Differential response of estrogen receptor subtypes to 1,3-diarylindene and 2,3-diarylindene ligands

Article abstract of DOI:10.1021/jm050226i

Estrogen receptors (ERs) control transcription of genes important for normal human development and reproduction. The signaling networks are complex, and there is a need for a molecular level understanding of the roles of receptor subtypes ERα and ERβ in normal physiology and as therapeutic targets. We synthesized two series of ER ligands, based on a common indene scaffold, in an attempt to develop compounds that can selectively modulate ER-mediated transcription. The 3-ethyl-1,2-diarylindenes, utilizing an amide linker for the 1-aryl extension, bind weakly to the ERs. The 2,3-diarylindenes bind with high affinity to the ER subtypes and demonstrate a range of different biological activities, both in transcriptional reporter gene assays and inhibition of estradiol-stimulated proliferation of MCF-7 cells. Several ligands differentiate between ERα and ERβ subtypes at an estrogen response element (ERE), displaying various levels of partial to full agonist activity at ERα, while antagonizing estradiol action at ERβ.

Full text of DOI:10.1021/jm050226i

J. Med. Chem. 2005, 48, 5989-6003
5989
Differential Response of Estrogen Receptor Subtypes to 1,3-Diarylindene and
2,3-Diarylindene Ligands
Nicola J. Clegg,†,‡ Sreenivasan Paruthiyil,|, Dale C. Leitman,|, and Thomas S. Scanlan†,‡,§,
*
Chemistry and Chemical Biology Graduate Program, Department of Pharmaceutical Chemistry, Department of Cellular and
Molecular Pharmacology, Department of Obstetrics, Gynecology, and Reproductive Sciences, and Center for Reproductive
Sciences, University of California, San Francisco, California 94143
Received March 11, 2005
Estrogen receptors (ERs) control transcription of genes important for normal human develop-
ment and reproduction. The signaling networks are complex, and there is a need for a molecular
level understanding of the roles of receptor subtypes ERR and ERâ in normal physiology and
as therapeutic targets. We synthesized two series of ER ligands, based on a common indene
scaffold, in an attempt to develop compounds that can selectively modulate ER-mediated
transcription. The 3-ethyl-1,2-diarylindenes, utilizing an amide linker for the 1-aryl extension,
bind weakly to the ERs. The 2,3-diarylindenes bind with high affinity to the ER subtypes and
demonstrate a range of different biological activities, both in transcriptional reporter gene assays
and inhibition of estradiol-stimulated proliferation of MCF-7 cells. Several ligands differentiate
between ERR and ERâ subtypes at an estrogen response element (ERE), displaying various
levels of partial to full agonist activity at ERR, while antagonizing estradiol action at ERâ.
Introduction
The estrogen receptor (ER) controls the transcription
of genes important for developmental, reproductive,
neural, skeletal, and cardiovascular processes. The
natural ligand, estradiol, modulates ER activity at
specific DNA response elements by binding to the two
ER subtypes, ERR and ERâ, which results in recruit-
ment of coregulatory complexes.1 Several compounds
have been developed as therapeutic agents to modulate
ER transcriptional activity for treatment of diseases
such as breast cancer and osteoporosis.2 Unlike estra-
diol, selective ER modulators (SERMs) display both
agonist and antagonist properties, depending on intrin-
sic cellular differences in processes or factors, such as
cell-signaling pathways, accessory cofactors, and tran-
scription factor modulatory proteins,3 as well as on the
DNA response element;4 however, the molecular details
for these selective ER modulating effects remains
unclear.
There is evidence that SERMs regulate ER-mediated
transcription by several different mechanisms,5 which
may involve the recruitment of distinct cofactor com-
plexes to the response element.6,7 The triphenylethylene
scaffold of tamoxifen is capable of producing a range of
tissue-specific effects that warrant further investigation
at a biological level.5,8 However, facile E/Z isomerization
via a tautomeric quinoid intermediate is encountered
in high-affinity derivatives containing the 4-hydroxy-
lated triphenylethylene structure, such as 4-hydroxy-
Figure 1. (a) Structures of prototypic SERMs. (b) Indenestrol
A is a metabolite of diethylstilbestrol. (c) Structures of the
* To whom correspondence should be addressed: Department of
Series I and Series II target indenes.
Pharmaceutical Chemistry, University of California San Francisco,
Genentech Hall, 600 16th Street Box 2280, San Francisco, CA 94143-
2280. Telephone: (415) 476-3620. Fax: (415) 502-7220. E-mail:
scanlan@cgl.ucsf.edu.
tamoxifen (OHT) (Figure 1a).9 The associated difficulties
in compound synthesis, purification, and biocharacter-
ization,5,10,11 prompted our search for a new scaffold that
was synthetically amenable to derivatization, for de-
velopment of novel SERMs.
† Chemistry and Chemical Biology Graduate Program.
‡ Department of Pharmaceutical Chemistry.
§ Department of Cellular and Molecular Pharmacology.
| Department of Obstetrics, Gynecology, and Reproductive Sciences.
Center for Reproductive Sciences.
10.1021/jm050226i CCC: $30.25 © 2005 American Chemical Society
Published on Web 08/19/2005

5990 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19
Clegg et al.
Scheme 1a
We desired a scaffold with a high likelihood of ER
activity, for which the pharmacology had not yet been
extensively explored, thereby maximizing the likelihood
of discovering compounds with novel properties. Several
derivatives have already been investigated to circum-
vent the problems with the 4-hydroxylated triphenyl-
ethylene scaffold.12-14 Precedence for our selection of an
indene-based scaffold arises from reports of indenestrol
A,15-18 a diethylstilbestrol (DES) metabolite which is
poorly uterotrophic,19 but has strong binding affinity for
ERR and acts as an agonist at an ERE.20 Limited SAR
studies of 1,3-dialkyl-2-arylindenes have highlighted the
importance of the C-3 stereochemistry and the phenolic
hydroxyls for binding and activity21-25 and demon-
strated a preference for an ethyl substituent at the
3-position of the indene ring.26 A limited series of basic
2,3-diarylindenes was examined for antifertility proper-
ties in rats during the 1960s,27,28 followed by analysis
of the binding orientation and fluorescent properties of
the 2,3-diarylindenes21,29-33 and limited SAR studies.34
Recently, basic 2-aryl-3-benzylindenes and -benzyl-
indanes have been investigated for their SERM effects
in bone and the cardiovascular system.35
a Reagents and conditions: (a) Lin(TMS)2, THF, -78 °C, p-
methoxybenzaldehyde; (b) (i) NaOH, EtOH, reflux, (ii) heat; (c) (i)
NaH, THF, (ii) 3-methoxybenzyle bromide, 50 °C; (d) BBr3, DCM,
-78 f 25 °C; (e) TlPSCl, imidazole, CH2Cl2.
Scheme 2a
The currently known SERMs, along with the indene
SAR data described above, provided a starting point for
synthesis of pathway-specific probes based on the uni-
fying diarylindene structure (Figure 1c). Aryl extensions
were built off the core indene scaffold at either the 1-
or the 3-position, analogous to the extensions of the
prototypical SERMs (Figure 1a). These extensions are
known to disrupt the conformation of the ERs by
displacing helix 12, thus preventing interaction with
coactivator proteins.36-38 The Series I indenes, incor-
porating an amide-linked aryl extension off carbon-1 of
the core scaffold, bind only weakly to the receptor,
whereas the 2,3-diarylindenes of Series II bind tightly
to the ERs. Subtle changes to the functional R-group of
Series II (Figure 1c) produce different biological activi-
ties, both in transcriptional reporter gene assays, and
inhibition of estradiol-stimulated proliferation of MCF-7
cells. Several of the Series II ligands can differentiate
between the ERR and ERâ subtypes at an ERE and
display various levels of partial to full agonist activity
at ERR, while antagonizing estradiol action at ERâ.
a Reagents and conditions: (a) (i) n-BuLi, THF, -78 f 25 °C,
(ii) CO2(s), (b) (i) HBTU, DMAP, CH2Cl2, (ii) (i-Pr2)EtN, 4-ami-
nophenol derivative, (c) Et3N-3HF, THF, (d) HCl, ether.
ing either a ketone or a direct carbon-carbon linker
between the aryl extension and the 1-position of the core
indene scaffold. Synthetic difficulties associated with the
high stability of the aromatic indenyl anion,45,46 pre-
vented derivatization of 4 to the requisite indenyl
bromide or boronic acid for Suzuki couplings. Attempts
at reaction of the indenyllithium with benzyl bromide
or an aryl Weinreb’s amide were also unsuccessful, and
coupling with benzaldehyde derivatives gave mixed
products. Changes to temperature or solvent polarity,
use of alternate metal cations, and addition of HMPA
had no effect on reactivity. However, conversion of
indene 4 to the indenyl carboxylate 6 was successful by
direct reaction of the anion with carbon dioxide (Scheme
2). Our plans for conversion of the indenyl carboxylate
6 to a Weinreb’s amide, for reaction with aryllithium
or arylmagnesium bromide derivatives, again proved
fruitless, due to unreactivity of the indenyl Weinreb’s
amide derivative. However, the ease of formation of the
amide by means of peptide coupling methods, convinced
us to redesign the compounds to target the amide-linked
diarylindenes of Series I.
Results and Discussion
Synthesis. The first series of indene ligands (Series
I) are based on the 3-ethyl-1,2-diarylindene structure
(Figure 1c). The 1-position of indenestrol A (Figure 1b)
was already known to accommodate an alkyl extension
of up to at least four carbons, without loss in binding
affinity,26,39 indicating a prime target site for derivati-
zation with aryl-based extensions. Because the target
indene structure contains a stereocenter at C-1, all com-
pounds were prepared as racemic mixtures.38,40-42 The
core 3-ethyl-2-arylindene scaffold (4) was synthesized
by Lewis-acid induced cyclodehydration of a ketone
intermediate (3), following adaptations to a published
synthetic route43,44 (Scheme 1). Boron tribromide was
used for the cyclodehydration of 3, because of its che-
lating effect, which sterically prevented formation of
undesirable isomers obtained using polyphosphoric acid.
Aniline derivatives were coupled to the indenyl car-
boxylate (6) using HBTU with catalytic DMAP (Scheme
2). Deprotection of the silyl ethers to give the target
indenes of Series I was accomplished with triethyl-
amine-trihydrogen fluoride, rather than tetrabutyl-
ammonium fluoride, which resulted in degradation. In
the case of the indenes with the basic side-chains (NC-
2, NC-4), this meant that the final products were
We initially envisaged the use of organometallic
coupling methods to synthesize target compounds hav-

Differential Response of Estrogen Receptor Subtypes
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19 5991
Scheme 3a
Scheme 5a
a Reagents and conditions: (a) MgX-PhR (X ) Br or Cl), THF,
(b) HCl, (c) BBr3, DCM.
a Reagents and conditions: (a) (i) n-BuLi, ether, -78 °C f 0
°C, (ii) EtI, (iii) TBAF, THF, (b) Et3N‚3HF, THF.
to the acid chloride, followed by Friedel-Crafts acyla-
tion to give indanone 11. Because the fused aryl ring
contains an activating group, ring closure is directed to
the 6-position, resulting in formation of only the desired
isomer. Indanone 11 can be derivatized using a variety
of aryl Grignard reagents (Schemes 4-6). Subsequent
acid workup and deprotection with boron tribromide
yields the target Series II indenes, which display a
variety of functional groups, differing in the charge,
electron-withdrawing effects, geometry, and polarity.
The indene products must be protected from air, humid-
ity, and light, due to facile air-oxidation resulting in
formation of 2,3-diarylindenones.29
Scheme 4a
The synthesis of NC-8 through NC-16 is depicted in
Schemes 4 and 5. The synthesis of NC-17, N-18, N-19,
and NC-20 proceeded via a methyl ester intermediate
(18), which was synthesized from triflate 17 via pal-
ladium-catalyzed alkoxycarbonylation49 (Scheme 6).
Reaction of ester 18 with boron tribromide resulted in
saponification and deprotection of the methyl ethers to
give the carboxylic acid derivative NC-17. Reduction of
the ester 18 with DIBAL-H produced the benzyl alcohol
19. Treatment of 19 with boron tribromide not only
deprotected the methyl ethers, but also converted the
alcohol to a bromide (NC-18), presumably as the result
of excess hydrogen bromide. We were unable to effect
selective deprotection of the phenolic methyl ethers of
ester 18, to produce NC-20. However, addition of
methanol to a solution of the acid (NC-17) and lithium
aluminum hydride resulted in formation of the methyl
ester (NC-20). The amide NC-19 was produced by
amination of ester 18 with trimethylaluminum-am-
monium chloride.
Estrogen Receptor Binding Affinity. All target
compounds were tested for binding affinity to full-length
ERR and ERâ in a competition assay with a fixed
concentration of fluorescent estradiol ligand, using
fluorescence polarization as a read-out. Results are
shown in Table 1. Both indenes NC-1 (ethyl indenestrol
A) and NC-3 (indenestrol A), which lack an aryl exten-
sion, show binding affinities for ERR and ERâ that are
comparable to, or better than, OHT. This is in general
agreement with previous studies for ERR.26 Although
NC-7 also lacks an aryl extension, it contains a charged
carboxylate group, which may explain its reduced af-
finity for both ER subtypes (only 10% of the affinity of
OHT).
Of note, is that the 3-ethyl-1,3-diarylindenes of Series
I that contain the amide-linked 1-aryl extension (NC-
2, -4, and -5), all show poor binding affinity for both ER
subtypes, less than 1% of the affinity of OHT. In
contrast, the 2,3-diarylindenes of Series II (NC-8 through
NC-20) show fair binding to both ERR and ERâ, on the
order of the affinity of raloxifene, and on average about
a Reagents and conditions: (a) KHMDS, THF, 3-methoxybenzyl
bromide, (b) (i) oxalyl chloride, CH2Cl2, (ii) AlCl3, CH2Cl2, (c) THF,
Grignard reagent, (d) HCl, (e) NDMBA, Pd(PPh3)4, CH2Cl2, (f)
Et3N‚3HF, THF, (g) BBr3, DCM.
isolated as the HCl salts. In addition to NC-3, which
was prepared during the synthesis of the core indene
scaffold, several other indenes lacking a bulky aryl
extension were prepared from intermediates in the
synthetic pathway (Scheme 3).
Series II indenes consist of a 2,3-diarylindene struc-
ture, in which the aryl extension is directly attached to
carbon-3 of the core scaffold (Figure 1c). The general
synthesis via an indanone intermediate 11 is based on
previously developed routes27,28 (Scheme 4). Literature
methods cite various Lewis acids as effective in the
cyclization of the acid;27,28,47,48 however, we found the
most effective route to be conversion of carboxylate 10

5992 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19
Clegg et al.
Scheme 6a
a Reagents and conditions: (a) 4-(triisopropylsiloxy)phenylmagnesium bromide, THF, (b) Et3N‚3HF, THF, (c) Tf2O, 2,6-lutidine, CH2Cl2,
0 f 25 °C, (d) CO, Pd(OAc)2/dppp. DMF, MeOH, Et3N, (e) AlMe3-NH4Cl, benzene, (f) BBr3, CH2Cl2, -78 f 25 °C, (g) DIBAL-H, THF, (h)
LiAlH4, THF, MeOH.
10-fold weaker than OHT. The major exception is NC-
17 that has a negatively charged carboxylate group, and
therefore very poor affinity for the ERs as a result of
unfavorable interactions with the receptor (1000-fold
worse than OHT).
The weak affinity of the Series I indenes was unan-
ticipated and suggests that the amide linkage may
interact unfavorably with residues in the ligand binding
pocket. Interestingly, research on the related spiroin-
dene ligand scaffold has also found that a carbonyl
linker was detrimental to binding affinity, despite
modeling predictions to the contrary.50
With the exception of NC-3, none of the indene
ligands showed significant binding preference for either
of the ER subtypes. NC-3 has a 9-fold preference in
binding affinity for ERâ in this assay, only slightly
higher than the 5-fold ERâ preference shown by estra-
diol. This increased affinity of NC-3 for ERâ may be due
to its lack of a functional group at the 1-position of the
indene ring, the absence of which may avoid potentially
unfavorable steric interactions with the Met354 residue
in the ERâ binding pocket. However, modeling studies
are necessary to further validate this hypothesis.
ER Subtype-Selective Ligand Activity at an
ERE. All the indene ligands were tested for their effects
on ER-mediated transcription in U2OS cells using high
throughput reporter gene assays. A luciferase reporter
gene construct regulated by a consensus ERE (the
vitellogenin A2 ERE-tk promoter) was used. On the
basis of the level of efficacy achieved in the dose-
response assays, the ligands were classified as ‘full
agonists’ (efficacy > 60%) or ‘partial agonists’ (10-59%
efficacy). Ligands that were able to abolish the agonist
activity of estradiol in competition assays, were termed
‘full antagonists’ (inhibit to basal activity levels). Ligands
defined as partial agonists can, in competition assays
with estradiol, display some partial antagonist proper-
ties (inhibit down to the levels of partial activation).
In dose-response assays, the ligands without an aryl
extension, NC-1 and NC-3, were potent agonists with
both ER subtypes at an ERE (EC50s in the range of 0.1
to 0.3 nM) (Table 2). NC-7 showed lower potency (EC50
) 50-160 nM), as a result of its weaker binding affinity.
The poor affinity of the Series I indenes (NC-2, -4, and
-5), translated into very weak agonist activity at ERR/
ERE in dose-response assays (EC50 ) 100-430 nM).
Despite the bulky aryl extensions of these ligands, the
weak interactions with the receptor must be insufficient
to induce either an optimal agonist conformation, or an
optimal antagonist conformation with full displacement
of helix 12.36-38 NC-4, with the dimethylamino exten-
sion analogous to OHT, is actually a full agonist of both
ERR and of ERâ, albeit at very high concentrations
(EC50 ) 150 nM for ERR and 500 nM for ERâ). NC-2
and NC-5 have little or no effect on ERâ, although high
concentrations of NC-2 do result in weak antagonism
of estradiol in competition assays (Ki ) 147 nM).
The most interesting activities are seen with the
Series II 2,3-diarylindenes. These ligands display a wide
range of activities at both ERR and ERâ and in several
cases are able to differentiate between the receptor
subtypes, acting as agonists at one subtype and antago-
nists at the other (Table 2). In agreement with the
binding data, the agonist potencies or EC50s of most of
the Series II ligands are on the order of 4-17 nM, which
is 100-fold less potent than E2, and the antagonist
potencies or Kis, range from 0.2 to15 nM which are 10-
100 fold weaker than OHT. However, these potencies
are sufficient for the ligands to be effective ER agonists
and antagonists in vitro. NC-17 is the only Series II
indene ligand that is virtually inactive for both ERs,
due to its weak affinity for the receptors.

Differential Response of Estrogen Receptor Subtypes
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19 5993
Table 1. Relative Binding Affinities of the Series I and Series II Indene Ligands for Full Length ERR and ERâ, Determined by a
Competitive Binding Assay Kit with Fluorescently Labeled Estradiol, Using Methods Described in the Experimental Section
a Relative binding affinities are expressed as a percentage of the potency of 4-hydroxytamoxifen, found to have a Ki of 5.7 nM for ERR
and 1.7 nM for ERâ.
Dose-response assays determine NC-14 to be a full
agonist at both ERR and ERâ. This is the only ligand
with a methylene-linked phenyl extension. The di-
rectly linked phenyl equivalent, NC-8, displays full
agonist activity at ERR and partial agonist activity at
ERâ. All the other Series II indenes show virtually no
agonist activity at ERâ, and competition experiments
versus a fixed estradiol concentration show them to
be partial to full antagonists at ERâ. The activity of
these same ligands at ERR ranges from full agonism
(NC-11, -13), to partial agonism (NC-9, -12, -15, -16,
-18), to full antagonism (NC-10, -19, -20). This suggests

5994 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19
Clegg et al.
Table 2. Transcriptional Activity of Series I and Series II Indenes on ERR and ERâ at a Classical ERE in U2OS Cells. Transient
Transfection Assays Were Performed as Described in the Experimental Section Using ERR and ERâ Expression Vectors and an
Estrogen-Regulated Luciferase Reporter Gene Plasmid
a Values represent the average of multiple experiments for which all data points were determined in triplicate. Agonist assays are
done with test compound alone; antagonist assays are done with compound together with 0.1 nM estradiol. EC50s and Kis were calculated
with Prism software analysis tools from dose-response and competition curves, as described in the Experimental Section. Compounds
are classified as either full agonists (efficacy > 60%), full antagonists versus 0.1 nM estradiol (inhibit to basal activity levels), or partial
agonists (10-59% efficacy) that can also display partial antagonist properties (inhibit down to the levels of partial activation in competition
assays). Refer to text for details.
that ERâ is more sensitive to ligand structure than ERR,
and that it can convert from an agonist to antagonist
conformation more readily.51 For example, the addition
of a phenolic hydroxyl to NC-8 converts it from an ERR
full agonist/ERâ partial agonist into an ERR partial
agonist/ERâ antagonist (NC-9). Interestingly, a similar
conversion is seen with 3-ethyl-2,4,5-tris(4-hydroxyphe-
nyl)furan, an ERR-selective furan ligand that is inactive
at ERâ, but converts to an ERâ agonist merely by
removal of the 5-phenolic hydroxyl.52
The above results demonstrate that very small changes
to the indene ligand structure can have drastic effects
on the transcriptional activity of the ER subtypes at
an ERE. The indene ligand scaffold has provided ac-
cess to a range of subtype-selective ER ligands. There
is no distinct correlation between the properties of the

Differential Response of Estrogen Receptor Subtypes
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19 5995
ligand (R,R-THC), which acts as an agonist on ERR at
ERE but an antagonist on ERâ.53 SAR data exploring
the role of substituent size and stereochemistry in
determining agonist versus antagonist activity of the
tetrahydrochrysenes, determined that all the THC
derivatives were agonists on ERR, and those with small
substituents were also agonists on ERâ.38 Of note is that
the induction of an antagonist conformation in ERâ was
achieved with less steric perturbation than in ERR, as
is also the case with the Series II indene ligands.
There are three Series II indene ligands that are full
antagonists of E2 activity at both ERR and ERâ. NC-10
has an N-piperidinyl extension group analogous to
raloxifene and is likely to function in the same way as
the prototypical SERMs, by displacing helix 12 as a
result of the bulky extension group, thereby disrupting
the coactivator binding interface.36-38 On the other
hand, NC-19 and NC-20 have much smaller aryl
extensions, containing an amide and a methyl ester,
respectively. In this respect, they are similar to R,R-
THC, which functions as an antagonist on ERâ but has
a structure that is very different from the bulky side-
chain extensions of the typical antiestrogens tamoxifen
and raloxifene.53 The crystal structures of R,R-THC
bound to the ERR and ERâ ligand binding domains
provide an interesting insight into its novel mechanisms
for selectivity.54 In the case of ERR, R,R-THC is buried
within the ligand binding pocket, however, for ERâ,
helix 12 is unable to close over the opening to the pocket
and adopts an antiestrogenic conformation, despite the
lack of a bulky side-chain as is the case for the other
SERMs. It appears that R,R-THC stabilizes a nonpro-
ductive conformation of key residues within the binding
pocket, that allosterically act to prevent progression to
an active agonist complex. This has been termed
‘passive antagonism’.54 We believe that the interactions
in the receptor binding pocket with the carbonyl func-
tionality of the extensions of NC-19 and NC-20 may
result in a similar allosteric effect on the receptor
conformation, which translates into a form of ‘passive
antagonism’.54,55 We hypothesize that the subtype selec-
tive effects of the Series II ligands on ER transcriptional
activity, result from differences between the interactions
of the ligand functional groups with residues within the
ligand binding pocket. This in turn leads to differences
in the allosteric communication with distal secondary
structural interactions, both within the ER, and with
other interacting proteins, thereby influencing cofactor
recruitment.55
Figure 2. Antiproliferative effect of Series II indene ligands
on estradiol-stimulated MCF-7 cells. Cells were treated with
0.01 nM estradiol alone (open bars), with test compound alone
(double cross-hatched bars), or with both 0.01 nM estradiol
and test compound, and incubated for 7 days, with replacement
of media and hormones after day 3. The extent of proliferation
was quantified by incorporation of tritiated thymidine, and
results were reported as fold proliferation compared to the no
hormone control.
functional groups and the efficacies or potencies of
activation at ERR. For example, the methyl (NC-15)
and fluoride (NC-11) derivatives are more efficacious
than the chloride (NC-12) and trifluoromethyl deriva-
tives (NC-15). The ethyl bromide (NC-18) shows lower
overall potencies, which may be due to covalent modi-
fication of lysines or other electron-donating residues.
In addition, there may be differences in the cell-
permeability of the indenes, which may also contribute
to apparent differences in potency. Indeed, NC-8, -9, -10,
-16, -18, -19, and -20 were soluble only to 10 mM in
ethanol stock solution (NC-16 and NC-18 were spar-
ingly so), whereas other ligands of Series II were soluble
at 100 mM.
Antiproliferative Effects in MCF-7 Cells. To study
the pharmacological activity of the Series II indene
ligands in MCF-7 breast cancer cells, we screened a
representative subset of ligands, to determine whether,
at saturating concentrations, they could block the 4-fold
proliferative effect induced by 0.01 nM estradiol. The
extent of proliferation was quantified by incorporation
of tritiated thymidine into the newly synthesized DNA
of the dividing cells. MCF-7 cells express ERR at high
levels; thus it follows that the ligands that show most
potent antagonism at ERR/ERE in reporter gene assays,
namely NC-9, -16, and -20, are likely to be most
effective at preventing E2-stimulated proliferation in
these cells. As expected, both prototypic SERMs, OHT
and raloxifene, inhibited proliferation completely at
saturating levels of ligand (Figure 2). Saturating levels
of NC-10 and NC-19 also resulted in close to complete
inhibition. NC-9, -16, and -20 induced 2-fold prolifera-
tion of cells upon ligand treatment in the absence of E2.
Therefore, in the presence of E2, the 4-fold proliferative
effect of E2 was inhibited by 2-fold. The remaining
indene ligands demonstrated no antiproliferative effects.
These results indicate that the data from reporter gene
assays do correlate to ligand effects in a more physi-
ological context using breast cancer cells that express
endogenous receptors.
Conclusion
Using a common indene scaffold, we have developed
two series of indene ligands with a range of selective
activities at the ER subtypes. This subtype selectivity
is particularly striking, because it results from only very
small changes in the ligand structure. Several functional
groups were explored in the side-chain of this series in
order to refine their selectivity and to gain insight into
how the chemical structure of a ligand relates to the
effect on ER mediated signaling. Previously character-
ized SERMs such as raloxifene possess a bulky basic
side-chain which displaces helix 12 of the receptor,
disrupting the interaction of coactivators necessary for
transcriptional activation. The indene ligands do not
The selective ERR agonist and ERâ antagonist prop-
erties of the Series II indenes (NC-11, -13, -9, -12, -15,
-16, and -18) are similar to the R,R-tetrahydrochrysene

5996 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19
Clegg et al.
per cuvette, together with 5 µg of the reporter plasmid and
2.5 µg of the ERR or ERâ expression vector, determined as
optimal for the ERE response. The electroporation buffer
consisted of 0.2 µm filtered PBS and 0.1% glucose. Cells were
transfected by electroporation at a potential of 0.25 kV and a
capacitance of 960 mF. Transfected cells were pooled and
immediately resuspended in assay medium. The assay medium
was prepared by the UCSF Cell Culture facility and consisted
of 1 µm-filtered DME/F12 Ham’s media 1:1 without phenol red,
with 15 mM HEPES and L-glutamine (Sigma #D-2906),
supplemented with 1.338 g/L NaHCO3, as well as with 100
mg/L streptomycin sulfate, 100 units/mL of penicillin G, and
10% heat-inactivated and hormone-depleted newborn calf
serum. This is serum that had been incubated at 56 °C for 30
min to inactivate complement, followed by treatment with
dextran-coated charcoal as described previously.4 Cells were
plated into 96-well Costar assay plates (#3917, opaque flat
bottom with lid, tissue culture treated) at 100 µL assay
medium per well, and at a density of approximately 1.5 × 104
cells per well for ERR, and 1.75 × 104 cells per well for ERâ.
After 6 to 8 h of incubation at 37 °C, to allow for adherence of
the cells, the media was removed, and fresh assay media was
added to the plates, along with 10-fold serial dilutions of
hormones, prepared in triplicate. The maximum concentration
of ethanol vehicle in all cases was 1% of the total volume. In
the case of the dose-response experiments, the serial dilutions
were done in the tissue culture plates, taking care not to
disturb the cells. In the case of the competition experiments,
a constant concentration of estradiol was required, so the serial
dilutions of test ligand were done on a separate plate and
transferred over to the wells of plated cells. Controls were
performed in all cases for no hormone response (ethanol vehicle
alone) and for maximal activation (10-7 M estradiol). In the
case of competition experiments, additional controls were
have the same large bulky side-chain and would appear
to disrupt the receptor conformation in a more subtle
way, by means of ‘passive antagonism’.
Several of the Series II indenes show various levels
of agonist activity at ERR, while possessing full antago-
nist activity at ERâ. In particular, NC-13 is a full
agonist at ERR, whereas it fully antagonizes activation
by estradiol at ERâ. These indenes can be used to
investigate the mechanisms of ligand-mediated tran-
scriptional activation or repression of ER target genes
and to study the biological roles of ERR and ERâ.
Understanding the molecular events that convert a drug
from being predominantly antiestrogenic to being es-
trogenic could potentially open the door to a better
understanding of the molecular mechanism of SERM
action, and ultimately, the use of our Series II selective
indene ligands in a combination of structural, biochemi-
cal, cellular, and physiological studies could enable a
greater understanding of how ligand-receptor confor-
mations relate to estrogen agonist or antagonist behav-
ior. This has significant implication for the design of
more effective drugs in the treatment of diseases such
as breast cancer and osteoporosis.
Experimental Section
Materials and Methods: Molecular Biology. The con-
struction of the expression vectors for both hERR (HEG0) and
hERâ have been previously described.4,56 The wild-type ERR
(HEG0) and the longer form of ERâ (ERâ1) with 149 residues
in the N-terminal domain were used in transfection experi-
ments. The ERE-driven luciferase reporter gene consists of two
repeats of the upstream region of the vitellogenin ERE
promoter from -331 to -289, followed by region -109 to +45
upstream of the thymidilate kinase, followed by the luciferase
gene.4,5,56
performed for maximal antagonism of estradiol (10-7
4-hydroxytamoxifen).
M
After 12 to 24 h of incubation at 37 °C, the cells were lysed
by first removing the media from the wells, washing with 100
µL/well of PBS (Ca2+- and Mg2+-free) and then adding 40 µL/
well of passive lysis buffer (Promega #E194A) at room tem-
perature. The plates were incubated for 20 to 30 min on a
shaker and then combined with reconstituted luciferase assay
buffer (Promega Luciferase Assay System #E1501) at room
temperature. Luminescence was immediately measured for 0.1
s/well on an Analyst AD plate-reader (LJL Biosystems), with
prior shaking of the plate (5 s) to ensure mixing of the
reagents.
In dose-response experiments, the transcriptional activa-
tion with increasing amounts of added ligand yielded an EC50
value, indicative of the potency of ER activation at the ERE.
Antagonist properties were measured in competition assays,
where an increasing amount of ligand is added to compete
away a constant level of 10-10 M E2 agonist, to calculate an
IC50 value, which is then converted to a Ki. Each hormone dose
was performed in triplicate, and the relative error was
determined by calculating the standard error of the three
values from the mean. The curves were fit with the Prism
statistical analysis software package, using either the sigmoi-
dal dose-response curve for dose-response experiments, or
the single binding site competition model for competition
experiments. The R2 values in all cases were greater than 0.8.
Experiments were conducted multiple times to ensure repro-
ducibility of the results, and the standard error of the mean
(SEM) was less than 0.2 log units from the average log EC50
or log IC50 value in all cases. EC50 or Ki values were reported
as averages across experiments along with SEM.
ER Binding Assays. The relative binding affinities of
compounds for full length ERR and ERâ were determined
using ERR and ERâ competitor assay kits (green), according
to the manufacturer’s instructions (PanVera Corp, Madison,
WI). Fluorescence polarization was used as a read-out, and
the experiments were performed in 96-well plates (Costar
black polystyrene round-bottom assay plates #3792) and read
(five readings per well; 0.1 s integration time; entire plate read
eight times; G factor ) 0.91) using the Analyst AD plate-reader
(LJL Biosystems) with fluorescein excitation (485 nM) and
emission (530 nM) filters. Each hormone dose was performed
in triplicate, and the relative error was determined by calcu-
lating the standard error of the three values from the mean.
In all cases we controlled for minimal competition (ethanol
vehicle alone), for no ER, for no fluorescent ligand, and for
maximal competition (10-5 M E2). The binding curves were fit
using a single binding site competition model, with the Prism
statistical analysis software package. The R2 values in all cases
were greater than 0.8. Experiments were conducted multiple
times to ensure reproducibility of the results, and the standard
error of the mean (SEM) was less than 0.2 log units from the
average logIC50 value in all cases. Ki values were reported as
averages across experiments along with SEM. Percent relative
binding affinity (RBA) was then determined by dividing the
Ki determined for unlabeled 4-hydroxytamoxifen by the test
ligand Ki and then multiplying the result by 100.
Tissue Culture, Transfection and Luciferase Assays.
U2OS human osteosarcoma cells were grown in 0.1 µm filtered
DME/H21 growth medium, supplemented with 4.5 g/L glucose,
0.876 g/L glutamine, 100 mg/L streptomycin sulfate, 100 units/
mL of penicillin G, and 10% newborn calf serum. Cells were
grown to a confluence of no more than 80%. For transient
transfection assays, cells were trypsinized, washed with PBS,
and resuspended in 0.5 mLof electroporation buffer in 0.4 cm
gap electroporation cuvettes at approximately 1.8 × 106 cells
Cell Proliferation Assays. Assays to determine the anti-
proliferative effects of selected indene test ligands on estradiol-
stimulated MCF-7 human breast cancer cells were performed
according to published procedures.57 Briefly, cells were treated
with both 10-11 M estradiol and a saturating concentration of
test ligand and incubated for 7 days, with replacement of
media and hormones after day 3. The extent of proliferation
was quantified by incorporation of tritiated thymidine into the

Differential Response of Estrogen Receptor Subtypes
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19 5997
newly synthesized DNA of the dividing cells, and results were
reported as fold proliferation compared to the no hormone
control.
CO3 followed by dilute NaCl solution, and dried over anhy-
drous MgSO4, and solvent was removed under reduced pres-
sure. The crude product was purified by flash column
chromatography (silica, 0%-25% EtOAc-hexanes) to give (()-
3-ethyl-N-(4-hydroxyphenyl)-6-triisopropylsiloxy-2-(4-triisopro-
pylsiloxyphenyl)-1H-indene-1-carboxamide as a white solid
(66.8 mg, 0.095 mmol) in 39% yield. Rf 0.47 (15% EtOAc-
1
Materials and Methods: Chemistry. H and 13C NMR
spectra were obtained on a Varian Model AS 400 (400 MHz)
instrument. 1H NMR chemical shifts are reported as δ values
in ppm downfield from internal tetramethylsilane or downfield
from residual H2O peak in CD3OD. 13C NMR chemical shifts
are reported as δ values with reference to the solvent peak.
High-resolution mass spectrometry (HRMS) using electrospray
ionization (EI) was performed by the National Bio-Organic,
Biomedical Mass Spectrometry Resource at the University of
California, San Francisco. In instances that fast atom bom-
bardment (FAB) was required, HRMS was performed by the
Mass Spectrometry Facility at the University of California,
Berkeley. Flash column chromatography on crude products
was performed using 230-400 mesh silica gel (Aldrich Chemi-
cal Co.) or 40-63 µm EMD Geduran Silica Gel 60 (VWR
International). Preparative TLC (pTLC) was run using glass
precoated silica gel plates (250 µm, particle size 5 to 17 µm,
pore size 60 Å, 20 × 20 cm; Aldrich Chemical Co. Z12272-6).
Purity of all compounds was determined by TLC using
commercial silica gel plates (Alltech, Alugram Sil B/UV 254)
and by 1H NMR and HRMS. Developed TLC plates were
visualized using short wave UV light (254 nm) and were
typically stained using a p-anisaldehyde solution followed by
heating with a heat gun. For final compounds, purity was
assessed as 95% pure or greater unless otherwise noted, by
analytical reverse phase high-pressure liquid chromatography
(RP-HPLC). This was done using an Alliance 2695 Separations
Module with a Waters 2996 photodiode array detector, together
with two different columns: a XTerra RP18 Column (3.5 µm;
4.6 × 50 mm) and a YMC-Pack Pro C4 Column (5-3 µm, 12
nm; 4.6 × 50 mm). The following solvent gradient was used
with a 1 mL/min flow rate: water:acetonitrile (0.05% trifluo-
roacetic acid) 90:10 to 5:95 over 10 min. Retention times for
the two different column conditions differed at least by
approximately 1 min, for each compound analyzed (see Sup-
porting Information).
1
hexanes); H NMR (400 MHz, CDCl3) δ 7.27 (d, J ) 8.79 Hz,
2 H), 7.17 (m, 3 H), 6.86 (d, J ) 8.79 Hz, 2 H), 6.80 (d, J )
8.79 Hz, 2 H), 6.61 (s, 1 H), 6.52 (d, J ) 8.79 Hz, 2 H), 5.52 (s,
1 H), 4.66 (s, 1 H), 2.71 (m, 2 H), 1.22 (m, 9 H), 1.03 (s, 18 H),
1.01 (s, 18 H); 13C NMR (100 MHz, CDCl3) 169.10, 155.48,
155.03, 153.06, 143.54, 143.06, 138.17, 135.81, 129.74, 129.11,
128.29, 122.98, 120.46, 120.11, 119.19, 115.94, 115.50, 30.31,
19.70, 17.94, 17.89. 13.58, 12.68, 12.64 ppm; HRMS (FAB)
exact mass calculated for C42H61NO4Si2: 699.4139, found:
699.4126.
Deprotection. To a solution of (()-3-ethyl-N-(4-hydroxy-
phenyl)-6-triisopropylsiloxy-2-(4-triisopropylsiloxyphenyl)-1H-
indene-1-carboxamide (60.5 mg, 0.086 mmol) in THF (3 mL)
at 25 °C was added triethylamine trihydrofluoride (42.26 µL,
0.259 mmol), and the reaction was stirred for 4 h 30 min. The
solution was poured into a 1 M sodium fluoride, extracted with
diethyl ether, washed with saturated NaHCO3, and dried over
anhydrous MgSO4, and solvent was removed under reduced
pressure. The crude yellowish-white solid was purified by
pTLC (silica, 15% CH3OH-CHCl3) and eluted off silica with
85% 2-propanol-CH2Cl2 to give NC-5 as a solid (26.5 mg, 0.053
1
mmol) in 62% yield. Rf 0.45 (10% CH3OH-CHCl3); H NMR
(400 MHz, CD3OD) δ 9.51 (s, 1 H), 7.22 (d, J ) 8.79 Hz, 2 H),
7.14 (d, J ) 8.30 Hz, 1 H), 7.08 (d, J ) 8.79 Hz, 2 H), 6.92 (s,
1 H), 6.74 (d, J ) 8.79 Hz, 2 H), 6.71 (d, J ) 5.86 Hz, 1 H),
6.60 (d, J ) 8.79 Hz, 2 H), 4.63 (s, 1 H), 2.61 (m, 2 H), 1.21 (m,
3 H); 13C NMR (100 MHz, CD3OD) 171.85, 157.61, 156.91,
155.63, 145.54, 143.36, 139.43, 138.17, 131.27, 130.63, 128.93,
123.81, 121.18, 116.30, 116.15, 115.24, 111.69, 60.77, 20.37,
13.83 ppm; HRMS (FAB) exact mass calculated for C24H21-
NO4: 387.1471, found: 387.1467.
(()-N-(4-(2-(Piperidinyl)ethoxy)phenyl)-3-ethyl-6-hy-
droxy-2-(4-hydroxyphenyl)-1H-indene-1-carboxamide
Monohydrochloride (NC-2). Coupling. To a solution of
carboxylate 6 (231.5 mg, 0.380 mmol) in CH2Cl2 (7.5 mL) at 0
°C were added HBTU (158.6 mg, 0.418 mmol) and 4-(dimeth-
ylamino)pyridine (2.32 mg, 0.019 mmol). The reaction was
stirred at 0 °C for 30 min, followed by 25 °C for 30 min. Aniline
8 (92.12 mg, 0.418 mmol) and diisopropylethylamine (132 µL,
0.76 mmol) were added, and the reaction was stirred at 25 °C
for 1 h. The deep orange-red solution was poured into a 5%
HCl solution, extracted with CH2Cl2, washed with saturated
NaHCO3, and dried over anhydrous MgSO4, and solvent was
removed under reduced pressure. The crude product was
purified by flash column chromatography (silica, 0%-10% CH3-
OH-CHCl3) to give (()-N-(4-(2-(piperidinyl)ethoxy)phenyl)-3-
ethyl-6-triisopropylsiloxy-2-(4-triisopropylsiloxyphenyl)-1H-
indene-1-carboxamide monohydrochloride as a pale yellowish-
white solid (273.7 mg, 0.34 mmol) in 89% yield. Rf 0.67 (20%
CH3OH-CHCl3); 1H NMR (400 MHz, CDCl3) δ 7.36 (d, J )
8.30 Hz, 2 H), 7.29 (s, 1 H), 7.23 (d, J ) 8.30 Hz, 1 H), 7.02 (d,
J ) 8.79 Hz, 2 H), 6.94 (d, J ) 8.30 Hz, 2 H), 6.90 (d, J ) 8.30
Hz, 1 H), 6.77 (s, 1 H), 6.72 (d, J ) 8.79 Hz, 2 H), 4.76 (s, 1 H),
3.99 (m, 2 H), 2.78 (m, 2 H), 2.68 (m, 2 H), 2.45 (s, 4 H), 1.57
(m, 4 H), 1.39 (m, 2 H), 1.35 (m, 3 H), 1.27 (m, 6 H), 1.11 (s, 18
H), 1.10 (s, 18 H); 13C NMR (100 MHz, CDCl3) 168.21, 155.47,
155.26, 154.80, 143.18, 143.01, 137.98, 135.74, 130.34, 128.95,
128.22, 121.93, 120.24, 119.89, 118.93, 115.77, 114.46, 65.98,
60.17, 57.66, 54.81, 25.71, 23.98, 19.52, 17.78, 17.72, 13.43,
12.51, 12.46 ppm; LRMS (EI) exact mass calculated for
C49H75N2O4Si2: 811.5265, found: 811.666.
Glassware was oven or flame-dried prior to use, and
reactions were performed under argon inert atmosphere that
was passed through a Drierite drying tube. CH2Cl2, diethyl
ether, DMF, CH3OH, THF, and toluene were dried using the
procedure recommended by Grubbs, using the solvent purifica-
tion system manufactured by Glass Contour, Inc. (Laguna
Beach, CA).58 Prior to the installation of this system, anhy-
drous solvents were purchased from Aldrich Chemical Co. and
in some instances, THF (anhydrous) was additionally purified
by distillation from sodium metal. All other reagents were
purchased from Aldrich Chemical Co. and were used without
further purification. The term ‘solvent was removed under
reduced pressure’ generally implies rotary evaporation, fol-
lowed by use of a high-vacuum pump. N,N-Diallyl-4-bro-
mobenzenamine (14) was prepared according to the published
procedure,59 and 2-(4-hydroxyphenyl)-1-phenyl-3H-inden-5-ol
(NC-8) was prepared from ketone 11 following published
methods.27,28 Improved syntheses of the core indene scaffold
62-64
(NC-3),33,43,47,60,61 and indanone 11
were achieved via
adaptations to published procedures (see Supporting Informa-
tion). Syntheses of the final test compounds are described
below, whereas preparation and characterization of intermedi-
ates is provided in the online supplement (see Supporting
Information).
(()-3-Ethyl-6-hydroxy-N,2-bis(4-hydroxyphenyl)-1H-
indene-1-carboxamide (NC-5). Coupling. To a solution of
carboxylate 6 (150 mg, 0.246 mmol) in CH2Cl2 (2 mL) at 0 °C
were added HBTU (102.76 mg, 0.271 mmol) and 4-(dimethy-
lamino)pyridine (1.5 mg, 0.0123 mmol). The reaction was
stirred at 0 °C for 50 min, followed by 25 °C for 30 min.
4-Aminophenol (29.57 mg, 0.271 mmol) and diisopropyleth-
ylamine (128.7 µL, 0.739 mmol) were added dropwise, and the
solution became a dark red which dissipated over time. The
reaction was stirred at 25 °C for 17 h, poured into a 5% HCl
solution, extracted with CH2Cl2, washed with saturated NaH-
Deprotection. To a solution of (()-N-(4-(2-(piperidinyl)-
ethoxy)phenyl)-3-ethyl-6-triisopropylsiloxy-2-(4-triisoprop-
ylsiloxyphenyl)-1H-indene-1-carboxamide monohydrochloride
(273.7 mg, 0.337 mmol) in THF (7.5 mL) at 25 °C was added
triethylamine trihydrofluoride (549.9 µL, 3.37 mmol), and the
reaction was stirred for 3 h. The reaction was poured into a

5998 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19
Clegg et al.
separatory funnel, and the two organic layers were separated.
The lower layer was saved, solvent was removed under reduced
pressure, and the residue was resuspended in diethyl ether
(2.5 mL). Excess HCl (2 M in diethyl ether) (3.7 mL) was
added, and the solution was stirred at 25 °C for 2 h, to form
the HCl salt, whereupon solvent was removed under reduced
pressure. The crude product was redissolved in CH3OH,
purified by repeated recrystallization (CH3OH-EtOAc), fil-
tered, and washed with cold EtOAc to give a 1:12 ratio of
product:triethylamine as a yellow crystalline solid (313.5 mg,
0.143 mmol) corresponding to a 43% yield of pure product
NC-2 (71.4 mg, 0.143 mmol): Rf 0.37 (40% CH3OH-CHCl3);
1H NMR (400 MHz, CD3OD) δ 10.20 (s, 1 H), 7.42 (d, J ) 7.33
Hz, 2 H), 7.33 (d, J ) 8.30 Hz, 2 H), 7.20 (d, J ) 7.81 Hz, 1 H),
7.01 (s, 1 H), 6.91 (d, J ) 8.30 Hz, 2 H), 6.82 (d, J ) 8.30 Hz,
2 H), 6.80 (m, 1 H), 4.93 (s, 1 H), 4.32 (m, 2 H), 3.53 (m, 4 H),
2.67 (m, 4 H), 1.80 (m, 6 H), 1.46 (m, 3 H); 13C NMR (100 MHz,
CD3OD) 171.61, 157.61, 156.91, 155.85, 145.66, 142.88, 139.44,
138.40, 133.68, 130.60, 128.80, 123.20, 121.15, 116.33, 115.93,
115.24, 111.80, 63.65, 60.38, 57.12, 54.87, 23.96, 22.49, 20.38,
13.93 ppm; HRMS (FAB) exact mass calculated for C31H35-
N2O4: 499.2597, found: 499.2607.
63.40, 60.73, 57.83, 44.00, 20.34, 13.80 ppm; HRMS (FAB)
exact mass calculated for C28H31N2O4: 459.2284, found:
459.2282.
(()-1,3-Diethyl-2-(4-hydroxyphenyl)-3H-inden-5-ol (eth-
yl indenestrol A, NC-1). Coupling. To a solution of indene
(4) (50 mg, 0.088 mmol) in diethyl ether (1.2 mL) at -78 °C
was added n-BuLi (2.5 M in hexanes) (38.9 µL, 0.0973 mmol)
dropwise. The clear yellow solution was stirred at -78 °C for
10 min, followed by 0 °C for 10 min, and the color became
orange, then bright yellow. The solution was then cooled to
-78 °C, iodoethane (7.8 µL, 0.973 mmol) was added dropwise,
and the solution was stirred at -78 °C for 1 h and then allowed
to warm slowly to 25 °C and stirred 24 h. The solution was
poured into water, extracted with diethyl ether, and dried over
anhydrous MgSO4, and solvent was removed under reduced
pressure. The crude product was purified by pTLC (silica, 5%
EtOAc-hexanes) to give a light orange oil (29.9 mg, 0.123
mmol) containing a 2:1 mixture of starting material to (()-
1,3-diethyl-2-(4-triisopropylsiloxyphenyl)-5-triisopropylsiloxy-
3H-indene product, which corresponds to a 38% yield of
product (29.9 mg, 0.033 mmol), deemed 87% pure by HPLC.
1
Rf 0.63 (5% EtOAc-hexanes); H NMR (400 MHz, CDCl3) δ
7.03 (m, 7 H), 3.80 (m, 1 H), 2.70 (m, 2 H), 1.87 (m, 1 H), 1.63
(m, 1 H), 1.23 (m, 9 H), 1.13 (s, 18 H), 1.11 (s, 18 H), 0.42 (t,
J ) 7.32 Hz, 3 H); the sample was too dilute for adequate
analysis by 13C NMR; HRMS (EI) exact mass calculated for
C37H60O2Si2: 592.4132, found: 592.4127.
(()-N-(4-(2-(Dimethylamino)ethoxy)phenyl)-3-ethyl-
6-hydroxy-2-(4-hydroxyphenyl)-1H-indene-1-carbox-
amide monohydrochloride (NC-4). Coupling. To a solution
of carboxylate 6 (231.5 mg, 0.380 mmol) in CH2Cl2 (7.5 mL)
at 0 °C were added HBTU (158.6 mg, 0.418 mmol) and
4-(dimethylamino)pyridine (2.32 mg, 0.019 mmol). The reac-
tion was stirred at 0 °C for 30 min, followed by 25 °C for 45
min. Aniline 9 (75.34 mg, 0.418 mmol) and diisopropylethy-
lamine (132 µL, 0.76 mmol) were added, and the reaction was
stirred at 25 °C for 1 h. The solution was poured into a 5%
HCl solution, extracted with CH2Cl2, washed with saturated
NaHCO3, and dried over anhydrous MgSO4, and solvent was
removed under reduced pressure. The crude product was
purified by flash column chromatography (silica, 0%-5% CH3-
OH-CHCl3) to give the pure (()-N-(4-(2-(dimethylamino)-
ethoxy)phenyl)-3-ethyl-6-triisopropylsiloxy-2-(4-triisopropyl-
siloxyphenyl)-1H-indene-1-carboxamide monohydrochloride
product (238.9 mg, 0.309 mmol) in 81% yield. Rf 0.71 (20% CH3-
Deprotection. To a partly pure solution of (()-1,3-diethyl-
2-(4-triisopropylsiloxyphenyl)-5-triisopropylsiloxy-3H-indene
(19.8 mg, 0.033 mmol) in THF (1.5 mL) at 25 °C was added
tetrabutylammonium fluoride (1 M in THF) (66 µL, 0.066
mmol). The solution immediately turned yellow, blue, and then
maroon and was stirred for 15 min and then quenched with
water. The organics were extracted with diethyl ether, washed
with saturated NaHCO3, and dried over anhydrous MgSO4,
and solvent was removed under reduced pressure. The crude
product was purified by pTLC (silica, 2.5% CH3OH-CHCl3)
to give NC-1 as a light yellow oil (2.0 mg, 0.0071 mmol) in
1
22% yield. Rf 0.27 (5% CH3OH-CHCl3); H NMR (400 MHz,
CD3OD) δ 7.09 (d, J ) 8.79 Hz, 2 H), 7.06 (d, J ) 7.81 Hz, 1
H), 6.83 (s, 1 H), 6.76 (d, J ) 8.79 Hz, 2 H), 6.64 (d, J ) 7.81
Hz, 1 H), 3.72 (m, 1 H), 2.54 (m, 2 H), 1.82 (m, 1 H), 1.55 (m,
1 H), 1.16 (m, 3 H), 0.32 (t, J ) 7.32 Hz, 3 H); 13C NMR (100
MHz, CD3OD) 157.22, 156.16, 149.73, 142.60, 140.03, 139.40,
130.88, 129.66, 120.33, 116.18, 114.05, 111.67, 52.23, 24.16,
20.16, 14.44, 8.24 ppm; HRMS (EI) exact mass calculated for
C19H20O2: 280.1463, found: 280.1470.
1
OH-CHCl3); H NMR (400 MHz, CDCl3) δ 7.37 (d, J ) 8.79
Hz, 2 H), 7.29 (s, 1 H), 7.23 (d, J ) 8.30 Hz, 1 H), 7.04 (d, J )
8.79 Hz, 2 H), 6.94 (d, J ) 8.79 Hz, 2 H), 6.90 (d, J ) 8.30 Hz,
1 H), 6.73 (d, J ) 8.79 Hz, 2 H), 4.76 (s, 1 H), 3.95 (m, 2 H),
2.77 (m, 2 H), 2.65 (t, J ) 5.86 Hz, 2 H), 2.28 (s, 6 H), 1.35 (m,
3 H), 1.26 (m, 6 H), 1.11 (s, 18 H), 1.10 (s, 18 H); 13C NMR
(100 MHz, CDCl3) 168.17, 155.39, 155.19, 154.71, 143.10,
143.01, 137.98, 135.75, 130.41, 128.91, 128.22, 121.84, 120.16,
119.84, 118.85, 115.68, 114.38, 65.91, 57.95, 45.55, 38.31,
19.46, 17.73, 17.67, 13.37, 12.46, 12.41 ppm; HRMS (FAB)
exact mass calculated for C46H71N2O4Si2: 771.4952, found:
771.4955.
(()-3-Ethyl-6-hydroxy-2-(4-hydroxyphenyl)-1H-indene-
1-carboxylic Acid (NC-7). To a solution of carboxylate 6 (83.0
mg, 0.136 mmol) in THF (3 mL) at 25 °C was added trieth-
ylamine trihydrofluoride (222 µL, 1.362 mmol), and the reac-
tion was stirred for 3 h 45 min. The solution was poured into
a 1 M sodium fluoride, acidified with 1 N HCl, extracted with
diethyl ether, washed with 0.5 M HCl, and dried over
anhydrous MgSO4, and solvent was removed under reduced
pressure. The crude brown solid was purified by pTLC (silica,
10% CH3OH-CHCl3) and eluted off silica with diethyl ether,
85% 2-propanol-CH2Cl2, and 5% CH3OH-CHCl3 to give a solid
(40.3 mg, 0.136 mmol) in 100% yield, deemed 61% pure by
HPLC. Rf 0.54 (20% CH3OH-CHCl3); 1H NMR (400 MHz, CD3-
OD) δ 7.18 (d, J ) 7.81 Hz, 2 H), 7.04 (d, J ) 8.30 Hz, 1 H),
6.92 (s, 1 H), 6.68 (d, J ) 7.81 Hz, 2 H), 6.63 (d, J ) 8.30 Hz,
1 H), 4.52 (s, 1 H), 2.55 (m, 2 H), 1.15 (m, 3 H); 13C NMR (100
MHz, CD3OD) 157.28, 156.50, 145.85, 141.81, 139.53, 139.03,
130.61, 129.69, 120.62, 120.62, 116.05, 114.74, 111.86, 49.64,
20.32, 13.80 ppm; HRMS (FAB) exact mass calculated for
C18H16O4: 296.1049, found: 296.1051.
Deprotection. To a solution of (()-N-(4-(2-(dimethyl-
amino)ethoxy)phenyl)-3-ethyl-6-triisopropylsiloxy-2-(4-triisoprop-
ylsiloxyphenyl)-1H-indene-1-carboxamide (87.5 mg, 0.113 mmol)
in THF (4 mL) at 25 °C was added triethylamine trihydro-
fluoride (73.97 µL, 0.45 mmol), and the reaction was stirred
for 3 h. The solvent was removed under reduced pressure, and
the residue was redissolved in CH3OH (1.5 mL). Excess HCl
(2 M in diethyl ether) (1.5 mL) was added and the solution
stirred at 25 °C for 2 h to form the HCl salt. The crude product
was purified by repeated trituration (CH3OH-EtOAc) to give
a 1:2 ratio of product:triethylamine as a yellowish-white
precipitate (87.1 mg, 0.113 mmol) corresponding to a 100%
yield of pure NC-4 product (61.1 mg, 0.113 mmol): Rf 0.16
(20% CH3OH-CHCl3); 1H NMR (400 MHz, CD3OD) δ??con-
centration dependent? 6.79 (d, J ) 8.30 Hz, 2 H), 6.73 (d, J )
8.30 Hz, 2 H), 6.63 (d, J ) 8.30 Hz, 1 H), 6.43 (s, 1 H), 6.30 (d,
J ) 8.30 Hz, 2 H), 6.25 (d, J ) 7.81 Hz, 2 H), 6.24 (m, 1 H),
4.20 (s, 1 H), 3.64 (m, 2 H), 2.85 (m, 2 H), 2.28 (m, 6 H), 2.09
(m, 2 H), 0.70 (m, 3 H); 13C NMR (100 MHz, CD3OD) 171.95,
157.61, 156.90, 155.84, 145.41, 143.36, 139.42, 138.11, 133.73,
130.60, 128.83, 123.47, 121.19, 116.25, 115.90, 115.22, 111.68,
1,2-Bis(4-hydroxyphenyl)-3H-inden-5-ol (NC-9). To a
solution of methyl 4-(6-methoxy-2-(4-methoxyphenyl)-1H-in-
27,28
den-3-yl)phenol
(31.3 mg, 0.091 mmol) in CH2Cl2 (3 mL)
at -78 °C was added boron tribromide (1 M in CH2Cl2) (300
µL, 0.3 mmol). The reaction mixture was stirred at -78 °C for
1 h, followed by further addition of boron tribromide (1 M in
CH2Cl2) (1 mL, 1 mmol). The solution was then stirred at 25

Differential Response of Estrogen Receptor Subtypes
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19 5999
°C for 3 h. The solution was poured into water, stirred 10 min,
and extracted with CH2Cl2. The aqueous layer was basified
with 1 N NaOH and extracted with CH2Cl2. The combined
organic layers were washed with saturated NaHCO3, and
EtOAc was used to dissolve the UV-active precipitate off the
sides of the glassware. The organic layers were combined, and
dried over anhydrous MgSO4, and solvent was removed under
reduced pressure. The crude oil was purified by pTLC (silica,
20% CH3OH-CH2Cl2) and eluted off silica with 10% CH3OH-
CH2Cl2 to give NC-9 as an orange oil (24.2 mg, 0.076 mmol)
in 84% yield, and deemed 81% pure by HPLC. Rf 0.48 (15%
CH3OH-CH2Cl2); 1H NMR (400 MHz, CD3OD) δ 7.04 (m, 4
H), 6.88 (s, 1 H), 6.84 (d, J ) 8.30 Hz, 1 H), 6.77 (d, J ) 8.30
Hz, 2 H), 6.59 (dd, J ) 8.30, 1.95 Hz, 1 H), 6.53 (d, J ) 8.79
Hz, 2 H), 3.66 (s, 2 H); 13C NMR (100 MHz, CD3OD) 157.71,
157.14, 156.38, 145.33, 140.94, 138.68, 131.53, 130.23, 129.96,
129.14, 121.18, 121.18, 116.59, 115.84, 114.11, 112.20, 41.62
ppm; HRMS (EI) exact mass calculated for C21H16O3: 316.1099,
found: 316.1096.
1-(4-(2-(Piperidin-1-yl)ethoxy)phenyl)-2-(4-hydroxy-
phenyl)-3H-inden-5-ol (NC-10). Coupling. To a solution of
bromide 12 (498 mg, 1.753 mmol) in THF (4 mL) were added
Mg turnings (39.4 mg, 1.753 mmol) and a catalytic amount of
iodine. The reaction mixture was refluxed for 4 h, until
disappearance of the Mg metal indicated formation of the
Grignard reagent. The solution was allowed to cool and was
added to a solution of ketone 11 (150 mg, 0.559 mmol) in THF
(2 mL). The reaction mixture was refluxed for 20 h overnight.
The solution was poured into 1.5N HCl, extracted with EtOAc
and then CH2Cl2, and washed with water. The aqueous layer
was basified with 1 N NaOH, extracted with EtOAc and CH2-
Cl2, and washed with water. The organic layers were combined
and dried over anhydrous MgSO4, and solvent was removed
under reduced pressure. The crude oil was purified by flash
column chromatography (silica, 0-2% CH3OH-CHCl3) to give
1-(2-(4-(6-methoxy-2-(4-methoxyphenyl)-1H-inden-3-yl)phenoxy)-
ethyl)piperidine as a red semisolid (181.5 mg, 0.398 mmol) in
71% yield. Rf 0.48 (15% CH3OH-CHCl3); 1H NMR (400 MHz,
CDCl3) δ 7.25 (d, J ) 7.81 Hz, 2 H), 7.18 (d, J ) 8.30 Hz, 2 H),
7.07 (d, J ) 11.72 Hz, 1 H), 7.06 (s, 1 H), 6.93 (d, J ) 7.81 Hz,
2 H), 6.78 (d, J ) 8.79 Hz, 1 H), 6.71 (d, J ) 8.79 Hz, 2 H),
4.12 (t, J ) 5.86 Hz, 2 H), 3.79 (s, 3 H), 3.76 (s, 2 H), 3.72 (s,
3 H), 2.78 (m, 2 H), 2.51 (m, J ) 4.88 Hz, 4 H), 1.60 (m, 4 H),
1.44 (m, 2 H); 13C NMR (100 MHz, CDCl3) 158.07, 157.80,
143.74, 140.41, 137.85, 137.40, 130.27, 129.24, 128.96, 128.53,
120.22, 114.74, 114.41, 113.42, 111.77, 109.94, 65.76, 57.88,
55.40, 54.95, 54.87, 40.81, 25.79, 24.07 ppm; HRMS (EI) exact
mass calculated for C30H33NO3: 455.2460, found: 455.2476.
130.27, 129.81, 123.40, 121.13, 115.89, 114.15, 112.27, 66.22,
58.84, 55.91, 41.70, 26.35, 24.93 ppm; HRMS (EI) exact mass
calculated for C28H29NO3: 427.2147, found: 427.2149.
1-(4-Fluorophenyl)-2-(4-hydroxyphenyl)-3H-inden-5-
ol (NC-11). Coupling. To a solution of ketone 11 (150 mg,
0.559 mmol) in THF (4 mL) at 25 °C was added 4-fluoro-
phenylmagnesium bromide (1 M in THF) (615 µL, 0.615 mmol).
The yellow solution was stirred at 25 °C for 2 h, poured into
1.5 N HCl, and extracted with diethyl ether, followed by CH2-
Cl2, the organic layers were washed with water, combined, and
dried over anhydrous MgSO4, and solvent was removed under
reduced pressure. The crude oil was purified by flash column
chromatography (silica, 0%-10% EtOAc-hexanes) to give 3-(4-
fluorophenyl)-6-methoxy-2-(4-methoxyphenyl)-1H-indene as a
pinkish orange crystalline solid (144.5 mg, 0.417 mmol) in 75%
yield. Rf 0.63 (40% EtOAc-hexanes); 1H NMR (400 MHz,
CDCl3) δ 7.30 (dd, J ) 8.30, 5.86 Hz, 2 H), 7.15 (d, J ) 8.79
Hz, 2 H), 7.09 (m, 2 H), 7.03 (d, J ) 8.30 Hz, 1 H), 7.03 (d, J
) 8.30 Hz, 1 H), 6.80 (dd, J ) 8.30, 2.44 Hz, 1 H), 6.73 (d, J
) 8.79 Hz, 2 H), 3.82 (s, 3 H), 3.80 (s, 2 H), 3.75 (s, 3 H); 13C
NMR (100 MHz, CDCl3) 158.39, 158.05, 143.82, 140.16, 138.85,
136.84, 132.34, 131.03, 130.95, 129.12, 120.16, 115.89, 115.68,
113.63, 111.96, 110.17, 55.55, 55.12, 41.03 ppm; HRMS (EI)
exact mass calculated for C23H19FO2: 346.1369, found:
346.1353.
Deprotection. To a solution of 3-(4-fluorophenyl)-6-meth-
oxy-2-(4-methoxyphenyl)-1H-indene (144.5 mg, 0.398 mmol)
in CH2Cl2 (4 mL) at -78 °C was added boron tribromide (1 M
in CH2Cl2) (834 µL, 0.834 mmol). The reaction mixture was
stirred at -78 °C for 2 h and then at 25 °C for a further 2 h,
followed by further addition of CH2Cl2 (4 mL) and boron
tribromide (1 M in CH2Cl2) (1 mL, 1 mmol). The semisolid
mixture was stirred at 25 °C for a further 26 h overnight. The
solution was poured into water, sonicated extensively to
dissolve the solids, extracted with CH2Cl2, and washed with
saturated NaHCO3, and solvent was removed under reduced
pressure. The crude oil was purified by flash column chroma-
tography (silica, 0%-40% EtOAc-hexanes), followed by pTLC
(40% EtOAc-hexanes), and eluted off silica with EtOAc to give
the pure NC-11 (62.1 mg, 0.195 mmol) in 49% yield. Rf 0.33
(40% EtOAc-hexanes); 1H NMR (400 MHz, CD3OD) δ 7.10
(m, 2 H), 6.96 (m, 2 H), 6.89 (d, J ) 8.79 Hz, 2 H), 6.82 (d, J
) 2.44 Hz, 1 H), 6.72 (d, J ) 8.30 Hz, 1 H), 6.54 (dd, J ) 8.30,
1.95 Hz, 1 H), 6.47 (d, J ) 8.30 Hz, 2 H), 3.55 (s, 2 H); 13C
NMR (100 MHz, CD3OD) 157.33, 156.48, 145.32, 140.32,
139.91, 137.52, 132.33, 132.25, 130.32, 129.38, 120.94, 116.62,
116.41, 115.93, 114.18, 112.34, 41.71 ppm; HRMS (EI) exact
mass calculated for C21H15FO2: 318.1056, found: 318.1056.
1-(4-Chlorophenyl)-2-(4-hydroxyphenyl)-3H-inden-5-
ol (NC-12). Coupling. To a solution of ketone 11 (150 mg,
0.559 mmol) in THF (4 mL) at 25 °C was added 4-chlorophen-
ylmagnesium bromide (1 M in diethyl ether) (615 µL, 0.615
mmol). The orange solution was stirred at 25 °C for 2 h, poured
into 1.5 N HCl, and extracted with diethyl ether, followed by
CH2Cl2, the organic layers were washed with water, combined,
and dried over anhydrous MgSO4, and solvent was removed
under reduced pressure. The crude oil was purified by flash
column chromatography (silica, 0%-5% EtOAc-hexanes) to
give 3-(4-chlorophenyl)-6-methoxy-2-(4-methoxyphenyl)-1H-
indene as a pinkish white solid (131.3 mg, 0.3626 mmol) in
65% yield. Rf 0.57 (40% EtOAc-hexanes); 1H NMR (400 MHz,
CDCl3) δ 7.36 (m, 2 H), 7.26 (d, J ) 8.30 Hz, 2 H), 7.14 (d, J
) 8.79 Hz, 2 H), 7.07 (s, 1 H), 7.02 (d, J ) 8.30 Hz, 1 H), 6.79
(dd, J ) 8.30, 1.95 Hz, 1 H), 6.73 (d, J ) 8.79 Hz, 2 H), 3.81
(s, 3 H), 3.78 (s, 2 H), 3.73 (s, 3 H); 13C NMR (100 MHz, CDCl3)
158.44, 158.04, 143.81, 139.84, 139.11, 136.58, 134.90, 132.94,
130.73, 129.12, 129.01, 128.91, 120.10, 113.65, 111.93, 110.17,
55.51, 55.09, 41.07 ppm; HRMS (EI) exact mass calculated for
C23H19ClO2: 362.1074, found: 362.1053.
Deprotection. To a solution of 1-(2-(4-(6-methoxy-2-(4-
methoxyphenyl)-1H-inden-3-yl)phenoxy)ethyl)piperidine (181.5
mg, 0.398 mmol) in CH2Cl2 (4 mL) at -78 °C was added boron
tribromide (1 M in CH2Cl2) (876 µL, 0.876 mmol). The reaction
mixture was stirred at -78 °C for 2 h and then at 25 °C for 1
h 30 min, followed by further addition of boron tribromide (1
M in CH2Cl2) (1.4 mL, 1.4 mmol). The solution was then stirred
at 25 °C for 1 h. The solution was poured into water (resulting
in precipitation of a brown residue), extracted with CH2Cl2,
washed with saturated NaHCO3, and dried over anhydrous
MgSO4. The residue from the glassware was redissolved in
acetone, the organics were combined, and solvent was removed
under reduced pressure. The crude oil was purified by flash
column chromatography (silica, 0%-5% CH3OH-CH2Cl2), fol-
lowed by pTLC (silica, 20% CH3OH-CH2Cl2), and eluted off
silica with 10% CH3OH-CH2Cl2, and solvent was removed
under reduced pressure. The residue was redissolved in 5%
CH3OH-CH2Cl2 followed by filtration through a fine silica frit
to give NC-10 as a purple oil (25.1 mg, 0.059 mmol) in 15%
yield and deemed as 61% pure by HPLC. Additionally 1,2-bis-
(4-hydroxyphenyl)-3H-inden-5-ol (122.3 mg, 0.387 mmol) was
retrieved as a side-product in 52% yield. Rf 0.29 (15% CH3-
Deprotection. To a solution of 3-(4-chlorophenyl)-6-meth-
oxy-2-(4-methoxyphenyl)-1H-indene (131.3 mg, 0.362 mmol)
in CH2Cl2 (4 mL) at -78 °C was added boron tribromide (1 M
in CH2Cl2) (876 µL, 0.876 mmol). The reaction mixture was
stirred at -78 °C for 2 h and then at 25 °C for a further 2 h,
1
OH-CH2Cl2); H NMR (400 MHz, CD3OD) δ 6.88 (m, 11 H),
4.07 (m, 2 H), 3.64 (s, 2 H), 2.74 (m, 2 H), 2.51 (m, 4 H), 1.56
(m, 4 H), 1.41 (m, 2 H); 13C NMR (100 MHz, CD3OD) 159.29,
157.23, 156.45, 145.35, 140.75, 139.08, 138.32, 132.24, 131.62,

6000 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19
Clegg et al.
after which additional CH2Cl2 (4 mL) and boron tribromide (1
M in CH2Cl2) (1 mL, 1 mmol) were added. The semisolid
mixture was stirred at 25 °C for a further 26 h overnight. The
solution was poured into water, sonicated extensively to
dissolve the solids, extracted with CH2Cl2, and washed with
saturated NaHCO3, and solvent was removed under reduced
pressure. The crude oil was purified by flash column chroma-
tography (silica, 0%-40% EtOAc-hexanes) to give NC-12 as
an orange oil (79 mg, 0.236 mmol) in 65% yield. Rf 0.33 (40%
EtOAc-hexanes); 1H NMR (400 MHz, CD3OD) δ 7.16 (d, J )
8.79 Hz, 2 H), 7.01 (d, J ) 8.30 Hz, 2 H), 6.84 (d, J ) 8.79 Hz,
2 H), 6.79 (s, 1 H), 6.69 (d, J ) 8.30 Hz, 1 H), 6.52 (dd, J )
8.30, 2.44 Hz, 1 H), 6.45 (d, J ) 8.30 Hz, 2 H), 3.48 (s, 2 H);
13C NMR (100 MHz, CD3OD) 157.33, 156.43, 145.32, 140.19,
140.01, 137.23, 136.74, 133.78, 132.04, 130.35, 129.84, 129.22,
120.90, 115.94, 114.19, 112.36, 41.75 ppm; HRMS (EI) exact
mass calculated for C21H15ClO2: 334.0761, found: 334.0772.
2-(4-Hydroxyphenyl)-1-p-tolyl-3H-inden-5-ol (NC-13).
Coupling. To a solution of ketone 11 (150 mg, 0.559 mmol)
in THF (4 mL) at 25 °C was added p-tolylmagnesium bromide
(1 M in diethyl ether) (615 µL, 0.615 mmol). The yellow
solution was stirred at 25 °C for 2 h, poured into 1.5 N HCl,
and extracted with diethyl ether, followed by CH2Cl2, the
organic layers were washed with water, combined and dried
over anhydrous MgSO4, and solvent was removed under
reduced pressure. The crude orange oil was purified by flash
column chromatography (silica, 0%-10% EtOAc-hexanes) to
give 6-methoxy-2-(4-methoxyphenyl)-3-p-tolyl-1H-indene as an
orange semisolid (149.2 mg, 0.436 mmol) in 78% yield. Rf 0.61
(40% EtOAc-hexanes); 1H NMR (400 MHz, CDCl3) δ 7.21 (m,
6 H), 7.07 (m, 2 H), 6.78 (dd, J ) 8.30, 2.44 Hz, 1 H), 6.71 (d,
J ) 8.30 Hz, 2 H), 3.79 (s, 3 H), 3.77 (s, 2 H), 3.71 (s, 3 H),
2.38 (s, 3 H); 13C NMR (100 MHz, CDCl3) 158.16, 157.86,
143.83, 140.44, 138.06, 137.81, 136.70, 133.35, 129.45, 129.38,
129.11, 129.03, 120.34, 113.46, 111.80, 109.99, 55.46, 55.02,
40.91, 21.29 ppm; HRMS (EI) exact mass calculated for
C24H22O2: 342.1620, found: 342.1624.
phenyl)-1H-indene as a red-orange oil (150.5 mg, 0.440 mmol)
in 79% yield. Rf 0.65 (40% EtOAc-hexanes); 1H NMR (400
MHz, CDCl3) δ 7.23 (d, J ) 8.79 Hz, 2 H), 7.14 (m, 3 H), 7.09
(m, 2 H), 6.95 (d, J ) 1.95 Hz, 1 H), 6.89 (d, J ) 8.30 Hz, 1 H),
6.76 (d, J ) 8.79 Hz, 2 H), 6.63 (dd, J ) 8.30, 2.44 Hz, 1 H),
3.98 (s, 2 H), 3.66 (s, 3 H), 3.65 (s, 3 H); 13C NMR (100 MHz,
CDCl3) 158.47, 157.70, 144.08, 140.13, 139.83, 139.56, 134.71,
129.72, 128.79, 128.49, 128.16, 125.95, 120.14, 113.87, 111.74,
109.94, 55.40, 55.14, 41.17, 32.19 ppm; HRMS (EI) exact mass
calculated for C24H22O2: 342.1620, found: 342.1623.
Deprotection. To a solution of 3-benzyl-6-methoxy-2-(4-
methoxyphenyl)-1H-indene (150.5 mg, 0.440 mmol) in CH2-
Cl2 (4 mL) at -78 °C was added boron tribromide (1 M in
CH2Cl2) (1.73 mL, 1.73 mmol). The reaction mixture was
stirred at -78 °C for 1 h 30 min, allowing the dry ice bath to
warm to 25 °C, after which the solution was stirred a further
2 h. Additional boron tribromide (1 M in CH2Cl2) (1.73 mL,
1.73 mmol) was added, and the solution was stirred 1 h at 25
°C. The solution was poured into water, stirred 30 min,
extracted with CH2Cl2, washed with saturated NaHCO3, and
dried over anhydrous MgSO4, and solvent was removed under
reduced pressure. The crude oil was purified by flash column
chromatography (silica, 0%-40% EtOAc-hexanes) to give the
pure NC-14 (130.8 mg, 0.416 mmol) in 95% yield. Rf 0.40 (40%
EtOAc-hexanes); 1H NMR (400 MHz, CD3OD) δ 7.04 (m, 7
H), 6.80 (d, J ) 2.44 Hz, 1 H), 6.71 (d, J ) 7.81 Hz, 1 H), 6.62
(d, J ) 8.79 Hz, 2 H), 6.47 (dd, J ) 8.30, 2.44 Hz, 1 H), 3.85
(s, 2 H), 3.52 (s, 2 H); 13C NMR (100 MHz, CD3OD) 157.41,
156.12, 145.56, 141.19, 140.93, 140.21, 135.55, 130.10, 129.95,
129.41, 129.23, 126.90, 121.09, 116.21, 113.99, 112.11, 41.89,
32.98 ppm; HRMS (EI) exact mass calculated for C22H18O2:
314.1307, found: 314.1324.
1-(4-(Trifluoromethyl)phenyl)-2-(4-hydroxyphenyl)-
3H-inden-5-ol (NC-15). Coupling. To a solution of 4-bromo-
benzotrifluoride (236 µL, 1.709 mmol) in THF (6 mL) were
added Mg turnings (41.55 mg, 1.709 mmol) and a catalytic
amount of iodine. The reaction mixture was refluxed for 50
min, until disappearance of the Mg metal indicated formation
of the Grignard reagent. The dark orange solution was allowed
to cool and added to a second flask, containing a solution of
ketone 11 (139 mg, 0.518 mmol) in THF (4 mL). The reaction
mixture was stirred at reflux for 1 h 30 min. The solution was
poured into 3 N HCl and stirred for 30 min, extracted with
diethyl ether, washed with water, and dried over anhydrous
MgSO4, and solvent was removed under reduced pressure. The
crude oil was purified by flash column chromatography (silica,
0%-10% EtOAc-hexanes) to give 3-(4-(trifluoromethyl)phenyl)-
6-methoxy-2-(4-methoxyphenyl)-1H-indene as a whitish crys-
talline solid (164.4 mg, 0.415 mmol) in 80% yield. Rf 0.64 (40%
Deprotection. To a solution of 6-methoxy-2-(4-methoxy-
phenyl)-3-p-tolyl-1H-indene (149.2 mg, 0.436 mmol) in CH2-
Cl2 (4 mL) at -78 °C was added boron tribromide (1 M in CH2-
Cl2) (959 µL, 0.959 mmol). The reaction mixture was stirred
at -78 °C for 1 h 30 min, allowing the dry ice bath to warm to
25 °C, after which the solution was stirred a further 2 h.
Additional boron tribromide (1 M in CH2Cl2) (1 mL, 1 mmol)
was added, and the solution was stirred 24 h at 25 °C. The
solution was poured into a dilute NaHCO3 solution, extracted
with CH2Cl2, washed with saturated NaHCO3, and again
extracted with CH2Cl2, and solvent from the combined organic
layers were removed under reduced pressure. The crude brown
oil was purified by flash column chromatography (silica, 0%-
40% EtOAc-hexanes), followed by pTLC (40% EtOAc-hex-
anes), and eluted off silica with EtOAc to give the pure NC-
13 (81.9 mg, 0.260 mmol) in 60% yield. Rf 0.32 (40% EtOAc-
1
EtOAc-hexanes); H NMR (400 MHz, CDCl3) δ 7.53 (d, J )
7.81 Hz, 2 H), 7.32 (d, J ) 8.30 Hz, 2 H), 6.98 (m, 3 H), 6.91
(d, J ) 8.30 Hz, 1 H), 6.67 (d, J ) 8.79 Hz, 1 H), 6.61 (d, J )
8.30 Hz, 2 H), 3.68 (s, 3 H), 3.67 (s, 2 H), 3.61 (s, 3 H); 13C
NMR (100 MHz, CDCl3) 158.57, 158.11, 143.84, 140.40, 139.85,
139.52, 136.39, 129.73, 129.31, 129.18, 128.62, 125.70, 125.67,
120.04, 113.69, 111.95, 110.19, 55.44, 55.05, 41.17 ppm; HRMS
(EI) exact mass calculated for C24H19F3O2: 396.1337, found:
396.1331.
1
hexanes); H NMR (400 MHz, CD3OD) δ 7.00 (m, 4 H), 6.91
(d, J ) 8.79 Hz, 2 H), 6.81 (d, J ) 2.44 Hz, 1 H), 6.73 (d, J )
8.30 Hz, 1 H), 6.53 (dd, J ) 8.30, 2.44 Hz, 1 H), 6.45 (d, J )
8.79 Hz, 2 H), 3.53 (s, 2 H), 2.19 (s, 3 H); 13C NMR (100 MHz,
CD3OD) 157.04, 156.23, 145.31, 140.74, 139.05, 138.58, 137.77,
135.08, 130.35, 130.25, 130.25, 129.71, 121.13, 115.80, 114.08,
112.23, 41.62, 21.36 ppm; HRMS (EI) exact mass calculated
for C22H18O2: 314.1307, found: 314.1310.
Deprotection. To a solution of 3-(4-(trifluoromethyl)phen-
yl)-6-methoxy-2-(4-methoxyphenyl)-1H-indene (164.4 mg, 0.415
mmol) in CH2Cl2 (4 mL) at -78 °C was added boron tribromide
(1 M in CH2Cl2) (913 µL, 0.913 mmol). The reaction mixture
was stirred at -78 °C for 1 h 30 min, allowing the dry ice bath
to warm to 25 °C, after which the solution was stirred a further
2 h. Additional boron tribromide (1 M in CH2Cl2) (913 µL, 0.913
mmol) was added, and the solution was stirred 1 h at 25 °C.
The solution was poured into water, stirred 30 min, extracted
with CH2Cl2, washed with saturated NaHCO3, and dried over
anhydrous MgSO4, and solvent was removed under reduced
pressure. The crude oil was purified by flash column chroma-
tography (silica, 0%-30% EtOAc-hexanes) to give NC-15 as
an orange oil (75.2 mg, 0.204 mmol) in 49% yield, and deemed
as 78% pure by HPLC. Rf 0.38 (40% EtOAc-hexanes); 1H NMR
(400 MHz, CD3OD) δ 7.52 (d, J ) 7.81 Hz, 2 H), 7.28 (d, J )
1-Benzyl-2-(4-hydroxyphenyl)-3H-inden-5-ol (NC-14).
Coupling. To a solution of ketone 11 (150 mg, 0.559 mmol)
in THF (4 mL) at 25 °C was added benzylmagnesium chloride
(2 M in diethyl ether) (307.5 µL, 0.615 mmol). The light yellow
solution was stirred at 25 °C for 1 h 30 min and then refluxed
for 2 h 30 min. Additional benzylmagnesium chloride (2 M in
diethyl ether) (307.5 µL, 0.615 mmol) was added, and the
solution was refluxed overnight for 16 h. The solution was
cooled, poured into 3 N HCl, extracted with diethyl ether,
washed with water, and dried over anhydrous MgSO4, and
solvent was removed under reduced pressure. The crude oil
was purified by flash column chromatography (silica, 0%-10%
EtOAc-hexanes) to give 3-benzyl-6-methoxy-2-(4-methoxy-

Differential Response of Estrogen Receptor Subtypes
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19 6001
8.30 Hz, 2 H), 6.87 (d, J ) 8.79 Hz, 2 H), 6.83 (d, J ) 1.95 Hz,
1 H), 6.73 (d, J ) 8.30 Hz, 1 H), 6.54 (dd, J ) 8.30, 2.44 Hz,
1 H), 6.48 (d, J ) 8.79 Hz, 2 H), 3.58 (s, 2 H); 13C NMR (100
MHz, CD3OD) 157.66, 156.70, 145.41, 142.48, 141.10, 139.67,
137.14, 131.19, 130.45, 130.21, 128.98, 126.63, 126.59, 120.86,
116.05, 114.29, 112.44, 41.98 ppm; HRMS (EI) exact mass
calculated for C22H15F3O2: 368.1024, found: 368.1026.
100% yield. Rf 0.45 (10% CH3OH-CH2Cl2); 1H NMR (400 MHz,
CD3OD) δ 7.25 (d, J ) 8.30 Hz, 2 H), 7.13 (d, J ) 8.79 Hz, 2
H), 7.01 (d, J ) 8.79 Hz, 2 H), 6.91 (d, J ) 1.95 Hz, 1 H), 6.83
(d, J ) 8.30 Hz, 1 H), 6.60 (dd, J ) 8.30, 2.44 Hz, 1 H), 6.52
(d, J ) 8.79 Hz, 2 H), 3.70 (s, 2 H); 13C NMR (100 MHz, CD3-
OD) 157.02, 156.29, 147.77, 145.36, 141.06, 138.96, 138.28,
131.14, 130.21, 130.13, 127.74, 121.22, 116.90, 115.78, 114.06,
112.17, 41.58 ppm; HRMS (EI) exact mass calculated for
C21H17NO2: 315.1259, found: 315.1251.
4-(6-Hydroxy-2-(4-hydroxyphenyl)-1H-inden-3-yl)ben-
zoic Acid (NC-17). To a solution of methyl ester 18 (20 mg,
0.052 mmol) in CH2Cl2 (2 mL) at -78 °C was added boron
tribromide (1 M in CH2Cl2) (228 µL, 0.228 mmol). The reaction
mixture was stirred at -78 °C for 1 h 30 min, additional boron
tribromide (1 M in CH2Cl2) (456 µL, 0.456 mmol) was added,
and the solution was stirred at 25 °C for 20 h. There was no
further reaction progress, so the solution was poured into
water, extracted with CH2Cl2, and then with diethyl ether,
after which the organic layers were combined and dried over
anhydrous MgSO4, and solvent was removed under reduced
pressure. The crude oil was purified by pTLC (silica, 15% CH3-
OH-CH2Cl2) and eluted off silica with diethyl ether to give
NC-17 as a yellow solid (15.9 mg, 0.046 mmol) in 89% yield,
deemed as 73% pure by HPLC. Rf 0.18 (10% CH3OH-CH2-
Cl2); 1H NMR (400 MHz, CD3OD) δ 8.00 (d, J ) 8.30 Hz, 2 H),
7.34 (d, J ) 8.30 Hz, 2 H), 6.99 (d, J ) 8.30 Hz, 2 H), 6.92 (d,
J ) 1.95 Hz, 1 H), 6.84 (m, 1 H), 6.61 (dd, J ) 8.30, 2.44 Hz,
1 H), 6.54 (d, J ) 8.79 Hz, 2 H), 3.73 (s, 2 H); 13C NMR (100
MHz, CD3OD) 169.78, 157.67, 156.73, 145.45, 143.47, 140.90,
139.87, 137.78, 131.24, 130.65, 130.45, 129.21, 126.12, 120.99,
116.05, 114.31, 112.41, 42.02 ppm; HRMS (EI) exact mass
calculated for C22H16O4: 344.1049, found: 344.1424.
1-(4-(Bromomethyl)phenyl)-2-(4-hydroxyphenyl)-3H-
inden-5-ol (NC-18). To a solution of alcohol 19 (18.6 mg, 0.052
mmol) in CH2Cl2 (4 mL) at -78 °C was added boron tribromide
(1 M in CH2Cl2) (229 µL, 0.229 mmol). The reaction mixture
was stirred at -78 °C for 30 min, allowing the dry ice bath to
warm to 25 °C, after which the solution was stirred a further
2 h 30 min. The solution was poured into water, stirred 30
min, extracted with CH2Cl2, washed with saturated NaHCO3,
and dried over anhydrous MgSO4, and solvent was removed
under reduced pressure. The crude oil was purified by pTLC
(50% EtOAc-hexanes) to give NC-18 as a yellowish solid (5.9
mg, 0.015 mmol) in 29% yield. Purity could not be ascertained
by HPLC, due to instability of the compound on the columns.
Rf 0.30 (30% EtOAc-hexanes); 1H NMR (400 MHz, CD3OD) δ
7.39 (d, J ) 8.30 Hz, 2 H), 7.20 (d, J ) 7.81 Hz, 2 H), 6.99 (d,
J ) 8.79 Hz, 2 H), 6.89 (s, 1 H), 6.82 (d, J ) 8.30 Hz, 1 H),
6.59 (m, 1 H), 6.52 (d, J ) 8.79 Hz, 2 H), 4.55 (s, 2 H), 3.69 (s,
2 H); 13C NMR (100 MHz, CD3OD) 157.45, 156.58, 145.39,
140.26, 140.07, 138.55, 138.44, 138.07, 130.84, 130.61, 130.36,
129.47, 121.07, 115.94, 114.24, 112.32, 41.88, 34.00 ppm;
HRMS (EI) exact mass calculated for C22H18O3: 392.0412,
found: 392.2022.
4-(6-Hydroxy-2-(4-hydroxyphenyl)-1H-inden-3-yl)-
benzamide (NC-19). Coupling. To a solution of flame-dried
ammonium chloride (43.1 mg, 0.806 mmol) in benzene (3 mL)
at °0 C was slowly added trimethylaluminum (2 M in hexanes)
(403 µL, 0.366 mmol). The reaction was stirred for 1 h 15 min
at 25 °C, until the evolution of bubbles ended. This trimeth-
ylaluminum-ammonium chloride complex49 was transferred
via cannula to a solution of methyl ester 18 (141.5 mg, 0.366
mmol) in benzene (2 mL), and the fluorescent yellow solution
was stirred at 60 °C for 4 h 30 min. Further reaction time did
not result in more than 50% conversion as monitored by TLC,
so the solution was cooled to 25 °C, poured into water, and
extracted with diethyl ether and then with dicholoromethane.
The organic layers were washed with saturated NaCl, com-
bined, and dried over MgSO4, and solvent was removed under
reduced pressure. The crude oil was purified by flash column
chromatography (silica, 0%-2% CH3OH-CHCl3), followed by
pTLC (silica, 5% CH3OH-CH2Cl2) and eluted off silica with
EtOAc and 5% CH3OH-CHCl3 to give 4-(6-methoxy-2-(4-
methoxyphenyl)-1H-inden-3-yl)benzamide as a whitish solid
1-(4-Aminophenyl)-2-(4-hydroxyphenyl)-3H-inden-5-
ol (NC-16). Coupling. To a solution of bromide 14 (577.5 mg,
2.29 mmol) in THF (4 mL) were added Mg turnings (56 mg,
2.29 mmol) and a catalytic amount of iodine. The reaction
mixture was refluxed for 2 h 30 min, until disappearance of
the Mg metal indicated formation of the Grignard reagent. The
solution was allowed to cool, and ketone 11 (409.7 mg, 1.53
mmol) was added. The yellow reaction mixture was stirred at
25 °C for 2 h. The solution was poured into 3 N HCl, and
extracted with EtOAc and then CH2Cl2, and washed with
saturated NaHCO3. The aqueous layer was basified with 1 N
NaOH, extracted with CH2Cl2, and washed with saturated
NaHCO3. The organic layers were combined and dried over
anhydrous MgSO4, and solvent was removed under reduced
pressure. The bright purple oil was purified by flash column
chromatography (silica, 0%-20% EtOAc-hexanes) to give N,N-
diallyl-4-(6-methoxy-2-(4-methoxyphenyl)-1H-inden-3-yl)ben-
zenamine as an orange oil (487.4 mg, 1.151 mmol) in 75% yield.
1
Rf 0.31 (40% EtOAc-hexanes); H NMR (400 MHz, CDCl3) δ
7.14 (d, J ) 8.79 Hz, 2 H), 7.08 (m, J ) 8.30, 6.84 Hz, 3 H),
6.96 (d, J ) 1.95 Hz, 1 H), 6.69 (dd, J ) 8.30, 1.95 Hz, 1 H),
6.63 (d, J ) 8.79 Hz, 4 H), 5.78 (m, 2 H), 5.09 (m, 4 H), 3.84
(m, J ) 5.37 Hz, 4 H), 3.70 (s, 3 H), 3.66 (s, 2 H), 3.63 (s, 3 H);
13C NMR (100 MHz, CDCl3) 157.96, 157.73, 147.80, 143.92,
140.73, 137.93, 137.03, 134.03, 129.98, 129.04, 123.84, 120.55,
116.13, 113.42, 112.42, 112.24, 111.77, 109.89, 55.47, 55.04,
52.66, 40.91 ppm; HRMS (EI) exact mass calculated for C29H29-
NO2: 423.2198, found: 423.2207.
Allyl Deprotection. Palladium tetrakis triphenylphos-
phine catalyst was prepared by adding tris(dibenzylidene-
acetone)dipalladium(0)-choloroform adduct (23.8 mg, 0.023
mmol) to a solution of triphenylphosphine (36.2 mg, 0.138
mmol) in CH2Cl2 (2 mL) and stirring 10 min until the solution
turned clear orange. The catalyst was transferred via cannula
to a solution of N,N-diallyl-4-(6-methoxy-2-(4-methoxyphenyl)-
1H-inden-3-yl)benzenamine (487.4 mg, 1.151 mmol) in CH2-
Cl2 (2 mL). N,N-1,3-Dimethylbarbituric acid (1.797 g, 11.51
mmol) was added, and the orange-brown reaction mixture was
stirred at 25 °C for 6 h.65 The solvent was removed under
reduced pressure, the residue was redissolved in diethyl ether,
washed with saturated NaHCO3, washed with water, and
dried over anhydrous MgSO4, and solvent was removed under
reduced pressure. The crude oil was purified by flash column
chromatography (silica, 0%-40% EtOAc-hexanes) to give 4-(6-
methoxy-2-(4-methoxyphenyl)-1H-inden-3-yl)benzenamine (353.1
mg, 1.028 mmol) in 89% yield. Rf 0.41 (40% EtOAc-hexanes);
1H NMR (400 MHz, CDCl3) δ 7.15 (d, J ) 8.79 Hz, 2 H), 7.07
(d, J ) 8.30 Hz, 2 H), 7.04 (d, J ) 8.79 Hz, 1 H), 7.00 (d, J )
2.44 Hz, 1 H), 6.72 (dd, J ) 8.30, 2.44 Hz, 1 H), 6.65 (m, 4 H),
3.75 (s, 3 H), 3.71 (s, 2 H), 3.68 (s, 3 H); 13C NMR (100 MHz,
CDCl3) 158.08, 157.82, 145.39, 143.90, 140.68, 137.89, 137.48,
130.24, 129.73, 129.06, 126.40, 120.41, 115.45, 113.49, 111.83,
110.02, 55.57, 55.11, 40.89 ppm; HRMS (EI) exact mass
calculated for C23H21 NO2: 343.1572, found: 343.1575.
Deprotection of Methyl Ether. To a solution of 4-(6-
methoxy-2-(4-methoxyphenyl)-1H-inden-3-yl)benzenamine (353.1
mg, 1.028 mmol) in CH2Cl2 (5 mL), at -78 °C was added boron
tribromide (1 M in CH2Cl2) (4.524 mL, 4.524 mmol). The
reaction mixture was stirred at -78 °C for 2 h and then at 25
°C for 2 h 30 min. The solution was poured into water and
stirred 30 min, and then the organic layer was separated from
the aqueous layer. Any remaining precipitate in the organic
or aqueous layer was collected via filtration, redissolved in
acetone, and added to the organic layer, and solvent was
removed under reduced pressure. The crude oil was purified
by flash column chromatography (silica, 0%-10% CH3OH-CH2-
Cl2) to give NC-16 as a white solid (324 mg, 1.028 mmol) in

6002 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19
(71.0 mg, 0.191 mmol) in 52% yield. Rf 0.23 (5% CH3OH-
Clegg et al.
(2) Jordan, V. C. Antiestrogens and selective estrogen receptor
modulators as multifunctional medicines. 1. Receptor interac-
tions. J. Med. Chem. 2003, 46, 883-908.
(3) Smith, C. L.; O’Malley, B. W. Coregulator function: a key to
understanding tissue specificity of selective receptor modulators.
Endocr. Rev. 2004, 25, 45-71.
(4) Paech, K.; Webb, P.; Kuiper, G. G.; Nilsson, S.; Gustafsson, J.
et al. Differential ligand activation of estrogen receptors ERalpha
and ERbeta at AP1 sites. Science 1997, 277, 1508-1510.
(5) Weatherman, R. V.; Clegg, N. J.; Scanlan, T. S. Differential
SERM activation of the estrogen receptors (ERalpha and ER-
beta) at AP-1 sites. Chem. Biol. 2001, 8, 427-436.
1
CHCl3); H NMR (400 MHz, CDCl3) δ 8.02 (d, J ) 8.42 Hz, 2
H), 7.52 (d, J ) 8.61 Hz, 2 H), 7.25 (d, J ) 8.97 Hz, 2 H), 7.22
(d, J ) 2.01 Hz, 1 H), 7.12 (d, J ) 8.42 Hz, 1 H), 6.90 (dd, J )
8.42, 2.38 Hz, 1 H), 6.82 (d, J ) 8.97 Hz, 2 H), 3.95 (s, 2 H),
3.93 (s, 3 H), 3.85 (s, 3 H); 13C NMR (100 MHz, CDCl3) 171.62,
159.39, 158.90, 144.74, 141.41, 140.56, 140.45, 137.63, 132.88,
130.23, 129.93, 129.68, 128.85, 120.71, 114.29, 112.70, 110.81,
55.92, 55.51, 41.81 ppm; HRMS (EI) exact mass calculated for
C24H21NO3: 371.1521, found: 371.1534.
Deprotection. To a solution of methyl 4-(6-methoxy-2-(4-
methoxyphenyl)-1H-inden-3-yl)benzoate (74.3 mg, 0.200 mmol)
in CH2Cl2 (2 mL) at -78 °C was added boron tribromide (1 M
in CH2Cl2) (880 µL, 0.880 mmol). The reaction mixture was
stirred at -78 °C for 45 min and then at 25 °C for 2 h 30 min.
The solution was poured into water and extracted with CH2-
Cl2 and then with EtOAc, after which the organic layers were
combined and dried over anhydrous MgSO4, and solvent was
removed under reduced pressure. The crude oil was purified
by pTLC (silica, 25% CH3OH-CH2Cl2) and eluted off silica
with EtOAc to give NC-19 as a yellow oil (45.5 mg, 0.133
(6) Rosenfeld, M. G.; Glass, C. K. Coregulator codes of transcrip-
tional regulation by nuclear receptors. J. Biol. Chem. 2001, 276,
36865-36868.
(7) Glass, C. K.; Rosenfeld, M. G. The coregulator exchange in
transcriptional functions of nuclear receptors. Genes Dev. 2000,
14, 121-141.
(8) Qiu, J.; Bosch, M. A.; Tobias, S. C.; Grandy, D. K.; Scanlan, T.
S. et al. Rapid signaling of estrogen in hypothalamic neurons
involves a novel G-protein-coupled estrogen receptor that acti-
vates protein kinase C. J. Neurosci. 2003, 23, 9529-9540.
(9) Katzenellenbogen, J. A.; Carlson, K. E.; Katzenellenbogen, B.
S. Facile geometric isomerization of phenolic nonsteroidal
estrogens and antiestrogens: limitations to the interpretation
of experiments characterizing the activity of individual isomers.
J. Steroid Biochem. 1985, 22, 589-596.
(10) Katzenellenbogen, B. S.; Norman, M. J.; Eckert, R. L.; Peltz, S.
W.; Mangel, W. F. Bioactivities, estrogen receptor interactions,
and plasminogen activator-inducing activities of tamoxifen and
hydroxy-tamoxifen isomers in MCF-7 human breast cancer cells.
Cancer Res. 1984, 44, 112-119.
(11) Robertson, D. W.; Katzenellenbogen, J. A.; Long, D. J.; Rorke,
E. A.; Katzenellenbogen, B. S. Tamoxifen antiestrogens. A
comparison of the activity, pharmacokinetics, and metabolic
activation of the cis and trans isomers of tamoxifen. J. Steroid
Biochem. 1982, 16, 1-13.
(12) Lednicer, D.; Lyster, S. C.; Aspergren, B. D.; Duncan, G. W.
Mammalian antifertility agents. 3. 1-Aryl-2-phenyl-1,2,3,4-
tetrahydro-1-naphthols, 1-aryl-2-phenyl-3,4-dihydronaphtha-
lenes, and their derivatives. J. Med. Chem. 1966, 9, 172-176.
(13) Lednicer, D.; Lyster, S. C.; Duncan, G. W. Mammalian antifertil-
ity agents. IV. Basic 3,4-dihydronaphthalenes and 1,2,3,4-
tetrahydro-1-naphthols. J. Med. Chem. 1967, 10, 78-84.
(14) McCague, R.; Kuroda, R.; Leclercq, G.; Stoessel, S. Synthesis
and estrogen receptor binding of 6,7-dihydro-8-phenyl-9-[4-[2-
(dimethylamino)ethoxy] phenyl]-5H-benzocycloheptene, a non-
isomerizable analogue of tamoxifen. X-ray crystallographic stud-
ies. J. Med. Chem. 1986, 29, 2053-2059.
(15) Adler, E.; Hagglund, B. Ark Kemi Mineral Geol. 1945, 19A, 1.
(16) Korach, K. S.; Metzler, M.; McLachlan, J. A. Estrogenic activity
in vivo and in vitro of some diethylstilbestrol metabolites and
analogs. Proc. Natl. Acad. Sci. U.S.A. 1978, 75, 468-471.
(17) Korach, K. S.; Metzler, M.; McLachlan, J. A. Diethylstilbestrol
metabolites and analogues. New probes for the study of hormone
action. J. Biol. Chem. 1979, 254, 8963-8968.
(18) Korach, K. S.; Fox-Davies, C.; Quarmby, V. E.; Swaisgood, M.
H. Diethylstilbestrol metabolites and analogs. Biochemical
probes for differential stimulation of uterine estrogen responses.
J. Biol. Chem. 1985, 260, 15420-15426.
(19) Korach, K. S.; Levy, L. A.; Sarver, P. J. Estrogen receptor
stereochemistry: receptor binding and hormonal responses. J.
Steroid Biochem. 1987, 27, 281-290.
(20) Curtis, S. W.; Shi, H.; Teng, C.; Korach, K. S. Promoter and
species specific differential estrogen-mediated gene transcription
in the uterus and cultured cells using structurally altered
agonists. J. Mol. Endocrinol. 1997, 18, 203-211.
(21) Chae, K.; Gibson, M. K.; Korach, K. S. Estrogen Receptor
Stereochemistry - Ligand Binding Orientation and Influence
on Biological Activity. Mol. Pharmacol. 1991, 40, 806-811.
(22) Kohno, H.; Bocchinfuso, W. P.; Gandini, O.; Curtis, S. W.;
Korach, K. S. Mutational analysis of the estrogen receptor
ligand-binding domain: influence of ligand structure and ster-
eochemistry on transactivation. J. Mol. Endocrinol. 1996, 16,
277-285.
(23) Mueller, S. O.; Hall, J. M.; Swope, D. L.; Pedersen, L. C.; Korach,
K. S. Molecular determinants of the stereoselectivity of agonist
activity of estrogen receptors (ER) alpha and beta. J. Biol. Chem.
2003, 278, 12255-12262.
(24) Korach, K. S.; Chae, K.; Levy, L. A.; Duax, W. L.; Sarver, P. J.
Diethylstilbestrol metabolites and analogs. Stereochemical probes
for the estrogen receptor binding site. J. Biol. Chem. 1989, 264,
5642-5647.
(25) Chae, K.; Johnston, S. H.; Korach, K. S. Multiple estrogen
binding sites in the uterus: stereochemistry of receptor and
nonreceptor binding of diethylstilbestrol and its metabolites. J.
Steroid Biochem. Mol. Biol. 1991, 38, 35-42.
1
mmol) in 66% yield. Rf 0.28 (10% CH3OH-CH2Cl2); H NMR
(400 MHz, CD3OD) δ 7.86 (d, J ) 8.79 Hz, 2 H), 7.34 (d, J )
8.30 Hz, 2 H), 7.00 (d, J ) 8.79 Hz, 2 H), 6.92 (d, J ) 1.95 Hz,
1 H), 6.84 (d, J ) 8.30 Hz, 1 H), 6.61 (m, 1 H), 6.53 (d, J )
8.79 Hz, 2 H), 3.73 (s, 2 H); 13C NMR (100 MHz, CD3OD)
172.26, 157.64, 156.72, 145.45, 142.32, 140.78, 139.94, 137.76,
133.60, 130.68, 130.45, 129.25, 129.18, 120.97, 116.03, 114.29,
112.41, 42.00 ppm; HRMS (EI) exact mass calculated for
C22H17NO3: 343.1208, found: 343.1214.
Methyl 4-(6-Hydroxy-2-(4-hydroxyphenyl)-1H-inden-
3-yl)benzoate (NC-20). To a slurry of NC-17 (39.1 mg, 0.114
mmol) in THF (3 mL) was added CH3OH (2 mL). The solution
was cooled to 0 °C, and lithium aluminum hydride (4.74 mg,
0.125 mmol) was added. The orange suspension was stirred
for 2 h at 0 °C and then overnight at 25 °C, upon which it
became brown. The reaction was poured into saturated sodium
potassium tartarate, and the CH3OH was removed under
reduced pressure. The solution was acidified to pH 3 with 3 N
HCl, extracted with ether, washed with water, and dried over
anhydrous MgSO4, and solvent was removed under reduced
pressure. The crude oil was purified by pTLC (10% CH3OH-
CHCl3) to give retrieved acid starting material (9 mg, 0.026
mmol) in 23% yield, and NC-20 product (7.6 mg, 0.0212 mmol)
in 19% yield, deemed as 72% pure by HPLC. Rf 0.40 (10% CH3-
1
OH-CHCl3); H NMR (400 MHz, CD3OD) δ 8.00 (d, J ) 8.30
Hz, 2 H), 7.36 (d, J ) 8.79 Hz, 2 H), 6.99 (d, J ) 8.79 Hz, 2 H),
6.92 (d, J ) 2.44 Hz, 1 H), 6.83 (d, J ) 8.30 Hz, 1 H), 6.61 (dd,
J ) 8.30, 2.44 Hz, 1 H), 6.54 (d, J ) 8.79 Hz, 2 H), 3.87 (s, 3
H), 3.74 (s, 2 H); the sample was too dilute for adequate
analysis by 13C NMR; HRMS (EI) exact mass calculated for
C23H18O4: 358.1205, found: 358.1206.
Acknowledgment. We are grateful for support of
this research through a grant from the National Insti-
tutes of Health (DK-57574) (T.S.S). N.J.C. was sup-
ported by a dissertation fellowship from the California
Breast Cancer Research Program (grant #8GB-0025)
and by a University of California Regents Fellowship.
Mass spectra were obtained on instruments supported
by the NIH (NCRR RR01614). We thank the laboratory
of Dr. R. Kip Guy for assistance with HPLC analysis,
and members of the Scanlan laboratory for helpful
discussions.
Supporting Information Available: Preparation of syn-
thetic intermediates and HPLC purity data for all final
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Weatherman, R. V.; Fletterick, R. J.; Scanlan, T. S. Nuclear-
receptor ligands and ligand-binding domains. Annu. Rev. Bio-
chem. 1999, 68, 559-581.

Differential Response of Estrogen Receptor Subtypes
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 19 6003
(26) Sadler, B. R.; Cho, S. J.; Ishaq, K. S.; Chae, K.; Korach, K. S.
Three-dimensional quantitative structure-activity relationship
study of nonsteroidal estrogen receptor ligands using the
comparative molecular field analysis/cross-validated r2-guided
region selection approach. J. Med. Chem. 1998, 41, 2261-2267.
(27) Lednicer, D.; Babcock, J. C.; Marlatt, P. E.; Lyster, S. C.; Duncan,
G. W. Mammalian Antifertility Agents. I. Derivatives of 2,3-
Diphenylindenes. J. Med. Chem. 1965, 39, 52-57.
(47) Silverman, T.; Bogert, M. T. The Synthesis of Some Indene and
Dihydronaphthalene Derivatives Related to Stilbestrol. J. Org.
Chem. 1948, 11, 34-49.
(48) Brown, D. W.; Denman, C.; O’Donnell, H. The Chemistry of
Indanones. Part 1. The Ring Closure of 2-(4-Methoxyphenyl)-3-
(3-Methoxyphenyl)propionic Acid. J. Chem. Soc. C 1971, 19,
3195-3198.
(49) Hwang, K. J.; Oneil, J. P.; Katzenellenbogen, J. A. 5, 6, 11, 12-
Tetrahydrochrysenes - Synthesis of Rigid Stilbene Systems
Designed to Be Fluorescent Ligands for the Estrogen-Receptor.
J. Org. Chem. 1992, 57, 1262-1271.
(50) Blizzard, T. A.; Mosley, R. T.; Birzin, E. T.; Chan, W.; Hammond,
M. L. Comparison of 2-phenylspiroindenes and 2-phenylspiroin-
denediones as estrogen receptor ligandssmodeling and binding
data don’t agree! Bioorg. Med. Chem. Lett. 2004, 14, 1317-
1321.
(51) 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.
(52) Mortensen, D. S.; Rodriguez, A. L.; Carlson, K. E.; Sun, J.;
Katzenellenbogen, B. S. et al. Synthesis and biological evaluation
of a novel series of furans: ligands selective for estrogen receptor
alpha. J. Med. Chem. 2001, 44, 3838-3848.
(53) 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.
(54) Shiau, A. K.; Barstad, D.; Radek, J. T.; Meyers, M. J.; Nettles,
K. W. et al. Structural characterization of a subtype-selective
ligand reveals a novel mode of estrogen receptor antagonism.
Nat. Struct. Biol. 2002, 9, 359-364.
(55) Nettles, K. W.; Sun, J.; Radek, J. T.; Sheng, S.; Rodriguez, A. L.
et al. Allosteric control of ligand selectivity between estrogen
receptors alpha and beta: implications for other nuclear recep-
tors. Mol. Cell 2004, 13, 317-327.
(56) Webb, P.; Lopez, G. N.; Uht, R. M.; Kushner, P. J. Tamoxifen
Activation of the Estrogen Receptor/AP-1 Pathway - Potential
Origin For the Cell-Specific Estrogen-Like Effects of Anti-
estrogens. Mol. Endocrinol. 1995, 9, 443-456.
(57) Paruthiyil, S.; Parmar, H.; Kerekatte, V.; Cunha, G. R.;
Firestone, G. L. et al. Estrogen receptor beta inhibits human
breast cancer cell proliferation and tumor formation by causing
a G2 cell cycle arrest. Cancer Res. 2004, 64, 423-428.
(58) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Safe and convenient procedure for solvent
purification. Organometallics 1996, 15, 1518-1520.
(59) Hone, N. D.; Baxter, A. D. Organosilicon Compounds and Their
Use in Combinatorial Chemistry; European Patent Office: Lon-
don, Great Britain, 1998.
(60) Myers, T. C.; Pratt, R. J.; Morgan, R. L.; O’Donnell, J.; Jensen,
E. V. Hormone Analogues. I. Preparation of Intermediate
Ketones for the Synthesis of Stilbestrol Analogues. J. Am. Chem.
Soc. 1955, 77, 5655-5657.
(61) Skaddan, M. B.; Katzenellenbogen, J. A. Integrated "3+1"
oxorhenium(V) complexes as estrogen mimics. Bioconjugate
Chem. 1999, 10, 119-129.
(28) Duncan, G. W.; Lednicer, D. Method for Controlling Birds and
Rodents; US Patent Office, 1967; p 46.
(29) Anstead, G. M.; Altenbach, R. J.; Wilson, S. R.; Katzenellen-
bogen, J. A. 2,3-Diarylindenes and 2,3-diarylindenones: syn-
thesis, molecular structure, photochemistry, estrogen receptor
binding affinity, and comparisons with related triarylethylenes.
J. Med. Chem. 1988, 31, 1316-1326.
(30) Anstead, G. M.; Katzenellenbogen, J. A. Fluorescence Properties
of 2,3-Diarylindenes. J. Phys. Chem. 1988, 92, 6249-6258.
(31) Anstead, G. M.; Peterson, C. S.; Pinney, K. G.; Wilson, S. R.;
Katzenellenbogen, J. A. Torsionally and hydrophobically modi-
fied 2,3-diarylindenes as estrogen receptor ligands. J. Med.
Chem. 1990, 33, 2726-2734.
(32) Anstead, G. M.; Wilson, S. R.; Katzenellenbogen, J. A. 2-Aryl-
indenes and 2-arylindenones: molecular structures and consid-
erations in the binding orientation of unsymmetrical non-
steroidal ligands to the estrogen receptor. J. Med. Chem. 1989,
32, 2163-2171.
(33) Anstead, G. M.; Peterson, C. S.; Katzenellenbogen, J. A. Hy-
droxylated 2,3-diarylindenes: synthesis, estrogen receptor bind-
ing affinity, and binding orientation considerations. J. Steroid
Biochem. 1989, 33, 877-887.
(34) Kirkiacharian, S.; Koutsourakis, P. G.; Philibert, D.; Bouchoux,
F.; Van de Velde, P. Synthesis and relative binding affinity to
steroid receptors and antiproliferative activity on MCF-7 cells
of 2,3-disubstituted indenes. Farmaco 1999, 54, 678-683.
(35) Bryant, H. U.; Jones, C. D.; Pennington, P. A. Indene compounds
having activity as SERMs; European Patent Office: London,
Great Britain, 1998; p 33.
(36) 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.
(37) Pike, A. C.; Brzozowski, A. M.; Hubbard, R. E.; Bonn, T.;
Thorsell, A. G. et al. Structure of the ligand-binding domain of
oestrogen receptor beta in the presence of a partial agonist and
a full antagonist. EMBO J. 1999, 18, 4608-4618.
(38) 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.
(39) Korach, K. S.; Chae, K.; Gibson, M.; Curtis, S. Estrogen Receptor
Stereochemistry - Ligand Binding and Hormonal Responsive-
ness. Steroids 1991, 56, 263-270.
(40) Chae, K.; Levy, L. A.; Korach, K. S. Chromatographic separation
and isolation of the enantiomers of diethylstilbestrol metabolites.
J. Chromatogr. 1988, 439, 484-487.
(41) Parker, C. E.; Levy, L. A.; Smith, R. W.; Yamaguchi, K.; Gaskell,
S. J. et al. Separation and detection of enantiomers of stilbestrol
analogues by combined high-performance liquid chromatogra-
phy-thermospray mass spectrometry. J. Chromatogr. 1985, 344,
378-384.
(62) Matsuo, J.; Koga, K. Enantioselective alpha-alkylation of phen-
ylacetic acid using a chiral bidentate lithium amide as a chiral
auxiliary. Chem. Pharm. Bull. 1997, 45, 2122-2124.
(63) Hauser, C. R.; Chambers, W. J. Alkylations of the R-Carbon of
Phenylacetic and Diphenylacetic Acids with Benzyl Chloride and
Related Halides by Alkali Amides. J. Am. Chem. Soc. 1956, 78,
4942-4944.
(42) Sadler, B. R.; Chae, K.; Ishaq, K. S.; Korach, K. S. Separation
of indenestrol A and B isomers and enantiomers by high-
performance liquid chromatography. J. Chromatogr. A 1998,
799, 117-124.
(43) Anstead, G. M.; Ensign, J. L.; Peterson, C. S.; Katzenellenbogen,
J. A. 2-Arylindenes and 2-arylindenones: synthesis of probes
to study the binding orientation of unsymmetrical nonsteroidal
ligands to the estrogen receptor. J. Org. Chem. 1989, 54, 1495-
1491.
(64) Carpino, L. A.; Lin, Y. Z. Benz[f]Indene. J. Org. Chem. 1990,
55, 247-250.
(65) Garro-Helion, F.; Merzouk, A.; Guibe, F. Mild and selective
palladium(0)-catalyzed deallylation of allylic amines. Allylamine
and diallylamine as very convenient ammonia equivalents for
the synthesis of primary amines. J. Org. Chem. 1993, 58,
6109-6113.
(44) Plentl, A. A.; Bogert, M. T. The Synthesis of Tricyclic Hydro-
carbons Related to Stilbestrol. J. Am. Chem. Soc. 1941, 63, 989-
995.
(45) Meurling, L. Alkylation of Indenyllithium and Alkylindenyl-
lithium Compounds. Acta Chem. Scand., Ser. B, Org. Chem.
Biochem. 1974, B28, 295-300.
(46) Cedheim, L.; Eberson, L. Simple Procedure for Preparation of
1-Alkylindenes. Synthesis (Stuttgart) 1973, 159-159.
JM050226I