Design, synthesis, biological evaluation and molecular docking studies of novel 3-aryl-4-anilino-2H-chromen-2-one derivatives targeting ERα as anti-breast cancer agents

Abstract

The estrogen receptor (ER) has played an important role in breast cancer development and progression and is a central target for anticancer drug discovery. In order to develop novel selective ERα modulators (SERMs), we designed and synthesized 18 novel 3-aryl-4-anilino-2H-chromen-2-one derivatives based on previously reported lead compounds. The biological results indicated that most of the compounds presented potent ERα binding affinity and possessed better anti-proliferative activities against MCF-7 and Ishikawa cell lines than the positive control tamoxifen. The piperidyl substituted compounds such as 16d and 18d demonstrated strong ERα binding affinities and excellent anti-proliferative activities respectively. Compound 18d displayed the most potent ERα binding affinity with RBA value of 2.83%, while 16d exhibited the best anti-proliferative activity against MCF-7 cells with IC₅₀ value of 4.52 ± 2.47 μM. Further molecular docking studies were also carried out to investigate binding pattern of the newly synthesized compounds with ERα. All these results together with the structure–activity relationships (SARs) indicated that these 3-aryl-4-anilino-2H-chromen-2-one derivatives with basic side chain could serve as promising leads for further optimization as novel SERMs.

Introduction

Breast cancer (BC) is the most frequent cancer among women worldwide. It had been estimated that 1.67 million new cases of BC were diagnosed in 2012 over the world according to a recent report.¹ Estrogen receptor α (ERα), a member of the large superfamily of nuclear receptors, is overexpressed and predominantly involved in more than 70% breast cancer patients.²

The estrogen receptor (ERα) has been an oncology target for the treatment of breast cancer since the 1960s.³ Since then, a number of selective ER modulators (SERMs) and down regulators (SERDs) have been discovered, and this field has been extensively reviewed.4, 5, 6 SERMs are a class of nonsteroidal compounds which act as antagonists in breast tissue but agonists in other tissues such as cardiovascular system and bone.7, 8 Due to their special action mode, SERMs remain as important anti-breast cancer agents with benefits in cardiovascular system and maintaining bone density compared with aromatase inhibitors and pure anti-estrogens.9, 10 Among them, tamoxifen (a SERM, 1a), containing a triphenylethylene scaffold and a basic side chain, was the first SERM approved in the US for the adjunctive treatment of metastatic breast cancer and have provided invaluable treatments for a number of patients.11, 12, 13 The search for new antiestrogen agents is still continuing, however, due to emerging drug-resistance mechanisms.¹⁴ More specifically, there is still an unmet medical need for a SERM with increased activity and less side effects.

The coumarin scaffold has been known to medicinal chemists for about 50 years for its antagonistic properties against the ER.15, 16 In recent years, 7-hydroxycoumarins, 7-hydroxychromenes and 7-hydroxychromanes, have been reported as selective ER ligands or modulators (SERMs) by many academic groups and pharmaceutical companies.17, 18, 19, 20, 21 SP500263 (3), a coumarin-based SERM, binds with high affinity to ERα and potently inhibits estrogen-dependent MCF-7 proliferation with similar IC₅₀ value to tamoxifen (Fig. 1).²² Recently, SS5020 (4) had been reported to downregulate the estrogen receptor with antitumor effects against chemically induced mammary tumors.²³ Structurally (Fig. 1), compound 4 can be viewed as a hybrid of the known SERM SP500263 (3) and SERD GW7604 (2b),²⁴ the latter of which is a descendant of 4-hydroxytamoxifen (1b). Furthermore, a related analogue of SS5020, compound 5 was also reported to be a selective ER down regulator (SERD) by the company AstraZeneca (Fig. 1).²⁵ However, 3-aryl-4-anilino-2H-chromen-2-one derivatives bearing basic side chain haven’t been investigated as SERMs till now.

The molecular basis of ER modulators and down regulators has been reported in a diverse variety of crystal structure determinations.26, 27 The crystal structure of GW5638 (2a) and compound 5, two SERDs, in complex with ERα shows that common phenyl acrylic acids overlay whereas the 3-aryl-2H-chromen-2-one skeleton overlay with the triphenylethylene scaffold (Fig. 2A) (PDB ID code: 1R5K). We thus hypothesized that replacing the phenyl acrylic acid of compound 5 with phenyl alkyl amine should be a reasonable approach to design novel coumarin containing SERMs whose basic side chain might overlay with the corresponding part of 4-hydroxyl tamoxifen.

In our process research of coumarin derivatives as SERMs, we assumed that it is possible to combine the hybrid fragments of compounds 3 and 5 to design a series of novel 3-aryl-4-anilino-2H-chromen-2-one derivatives bearing flexible basic side chains as ligands of ER. Further modification will focus on the substituents at C-7 position of coumarin skeleton and flexible basic side chains to obtain diverse derivatives (Fig. 2B). All the title compounds were evaluated for their ERα binding affinities and anti-proliferative activities against two human cancer cell lines (MCF-7 and Ishikawa). Several molecules exhibited comparable ERα binding affinity and anti-proliferative potency with Tamoxifen. Furthermore, molecular docking has been carried out. The possible binding modes with ERα and preliminary structure–activity relationships were also discussed.

The synthesis of the 4-[(2-aminoethoxy)]-substituted anilines 11a–f and the desired 4-anilino-3-aryl-2H-chromen-2-ones 15a–f, 16a–f and 18a–f is outlined in Scheme 1. The 4-[(2-aminoethoxy)]-substituted anilines 11a–f were obtained in four steps involving: i) reaction of p-nitrophenol 6 with 2-chloroethanol; ii) chlorination of the resulting 2-hydroxyethyl ether 8; iii) treatment of the 2-chloroethyl derivative 9 with selected amines; and, finally iv) reduction of the nitro group in the aminated derivatives 10a–f. As for intermediates 12–14, the synthesis procedure has been previously reported by our group.²⁸ Target compounds 15a–f, 16a–f and intermediates 17a–f were then furnished from their intermediates 12–14 by nucleophilic substitution with a variety of anilines 11a–f. Finally, compounds 18a–f were prepared from their corresponding intermediates 17a–f via Bn-deprotection in the presence of Pd-C in THF (Scheme 1).

The ERα binding affinities of test compounds were determined by a fluorescence polarization displacement assay using commercial estrogen receptor competitor assay kits with Tamoxifen as the positive control.29, 30, 31 The binding affinities are listed in Table 1, Table 2, Table 3 and are expressed as relative binding affinity (RBA) values, that is, relative to the affinity of estradiol (E2), which is set at 100%. The data can be analyzed in three groupings, compounds 15a–f, which bearing an H at C-7 position with variations in the basic side chain, compounds 16a–f with a methoxy group and 18a–f with a hydroxyl group. As shown in Table 1, most of title compounds 15a–f exhibited fair binding affinities towards ERα in comparison with Tamoxifen. Compounds 15c and 15d exhibited relatively higher potency with RBA values of 0.48% and 052% on ERα, respectively. The results indicated that the skeleton of 3-aryl-4-anilino-2H-chromen-2-one could favorably mimic that of estradiol while H substitution at C-7 position is unfavorable for ERα binding affinity.

To prove the previous hypothesis, we subsequently introduced a methoxy group into C-7 position of coumarin skeleton and evaluated the binding affinities of 16a–f. The results are shown in Table 2. Among compounds 16a–f, 16d displayed the most potent ERα binding affinity activity with RBA value of 0.95% while morpholinyl side chain containing compound 16e showed the lowest binding affinity in accordance with previous 15f. The improvement of the binding affinity on ERα was observed in this series compared with 15a–f (16a vs 15a, 16b vs 15b and 16d vs 15d). It was supposed that these ring A-expanded analogues 16a–f were assumed to be more efficiently filling the larger cavity of ERα.

As is known that the hydrogen bonds formed between the two hydroxyl group of estradiol and Glu353, Arg394 and His524 contribute a lot to keep the conformation stable when estradiol binds to ERα.³⁴ As the ability to form hydrogen bonds with these key amino acids is widely accepted to identify ERα ligands, we next evaluated the hydroxyl containing compounds 18a–f. Expectedly, it’s obviously to note that compounds 18a–f possess better binding affinity than the other two series. It could be inferred that compounds 18a–f containing hydroxyl groups on the C-7 position to mimic estradiol were able to form more hydrogen bonds with ERα than H substituted compounds 15a–f and methoxyl substituted compounds 16a–f, thus presenting better binding affinities.

Different amino substituents were also investigated, among all the compounds 15a–f, 16a–f and 18a–f, it seemed that the bulkier group (morpholinyl, N-methyl piperazinyl) were inferior to the others. It was unexpected that compound 15e and 16e which possessed morpholinyl exerted no binding activity. Dimethylamine, pyrrolidinyl and piperidyl were more preferable in the side chain. 18d with piperidyl as the basic group was the most potent compound with RBA value of 2.83% which is close to Tamoxifen (RBA = 7.64%).

To evaluate the anticancer activity, all synthesized compounds were screened against human breast cancer cell line MCF-7 (ER+) and human endometrial cancer cell line Ishikawa by MTT method with Tamoxifen as positive control. It is known that Tamoxifen will stimulate endometrial cell growth and induce endometrial cancer after long term treatment. Therefore it is of concern to evaluate the influence of new compounds targeting ER on endometrial cells. Results are summarized in Table 4.

As a global observation, most of the title compounds presented significant anti-proliferative activity on these two cancer cell lines. Moreover, the active compounds demonstrated better activity in breast cancer than in endometrial cancer cells, and most of them were more active than Tamoxifen in the two breast cancer cells. Compounds 16e and 18e with morpholinyl group in the side chain which showed no binding affinity in ERα assay were also not active in this test with IC₅₀ values of more than 100 μM, which indicated an ERα pathway on the anti-proliferative activity. However compound 15e exhibited moderate anti-proliferative activity on these two cancer cell lines but showed no binding affinity in ERα assay, whose anti-proliferative activity maybe acquired not only through ERα but a multi-target effect. Compounds 15f, 16f and 18f with N-methyl piperazinyl showed moderate activity on MCF-7 and Ishikawa but yet were less potent in their respective groups, which indicated the bulkier basic group was unfavorable in suppressing cancer cell growth.

The most potent compound against MCF-7 was 16d with an IC₅₀ of 4.52 μM, about 3-fold more potent than Tamoxifen (11.35 μM), but only showed moderate activity compared with other compounds against Ishikawa with the IC₅₀ valued at 11.58 μM. It was interesting to find that dimethoxy compounds (16a, 16c, 16d and 16f) which were less potent in ERα binding assay manifested slightly greater potency than the corresponding 7-hydroxyl compounds (18a, 18c, 18d and 18f) against MCF-7 which may also confirm that the obtained anti-proliferative activity was not only through ERα.

Analysis of the influence of different basic group on anti-proliferative activity suggested a preference for three medium size basic group, diethylamine, pyrrolidinyl and piperidyl (15b, 15c, 15d, 16b, 16c, 16d, 18b, 18c and 18d). For example, 16d, the most potent compound against MCF-7 and 18d against Ishikawa were both piperidyl substituted. Overall, compounds with the above three basic groups displayed the best activity against all of the two cancer cell lines.

In summary, replaced the H or OMe on C-7 position of coumarin framework with OH, most of the title compounds slightly decreased cell growth inhibitory potency in MCF-7 cell line. However, completely reverse trend was seen in ERα binding affinity. Interestingly, introduction of medium size basic groups such as pyrrolidinyl and piperidyl on terminal aniline could increase both the ERα binding affinity and anti-proliferative activities against two cancer cell lines. Furthermore, a significant decrease of anti-proliferative activity was observed by introduction of morpholinyl group indicating that this group is unfavorable in these series. The preliminary structure–activity relationship (SAR) results were summarized in Fig. 3.

The docking study was carried out using the Discovery Studio 3.0/CDOCKER protocol. The X-ray crystallographic structure of ERα complexed with 4-hydroxytamoxifen (OHT) was obtained from the RCSB Protein Data Bank (http://www.pdb.org/pdb/home/home.do) with PDB code 3ERT. The details about molecular docking process are described in the experimental section.

From the previous work, we hypothesized that the basic side chains of our target compounds and OHT should overlay whereas coumarin core should overlay with the triphenylethylene scaffold of OHT. To test this hypothesis, representative compounds 15d, 16d and 18d were docked into the active site of ERα. As is shown in Fig. 4, in the binding pocket of ERα, the triphenylethylene scaffold of 4-hydroxytamoxifen (red stick) is located into the binding pocket and the side chain stretches toward the edge of the pocket. Furthermore, an excellent superimposition of 15d, 16d, and 18d over the structure of OHT was observed, which suggested that the designed compounds exhibit the same binding patterns as that of OHT.

To further support our structural modifications, a deeper binding mode analysis was performed on the promising compounds 15d, 16d and 18d with structural diversities on the C-7 position of coumarin skeleton. In the docking study, the core skeleton of 3-aryl-4-anilino-2H-chromen-2-one is favorably positioned similar to 4-hydroxytamoxifen (OHT). The basic side chains of 15d, 16d and 18d pointed toward Asp351 to generate an antagonistic conformation as that of 4-OHT. The docking orientation and interactions of OHT within the ligand binding domain (LBD) of ERα are shown in Fig. 5A. The hydrogen bond interaction of 4-OH of OHT with Glu353 and Arg394 plays important role in stabilizing the binding mode. Compound 15d with H on the C-7 position of coumarin skeleton didn’t form any hydrogen bond with ERα, which is in accordance with its low binding affinity (Fig. 5B). However, the two methoxys containing 16d formed two hydrogen bonds with Arg394 and His524 respectively (Fig. 5C). Furthermore, these results also confirmed that the obtained anti-proliferative activities of 15d and 16d were not only through ERα but by a multiple way.

The plausible binding mode of 18d in the active site of ERα was shown in Fig. 5D. As we expected, the 7-OH of 18d formed hydrogen bonds with two amino acid residues in the hinge region, Glu353 and Arg394. It is noting that these two hydrogen bonds were consistent with OTH and contributed significantly to the good binding affinity with ERα. In addition, the p-methoxy of 18d interacted with His524 and formed a hydrogen bond. Molecular docking simulation supported the initial design strategy and suggested possible basis for the observed activities. Both biological date and docking results demonstrated that compounds 15d, 16d and 18d could be promising ERα inhibitors and their other potential mechanisms deserve further exploration.

By structural modifications on the coumarin and basic side chain moieties in prior leads 3 and 5, a series of 3-aryl-4-anilino-2H-chromen-2-one derivatives with basic side chain have been designed, synthesized and evaluated in ERα binding and anti-proliferative activity assays. Most of them presented moderate to excellent ERα binding affinities and better anti-proliferative effects in comparison with the positive control tamoxifen. The hydroxy containing compound 18d with piperidyl was found to be the most potential inhibitor on ERα while the methoxy containing compound 16d exhibited the best in vitro proliferation inhibition activities in cellular assays with IC₅₀ value of 4.52 μM on the MCF-7 cell line, which suggested that the obtained anti-proliferative activity might not be only through ERα. All these results together with molecular docking analysis led to a new class of promising ERα targeting candidates for the development of anti-breast cancer agents. Further derivatization of this series of compounds and investigation of multi-targeted mechanisms will appear in due course.