source was an argon/krypton laser providing spectral lines for excitation at 488 nM (GFP) and 568 nM [nuclear stain with propidium iodide (PI)]. Fluorescence emission was filtered by a 530 nM band pass filter for GFP and a 590 nM long pass filter for PI.
RESULTS
A triply PDR5, SNQ2, and YOR1 deleted S. cerevisiae yeast strain was used as host for the heterologous expression of hER and hERß. The expression deployed a two plasmid strategy, the episomal expression of the individual receptors whereby the receptor cDNA was set under control of the constitutive CUP1 promoter and the episomal expression of the cis-acting estrogen hormone-responsive element (ERE) fused to the yeast-optimized green fluorescent protein (yEGFP; ref 24
). The advantage of using a S. cerevisiae mutant devoid of the three major transporters conferring pleiotropic drug resistance and thus inability to efficiently export small, hydrophobic molecules can be derived from the experiments given in Fig. 1
. Wild-type and the triply deleted mutant S. cerevisiae cells expressing the hERß were challenged with increasing tibolone concentrations (Fig. 1A
). The transactivation assay revealed a fivefold increased response and significantly higher sensitivity on application of tibolone (up to 50,000 pg/ml) with the genetically modified host. With wild-type expressed hER similar observations were obtained for 3-OH-tibolone (3-fold decrease in sensitivity), the phyto- and mykoestrogens (6- to 240-fold decrease, Fig. 1B
), and for E3 with the EC50 value of 67.8 vs. 3.46 nM (20-fold) in comparison to the mutant. A minor difference was observed for estradiol-3-benzoate with the EC50 value of 9.0 nM (wild-type) vs. 3.98 nM (mutant) and no difference for ethinyl-estradiol and mestranol (Fig. 1C
).
|
For visual inspection of subcellular localization and distribution, the hER-receptor was N terminally fused to GFP. The multicopy GFP tagged hER was then investigated in S. cerevisiae pdr5 snq2 yor1 cells (Fig. 2
) revealing an evenly cytosolic distributed fluorescence labeling except for the nucleus. No accumulated vacuolar labeling was observed, indicative of stable receptor fusion proteins not subjected to premature vacuolar degradation.
|
Test conditions
For dose-response curves, fluorescence readouts were obtained by measuring emission at 535 nM and normalization to cell growth on 16.5 h compound exposure in liquid media. Thereby, possible strain-specific and compound-induced altered growth characteristics with variations of maximum cell density in the stationary phase were taken into account, enabling consistent curve fits. Blank and negative control cultures served to correct for the accumulation of oxidized flavines in the late growth phase. Test validity criteria of at least fivefold increase in turbidity with distinguishable fluorescence emission were investigated by different E2 concentrations on hER in complete synthetic SDAP- and YNB-minimal medium (data not shown). Comparative data set evaluation revealed the latter as most suitable, which was used for all further investigations. Two different external pH values were investigated, pH 4.5 as the optimal yeast growth condition and pH 6.4 as related to complex, native (e.g., environmental origin or blood serum) samples. E2 exposure at pH 4.5, 6.4, and in 50 mM citric-acid buffered YNB growth medium (pH 6.4) did not alter the steroid induced response (data not shown), the latter was thus routinely adopted. In contrast, by investigation of incubation temperature and culture shaking (room temperature: 25°C vs. 30°C with and without shaking) conditions, constant shaking of the microplates turned out to be an essential factor. Routine application may thus be performed at room temperature or 30°C whereby incubation shaking is indispensable.
Comparison of compound mediated (anti) estrogenic bioactivity on hER and hERß
Several natural and synthetic compounds were analyzed for their (anti) estrogenic bioactivity. The sensitivity of the hER- and ß-bioactivity assay, represented by the calculated EC50 values as obtained from the fit of the Hill function by nonlinear regression, is given in comparison to other test systems in Table 3
. Among the natural hormonal compounds for E2 as perfect ligand, EC50 values of 0.22 nM and 0.41 nM with the hER- and ß-receptor, respectively, were determined. The stereo isomer 17-estradiol, previously shown as of 15-fold weaker potency compared with E2 with the hER-receptor (18)
, induced a similar weak response with the hERß-receptor (Fig. 3
and Table 3
). For the second main E2 metabolite E3, different sensitivities with EC50 values of 3.46 and 1.87 were determined for the hER- and ß-receptor, respectively.
|
|
Among the plant secondary metabolites for genistein as isoflavone, EC50 values of 4.47 vs. 76.4 nM were obtained with hER and hERß, respectively. Coumestrol as a member of the coumestanes was compared with the Fusarium spp. produced mycotoxin zearalenone, and only small differences in EC50 values between the receptors were observed (Table 3)
. Of the synthetic hormones, 17-ethinylestradiol exhibited the strongest estrogenic bioactivity, followed by diethylstilbestrol (DES), ß-estradiol-3-benzoate, mestranol, and tibolone and its derivatives with the hER-receptor. Interestingly, the hERß-receptor exhibited a fivefold higher sensitivity to mestranol in comparison to hER with EC50 values of 2.85 vs. 15.8 (Fig. 4
). The synthetic gestagens 17-hydroxyprogesterone and norethisteron did with neither the hER nor the hERß induce transcriptional activation of the fluorophore; correspondingly, no EC values could be determined as for the ER-receptor antagonists tamoxifen and 3-hydroxytamoxifen. All experiments were paralleled by the nonfunctional hER control strain (SCTD-10/04, truncated hER; ref. 18
) lacking most of the C-terminal ligand binding domain E and not responding to compound induced transactivation (data not shown).
|
As the synthetic steroid, the selective tissue estrogenic activity regulator (STEAR) tibolone and its 3- and 3ß-OH metabolites were tested with the yeast assay. As shown in Fig. 5
all three compounds, the origin substance as well as the active metabolites induced similar transactivation responses with both the hER- and ß-receptors. EC50 values between 88.4 nM for 3-hydroxyl-tibolone and 107 nM for both tibolone and 3ß-hydroxyl-tibolone with hER were however in marked contrast to approximately threefold lower EC50 values of tibolone (34 nM) and 3-hydroxyl-tibolone (45.4 nM) with hERß (Table 3)
. The EC50 value for 3ß-hydroxyl-tibolone was with 61.6 nM approximately twofold higher than for tibolone. The relative potencies of tibolone and its metabolites were 500-fold lower than E2 for hER and 100-fold lower for hERß.
|
Ex vivo samples
To compare human serum total estrogenic bioactivity with E2 content, parallel investigations were performed. Estrogenic bioactivity was determined with the S. cerevisiae pdr5 snq2 yor1 hER/ERE-GFP indicator strain assay containing 50% complex human serum samples (premenopausal) against a E2 calibration in 20% CSS and by standard receptor immune determination. Sample 1 corresponded to day 19 of the menstrual cycle (luteal phase) and was determined as 239 pg/ml E2 (5.9 pg/ml FSH) by receptor immune assay. Sample 2 corresponded to day 4 of the cycle (follicular phase), with receptor immune assay determined as 46 pg/ml E2 (15.1 pg/ml FSH). In contrast, for sample 1 total estrogenic bioactivity was, based on triplicate results, calculated to correspond to 650 pg/ml E2 and for sample 2 to correspond 610 pg/ml E2.
DISCUSSION
Current investigations focused on sensitized S. cerevisiae strains expressing either the hER or hERß receptor and pERE-GFP reporter plasmids to identify and assess (anti) estrogenic bioactivity. Since yeast do not contain endogenous steroid receptors, the indicator strains expressing the full functional hERs enable quantification of both the DNA binding and transcriptional activation function because the receptors are estrogen-induced and bind their own response element. Investigations comprised both sensitivity and specificity of the system. Implementation of a S. cerevisiae host strain devoid of three endogenous xenobiotic-transporting plasma membrane ATPases enhanced the sensitivity for steroids (Fig. 1)
except ethinyl-estradiol and mestranol (Fig. 1C
). Thus most of the tested natural and synthetic compounds appear to be substrates for either or all of Pdr5p, Snq2p, or Yor1p. However, for ethinyl-estradiol and mestranol, access to the cell interior appears to be not limited by PDR export. Since all tested compounds are between 260 and 320 kDa and zearalenone as largest is clearly a substrate for the PDRs (Fig. 1B
), size alone may not be the relevant substrate determining parameter. Side chains like the ethinyl group (ethinyl-estradiol and mestranol) and the additional 3-methyl-ether of mestranol may render them poor efflux substrates and enable cellular uptake. However, generally increased sensitivity of the S. cerevisiae mutant may assist adoption of this short term in vitro assay for environmental or ex vivo test settings.
Subcellular localization of the GFP-tagged hER in the yeast cells revealed an even cytosolic fluorescence pattern similar to the data of Kousteni et al. (30)
as could be expected for multicopy, constitutive CUP promoter driven expression of the fusion construct (Fig. 2)
, indicative of 1) no predominant retrograde transport or premature degradation of the heterologously expressed hER cDNA, and 2) correct yeast chaperone molecular interaction (10)
necessary to mediate the fit of hormones to the binding pocket and production of complexes to exert the transactivation effects.
Applicability
Various incubation conditions were investigated to identify the most suitable and practical protocol for the different requirements of (e.g., yeast growth, detection limits, handling, and robustness). Growth in YNB met the defined validity criteria as fitness parameter. High speed culture shaking (950 rpm) was identified as an essential factor. serving both oxygen supply and even cell distribution within the well and thus eventually growth and, if induced, GFP production. The yeast plasma membrane H+-ATPase Pma1 (31)
is the major source for cytosolic proton extrusion and generation of the electrochemical gradient (proton motive force) across the cellular membrane. In logarithmically growing yeast cultures the external pH can thus decline down to values of 3 or 3.5 in unbuffered media. Since no significant differences of E2 induced hER responses at external pH values of 4.5, 6.4, and buffered 6.4 were observed, we conclude the robustness of the assay.
Specificity and sensitivity
The specificity of the assay was confirmed by the nil transactivation response on application of the synthetic gestagens 17-hydroxyprogesteron and norethisteron and the ER receptor antagonists tamoxifen and 3-OH-tamoxifen with both hER and hERß (Table 3)
. In humans and other mammalians the response of different organs, tissues, and cell types on estradiol is dependent on the differential expression of the two estradiol receptors (and coactivators) and their intrinsic (binding) properties. Subtle differences were observed between the hER and -ß mediated responses for the endogenous hormones E2 and E3, for which in this assay approximately twofold different and opposed EC50 values with 0.22 vs. 0.41 nM (E2) and 3.46 vs. 1.87 nM (E3) for hER and hERß, respectively, were determined. Obtained values were in good agreement with other yeast assays (16
, 32
, 33)
and appeared considerably more sensitive than receptor binding assay results (34)
but less sensitive than the E-screen-assay with MCF-7 cells (34)
. For 17-estradiol, generally considered as being devoid of classical biological estrogen activity and only rarely detected in human serum or urine (35)
, the threefold lower EC50 of the hERß compared with the hER receptor (Table 3)
, also observed by Bovee et al. (33)
, emphasizes the different receptor affinities.
The most remarkable difference was among the natural compounds detected for genistein with 17-fold higher sensitivity of hER than hERß, similar to other yeast and receptor binding tests (33
, 34)
. Since for coumestrol related and only for zearalenone, twofold lower hERß EC50 values were determined and these data are in tendency comparable to other in vitro data (33
, 34
for coumestrol), the notion of greater affinity of the phytoestrogens for hERß (36)
can, at least by yeast tests, not be confirmed.
Among the synthetic estrogens hERß was, except for ß-estradiol-3-benzoate, found more sensitive whereby the strongest difference was observed for DES (12-fold higher sensitivity) followed by mestranol (5-fold) and 17-ethinylestradiol (2.7-fold). ß-Estradiol-3-benzoate, which itself does not bind to the receptor, requires metabolic conversion. The rather low EC50 value of 3.9 nM for both receptors is indicative of metabolic conversion by yeast esterases.
Tibolone and its metabolites
For the STEAR tibolone and its 3- and 3ß-OH metabolites, clear estrogenic effects, though considerably less potent than all other natural and synthetic compounds, were detected. Interestingly, twofold differences in sensitivity of hERß over hER were observed similar to receptor binding tests (37)
. In contrast to other in vitro tests (37)
, tibolone and the 3ß-OH metabolite induced similar transactivation activity with hER but with hERß the tibolone EC50 was twice as low as with the 3ß-OH metabolite (Table 3)
indicative of lower efficacy of this metabolite. The 
4-isomer was, due to its androgenic properties, not tested. The known estrogenic effects of tibolone on brain, bone, and vagina are mediated by the two major free 3- and 3ß-OH metabolites as a result of metabolic conversion by 3- and 3ß-hydroxysteroid dehydrogenase (3-/3ßHSD) conversion of the parent compound (38)
. 17ß- and 20HSD activity has been reported for S. cerevisiae (39)
but not yet for 3-/3ßHSD activity. The observed transactivation response might thus be either due to a so far uncharacterized specific 3-/3ßHSD activity, unspecific activity of yeast HSDs, or at least partial binding of tibolone to the estradiol receptors. Kinetic fluorescence development analysis (1 and 15 min interval resolution, data not shown) revealed almost identical inductions for tibolone and the 3-OH/3ß-OH-metabolites, indicative that conversion into the proposed active derivatives is not necessary to exert transcriptional activation in the yeast bioactivity assay.
Relative potencies
Based on the relative potencies (Table 3
, Fig. 6
), a ranking of compounds with the hER was determined as E2, 17-ethinylestradiol, DES, coumestrol, E3, ß-estradiol-3-benzoate, genistein, zearalenone, 17-estradiol, mestranol, 3-OH-tibolone, tibolone, and 3ß-OH-tibolone.
|
For hERß, the compound ranking was DES, 17-ethinylestradiol, E2, 17-estradiol, E3, zearalenone, mestranol, coumestrol, ß-estradiol-3-benzoate, tibolone, 3-OH-tibolone, 3ß-OH-tibolone, and genistein, revealing considerable differences between the receptor affinities and thus transactivation activity. Except for genistein all compounds tested in this assay displayed lower relative potencies (relative to E2) with hER than with hERß (Fig. 6)
. Taken together, such spectrum of estrogenic activity may characterize the estrogen receptor subtypes, potentially indicative for the structural differences within the C-terminal ligand binding (59% homology) domain.
Ex vivo samples
The potential role of the yeast test in routine laboratory diagnostics was investigated by correlation of total estrogenic bioactivity and E2 content in human serum samples. The E2 levels well reflected the luteal and follicular phases of the menstrual cycle. However, the equivalent total estrogenic bioactivity at both time points was 2.7 and 13.2 times higher, respectively, and with only little variation. Whether such high and rather constant estrogenic activity is common remains to be determined. Clinically, such information may be useful in the future both for monitoring total estrogenicity of women before and within menopausal hormone replacement therapy and pre- and postendocrinological disease treatment.
The results of the functional assay for estrogenic bioactivity, based on a receptor/reporter simple optical readout assay, demonstrated the versatility of applying a "simple" single-celled eukaryote for analyses of compounds belonging to different chemical classes as well as ex vivo samples.
ACKNOWLEDGMENTS
The authors thank Dr. H. J. Kloosterboer (N. V. Organon, The Netherlands) for the gift of tibolone and derivatives and critical reading of the manuscript and B. Kirberg for excellent technical assistance. The work was in part funded by EU QLK3-CT-2001–00401.
Received for publication November 30, 2005. Accepted for publication February 27, 2006.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and pdr5 snq2 yor1 (
) strains expressing the hERß receptor and ERE-GFP reporter constructs on exposure to increasing tibolone concentrations. Fluorescence emission was measured (excitation at 485 nM) at 535 nM and normalized to OD (OD600). Tibolone exerted its effects much stronger in the triply PDR transporter deleted strain resulting in a 5-fold increase in maximum transactivation response. B, C) Dose–response curve fits of S. cerevisiae wild type expressing hER
, EC50 of 1.1 µM), coumestrol (
, EC50 of 29.3 nM) and zearalenone (
, EC50 of 33.4 nM) EC50 values of mutant were considerably lower with 4.5, 2.5, and 5.3 nM, respectively. C) For ethinyl-estradiol (
, EC50 of 0.25 nM) and mestranol (
), mestranol (
), ß-estradiol-3-benzoate (
) and 3-beta-OH-tibolone (