Lead Optimization of Benzoxepin-Type Selective Estrogen Receptor (ER) Modulators and Downregulators with Subtype-Specific ERα and ERβ Activity

Article abstract of DOI:10.1021/acs.jmedchem.6b01917

Estrogen receptor α (ERα) is an important target for the design of drugs such as tamoxifen (2a) and fulvestrant (5). Three series of ER-ligands based on the benzoxepin scaffold structure were synthesized: series I containing an acrylic acid, series II with an acrylamide, and series III with a saturated carboxylic acid substituent. These compounds were shown to be high affinity ligands for the ER with nanomolar IC₅₀ binding values. Series I acrylic acid ligands were generally ERα selective. In particular, compound 13e featuring a phenylpenta-2,4-dienoic acid substituent was shown to be antiproliferative and downregulated ERα and ERβ expression in MCF-7 breast cancer cells. Interestingly, from series III, the phenoxybutyric acid derivative compound 22 was not antiproliferative and selectively downregulated ERβ. A docking study of the benzoxepin ligands was undertaken. Compound 13e is a promising lead for development as a clinically relevant SERD, while compound 22 will be a useful experimental probe for helping to elucidate the role of ERβ in cancer cells.

Full text of DOI:10.1021/acs.jmedchem.6b01917

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
Article
Lead Optimization of Benzoxepin-Type Selective Estrogen Receptor (ER)
Modulators and Downregulators with Subtype-Specific ER# and ER# Activity
Niamh Mary O'Boyle, Irene Barrett, Lisa M Greene, Miriam Carr, Darren Fayne, Brendan
Twamley, Andrew J.S. Knox, Niall O. Keely, Daniela M. Zisterer, and Mary Jane Meegan
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 20 Apr 2017
Downloaded from http://pubs.acs.org on April 20, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Lead Optimization of BenzoxepinꢀType Selective Estrogen Receptor (ER) Modulators and
Downregulators with SubtypeꢀSpecific ERα and ERβ Activity
1
,2
1
2
1
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Niamh M. O’Boyle, * Irene Barrett, Lisa M. Greene, Miriam Carr, Darren Fayne,
3
2
1
2
Brendan Twamley, Andrew J.S. Knox, Niall O. Keely, Daniela M. Zisterer and Mary J.
1
Meegan
1
School of Pharmacy and Pharmaceutical Sciences, Trinity Biomedical Sciences Institute, Trinity
College Dublin, 152 ꢀ 160 Pearse St, Dublin 2, D02 R590, Ireland.
2
School of Biochemistry and Immunology, Trinity Biomedical Sciences Institute, Trinity College
Dublin, 152 ꢀ 160 Pearse St, Dublin 2, D02 R590, Ireland.
3
School of Chemistry, Trinity Biomedical Sciences Institute, Trinity College Dublin, 152 ꢀ 160
Pearse St, Dublin 2, D02 R590, Ireland.
Abstract
Estrogen receptor α (ERα) is an important target for the design of drugs such as tamoxifen (2a)
and fulvestrant (5). Three series of ERꢀligands based on the benzoxepin scaffold structure were
synthesised – series I containing an acrylic acid, series II with an acrylamide and series III with a
saturated carboxylic acid substituent. These compounds were shown to be high affinity ligands
for the ER with nanomolar IC binding values. Series I acrylic acid ligands were generally
5
0
ERα selective. In particular, compound 13e featuring a phenylpentaꢀ2,4ꢀdienoic acid substituent
was shown to be antiproliferative and downregulated ERα and ERβ expression in MCFꢀ7 breast
cancer cells. Interestingly, from series III, the phenoxybutyric acid derivative compound 22 was
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
not antiproliferative and selectively downregulated ERβ. A docking study of the benzoxepin
ligands was undertaken. Compound 13e is a promising lead for development as a clinically
relevant SERD, whilst compound 22 will be a useful experimental probe for helping to elucidate
the role of ERβ in cancer cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Key words:
Antiproliferative activity, Benzoxepin, Breast Cancer, ER binding, Estrogen receptor ligand,
Selective Estrogen Receptor Downregulator, SERD
2
ACS Paragon Plus Environment

Page 3 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
The two nuclear estrogen receptors (ERα and ERβ) mediate the biological effects of the estrogen
hormones and ERα is an attractive therapeutic target for diseases including breast cancer and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
osteoporosis. Estrogens including estradiol (1, Figure 1) are known to have tissue selective
effects, and there is considerable interest in the therapeutic use of selective estrogen receptor
2
3
modulators (SERMs). A number of SERMs are currently in clinical use, including tamoxifen
(2a) for treatment of hormoneꢀdependent breast cancer, and raloxifene (3a), lasofoxidine and
4
basidoxifene for the prevention of osteoporosis (Figure 1). The clinical successes of 2a and 3a
has provided the driving force to discover new, multifunctional SERMs. There is ongoing
debate about the role of ERβ in cancer. It is generally thought that expression of ERβ has
5
, 6
antiproliferative effects in breast cancer cells. In prostate cancer, its role is still unclear and
7
, 8
there is some evidence that certain isoforms of ERβ are oncogenic. ERβ expression has been
reported have a potentially protective effect in normal cells on ERα promoted
9
hyperproliferation. There is much research being undertaken to fully elucidate the effects of
ERβ in cancers.
The selective ER downregulator (SERD) 5 is in clinical use for the treatment of postmenopausal
women with ERꢀpositive, locally advanced or metastatic breast cancer for disease relapse on or
after adjuvant antiꢀestrogen therapy, or disease progression on therapy with an antiꢀestrogen
(Figure 1). The acrylic acid GW5638 (4a) and its presumed metabolite GW7604 (4b, Figure 1)
have also been identified as SERDs mechanistically different to 2a and 3a, with full agonist
activity in bone and in the cardiovascular system, and antagonistic activity in human endometrial
10, 11,
adenocarcinoma Ishikawa cells and in the rat uterus with minimal residual agonist activity.
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
12
Breast cancers resistant to 2a are not crossꢀresistant to 4a indicating that this type of SERD
has potential as a therapeutic agent. The SERD action of 4b causes a decrease in cellular ERα
1
3
levels.
Related compounds showed low stimulation of uterine cell proliferation with good
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
14
14
15
ERα and ERβ binding affinities.
Acrylic acidꢀsubstituted quinoline, naphthalene,
1
6
17
18
19
tetrahydroisoquinoline, bicyclo[3.3.1]nonane, coumarin,
benzopyranobenzoxepanes and
1
5, 20, 21
benzosuberone scaffolds have also been investigated as SERMs.
The 1Hꢀpyrido[3,4ꢀ
22
23
b]indolꢀ1ꢀylphenylacrylic acid AZD9496 (4c), coumarin, and the indazole ERα modulator
2
4
GDCꢀ0810 (4d) were recently reported as potent and orally bioavailable SERDs and
antagonists (Figure 1).
We have previously identified novel SERMs [e.g. compound 3b (Figure 1)] which demonstrated
2
5
potential as ER binding ligands, and now we report the development of this chemical template
for nonsteroidal SERDs where the ER degradation was optimised. The conformationally
restricted benzoxepin template has been elaborated to incorporate substitution with acrylic acids,
and related structural modifications. An overview of the three distinct series of compounds
described in this study is provided in Figure 2. The evaluation of antiproliferative activity and
relative binding affinity of these ligands for ERα and ERβ together with their stimulatory effect
on uterine tissue is examined. A molecular modelling study was investigated to rationalise the
binding selectivity of these ligands for the ERα and ERβ receptors.
Chemistry
The benzoxepins synthesised in this study are arranged in three different structural classes
(Series I, II and III; Figure 2) to investigate the response of the ER ligandꢀbinding domain to
4
ACS Paragon Plus Environment

Page 5 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
targeted structural alterations (Schemes 1ꢀ3). The first group (Series I, compounds 11aꢀ11d and
related alkenes 13aꢀc and 13e, Scheme 1) are ringꢀfused analogues of 4a/4b, with variation in
substitution at the Cꢀ8 position of the benzoxepin or benzothiepin scaffold structure. The
fluorine substituent at Cꢀ8 of the benzoxepin and benzothiepin structure would be expected to
increase the lipophilicity of these compounds and also block expected metabolic inactivation,
contributing to a longer plasma halfꢀlife. It has been reported that the inclusion of a fluorine
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
6
contributes to the SERM activity of ER ligands such as oxachrysenol and we have observed a
25
similar effect for compound 5.
The second group of compounds investigated (Series II, 14aꢀ
1
4i, Scheme 2) contain amideꢀmodified derivatives of the core acrylic acid compound 11c.
Compounds 19a, 19b and 22 (Series III, Scheme 3) are distinguished by the inclusion of the 4ꢀ
oxyacetic and 4ꢀoxybutyric acid substituent in place of the acrylic acid in Ring B of 11c.
The synthetic route that produced the required products 11aꢀ11d (series I) most efficiently is
shown in Scheme 1. The benzoxepinꢀ4ꢀones 6aꢀd were treated with trifluoromethanesulfonic
anhydride to afford the intermediate triflates which were then coupled with the arylboronic ester
to afford the aldehydes 7aꢀd in good yield (steps i and ii). Vinyl bromination of the alkenes 7aꢀ
d with pyridinium tribromide gave the bromides 8aꢀd (Scheme 1, step iii). Bromides 8aꢀc were
then coupled with 4ꢀhydroxyboronic acid in a second Suzuki reaction to afford the phenolic
substituted benzoxepins 9aꢀc (Scheme 1, step iv). A WittigꢀHorner reaction of 9aꢀc with
(ethoxycarbonyl methylene)triphenylphosphorane, followed by saponification of the esters 10aꢀc
afforded the required acids 11aꢀc in good yield (Scheme 1, steps v and vi). The benzothiepin
compound 11d was obtained in an alternative reaction sequence from the bromide 8d. Initial
WittigꢀHorner reaction with (ethoxycarbonylmethylene)triphenylphosphorane followed by
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
hydrolysis of the ester in situ afforded the acrylic acid 12. Suzuki coupling of the acid 12 with
4
ꢀhydroxyphenylboronic acid gave the required product 11d. The related alkene and α,βꢀ
unsaturated ketone products 13aꢀc were also synthesised by similar Wittig type reactions from
aꢀ9c (Scheme 1, step vii). The extended unsaturated acid compound 13e was obtained from
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
9
aldehyde 9c by treatment with (E)ꢀethyl 4ꢀ(diethoxyphosphoryl)butꢀ2ꢀenoate (prepared from
ethyl 4ꢀbromocrotonate and triethylphosphite) followed by saponification of the ester 13d
(Scheme 1, steps vii and viii).
Series II consisting of eight α,βꢀunsaturated amides 14aꢀh was obtained from the unsaturated
acid 11c by coupling with various amines in the presence of HOBt, using EDCI as the coupling
agent (Scheme 2, step i). Surprisingly, the coupling of 11c with aniline proved to be difficult.
An alternative method to obtain the desired product 14i required the initial preparation of the
phosphonate 15 (Scheme 2).
2ꢀBromoꢀNꢀphenylacetamide (prepared from aniline and
acetylbromide) was treated with triphenylphosphite to afford the corresponding phosphonate
2
7
1
5, which was reacted with aldehyde 9c to afford the required product 14i in 83% yield
Scheme 2, step ii). An alternative reaction of 9c with the phosphonium bromide prepared from
ꢀbromoꢀNꢀphenylacetamide and triphenylphosphine was unsuccessful.
(
2
Related compounds 19a and 19b (series III) containing an 8ꢀoxyacetic acid side chain were
2
5
obtained by Oꢀalkylation of the vinyl bromides 16a and 16b (prepared as previously reported)
with ethyl bromoacetate to afford esters 17a and 17b (Scheme 3, step i). Subsequent Suzuki
coupling with 4ꢀhydroxyphenylboronic acid followed by hydrolysis of esters 18a and 18b
afforded the desired acids 19a and 19b (Scheme 3, steps ii and iii). A similar alkylation reaction
6
ACS Paragon Plus Environment

Page 7 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of 16a with ethyl bromobutyrate afforded the ester product 20, which was subsequently arylated
with 4ꢀhydroxyphenylboronic acid to yield 21 (Scheme 3, steps i and ii). Hydrolysis of ester 21
gave the product 22 containing the required oxybutyric acid side chain (Scheme 3, step iii).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Acrylic acid 4b, required for comparative biochemical studies, was obtained in a novel threeꢀstep
route by initial McMurray coupling of ketone 23 and propiophenone to afford the iodoꢀ
substituted triphenylethylene 24 as the major product (Scheme 4, step i), which could be
separately converted into either the acrylic acid 25 (94%) or the acrylate ester 26 (93%) by Heckꢀ
type reactions with acrylic acid and ethyl acrylate respectively in the presence of palladium
acetate (Scheme 4, steps ii and iii respectively). Deprotection of 25 with boron tribromide
directly afforded 4b (Scheme 4, step iv; 28% overall yield) as the major Z isomer was obtained
(E/Z mixture: 1:3.5), while demethylation of 26 followed by in situ hydrolysis of intermediate 27
also yielded acrylic acid 4b (22% overall yield) (Scheme 4, steps iv and v). Previous synthetic
routes to this compound and analogues have relied on FriedelꢀCrafts acylation, Grignard or
10, 13, 28
Suzuki coupling reactions to generate the triphenylethylene structure,
followed by a Wittig
or HornerꢀEmmons reaction for introduction of the acrylate side chain [15% overall yield
FriedelꢀCrafts, Grignard and Wittig for synthesis of 4b)] and 48% overall yield (Suzuki
(
coupling followed by HornerꢀEmmons reaction for synthesis of 4a)].
XꢀRay Crystallography
An XꢀRay crystallography study of two of the benzoxepins, ester 18a and acid 22, was
undertaken to confirm the structural assignments and also to explore potentially important
structural features for potent ER activity. The 7ꢀmembered ring displays a puckered
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
conformation in both structures (Figure 3). This results in a molecular arrangement in which the
three aromatic rings attached to the 7ꢀmembered ring are not coplanar. Calculation of the ring
plane and torsional angles between the aromatic rings of these compounds further demonstrates
that these rings are arranged outꢀofꢀplane with respect to each other (Table 1). For example, the
ring angle between the planes of ring A and ring B of compound 22 is 64.3°, and the torsional
angle is 42°. Bond lengths between C8 and C18 (compound 18a) and C8 and C19 (compound
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2) of 1.345 Å and 1.350 Å, respectively, indicate the position of the double bond in the 7ꢀ
membered ring.
Biochemical Results: Antiproliferative activity in MCFꢀ7 breast cancer cells. The
antiproliferative activity of the benzoxepin compounds was firstly evaluated using the ERꢀ
expressing (ERꢀdependent) MCFꢀ7 human breast cancer cell line. Compound 2a (IC = 4.1
5
0
ꢀM) and 4b (IC = 1.3 ꢁM) were used as positive controls and the IC values obtained are in
5
0
50
29, 10, 13, 30
agreement with previously reported values for these compounds (Table 2).
In the initial series of acrylic acids 11aꢀ11d the presence of a fluorine at Cꢀ8 in benzoxepin
compound 11c, together with the sulfurꢀcontaining ring in compound 11d, resulted in a marked
improvement in antiproliferative activity (IC values of 0.26 and 0.095 ꢁM respectively, Table
5
0
2) when compared with nonꢀfluorinated compound 11a (IC = 21 ꢁM). The α,βꢀunsaturated
5
0
ketones 13a and 13c also retained subꢀmicromolar activity (IC50 = 0.89 and 0.97 ꢁM
respectively). Compound 13e with an extended pentaꢀ2,4ꢀdienoic acid substituent in Ring B,
retains moderate antiproliferative activity (IC = 1.6 μM).
5
0
8
ACS Paragon Plus Environment

Page 9 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction of the amide modification (series II) resulted in a decrease in the antiproliferative
activity of the fluorinated products 14aꢀi when compared with acrylic acid compound 11c.
Piperidineꢀsubstituted amides 14e and 14g proved to be the most active, together with 14h (IC50
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
=
0.48, 1.05 and 0.70 ꢁM respectively). Oxyacetic analogues of 2a had been reported to act as
31, 32
estrogen antagonists in MCFꢀ7 cells.
In series III, compounds 19a and 19b, containing a 4ꢀ
oxyacetic acid substituent in place of the acrylic acid in 11b and 11c, along with compound 22,
did not have antiproliferative activity at concentrations up to 20 ꢁM (as was observed for the
32
oxyacetic analogue of 2a ).
Cytotoxicity was evaluated by use of the standard LDH assay to establish that the observed
antiproliferative effects were attributable to cytostasis rather than cellular necrosis (Table 2).
The majority of the compounds demonstrated minimal cytotoxicity (<5% at 10 ꢀM
concentration), considerably below that obtained for 2a (13.4%). Compounds 13e and 22 were
also assessed for toxicity using the nonꢀtumorigenic MCFꢀ10a mammary epithelial cell line.
Compound 22 did not have any effect on the viability of MCFꢀ10a cells at concentrations of 1
and 10 ꢀM over 24 and 48 h periods (Figure 4). No effects were observed for compound 13e at 1
and 2 μM; however, MCFꢀ10a cell viability was reduced at higher concentrations of 5 and 10
ꢀ
M, with reduction of 50% at 5 ꢀM over 24 h. As the IC value of 13e in MCFꢀ7 cells is 1.6
5
0
ꢁM, there is a therapeutic window available for this compound at which it may cause
antiproliferative effects in breast tumour cells without significant toxicity to nonꢀtumorigenic
cells.
Biochemical Results: ER Binding Studies. ERꢀbinding studies were carried out with both
ERα and ERβ to confirm receptor involvement in the observed antiproliferative effect. A
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
fluorescence polarisation procedure was employed for this competitive binding assay which
measures the displacement of fluoresceinꢀlabelled estradiol (fluoromone) from the human
recombinant full length ERα and ERβ. All ERꢀbinding values are expressed in nM. The relative
binding affinity (RBA) of ER ligands is often reported. Compound 1 is typically used as the
reference ligand and is taken as the 100 % binding value. Using the reported reference IC50
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
34
values obtained for 1 in ERα (5.7 nM) and ERβ (5.6 nM) , the RBAs of the selected
conjugates were calculated (Table 2). Values greater than 100 % indicate a greater affinity for
the ER than 1; values less than 100 % indicate a diminished affinity for the ER.
All compounds (with the exception of 13b) displayed potent ERα and ERβ binding activities,
with the majority of the more potent compounds showing selectivity for ERβ. Introduction of
the 8ꢀfluoro substituent in series I, as in (11c), and also the benzothiepin ring scaffold (11d) gave
increased ER binding activity for both ERα and ERβ. As an example of the series, compound
1
1d exhibited potent binding to both ERα and ERβ with IC50 = 4.1 nM (ERα) and 3.1 nM
ERβ). Compound 13e, containing the extended pentaꢀ2,4ꢀdienoic acid substituent in Ring B,
was found to display potent ERꢀbinding activity with IC = 71.6 nM (ERα) and 0.55 nM (ERβ),
(
5
0
equivalent to 129ꢀfold ERβ selectivity. Compound 4a and related amides have previously been
10
reported as SERMs with antagonist activity in rat uterus and acting as agonists in bone.
In our
acrylamide series II (compounds 14aꢀi) compound 14i demonstrated the most effective binding
activity [IC = 11.7 nM (ERα) and 0.94 nM (ERβ)], with 11ꢀfold ERβ selectivity. Acrylamide
5
0
1
4b, possessing moderate antiproliferative activity, demonstrated good ERꢀbinding activity (IC50
67 nM for ERα and 2.4 nM for ERβ, with 27ꢀfold ERβ selectivity).
=
10
ACS Paragon Plus Environment

Page 11 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Triarylethylene oxyalkanoic acid modifications of 2a have been reported as bone selective
3
1
estrogen mimetics. Of our compounds in series III, 19a and 19b had high IC binding values
5
0
indicative of poor affinity for both ERα and ERβ, perhaps explaining their poor antiproliferative
activity. The related compound 22, containing the 4ꢀoxybutyric acid substituent, demonstrated
extremely interesting ERꢀbinding properties with IC50 = 147 nM (ERα) and 1.23 nM (ERβ),
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
which is 117ꢀfold selectivity for ERβ, despite its lack of antiproliferative potency. This binding
value for 22 is an improvement on reported values for the corresponding oxybutyric acid
3
2
analogue of 2a.
In order to assist us in the rationalisation of the α/β selectivity observed for the ER binding
results obtained for these benzoxepin compounds, the ER binding effects of the core structure 9c
was determined as IC = 407 nM for ERα and 395 nM for ERβ. This indicates that ER binding
5
0
is related to the presence of 4ꢀhydroxyphenyl at Cꢀ4 or Cꢀ5 of the 4,5ꢀdiarylꢀ8ꢀfluorobenzoxepin
scaffold. ER selectivity is conferred by the specific type of acidic side chain substituent present
at Cꢀ4 of Ring B. Small modifications to this side chain lead to dramatic differences in ERꢀ
binding affinity, subtype selectivity and antiproliferative potency.
Biochemical Results: Estrogenic Stimulation. The estrogenic stimulation and antagonistic
properties of a number of compounds were measured in an in vitro assay using stimulation of
35
alkaline phosphatase (AlkP) in Ishikawa human endometrial adenocarcinoma cells (Table 3).
Compounds 11aꢀ11d and 13a from series I were investigated alongside a representative example
from series II (14a) and series III (22). Antiestrogenic (IC = 0.0016 ꢁM) and estrogenic activity
5
0
14
(
4 % stimulation at 1 ꢁM) have been reported for 4b in Ishikawa cells. Compound 11d
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
possessed an improved potency over 2a as an ER antagonist (IC = 0.18 ꢁM) while compounds
5
0
1
1a, 11c, 13a and 22 also displayed good antiestrogenic activity, with IC values in the range
50
0.43 ꢀ 1.39 ꢀM. In the acrylamide series, compound 14a was found to be extremely potent as an
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
14
estrogen antagonist with an IC value of 0.0098 ꢁM, comparable to the value reported for 4b.
5
0
This compound does not demonstrate significant antiproliferative activity in the estrogen
dependent MCFꢀ7 cells (IC = 13 ꢁM) which is 10ꢀfold less potent than 4b (IC = 1.3 ꢁM); ER
5
0
50
binding is also less potent than observed for 4b. This may be attributed to the differences in
interaction of 14a with the ER on binding due to replacement of carboxylic acid of 4b, 11c and
11d with the more lipophilic N,Nꢀdiethylamide substituent. The estrogenic stimulatory effect is
determined by measuring AlkP stimulation in the absence of 1 (Table 3). Compounds 11c, 11d
and 22, which were among the most potent ERꢀbinding compounds, demonstrated an absence of
or very low stimulatory effects (0 %, 1.6 % and 3.2 % at concentrations of 1 ꢀM, 10 ꢀM and 1
1
4
ꢀ
M respectively), when compared with 4b (4%) and 2a (10%). The most potent antagonist
compound, acrylamide 14a (evaluated for stimulatory activity at a 100ꢀfold higher concentration
of 100 ꢁM, as 1 ꢁM had no effect), together with the ketone 13a (1 ꢁM), and acrylic acid 11a (1
ꢁ
M), showed relatively low stimulatory values of 10 %, 9.6 % and 4.6 % respectively.
Selection of the optimal structural features for antiestrogenic activity for benzoxepin and
benzothiepin ring scaffolds, without adverse estrogenic effects on tissues such as the uterus, is
possible by reference to the results of these AlkP assays.
Biochemical Results: Effect of compounds 13e and 22 on the expression levels of ERα and
ERβ in MCFꢀ7 cells
12
ACS Paragon Plus Environment

Page 13 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The expression of ERα and ERβ in MCFꢀ7 cells was examined by Western blotting (Figure 5).
Compounds 13e (series I) and 22 (series III) were chosen for further biochemical evaluation;
compound 22 showed an interesting profile of activity due to its lack of antiproliferative effect
whilst having an extremely selective effect on ERβ. Compound 13e has potent antiproliferative
activity in MCFꢀ7 cells and good potency against both ERα and ERβ (Table 2). The pure
estrogen antagonist 5, a known SERD, downregulates the expression of ERα and was used as a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
positive control.
Compound 5 causes rapid proteasomal degradation of ERα via
ubiquitinylation, resulting in shutdown of the estrogenꢀsignaling process and thus inducing
3
6
proliferation arrest and apoptosis of estrogenꢀdependent breast cancer cells.
SERMs such as 4ꢀhydroxytamoxifen (2b, Figure 1) bind to ERα as antagonists or partial
agonists depending on the target tissue. MCFꢀ7 breast cancer cells were treated with
In contrast,
compounds 13e and 22 (10 ꢀM), and after 24 h whole cell lysates were prepared and analysed by
SDSꢀPAGE and Western blotting for expression levels of ERα or ERβ. The known SERD 5
reduced ERα protein levels, with little or no effect on ERβ (Figure 5). Compound 13e, which
possessed good antiproliferative activity, was found to downregulate both ERα and ERβ (Figure
5
). Compound 22 selectively downregulated ERβ in MCFꢀ7 cells, with little effect on the
expression of ERα. This result is consistent with the ERꢀbinding assay, in which compound 22
was ERβ selective.
Compound 22 is the first reported ERβꢀselective SERD. It does not have antiproliferative effects
in either MCFꢀ7 or MCFꢀ10a cells, despite possessing antiestrogenic activity (Table 2 and Figure
4
). There is debate about the role of ERβ in cancer; in prostate cancer, there is some evidence
7
, 8
that certain isoforms of ERβ are oncogenic whilst generally thought that expression of ERβ
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5, 6
has antiproliferative effects in breast cancer cells. Due to its unique combination of cellular
effects, compound 22 is a useful tool for investigation of the role of ERβ in cancer cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Molecular Modelling of Benzoxepins in ERα and ERβ
A number of structurally related benzoxepins were discovered to have diverse effects on ERα
and ERβ, in particular compounds 13e and 22. Hence, a molecular modelling study of 11c, 11d,
1
3e, 14b, 14i and 22 was carried out to investigate their potential interactions with ERα and ERβ.
The two ER subtypes differ significantly in size: 595 amino acids in ERα compared to 485
amino acids (ERβ). The conservation of amino acid sequence in the ligand binding sites of ERα
and ERβ is only 59%, with the most notable differences being replacement of Met412 and
Leu384 in ERα with Ile and Met in ERβ. The 3ERT Xꢀray structure of hERα coꢀcrystallised
37
with 2b was downloaded from the PDB website. For ERβ the 1NDE Xꢀray structure coꢀ
38
crystallised with a triazine modulator was used. After validating the docking protocol and
determining receiver operating characteristic (ROC) values of 0.896 and 0.819 for the ERα and
ERβ haystack docking respectively, we undertook a more inꢀdepth binding analysis on
compounds 13e and 22 as they demonstrated the optimal ERα:ERβ binding ratio in favour of
ERβ. In addition to 1125 inactive compounds, the ERα and ERβ haystacks contained 39 and 32
known active compounds respectively. Amino acid numbering corresponds to
AA(#ERα)/(#ERβ) unless specified in the text. All of our compounds had topꢀranked binding
poses for both ERα and ERβ, excluding 22 in ERβ, placed the ring A fluorine atom at the same
location as the hydroxyl group of 2b. The comparative ranking of the most potent compounds
in ERα and ERβ for the docking analysis demonstrates a high degree of correlation with the
experimental binding affinity, particularly for ERα (Table 4). All the new compounds described
14
ACS Paragon Plus Environment

Page 15 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in this study ranked in the top 5% of the ordered hit list and are comparable to the known active
compounds.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In ERβ, the acid 22 has a 180 degree flipped orientation compared to other compounds in this
study, in that the phenolic hydroxyl group mimics the position adopted by 2b (Figure 7). The
fluorineꢀcontaining ring no longer clashed with Leu525 of ERα, as this sidechain has rotated
1
80° and is directed outside the LBP, but is ideally positioned to accept a HB from His475 of
ERβ. The ortho hydrogen of the ether appended phenyl ring B is located adjacent to the coꢀ
crystallised water molecule thereby enabling a HBD interaction. The oxygen atom of the
benzoxepin ring is adjacent to the Met421 of ERα which protrudes into the pocket and would
lead to electrostatic repulsion but in ERβ this residue is replaced by the Ile373 and moves to the
side of the pocket to form a hydrophobic layer below the ligand’s fluorinated phenyl group. In
ERα, 22 overlays well on the core structure of 2b except that the ring A fluorophenyl fluorine
maps to the hydroxyl group of 2b (Figure 7). The carboxylic acid side chain is not oriented
towards the Asp351 group but this may result from only generating 50 conformers of each ligand
as the sideꢀchain has many degrees of freedom. This series of LBP amino acid positional
changes between ERα and ERβ all favour binding of 22 to ERβ over ERα which is
experimentally reflected in the IC50 ER binding values and in the effects on ER expression
demonstrated in the western blot in MCFꢀ7 cells.
The 180 degree rotated orientation of a compound’s binding pose was previously highlighted by
3
9
our group and is reminiscent of earlier studies on 3a. In an influential coꢀcrystal structure of 3a
with ERα (1ERR) the phenolic group of the benzothiophene is interacting with Glu353 and
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
40
Arg394. However, a later 2002 Xꢀray structure of the aroylbenzothiophene core of 3a with no
4
1
sidechain (1GWQ) places the A ring in this position and the phenolic moiety interacts with
Glu353 and Arg394 which is analogous to the interactions made by 2b (3ERT). This illustrates
the possible symmetry of binding mode of the core scaffold which can rotate 180 degrees
depending on the nature of the appended sideꢀchain. While the ability of fluorine to act as a HBA
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
2
is still debated there is evidence that it is possible, although to a lesser extent than oxygen.
Crystallographic evidence in the case of ER binding to fluorinated A ring ligands is not
available. However, a very recent publication detailing the Xꢀray structure of 4,4ꢀ
dichlorodiphenyltrichloroethane (DDT) in ER (5KRA) places a different halogen, chlorine, in
4
3
position to partly mimic the Ring A phenolic groups’ interactions with Glu353 and Arg394. In
addition to the capacity to accept hydrogen bonds, chlorine can also interact with Glu353
4
4
through a halogen bond. This data adds confidence to our postulated binding modes wherein
the fluorine atom can occupy the traditional “Ring A” hydroxyl group position and engage in
HBA or dipole interactions with Arg394 and a bridging water molecule.
For both ERα and ERβ, 13e coꢀlocates the ring A fluorinated phenyl group adjacent to the 4ꢀ
hydroxy group of 2b and is positioned to accept HBs from Arg394/346 (Figure 6). The flipped
orientation as observed with 22 docked in ERβ is not possible in this case due to the linear
delocalised nature of the “antagonistic” sidechain which would clash with Asp351/303. For the
ERβ binding pose of 13e, the phenolic ring C clashes with Leu525 of ERα which is involved
with generating a lipophilic hole with Leu384 (Figure 6). In ERβ, Leu384 is replaced with
Met336 which locks a coꢀcrystallised water in place for a HBA interaction with the acyclic
secondary amine of the 1NDE coꢀcrystallised piperazinylꢀ1,3,5ꢀtriazine. Leu476 of ERβ can
16
ACS Paragon Plus Environment

Page 17 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
therefore relocate enabling the phenol group to occupy the pocket and act as a HBD with
Met295, in addition to potentially acting as a HBA with His475. In ERβ, the ortho hydrogen
atoms of both the phenol C ring and phenyl B ring are adjacent to the coꢀcrystallised water
molecule, bridging from Met336, thereby enabling HBD interactions.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
22
Consistent with previously reported XꢀRay studies of 4a and 4c in ERα, the acrylic acid
moiety of 13e is located adjacent to Asp351/303 in both the ERα and ERβ docked structures,
demonstrating an unusual acid−acid interaction which causes helix 12 to adopt an unexpected
conformation and has been proposed to be important for achieving a downregulatorꢀantagonist
profile for these acrylic acid ligands. With these studies in mind, the docked compounds were
not deprotonated. This observation is also consistent with the XꢀRay structure of 4a in the LBP
of ERα which has revealed that the compound occupies an orientation similar to 2b in the ligand
1
2
binding pocket. The unionized carboxylic acid groups of 4a and Asp351 form a hydrogen bond
observed in the crystal structure at pH 5.6. The acrylic acid substituent induces a conformation
of H12 in which it is displaced from the hydrophobic cleft, and stabilized by the formation of
hydrogen bonds from the carboxylic acid of 4a to the amide backbone NH groups of Leu536 and
Try537 at the N terminus of H12. This relocation causes a significant increase (27%) in the
exposed hydrophobic surface of H12 and induces destabilization and ERα protein degradation.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Discussion
In this work, we have identified novel structures based on the benzoxepin scaffold which have a
distinct profile of activity compared to 2a, implying that they may be useful in the treatment of
ERꢀpositive breast cancer tumours which may have developed resistance to conventional
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
5, 46
hormonal treatments.
Breast tumorigenesis can be associated with an increase in ERα
expression and a decrease in ERβ. Thus, it is hypothesized that ERβ plays a role as a tumor
suppressor in breast cancer. Breast tumours that become resistant to one antiestrogen class often
continue to maintain sensitivity to another class of antiestrogen. SERDs can be considered as a
further development in the design of antiestrogen type treatments for breast cancer, and agents
47
such as 5 have been shown to be effective against ERαꢀpositive, 2aꢀresistant breast cancers.
Compounds that induce ERα degradation could be used to increase the period of time that
4
8
patients can be successfully treated with antiestrogen therapies.
There is currently much interest in the discovery of novel molecular scaffolds with SERM or
SERD profile properties which could be suitable for development of new therapies for the
treatment of breast cancer, osteoporosis, and related hormoneꢀdependent conditions.
The
profound effect of the acrylic acid Hꢀbonding with the Hꢀ12 residues observed in the XꢀRay
structure of 4a prompted us to investigate the effect of a number of conformationallyꢀconstrained
benzoxepin templates on the ER affinity and antiproliferative effects in MCFꢀ7 ERꢀpositive
breast cancer cells. In the design of the compounds we have included functional groups such as
acrylic acids, acrylamides, methyl ketones, oxyacetic acids, oxybutyric acids and extended
pentadienoic acids, which can form HBs to Helix 12 Asp351, and also interact with the backbone
Helix12 amide groups. The recently reported benzopyranobenzoxepanes were identified as
18
ACS Paragon Plus Environment

Page 19 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potent SERMs for treatment of postmenopausal symptoms, behaving as antagonists in the uterus,
1
9
whilst exhibiting potent agonist activity in bone and plasma lipids. We have identified a
number of fluoroꢀsubstituted benzoxepin compounds containing an acrylic acid side chain which
demonstrate potent antiproliferative activity against the MCFꢀ7 human breast cancer cell line
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Table 2). The presence of the fluorine substituent at Cꢀ8 in Ring A (Figure 2) is expected to
block metabolism of these compounds, ensuring a longer plasma halfꢀlife. These ERꢀtargeting
benzoxepins are designed to circumvent the metabolic complications introduced by E, Z
isomerisation of the 2aꢀtype triarylethylene antiestrogen structures. These compounds were
demonstrated to be high affinity ligands for the ER with the majority of IC50 values in the
nanomolar range. The acrylic acid ligands in series I were generally ERα selective, while the
phenylpentaꢀ2,4ꢀdienoic acid derivative 13e, acrylamide 14i and the phenoxybutyric acid
derivative 22 demonstrated selectivity for ERβ of 11ꢀ, 117ꢀ and 129ꢀfold respectively, with low
Ishikawa cell stimulation. In series III, it was apparent that two–carbon homologation of the
oxyacetic side chain of 19a to produce compound 22 resulted in a notable improvement in the
ER affinity and reduced estrogenic stimulatory effect as demonstrated in the Ishikawa cell assay.
Compound 13e was shown to be a downregulator of ERα expression in MCFꢀ7 breast cancer
cells, while compound 22 was nonꢀantiproliferative and selectively downregulated expression of
ERβ. The compounds demonstrated low cytotoxicity in both the LDH assay and in the MCFꢀ
10a nonꢀtumorigenic cell line. The receptor selectivity for ERβ for the most potent compounds
1
3e and 22 was examined in docking studies and suggest a correlation between slight change in
binding site amino acids between the two isoforms and enhanced ERβ binding for this
benzoxepin ring template. The downregulation effect in ERα of 13e may be attributed to the
additional hydrogen bonding of the carboxyl group with the Leu536, Tyr537 (via a water
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
molecule) and the expected interaction with Asp 351.
The novel SERDs 13e and 22 based on
the highꢀaffinity benzoxepin ligand core structure with phenoxybutyric acid and pentaꢀ2,4ꢀ
dienoic acid substituents respectively resulted in differing profiles of downregulation of the ER,
and are potentially useful scaffolds for future development.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Molecular modelling studies confirm that inclusion of the benzoxepin scaffold in compounds 13e
and 22 facilitates a number of key interactions with the ER LBD residues. The low oral
bioavailability of 5 indicates the potential for the development of potent orally available
49
SERDs.
Further biochemical studies are necessary to determine the effects of these novel
analogues on estrogen response element (ERE) transcription and ERα stability in MCFꢀ7 cells,
and to determine the mechanistic differences between their activity and that of 2a. Molecular
modifications which facilitate varied interactions between the ligand and the receptor residues,
resulting in possible unfolding of the LBD and increasing its hydrophobicity are seen to result in
the discovery of new types of antiestrogens with potential clinical applications.
Conclusions
This study demonstrates that an alternative ligand core structure, i.e. the benzoxepin scaffold,
can be used in place of the triarylethylene core characteristic of 4a and 4b. This study extends
our current understanding of the pharmacophore requirements for SERD activity, and probes the
effect of the benzoxepin ligand core structure and the amido, phenylacrylic acid, phenoxyacetic
acid, phenoxybutyric acid and pentaꢀ2,4ꢀdienoic acid substituents on the activity of the ER.
Specific structural modifications which facilitate additional interactions between the ligand and
the ER have been shown to be key factors in the design of novel SERDs which have useful
clinical applications. Compound 22 shows significant potential for utility in probing the role of
ERβ in the development of cancer. Compound 13e, which features the phenylpentaꢀ2,4ꢀdienoic
20
ACS Paragon Plus Environment

Page 21 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acid substituent on the benzoxepin core, was antiproliferative and effectively downregulated
ERα in MCFꢀ7 breast cancer cells. These novel ligands can be used to probe the size, shape and
flexibility of the ERα and ERβ ligandꢀbinding domains and will be investigated further as
potential drugs.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Experimental Section
Materials and methods: Chemistry
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
All reagents were commercially available and were used without further purification unless
otherwise indicated. Anhydrous DCM was obtained by distillation from calcium hydride, and
anhydrous THF by distillation from benzophenoneꢀsodium, both in inert atmospheres
immediately before use. IR spectra were recorded as thin films on NaCl plates or as KBr discs
1
13
on a PerkinꢀElmer Paragon 100 FTꢀIR spectrometer. H and C NMR spectra were obtained on
o
1
13
a Bruker Avance DPX 400 instrument at 20 C, 400.13 MHz for H spectra, 100.61 MHz for C
19
spectra, or 376 MHz for F spectra, in either CDCl or CD OD (internal standard
3
3
tetramethylsilane). High resolution molecular ion determinations (HRMS) (mass measurement
accuracies of < ± 5 ppm) were obtained at the Mass Spectrometry Laboratory, Centre for
Synthesis and Chemical Biology (CSCB), University College Dublin, by Dr. Dilip Rai.
Additional mass measurements were obtained by Dr. Martin Feeney at the HRMS Laboratory in
the Department of Chemistry, Trinity College Dublin. TLC was performed using Merck Silica
gel 60 TLC aluminium sheets with fluorescent indicator visualizing with UV light at 254nm.
Flash chromatography was carried out using standard silica gel 60 (230ꢀ400 mesh) from Merck.
All products isolated were homogenous on TLC. Samples were tested for purity using reversed
phase HPLC (Waters Alliance system) at 254 nm using a Phenomenex column (4 µ micron, 250
x 4.60 mm) using a mobile phase of acetonitrile: water (0.1 % TFA) (70:30) delivered at a flow
rate of 1.0 mL/min. Final compounds had a purity of ≥ 95%. Compounds 6b, 16a and 16b were
2
5
synthesized and characterized as previously described by us.
22
ACS Paragon Plus Environment

Page 23 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
General Method I. Preparation of 6a, 6c, 6d. A mixture of appropriately substituted butyric
acid (27.8 mmol) and polyphosphoric acid (51 g) was heated at 70 °C for 4 h. The brown syrup
was poured into iceꢀwater (100 mL) and the aqueous layer extracted with ethyl acetate (4 × 200
mL). The combined organic layers were washed with 10% NaOH (100 mL), brine (100 mL),
dried over sodium sulfate and the solvent removed under reduced pressure. Compounds 6a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
51
52
(
waxy white solid, 35% yield), 6c (yellow oil, 31% yield) and 6d (yellow oil, 77% yield)
were prepared and characterized as previously reported.
General Method II: Synthesis of 7aꢀ7d. Trifluoromethanesulfonic anhydride (4.38 mmol) in
dichloromethane (20 mL) was added to a suspension of 6aꢀ6d (2.19 mmol) and sodium sulphate
(4.38 mmol) in dichloromethane (20 mL) at 0 °C. The suspension was stirred at rt for 18 h,
filtered and the filtrate washed with water (50 mL), brine (50 mL), dried over sodium sulfate and
the solvent removed under reduced pressure. The residue was dissolved in THF (10 mL), and 4ꢀ
formylphenylboronic acid (7.2 mmol) and Na CO (2M aq., 2.93 mL) were added. The mixture
2
3
was stirred under nitrogen for 10 min. Pd(PPh ) (138 mg) was added and the reaction refluxed
3
4
at 85 °C for 6 h. The solution was cooled to 20 °C and acidified with HCl (2M). The aqueous
layer was extracted with dichloromethane (3 × 30 mL) and the combined organic layers were
washed with water (30 mL), brine (30 mL), dried over sodium sulfate and the solvent removed
under reduced pressure.
General Method III: Preparation of 8aꢀ8d. Pyridinium tribromide (5.46 mmol, 90% grade)
was added portionwise to a solution of 7aꢀ7d (3.64 mmol) in dichloromethane (10 mL) at 0 °C.
The solution was warmed to rt and stirred for 18 h. NaHCO (10%, 20 mL) was added and the
3
aqueous layer was extracted with dichloromethane (3 × 50 mL). The combined organic layers
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
were washed with water (2 × 20 mL), and dried over sodium sulfate. The solvent was removed
under reduced pressure and the crude product purified by flash column chromatography.
General method IV: Preparation of 9aꢀ9c, 18aꢀ18b and 21. To a solution of 8aꢀ8c, 17a, 17b
or 20 (2.63 mmol) in THF (20 mL) was added 4ꢀhydroxyphenylboronic acid (3.94 mmol),
sodium carbonate (2M, 6.58 mL, 13.2 mmol), and Pd(PPh ) (0.05 g) and the solution refluxed at
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
4
9
0 °C for 22 h. The solution was cooled, water was added (20 mL) and the aqueous layer
extracted with dichloromethane (4 × 50 mL). The combined organic layers were washed with
brine (20 mL), dried over sodium sulfate and the solvent removed under reduced pressure.
General Method V. Preparation of 10aꢀ10c. A solution of 9aꢀ9c (2.05 mmol) and
(ethoxycarbonylmethylene)triphenylphosphorane (2.67 mmol) in dry CH Cl (20 mL) was
2
2
refluxed for 6 h.
The solution was cooled to rt and the organic layer diluted with
dichloromethane (20 mL), washed with water (30 mL) and brine (30 mL), dried over sodium
sulfate and the solvent removed under reduced pressure. The residue was purified by flash
column chromatography (silica gel, 5% ethyl acetate/95% hexane) to remove triphenylphosphine
oxide.
General Method VI: Preparation of 11aꢀ11c. To a solution of 10aꢀ10c (1.64 mmol) in
methanol (10 mL) was added NaOH (2M aq., 10 mL) and the mixture was refluxed for 3 h,
cooled to rt and then acidified with conc. HCl.
The aqueous layer was extracted with
dichloromethane (3 × 30 mL) and the combined organic layers were washed with water, brine,
dried over sodium sulfate and the solvent removed under reduced pressure. The crude product
was purified by flash column chromatography (silica gel, 5% ethyl acetate/hexane).
General Method VII: Preparation of 13aꢀ13d. A solution of the appropriate bromoketone or
bromoester (10 mmol) and triethylphosphite (11.4 mmol) was refluxed at 120ꢀ130 °C for 2 h.
24
ACS Paragon Plus Environment

Page 25 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The product was isolated as a colourless oil (100%) and used in the following reaction. To a
suspension of sodium hydride (1.5 mmol, 60% dispersion) in THF (10 mL) at 0 °C was added
the appropriate phosphorane or triphenylphosphonium salt (1.5 mmol) and the solution stirred
for 10 min. Compound 9c (1.0 mmol) was added to this mixture at 0 °C and the solution warmed
to room temperature and stirred overnight. Water (30 mL) was added and the aqueous layer
extracted with dichloromethane (2 × 30 mL). The combined organic layers were washed with
brine (30 mL), dried over sodium sulfate and the solvent removed under reduced pressure.
General Method VIII: Synthesis of 14aꢀ14h. To a solution of 11c (0.62 mmol) in CH Cl (30
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
mL) at 0 °C was added HOBt (1.10 mmol), EDCI (1.10 mmol) and triethylamine (1.36 mmol)
and the mixture was stirred for 10 min. The appropriate amine (1.10 mmol) was added and the
mixture stirred overnight at rt. Water (30 mL) was added and the aqueous layer extracted with
CH Cl (2 × 30 mL). The combined organic layers were washed with brine (30 mL), dried over
2
2
sodium sulfate and the solvent removed under reduced pressure.
General Method IX: Preparation of 13e, 19a, 19b, and 22. A solution of 13d, 18a, 18b or 21
(100 mg) was refluxed in ethanol (5 mL) with NaOH (aq., 1M, 2 mL) for 1 h. The solution was
cooled, acidified and the aqueous layer extracted with dichloromethane (2 × 30 mL). The
combined organic layers were washed with brine, dried over sodium sulfate and the solvent
removed under reduced pressure. The product was purified by flash column chromatography (1:1
ethyl acetate/hexane).
(E)ꢀ3ꢀ{4ꢀ[(E/Z)ꢀ1ꢀ(4ꢀHydroxyphenyl)ꢀ2ꢀphenylbutꢀ1ꢀenyl]phenyl}acrylic acid (4b)
Method 1: A solution of acid 25 (0.35 mmol, 1 eq.) was stirred in anhydrous DCM (5 mL) and
the reaction mixture was cooled to ꢀ78 °C. Boron tribromide solution (1.0 M, 1.40 mmol, 4 eq)
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
was added slowly to the reaction mixture. The reaction was stirred at ꢀ78 °C for 45 min then at
rt for a further three h. The reaction was quenched by addition of methanol (5 mL). The
mixture was evaporated to dryness in vacuo and the residue was purified via flash
2
8
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chromatography (DCM:MeOH) to afford 4b as a resin (60%).
Method 2: A solution of the
ester 26 (0.38 mmol) was stirred in anhydrous DCM (5 mL) and the reaction mixture was cooled
to ꢀ78 °C. Boron tribromide solution (1.0 M, 1.52 mmol, 4 eq.) was added slowly to the
solution. The reaction was stirred at ꢀ78 °C for 45 min then at rt for a further three h. The
reaction was quenched by addition of methanol (5 mL). The mixture was evaporated to dryness
in vacuo. The residual product (27) was dissolved in THF (10 mL) and 0.1 M NaOH solution
was added. The mixture was refluxed for 1 h., then diluted with ethyl acetate and the solution
was acidified with HCl (10 %). The aqueous phase was extracted with ethyl acetate (3 × 50
mL). The combined organic phases were dried over sodium sulfate and evaporated to dryness in
vacuo. The residue was purified via flash chromatography (DCM:MeOH) to afford 4b as a
ꢀ1
resin (50%, Z/E = 3.5/1). IR: νmax (KBr) cm : 3396, 3189, 2965, 1685, 1635, 1603, 1511, 1441.
1
H NMR (CDCl ): δ 10.84 (1H, bs), 7.56 ꢀ 7.60 (1H d, J = 16 Hz), 6.56 – 7.38 (13H, m), 6.42 –
3
1
3
6
.48 (1H, d, J = 16 Hz), 3.02 (1H, bs), 2.53 – 2.59 (2H, m), 0.96 – 1.00 (3H, m). C NMR
(
CDCl ): δ 167.7, 157.2, 146.7, 145.1, 143.4, 143.0, 139.1, 134.9, 132.7, 132.6, 132.0, 130.7,
3
1
30.4, 128.9, 128.7, 128.6, 128.0, 127.1, 126.9, 118.8, 118.3, 115.9, 115.2, 30.3, 29.6, 13.7.
+
HRMS (ESI): Found 369.1490 (M–H) , C H O requires 369.1491.
2
5
21
3
4
ꢀ(2,3ꢀDihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (7a) was prepared from 6a by general
1
method II to give a yellow oil, which was used without further purification (83%). H NMR
(
CDCl ): δ 10.03 (1H, s), 7.97 (2H, d, J = 8.04 Hz), 7.41 (2H, d, J = 8.04 Hz), 7.21 (1H, t, J =
3
26
ACS Paragon Plus Environment

Page 27 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7
.78 Hz), 7.09 (1H, d, J = 8.00 Hz), 6.91 (1H, t, J = 7.54 Hz), 6.67 (1H, d, J = 7.52 Hz), 5.28
(1H, s), 4.56 (2H, t, J = 5.76 Hz), 3.06 (2H, t, J = 5.78 Hz)
4ꢀ(8ꢀMethoxyꢀ2,3ꢀdihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (7b) was prepared from 6b by
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
general method II. The product was purified by flash column chromatography (silica gel, 5%
diethyl ether/hexane) to give the product as a yellow oil (40%) which was used without further
1
purification in the following reaction. H NMR (CDCl ): δ 10.03 (1H, s), 7.86 (2H, d, J = 8.00
3
Hz), 7.46 (2H, d, J = 8.00 Hz), 6.81 (1H, d, J = 8.52 Hz), 6.69 (1H, d, J = 2.48 Hz), 6.59 (1H,
d,d, J = 2.50 Hz, 6.04 Hz), 6.27 (1H, t, J = 6.02 Hz, CH), 4.50 (2H, t, J = 5.78 Hz, OCH ), 3.82
2
13
(3H, s, OCH ), 2.59 (2H, q, J = 6.01 Hz, OCH CH ); C NMR (CDCl ): δ 191.5, 159.6, 158.9,
3 2 2 3
149.1, 139.7, 134.7, 131.3, 129.3, 128.8, 128.5, 123.1, 109.1, 106.5, 76.0, 54.9, 30.3.
ꢀ(8ꢀFluoroꢀ2,3ꢀdihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (7c) was prepared from 6c by
4
general method II. The product was purified by flash column chromatography (silica, 2.5%
methanol/dichloromethane) to give the product as yellow oil (80%). IR νmax(KBr): 1698, 1601,
ꢀ1
1
1
151 cm ; H NMR (CDCl ): δ 10.05 (1H, s), 7.91 (2H, d, J = 8 Hz), 7.40 (2H, d, J = 8 Hz),
3
6
.82 (1H, d, J = 8 Hz), 6.65 (2H, d, J = 8 Hz), 4.60 (2H, t, J = 6.0 Hz), 3.10 (2H, t, J = 6.0 Hz);
C (101 MHz, CDCl ): δ 191.5, 162.0 (J = 250 Hz), 158.8 (J = 12 Hz), 148.4, 139.5, 134.8,
1
3
3
F
F
1
31.6 (J = 10 Hz), 129.6, 129.4, 128.7, 127.1, 110.2 (J = 21 Hz), 109.0 (J = 23 Hz), 107.4 (J
F F F F
1
9
+
=
23 Hz), 76.8, 29.8; F NMR (CDCl ): δ ꢀ112.7. HRMS (ESI) Found 269.0987(M+H) ;
3
C H O F requires 269.0978.
17
14
2
4
ꢀ(8ꢀFluoroꢀ2,3ꢀdihydrobenzo[b]thiepinꢀ5ꢀyl)benzaldehyde (7d) was prepared from 6d by
general method II. The product was purified by flash column chromatography (silica gel, 5%
diethyl ether/hexane) to give the product as a yellow solid (75%). Mp 140 °C; IR νmax(KBr):
ꢀ1
1
3
420.3, 1695.9, 1600.8, 1211.2 cm ; H NMR (CDCl ): δ 9.99 (1H, s), 7.80 (2H, d, J = 8.52
3
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Hz), 7.42 (1H, q (d,d), J = 2.52 Hz, 6.00 Hz), 7.36 (2H, d, J = 8.52 Hz), 7.02 – 6.98 (2H, m),
1
3
6
.69 (1H, t, J = 7.78 Hz), 3.49 (2H, t, J = 6.52 Hz), 2.35 (2H, q, J = 7.03 Hz); C NMR (CDCl ):
3
δ 191.4, 161.9, 147.2, 142.0, 139.8, 136.4, 134.9, 131.7, 131.6, 131.4, 129.4, 128.0, 121.3, 121.0,
19
+
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
15.2, 115.0; F NMR (CDCl ): δ ꢀ113.593; HRMS (ESI) Found 283.0584(MꢀH) ; C H FOS
3 17 12
requires 283.0593.
4
ꢀ(4ꢀBromoꢀ2,3ꢀdihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (8a) was prepared from 7a by
general method III. The crude product purified by flash column chromatography (silica gel, 5%
diethyl ether/hexane) to give 8a as a brown oil (74%) which was used without further
1
purification. H NMR (CDCl ): δ 10.01 (1H, s), 7.89 (2H, d, J = 8.0 Hz), 7.40 (2H, d, J = 8.0
3
Hz), 7.22ꢀ7.13 (2H, m), 6.90 (1H, d, J = 7.5 Hz), 6.66 (1H, d, J = 7.5 Hz), 4.54 (2H, t, J = 6.0
Hz), 3.05 (2H, t, J = 5.8 Hz).
4
ꢀ(4ꢀBromoꢀ8ꢀmethoxyꢀ2,3ꢀdihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (8b) was prepared
from 7b by general method III. The crude product was obtained as a dark brown solid (93%)
1
which was used without further purification; Mp 94 °C; H NMR (CDCl ): δ 10.06 (1H, s), 7.93
3
(2H, d, J = 8.0 Hz), 7.42 (2H, d, J = 8.0 Hz), 6.65 (1H, d, J = 2.5 Hz), 6.57 (1H, d, J = 9.0 Hz),
6
.50 (1H, dd, J = 2.5 Hz, 6.5 Hz), 4.57 (2H, t, J = 5.7 Hz), 3.79 (3H, s), 3.15 (2H, t, J = 5.7 Hz);
1
3
C NMR (CDCl ): δ 191.4, 159.8, 157.8, 148.7, 137.9, 134.8, 131.8, 130.2, 129.3, 123.3, 121.6,
3
1
09.3, 106.5, 75.3, 54.9, 41.7.
4
ꢀ(4ꢀBromoꢀ8ꢀfluoroꢀ2,3ꢀdihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (8c) was prepared from
7c by general method III. The crude product was purified by flash column chromatography
(silica, 5% diethyl ether/hexane) to give 8c as a white solid (91%). Mp 96 °C; IR νmax (KBr):
ꢀ1
1
1
698, 1601, 1151 cm ; H NMR (CDCl ): δ 10.05 (1H, s), 7.91 (2H, d, J = 8.5 Hz), 7.40 (2H, d,
3
J = 8.0 Hz), 6.82 (1H, m), 6.65 (2H, q, J = 3.2 Hz), 4.60 (2H, t, J = 6.0 Hz), 3.10 (2H, t, J = 6.0
28
ACS Paragon Plus Environment

Page 29 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
13
Hz); C NMR (CDCl ): δ 191.4, 162.1 (d, J = 251 Hz), 157.5 (d, J = 10 Hz), 148.1, 137.6,
3
F
F
19
1
34.9, 132.0, 129.8, 127.5 (J = 10 Hz) 122.9, 110.4 (J = 20 Hz), 109.2, 108.9, 76.2, 41.2; F
F F
+
NMR (CDCl ): δ ꢀ111.71. HRMS (ESI): Found; 347.0076 (M+H) ; C H O BrF requires
3
17 13
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
4
7
47.0083.
ꢀ(4ꢀBromoꢀ8ꢀfluoroꢀ2,3ꢀdihydrobenzo[b]thiepinꢀ5ꢀyl)benzaldehyde (8d) was prepared from
d by general method III. The crude product was purified by chromatography (silica gel, 10%
1
diethyl ether/hexane) to give 8d as a yellow oil (76%). H NMR (CDCl ): δ 10.02 (1H, s), 7.86
3
(
2H, d, J = 8.5 Hz), 7.42ꢀ7.38 (3H, m), 6.95ꢀ6.91 (1H, m), 6.84 (1H, q (d,d), J = 5.5 Hz, 3.0 Hz),
13
3
.49 (2H, t, J = 6.5 Hz), 2.88 (2H, t, J = 6.3 Hz) ppm; C NMR (CDCl ) δ: 191.3, 162.0, 147.2,
3
140.8, 140.7, 135.1, 135.0, 134.9, 131.7, 131.6, 130.0, 129.2, 124.8, 121.6, 121.3, 115.4, 115.2;
19
+
F (CDCl ): ꢀ112.599 ppm; HRMS (ESI): Found; 362.9877(M+H) ; C H OFSBr requires
3
17 13
3
62.9855.
4ꢀ(4ꢀ(4ꢀHydroxyphenyl)ꢀ2,3ꢀdihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (9a) was prepared
from 8a by general method IV. The crude product was purified by flash column chromatography
(
silica gel, 5% diethyl ether/hexane) to give 9a as a bright yellow powder (85%). Mp 164 °C;
which was used without further purification in the subsequent reaction. IR νmax(KBr): 3334,
ꢀ1 1
1
7
675, 1610, 1598, 1190, 1169 cm ; H NMR (CDCl ): δ 9.90 (1H, s), 7.67 (2H, d, J = 8.0 Hz),
3
.23ꢀ7.17 (4H, m), 7.03 (2H, d, J = 8.0 Hz), 6.88 (2H, d, J = 8.5 Hz), 6.82 (1H, d, J = 7.6 Hz),
1
3
6.70 (2H, d, J = 8.5 Hz), 4.66 (2H, t, J = 6.0), 3.82 (2H, t, J = 5.5 Hz); C NMR (CDCl ): δ
3
1
1
92.3, 163.6, 161.1, 157.1, 155.1, 148.5, 140.5, 135.0, 134.2, 133.3, 131.7, 131.6, 130.8, 129.3,
15.2, 111.1, 110.9, 109.8, 109.6, 80.6, 35.8.
4ꢀ(4ꢀ(4ꢀHydroxyphenyl)ꢀ8ꢀmethoxyꢀ2,3ꢀdihydrobenzo[b]oxepinꢀ5ꢀyl)benzaldehyde (9b) was
prepared from 8b by general method IV. The crude product was purified by flash column
29
ACS Paragon Plus Environment