A yeast screen system for aromatase inhibitors and ligands for androgen receptor: yeast cells transformed with aromatase and androgen receptor.

Endocrine disruptors are hormone mimics that modify hormonal action in humans and animals. It is thought that some endocrine disruptors modify estrogen and androgen action in humans and animals by suppressing aromatase activity. Aromatase cytochrome P450 is the key enzyme that converts C19 androgens to aromatic C18 estrogenic steroids. We have developed a novel aromatase inhibitor screening method that allows us to identify antiaromatase activity of various environmental chemicals. The screen was developed by coexpressing the human aromatase and the mouse androgen receptor in yeast cells, which carry the androgen-responsive ss-galactosidase reporter plasmid. Functional expression of aromatase in yeast has been demonstrated using the [3H]-water release assay with intact cells as well as with yeast microsomes. The aromatase activity could be blocked by known aromatase inhibitors such as aminoglutethimide (AG). Yeast-produced androgen receptors were able to transactivate a yeast basal promoter linked to an androgen-responsive element in response to androgens. The resultant triple yeast transformant responded to the treatment of testosterone, androstenedione, or 5 alpha-dihydrotestosterone (5 alpha-DHT). In the absence of the aromatase inhibitor AG, transcriptional activation was observed only for the nonaromatizable androgen 5 alpha-DHT. However, the two aromatizable androgens (testosterone and androstenedione) induced the reporter activity in the presence of AG. Using this yeast-based assay, we confirmed that two flavones, chrysin and alpha-naphtholflavone, are inhibitors of aromatase. Thus, this yeast system allows us to develop a high-throughput screening method, without using radioactive substrate, to identify aromatase inhibitors as well as new ligands (nonaromatizable androgen mimics) for the androgen receptors. In addition, this screening method also allows us to distinguish nonandrogenic aromatase inhibitors from inhibitors with androgenic activity. This yeast screening method will be useful to screen environmental chemicals for their antiaromatase activity and for their interaction with androgen receptor. ImagesFigure 2Figure 3Figure 4Figure 5Figure 6Figure 7Figure 8Figure 9

The conversion of androgens to estrogens is catalyzed by the enzyme aromatase, which belongs to the cytochrome P450 superfamily. Aromatase plays an important role in the expression of secondary sexual characteristics, maintenance of pregnancy, and brain sex differentiation (1). Furthermore, the enzyme has received considerable attention in breast cancer development because high expression of the enzyme activity has been associated with a significant number of breast tumors (2)(3)(4). Thus, aromatase inhibitors are used as therapeutic agents for the treatment of estrogen-dependent breast cancer by inhibition of estrogen production in addition to antiestrogen treatment.
Aromatase inhibitors can be separated into two classes based on their mechanisms of action. Aromatase inhibitors such as 4hydroxyandrostenedione (4-OHA) and 1,4,6androstatrien-3,17-dione (ATD) are steroidal substrate analogs, which inhibit the enzyme activity by competing for the active site of the enzyme (5). The other class of inhibitors is nonsteroidal inhibitors such as aminoglutethimide (AG), which interact with the heme prosthetic group ofaromatase (6).
Endocrine disruptors are hormone mimics that modify hormonal action in humans and animals (7)(8)(9). These compounds indude chemicals isolated from plants (such as phytoestrogens) and man-made chemicals (such as dioxins and polychlorinated biphenyls).
There is increasing concern about how many such chemicals are present in our environment. There is a need for high throughput approaches to examine various environmental chemicals. Endocrine disruptors bind to the estrogen receptor or androgen receptor and induce many components of estrogen or androgen action (10)(11)(12). We feel that it is also possible that some of these compounds may act in an indirect fashion by inhibiting aromatase activity, resulting in a decrease in the level of estrogens or an increase in the level of androgens in cells. Research from other laboratories as well as our own has demonstrated that phytoestrogens such as flavones are competitive inhibitors of aromatase (13)(14)(15). Some flavones can inhibit aromatase with A? values similar to that of AG, an approved drug for treating breast cancer. It is possible that there are additional environmental chemicals that can suppress aromatase activity by competing for the binding of the androgen substrates or by interacting with the heme group ofaromatase.
The conventional method for aromatase inhibitor screening uses an in vitro enzyme assay with human placental microsomes as sources of the enzyme. However, the aromatase assay in such preparations depends on several components including short incubation time, the stability of NADPHcytochrome P450 reductase and aromatase in such preparations. Furthermore, the assay procedure is tedious and it would not be easy to adapt this assay for high-throughput screening. Although certain intact cells expressing aromatase, including human aromatase transfected mammalian cell lines (16), provide an ideal in vivo model for aromatase inhibitor screening, the cost of mammalian cell culture is a concern. In addition, these expression systems require an enzyme assay that uses radioactive substrates, which is a limiting factor for high-throughput screening. To circumvent these problems, we developed a novel method, which does not involve a radioactive assay, by coexpressing the human aromatase enzyme and the mouse androgen receptor in yeast cells carrying the androgen-responsive reporter plasmid. We also provide evidence that this high-throughput screening method can differentiate aromatase inhibitors with or without androgenic activity.

Materials and Methods
Chemicals and buffers. Restriction enzymes, T4 polynucleotide kinase, and T4 DNA ligase were purchased from New England Biolab (Beverly, MA). Taq polymerase was from Perkin-Elmer Cetus (Norwalk, CT).
Yeast strains. We used the Saccharomyces cerevisiae strain BJ2168 (Mat a, a prcl-407 prbl-1122 pep4-3 leu2 trpl ura3-52) (17). Expression of the human aromatase was under the control of a constitutive yeast promoter, triosephosphate dehydrogenase (TDH3); expression of the mouse androgen receptor was under the control of a regulated yeast promoter, metallothionein promoter (CUP1) as described previously (18). Growth and transformation of yeast cells were performed according to standard procedures (19). Construction ofyeast expression vector for the mouse androgen receptor. A cassette that contained the metallothionein promoter (CUP 1)-ubiquitin gene-linker containing unique Eagi and BssHII sites followed by the iso-1-cytochrome C (CYCI) terminator was inserted into the BamHI and the SphI sites (multiple cloning sites) of the yeast vector (YEp351). The mouse androgen receptor cDNA was excised from the previously described yeast expression vector (YEpmAR) (20) after complete digestions with Eag and BssHII and was subsequently ligated to YEp351. The resultant androgen receptor expression vector (YEpARc) ( Figure IA) was used to transform the yeast strain, BJ2168, using lithium acetate protocol (19). Transformants were selected by leucine auxotrophy.
Construction of androgen-responsive reporter plasmid. The reporter plasmid (YEpA) was used to study the transcriptional activation of the mouse androgen receptors produced in yeast cells (20). Briefly, an oligonucleotide containing two copies of the glucocorticoid/progesterone response element (GRE/PRE) from the tyrosine aminotransferase gene (21) was inserted into the unique XboI site of the yeast reporter plasmid, pCl, as described previously (18,20). The resulting reporter plasmid (YEpA) ( Figure 1B) contains an androgen response element (GRE/PRE) upstream of the CYCI promoter that is fused to the lac-Z gene of Escherichia coli. These plasmids were used to transform yeast cells expressing the mouse androgen receptor and the final transformants (YEpARc/YEpA) were selected by leucine and uracil auxotrophy.
Construction ofyeast expression vector for human aromatase. Two unique restriction sites were engineered at the 5' and the 3' ends of the human placental aromatase coding sequence (16,22) for insertion into the yeast expression vector (YEpV5) described previously (18). This was achieved by PCRmediated mutagenesis.
The resultant aromatase cDNA contained an Eagl site in front of the start codon and a DraIll site right after the stop codon. To produce an in-frame ubiquitin-fusion receptor under the control of the yeast TDG3 promoter, an oligonucleotide containing the last six amino acids of the ubiquitin gene followed by an Eagl and a DraIII site was inserted into the AflII-Kpn sites of the yeast expression vector, YEpV5. After complete digestions with Eagl and DraIII, the aromatase cDNA was inserted into the yeast expression vector (YEpV5). The resultant aromatase expression vector (YEpARO) ( Figure  IC) was used to transform BJ2168 yeast strain carrying the androgen receptor expression construct and the androgen-responsive reporter plasmid (YEpARc/YEpA). The final triple transformant yeast strain (YEpARc/ YEpARO/YEpA) was selected by leucine, uracil, and tryptophan auxotrophy. Aromatase assay. The aromatase activity was measured by [3H]-water assay (23) directly on yeast cultures. The [3H]-water release assay for human aromatase expressed in mammalian cells and yeast was validated in our laboratory by a product isolation assay (16,22). [3H]-Androstenedione at a final concentration of 46 nM was added to 1 mL transformed yeast strain (YEpARc/YEpARO/ YEpA) when cell density reached the late log phase. After 1-hr incubation at 30°C, the reaction was stopped by addition of 200 1sL 10% (weight/volume) trichloracetic acid and 100 pL dextran-coated charcoal (2% norit A and 0.2% dextran in water). After 10 min incubation, the suspension was centrifuged and the supernatant was again treated with dextran-coated charcoal. After centrifugation, the supernatant was filtered through a small column fitted with glass wool and the tritiated water content was determined by scintillation counting. For the in vitro enzyme assay, microsomes were prepared from yeast cells as follows: Yeast cells were washed with enzyme buffer (50 mM Tris-HCI and 1 mM EDTA, pH 7.0) and disrupted with glass beads. This and subsequent procedures were performed at 40C. The cell suspension was centrifuged at 1,000 x g for 10 min, and the supernatant was centrifuged at 10,000 x g for 30 min. The resulting supernatant was further centrifuged at 100,000 x g for 30 min to obtain the yeast microsomal pellet. Yeast microsomes (20 pL) were added to 180 jL enzyme buffer containing NADPH (0.2 mg/mL) and [3H]-androstenedione at concentration as indicated. After 20 min incubation at 37°C, the reaction was stopped by the addition of 20 pL 50% trichloroacetic acid. The rest of the procedure was identical to that described for the yeast culture assay.
Preparation ofyeast extracts and receptor-binding assays. Yeast extracts were prepared for androgen receptor-binding assay as described previously (18,20).
Volume 107, Number 1 1, November 1999 * Environmental Health Perspectives Articles * Yeast screening methods of aromatase inhibitors Transcription assays. The triple transformant yeast strain was grown in enriched medium (without tryptophan, uracil, and leucine) containing 20 nM androstenedione in the presence or absence of aromatase inhibitors. When cell density reached the late log phase, cells were harvested by centrifuigation and subsequently disrupted by vortexing with glass beads. Cell suspension was centrifuged at 10,000 x g for 10 min and the supernatant (yeast extract) was assayed for [3-galactosidase activity as described previously (18,20).
Plate assayfor high-throughput screening. An overnight culture yeast strain (YEpmARc/ YEpARONYEpA2) (5%) was inoculated into dextrose-agar medium maintained at 50°C. After the addition of vitamins, amino acids, x-gal, and 20 nM androstenedione, the final culture was poured onto a 15-cm petri plate and allowed to cool at room temperature. Discs containing different aromatase inhibitors were placed on top of the agar and the plate was incubated for 2 days at 300C. As a control, another culture plate containing the same constituents without 20 nM androstenedione was also set up simultaneously to detect the androgenic activity of aromatase inhibitors.

Results and Discussion
Concept ofcoexpressing androgen receptor and aromatase in yeast as a novel screening methodfor antiaromatase activity. In the present study, we developed a screening method to identify aromatase inhibitors without using an enzyme assay with radioactive substrates. This concept was achieved by coexpressing androgen receptor and aromatase in yeast cells carrying the androgen-responsive reporter plasmid. If the triple transformant yeast cells carrying the three plasmids as described in Figure 1 are incubated with an aromatizable androgen (testosterone or androstenedione), the androgen diffuses into the cell, where it will either bind to the yeast-produced androgen receptor or will be converted to estrogen by aromatase produced in yeast. If the conditions (concentrations and choice of substrates) favor the enzymatic reactions, most of the androgen will be converted into estrogen but not bind to the androgen receptor and therefore cannot transactivate the yeast basal promoter linked to the androgenresponsive element (GRE/PRE). The end result of this androgen metabolism within the yeast cells is the inhibition or reduction of reporter enzyme 3-galactosidase induction. However, androgen-dependent transcriptional activation will be apparent as reflected by the reporter enzyme induction if an aromatase inhibitor is included in the yeast medium. As a first step toward the development of this novel screen, the two expression systems for androgen receptor and aromatase in yeast must be characterized and validated separately.
Androgen-dependent transcriptional activation in yeast. We have previously shown that many of the steroid/retinoid receptor superfamily can function as ligandactivated transcription factors in yeast cells (18,20,24,25). In this study, a new yeast vector carrying the CUPI promoter and a LEU2 gene was constructed to express the mouse androgen receptor. To confirm receptor expression, yeast extracts were prepared from the yeast strain carrying this expression vector (YEpARc) under copper induction and analyzed for receptor-binding activities. Scatchard analysis indicated that the yeastproduced mouse androgen receptor binds to androgen with high affinity with a Kd value of 0.96 nM and a maximum binding capacity of 135 fmol/mg ( Figure 2). The binding specificity of these receptors was also examined by labeling the extracts with 10 nM [3H]-5a-DHT in the presence or absence of 1-, 10-and 100-fold molar excess of radioinert competitors (Figure 3). Yeast-produced androgen receptors exhibit steroid-binding specificity characteristics of the authentic mammalian androgen receptor. The functional expression of the androgen receptor in yeast was further confirmed by transforming the yeast strain (YEpARc) with a reporter plasmid containing two copies of a GRE/PRE of the glucocorticoid responsive tyrosine aminotransferase gene upstream of the yeast promoter, CYCI, which is fused to the lac-Z gene of E. coli. The resulting double transformant yeast strain (YEpARcIYEpA) was analyzed for ligand-dependent transcriptional activation. In the presence of 10 pM cupric sulfate, the three androgens (androstenedione, testosterone, or 5a-DHT) were able to induce the P-galactosidase activity in the absence or presence of AG (1 mM) ( Figure  4). These results indicate that aromatase inhibitors such as AG do not interfere with the binding of androgens to the expressed androgen receptor. We also observed that the reporter enzyme could be induced (3to 4-fold) by high concentrations of androgen (. 1 pM) in yeast cells in the absence of copper, indicating that the CUP promoter is slightly leaky (data not shown).
Expression ofaromatase in yeast. Baker's yeast (S. cerevisiae) is an ideal host for expressing mammalian cytochrome P450s, including aromatase (22). In the present study, we used the previously described ubiquitin-fusion technology (18,20,21) Figure 5). The Km value is similar to those reported for aromatase in human placental microsomes, but the Vma. is lower than those reported previously in placental microsomes (26). Paradoxically, this lower level of enzyme expression in yeast actually enhances the sensitivity of the screen because reporter enzyme induction can be detected in the presence of lower concentrations of aromatase inhibitor. The three aromatase inhibitors (4-OHA, ATD, and AG) were tested on the yeast strain carrying the aromatase expression construct (YEpARO). As shown in Figure 6, 4-OHA and ATD were effective inhibitors, with a median inhibitory concentration of 0.1 pM. These two inhibitors were able to inhibit 90% of the aromatase enzyme activity at a concentration of 1 pM, whereas the nonsteroidal inhibitor AG required 1 mM to achieve a similar inhibitory effect. Relatively high concentrations of aromatase inhibitors were needed to suppress aromatase in yeast. This observation is likely because of the poor ability of the compounds to diffuse into the yeast cells. We are evaluating approaches to increase the sensitivity of the assay. Although aromatase inhibitor screening using the yeast strain expressing the human aromatase would be a good approach because of rapid cell growth and the stability of the expression plasmid attributed by the auxotrophic marker, this screening method involves an enzyme assay, which will require a large quantity of radioactive substrates when high-volume screening is desired. An ideal aromatase inhibitor screening should handle a large number of samples with good selectivity, sensitivity, and capacity. To accomplish this goal, we coexpressed the androgen receptor and the aromatase enzyme in yeast cells carrying the androgen-responsive reporter plasmid.
Coexpressing aromatase and androgen receptor in yeast. The yeast strain expressing the aromatase enzyme was cotransformed with the androgen receptor expression vector and the androgen-responsive reporter plasmid. The three androgens (testosterone, androstenedione, and 5a-diydrotestosterone) were tested for reporter enzyme induction in the triple transfromant yeast strain (YEpARc/YEpAROIYEpA). As shown in Figure 7, the aromatizable androgens androstenedione and testosterone were not able to induce the reporter enzyme in the absence of the aromatase inhibitor, AG. However, ligand-dependent transcriptional activation was apparent in the presence of AG (1 mM). Addition of AG alone in the yeast culture did not induce reporter enzyme activity. In contrast to aromatizable androgens, the nonaromatizable androgen, 5a-DHT, induced the reporter gene efficiently in the presence or absence of AG. Thus, these data clearly support our concept of coexpressing the androgen receptor and the 18 To develop a high-throughput screening assay, the yeast strain carrying the three plasmids (YEpARc, YEpARO, and YEpA) were inoculated on agar plates, which contained the chromogenic substrate, x-gal, and androstenedione (20 nM). Discs containing 10 pg of the three aromatase inhibitors, 4-OHA, ATD, and AG, were placed on top of the agar ( Figure 8A). After 48 hr incubation at 300C, blue halo formations as a result of 3-galactosidase induction were detected surrounding these inhibitors. To distinguish androgenic from nonandrogenic aromatase inhibitors, the three inhibitors were also tested with the same yeast strain in the absence of androstenedione ( Figure 8B). Both 4-OHA and ATD exhibited transcriptional activation as reflected by the formation of  blue halos surrounding the discs, whereas the nonsteroidal inhibitor, AG, was inactive on this control plate. The observation indicates that 4-OHA and ATD are androgen mimics. The detection of an androgenic activity of these inhibitors may be due to the fact that a rather high amount of compounds was used. These compounds reportedly have weak androgenic activity (27)(28)(29)(30).
The binding affinity of 4-OHA to the androgen receptor is 0.1% that of mibolerone (30). Using this yeast screening method, we confirmed our previous report (13) that two endocrine disruptors, chrysin and a-naphtholflavone, are aromatase  (Figure 9). Naringenin was weakly positive in the yeast screening method. We will evaluate the method using other compounds that inhibit aromatase. For a highthroughput screening, compounds from a 96-well plate can be delivered by a robot onto a large agar plate that can hold 576 samples. Thus, a large number of compounds can be screened each day. The sensitivity of this screening assay also depends on the choice and concentrations of substrates used because residual substrates will lead to high background. We found that androstenedione is a better substrate than testosterone and that the optimal concentration of androstenedione used is 20 nM. In theory, the androgen receptor and the androgen response element can be substituted with the estrogen receptor and the estrogen response element for this nonradioactive screen. In the latter case, an aromatizable androgen is converted to estrogen, which binds to the estrogen receptor and subsequently induces the reporter enzyme. On a plate assay, the yeast cells will turn blue as a result of reporter enzyme induction. The dear halo will appear in the presence of an aromatase inhibitor. In practice, it is more difficult to identify activity against a blue background. Furthermore, the screen using estrogen receptor expression does not provide information on androgenicity of the inhibitor. In addition, it seems that there are more estrogenic environmental chemicals identified than androgenic chemicals. This will create problems in such yeast screening in that aromatase noninhibitors cannot be distinguished from estrogen mimics. Thus, Figure 9. Testing of endocrine disruptors in the yeast screening assay. Abbreviations: AG, aminoglutethimide; C, chrysin; N, naringenin; a-N, naphtholflavone. Three flavones (C, N, and a-N) and the aromatase inhibitor (AG) were tested in the assay in the presence of 20 nM androstenedione, under conditions described in Figure 8. The numbers in the brackets indicate micrograms of compounds on the discs. Blue halos surrounding the discs indicate p-galactosidase enzyme induction.
we have chosen the androgen-responsive transcription unit for this screen development.
In summary, we have reconstituted for the first time a yeast transcription unit to study androgen action and aromatase simultaneously. This reconstituted system allows us to develop a high-throughput screening method without using radioactive substrate to identify aromatase inhibitors as well as new ligands (nonaromatizable androgen mimics) for the androgen receptors. In addition, this screening method also allows us to distinguish nonandrogenic aromatase inhibitors from inhibitors with androgenic activity. This yeast screening method will be useful to screen environmental chemicals for their antiaromatase activity and for their interaction with androgen receptor.