Estrogens (Es) play a crucial role in breast cancer (BC) development. Aromatase (CYP19), a cytochrome P450, is the enzyme that synthesizes Es. Using enzyme activity measurements (1–3), immunocytochemistry (4–7), and RT-PCR analysis (8, 9), aromatase is expressed at a higher level in human BC tissue than in normal breast tissue. Cell culture (10,11), animal experiments using aromatase-transfected BC cells (12, 13), and transgenic mouse studies (14) have demonstrated that in-situ produced E plays a more important role than circulating Es in breast tumor promotion. In addition, tumor aromatase has been shown to stimulate BC growth in both an autocrine and aparacrine manner (15). RT-PCR and gene transcriptional studies have revealed that in normal tissue, the aromatase promoter switches from a glucocorticoid-stimulated promoter, 1.4, to the cAMP-stimulated promoters, 1.3 and II, in cancerous tissue (9, 16, 17). Suppression of in-situ E biosynthesis can be achieved by the prevention of aromatase expression, or by the inhibition of aromatase activity in breast tumors. Our laboratory has devoted significant effort to understand the regulatory mechanism of aromatase expression in BC tissue and to determine the structure- function relationship of aromatase. The information obtained from our studies will help in developing approaches to repress aromatase/E biosynthesis in BC tissue.

Aromatase, a cytochrome P450, catalyzes three consecutive hydroxylation reactions converting C19 androgens to aromatic C18 estrogenic steroids. In this study, eight human aromatase mutants (I133Y, I133W, F235L, I395F, I474Y, I474W, I474M, and I474N) were prepared to evaluate the active site and a proposed hydrophobic pocket of the enzyme that exists in an aromatase model based on the X-ray structure of cytochrome P450cam. In addition, the binding characteristics of three steroidal inhibitors [4-hydroxyandrostenedione, 7alpha-(4'-amino)phenylthio-1,4-androstandiene-3,17-dione, and bridge (2,19-methyleneoxy)androstene-3,17-dione (MDL 101,003)] and four nonsteroidal inhibitors [aminoglutethimide, CGS 20267, ICI D1033, and vorozole (R83842)] were investigated through inhibitory profile studies on the eight new and three previously generated mutants (P308F, D309A, and T310S). The latter analyses have provided a molecular basis regarding how seven aromatase inhibitors with different structures bind to the active site of aromatase.

MCF-7 cells transfected with human placental aromatase gene (MCF-7Ca cells) or cells transfected with plasmid vector only (MCF-7Cc cells) were inoculated into nude mice with Matrigel. Tumors formed from both MCF-7Ca and MCF-7Cc cells grew faster in intact mice than in ovariectomized mice, suggesting that the tumors maintained their responsiveness to estrogen stimulation and that their growth was supported by ovarian estrogen. Injections of androstenedione (0.1 mg/mouse/day) to provide the substrate for aromatization to ovariectomized mice bearing MCF-7Ca tumors accelerated their growth but did not affect growth of MCF-7Cc tumors. This result indicates that local production of estrogen by intratumoral aromatase was sufficient to stimulate tumor growth. When ovariectomized mice with MCF-7Ca tumors supplemented with androstenedione were treated with aromatase inhibitors 4-hydroxyandrostenedione (1 mg/mouse/day, s.c.) or CGS 16949A (0.5 mg/mouse/day, s.c.), or with the antiestrogen tamoxifen (10 micrograms/mouse/day, s.c.), tumor growth was significantly inhibited. Tumor aromatase activity measured at the end of treatment was also inhibited by 4-hydroxyandrostenedione when the mice were sacrificed 4 h after the last injection. The tumors of this mouse model are dependent for their growth on estrogens from an endogenous nonovarian source. Thus, it simulates the situation in the postmenopausal breast cancer patient and could be used to evaluate the effect of aromatase inhibitors and antiestrogens.

Aromatase is present in human breast tumors and in breast cancer cell lines suggesting the possibility of in-situ estrogen production via the androstenedione to estrone and estradiol pathway. However, proof of the biologic relevance of aromatase in breast cancer tissue requires the demonstration that this enzyme mediates biologic effects on cell proliferation. Accordingly, we studied the effects of the aromatase substrate, androstenedione, on the rate of proliferation of wild-type and aromatase-transfected MCF-7 breast cancer cells. Androstenedione did not increase cell growth in wild-type MCF-7 cells which contained relatively low aromatase activity and produced 4-fold more estrone than estradiol. In contrast, aromatase-transfected cells contained higher amounts of aromatase, produced predominantly estradiol, and responded to androstenedione with enhanced growth. An aromatase inhibitor fadrozole hydrochloride, blocked the proliferative effects of androstenedione providing evidence for the role of aromatase in this process. As further evidence of the requirement for aromatase, cells transfected with the neomycin resistance expression plasmid but lacking the aromatase cDNA did not respond to androstenedione. These studies provide evidence that aromatase may have a biologic role for in-situ synthesis of estrogens in breast cancer tissue.

Aromatase catalyzes the conversion of 4-androstene-3,17-dione to estrogen with the concomitant formation of the minor metabolites 4-androstene-19-hydroxy-3,17-dione(19-hydroxyandrostenedione) and 4-androstene-3,17,19-trione(19-oxoandrostenedione). Microsomes of chinese hamster ovary (CHO) cells expressing human aromatase were isolated to investigate androstenedione metabolism. Relatively greater amounts of the minor metabolites result after limitation of electron flux from NADPH-cytochrome P450 reductase to aromatase. Substitution of NADH for NADPH or limitation of NADPH availability increased minor metabolite formation relative to estrogen formation. Similar changes in metabolite ratios were observed when metabolism was conducted either at high pH (8.3) or in the presence of n-alcohols in the range of 5-200 mM alcohol concentrations. However, conditions of low pH (5.5) or high ionic strength (1 M KCl) resulted in minor changes in metabolite ratios, suggesting little or no effect on electron flux between NADPH-cytochrome P450 reductase and aromatase. Theoretical molar ratios of the resulting metabolites were predicted using a reaction scheme assuming sequential substrate oxidations without reversible intermediate release from the aromatase active site. This model was supported by a close agreement between theoretical and experimental metabolite ratios for a broad range of NADPH concentrations. The results indicate that metabolite ratios provide a sensitive indicator of aromatase-oxidoreductase interactions in the microsomal environment.

Mutant aromatase cytochrome P-450s, expressed in CHO cells after transfection with cDNAs, have been characterized in terms of their catalytic efficiencies. After solubilization from microsomes, specific aromatase P-450 content of wild-type and mutants Pro308Phe, Asp309Asn, Asp309Ala and Phe406Arg was quantitated by a sandwich enzyme-linked immunosorbent assay (ELISA). Microsomal aromatase activity was determined by the 3H-water method using [1 beta-3H]androstenedione as substrate. Estimations of the actual turnover rate (catalytic efficiency) were derived from the combined data. The P-450 content in the mutants varied but was always less than that in the wild type. Hence, the decreases in the Vmax observed in the mutant enzymes did not correlate completely with reductions in catalytic effectiveness. In recent studies on the structure-function relationship of aromatase cytochrome P-450, the observed reduction of enzyme activity in terms of Vmax following site-directed mutagenesis led to the assumption that there was a corresponding loss of catalytic effectiveness. The present study reveals that a lower P-450 content can contribute significantly to decreasing catalytic activity in the mutants. In fact, in mutant Phe406Arg which exhibited virtually no catalytically active aromatase, the specific P-450 content was below the detectable level. Because of its location, the result of this latter mutation could be a major structural perturbation of the heme-binding property. Thus, interpretation of losses and reductions in aromatase activity resulting from single amino-acid replacement should take into account changes in the specific content of aromatase cytochrome P-450.

Several functional domains, especially the active site regions, in aromatase cytochrome P450 were inferred by alignment of amino acid sequences of the enzyme from five species, human, rat, mouse, chicken, and trout, and that of Pseudomonas putida cytochrome P450cam, whose x-ray structure has been determined (Poulos, T. L., Finzel, B. C., and Howard, A. J. (1987) J. Mol. Biol. 195, 687-700). The predicted functions of these domains have been evaluated by site-directed mutagenesis. Eighteen mutants, including seven new mutants, have been generated in this laboratory. The seven newly prepared mutants are Q123E, Q123H, T310S, T310C, R365K, R365A, and NDELTA20 (a mutant without the first 20 amino acids). The preparation and characterization of these new mutants are described. The structural model described in this paper should be very useful for future structure-function studies of aromatase by site-directed mutagenesis.

Mutant forms of aromatase cytochrome P-450 bearing modifications of amino acid residues Pro308 and Asp309 and expressed in transfected Chinese hamster ovary cells were subjected to kinetic analysis and inhibition studies. The Km for androstenedione for expressed wild type (11.0 +/- 0.3 nM SEM, n = 3) increased 4-, 25- and 31-fold for mutants Pro308Phe, Asp309Asn and Asp309Ala, respectively. There were significant differences in sensitivity among wild type and mutants to highly selective inhibitors of estrogen biosynthesis. 4-Hydroxyandrostenedione (4-OHA) a strong inhibitor of wild type aromatase activity (IC50 = 21 nM and Ki = 10 nM), was even more effective against mutant Pro308Phe (IC50 = 13 nM and Ki = 2.8 nM), but inhibition of mutants Asp309Asn and Asp309Ala was considerably less (IC50 = 345 and 330 nM and Ki = 55 and 79 nM, respectively). Expressed wild type aromatase and Pro308Phe aromatase were strongly inhibited by CGS 16949A (IC50 = 4.0 and 4.6 nM, respectively) whereas mutants Asp309Asn and Asp309Ala were markedly less sensitive (IC50 = 140 and 150 nM, respectively). CGS 18320B produced similar inhibition. Kinetic analyses produced Ki = 0.4 nM for CGS 16949A inhibition of wild type versus 1.1, 37 and 58 nM, respectively, against Pro308Phe, Asp309Asn and Asp309Ala. The results demonstrate significant changes in function resulting from single amino acid modifications of the aromatase enzyme. Our data indicate that mutation in Asp309 creates a major distortion in the substrate binding site, rendering the enzyme much less efficient for androstenedione aromatization. The substitution of Pro308 with Phe produces weaker affinity for androstenedione in the substrate pocket, but this alteration favors 4-OHA binding. Similarly, mutant Pro308Phe exhibits a slightly greater sensitivity to inhibition by CGS 18320B than does the wild type. These results indicate that residues Pro308 and Asp309 play critical roles in determining substrate specificity and catalytic capability in aromatase.

Aromatase, a cytochrome P-450, catalyzes the formation of aromatic C18 estrogenic steroids from C19 androgens. Using the x-ray structure of cytochrome P-450 cam as the model, seven mutants of human aromatase were designed and expressed in Chinese hamster ovary cells by a stable expression method. They are His-128 --> Gln, His-128 --> Ala, Cys-299 --> Ala, Glu-302 Leu, Asp-309 --> Asn, Asp-309 --> Ala, and Ser-312 --> Cys. The presence of the aromatase mutants in the transfected Chinese hamster ovary cells were confirmed by immunoprecipitation analysis. The kinetic parameters of these mutants using [1-beta, 2-beta-H-3] androstenedione (or [1-beta-H-3]androstenedione), and [1-beta, 2-beta-H-3]testosterone as substrates were determined. In addition, inhibition profiles for these mutants with two aromatase inhibitors, 4-hydroxyandrostenedione and aminoglutethimide were obtained. Furthermore, the reactions catalyzed by these mutants were examined by evaluating the levels of the product estrone, and two intermediates, 19-hydroxyandrostenedione and 19-oxoandrostenedione by reverse phase high performance liquid chromatography using [7-H-3]androstenedione as the substrate. Our results indicate that among the positions we modified, Asp-309 appears to be very important for the enzyme catalysis.

A mammalian cell expression plasmid, pH beta-Aro, containing the human placenta aromatase complementary DNA was constructed. The prepared plasmid was used to transfect breast cancer cells (MCF-7), noncancerous breast cells (HBL-100), and Chinese hamster ovary cells by a stable expression method. While the maximum velocities for aromatase expressed in three types of cells were different (10-201 pmol of [3H2O] formed/h/mg) using [1 beta, 2 beta-3H]androst-4-ene-3,17-dione as the substrate, the apparent Michaelis-Menten constants were found to be similar (39.9-57.8 nM) and were within the range determined for the enzyme existing in human placenta. The expressed activities were inhibited by the known aromatase inhibitors, 4-hydroxyandrostenedione and aminoglutethimide, at concentrations that normally inhibit the human placental aromatase. However, it was found that the inhibition profiles were different for aromatase expressed in different types of cells, suggesting that other factors, such as the uptake of the inhibitor, may also play a role in determining the inhibition efficiency. These constructed aromatase expressing mammalian cell lines will be very useful tools for aromatase inhibitor screening.