Molecular basis for the interaction of four different classes of substrates and inhibitors with human aromatase

Abstract

Aromatase cytochrome P450 (CYP19) converts androgen to estrogen. In this study, the interactions of four classes of compounds, 17β-estradiol (the product of aromatase), 17-methyltestosterone (a synthetic androgen), dibenzylfluorescein (a synthetic substrate of aromatase), and coumestrol (a phytoestrogen), with aromatase were investigated through spectral analysis using purified human recombinant aromatase and site-directed mutagenesis studies using CHO cells expressing wild-type human aromatase or five aromatase mutants, E302D, D309A, T310S, S478T and H480Q. Spectral analysis showed that a type I binding spectrum was produced by the binding of 17-methyltestosterone to aromatase and a novel binding spectrum of aromatase was induced by dibenzylfluorescein. Mutagenesis experiments demonstrated that residues S478 and H480 in the β-4 sheet play an important role in the binding of all four compounds. Computer-assisted docking of these compounds into the three-dimensional model of aromatase revealed that: (1) weak interaction between 17β-estradiol and the β-4 sheet of aromatase facilitates the release of 17β-estradiol from the active site of aromatase; (2) 17-methyl group of 17-methyltestosterone affects its binding to aromatase; (3) dibenzylfluorescein binds to the active site of aromatase with its O-dealkylation site near the heme iron and residue T310; and (4) coumestrol binds to aromatase in a manner such that rings A and C of coumestrol mimic rings A and B of steroid. These structure–function studies help us to evaluate the structural model of aromatase, and to accelerate the structure-based design for new aromatase inhibitors.

Introduction

The female hormone, estrogen, plays an important role in breast cancer development. Upon binding to estrogen, estrogen receptor (ER) activates transcription of its target genes, which are responsible for cancer cell proliferation in estrogen-dependent breast tumor. The most important estrogens in humans are estrone (E1) and 17β-estradiol (E2), which are synthesized from androstenedione and testosterone, respectively, by aromatase cytochrome P450 (CYP19) enzyme. The third-generation aromatase inhibitors (AIs), which include two triazole derivatives, anastrozole (Arimidex) [1] and letrozole (Femara) [2], and one steroid analogue, exemestane (Aromasin) [3], are currently used clinically for the treatment of hormone-dependent breast cancer in postmenopausal women [4], [5], [6], [7].

The structure–function characterization of aromatase would provide useful information for the understanding of the reaction mechanisms of the enzyme and for the design of new potent AIs. However, the structural basis of drug binding to aromatase is not well understood because the three-dimensional (3-D) crystal structure of aromatase is unknown. Previous studies from our laboratory and others have made considerable progress towards characterizing the structure–function relationship of aromatase [8], [9], [10], [11], [12], [13], [14]. Recently, a theoretical 3-D model of aromatase (PDB code 1TQA) was built based on the crystal structure resolved at 2.55 Å of warfarin-bound human CYP2C9 [14]. We have proposed a new clamping mechanism of androstenedione and exemestane binding to the active site of aromatase, based on the results from site-directed mutagenesis, proteomic analysis, and computer-assisted protein–ligand docking using the 3-D model of aromatase [15]. This clamping mechanism provides insight into the interaction of aromatase with its steroidal ligands.

In this study we investigated the interaction of aromatase with four classes of compounds, E2, 17-methyltestosterone (17MT), dibenzylfluorescein (DBF), and coumestrol, each having its own unique structural features (Fig. 1). E2, the aromatization product of testosterone by aromatase, has the same steroid scaffold as testosterone. The major structural difference between E2 and testosterone is that E2 has a planar aromatic A ring, while testosterone has a non-aromatic six-carbon A ring. In a recent publication [15], we discussed how the androgen substrate binds to aromatase and how the aromatization reaction occurs; however, how this structural conversion by the aromatization reaction leads to the release of E2 from aromatase remains a mystery. 17MT is a synthetic androgen used to treat men with a testosterone deficiency. It is also used as a component in contraceptive. 17MT has an extra 17 alpha-methyl group compared to androstenedione and testosterone, whereas this 17 alpha-methyl makes it a poor substrate of aromatase [16]. DBF is a fluorescein derivative and a substrate for CYP enzymes, such as CYP3A4, CYP2C8, and aromatase. The benzyl ether group at position 6 of DBF (as shown in Fig. 1) can be O-dealkylated by CYP enzymes [17]. DBF is an effective substrate for aromatase. The relatively large molecular size of DBF when compared to the steroidal substrates of aromatase leads us to question how DBF binds to the active site of aromatase. Coumestrol, a phytoestrogen, has a similar structural scaffold to a steroid. It is a potent antagonist for ER; however, it was found to be a weak competitive inhibitor of aromatase [18].

These four compounds contain special structural characters. The structure–function characterization of aromatase in this study helps us to examine the accuracy of the predicted active site pocket of aromatase and the hypothetical clamping mechanism of ligand binding.