European Journal of Medicinal Chemistry 240 (2022) 114569

Contents lists available at ScienceDirect

European Journal of Medicinal Chemistry

journal homepage: www.elsevier.com/locate/ejmech

4th generation nonsteroidal aromatase inhibitors: An iterative SAR-guided

design, synthesis, and biological evaluation towards picomolar dual
binding inhibitors
Ahmed G. Eissa a, 1, Denise Barrow a, Julia Gee a, Lauren E. Powell b, Paul A. Foster b, c,
Claire Simons a, *
a
School of Pharmacy and Pharmaceutical Sciences, Cardiff University, King Edward VII Avenue, Cardiff, CF10 3NB, UK
b
Institute of Metabolism and Systems Research, College of Medical and Dental Sciences, University of Birmingham, Birmingham, B15 2TT, UK
c
Centre for Endocrinology, Diabetes, and Metabolism, Birmingham Health Partners, Birmingham, B15 2TT, UK

A B S T R A C T

One in every eight women will be diagnosed with breast cancer during their lifetime and approximately 70% of all patients are oestrogen receptor (ER) positive
depending upon oestrogen for their growth accounting for third generation aromatase (CYP19A1) inhibitors being the mainstay in the treatment of ER-positive breast
cancer. Despite the success of current aromatase inhibitors, acquired resistance occurs after prolonged therapy. Although the precise mechanisms of resistance are not
known, lack of cross resistance among aromatase inhibitors drives the need for a newer generation of inhibitors to overcome this resistance alongside minimising
toxicity and adverse effects. Novel triazole-based inhibitors were designed based on previously published parent compound 5a, making use of the now available
crystal structure of CYP19A1 (PDB 3S79), to make modifications at specific sites to explore the potential of dual binding at both the active site and the access channel.
Modifications included adding long chain substituents e.g. but-2-ynyloxy and pent-2-ynyloxy at different positions including the most active compound 13h with IC50
value in the low picomolar range (0.09 nM). Aromatase inhibition results paired with molecular dynamics studies provided a clear structure activity relationship and
favourable dual binding mode was verified. Toxicity assays and CYP selectivity profile studies for some example compounds were performed to assess the safety
profile of the prepared inhibitors providing the basis for the 4th generation nonsteroidal aromatase inhibitors.

1. Introduction (SERDs) e.g. fulvestrant (Fig. 1) [10]. The second strategy is oestrogen
deprivation via inhibition of aromatase (CYP19A1) responsible for
Breast cancer is a complex collection of heterogenous neoplastic, oestrogen synthesis e.g. anastrozole, letrozole and exemestane (Fig. 1)
often recurrent, diseases affecting 1 in 8 women [1–4]. The number of [9,10]. Despite the success, high efficacy and selectivity of the currently
women diagnosed with breast cancer has risen from 1.68 million to 2.1 available third generation aromatase inhibitors (AIs), acquired resis­
million between 2012 and 2018 ranking first place of highest incidence tance eventually occurs after prolonged therapy along with some
malignances among women with 24% and only second to lung cancer cross-activity to other cytochrome P450 (CYP) family members (e.g.
among all populations [3–5]. Different subclasses of breast cancer are anastrozole’s inhibition of CYP1A2, letrozole’s inhibition of CYP2A6)
responsible of 14.3% of all female cancer related deaths [5,6], approx­ and some androgenic and/or weak ERα agonistic activity [11–13].
imately 70% of which show dependence on oestrogen/ER signaling for Considerable efforts have been made to date in relation to designing
their growth, thus called ER positive [7]. further compounds, some with improved IC50 values compared with the
As early as the 1890s, hormonal targeted therapy gained a crucial clinically-approved reference compounds and so with promising AI ac­
role over the years in oestrogen deprivation and control of breast cancer tivity. This is the case either for steroidal or nonsteroidal AIs especially
[8,9]. ER-α is often directly involved in tumour growth providing the after the crystal structure of aromatase (PDB 3EQM) was published [14].
basis for two different classes of antihormonal therapy; interference with Therefore the design and synthesis of a new generation of inhibitors is
binding of oestrogen to the receptor, which is divided into selective needed to widen the therapeutic options available owing to the risk of
oestrogen receptor modulators (SERMs) that are competitive inhibitors resistance towards available drugs, and also to minimise toxicity and
of the ER e.g. tamoxifen, and selective oestrogen receptor degraders reduce the non-specific and adverse effects by increasing aromatase

* Corresponding author.
E-mail address: simonsc@cardiff.ac.uk (C. Simons).
1
Present Addresses: Ahmed G. Eissa, Department of Medicinal Chemistry, Faculty of Pharmacy, Zagazig University, Zagazig P.C. 44519, Egypt.

https://doi.org/10.1016/j.ejmech.2022.114569
Received 13 May 2022; Received in revised form 17 June 2022; Accepted 23 June 2022
Available online 6 July 2022
0223-5234/© 2022 The Author(s). Published by Elsevier Masson SAS. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

selectivity [7,11,15]. This led to a continuous exploration of alternative

strategies with allosteric inhibition rising as a possibility especially after
letrozole and one of the metabolites of tamoxifen; namely endoxifen,
were reported to have the potential of non-competitive/mixed inhibi­
tion for aromatase [16,17]. This type of inhibition indicates the presence
and identification of allosteric sites in aromatase. Three potential allo­
steric sites were identified through computational studies including the
haem proximal site along with two access channels connected to the
active site [18–20]. The front door access channel gated by Asp309 was
identified in the crystal structure published in 2009 (PDB 3EQM) [14].
On the other side of the active site, a backdoor channel gated with
Ser314 was discovered through the crystal structures published in 2018 Fig. 2. Reported (A) steroidal and (B) xanthone O-pent-2-yne derivatives with
(PDB 5JKV, 5JKW, 5JL6, 5JL7 and 5JL9), which is postulated to be aromatase dual binding capacity.
involved in the passage of catalytic water [14,21].
Allosteric non-competitive inhibition of aromatase offers major ad­ so this concept is gaining more popularity and may present the molec­
vantages over conventional inhibitors in terms of reaching maximum ular basis for the design of 4th generation non-steroidal AIs [9,18,22].
inhibition without the complete blockage of oestrogen production, As an early part of this project, 1-[(6-methoxybenzofuran-2-yl)phe­
which would reduce side effects and delay or avoid the onset of resis­ nylmethyl]triazoles (Fig. 3, R1 = H, R2 = OCH3, R3 = varying func­
tance along with better selectivity owing to allosteric sites being less tional groups) were previously reported as potent aromatase inhibitors,
conserved across other enzymes. Also, they are not outcompeted by high presenting a suitable scaffold for incorporation of the hydrophobic tail
concentrations of androgens as natural substrate, making this approach required for filling the access channel in a dual orthosteric/allosteric
appealing for rational drug discovery [18]. Even though there are no inhibition [23]. In this paper, three main variants were studied, the first
reported effective allosteric inhibitors of aromatase, some novel steroi­ of which was the position of the hydrophobic tail (Fig. 3, R1, R2 or R3) to
dal compounds with a dual-binding ability for the access channel and provide an answer to the preferable binding mode. Then the length of
the active site were reported in 2012 [22]. the hydrophobic tail was investigated as it was not guaranteed that a
These potent inhibitors were characterised by a long alkyne side five-carbon chain length would be the optimal length for activity in a
chain at C6 of the steroidal scaffold that could fit through the narrow non-steroidal compound. Along with the nature of the secondary sub­
hydrophobic access channel, which is the main transport route for ste­ stitution, a more comprehensive SAR was developed with toxicity
roids, water molecules and proton sources, giving rise to the concept of assessment and CYP selectivity of the compounds with promising aro­
4th generation steroidal aromatase inhibitors with dual binding capacity matase inhibitory activity. These questions were addressed through an
(Fig. 2A). The alkyne side chains are proposed to sit in the access iterative process of a design/synthesis/inhibitory evaluation cycle to
channel preventing entry of other molecules and to displace water guide the rational drug design in three distinct comparisons (Fig. 3).
molecules believed to act catalytically, by proton relay through the
Arg192-Glu483 salt bridge pair at the channel entrance and Asp309 2. Results and discussion
within the channel, in the enolization of androstenedione [22]. This is
extendable to the non-steroidal AI class as some nonsteroidal xanthone 2.1. Chemistry
compounds with the same pent-2-yne arm (Fig. 2B) were reported to
have aromatase inhibitory activity in low micromolar range in 2020 and The synthetic pathways for the previously reported parent AIs (5)
and the newly synthesized modified compounds (6 and 7) are outlined

Fig. 1. Clinically used selective oestrogen receptor modulators (SERMs, e.g. tamoxifen), selective oestrogen receptor degraders (SERDs e.g. fulvestrant) and aro­
matase inhibitors (AIs, e.g. steroidal exemestane, non-steroidal anastrozole and letrozole).

2
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

Fig. 3. Proposed process of identifying optimal non-steroidal [(benzofuran-2-yl)phenylmethyl]triazole derivatives as aromatase dual binding site inhibitors.

in Scheme 1. Initially, the parent methoxy substituted compounds were yields (52–95%) and the alcohols (10) obtained in quantitative yield.
prepared using the reported synthetic pathway as indicated in Scheme 1 The introduction of the triazole proceeded with loss of the pyran pro­
with modifications to the methods made where necessary to optimize tection to give the triazole phenol compounds (11 and 12, 44–86%). A
the reactions [23]. final nucleophilic substitution step was required at the end of the syn­
The Rap-Stoermer formation of the benzofurans was performed thetic pathway to add the longer alkyloxy chains through Williamson
using the original method with NaH as the base, however it was replaced ether synthesis to give the 6-O-alkyl/alkyne (13) and 5-O-pent-2-yny­
by K2CO3 as reported by Mahboobi et al., to improve the yield (80–96%) loxy (14) derivatives. The fluoro and chloro derivatives were obtained in
[24]. Also, the equivalents of SOCl2 in the third step were increased from good yields (50–90%), however the nitrile derivatives (13g and 13h)
equimolar to 1.6 equivalents to optimize the method by reducing the were obtained in low yields (7 and 17% respectively) owing to complex
reaction time from 4 days to only 16 h with yields ranging from 31 to reaction mixtures.
63%. The 4′ -pent-2-ynyloxy compound (19) with the extended substitu­
The new molecules with extended substitutions on the benzofuran tion on the phenyl side of the compound required a slightly modified
ring, ranging from ethoxy to pent-2-ynyloxy, were prepared via a pathway (Scheme 3) with the first step being a demethylation of 3c to
divergent pathway (Scheme 2) starting with the hydroxy salicylalde­ form 15, which was then pyran protected (16) and proceeded as
hyde derivative after being pyran protected (8) [23,25]. Following the described in Scheme 2 from the reduction step (17).
methods described in Scheme 1, the ketones (9) were obtained in good The final triazole products were confirmed by 1H and 13C NMR with

Scheme 1. Synthesis of methoxy substituted compounds (5, 6 and 7). Reagents and conditions: (i) K2CO3, CH3CN, 70 ◦ C, 3 h, 80–96%; (ii) NaBH4, dioxane, rt, 2h,
quantitative; (iii) SOCl2, triazole, CH3CN, 0 ◦ C, 1 h, then 4, CH3CN, K2CO3, rt, 16 h, 31–63%.

3
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

Scheme 2. Synthesis of the 6- and 5-hydroxy and extended compounds (12–15). Reagents and conditions: (i) K2CO3, CH3CN, 70 ◦ C, 3 h, 52–95%; (ii) NaBH4, dioxane,
rt, 2 h, quantitative; (iii) SOCl2, triazole, CH3CN, 0 ◦ C, 1 h, then 10, CH3CN, K2CO3, rt, 16 h, 44–86%; (iv) K2CO3, CH3CN, 40 ◦ C, 1h, then alkyl bromide, rt, 16
h, 7–90%.

purity determined by HPLC with all compounds ≥95% pure. The two 2.2. Aromatase inhibition
characteristic singlets of the triazole group were observed at ~8.2 and
8.0 ppm in 1H NMR and the CH-triazole as either a singlet or finely split Aromatase activity was assayed using a modified titrated water assay
triplet (J = 0.9 Hz) at ~ 6.5 ppm in 1H NMR and at ~ 61.5 ppm in 13C previously reported [26]. Briefly, placental choriocarcinoma JEG-3
NMR. cells, known to have high aromatase activity, were grown to approxi­
mately 80% confluence in six-well culture plates. Once established, cells
were treated with androst-4-ene-3,17-dione[1β-3H] as aromatase sub­
strate. Aromatase activity was measured in the absence and presence of

4
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

Scheme 3. Synthesis of the 4′ -hydroxy and pent-2-ynyloxy derivatives (18 and 19). Reagents and conditions: (i) HBr, CH3COOH, 110 ◦ C, 16 h, 57%; (ii) 3,4-dihydro-
2H-pyran, p-toluenesulfonic acid, EtOAc, rt, 18 h 73%; (iii) NaBH4, dioxane, rt, 2 h, quantitative; (iv) SOCl2, triazole, CH3CN, 0 ◦ C, 1 h, then 17, CH3CN, K2CO3, rt,
16 h, 45%; (v) K2CO3, CH3CN, 40 ◦ C, 1 h, then 1-bromopent-2-yne, rt, 16 h, 44%.

Table 1
Comparison of 6-, 5- and 4′ -hydroxy, methoxy and pent-2-ynyloxy substitution.

Compound R CYP19A1 IC50 (nM) 95% Confidence interval (nM)

6-substitution
11a H 0.79 0.724–0.857
5a CH3 0.47 0.449–0.494
13a CH2C–
–CCH2CH3
– 2.76 2.296–3.314
5-substitution
12 H 5.67 4.965–6.468
6 CH3 2.01 1.646–2.452
14 CH2C–
–CCH CH
– 2 3 21.05 16.35–27.11
4′ -substitution
18 H 30.25 25.52–35.86
7 CH3 >1000 –
19 CH2C–
–CCH CH
– 2 3 >1000 –
Letrozole 0.70 0.556–0.883

Table 2
Comparison of the chain length at the optimal 6-position.

Compound R CYP19A1 IC50 (nM) 95% Confidence interval (nM)

5a CH3 0.47 0.449–0.494

13c CH2CH3 0.46 0.378–0.562
13d CH2C–
–CH
– 1.03 0.674–1.573
13b CH2C–
–CCH3
– 0.53 0.479–0.589
13a CH2C–
–CCH2CH3
– 2.79 2.296–3.314
Letrozole 0.70 0.556–0.883

5
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

Table 3
Comparison of secondary substitution (R3) of 6-hydroxy, 6-methoxy, 6-butynyloxy and 6-pentynyloxy derivatives.

Compound R R3 CYP19A1 IC50 (nM) 95% Confidence interval (nM)

Cl substitution
11a H Cl 0.79 0.724–0.857
5a CH3 Cl 0.47 0.449–0.494
13b CH2C–
–CCH
– 3 Cl 0.53 0.479–0.589
13a CH2C–
–CCH CH
– 2 3 Cl 2.76 2.296–3.314
F substitution
11b H F 0.39 0.359–0.431
5b CH3 F 0.15 0.101–0.215
13f CH2C–
–CCH3
– F 4.1 3.283–5.270
13e CH2C–
–CCH CH
– 2 3 F 0.51 0.419–0.619
CN substitution
11c H CN 0.56 0.504–0.612
5c CH3 CN 0.11 0.092–0.126
13h CH2C–
–CCH3
– CN 0.09 0.078–0.126
13g CH2C–
–CCH2CH3
– CN 0.72 0.677–0.759
Letrozole 0.70 0.556–0.883

inhibitors (0.001pM–100pM). Aromatase activity results were deter­ aromatase inhibition also observed for the propynyloxy (13d, IC50 1.03
mined as a concentration of product formed per mg of protein per hour. nM) compared with the pentynyloxy (13a, IC50 2.79 nM), however for
Each data point was measured in triplicates and the error in the IC50 binding in both haem binding site and access channel the longer alky­
calculations represented as 95% confidence interval. nyloxy substitutions were preferred.
The aromatase inhibitory activity of the 6-, 5- and 4′ -hydroxy, The effect of the secondary substitution, i.e. substitution of the
methoxy and pent-2-ynloxy derivatives, with the secondary chloro phenyl ring, on the 6-but-2-ynyloxy and 6-pent-2-ynyloxy derivatives
substituent were first evaluated to determine the effect of the position of was then evaluated. The choice of secondary substituents, Cl, F and CN,
the hydroxy/methoxy/pent-2-ynyloxy groups and the chain length on was determined from our previous published research [23] as the most
aromatase inhibition (Table 1). The 6-substituted hydroxy (11a),
methoxy (5a) and pent-2-ynylkoxy (13a) derivatives showed optimal
aromatase inhibition (IC50 0.79, 0.47 and 2.76 nM respectively) with a
reduction in activity observed for the 5-substituted hydroxy (12), Table 4
methoxy (6) and pent-2-ynylkoxy (14) (IC50 5.67, 2.01 and 21.05 nM CYP IC50 (μM) profile of lead compounds 13h and 13e.
respectively). A significant reduction was observed for the 4′ -substituted
Compound 1A2 2C9 2C19 2D6 3A4 19A1
hydroxy (18), methoxy (7) and pent-2-ynylkoxy (19) (IC50 30.25 nM, >
1 μM and >1 μM respectively), which would indicate the 4′ -position to 13e 5.3 ± 9.2 ± 5.21 ± >25 21 ± 0.00051
0.33 1.87 0.88 4.82
be unfavourable for binding and orientation within the aromatase
13h 2.88 ± >25 2.17 ± >25 >25 0.00009
binding sites. 0.16 0.18
Having established that the 6-position was optimal, a broader range
Control standards: CYP1A2 α-naphthoflavone IC50 0.02 ± 0.002 μM, CYP2C9
of chain lengths at the 6-position of the benzofuran ring was investigated
sulfaphenazole IC50 0.245 ± 0.05 μM, CYP2C19 tranylcypromine IC50 14.4 ±
(Table 2). The methoxy (5a, IC50 0.47 nM), ethoxy (13c, IC50 0.46 nM)
1.62 μM, CYP2D6 quinidine IC50 0.137 ± 0.015 μM, CYP3A4 ketoconazole IC50
and but-2-ynyloxy (13b, IC50 0.53 nM) were optimal with good 0.076 ± 0.002 μM.

Fig. 4. Toxicity of final compounds tested at 1 μM for 48 h treatment followed by BrdU proliferation assay in A) MCF-10A and B) MDA-MB-231 cells. Stats are one-
way ANOVA followed by a Tukey’s Multiple Comparison test comparing all compounds against control (Con) and doxorubicin (Dox, 1 μM). Data represents n = 5–6
technical replicates ±SEM. ***p < 0.001 compared to control. NS – Non-significant compared with control.

6
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

promising substitutions for aromatase inhibition. 2.3. Toxicity (BrdU) assays

For the chloro and nitrile secondary substitutions, the but-2-ynyloxy
derivatives (13b and 13h respectively) performed best with low pico­ Examples of the compounds, namely 5b, 11b, 13e, and 13h, were
molar inhibitory activity observed for the nitrile derivative (13h, IC50 tested at 1 μM over 48 h along with doxorubicin (Dox) as positive control
0.09 nM), while for the fluoro secondary substitution, the pent-2- by bromodeoxyuridine (BrdU) proliferation assay to evaluate the
ynyloxy derivative (13e, IC50 0.51 nM) was optimal (Table 3). toxicity against non-cancerous breast epithelial cells (MCF-10A) and
non-oestrogen dependent breast cancer cells (MDA-MB-231). Statistics
using one-way ANOVA followed by a Tukey’s Multiple Comparison test

Fig. 5. 3D fitting of the final frame after 150 ns MD simulation in the active site of the aromatase enzyme for (A) the 6-methoxy derivative S-5a (B) the 6-but-2-
ynyloxy derivative S-13b.

Fig. 6. Positioning of ligands with respect to the haem of the final frame of (A) R-13e (B) S-13e (C) R-13h (D) S-13h after 400 ns MD simulations.

7
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

Fig. 7. Protein-ligand interactions of R- and S-13h within the haem and access channel binding sites of CYP19A1.

comparing all compounds against control showed no significant differ­ 14) the S-enantiomer also showed a more favourable binding profile,
ence between the tested compounds and the negative control indicating while the R-enantiomer of 14 binds to the haem through the N2 of tri­
that the compounds had no impact on MCF-10A or MDA-MB-231 growth azole. In contrast, the R-enantiomers of the phenyl substituted de­
(Fig. 4). These results suggest possible limited toxicity in normal breast rivatives (e.g. 19) bound with the haem through the N4 of the triazole
tissue and little off-target effects. ring, while the S-enantiomers bound with the haem through the N2 of
the triazole ring. For optimal haem binding the azole should be
2.4. CYP panel and selectivity perpendicular (~90◦ ) to the plane of the haem group, which is normally
achieved through binding with the N4 of the triazole, however binding
The two lead compounds, 1-((4-fluorophenyl)(6-(pent-2-yn-1-yloxy) via the N2 of the triazole does not allow a perpendicular interaction,
benzofuran-2-yl)methyl)-1H-1,2,4-triazole (13e) and 4-((6-(but-2-yn-1- which would result in a less favourable conformation for binding.
yloxy)benzofuran-2-yl)(1H-1,2,4-triazol-1-yl)methyl)benzonitrile Varying the length of the alkyl chain substituent at the 6-position of
(13h), were tested for inhibitory activity against a CYP panel (1A2, 2C9, the benzofuran group did not show a significant difference in the sta­
2C19, 2D6 and 3A4) by Cyprotex Discovery Limited using a human liver bility of the complex and all compounds showed good 3D fitting, with
microsomal assay with a CYP isoform specific probe substrate [27]. The the methoxy derivatives fitting in the haem active site pocket e.g. 5a
lead fluoro pent-2-ynyloxy lead compound (13e) displayed excellent (Fig. 5A), and the longer chain substituted derivatives fitting both the
selectivity for CYP19A1 compared with CYPs 3A4 and 2D6 (>277,000 haem active site and the access channel gated by Arg192, Asp309,
and 49,000 respectively) and very good selectivity compared with CYPs His480 and Glu483 e.g. 13b (Fig. 5B) with the chloro substituent
1A2, 2C9 and 2C19 (10,000, 18,000 and 10,000 respectively) (Table 4). forming a binding interaction with the key amino acid Met374.
Likewise, the nitrile but-2-ynyloxy lead compound (13h) showed The S-enantiomer of the lead compounds, 6-pent-2-ynyloxy 4′ fluoro-
excellent selectivity for CYP19A1 compared with CYPs 3A4, 2C9 and substituted (13e) and 6-but-2-ynyloxy 4’ nitrile-substituted derivative
2D6 (>277,000) and very good selectivity compared with CYPs 1A2 and (13h) were optimally positioned within the haem and access channel
2C19 (32,000 and 24,000 respectively) (Table 4). binding sites and formed a direct N4-triazole haem binding interaction
with a distance of 2.37 and 2.38 Å respectively (Fig. 6C and D). The pent-
2-ynyloxy chain of R-13e was not optimally positioned in the access
2.5. Computational studies channel and, although the triazole N4 was close to the haem iron (2.43
Å) a direct bond was not observed (Fig. 6A). For both enantiomers of
Using the crystal structure of human placental aromatase (CYP19A1) 13h the nitrile group formed binding interactions with Met374 and
refined at 2.75 Å (PDB 3S79) [21], the compounds (R- and S-enantio­ directly or indirectly with Arg115 (Fig. 6B and D) and the but-2-ynyloxy
mers) were docked using molecular operating environment software chain was optimally positioned within the access channel.
(MOE) [28], and the best poses were selected based upon the 3D visual
inspection and the score value of the ligand-protein complex (Table S1), 3. Conclusion
which were all subjected to 150 ns MD simulations with longer 400 ns
simulations run for exemplar compounds, using the Desmond pro­ The 6-O-alkyl/alkyne benzofurans (5, 13) were optimal with respect
gramme of Schrödinger software [29,30] to equilibrate and establish an to aromatase inhibitory activity compared with the 5-O-alkyl/alkyne
optimal complex. Regarding the position of the substituent, the benzofurans (6, 14) and 4′ -O-alkyl/alkyne phenyl derivatives (7, 19)
6-substituted benzofurans (13) and the 4-substituted phenyl derivatives (Table 1) and the 6-O-alkyl/alkyne benzofurans equilibrated with more
(19) formed more stable complexes than the 5-substituted analogues stable aromatase-ligand complexes over the MD simulations (Fig. S1).
(14) (Fig. S1). Further investigation of the 6-O-alkyl/alkyne benzofuran derivatives
Studying the binding profile through the simulation time (Fig. S2) with either Cl, F or CN substitutions in the 4′ -position of the phenyl ring
and the ligand interactions of the final frame of the MD simulation determined that the 6-O-but-2-yne chain was optimal for aromatase
(Fig. S3) indicated that both enantiomers of the simpler methoxy de­ inhibitory activity of the chloro (13b) and nitrile (13h) substituted
rivative (5a) can comfortably fit in the haem binding pocket. Intro­ phenyl rings (IC50 0.53 and 0.09 nM respectively), while the 6-O-pent-2-
duction of the extended pentynyloxy group restricted fit within the yne chain was optimal for aromatase inhibitory activity of the F (13e)
enzyme with the 6-pentynyloxy S-13a more optimally positioned for the substituted phenyl ring (IC50 0.51 nM) (Table 3). Computational studies
access channel compared with R-13a, although both bind with the haem indicated that the alkyne chain was positioned in the front door access
bind through the N4 of triazole. For the 5-pentynyloxy benzofurans (e.g.

8
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

channel gated by Arg192, Asp309, His480 and Glu483 with additional 152.36 (CH), 150.35 (C), 143.26 (CH), 135.25 (C), 134.22 (C), 129.33 (2
stabilizing hydrophobic interactions. Binding interaction with Met374 x CH), 129.02 (2 x CH), 127.92 (C), 114.33 (CH), 112.08 (CH), 108.14
was also noted for the chloro (13b) and nitrile (13h) derivatives (Fig. 6 (CH), 103.82 (CH), 61.57 (CH), 55.93 (CH3). HPLC (A): 99.9% at R.T. =
and S3). 8.9 min.
From computational studies it is proposed that the improved aro­ 1-((5-Chlorobenzofuran-2-yl)(4-methoxyphenyl)methyl)-1H-
matase inhibitory activity of the nitrile derivative (13h) may be owing 1,2,4-triazole (7). Prepared from (5-chlorobenzofuran-2-yl)(4-
to the ability of both R- and S-enantiomers to fit optimally within both methoxyphenyl)methanol (4c) [23] (0.5 g, 1.74 mmol). Purification by
binding pockets and interact with the haem through the N4 of the tri­ gradient column chromatography afforded the product (7) at 60%
azole ring with anchoring of the ligands through a H-bonding interac­ EtOAc in petroleum ether (v/v) as a yellow oil. Yield: 0.18 g (31%); Rf =
tion between CN and Met374/Arg115 (Fig. 7), and future experimental 0.25 (petroleum ether - EtOAc 3:1 v/v). 1H NMR (CDCl3): δ 8.30 (s, 1H,
studies will involve experimental studies to validate the computational CH-triazole), 8.01 (s, 1H, CH-triazole), 7.43 (d, J = 1.7 Hz, 1H, Ar), 7.30
findings. The excellent toxicity and selectivity profile of these com­ (d, J = 8.6 Hz, 1H, Ar), 7.23 (d, J = 8.7 Hz, 2H, Ar), 7.20 (dd, obscured
pounds also supports the basis and rationale for continuing research, by CDCl3, 1H, Ar), 6.88 (d, J = 8.7 Hz, 2H, Ar), 6.75 (s, 1H, Ar), 6.45 (s,
subject to funding, for these 4th generation nonsteroidal aromatase 1H, CH), 3.75 (s, 3H, OCH3). 13C NMR (CDCl3): δ 160.47 (C), 154.29 (C),
inhibitors. 153.66 (C), 151.06 (CH), 129.35 (2 x CH), 128.99 (C), 128.84 (C),
126.64 (C), 125.51 (CH), 121.08 (CH), 114.66 (2 x CH), 112.56 (CH),
4. Experimental section 107.08 (CH), 62.02 (CH), 56.35 (CH3). HPLC (B): 97.8% at R.T. = 6.33
min.
4.1. Materials and instrumentation for the chemical synthesis 2-((4-Fluorophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-
6-ol (11b, R3 = F). Prepared from (4-fluorophenyl)(6-((tetrahydro-2H-
All commercially available starting materials and solvents were of pyran-2-yl)oxy)benzofuran-2-yl)methanol (10c) [23] (0.71 g, 1.97
general purpose or analytical grade and used without further purifica­ mmol). Purification by gradient column chromatography afforded the
tion. Melting points were determined using a Gallenkamp melting point product (11b) at 70% EtOAc in petroleum ether (v/v) as a yellow wax.
apparatus and are uncorrected. 1H and 13C NMR spectra were recorded Yield: 0.46 g (86%); Rf = 0.22 (petroleum ether - EtOAc 1:1 v/v). 1H
on a Bruker Advance DP500 spectrometer operating at 500 MHz and NMR (DMSO‑d6): δ 9.61 (bs, 1H, OH), 8.70 (s, 1H, CH-triazole), 8.08 (s,
125 MHz, respectively. Chemical shifts are given in parts per million 1H, CH-triazole), 7.52 (m, 2H, Ar), 7.38 (d, J = 8.5 Hz, 1H, Ar), 7.29 (t,
(ppm) relative to the internal standard tetramethylsilane (Me4Si). J = 8.9 Hz, 3H, Ar), 6.75 (d, J = 1.9 Hz, 1H, Ar), 6.75 (dd, J = 2.1, 8.4
Analytical thin layer chromatography (TLC) was carried out on pre­ Hz, 1H, Ar), 6.49 (t, J = 1.0 Hz, 1H, CH). 13C NMR (DMSO‑d6): δ 161.42
coated silica plates (ALUGRAM® SIL G/UV254) with visualisation via (C), 159.47 (C), 154.34 (C), 154.18 (C), 150.32 (CH), 150.08 (C), 142.53
UV light (254 nm). HPLC were either performed by the Department of (CH), 131.14 (C), 128.49 (CH), 128.42 (CH), 120.01 (CH), 117.60 (C),
Pharmacy & Pharmacology, University of Bath, Bath, UK on a Zorbax 114.10 (CH), 113.93 (CH), 110.90 (CH), 105.35 (CH), 95.91 (CH), 57.77
Eclipse plus C18 Rapid resolution 2.1 × 50 mm, 1.8 μm particle size (CH). HPLC (B): 98.3% at R.T. = 4.04 min.
using gradient (methanol: H2O) with 0.1% formic acid (method A) or in 2-((4-Chlorophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-
house on a Shimadzu LC-2030C Plus C18 Rapid resolution 250 × 4.6 5-ol (12). Prepared from (4-chlorophenyl)(5-((tetrahydro-2H-pyran-2-
mm, 5 μm particle size using isocratic 80:20 (methanol: H2O) (method yl)oxy)benzofuran-2-yl)methanol (10b) [23] (0.71 g, 1.97 mmol). Pu­
B). All biologically evaluated compounds are ≥95% pure by HPLC rification by gradient column chromatography afforded the product
analysis. (12) at 70% EtOAc in petroleum ether (v/v) as a thick yellow oil. Yield:
Methods for the preparation of the ketones (3, 9), alcohols (4, 10, 0.5 g (78%); Rf = 0.2 (petroleum ether - EtOAc 1:1 v/v). 1H NMR
17) and compounds 15 and 16 are described in the Supporting infor­ (DMSO‑d6): δ 9.22 (bs, 1H, OH), 8.72 (s, 1H, CH-triazole), 8.09 (s, 1H,
mation. Triazoles 5a-c, 11a and 11c were previously reported by us CH-triazole), 7.52 (d, J = 8.6 Hz, 2H, Ar), 7.47 (d, J = 8.5 Hz, 2H, Ar),
[23]. 7.33 (m, 2H, Ar), 6.92 (d, J = 2.46 Hz, 1H, Ar), 6.75 (dd, J = 2.5, 8.7 Hz,
1H, Ar), 6.51 (t, J = 0.95 Hz, 1H, CH). 13C NMR (DMSO‑d6): δ 154.20
4.2. General method for the preparation of the 1-(substituted benzofuran- (C), 153.97 (C), 152.53 (CH), 149.23 (C), 144.82 (CH), 135.83 (C),
2-yl)(4-substituted phenyl)methyl)-1H-1,2,4-triazole (5, 6, 7, 11, 12 and 133.94 (C), 130.30 (2 x CH), 129.32 (2 x CH), 128.62 (C), 114.19 (CH),
18) 111.94 (CH), 107.47 (CH), 106.32 (CH), 59.86 (CH). HPLC (B): 95.6% at
R.T. = 4.36 min.
To a cooled suspension of triazole (4 m.eq.) in dry CH3CN (3 mL/ 4-((5-Chlorobenzofuran-2-yl)(1H-1,2,4-triazol-1-yl)methyl)
mmol of the carbinol (4, 10 or 17) was added a solution of thionyl phenol (18). Prepared from (5-chlorobenzofuran-2-yl)(4-((tetrahydro-
chloride (1.6 m.eq.) in dry CH3CN (2 mL/mmol of carbinol). The 2H-pyran-2-yl)oxy)phenyl)methanol (17) (0.47 g, 1.31 mmol). Purifi­
mixture was stirred at 0 ◦ C for 1 h then K2CO3 (1 m.eq.) was added cation by gradient column chromatography afforded the product (18) at
followed by a solution of the carbinol (4, 10 or 17) (1 m.eq.) in dry 70% EtOAc in petroleum ether (v/v) as a white solid. Yield: 0.19 g
CH3CN (3 mL/mmol of carbinol) and the reaction stirred at room tem­ (45%); m.p. 240–242 ◦ C; Rf = 0.17 (petroleum ether - EtOAc 1:1 v/v). 1H
perature for 16 h. The reaction mixture was then filtered to remove any NMR (DMSO‑d6): δ 9.69 (bs, 1H, OH), 8.68 (s, 1H, CH-triazole), 8.06 (s,
insoluble substances. The filtrate was diluted with EtOAc (100 mL) and 1H, CH-triazole), 7.71 (d, J = 2.1 Hz, 1H, Ar), 7.59 (d, J = 8.7 Hz, 1H,
washed with H2O (3 × 50 mL). The organic layer was dried (MgSO4) and Ar), 7.34 (dd, J = 2.2, 8.7 Hz, 1H, Ar), 7.31 (d, J = 8.6 Hz, 2H, Ar), 7.20
concentrated under reduced pressure. (s, 1H, Ar), 6.81 (d, J = 8.6 Hz, 1H, Ar), 6.61 (t, J = 1.0 Hz, 1H, CH). 13C
1-((4-Chlorophenyl)(5-methoxybenzofuran-2-yl)methyl)-1H- NMR (DMSO‑d6): δ 158.36 (C), 156.80 (C), 153.49 (C), 152.37 (CH),
1,2,4-triazole (6). Prepared from (4-chlorophenyl)(5-methox 144.55 (CH), 129.91 (2 x CH), 129.67 (C), 127.96 (C), 126.68 (C),
ybenzofuran-2-yl)methanol (4b) (0.3 g, 1.04 mmol). Purification by 125.21 (CH), 121.46 (CH), 115.97 (2 x CH), 113.31 (CH), 106.75 (CH),
gradient column chromatography afforded the product (6) at 60% 60.15 (CH). HPLC (B): 100% at R.T. = 4.69 min.
EtOAc in petroleum ether (v/v) as a yellow oil. Yield: 0.26 g (63%); Rf =
0.1 (petroleum ether - EtOAc 3:1 v/v). 1H NMR (CDCl3): δ 8.16 (s, 1H, 4.3. General method for the preparation of the longer chain 1-(substituted
CH-triazole), 8.05 (s, 1H, CH-triazole), 7.41 (d, J = 8.6 Hz, 2H, Ar), 7.37 benzofuran-2-yl)(4-substituted phenyl)methyl)-1H-1,2,4-triazole (13, 14
(d, J = 8.9 Hz, 1H, Ar), 7.25 (d, J = 8.4 Hz, 2H, Ar), 7.01 (d, J = 2.6 Hz, and 19)
1H, Ar), 6.96 (dd, J = 2.6, 8.9 Hz, 1H, Ar), 6.81 (s, 1H, Ar), 6.55 (s, 1H,
CH), 3.84 (s, 3H, OCH3). 13C NMR (CDCl3): δ 156.38 (C), 152.66 (C), To a solution of the phenolic compound (11, 12 or 18) (1 m.eq.) in

9
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

dry CH3CN (13 mL/mmol of phenolic compound), K2CO3 (1.2 m.eq.) δ 156.55 (C), 156.15 (C), 152.39 (CH), 151.29 (C), 143.23 (CH), 135.23
was added and the mixture stirred for 1 h at 40 ◦ C then the alkyl/alkyne (C), 134.27 (C), 129.32 (2 x CH), 128.95 (2 x CH), 121.83 (CH), 121.36
bromide (1.2–6 m.eq.) was added and the reaction mixture stirred at (C), 113.35 (CH), 108.02 (CH), 97.52 (CH), 78.24 (C), 75.93 (CH), 61.53
room temperature for 16 h. The reaction mixture was concentrated (CH), 56.41 (CH2). HPLC (B): 100% at R.T. = 5.78 min.
under reduced pressure and the residue dissolved in EtOAc (100 mL), 1-((4-Fluorophenyl)(6-(pent-2-yn-1-yloxy)benzofuran-2-yl)
washed with H2O (3 × 50 mL), dried (MgSO4) and concentrated under methyl)-1H-1,2,4-triazole (13e, R = pent-2-yne R3 = F). Prepared
reduced pressure. from 2-((4-fluorophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-6-ol
1-((4-Chlorophenyl)(6-(pent-2-yn-1-yloxy)benzofuran-2-yl) (11b) (0.2 g, 0.64 mmol) and 1-bromopent-2-yne (0.13 mL, 1.29 mmol).
methyl)-1H-1,2,4-triazole (13a, R = pent-2-yne, R3 = Cl). Prepared Purification by gradient column chromatography afforded the product
from 2-((4-chlorophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-6-ol (13e) at 60% EtOAc in petroleum ether (v/v) as a yellow oil. Yield: 0.12
(11a) [23] (0.2 g, 0.61 mmol) and 1-bromopent-2-yne (0.07 mL, 0.67 g (50%); Rf = 0.43 (petroleum ether - EtOAc 1:1 v/v). 1H NMR
mmol). Purification by gradient column chromatography afforded the (DMSO‑d6): δ 8.72 (s, 1H, CH-triazole), 8.08 (s, 1H, CH-triazole), 7.54
product (13a) at 60% EtOAc in petroleum ether (v/v) as a yellow oil. (m, 3H, Ar), 7.32 (s, 1H, Ar), 7.30 (t, J = 8.9 Hz, 2H, Ar), 7.21 (d, J = 1.9
Yield: 0.18 g (75%); Rf = 0.55 (petroleum ether - EtOAc 1:1 v/v). 1H Hz, 1H, Ar), 6.91 (dd, J = 2.2, 8.5 Hz, 1H, Ar), 6.56 (t, J = 0.9 Hz, 1H,
NMR (CDCl3): δ 8.24 (s, 1H, CH-triazole), 7.98 (s, 1H, CH-triazole), 7.34 CH), 4.77 (t, J = 2.1 Hz, 2H, CH2), 2.24 (qt, J = 2.2, 7.5 Hz, 2H, CH2),
(m, 3H, Ar), 7.17 (d, J = 8.4 Hz, 2H, Ar), 7.00 (d, J = 2.0 Hz, 1H, Ar), 1.06 (t, J = 7.5 Hz, 3H, CH3). 13C NMR (DMSO‑d6): δ 163.57 (C), 161.62
6.87 (dd, J = 2.1, 8.5 Hz, 1H, Ar), 6.75 (s, 1H, Ar), 6.45 (s, 1H, CH), 4.62 (C), 156.52 (C), 155.89 (C), 153.26 (C), 152.49 (CH), 144.71 (CH),
(t, J = 2.1 Hz, 2H, CH2), 2.16 (qt, J = 2.1, 7.5 Hz, 2H, CH2), 1.07 (t, J = 133.15 (C), 130.67 (CH), 130.60 (CH), 122.20 (CH), 121.40 (C), 116.26
7.5 Hz, 3H, CH3). 13C NMR (CDCl3): δ 156.98 (C), 156.26 (C), 151.81 (CH), 116.08 (CH), 113.22 (CH), 107.33 (CH), 97.65 (CH), 89.58 (C),
(CH), 150.87 (C), 143.19 (CH), 135.27 (C), 134.17 (C), 129.33 (2 x CH), 75.27 (C), 59.79 (CH), 56.93 (CH2), 13.98 (CH3), 12.14 (CH2). HPLC
128.08 (2 x CH), 121.50 (CH), 120.98 (C), 113.44 (CH), 108.20 (CH), (B): 100% at R.T. = 6.70 min.
97.37 (CH), 89.98 (C), 73.87 (C), 61.70 (CH), 57.11 (CH2), 13.59 (CH3), 1-((4-Fluorophenyl)(6-(but-2-yn-1-yloxy)benzofuran-2-yl)
12.51 (CH2). HPLC (A): 99% at R.T. = 4.89 min. methyl)-1H-1,2,4-triazole (13f, R = but-2-yne R3 = F). Prepared from
1-((6-(But-2-yn-1-yloxy)benzofuran-2-yl)(4-chlorophenyl) 2-((4-fluorophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-6-ol
methyl)-1H-1,2,4-triazole (13b, R = but-2-yne, R3 = Cl). Prepared (11b) (0.17 g, 0.57 mmol) and 1-bromobut-2-yne (0.1 mL, 1.12 mmol).
from 2-((4-chlorophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-6-ol Purification by gradient column chromatography the product (13f) at
(11a) [23] (0.22 g, 0.67 mmol) and 1-bromobut-2-yne (0.12 mL, 1.35 60% EtOAc in petroleum ether (v/v), which was further purified by
mmol). Purification by gradient column chromatography afforded the preparative TLC to afford the product as a yellow oil. Yield: 0.07 g
product (13b) at 60% EtOAc in petroleum ether (v/v) as a yellow oil. (36%); Rf = 0.50 (petroleum ether - EtOAc 1:1 v/v). 1H NMR (CDCl3): δ
Yield: 0.19 g (76%); Rf = 0.3 (petroleum ether - EtOAc 1:1 v/v). 1H NMR 8.16 (s, 1H, CH-triazole), 8.05 (s, 1H, CH-triazole), 7.44 (d, J = 8.6 Hz,
(CDCl3): δ 8.15 (s, 1H, CH-triazole), 8.05 (s, 1H, CH-triazole), 7.44 (d, J 1H, Ar), 7.32 (m, 2H, Ar), 7.15 (m, 3H, Ar), 6.96 (dd, J = 2.2, 8.5 Hz, 1H,
= 8.6 Hz, 1H, Ar), 7.41 (d, J = 8.5 Hz, 2H, Ar), 7.25 (d, J = 8.2 Hz, 2H, Ar), 6.81 (s, 1H, CH), 6.52 (t, J = 0.9 Hz, 1H, Ar), 4.70 (q, J = 2.3 Hz, 2H,
Ar), 7.10 (d, J = 2.1 Hz, 1H, Ar), 6.96 (dd, J = 2.2, 8.6 Hz, 1H, Ar), 6.81 CH2), 1.89 (t, J = 2.3 Hz, 3H, CH3). 13C NMR (CDCl3): δ 163.99 (d, 1JC,F
(s, 1H, Ar), 6.54 (t, J = 0.8 Hz, 1H, CH), 4.70 (q, J = 2.2 Hz, 2H, CH2), = 247.51 Hz, C), 156.87 (C), 156.23 (C), 152.29 (CH), 151.36 (C),
1.89 (t, J = 2.3 Hz, 3H, CH3). 13C NMR (CDCl3): δ 156.91 (C), 156.25 143.03 (CH), 131.64 (d, 4JC,F = 3.36 Hz, C), 129.53 (d, 3JC,F = 8.40 Hz, 2
(C), 152.38 (CH), 151.05 (C), 143.26 (CH), 135.20 (C), 134.33 (C), x CH), 121.69 (CH), 121.05 (C), 116.22 (d, 2JC,F = 21.75 Hz, 2 x CH),
129.31 (2 x CH), 128.94 (2 x CH), 121.72 (CH), 121.01 (C), 113.39 (CH), 113.34 (CH), 107.92 (CH), 97.30 (CH), 84.15 (C), 73.72 (C), 61.56 (CH),
108.08 (CH), 97.29 (CH), 84.17 (C), 73.71 (C), 61.56 (CH), 57.01 (CH2), 57.01 (CH2), 3.74 (CH3). HPLC (A): 99.0% at R.T. = 4.73 min.
3.74 (CH3). HPLC (B): 100% at R.T. = 7.35 min. 4-((6-(Pent-2-yn-1-yloxy)benzofuran-2-yl)(1H-1,2,4-triazol-1-
1-((4-Chlorophenyl)(6-ethoxybenzofuran-2-yl)methyl)-1H- yl)methyl)benzonitrile (13g, R = pent-2-yne R3 = CN). Prepared from
1,2,4-triazole (13c, R = ethyl R3 = Cl). Prepared from 2-((4-chlor­ 4-((6-hydroxybenzofuran-2-yl)(1H-1,2,4-triazol-1-yl)methyl)benzoni­
ophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-6-ol (11a) [23] (0.2 trile (11c) [23] (0.1 g, 0.31 mmol) and 1-bromopent-2-yne (0.07 mL,
g, 0.61 mmol) and ethylbromide (0.25 mL, 3.4 mmol). Purification by 0.68 mmol). Purification by gradient column chromatography afforded
gradient column chromatography afforded the product (13c) at 60% the product (13g) at 70% EtOAc in petroleum ether (v/v) as a yellow oil.
EtOAc in petroleum ether (v/v) as a yellow oil. Yield: 0.147 g (67%); Rf Yield: 0.02 g (17%); Rf = 0.27 (petroleum ether - EtOAc 1:1 v/v). 1H
= 0.4 (petroleum ether - EtOAc 1:1 v/v). 1H NMR (CDCl3): δ 8.05 (s, 1H, NMR (CDCl3): δ 8.11 (s, 1H, CH-triazole), 7.98 (s, 1H, CH-triazole), 7.64
CH-triazole), 7.94 (s, 1H, CH-triazole), 7.32 (m, 3H, Ar), 7.14 (d, J = 8.2 (d, J = 8.5 Hz, 2H, Ar), 7.36 (d, J = 8.6 Hz, 1H, Ar), 7.30 (d, J = 8.2 Hz,
Hz, 2H, Ar), 6.89 (d, J = 2.1 Hz, 1H, Ar), 6.81 (dd, J = 2.2, 8.6 Hz, 1H, 2H, Ar), 7.02 (d, J = 1.99 Hz, 1H, Ar), 6.89 (dd, J = 2.1, 8.5 Hz, 1H, Ar),
Ar), 6.70 (s, 1H, Ar), 6.42 (t, J = 0.9 Hz, 1H, CH), 3.99 (q, J = 6.9 Hz, 2H, 6.78 (s, 1H, Ar), 6.49 (app. s, 1H, CH), 4.63 (t, J = 2.1 Hz, 2H, CH2), 2.19
CH2), 1.37 (t, J = 6.9 Hz, 3H, CH3). 13C NMR (CDCl3): δ 158.10 (C), (qt, J = 2.1, 7.5 Hz, 2H, CH2), 1.08 (t, J = 7.5 Hz, 3H, CH3). 13C NMR
156.50 (C), 152.35 (CH), 150.68 (C), 143.22 (CH), 135.15 (C), 134.41 (CDCl3): δ 157.18 (C), 156.34 (C), 152.63 (CH), 149.86 (C), 140.95 (C),
(C), 129.28 (2 x CH), 128.92 (2 x CH), 121.67 (CH), 120.36 (C), 113.24 132.82 (2 x CH), 128.23 (2 x CH), 121.83 (CH), 120.78 (C), 118.06 (C),
(CH), 108.12 (CH), 96.56 (CH), 64.02 (CH2), 61.57 (CH), 14.77 (CH3). 113.65 (CH), 113.16 (C), 108.63 (CH), 97.35 (CH), 90.04 (C), 73.79 (C),
HPLC (B): 100% at R.T. = 7.56 min. 61.60 (CH), 57.11 (CH2), 13.58 (CH3), 12.51 (CH2). HPLC (B): 97.4% at
1-((4-Chlorophenyl)(6-(prop-2-yn-1-yloxy)benzofuran-2-yl) R.T. = 5.45 min.
methyl)-1H-1,2,4-triazole (13d, R = prop-2-yne R3 = Cl). Prepared 4-((6-(But-2-yn-1-yloxy)benzofuran-2-yl)(1H-1,2,4-triazol-1-yl)
from 2-((4-chlorophenyl)(1H-1,2,4-triazol-1-yl)methyl)benzofuran-6-ol methyl)benzonitrile (13h, R = but-2-yne R3 = CN). Prepared from 4-
(11a) [23] (0.2 g, 0.61 mmol) and propargylbromide (80% wt in ((6-hydroxybenzofuran-2-yl)(1H-1,2,4-triazol-1-yl)methyl)benzonitrile
toluene) (0.2 mL, 1.84 mmol). Purification by gradient column chro­ (11c) [23] (0.2 g, 0.63 mmol) and 1-bromobut-2-yne (0.11 mL, 1.26
matography afforded the product (13d) at 60% EtOAc in petroleum mmol). Purification by gradient column chromatography at 70% EtOAc
ether (v/v) as a yellow oil. Yield: 0.2 g (90%); Rf = 0.37 (petroleum ether in petroleum ether (v/v), followed by preparative TLC afforded the
- EtOAc 1:1 v/v). 1H NMR (CDCl3): δ 8.06 (s, 1H, CH-triazole), 7.95 (s, product (13h) as a yellow oil. Yield: 0.017 g (7%); Rf = 0.27 (Petroleum
1H, CH-triazole), 7.36 (d, J = 8.6 Hz, 1H, Ar), 7.32 (d, J = 8.5 Hz, 2H, ether/ethyl acetate 1:1). 1H NMR (CDCl3): δ 8.11 (s, 1H, CH-triazole),
Ar), 7.15 (d, J = 8.4 Hz, 2H, Ar), 7.02 (d, J = 2.1 Hz, 1H, Ar), 6.88 (dd, J 7.98 (s, 1H, CH-triazole), 7.64 (d, J = 8.5 Hz, 2H, Ar), 7.36 (d, J =
= 2.3, 8.5 Hz, 1H, Ar), 6.71 (s, 1H, Ar), 6.45 (t, J = 0.9 Hz, 1H, CH), 4.65 8.6 Hz, 1H, Ar), 7.30 (d, J = 8.2 Hz, 2H, Ar), 7.02 (d, J = 2.2 Hz, 1H, Ar),
(d, J = 2.4 Hz, 2H, CH2), 2.47 (t, J = 2.4 Hz, 1H, CH). 13C NMR (CDCl3): 6.89 (dd, J = 2.3, 8.6 Hz, 1H, Ar), 6.78 (s, 1H, Ar), 6.49 (s, 1H, CH), 4.62

10
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

(q, J = 2.3 Hz, 2H, CH2), 1.80 (t, J = 2.3 Hz, 3H, CH3). 13C NMR (CDCl3): ns and 400 ns simulations) were performed using OPLS_2005 forcefield
δ 176.61 (C), 157.13 (C), 156.33 (C), 149.85 (C), 140.94 (C), 132.82 (2 x at 300 K and constant pressure (1 bar).
CH), 128.23 (2 x CH), 121.86 (CH), 120.66 (C), 118.06 (C), 113.62 (CH),
113.15 (C), 108.62 (CH), 97.28 (CH), 84.27 (C), 73.57 (C), 61.60 (CH), 4.5. Cell culture
57.01 (CH2), 3.74 (CH3). HPLC (A): 98.6% at R.T. = 4.60 min.
1-((4-Chlorophenyl)(5-(pent-2-yn-1-yloxy)benzofuran-2-yl) JEG-3 cells were purchased from ATCC and grown in Eagle’s Mini­
methyl)-1H-1,2,4-triazole (14). Prepared from 2-((4-chlorophenyl) mal Essential Medium (EMEM) supplemented with 10% fetal calf serum
(1H-1,2,4-triazol-1-yl)methyl)benzofuran-5-ol (12) (0.2 g, 0.61 mmol) (FCS). MCF-10A cells were a gift from Prof. Christopher McCabe (Uni­
and 1-bromopent-2-yne (0.07 mL, 0.67 mmol). Purification by gradient versity of Birmingham) and were grown in Dulbecco’s Modified Eagle
column chromatography afforded the product (14) at 60% EtOAc in Medium (DMEM) supplemented with 20 ng/ml epidermal growth factor
petroleum ether (v/v) as a yellow oil. Yield: 0.12 g (50%); Rf = 0.42 (EGF), 100 ng/ml cholera toxin, 0.01 mg/ml insulin, 500 ng/ml hy­
(petroleum ether - EtOAc 1:1 v/v). 1H NMR (CDCl3): δ 8.70 (s, 1H, CH- drocortisone, and 5% horse serum (Sigma). MDA-MB-231 cells were
triazole), 8.08 (s, 1H, CH-triazole), 7.34 (d, J = 8.3 Hz, 2H, Ar), 7.28 (d, purchased from ATCC and grown in Roswell park Memorial Institute
J = 9.1 Hz, 1H, Ar), 7.25 (d, J = 8.3 Hz, 2H, Ar), 7.02 (d, J = 2.6 Hz, 1H, medium (RPMI1690) supplemented with 10% FCS. All cells were
Ar), 6.94 (dd, J = 2.6, 9.0 Hz, 1H, Ar), 6.86 (s, 1H, Ar), 6.51 (s, 1H, CH), cultured at 37 ◦ C under 5% CO2 in a humidified incubator.
4.60 (t, J = 2.2 Hz, 2H, CH2), 2.18 (qt, J = 2.1, 7.5 Hz, 2H, CH2), 1.07 (t,
J = 7.5 Hz, 3H, CH3). 13C NMR (CDCl3): δ 156.62 (C), 151.90 (C), 150.72 4.6. Aromatase activity assay
(C), 135.66 (C), 133.38 (C), 129.46 (2 x CH), 129.28 (2 x CH), 127.74
(C), 115.37 (CH), 112.11 (CH), 108.66 (CH), 105.64 (CH), 89.69 (C), Aromatase activity was assayed using a modified titrated water assay
74.16 (C), 62.16 (CH), 57.38 (CH2), 13.62 (CH3), 12.50 (CH2). HPLC as previously reported [28]. JEG-3 cells were grown in 1 mL EMEM to
(A): 97.7% at R.T. = 4.89 min. approximately 80% confluence in six-well cell culture plates. Andros­
1-((5-Chlorobenzofuran-2-yl)(4-(pent-2-yn-1-yloxy)phenyl) t-4-ene-3,17-dione[1β-3H] was dissolved in serum-free cell culture me­
methyl)-1H-1,2,4-triazole (19). Prepared from 2 4-((5-chlor­ dium and added into each well. Aromatase activity was measured in the
obenzofuran-2-yl)(1H-1,2,4-triazol-1-yl)methyl)phenol (18) (0.15 g, absence and presence of inhibitor (0.001pM–100pM). After a 1 h incu­
0.69 mmol) and 1-bromopent-2-yne (0.07 mL, 0.67 mmol). Purification bation at 37 ◦ C followed by a 5-min incubation on ice, 500 μL of culture
by gradient column chromatography afforded the product (19) at 60% medium was taken from each well. Medium was vortexed with 2%
EtOAc in petroleum ether (v/v) as a yellow oil. Yield: 0.08 g (44%); Rf = dextran-treated charcoal (Sigma-Aldrich) in PBS and centrifuged at
0.42 (petroleum ether - EtOAc 1:1 v/v). 1H NMR (CDCl3): δ 8.21 (s, 1H, 4000 rpm. The supernatant containing the product, [3H] H2O, was
CH-triazole), 7.99 (s, 1H, CH-triazole), 7.43 (d, J = 1.9 Hz, 1H, Ar), 7.30 quantified by scintillation counting. Cell protein concentrations were
(d, J = 8.7 Hz, 1H, Ar), 7.22 (m, 3H, Ar), 6.95 (d, J = 8.8 Hz, 2H, Ar), determined using Pierce BCA assay kit (Thermo Fisher Scientific). Aro­
6.73 (s, 1H, Ar), 6.45 (s, 1H, CH), 4.61 (t, J = 2.1 Hz, 2H, CH2), 2.18 (qt, matase activity results were determined as a concentration of product
J = 2.1, 7.5 Hz, 2H, CH2), 1.07 (t, J = 7.5 Hz, 3H, CH3). 13C NMR formed per mg of protein per hour (pmol/mg/h). Results were shown as
(CDCl3): δ 158.71 (C), 154.33 (C), 153.66 (C), 151.39 (CH), 142.96 a % change in activity compared to control. Each data point was
(CH), 129.24 (2 x CH), 128.99 (C), 128.84 (C), 127.26 (C), 125.50 (CH), measured in triplicates and the error in the IC50 calculations represented
121.08 (CH), 115.57 (2 x CH), 112.56 (CH), 107.07 (CH), 90.06 (C), as 95% confidence interval.
73.71 (C), 61.91 (CH), 56.63 (CH2), 13.57 (CH3), 12.49 (CH2). HPLC:
(B) 96.1% at R.T. = 9.17 min. 4.7. BrdU-based cell proliferation assay to assess drug toxicity

4.4. Computational studies MCF-10A and MDA-MB-231 cells were plated onto 96-well microti­
ter tissue culture plates in RPMI1690 medium at a density of 8 × 103
4.4.1. Docking cells/well (for MCF-10A) or 5 × 103 cells/well (for MDA-MB-231.
The crystal structure of human placental aromatase (CYP19A1) Groups were treated with either DMSO alone (at no greater than
refined at 2.75 Å (PDB 3S79) [21], was downloaded from the protein 0.01%) as a vehicle control, or at a dose of 1 μM of inhibitor or doxo­
data bank (https://www.rcsb.org). Missing hydrogens were added, and rubicin control, for 48 h. Effects of drug treatment on cell growth were
the charge and geometry of the iron atom were adjusted as previously detected using the BrdU cell proliferation assay (Roche) according to the
described [31]. Using the site finder tool in molecular operating envi­ manufacturer’s recommendations. The BrdU colorimetric immunoassay
ronment (MOE) 2015-10 software [28], the active site was chosen to is a quantitative cell proliferation assay based on the measurement of
contain the main amino acid residues and the haem molecule. The BrdU incorporation during DNA synthesis. After treatments 20 μL/well
amino acids constituting the wall of the active site contained Arg115, of BrdU were added to each well, followed by an incubation of 2 h at 37
Ile133, Phe134, Phe221, Trp224, Ile305, Ala306, Asp309, Thr310, ◦
C. The cells were subsequently fixed, and the DNA denatured. Anti-
Val370, Leu372, Val373, Met374, Leu477, Ser478. The 3D structures of BrdU-peroxidase immune complexes were detected by substrate reac­
the ligands (R- and S-enantiomers) were generated using MOE builder, tion and quantified in an ELISA reader at 370 nm.
energy minimised and saved in a dataset ready for docking studies. The
complexes for molecular dynamics (MD) studies were prepared by Notes
docking the compounds using MOE.
The authors declare no competing financial interest.
4.4.2. Molecular dynamics
Molecular dynamics simulations were performed using Schrödinger Declaration of competing interest
2020–1 Desmond programme [29,30] as previously described [32].
Briefly, using the pdb files containing the selected docking poses, the The authors declare that they have no known competing financial
structures were optimised with protein preparation wizard. The volume interests or personal relationships that could have appeared to influence
of space in which the simulation takes place, the global cell, is built up the work reported in this paper.
by regular 3D simulation boxes. The orthorhombic water box allowed
for a 10 Å buffer region between protein atoms and box sides. Over­ Acknowledgment
lapping water molecules were deleted, and the systems were neutralised
with Na + ions and salt concentration 0.15 M. Molecular dynamics (150 We thank the Egyptian Ministry of Higher Education-Mission Sector

11
A.G. Eissa et al. European Journal of Medicinal Chemistry 240 (2022) 114569

and Zagazig University, Egypt for funding this research in Cardiff Uni­ [16] W. Jessie, L. Zeruesenay, D.A. Flockhart, Tamoxifen metabolites as active
inhibitors of aromatase in the treatment of breast cancer, Breast Cancer Res. Treat.
versity through a PhD scholarship to Ahmed Eissa. Molecular dynamics
131 (2) (2012) 473–481, https://doi.org/10.1007/s10549-011-1428-z.
simulations were undertaken using the supercomputing facilities at [17] C. Egbuta, J. Lo, D. Ghosh, Mechanism of inhibition of estrogen biosynthesis by
Cardiff University operated by Advanced Research Computing at Cardiff azole fungicides, 155, 2014, pp. 4622–4628, https://doi.org/10.1210/en.2014-
(ARCCA) on behalf of the Cardiff Supercomputing Facility and the HPC 1561. December.
[18] A. Spinello, S. Martini, F. Berti, M. Pennati, M. Pavlin, J. Sgrignani, G. Grazioso,
Wales and Supercomputing Wales (SCW) projects. We acknowledge G. Colombo, N. Zaffaroni, A. Magistrato, European journal of medicinal Chemistry
support of the latter, which is part-funded by the European Regional rational design of allosteric modulators of the aromatase enzyme : an
Development Fund (ERDF) via the Welsh Government. unprecedented therapeutic strategy to fight breast cancer, Eur. J. Med. Chem. 168
(2019) 253–262, https://doi.org/10.1016/j.ejmech.2019.02.045.
[19] A. Magistrato, J. Sgrignani, R. Krause, A. Cavalli, Single or Multiple access
Appendix A. Supplementary data channels to the CYP450s active site? An answer from free energy simulations of the
human aromatase enzyme, J. Phys. Chem. Lett. 8 (9) (2017) 2036–2042, https://
doi.org/10.1021/acs.jpclett.7b00697.
Supplementary data to this article can be found online at https://doi. [20] J. Sgrignani, M. Bon, G. Colombo, A. Magistrato, Computational approaches
org/10.1016/j.ejmech.2022.114569. elucidate the allosteric mechanism of human aromatase inhibition: a novel possible
route to small-molecule regulation of CYP450s activities? J. Chem. Inf. Model. 54
(10) (2014) 2856–2868, https://doi.org/10.1021/ci500425y.
References [21] D. Ghosh, C. Egbuta, J. Lo, Testosterone complex and non-steroidal ligands of
human aromatase, J. Steroid Biochem. Mol. Biol. 181 (2018) 11–19, https://doi.
[1] T. Vargo-Gogola, J.M. Rosen, Modelling breast cancer: one size does not fit all, Nat. org/10.1016/j.jsbmb.2018.02.009.
Rev. Cancer 7 (9) (2007) 659–672, https://doi.org/10.1038/nrc2193. [22] D. Ghosh, J. Lo, D. Morton, D. Valette, J. Xi, J. Griswold, S. Hubbell, C. Egbuta,
[2] K. Polyak, Heterogeneity in breast cancer, J. Clin. Invest. 121 (10) (2011) W. Jiang, J. An, H.M.L. Davies, Novel aromatase inhibitors by structure-guided
3786–3788, https://doi.org/10.1172/JCI60534.3786. design, J. Med. Chem. 55 (19) (2012) 8464–8476, https://doi.org/10.1021/
[3] J. Ferlay, I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, M. Rebelo, D.M. Parkin, jm300930n.
D. Forman, F. Bray, Cancer incidence and mortality worldwide: sources, methods [23] M.R. Saberi, T.K. Vinh, S.W. Yee, B.J.N. Griffiths, P.J. Evans, C. Simons, Potent
and major patterns in GLOBOCAN 2012, Int. J. Cancer 136 (5) (2015) E359–E386, CYP19 (aromatase) 1-(Benzofuran-2-yl)(Phenylmethyl)Pyridine, -imidazole, and
https://doi.org/10.1002/ijc.29210. -triazole inhibitors: synthesis and biological evaluation, J. Med. Chem. 49 (3)
[4] Cancer Research UK, Breast cancer statistics. http://www.wcrf.org/int/cancer (2006) 1016–1022, https://doi.org/10.1021/jm0508282.
-facts-figures/data-specific-cancers/breast-cancer-statistics. (Accessed 19 [24] S. Mahboobi, A. Sellmer, H. Höcher, C. Garhammer, H. Pongratz, T. Maier,
December 2021). T. Ciossek, T. Beckers, 2-Aroylindoles and 2-aroylbenzofurans with N-
[5] F. Bray, J. Ferlay, I. Soerjomataram, R.L. Siegel, L.A. Torre, A. Jemal, Global cancer hydroxyacrylamide substructures as a novel series of rationally designed histone
statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 deacetylase inhibitors, J. Med. Chem. 50 (18) (2007) 4405–4418, https://doi.org/
cancers in 185 countries, CA A Cancer J. Clin. 68 (6) (2018) 394–424, https://doi. 10.1021/jm0703136.
org/10.3322/caac.21492. [25] B. Schmidt, M. Riemer, U. Schilde, Chroman-4-ones via microwave-promoted
[6] J. Ferlay, M. Ervik, F. Lam, M. Colombet, L. Mery, M. Piñeros, A. Znaor, domino claisen rearrangement–oxa-Michael addition: synthesis of tabchromones A
I. Soerjomataram, F. Bray, Global cancer observatory: cancer Today.Lyon, France: and B, Synlett 25 (20) (2014) 2943–2946, 2014.
international agency for research on cancer, Available from: https://gco.iarc. [26] P.A. Foster, S.K. Chander, S.P. Newman, L.W.L. Woo, O.B. Sutcliffe, C. Bubert,
fr/today. (Accessed 11 June 2019). https://gco.iarc.fr/today (accessed December D. Zhou, S. Chen, B.V.L. Potter, M.J. Reed, A. Purohit, A new therapeutic strategy
19, 2021). against hormone-dependent breast cancer: the preclinical development of a dual
[7] H. Kang, X. Xiao, C. Huang, Y. Yuan, D. Tang, X. Dai, X. Zeng, Potent aromatase aromatase and sulfatase inhibitor, Clin. Cancer Res. 14 (20) (2008) 6469–6477,
inhibitors and molecular mechanism of inhibitory action, Eur. J. Med. Chem. 143 https://doi.org/10.1158/1078-0432.CCR-08-1027.
(2018) 426–437, https://doi.org/10.1016/j.ejmech.2017.11.057. [27] Cyprotex Discovery Ltd, Cytochrome P450 (CYP) inhibition assay (IC50). https
[8] G.T. Beatson, On the treatment of inoperable cases of carcinoma of the mamma: ://www.cyprotex.com/admepk/in-vitro-metabolism/cytochrome-p450-inhibition
suggestions for a new method of treatment with illustrative cases, Lancet II (1896) /, 29th November 2021.
104–107, https://doi.org/10.1016/S0140-6736(01)72384-7. [28] Molecular Operating Environment (MOE) CCI, 2016, 1010 Sherbooke St. West,
[9] J. Caciolla, A. Spinello, S. Martini, A. Bisi, N. Zaffaroni, S. Gobbi, A. Magistrato, Suite #910, Montreal, QC, Canada,H3 A 2R7.
Targeting orthosteric and allosteric pockets of aromatase via dual-mode novel [29] Schrödinger Release 2020–1: Desmond Molecular Dynamics System, D. E. Shaw
azole inhibitors, ACS Med. Chem. Lett. 11 (5) (2020) 732–739, https://doi.org/ Research, New York, NY, 2020. Maestro-Desmond Interoperability Tools,
10.1021/acsmedchemlett.9b00591. Schrödinger, New York, NY, 2020, https://www.schrodinger.com/products/des
[10] F.B. De Abreu, G.N. Schwartz, W.A. Wells, G.J. Tsongalis, Personalized therapy for mond.
breast cancer, Clin. Genet. 86 (1) (2014) 62–67, https://doi.org/10.1111/ [30] K.J. Bowers, E. Chow, H. Xu, R.O. Dror, M.P. Eastwood, B.A. Gregersen, J.
cge.12381. L. Klepeis, I. Kolossvary, M.A. Moraes, F.D. Sacerdoti, K.K. Salmon, Y. Shan, D.
[11] D. Ghosh, J. Lo, C. Egbuta, Recent progress in the discovery of next generation E. Shaw, Scalable algorithms for molecular dynamics simulations on commodity
inhibitors of aromatase from the structure-function perspective, J. Med. Chem. 59 clusters, in: Proceedings of the ACM/IEEE Conference on Supercomputing (SC06),
(11) (2016) 5131–5148, https://doi.org/10.1021/acs.jmedchem.5b01281. Tampa, Florida, 2006, pp. 11–17. November, https://dl.acm.org/doi/proceedi
[12] S. Hiscox, E.L. Davies, P. Barrett-Lee, Aromatase inhibitors in breast cancer, ngs/10.1145/1188455, 29th November 2021.
Maturitas 63 (4) (2009) 275–279, https://doi.org/10.1016/j. [31] F.A.S. Binjubair, J.E. Parker, A.G. Warrilow, K. Puri, P.J. Braidley, E. Tatar, S.
maturitas.2009.05.008. L. Kelly, D. Kelly, C. Simons, Small molecule inhibitors targeting 14α-demethylase
[13] J.M. Dixon, Endocrine resistance in breast cancer, New J. Sci. 2014 (10) (2014) (CYP51): synthesis, molecular modeling and evaluation against Candida albicans,
1–27, https://doi.org/10.1155/2014/390618. ChemMedChem 15 (2020) 1294–1309, https://doi.org/10.1002/
[14] D. Ghosh, J. Griswold, M. Erman, W. Pangborn, Structural basis for androgen cmdc.202000250.
specificity and oestrogen synthesis in human aromatase, Nature 457 (7226) (2009) [32] N.A. Noureldin, J. Richards, H. Kothayer, M.M. Baraka, S.M. Eladl, W. Wootton,
219–223, https://doi.org/10.1038/nature07614. C. Simons, Design, computational studies, synthesis and in vitro antimicrobial
[15] A. Mojaddami, A. Sakhteman, M. Fereidoonnezhad, Z. Faghih, A. Najdian, evaluation of benzimidazole based thio-oxadiazole and thio-thiadiazole analogues,
S. Khabnadideh, H. Sadeghpour, Z. Rezaei, Binding mode of triazole derivatives as BMC Chem. 15 (1) (2021) 58, https://doi.org/10.1186/s13065-021-00785-8.
aromatase inhibitors based on docking, protein ligand interaction fingerprinting,
and molecular dynamics simulation studies, Res. Pharm. Sci. 12 (1) (2017) 21–30,
https://doi.org/10.4103/1735-5362.199043.

12