Synthesis and evaluation of estrogen receptor ligands with bridged oxabicyclic cores containing a diarylethylene motif: Estrogen antagonists of unusual structure

Article abstract of DOI:10.1021/jm0506773

A new series of ligands for the estrogen receptor (ER) based on a three-dimensional structural motif consisting of a bridged oxabicyclic core (7-oxabicyclo[2.2.1]heptene or heptadiene) were synthesized and examined for their receptor binding activity and as regulators of transcription through the two ER subtypes, ERα and ERβ. The prototypical ligands also contain a 1,2-diarylethylene motif, common to many nonsteroidal estrogens, as an embellishment on the oxabicyclic core. Thus, these ligands bear peripheral groups typically found in ER ligands, built here upon an overall three-dimensional core topology that is unusual for these targets. Most of these compounds were conveniently synthesized by a Diels-Alder reaction of various 3,4-diarylfurans with a variety of dienophiles, neat and under mild conditions in the absence of catalysts. Some of the synthesized compounds display good binding affinity for the ER, and in transcription assays, the highest affinity compounds are antagonists on both ERs. Molecular modeling studies suggest a structural basis for the antagonist activity of these compounds. These compounds, based on the bicyclo[2.2.1]core system, expand the structural diversity of ligands that can be antagonists for the estrogen receptors.

Full text of DOI:10.1021/jm0506773

J. Med. Chem. 2005, 48, 7261-7274
7261
Synthesis and Evaluation of Estrogen Receptor Ligands with Bridged
Oxabicyclic Cores Containing a Diarylethylene Motif: Estrogen Antagonists of
Unusual Structure
Hai-Bing Zhou,† John S. Comninos,† Fabio Stossi,‡ Benita S. Katzenellenbogen,‡ and
John A. Katzenellenbogen*,†
Department of Chemistry and Department of Molecular and Integrative Physiology, University of Illinois, 600 South Mathews
Avenue, Urbana, Illinois 61801
Received July 15, 2005
A new series of ligands for the estrogen receptor (ER) based on a three-dimensional structural
motif consisting of a bridged oxabicyclic core (7-oxabicyclo[2.2.1]heptene or heptadiene) were
synthesized and examined for their receptor binding activity and as regulators of transcription
through the two ER subtypes, ERR and ERâ. The prototypical ligands also contain a
1,2-diarylethylene motif, common to many nonsteroidal estrogens, as an embellishment on
the oxabicyclic core. Thus, these ligands bear peripheral groups typically found in ER ligands,
built here upon an overall three-dimensional core topology that is unusual for these targets.
Most of these compounds were conveniently synthesized by a Diels-Alder reaction of various
3,4-diarylfurans with a variety of dienophiles, neat and under mild conditions in the absence
of catalysts. Some of the synthesized compounds display good binding affinity for the ER, and
in transcription assays, the highest affinity compounds are antagonists on both ERs. Molecular
modeling studies suggest a structural basis for the antagonist activity of these compounds.
These compounds, based on the bicyclo[2.2.1]core system, expand the structural diversity of
ligands that can be antagonists for the estrogen receptors.
Scheme 1
Introduction
The estrogen receptors (ERs), members of the family
of nuclear receptors, have emerged as attractive phar-
maceutical targets for therapeutic intervention in a wide
variety of diseases, including osteoporosis and breast
cancer. Two receptor subtypes are now known, ERR and
ERâ,1,2 and they have different tissue distribution
patterns. For example, ERR predominates in the breast
and in reproductive tissues such as the uterus, whereas
ERâ is the principal subtype in the ovary and certain
regions of the brain.3-5 Because of the known impor-
tance of ERR as a pharmaceutical target and the
potential importance of ERâ as well,6 molecules that act
as agonists or antagonists selectively on one or the other
of the ER subtypes are currently being investigated for
their therapeutic potential; those whose activity also
shows tissue selectivity, termed selective estrogen re-
ceptor modulators (SERMs),7,8 are of particular interest.
As part of our long-term interest in ER ligands, we
have undertaken exploratory studies aimed at preparing
new compounds wherein the central hydrophobic core
has, overall, a more three-dimensional topology than is
typically found in both steroidal and nonsteroidal
estrogen ligands. This design strategy was based on
structural studies of the ligand binding pockets of both
ERR and ERâ that reveal substantial unoccupied space
above and below the mean plane of the endogenous
ligand, estradiol (E2), particularly near the middle of
this molecule (namely, below the B ring and above the
C ring),9 as well as the apparent flexibility of the ligand
binding pocket. By incorporating a hydrophobic bicyclic
unit as the core structure of new ligands, we hoped to
exploit this unfilled space in the ER binding pocket and
thereby, potentially, to enhance binding affinity or ER
subtype selectivity. In fact, a number of recent reports
have described both steroidal and nonsteroidal estro-
gens, such as those derived from ferrocene,10-12 carbo-
ranes,13,14 cyclopentadienyl metal tricarbonyls,15 and
polycyclics,16 made up of structural elements having a
pronounced overall three-dimensional topology. In light
of these recent reports and in continuation of our
interest in nonsteroidal estrogens, we have taken a
novel approach to probe the ligand binding pockets in
the two ERs with a new bicyclic core structure.
* To whom correspondence should be addressed. Phone: 217 333
6310. Fax: 217 333 7325. E-mail: jkatzene@uiuc.edu.
† Department of Chemistry.
In previous studies, we have reported three different
structural motifs shown in Scheme 1, each containing
‡ Department of Molecular and Integrative Physiology.
10.1021/jm0506773 CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/18/2005

7262 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Zhou et al.
Scheme 2
a bicyclo[3.3.1]nonane core, as possible estradiol mimics,
and we have conducted a limited structure-activity
relationship (SAR) study of each of these types.17,18 Type
III compounds emerged as the most promising leads for
developing high-affinity ER ligands, but they showed
little selectivity for either ER subtype. Type II com-
pounds, on the other hand, despite their lower affinity,
exhibited significant ERâ binding selectivity. In addi-
tion, our examinations of different classes of synthetic
heterocyclic ER ligands, in which, for example, the core
structure is a simple five-membered ring heterocycle,
such as a furan19 or a pyrazole,20-22 furnish examples
of other compounds with high binding affinity, subtype
selectivity, or both. Thus, we wondered whether the
introduction of a heteroatom in a bicyclic core (e.g., an
oxa- or azabicyclic bridged compound) might provide ER
ligands with interesting biological character.
Scheme 3
A brief survey of the various available bicyclic systems
suggested that a 7-oxabicyclo[2.2.1]hept-5-ene system
would be an appropriate three-dimensional hydrophobic
core element. In addition, in a ligand of this design, the
oxygen group in the heterocyclic moiety would mimic
an element of the core of the furan system; furan-core
ER ligands are known to have exceptionally high ER
binding affinity and ERR subtype selectivity.19 Recently,
ER ligands having a related oxabicyclo[3.3.1]nonene
core structure have been reported by investigators at
Bayer,23 and researchers at Ligand Pharmaceuticals
have further developed structure-activity relationship
(SAR) studies on this class of compounds.24 However,
to the best of our knowledge, there has as yet been no
report of the synthesis of 7-oxabicyclo[2.2.1]hept-5-ene
compounds as ER ligands.
Scheme 4
In this report, we describe a novel type of ER ligand
constituted of a 7-oxabicyclic[2.2.1]hept-5-ene or hepta-
2,5-diene core that can be prepared conveniently by a
Diels-Alder reaction of a furan with a suitable dieno-
phile. By starting with appropriately substituted furans,
one can produce by this approach bicyclic[2.2.1]heptene
systems that are adorned with a 1,2-diarylethylene unit,
a motif found in many high-affinity nonsteroidal estro-
gens. Thus, the bicyclic core ER ligands of this design
have the overall three-dimensional topology that we
have sought in a form that is adorned with typical ER
ligand peripheral elements. These novel bicyclic core
compounds present ligands of unusual sizes and geom-
etries that can be used to probe the size, shape, and
flexibility of the ERR and ERâ ligand binding pockets
and potentially could be further developed as pharma-
cological agents.
by the general route shown in Scheme 3.26 Condensation
of the potassium salts of arylacetic acids 6 with R-bromo-
4-methoxyacetophenone 5 and 18-crown-6 as catalyst
gave the corresponding acetates 7. Treatment of these
esters (7) with NaH in anhydrous DMSO gave the
2(5H)-furanones 8, which were demethylated with py-
ridinium chloride at 220 °C to give the phenolic 2(5H)-
furanones 9. Diisobutylaluminum hydride reduction of
these furanones at -78 °C gave, after acidic workup,
the corresponding 3,4-diarylfurans 1b and 1c.
When the demethylation reaction of the 2(5H)-fura-
nones 8 was carefully controlled at 200 °C, the mono-
demethylated product 10a was obtained as a mixture
of two isomers, accompanied with almost the same
amount of the fully demethylated product 9a. Diisobu-
tylaluminum hydride reduction of the butenolide 10 at
-78 °C gave, after acidic workup, the corresponding
monophenolic furan 1d (Scheme 4).
Results and Discussion
Chemical Synthesis. The synthetic route for the
preparation of the oxabicyclic bridged compounds in-
volved a Diels-Alder reaction of aryl-substituted furans
with various dienophiles. The synthesis of the basic 3,4-
diphenyl furan 1a was accomplished as shown in
Scheme 2. Carbometalation of 3-phenyl-2-propyn-1-ol 2
with phenylmagnesium chloride gave the magnesium
chelate 3, which was allowed to react with DMF;
acidification of the intermediate hemiacetal 4 without
isolation gave the furan 1a.25
3,4-Bis(4-hydroxyphenyl)furan 1b and 3-(4-hydroxy-
phenyl)-4-(3-hydroxyphenyl)furan 1c were synthesized
The synthesis of the oxabicyclic bridged compounds
was effectively accomplished by a Diels-Alder reaction

Oxabicyclic Core Estrogen Receptor Ligands
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7263
Figure 1. ORTEP of exo-12a.
of furans 1 and various dienophiles (eq 1). As dieno-
Figure 2. 1H NMR assignments of exo- and endo-12a.
affinity of all of the compounds that we have prepared
in this report.
To compare the biological property of the exo isomer
with the endo one, we undertook a brief investigation
on the stereoselectivity of the reaction of 1b with 11a,
and we found that when the reaction was conducted at
35 °C for 24 h, the endo isomer of 12a was formed in
18% yield, in a ratio of 1:3 to the exo isomer (Table 1,
entry 2). This endo isomer could be isolated and
characterized spectroscopically. Identifying features in
philes, it proved to be particularly convenient to use
vinyl sulfones and sulfonates, acetylene dicarboxylates,
maleates, and maleimides. The results from reaction
with these dienophiles are summarized in Table 1.
While we did explore a range of dienophiles, certain
ones, for example, methyl vinyl ketone, gave unstable
products; others, especially moderately activated sty-
rene, vinyl ester, and phenylacetylene dienophiles,
proved to be rather unreactive, as discussed below.
A series of compounds having the 1,2-diphenylene
motif were synthesized starting from unprotected phe-
nolic or phenyl furans and the appropriate dienophiles.
It is noteworthy that high stereoselectivity was observed
in the reaction of phenolic furans with the dienophiles,
the exo product being obtained nearly exclusively; only
traces, if any, of endo products were observed. For
example, in the reaction of furan 1b with vinyl phenyl
sulfonate 11a, vinyl phenyl sulfone 11b, and vinyl ethyl
sulfone 11c, the reaction takes place smoothly at 90 °C,
neat (without solvent). The products 12a-12c were
obtained exclusively as the exo diastereomers in 80%,
81%, and 91% yields, respectively (Table 1, entries 1,
3, and 4).
1
the H NMR spectra of exo and endo 12a are given in
Figure 2.28 Assignments are facilitated by the fact that
the bridgehead protons do not couple with the endo
protons, because their dihedral angles are very close to
90°; this is evident from the X-ray structure of com-
pound exo 12a (H1-H2, 89° and H3b-H4, 76°). The most
characteristic signals are those of the proton on C-2
(H2): in the case of exo 12a, the H2 proton at 3.59 ppm
is endo, appearing as a doublet of doublets having a cis
and a gauche coupling of 8.3 and 4.6 Hz, respectively
(no coupling to H1), whereas in endo 12a the H2′ proton
at 4.15 ppm is exo, appearing as an ABX multiplet
having one cis and two gauche couplings of 9.6, 4.8, and
4.4 Hz, respectively. The other couplings and chemical
shifts are given in Figure 2.
Despite the generally good yields that we obtained
with the vinyl sulfone and sulfonyl dienophiles, we
found that the reaction of 1b with diethyl maleate 11d
and dimethyl acetylenedicarboxylate 11e gave the
products 12d and 12e in lower yields, 42% and 75%,
respectively (Table 1, entries 5 and 6). Part of the
reduced yield appears to be the sensitivity of the
products to purification by preparative TLC on silica gel.
In the reaction of 1b with 11f and 11g, the products
12f and 12g were obtained in quantitative yield, and
upon completion of the reaction, they were purified
simply by filtration and recrystallization from THF
(Table 1, entries 7 and 8). Similar to the reactions of
furan 1b, reaction of furans 1c and 1d with dienophiles
again proceeded with exo selectivity but produced a 1:1
mixture of two inseparable regioisomers (Table 1,
entries 9-12). In comparison, compounds 12l, 12m and
12n, which have no hydroxyl group on the phenyl ring,
were also prepared by the reaction of furan 1a with 11a,
11b, and 11h in 78%, 81%, and 91% yields, respectively,
accompanied with small amounts (∼5%) of endo isomers
(Table 1, entries 13-15).
Endo diastereomers typically predominate as the
products of Diels-Alder reactions conducted under
conditions of kinetic control because of stabilizing
secondary orbital interactions between the diene and
dienophile that are only possible in the endo transition
state. Nevertheless, the predominance of exo products
in related furan Diels-Alder reactions is precedented.27
This unexpected exo selectivity probably arises because
the secondary interactions with a sulfonate-based di-
enophile are likely to be weaker than those with a
carbonyl-based dienophile and also because this facile
cycloaddition reaction is probably readily reversible,
which results in the rapid accumulation of the thermo-
dynamically more stable exo product.
Although the endo and exo stereoisomers have char-
acteristic 1H NMR signals (see below), to verify our
stereoisomeric assignments, we obtained an X-ray crys-
tal structure of compound 12a, which confirmed that it
is exo-5,6-bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-
5-ene-2-sulfonic acid phenyl ester (Figure 1). As noted
below, this compound, in fact, has the highest binding

7264 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Table 1. Diels-Alder Reaction of Furan 1 with Dienophile 11a
Zhou et al.
a To a round-bottom flask was added furan (0.2 mmol) and dienophile (0.26 mmol), and the mixture was stirred at the specified
temperature and time given in the table. b The conversion was calculated based on the crude 1H NMR of the mixture. c Isolated yield
based on the furan used. d The reaction was conducted at 35 °C for 24 h, and 57% of exo-12a was obtained.

Oxabicyclic Core Estrogen Receptor Ligands
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7265
Scheme 5
Scheme 6
In our previous work on furan-core ER ligands, we
found that triaryl, especially trisphenol, furans always
show higher binding affinity and subtype selectivity
than the corresponding bisphenol analogues.19 Here, we
also wanted to introduce a third phenol ring into the
bridged compound, because this design might lead to
an ER ligand with interesting biological activity. First,
we tried to introduce a third phenol in the bicyclic core
as shown in Scheme 5. However, due to the hindrance
of the diphenyl groups, attempts to achieve the reaction
of furan with dienophiles that bear a phenyl group, for
example, methyl 3-phenylpropiolate, diphenylacetylene,
styrene, ethyl cinnamate, or vinyl benzoate, failed, even
with Lewis acid catalysts. Thus, we took an alternative
synthetic strategy.
Methyl 3-bromopropiolate 13, an active dienophile,
was synthesized as described in the literature (Scheme
5).29 Reaction of 13 with furan 1a afforded oxabicyclo-
heptene 14 in high isolated yield. The triaryl oxabicyclic
bridged compounds 15ab were prepared by standard
Suzuki coupling reactions with the appropriate aryl
boronic acids. Unfortunately, the product of the diphe-
nolic furan 1b with 13 was very unstable, and attempts
to isolate the product failed.
In all the compounds that we have prepared, the
ligand bearing a phenyl sulfonate in the 2-position of
the core always had a higher ER binding affinity (see
below). We wondered whether the replacement of the
phenyl ring on the sulfonate moiety with a p-hydroxy-
phenyl group might lead to ligands with increased
binding affinity. Thus, we investigated the synthesis of
compound 12q, as shown in Scheme 6. It turned out that
the synthesis of the ethenesulfonic acid 4-hydroxyphenyl
ester (17) from the reaction of chloroethanesulfonyl
chloride with hydroquinone was unsuccessful. As an
alternate approach, 2-chloroethanesulfonic acid 4-meth-
oxyphenyl ester (16) was synthesized in 50% yield by
the reaction of 2-chloroethanesulfonyl chloride with
4-methoxyphenol in the presence of 25% aqueous NaOH
in 1,2-dichloroethane.30 Reaction of the vinyl sulfone 16
with furan 1b, using the same reaction conditions as
described above, afforded the bicycle 12o in 90% yield.
However, our attempts to demethylate this product
(12o) gave only products from decomposition. An alter-
nate route was successful however. Demethylation of
the vinyl sulfone 16 using 20 equiv of BF3‚SMe2 in
dichloromethane gave the deprotected vinyl sulfone 17
in 65% yield, and reaction of the latter with furans 1a
and 1b afforded products 12p and 12q in 75% and 67%
yields, respectively. At the same time, a significant
amount of the endo isomer was also observed in these
reactions, and the endo isomer of 12p could be isolated
in 15% yield.
Binding Affinity for Estrogen Receptors ERR
and ERâ. We examined the binding affinity of these
novel bicyclic ligands to both full-length, purified human
ERs, ERR and ERâ, as well as ER in uterine cytosol
preparations (which contains principally ERR31 and is
included because it is representative of ER in a more
natural protein context), by a competitive radiometric
binding assay, using methods that have been described
elsewhere in detail.32,33 The binding affinities of all the
ligands in this study are expressed as relative binding
affinity (RBA) values, where estradiol has an affinity
of 100%; the RBA values determined with the three
receptor preparations are given in Table 2.
As a global observation, it is noteworthy that the
binding affinity of the ligands in this series depends on
the position of the hydroxyl group in the phenyl ring
and the substituents at the 2 and 3 positions of the
7-oxabicycloheptane core. The compound that has the
highest binding affinity in all three ER preparations is
exo-5,6-bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-5-
ene-2-sulfonic acid phenyl ester 12a, a compound that
possesses a p-hydroxyl group in both of the core phenyl
substituents and a phenyl sulfonate moiety at the
2-position of the bicyclic unit. The RBA values of this
compound are 18 on uterine ER and 9.3 and 1.7 for ERR
and ERâ, respectively (Table 2, entry 1). Although lamb
uterine cytosol consists almost exclusively of ERR,

7266 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Zhou et al.
Table 2. Relative Binding Affinity (RBA) of Oxabicyclic Non-Steroidal Analogues for Estrogen Receptor ERR and ERâa

Oxabicyclic Core Estrogen Receptor Ligands
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7267
Table 2. (Continued)
a Relative binding affinitiy (RBA) values are determined by competitive radiometric binding assays and are expressed as ICe5s0tradiol
/
compound
IC
× 100 ( the range or the standard deviation (RBA, estradiol ) 100%). In these assays, the Kd for estradiol is 0.2 nM on ERR
50
and 0.5 nM on ERâ. For details, see the Experimental Section.
Scheme 7
binding affinities to the ER in this complex protein
mixture are not always equal to that of purified human
ERR, although it is not clear whether this is due to
species differences, preferential binding of tracer, [3H]-
estradiol, by the other, nonreceptor proteins present in
the cytosol preparation, or the influence of endogenous
coregulator proteins. Surprisingly, the endo isomer of
12a shows 27-fold selectivity for ERR over ERâ; al-
though with a decrease in overall binding, particularly
for ERâ, with an RBA of 4.1 and 0.15 for ERR and ERâ,
respectively, it has the highest ERR binding preference
in the series (Table 2, entry 2).
preference (14-fold) of all the exo ligands in this series
(RBA ERR 2.2 and ERâ 0.16) (Table 2, entry 19). Also,
both exo and endo isomers of compound 12p gave very
low binding affinities on ERR and ERâ (Table 2, entries
17 and 18). It is of note that replacement of the phenyl
sulfonate group with a phenyl sulfone, as the compound
12b, causes a great decrease in binding, particularly for
ERâ, only giving RBA values of 0.64 and 0.059 for ERR
and ERâ, respectively; thus, the sulfone diphenol 12b
has an ERR binding preference (11-fold) that is compa-
rable to that of the sulfone triphenol 12q (Table 2, entry
3).
Another high-affinity ER ligand, identified by us
earlier, has a phenyl sulfonate substituent, as does the
bicycle 12a. Compound 18 (Scheme 7) is one of several
trace impurities that are present in phenol red prepara-
tions, the pH indicator dye that is widely used in cell
culture, and it is responsible for the estrogenic activity
of phenol red-containing cell culture medium.34 This
phenyl sulfonate 18 has an affinity for ER that is about
half that of estradiol, yet close analogues in which the
sulfonate is replaced by carboxylate or carboxamide
links have much lower ER binding affinities,35,36 as is
the case with the 7-oxabicyclo[2.2.1]heptenes.
When the phenyl group on the sulfone in 12b was
replaced with an ethyl group, as compound 12c, there
is a dramatic drop in binding affinity on both ER
subtypes, which is probably due to the increased polarity
of this compound or a reduced steric effect (Table 2,
entry 4). The presence of diethoxyl carbonyl groups at
both the C-2 and C-3 positions of the core, as in
compound 12d, gave an RBA value of 0.02% for ERR
(Table 2, entry 5). Conversely, replacing the single bond
with a rigid ethylene spacer, as in compound 12e,
reduced binding affinities to nondetectable levels. This
The effect of the C-5 and C-6 phenolic functions on
binding affinity is evident from a comparison of exo 12a
and 12b with corresponding phenyl sulfonates 12h-
12k, which have protected or altered phenol positions.
When one of the hydroxyl groups was moved to the meta
position of the phenyl ring (e.g., 12h and 12i), binding
affinity to all receptor preparations decreased 10-fold
compared to compounds 12a and 12b (Table 2, entries
9 and 10). Compounds having only one hydroxyl group
(e.g., 12j and 12k) also had substantially lower binding
affinities than 12a and 12b (Table 2, entries 11 and 12).
A chemical feature common to nearly all synthetic ER
ligands having good binding affinity is the presence of
a phenolic ring that seems to mimic the steroid “A ring”
present in natural estrogens. This phenol forms impor-
tant hydrogen bonds with residues Glu353 and Arg394
in ERR (or with the corresponding residues in ERâ), as
well as with a water molecule.9 Consistent with this is
the very low affinity of compounds 12l-12n, which have
no phenolic hydroxyl groups (Table 2, entries 13-15).
In the ligands studied here, the substituent on the
C-2 or C-3 position of the bridged core has a significant
effect on the binding affinity and selectivity. The highest
affinity ligand, 12a (exo), bears a phenyl sulfonate group
at this position. Because many nonsteroidal estrogens
are triphenols, it was somewhat surprising that place-
ment of a methoxyl or hydroxyl group at the para
position of the sulfonate phenyl ring, as in compounds
12o-12q, caused a decrease, rather than an increase,
in affinity for both ERR and ERâ binding (Table 2,
entries 16-19). Even the triol compound 12q showed a
marked drop in binding, particularly for ERâ, thus
giving a compound that has the highest ERR binding

7268 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Zhou et al.
might be due in part to the decreased flexibility of the
ligand (Table 2, entry 6). However, the compounds 12f
and 12g, produced by the reaction of furan 1b with
N-methyl and phenylmaleimide, also exhibit little or no
binding affinity. This suggests that the polar substituent
(i.e., the nitrogen atom in these compounds) might be
responsible for their immeasurably low binding affini-
ties (Table 2, entries 7 and 8). Also, when steric
hindrance is introduced at the C-3 position of the core
by an aryl group, as in compounds 15a,b, one obtains
analogues that possess very poor binding properties,
presumably because of an unfavorable interaction of the
third phenyl ring in 15a,b with the receptor (Table 2,
entries 20 and 21).
Note that all the furans synthesized in this report
(1a-d) exhibit little or no binding affinity (data not
shown) except furan 1c, which has an RBA of 0.087%
for ERR and 0.241% for ERâ. This is consistent with
our earlier work on furan-core ER ligands where we
found that only tetrasubstituted furans had high bind-
ing affinities (Scheme 1).19 Thus, it appears that the
other carbocyclic bridge of the oxabicycloheptene ana-
logues is making a strong contribution to the ERR and
ERâ binding affinity. This result also indicates that our
design for the three-dimensional hydrophobic core ele-
ment is a reasonable one and that the core bicyclic unit
plays an important role in engendering favorable bind-
ing to the receptor and for achieving, in some cases, good
subtype selectivity.
Overall then, the disposition of the appended phenols,
the substituents in the C-2 and C-3 positions of the
bicyclic core unit in these ER ligands, and the electron-
withdrawing group derived from the dienophile all prove
to be factors in determining the affinity and selectivity
of these new ligands for the ER subtypes. High affinity
is generally found when the bicyclic core is embellished
with the diarylethylene unit at the C-5 and C-6 sites,
together with an appropriate third substituent, produc-
ing a molecule that is, overall, neither excessively
crowded nor polar.
Figure 3. Antagonism (ERR, 9, and ERâ, 2, shown as dotted
lines) and agonism (ERR, b, and ERâ, 0, shown as solid lines)
of estradiol-activated transcription activation through ERR and
ERâ by compounds 12a and 12b. Human endometrial cancer
cells (HEC-1) were transfected with expression plamids for
ERR or ERâ and the estrogen-responsive reporter gene 2ERE-
pS2-Luc and were then incubated with the indicated concen-
tration of ligand, together with 1 nM estradiol (E2), for 24 h.
Values are the mean ( range or SD from two or more
experiments and are expressed as a percent of the activity of
ERR and ERâ with 10-9 M E2. Further details are given in
the Experimental Section and in our prior publication.37
Figure 3. From these assays, it is clear that compounds
12a and 12b are antagonists on both ERR and ERâ.
They have a slight potency preference for the ERâ
subtype, similar to that observed with the ICI anties-
trogen (data not shown), and overall they are less
potent.
Transcription Activation through Estrogen Re-
ceptors ERR and ERâ. Two of the compounds showing
higher binding affinity values for either one of the two
ERs (exo-12a and -12b) were assayed for gene tran-
scriptional activity through ERR and ERâ. These cotrans-
fection assays were conducted in human endometrial
cancer (HEC-1) cells, using expression plasmids for full-
length human ERR or ERâ and an estrogen-responsive
reporter gene construct.37 In an initial screen, the
transcriptional activity of these compounds was tested
at 10-8 and 10-6 M alone (as agonists) or in the presence
of 10-9 M estradiol (as antagonists). It was clear from
this screen that both compounds were essentially devoid
of agonist activity (showing stimulatory activity less
than 5% that of estradiol) and functioned as antagonists.
To further characterize their activity as ER antago-
nists, compounds 12a and 12b were assayed in an
antagonist dose-response mode at four concentrations,
10-8 to 10-5 M, in the presence of 10-9 M E2. In all cases,
transcriptional activity is expressed relative to that
obtained with 10-9 M estradiol alone, which is set at
100%. The well-known antiestrogen faslodex (fulves-
trant; ICI 182,780) was also assayed under these
conditions. The dose response curves are shown in
Because they lack the typical extended and polar or
basic side chain so commonly found in ER antagonists,
compounds 12a and 12b represent a novel structural
class as ER antagonists. It is also curious that their
binding affinity preference is for the ERR subtype yet
their potency preference as transcriptional antagonists
is for the ERâ subtype, although it has long been
recognized that binding affinities and transcriptional
potencies do not correlate in a strictly quantitative
fashion.38
Considerations of Ligand Orientation in the
Estrogen Receptor Binding Pocket. Because the
oxabicyclo[2.2.1]heptene core compounds that we have
prepared represent a new structural class of ER ligands,
we undertook a molecular modeling study of the highest
affinity ligand, the sulfonate 12a (exo). We hoped that
this investigation would point to how ligands in this
class might be oriented in the ligand binding pockets of
the receptors and why they behave as antagonists,
rather than agonists.
To build models, we selected the ERR-Ral (raloxifene)
X-ray crystal structure (1ERR in the Protein Data Bank)
for our ER model, because raloxifene is also an antago-
nist of ER action and has an ERR/ERâ selectivity similar

Oxabicyclic Core Estrogen Receptor Ligands
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7269
Figure 4. (A) Model of (1R,2S,4R)-oxabicyclo[2.2.1]heptene 12a in the ERR ligand binding pocket. The surface of the ligand is
shown as the electrostatic potential. The ERR pocket is displayed as dots mapped with the electrostatic potential. (B) Comparison
of 12a (element colors) overlaid with raloxifene (purple).
to that of 12a. Determining the orientation of 12a
within the ligand binding pocket is challenging, because
two pendant phenolic rings are attached to the bicyclic
core, each of which could serve as the crucial “A ring”
mimic of natural estrogens. Also, each phenolic moiety
possesses two possible binding orientations of the bicy-
clic core within the receptor through a 180° rotation
around the aryl-oxabicyclic core bond. Binding orienta-
tion determination is further complicated in that 12a
was prepared as a racemic mixture of enantiomers.
With eight possible orientations, four each of the two
enantiomers of 12a, we decided to use the FlexX routine
for each enantiomer (Sybyl 7.0, Tripos Inc.) instead of
individually prepositioning each of the eight possible
binding modes as required for use of the FlexiDoc
routine. FlexX uses an incremental construction algo-
rithm to build a ligand into a binding pocket and
calculate a docking score based on interactions with the
binding pocket. We then selected the top-ranked docking
orientation and minimized the complex (see Figure 4).
Interestingly, only one enantiomer, that possessing the
1R,2S,4R absolute configuration, shown in this figure,
could be docked effectively into the ligand binding
pocket, and a single binding orientation was predicted
by the FlexX routine.
The ligand binding pocket of the ERs is largely
hydrophobic and rather open, having polar residues only
at the regions where the phenol and 17â-hydroxyl group
of estradiol are accommodated; the phenol hydrogen
bonds with glutamate and arginine residues and a
structured water in the rather tight A-ring binding
subpocket, while a histidine is the hydrogen bonding
partner of the 17â-hydroxyl group.9 It was encouraging
that FlexX predicted a phenol of the bicycle 12a as an
“A ring” mimic, since the phenol moiety is estimated to
provide 1.9 kcal/mol of stabilization through its three
hydrogen bonds.39 The second phenol of 12a was posi-
tioned in the 7R region of the ligand binding pocket, a
position that is generally tolerant of large substituents.39
Many ER ligands form a significantly weaker hydrogen
bond with His524, which is estimated to provide 0.6
kcal/mol of stabilization energy.39 The second phenol of
12a, however, is not oriented properly to hydrogen bond
with His524, due to the rigidity of the oxabicyclic core,
and it appears to have no hydrogen bonding partner in
our model.
The phenylsulfonate ester in our model is positioned
into the 11â groove of the ligand binding pocket with
the phenyl group extending outward. This could help
to explain the antagonistic activity of the 12a. Antago-
nists such as raloxifene appear to inhibit coactivator
recruitment through displacement of helix 12 with a
large substituent that protrudes through this same
groove in the 11â-subpocket.9,40 Analogous to known ER
antagonists, the phenylsulfonate ester of 12a is posi-
tioned properly to prevent helix 12 from binding back
over the ligand pocket, as it must to form the hydro-
phobic cleft into which the LXXLL motif of coactivators
binds.40 A portion of the three-dimensional oxabicyclic
core also occupies space in the 11â region of the binding
pocket that is typically void of substituents with ste-
roidal estrogens.
Molecular modeling studies, of course, are not defini-
tive, so it would prove interesting to perform X-ray
crystallographic structure studies in the future to
understand in greater detail the molecular basis for the
antagonist activity of compounds 12a and 12b, because
their structures are so unlike that of typical ER antago-
nists.
Conclusion
In summary, to examine the tolerance of the estrogen
receptors ERR and ERâ for “thickness” at the core of a
ligand, we have prepared a series of novel ligands for
these receptors based on an inherently three-dimen-
sional 7-oxabicyclo[2.2.1]hept-5-ene core. Ligands in this
series can be readily prepared by a Diels-Alder reaction

7270 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Zhou et al.
Hz, 1H), 2.57 (ddd, J ) 12.2, 4.6, 4.6 Hz, 1H), 2.16 (dd, J )
12.2, 8.3 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 155.96, 155.90,
149.48, 141.6, 137.50, 130.16, 129.34, 129.01, 122.37, 116.13,
116.0, 84.56, 83.21, 60.89, 30.81; HRMS (ESI) calcd for
C24H20O6SH, 437.1059 (M + H+); found, 437.1063.
between a 3,4-disubstituted furan and various dieno-
philes without Lewis acid catalysis; the exo stereoiso-
mers predominate. Those that have 4-hydroxyphenyl
substituents at C-5 and C-6 and a phenyl sulfonate at
C-2 have substantial affinity for both ERR and ERâ, but
with an affinity preference for ERR. Minor changes to
this pattern of substitution generally lead to analogues
of lower affinity. In cell-based transcription assays, two
compounds with high ER binding affinity were found
to be antagonists on both ER subtypes, with a modest
potency preference for ERâ. These ER ligands have
unusual bicyclic core structures, and although they are
antagonists, they do not have substituents that are
typically found in ER antagonists; molecular modeling,
however, shows that the phenyl sulfonate moiety of the
antagonist ligands is capable of stabilizing an ER
conformation that is likely not to be able to bind
coactivators, and this might explain their antagonist
character. The generation of this new series of ER
ligands provides important insight into the diversity of
structures that can function as ligands for the estrogen
receptors and function as antagonists on these impor-
tant pharmaceutical targets.
endo-5,6-Bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-
5-ene-2-sulfonic Acid Phenyl Ester (12a). endo-12a was
purified by preparative thin-layer chromatography (45% EtOAc/
hexanes) to give a gray solid (18% yield) (mp 130-132 °C); 1H
NMR (400 MHz, CDCl3) δ 7.39-7.35 (m, 5H), 7.31-7.21 (m,
2H), 7.13 (d, J ) 8.8 Hz, 2H), 6.75 (d, J ) 8.8 Hz, 2H), 6.70 (d,
J ) 8.8 Hz, 2H), 5.57 (dd, J ) 4.4, 1.0 Hz, 1H), 5.31 (dd, J )
4.8, 1.0 Hz, 1H), 4.87 (s, 1H), 4.80 (s, 1H), 4.15 (ABX, J ) 9.6,
4.8, 4.4 Hz, 1H), 2.57 (ABX, J ) 12.0, 9.6, 4.8 Hz, 1H), 2.06
(dd, J ) 12.0, 4.8 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ
155.77, 154.29, 148.00, 141.29, 138.27, 136.61, 130.16, 129.88,
129.72, 127.58, 125.78, 122.40, 115.97, 115.41, 84.63, 83.18,
59.56, 29.97; HRMS (ESI) calcd for C24H20O6SH, 437.1059 (M
+ H+); found, 437.1057.
2-Benzenesulfonyl-5,6-bis-(4-hydroxyphenyl)-7-
oxabicyclo[2.2.1]hept-5-ene (12b). Product 12b was ob-
tained as orange oil that solidified on standing (81% yield).
The product was then recrystallized in ethyl acetate/hexanes
1
to give an off-white crystal (mp 174-175 °C); H NMR (400
MHz, CDCl3) δ 7.97 (d, J ) 7.2 Hz, 2H), 7.66-7.71 (m, 1H),
7.59 (t, J ) 7.6 Hz, 2H), 7.12 (d, J ) 8.4 Hz, 2H), 7.03 (d, J )
8.4 Hz, 2H), 6.73 (d, J ) 8.8 Hz, 2H), 6.69 (d, J ) 8.8 Hz, 2H),
5.59 (s, 1H), 5.27 (d, J ) 3.6 Hz, 1H), 5.01 (s, 1H), 4.96 (s,
1H), 3.42 (dd, J ) 8.0, 4.4 Hz, 1H), 2.38 (ddd, J ) 12.4, 3.6,
3.6 Hz, 1H), 1.94 (dd, J ) 12.0, 8.4 Hz, 1H); 13C NMR (125
MHz, acetone-d6) δ 157.60, 157.51, 141.47, 139.86, 137.75,
134.73, 133.86, 129.55, 129.04, 128.91, 128.76, 123.80, 115.70,
115.68, 83.78, 83.13, 65.30, 19.98; HRMS (ESI) calcd for
C24H20O5SH, 421.1109 (M + H+); found, 421.1118.
Experimental Section
Materials and Methods. All reagents and solvents were
obtained from Aldrich. Tetrahydrofuran, diethyl ether, toluene,
and dichloromethane were obtained prior to use from a solvent-
dispensing system (SDS) built by J. C. Meyer. Glassware was
oven-dried, assembled while hot, and cooled under an inert
atmosphere. Unless otherwise noted, all reactions were con-
ducted in an inert atmosphere. Reaction progress was moni-
tored using analytical thin-layer chromatography (TLC) on
0.25 mm Merck F-254 silica gel glass plates. Visualization was
achieved by either UV light (254 nm) or potassium perman-
ganate indicator. Flash chromatography was performed with
Woelm silica gel (0.040-0.063 mm) packing.
2-Ethylsulfonyl-5,6-bis-(4-hydroxyphenyl)-7-oxabicyclo-
[2.2.1]hept-5-ene (12c). A white solid was formed during the
reaction, and the pure 12c was obtained by recrystallization
(75%) from THF (mp 178-180 °C); 1H NMR (500 MHz,
acetone-d6) δ 8.60 (s, br, 2H), 7.23 (d, J ) 3.0 Hz, 2H), 7.21 (d,
J ) 3.0 Hz, 2H), 6.82 (d, J ) 3.0 Hz, 2H), 6.80 (d, J ) 3.0 Hz,
2H), 5.58 (d, J ) 4.4 Hz, 1H), 5.36 (dd, J ) 4.4, 1.0 Hz, 1H),
3.41 (dd, J ) 8.5, 4.5 Hz, 1H), 3.10, (q, J ) 7.0 Hz, 2H), 2.37
(ddd, J ) 12.0, 4.5, 4.5 Hz, 1H), 2.13 (dd, J ) 12.5, 8.5 Hz,
1H), 1.31 (t, J ) 7.0 Hz, 3H); 13C NMR (125 MHz, acetone-d6)
δ 157.55, 141.40, 137.78, 129.04, 128.97, 124.60, 123.93,
115.77, 115.69 (one carbon missing as a result of overlap),
83.45, 83.11, 62.65, 45.64, 30.11, 5.29; HRMS (ESI) calcd for
C20H20O5SH, 373.11096 (M + H+); found, 373.1099.
5,6-Bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-5-
ene-2,3-dicarboxylic Acid Diethyl Ester (12d). Compound
12d was purified by preparative thin-layer chromatography
(30% EtOAc/hexanes) to give a light-yellow solid (42% yield)
that was recrystallized from ether/hexanes (mp 89-91 °C); 1H
NMR (400 MHz, CDCl3) δ 7.15 (d, J ) 8.8 Hz, 4H), 6.75 (d, J
) 8.8 Hz, 4H), 5.45 (s, 2H), 5.38 (s, 2H), 4.18 (q, J ) 7.0 Hz,
4H), 3.11 (s, 2H), 1.28 (t, J ) 7.0 Hz, 6H); 13C NMR (100 MHz,
CDCl3) δ 171.74, 155.87, 139.90, 129.13, 125.09, 115.95, 85.55,
61.51, 48.92, 14.34; HRMS (ESI) calcd for C24H24O7H, 425.16001
(M + H+); found, 425.1598.
5,6-Bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hepta-
2,5-diene-2,3-dicarboxylic Acid Dimethyl Ester (12e). A
yellow solid formed during the reaction, and the pure 12e was
obtained as a yellow solid (75%) that was recrystallized from
ether (mp 214-216 °C);1H NMR (500 MHz, acetone-d6) δ 8.64
(s, br, 2H), 7.31 (d, J ) 9.0 Hz, 4H), 6.81 (d, J ) 9.0 Hz, 4H),
5.93 (s, 2H), 3.77 (s, 6H); 13C NMR (125 MHz, acetone-d6) δ
176.40, 157.71, 139.37, 129.02, 123.94, 115.82, 85.63, 61.52,
49.10; HRMS (ESI) calcd for C22H18O7H, 395.1131 (M + H+);
found, 395.1131.
8,9-Bis-(4-hydroxyphenyl)-4-methyl-10-oxa-4-aza-
tricyclo[5.2.1.02,6]dec-8-ene-3,5-dione (12f). A white solid
formed during reaction and was washed by CH2Cl2 and ether;
pure 12f was obtained as white solid (96%) yield that was
recrystallized from THF (mp 224-226 °C); 1H NMR (400 MHz,
1H NMR and 13C NMR spectra were obtained on a 400 or
500 MHz instrument. The chemical shifts are reported in ppm
and are referenced to either tetramethylsilane or the solvent.
Mass spectra were recorded under electron impact conditions
at 70 eV. Melting points were obtained on a Thomas-Hoover
MelTemp apparatus and are uncorrected. Elemental analyses
were performed by the Microanalytical Service Laboratory of
the University of Illinois. Those final components that did not
give satisfactory combustion analysis gave satisfactory exact
mass determinations and were found to be at least 96% pure
by HPLC analysis. Furan 1a,25 1b, and 1c26 and methyl
3-bromopropiolate 1329 were synthesized according to pub-
lished procedures.
General Procedure for the Diels-Alder Reaction.
Furan 1 (0.2 mmol) and dienophiles (0.26 mmol) were placed
in a round flask, and the mixture was stirred under a N2
atmosphere at 90 °C for 6-24 h. The crude product was
purified by flash chromatography (10-30% EtOAc/hexanes),
preparative thin-layer chromatography, or recrystallization.
After isolation, the sample was dried under vacuum in an
Aberhalden apparatus in refluxing benzene for 24-48 h. Even
after this process some samples still contained water in the
crystals. This was confirmed by 1H NMR, and suitable cor-
rections are made for this water in evaluation of combustion
microanalysis data.
exo-5,6-Bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-
5-ene-2-sulfonic Acid Phenyl Ester (12a). exo-12a was
purified by flash chromatography (30% EtOAc/hexanes) to give
a light yellow needle (80% yield) that was recrystallized from
1
ethyl acetate/hexanes (mp 142-143 °C); H NMR (400 MHz,
CDCl3) δ 7.66-7.36 (m, 9H), 6.77 (d, J ) 8.8 Hz, 2H), 6.76 (d,
J ) 8.6 Hz, 2H), 5.70 (d, J ) 1.0 Hz, 1H), 5.40 (dd, J ) 4.6,
1.0 Hz, 1H), 5.06 (s, 1H), 5.02 (s, 1H), 3.59 (dd, J ) 8.3, 4.6

Oxabicyclic Core Estrogen Receptor Ligands
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7271
acetone-d6) δ 8.67 (s, br, 2H), 7.25 (d, J ) 8.8 Hz, 4H), 6.84 (d,
J ) 8.8 Hz, 4H), 5.39 (s, 2H), 3.23 (s, 2H), 2.90 (s, 3H); 13C
NMR (100 MHz, acetone-d6) δ 163.53, 159.22, 152.53, 145.21,
128.73, 125.38, 115.67, 89.32, 51.85; HRMS (ESI) calcd for
C21H17O5NH, 364.1185 (M + H+); found, 364.1190.
thin-layer chromatography (40% EtOAc/hexanes) to give a pale
yellow solid (85% yield) that was recrystallized from ether/
hexanes; 1H NMR (400 MHz, CDCl3) δ 7.96-7.98 (m, 2H),
7.57-7.68 (m, 3 H), 7.17 (d, J ) 8.8 Hz, 0.6H), 7.12 (d, J ) 8.8
Hz, 0.4H), 7.08 (d, J ) 8.8 Hz, 0.4H), 7.02 (d, J ) 8.8 Hz, 0.6H),
6.80 (d, J ) 8.8 Hz, 0.6H), 6.75 (d, J ) 8.8 Hz, 0.4H), 6.73 (d,
J ) 8.8 Hz, 0.4H), 6.80 (d, J ) 8.8 Hz, 0.6H), 6.69 (d, J ) 8.8
Hz, 0.6H), 5.60 (d, J ) 0.8 Hz, 1H), 5.41 (s, br, 1H), 5.27 (d, J
) 4.4 Hz, 1H), 3.79 (s, 3 H), 3.42 (dd, J ) 8.4, 4.4 Hz, 1H),
2.38 (ddd, J ) 12.4, 4.6, 4.6 Hz, 1H), 1.94 (dd, J ) 12.0, 8.4
Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 159.69, 159.55, 155.93,
155.91, 141.50, 138.92, 138.82, 137.77, 134.11, 130.29, 130.04,
129.59, 129.53, 129.49, 129.43, 129.19, 129.01, 128.94, 128.73,
128.56, 128.11, 125.11, 124.41, 115.95, 115.15, 114.41, 113.82,
83.86, 83.51, 66.13, 65.52, 65.48, 55.5, 30.35; HRMS (ESI) calcd
for C25H22O5SH, 435.1266 (M + H+); found, 435.1265.
4,8,9-Triphenyl-10-oxa-4-aza-tricyclo[5.2.1.02,6]dec-8-
ene-3,5-dione (12g). A white solid formed during reaction and
was washed by CH2Cl2 and ether; the pure 12g was obtained
as white solid (97% yield) that was recrystallized from THF
1
(mp 237-239 °C); H NMR (400 MHz, acetone-d6) δ 8.66 (s,
br, 2H), 7.48-7.52 (m, 2H), 7.40-7.44 (m, 1H), 7.29-7.31 (m,
2H), 7.30 (d, J ) 8.0 Hz, 4H), 6.85 (d, J ) 8.0 Hz, 4H), 5.53 (s,
2H), 3.41 (s, 2H); 13C NMR (100 MHz, acetone-d6) δ 175.58,
157.87, 139.56, 133.04, 129.06, 129.02, 128.45, 127.04, 123.91,
115.82, 86.21, 49.27; HRMS (ESI) calcd for C26H19O5H,
426.13414 (M + H+); found, 426.1354.
5,6-Diphenyl-7-oxabicyclo[2.2.1]hept-5-ene-2-sulfon-
ic Acid Phenyl Ester (12l). Compound 12l was purified by
preparative thin-layer chromatography (2% ether/CH2Cl2) to
give a colorless needle (78% yield) that was recrystallized from
6-(3- or 4-Hydroxyphenyl)-5-(4- or 3-hydroxyphenyl)-
7-oxabicyclo[2.2.1]hept-5-ene-2-sulfonic Acid Phenyl Es-
ter (12h, Mixture of 1.2:1 Isomers). Compound 12h was
purified by flash chromatography to give a pale yellow solid
(90% yield) that was recrystallized from ether/hexanes; 1H
NMR (400 MHz, CDCl3) δ 7.16-7.37 (m, 6H), 6.72-6.85 (m,
7H), 5.72 (d, J ) 1.2 Hz, 0.45H), 5.69 (d, J ) 1.2 Hz, 0.55H),
5.42 (dd, J ) 4.4, 0.4 Hz, 0.55H), 5.40 (dd, J ) 4.4, 0.4 Hz,
0.45H), 5.11 (m, br, 2 H), 3.64 (dd, J ) 8.4, 4.4 Hz, 0.55H),
3.57 (dd, J ) 8.4, 4.4 Hz, 0.45H), 2.54-2.59 (m, 1H), 2.19 (dd,
J ) 12.0, 8.4 Hz, 0.45H), 2.13 (dd, J ) 12.4, 8.4 Hz, 0.55H);
13C NMR (100 MHz, CDCl3) δ 156.22, 156.15, 156.12, 155.97,
149.41, 143.59, 141.53, 139.50, 139.32, 1137.38, 136.36, 135.06,
134.34, 133.55, 130.60, 130.40, 130.20, 129.32, 127.49, 124.60,
123.81, 122.36, 120.09, 119.81, 116.19, 116.05, 115.73, 115.62,
114.17, 113.96, 84.62, 84.51, 83.32, 83.20, 60.95, 60.62, 30.92,
30.53; HRMS (ESI) calcd for C24H20O6SH, 437.10597 (M + H+);
found, 437.1064.
2-Benzenesulfonyl-6-(3- or 4-hydroxyphenyl)-5-(4- or
3-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-5-ene (12i, Mix-
ture of 1.3:1 Isomers). Compound 12i was purified by flash
chromatography to give a pale yellow solid (66% yield) that
was recrystallized from ether/hexanes; 1H NMR (400 MHz,
CDCl3) δ 7.93-7.97 (m, 2H), 7.50-7.68 (m, 3H), 7.00-7.13 (m,
1H), 7.10 (d, J ) 8.8 Hz, 2H), 6.73 (d, J ) 8.8 Hz, 2H), 6.61-
6.75 (m, 3H), 5.70-5.90 (m, br, 2 H), 5.72 (d, J ) 1.2 Hz,
0.44H), 5.57 (d, J ) 1.2 Hz, 0.56H), 5.27 (dd, J ) 4.4, 0.4 Hz,
0.56H), 5.23 (dd, J ) 4.4, 0.4 Hz, 0.44H), 3.47 (dd, J ) 8.0, 4.4
Hz, 0.56H), 3.39 (dd, J ) 8.4, 4.4 Hz, 0.44H), 2.32-2.38 (m,
1H), 1.96 (dd, J ) 12.0, 8.4 Hz, 0.44H), 1.89 (dd, J ) 12.4, 8.4
Hz, 0.56H); 13C NMR (100 MHz, CDCl3) δ 156.35, 156.15,
143.66, 141.43, 139.60, 138.63, 138.58, 137.58, 134.45, 134.22,
133.61, 130.33, 129.67, 129.63, 129.06, 128.93, 128.90, 124.44,
123.68, 119.79, 119.32, 116.04, 115.60, 115.51, 114.21, 113.66
(six carbons missing as a result of overlap), 83.92, 83.72, 83.64,
83.49, 65.56, 65.22, 30.48, 30.00; HRMS (ESI) calcd for
C24H20O5SH, 421.11096 (M + H+); found, 421.1109.
5(or 6)-(4-Hydroxyphenyl)-6(or 5)-(4-methoxyphenyl)-
7-oxabicyclo[2.2.1]hept-5-ene-2-sulfonic Acid Phenyl Es-
ter (12j, Mixture of 1:1 Isomers). Compound 12j was
purified by preparative thin-layer chromatography (30% EtOAc/
hexanes) to give a pale yellow solid (77% yield) that was
recrystallized from ether/hexanes; 1H NMR (400 MHz, CDCl3)
δ 7.16-7.36 (m, 9H), 6.84 (d, J ) 8.8 Hz, 1H), 6.83 (d, J ) 8.8
Hz, 1H), 6.77 (d, J ) 8.8 Hz, 1H), 6.77 (d, J ) 8.8 Hz, 1H),
5.71,5.70 (d, J ) 0.8 Hz, 0.5H), 5.71 (d, J ) 0.8 Hz, 0.5H),
5.41 (d, J ) 4.4 Hz, 0.5H), 5.40 (d, J ) 4.4 Hz, 0.5H), 5.04 (s,
br, 0.5 H), 5.01 (s, br, 0.5 H), 3.82 (s, 1.5 H), 3.81 (s, 1.5 H),
3.60 (dd, J ) 8.4, 4.8 Hz, 1H), 2.57 (ddd, J ) 12.4, 4.6, 4.6 Hz,
1H), 2.16 (dd, J ) 12.4, 8.4 Hz, 1H); 13C NMR (100 MHz,
CDCl3) δ 159.85, 159.25, 155.92, 155.86, 149.46, 141.64, 141.48,
137.55, 137.37, 130.15, 129.87, 129.32, 129.13, 129.00, 127.43,
125.23, 124.46, 124.16, 122.37, 116.11, 115.96, 114.61, 114.46,
113.89, 84.58, 83.22, 60.91, 55.56, 55.53, 30.83; HRMS (ESI)
calcd for C25H22O6SH, 451.12152 (M + H+); found, 451.1228.
1
ether/hexanes (mp 111-112 °C); H NMR (500 MHz, CDCl3)
δ 7.20-7.35 (m, 15H), 5.77 (d, J ) 1.5 Hz, 1H), 5.47 (dd, J )
4.5, 0.8 Hz, 1H), 3.64 (dd, J ) 8.5, 4.5 Hz, 1H), 2.61 (ddd, J )
12.0, 4.5, 4.4 Hz, 1H), 2.21 (dd, J ) 12.0, 8.5 Hz, 1H); 13C NMR
(100 MHz, CDCl3) δ 149.46, 143.67, 139.56, 132.46, 131.69,
130.18, 129.25, 129.09, 128.85, 128.79, 127.81, 127.49, 127.46,
122.35, 84.62, 83.32, 60.69, 30.64; MS (EI) m/z 404 (M+, 28).
HRMS (EI) calcd for C24H20O4S, 404.10822; found, 404.1089.
5-Benzenesulfonyl-2,3-diphenyl-7-oxabicyclo[2.2.1]hept-
2-ene (12m). Compound 12m was purified by flash chroma-
tography (30% EtOAc/hexanes) to give a colorless needle (81%
yield) that was recrystallized from ether/hexanes (mp 138-
139 °C); 1H NMR (400 MHz, CDCl3) δ 7.98-8.00 (m, 2H),
7.67-7.71 (m, 1H), 7.58-7.62 (m, 2H), 7.21-7.29 (m, 8H),
7.11-7.13 (m, 2H), 5.66 (d, J ) 1.2 Hz, 1H), 5.34 (dd, J ) 4.4,
0.8 Hz, 1H), 3.47 (dd, J ) 8.4, 4.4 Hz, 1H), 2.43 (ddd, J ) 12.4,
3.6, 3.6 Hz, 1H), 1.99 (dd, J ) 12.0, 8.4 Hz, 1H); 13C NMR
(100 MHz, CDCl3) δ 143.78, 139.80, 138.93, 134.09, 132.65,
131.84, 129.59, 129.05, 129.03, 128.98, 128.69, 128.53, 127.72,
127.25, 83.92, 83.64, 65.32, 30.13; MS (EI) m/z 388 (M+, 10).
HRMS (EI) calcd for C24H20O3S 388.11331; found, 388.1125.
8,9-Diphenyl-4,10-dioxa-tricyclo[5.2.1.02,6]dec-8-ene-
3,5-dione (12n). Product 12n was not stable on the silica gel
and was obtained as a pale yellow solid (92%) that was
recrystallized from ethyl acetate and hexanes (mp 69-71 °C);
1H NMR (400 MHz, acetone-d6) δ 7.29-7.35 (m, 10H), 5.72 (s,
2H), 3.50 (s, 2H); 13C NMR (100 MHz, acetone-d6) δ 170.17,
141.83, 131.33, 129.32, 129.02, 127.55, 87.19, 50.0; MS (EI)
m/z 318 (M+, 6); HRMS (EI) calcd for C20H14O4, 318.0892;
found, 318.0884.
Synthesis of Ethenesulfonic Acid 4-Methoxyphenyl
Ester (16). A mixture of 4-methoxyphenol (2 g, 16.5 mmol)
dissolved in 3 mL of H2O and 3 mL of 1,2-dichloroethane was
stirred at 0 °C, and 11.4 mL of 25% aqueous NaOH and 2.7 g
of chloroethanesulfonyl chloride were added simultaneously
and slowly. The mixture was stirred for another 2 h at 0 °C
and then extracted with CH2Cl2. The combined organics were
washed with saturated NaCl, dried over Na2SO4, and filtered,
and the solvent was evaporated. Distillation under vacuum
1
gave pure 16 as a pale yellow oil in 57% yield; H NMR (400
MHz, CDCl3) δ 7.14 (d, J ) 9.2 Hz, 2H), 6.87 (d, J ) 9.2 Hz,
2H), 6.65 (dd, J ) 16.6, 10.0 Hz, 1H), 6.34 (d, J ) 17.2 Hz,
1H), 6.15 (d, J ) 10.0 Hz, 1H) 3.8 (s, 3H); 13C NMR (100 MHz,
CDCl3) δ 158.62, 142.99, 132.15, 131.98, 123.53, 114.94, 55.86.
Synthesis of Ethenesulfonic Acid 4-Hydroxyphenyl
Ester (17). Boron trifluoride/dimethyl sulfide (15 equiv) was
added slowly to a solution of ethenesulfonic acid 4-methoxy-
phenyl ester (2.34 mmol) in 5 mL of dry CH2Cl2 at 0 °C. The
solution was allowed to warm to room temperature and was
stirred for a total of 18 h. The reaction was quenched by the
addition of 10 mL of water and 5 mL of methanol, and the
solution was extracted with ethyl acetate (3 × 20 mL). The
combined organic phases were washed with saturated NaCl
and dried with Na2SO4. The crude product was purified by
4-[5(or
6)-Benzenesulfonyl-3-(4-methoxyphenyl)-7-
oxabicyclo[2.2.1]hept-2-en-2-yl]-phenol (12k, Mixture of
3:2 Isomers). Compound 12k was purified by preparative

7272 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Zhou et al.
flash chromatography using 30% EtOAc/hexanes as eluent,
which afforded the pure 17 as a pale yellow viscous oil in 75%
2.0 Hz, 1H), 5.61 (d, J ) 2.0 Hz, 1H), 3.82 (s, 3H); 13C NMR
(100 MHz, CDCl3) δ 163.08, 148.72, 148.47, 146.06, 142.94,
133.54, 133.25, 129.05, 128.87, 128.62, 128.59, 127.50, 127.34,
94.63, 89.29, 52.17.
General Procedure for the Suzuki Coupling Reaction.
A 25 mL flask was charged with Pd(PPh3)4 (0.01 mmol),
benzene (5 mL), 14 (0.1 mmol), and an aqueous solution of
Na2CO3 (0.1 mL of 2 M solution) under nitrogen atmosphere,
and then aryl boronic acid (0.15 mmol) in ethanol (1 mL) was
added. The mixture was stirred at 90 °C for 12 h. After the
reaction was complete, the mixture was extracted with ethyl
acetate, washed with saturated NaCl solution, and finally
dried over Na2SO4. The crude product was purified by pre-
parative thin-layer chromatography (30% EtOAc/hexanes).
3,5,6-Triphenyl-7-oxabicyclo[2.2.1]hepta-2,5-diene-2-
carboxylic Acid Methyl Ester (15a). Product 15a was
purified by preparative thin-layer chromatography (20% EtOAc/
hexanes) to give a colorless crystal (71% yield) that was
recrystallized from ether/hexanes (mp 122-123 °C); 1H NMR
(500 MHz, CDCl3) δ 7.45-7.50 (m, 3H), 7.23-7.32 (m, 12H),
6.09 (d, J ) 2.0 Hz, 1H), 5.94 (d, J ) 2.0 Hz, 1H), 3.83 (s, 3H);
13C NMR (125 MHz, CDCl3) δ 166.85, 164.61, 149.70, 146.64,
138.28, 133.97, 133.96, 132.46, 129.94, 129.15, 128.85, 128.73,
128.46, 128.39, 128.14, 127.99, 127.24, 92.97, 89.56, 51.86. MS
(EI) m/z 380 (M+, 12). HRMS (EI) calcd for C26H20O3, 380.14123;
found, 380.14130.
1
yield that solidified on standing; H NMR (400 MHz, CDCl3)
δ 7.06 (d, J ) 9.0 Hz, 2H), 6.79 (d, J ) 9.0 Hz, 2H), 6.64 (dd,
J ) 16.6, 10.0 Hz, 1H), 6.34 (d, J ) 16.6 Hz, 1H), 6.17 (d, J )
10.0 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 155.12, 142.75,
132.38, 131.90, 123.68, 116.56.
5,6-Bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-5-
ene-2-sulfonic Acid 4-Methoxyphenyl Ester (12o). Com-
pound 12o was purified by preparative thin-layer chromatog-
raphy (30% EtOAc/hexanes) to give a pale yellow solid (61%
yield) that was recrystallized from ethyl acetate/hexanes (mp
1
95-97 °C). H NMR (400 MHz, CDCl3) δ 7.17 (d, J ) 8.8 Hz,
2H), 7.15 (d, J ) 8.4 Hz, 2H), 7.12 (d, J ) 9.0 Hz, 2H), 6.82 (d,
J ) 8.8 Hz, 2H), 6.76 (d, J ) 8.4 Hz, 2H), 6.75 (d, J ) 8.4 Hz,
2H), 5.68 (d, J ) 1.2 Hz, 1H), 5.58 (br, s, 2H), 5.39 (dd, J )
4.5, 1.0 Hz, 1H), 3.78 (s, 3H), 3.56 (dd, J ) 8.2, 4.0 Hz, 1H),
2.56 (ddd, J ) 12.5, 6.3, 6.3 Hz, 1H), 2.14 (dd, J ) 12, 8.5 Hz,
1H); 13C NMR (100 MHz, CDCl3) δ 156.02, 155.93, 155.88,
155.46, 141.54, 137.49, 129.86, 129.74, 129.37, 128.98, 123.41,
123.37, 116.12, 115.97, 84.62, 83.18, 60.53, 55.86, 30.79; HRMS
(ESI) calcd for C25H22O7SH, 467.11644 (M + H+); found,
467.1171.
exo-5,6-Diphenyl-7-oxabicyclo[2.2.1]hept-5-ene-2-sul-
fonic Acid 4-Hydroxyphenyl Ester (12p). exo-12p was
purified by preparative thin-layer chromatography (30% EtOAc/
hexanes) to give a colorless needle (60% yield) that was
recrystallized from ether/hexanes (mp 59-60 °C); 1H NMR
(400 MHz, CDCl3) δ 7.27 (m, 10H), 7. 05 (d, J ) 8.8 Hz, 2H),
6.74 (d, J ) 8.4 Hz, 2H), 5.76 (d, J ) 1.2 Hz, 1H), 5.47 (dd, J
) 4.4, 1.2 Hz, 1H), 5.10 (s, br, 1H), 3.61 (dd, J ) 8.4, 4.4 Hz,
1H), 2.60 (ddd, J ) 12.0, 4.4, 4.4 Hz, 1H), 2.20 (dd, J ) 12.0,
8.4 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 154.62, 143.62,
142.93, 139.53, 132.43, 131.67, 129.25, 129.09, 128.87, 128.78,
127.84, 127.46, 123.62, 116.55, 84.64, 83.31, 60.35, 30.65;
HRMS (ESI) calcd for C24H20O5SH, 421.11096 (M + H+); found,
421.1109.
endo-5,6-Diphenyl-7-oxabicyclo[2.2.1]hept-5-ene-2-sul-
fonic Acid 4-Hydroxyphenyl Ester (12p). Product 12p was
purified by preparative thin-layer chromatography (30% EtOAc/
hexanes) to give a colorless viscous oil (15%); 1H NMR (400
MHz, CDCl3) δ 7.22-7.46 (m, 10H), 6.92 (d, J ) 8.8 Hz, 2H),
6.75 (d, J ) 9.0 Hz, 2H), 5.65 (dd, J ) 4.2, 1.0 Hz, 1H), 5.35
(dd, J ) 4.6, 1.0 Hz, 1H), 4.84 (s, br, 1H), 4.16 (ABX, J ) 9.2,
4.6, 4.2 Hz, 1H), 2.60 (ABX, J ) 12.0, 9.2, 4.6 Hz, 1H), 2.12
(dd, J ) 12.0, 4.6 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 54.63,
149.68, 142.08, 138.33, 132.74, 129.06, 128.65, 128.49, 128.37,
128.22, 128.18, 123.69, 116.51, 116.40, 84.77, 83.15, 59.07,
29.79; HRMS (ESI) calcd for C24H20O5SH, 421.11096 (M + H+);
found, 421.1110.
5,6-Bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-5-
ene-2-sulfonic Acid 4-hydroxyphenyl Ester (12q). Product
12q was purified by preparative thin-layer chromatography
(45% EtOAc/hexanes) to give a pale yellow solid (67% yield)
that was recrystallized from ether/hexanes (mp 110-112 °C);
1H NMR (400 MHz, acetone-d6) δ 8.66 (s, 1H), 8.64 (s, 1H),
8.60 (s, 1H), 7.24 (d, J ) 8.8 Hz, 2H), 7.22 (d, J ) 8.4 Hz, 2H),
7.07 (d, J ) 9.2 Hz, 2H), 6.84 (d, J ) 8.4 Hz, 2H), 6.80 (d, J )
8.8 Hz, 2H), 6.79 (d, J ) 8.8 Hz, 2H), 5.62 (d, J ) 1.2 Hz, 1H),
5.58 (br, s, 2H), 5.42 (dd, J ) 4.4, 0.8 Hz, 1H), 3.70 (dd, J )
8.4, 4.8 Hz, 1H), 2.40 (ddd, J ) 12.0, 4.4, 4.4 Hz, 1H), 2.25
(dd, J ) 12.0, 8.4 Hz, 1H); 13C NMR (100 MHz, acetone-d6) δ
157.76, 157.61, 156.45, 142.42, 141.52, 137.30, 129.50, 128.79,
124.41, 123.62, 123.54, 116.19, 115.91, 115.70, 84.61, 82.95,
60.26, 30.76; HRMS (ESI) calcd for C24H20O7SH, 453.1008 (M
+ H+); found, 453.1008.
3-Bromo-5,6-diphenyl-7-oxabicyclo[2.2.1]hepta-2,5-di-
ene-2-carboxylic Acid Methyl Ester (14). Furan 1a (0.2
mmol) was dissolved in benzene (5 mL), and methyl 3-bro-
mopropiolate 13 (0.26 mmol) was added. The reaction mixture
was stirred at 90 °C for 24 h. The crude product was purified
by flash chromatography (10-30% EtOAc/hexanes) to give a
pale yellow oil that was essentially pure; 1H NMR (400 MHz,
CDCl3) δ 7.43-7.45 (m, 2H), 7.25-7.38 (m, 8H), 6.01 (d, J )
3-(4-Hydroxyphenyl)-5,6-diphenyl-7-oxabicyclo[2.2.1]-
hepta-2,5-diene-2-carboxylic Acid Methyl Ester (15b).
Product 15b was purified by preparative thin-layer chroma-
tography (40% EtOAc/hexanes) to give a light-yellow crystal
(65% yield) that was recrystallized from ether/hexanes (mp
1
80-82 °C); H NMR (400 MHz, CDCl3) δ 7.52 (d, J ) 8.8 Hz,
2H), 7.45-7.48 (m, 2H), 7.25-7.29 (m, 8H), 6.74 (d, J ) 8.8
Hz, 2H), 6.09 (d, J ) 2.0 Hz, 1H), 5.96 (d, J ) 2.0 Hz, 1H),
3.84 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 166.74, 164.95,
157.53, 149.77, 146.33, 135.45, 133.97, 134.00, 131.26, 129.13,
128.71, 128.38, 128.10, 127.99, 127.22, 124.95, 115.47, 92.69,
89.47, 51.82. HRMS (ESI) calcd for C26H20O4H, 397.14397 (M
+ H+); found, 397.1434.
Estrogen Receptor Binding Affinity. Relative binding
affinities were determined by a competitive radiometric bind-
ing assay as previously described,32,33 using 10 nM [3H]-
estradiol as tracer ([6,7-3H]estra-1,3,5(10)-triene-3,17-â-diol,
51-53 Ci/mmol, Amersham BioSciences, Piscataway, NJ), and
purified full-length human ERR and ERâ were purchased from
PanVera/Invitrogen (Carlsbad, CA). Incubations were for 18-
24 h at 0 °C. Hydroxyapatite (BioRad, Hercules, CA) was used
to absorb the receptor-ligand complexes, and free ligand was
washed away. Assays with uterine cytosol preparations were
done in a related manner, as previously described,32 but
charcoal-coated dextran was used to adsorb free tracer. The
binding affinities are expressed as relative binding affinity
(RBA) values with the RBA of estradiol set to 100%. The values
given are the average ( range or SD of two to three indepen-
dent determinations. Estradiol binds to ERR and uterine
cytosol ER with a Kd of 0.2 nM and to ERâ with a Kd of 0.5
nM.
Gene Transcriptional Activity. Assays were performed
as previously described.37,41 Human endometrial cancer (HEC-
1) cells were maintained in minimum essential medium
(MEM) plus phenol red supplemented with 5% calf serum and
5% fetal calf serum. Cells were plated in phenol red-free
Improved MEM and 5% charcoal dextran-treated calf serum
(CDCS) and were given fresh medium 24 h before transfection.
Transfection assays were performed in 24-well plates using a
mixture of 0.35 mL of serum-free improved MEM medium and
0.15 mL of Hank’s balanced salt solution containing 5 µL of
lipofectin (Life Technologies, Inc., Gaithersburg, MD), 1.6 µg
of transferrin (Sigma, St. Louis, MO), 200 ng of pCMV
â-galactosidase as internal control, 1 µg of 2ERE-pS2-Luc, and
100 ng of ER expression vector per well. The cells were
incubated at 37 °C in a 5% CO2-containing incubator for 5 h.
The medium was then replaced with fresh Improved MEM
supplemented with 5% CDCS plus the desired concentrations

Oxabicyclic Core Estrogen Receptor Ligands
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7273
(2) Kuiper, G. G.; Enmark, E.; Pelto-Huikko, M.; Nilsson, S.;
Gustafsson, J. A. Cloning of a novel receptor expressed in rat
prostate and ovary. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 5925-
5930.
(3) Shughrue, P. J.; Lane, M. V.; Merchenthaler, I. Comparative
distribution of estrogen receptor-alpha and -beta mRNA in the
rat central nervous system. J. Comp. Neurol. 1997, 388, 507-
525.
(4) Shughrue, P. J.; Lane, M. V.; Scrimo, P. J.; Merchenthaler, I.
Comparative distribution of estrogen receptor-alpha (ER-alpha)
and beta (ER-beta) mRNA in the rat pituitary, gonad, and
reproductive tract. Steroids 1998, 63, 498-504.
(5) Couse, J. F.; Lindzey, J.; Grandien, K.; Gustafsson, J. A.; Korach,
K. S. Tissue distribution and quantitative analysis of estrogen
receptor-alpha (ERalpha) and estrogen receptor-beta (ERbeta)
messenger ribonucleic acid in the wild-type and ERalpha-
knockout mouse. Endocrinology 1997, 138, 4613-4621.
(6) Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Estrogen
receptor transcription and transactivation: Estrogen receptor
alpha and estrogen receptor beta: regulation by selective
estrogen receptor modulators and importance in breast cancer.
Breast Cancer Res. 2000, 2, 335-344.
(7) Shang, Y.; Brown, M. Molecular determinants for the tissue
specificity of SERMs. Science 2002, 295, 2465-2468.
(8) Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Biomedicine.
Defining the “S” in SERMs. Science 2002, 295, 2380-2381.
(9) Brzozowski, A. M.; Pike, A. C.; Dauter, Z.; Hubbard, R. E.; Bonn,
T.; et al. Molecular basis of agonism and antagonism in the
oestrogen receptor. Nature 1997, 389, 753-758.
(10) Top, S.; Vessieres, A.; Leclercq, G.; Quivy, J.; Tang, J.; et al.
Synthesis, biochemical properties and molecular modelling
studies of organometallic specific estrogen receptor modulators
(SERMs), the ferrocifens and hydroxyferrocifens: evidence for
an antiproliferative effect of hydroxyferrocifens on both hormone-
dependent and hormone-independent breast cancer cell lines.
Chem.sEur. J. 2003, 9, 5223-5236.
(11) Top, S. V., A.; Cabestaing, C.; Laios, I.; Leclercq, G.; Provot, C.;
Jaouen, G. Studies on organometallic selective estrogen receptor
modulators. (SERMs) Dual activity in the hydroxy-ferrocifen
series. J. Organomet. Chem. 2001, 637-639, 500-506.
(12) Plazuk, D.; Vessieres, A.; Le Bideau, F.; Jaouen, G.; Zakrzewski,
J. Synthesis of benzyl- and benzhydrylferrocenes via Friedel-
Crafts alkylation of ferrocene. Access to ferrocenyl bisphenols
with high affinities for estrogen receptors. Tetrahedron Lett.
2004, 45, 5425-5427.
(13) Endo, Y.; Iijima, T.; Yamakoshi, Y.; Fukasawa, H.; Miyaura, C.;
et al. Potent estrogen agonists based on carborane as a hydro-
phobic skeletal structure. A new medicinal application of boron
clusters. Chem. Biol. 2001, 8, 341-355.
of ligands. Cells were harvested 24 h later. Luciferase and
â-galactosidase activity were assayed as described.41
Molecular Modeling. The protein structure used in dock-
ing simulations was based on the X-ray crystallographic
structure of the human estrogen receptor ligand binding
domain bound to raloxifene (Protein Data Bank entry 1ERR).
The crystal structure contains one homodimer with 222
resolved residues, but only one monomer (chain A) was chosen
for modeling purposes. The crystal structure contains several
residues with missing atoms, and the flexible loops between
helices H9 and H10 (460-469) and H11 and H12 (529-534)
are completely missing. We used the molecular modeling
program Sybyl 7.0 (Tripos Inc., St. Louis, MO) for all manipu-
lations. Initially, we preformed a loop search of protein
fragments from the Protein Data Bank to insert into the
missing regions of the original X-ray structure. The optimal
loop was selected based on a high RMS fit onto three resolved
residues flanking the missing sequence and minimal van der
Waals contacts with the rest of the protein. Finally, missing
atoms were added to the ligand, protein, and all water
molecules within a 4 Å radius of the protein.
The ERR-RAL complex later used for docking studies was
minimized by a four-step procedure where each step used the
Powell algorithm along with the MMFF94s molecular force
field until convergence was reached (RMS < 0.05). First, the
inserted loops were minimized, while the rest of the protein
was fixed in space. Next, hydrogen atoms on the ligand,
protein, and water molecules were allowed to reach a mini-
mum energy. Third, the amino acid side chains were allowed
to minimize, except for Glu353 and Arg394, which remained
fixed along with the water molecules, the ligand, and the
protein backbone. Finally, all atoms were released and allowed
to reach a minimum energy. The minimized structure was
overlaid with the original X-ray structure and showed very
little movement of residues to ensure our model was reason-
able.
Raloxifene was deleted from the minimized ERR-RAL
complex, and SiteID was used to identify the ligand binding
site using a flood-fill solvation technique. FlexX was used
incrementally to construct one enantiomer of 12a in the
binding pocket then identify and rank the most favored binding
orientations in the ERR complex. The highest-ranked complex
was minimized using the last two steps in the minimization
procedure mentioned previously (see above).
(14) Valliant, J. F.; Schaffer, P.; Stephenson, K. A.; Britten, J. F.
Synthesis of boroxifen, a nido-carborane analogue of tamoxifen.
J. Org. Chem. 2002, 67, 383-387.
Acknowledgment. This work was supported through
grants from the National Institutes of Health (Grant
PHS 5R37 DK15556 to J.A.K. and Grant PHS 5R01
CA18119 to B.S.K.). We are grateful to Kathryn Carlson
for binding assays and for comments on the manuscript.
We thank Dr. Scott Wilson for performing the X-ray
crystallographic analysis of 12a. NMR spectra were
obtained in the Varian Oxford Instrument Center for
Excellence in NMR Laboratory. Funding for the instru-
mentation was provided in part by the W.M. Keck
Foundation, the National Institutes of Health (Grant
PHS 1 S10 RR104444-01), and the National Science
Foundation (Grant NSF CHE 96-10502). Mass spectra
were obtained on instruments supported by grants from
the National Institute of General Medical Sciences
(Grant GM 27029), the National Institutes of Health
(Grant RR 01575), and the National Science Foundation
(Grant PCM 8121494).
(15) Mull, E. S.; Sattigeri, V. J.; Rodriguez, A. L.; Katzenellenbogen,
J. A. Aryl cyclopentadienyl tricarbonyl rhenium complexes:
novel ligands for the estrogen receptor with potential use as
estrogen radiopharmaceuticals. Bioorg. Med. Chem. 2002, 10,
1381-1398.
(16) Yamakoshi, Y.; Otani, Y.; Fujii, S.; Endo, Y. Dependence of
estrogenic activity on the shape of the 4-alkyl substituent in
simple phenols. Biol. Pharm. Bull. 2000, 23, 259-261.
(17) Muthyala, R. S.; Sheng, S.; Carlson, K. E.; Katzenellenbogen,
B. S.; Katzenellenbogen, J. A. Bridged bicyclic cores containing
a 1,1-diarylethylene motif are high-affinity subtype-selective
ligands for the estrogen receptor. J. Med. Chem. 2003, 46, 1589-
1602.
(18) Muthyala, R. S.; Carlson, K. E.; Katzenellenbogen, J. A. Explo-
ration of the bicyclo[3.3.1]nonane system as a template for the
development of new ligands for the estrogen receptor. Bioorg.
Med. Chem. Lett. 2003, 13, 4485-4488.
(19) Mortensen, D. S.; Rodriguez, A. L.; Carlson, K. E.; Sun, J.;
Katzenellenbogen, B. S.; et al. Synthesis and biological evalu-
ation of a novel series of furans: ligands selective for estrogen
receptor alpha. J. Med. Chem. 2001, 44, 3838-3848.
(20) Fink, B. E.; Mortensen, D. S.; Stauffer, S. R.; Aron, Z. D.;
Katzenellenbogen, J. A. Novel structural templates for estrogen-
receptor ligands and prospects for combinatorial synthesis of
estrogens. Chem. Biol. 1999, 6, 205-219.
Supporting Information Available: Elemental analysis
results, HPLC results, HPLC spectra, and X-ray data for
compound 12a. This material is available free of charge via
the Internet at http://pubs.acs.org.
(21) Stauffer, S. R.; Coletta, C. J.; Tedesco, R.; Nishiguchi, G.;
Carlson, K.; et al. Pyrazole ligands: structure-affinity/activity
relationships and estrogen receptor-alpha-selective agonists. J.
Med. Chem. 2000, 43, 4934-4947.
(22) Stauffer, S. R.; Huang, Y.; Coletta, C. J.; Tedesco, R.; Katzenel-
lenbogen, J. A. Estrogen pyrazoles: defining the pyrazole core
structure and the orientation of substituents in the ligand
binding pocket of the estrogen receptor. Bioorg. Med. Chem.
2001, 9, 141-150.
References
(1) Mosselman, S.; Polman, J.; Dijkema, R. ER beta: identification
and characterization of a novel human estrogen receptor. FEBS
Lett. 1996, 392, 49-53.

7274 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Zhou et al.
(23) Sibley, R.; Hatoum-Mokdad, H.; Schoenleber, R.; Musza, L.;
Stirtan, W.; et al. A novel estrogen receptor ligand template.
Bioorg. Med. Chem. Lett. 2003, 13, 1919-1922.
(24) Hamann, L. G.; Meyer, J. H.; Ruppar, D. A.; Marschke, K. B.;
Lopez, F. J.; et al. Structure-activity relationships and subtype
selectivity in an oxabicyclic estrogen receptor alpha/beta agonist
scaffold. Bioorg. Med. Chem. Lett. 2005, 15, 1463-1466.
(25) Forgione, P.; Wilson, P. D.; Fallis, A. G. Magnesium mediated
carbometalation of propargyl alcohols: direct routes to furans
and furanones. Tetrahedron Lett. 2000, 41, 17-20.
(26) Dikshit, D. K.; Singh, S.; Singh, M. M.; Kamboj, V. P. Synthesis
and biological activity of 2,3- and 3,4-diarylfurans and 2,3,4-
triaryl-2,5-dihydrofurans. Indian J. Chem., Sect. B: Org. Chem.
Incl. Med. Chem. 1990, 29, 954-960.
(27) Brion, F. On the Lewis acid-catalyzed Diels-Alder reaction of
furan. Regio- and stereospecific synthesis of substituted cyclo-
hexenols and cyclohexadienols. Tetrahedron Lett. 1982, 23,
5299-5302.
(28) Klein, L. L.; Deeb, T. M. Mild and practical cycloaddition
reactions of furans with phenyl vinyl sulfonate. Tetrahedron Lett.
1985, 26, 3935-3938.
(29) Leroy, J. A convenient procedure for the preparation of 3-bro-
mopropiolic esters. Synth. Commun. 1992, 22, 567-572.
(30) Morris, J.; Wishka, D. G. Vinyl sulfonyl esters and amides in
the synthesis of substituted delta-sultams and delta-sultones.
J. Org. Chem. 1991, 56, 3549-3556.
(31) Pettersson, K.; Gustafsson, J. A. Role of estrogen receptor beta
in estrogen action. Annu. Rev. Physiol. 2001, 63, 165-192.
(32) Katzenellenbogen, J. A.; Johnson, H. J., Jr.; Myers, H. N.
Photoaffinity labels for estrogen binding proteins of rat uterus.
Biochemistry 1973, 12, 4085-4092.
(34) Berthois, Y.; Katzenellenbogen, J. A.; Katzenellenbogen, B. S.
Phenol red in tissue culture media is a weak estrogen: implica-
tions concerning the study of estrogen-responsive cells in culture.
Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 2496-2500.
(35) Bindal, R. D.; Carlson, K. E.; Katzenellenbogen, B. S.; Katzenel-
lenbogen, J. A. Lipophilic impurities, not phenolsulfonphthalein,
account for the estrogenic activity in commercial preparations
of phenol red. J. Steroid Biochem. 1988, 31, 287-293.
(36) Bindal, R. D.; Katzenellenbogen, J. A. Bis(4-hydroxyphenyl)[2-
(phenoxysulfonyl)phenyl]methane: isolation and structure elu-
cidation of a novel estrogen from commercial preparations of
phenol red (phenolsulfonphthalein). J. Med. Chem. 1988, 31,
1978-1983.
(37) Sun, J.; Meyers, M. J.; Fink, B. E.; Rajendran, R.; Katzenellen-
bogen, J. A.; et al. Novel ligands that function as selective
estrogens or antiestrogens for estrogen receptor-alpha or estro-
gen receptor-beta. Endocrinology 1999, 140, 800-804.
(38) Katzenellenbogen, J. A.; O’Malley, B. W.; Katzenellenbogen, B.
S. Tripartite steroid hormone receptor pharmacology: interac-
tion with multiple effector sites as a basis for the cell- and
promoter-specific action of these hormones. Mol. Endocrinol.
1996, 10, 119-131.
(39) Anstead, G. M.; Carlson, K. E.; Katzenellenbogen, J. A. The
estradiol pharmacophore: ligand structure-estrogen receptor
binding affinity relationships and a model for the receptor
binding site. Steroids 1997, 62, 268-303.
(40) Shiau, A. K.; Barstad, D.; Loria, P. M.; Cheng, L.; Kushner, P.
J.; et al. The structural basis of estrogen receptor/coactivator
recognition and the antagonism of this interaction by tamoxifen.
Cell 1998, 95, 927-937.
(41) McInerney, E. M.; Tsai, M. J.; O’Malley, B. W.; Katzenellenbo-
gen, B. S. Analysis of estrogen receptor transcriptional enhance-
ment by a nuclear hormone receptor coactivator. Proc. Natl.
Acad. Sci. U.S.A. 1996, 93, 10069-10073.
(33) Carlson, K. E.; Choi, I.; Gee, A.; Katzenellenbogen, B. S.;
Katzenellenbogen, J. A. Altered ligand binding properties and
enhanced stability of a constitutively active estrogen receptor:
evidence that an open pocket conformation is required for ligand
interaction. Biochemistry 1997, 36, 14897-14905.
JM0506773