Isocoumarins as estrogen receptor beta selective ligands: Isomers of isoflavone phytoestrogens and their metabolites

Article abstract of DOI:10.1016/j.bmc.2005.07.014

In a search for new ligands selective for the estrogen receptor beta (ERβ), we prepared a series of non-steroidal compounds having an isocoumarin core structure. An interesting feature of these derivatives is that they bear the same functionalities as the well-known ERβ -selective, isoflavone phytoestrogens daidzein and genistein, but in an isomeric arrangement. These compounds could be prepared efficiently by electrophilic cyclization of acetylenic ester precursors, followed by simple manipulations to introduce additional substituents. Through a reduction of some of the isocoumarins, we also obtained isomeric analogs of the isoflavone metabolites equol and dehydroequol. The compounds we prepared were evaluated in ER binding assays, and selected compounds were studied further in cell-based gene transcription assays. Several of the isocoumarins and their analogs are high-affinity ligands that show considerable selectivity for ERβ in terms of binding affinity, and strikingly high ERβ selectivity in terms of potency in gene transcription assays. Two of the best compounds, which combine high transcriptional potency with an ERβ selectivity greater than 1000, should prove to be excellent probes of ERβ function in vivo.

Full text of DOI:10.1016/j.bmc.2005.07.014

Bioorganic & Medicinal Chemistry 13 (2005) 6529–6542
Isocoumarins as estrogen receptor beta selective ligands:
Isomers of isoflavone phytoestrogens and their metabolites
Meri De Angelis,a Fabio Stossi,b Michael Waibel,a
Benita S. Katzenellenbogenb and John A. Katzenellenbogena,*
aDepartment of Chemistry, University of Illinois, 600 South Mathews Avenue, Urbana, IL 61801, USA
bDepartment of Molecular and Integrative Physiology, University of Illinois, 600 South Matthews Avenue, Urbana, IL 61801, USA
Received 16 June 2005; revised 1 July 2005; accepted 5 July 2005
Available online 15 August 2005
Abstract—In a search for new ligands selective for the estrogen receptor beta (ERb), we prepared a series of non-steroidal com-
pounds having an isocoumarin core structure. An interesting feature of these derivatives is that they bear the same functionalities
as the well-known ERb -selective, isoflavone phytoestrogens daidzein and genistein, but in an isomeric arrangement. These com-
pounds could be prepared efficiently by electrophilic cyclization of acetylenic ester precursors, followed by simple manipulations
to introduce additional substituents. Through a reduction of some of the isocoumarins, we also obtained isomeric analogs of the
isoflavone metabolites equol and dehydroequol. The compounds we prepared were evaluated in ER binding assays, and selected
compounds were studied further in cell-based gene transcription assays. Several of the isocoumarins and their analogs are high-
affinity ligands that show considerable selectivity for ERb in terms of binding affinity, and strikingly high ERb selectivity in terms
of potency in gene transcription assays. Two of the best compounds, which combine high transcriptional potency with an ERb selec-
tivity greater than 1000, should prove to be excellent probes of ERb function in vivo.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
are almost identical, the principal differences being a
somewhat smaller internal volume for ERb and the sub-
stitution of just two amino acids, with Met421 in ERa
corresponding to Ile373 in ERb, and Leu384 in ERa
corresponding to Met336 in ERb.6
The estrogen receptors (ERs), members of the nuclear
hormone receptor superfamily, mediate the activity of
estrogens in many different organs, including the repro-
ductive, skeletal, cardiovascular, and central nervous
systems.1–3 Two different ER subtypes, the product of
different genes, are known: ERa, the first discovered,
and ERb, which was only discovered in 1996.4,5 Both
estrogen receptors bind 17b-estradiol with high affinity
and bind to classical estrogen response elements in a
similar way; however, there are differences in their ami-
no acid sequence, transcriptional activity, and pattern of
tissue distribution.
The tissue distribution patterns of ERa and ERb are
also quite different. Since ERb is mostly expressed in
the prostate, ovary, colon, urinary tract, and some brain
regions, but is less expressed in certain reproductive or-
gans,7 a selective ERb agonist might maintain the bene-
ficial effects of estrogen therapy in these ERb-rich tissues
without increasing the risk of cancer in other organs that
are ERa rich, such as the breast and uterus. In addition,
ERb has been shown to be antiproliferative when pres-
ent along with ERa in breast cancer cells.8 For this rea-
son, many researchers have focused their attention on
the synthesis of compounds selective for ERb.
The overall amino acid sequence identity of the two ER
subtypes is 44%; the DNA-binding domain is very well
conserved, while the amino acid sequence identity of
the ligand-binding domain (LBD) is 59%. Despite these
differences, the two ER ligand-binding pockets (LBP)
Among non-steroidal compounds, we have reported that
the diarylpropionitrile derivative 1 (DPN)9 and the inda-
zole compound 2 are very ERb selective;10 others have
reported the benzoxazole product 3 also to be highly
ERb selective (Fig. 1).11 It is interesting that the soy iso-
flavones, such as daidzein (4a, Fig. 2) and genistein (4b),
Keywords: Estrogen receptor alpha; Estrogen receptor beta; Isocoum-
arin; Equol; Genistein.
*
Corresponding author. Tel.: +1 217 333 6310; fax: +1 217 333
7325; e-mail: jkatzene@uiuc.edu
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.07.014

6530
M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
OH
Y
Y
OH
Cl
N
O
N
X
O
O
X
CN
HO
O
Coumarin
Isocoumarin
HO
OH
1 DPN
2
OH
N
Core structure
Substituent
HO
O
HO
F
Figure 3. Coumarin and isocoumarin cores; ER pharmacophore
model.
3
Figure 1. ER-selective ligands.
tro.14 Thus, a series of C-4 substituted isocoumarins
was also synthesized.
OH
OH
O
O
X
O
O
The isocoumarin scaffold conforms to a pharmacophore
model for ERb (Fig. 3) that has emerged from our work
and that of others.10 This model—in which the size of
the central region of the ligand is rather small—is based
on the fact that the ERb binding pocket is smaller by
about 100 A than that of ERa, so that ERb selectivity
can often be engendered by reducing ligand size in this
region, as well as by including some polar functions.9,10
In this regard, it is of note that one of the two phenol
rings in the isocoumarin structure is fused to form a het-
erocycle, which reduces size and introduces core polari-
ty. Ideally, as well, the nature of substituents in this ERb
pharmacophore should be such that they interact favor-
ably with the Ile373 or Met336 residues in ERb and/or
unfavorably with the Met421 and Leu384 in ERa, in
this way also engendering better selectivity for ERb.15,16
HO
O
HO
X = H Daidzein (4a)
X = OH Genistein (4b)
Coumestrol 5
3
6
˚
OH
OH
HO
O
HO
O
Dehydroequol 7
Equol 6
Figure 2. Phytoestrogens and analogs.
and the clover coumarin coumestrol (5), long known to be
estrogenic (i.e., phytoestrogens), were among the first
compounds noted to be ligands with selective affinity for
ERb, a property shared by the daidzein enteric metabo-
lite, (S)-equol (6).9,12 Dehydroequol (7) is also known to
show estrogenic activity,13 but its selectivity for ERb
has not been reported. Despite their considerable selectiv-
ity for ERb in terms of binding,9,12 however, the phytoes-
trogens 4–6 showed only modest selectivity for ERb in
terms of potency in transcription assays.12
Herein, we present the synthesis and biological evalua-
tion of a series of isocoumarins and derivatives, some
of which show high selectivity for ERb both in terms
of binding affinity and transcriptional potency.
2. Results and discussion
2.1. Chemical synthesis
All the compounds synthesized by us can be divided in
to two groups: isocoumarins and their reduced analogs.
Since we are interested in developing compounds that
are selective for ERb not just in terms of their binding
affinity for the ERs but also in terms of their potency
in cell-based assays and ultimately in vivo, we decided
to investigate a series of compounds that possess a scaf-
fold related to daidzein (4a), coumestrol (5), and ( )-
equol (6), but have an isomeric arrangement of the cen-
tral atoms of the heterocyclic core. For this reason we
turned our attention to compounds having an isocoum-
arin core structure, a system that can also be reduced to
give isomeric analogs of ( )-equol (6) and dehydroequol
(7) (Fig. 3). The isocoumarin core structure is found in
many natural products that exhibit a broad range of
biological activities. For example, unsymmetrical 3-ar-
yl-4-substituted isocoumarins are reported to possess
cytotoxic activity against human cancer cell lines in vi-
2.1.1. Isocoumarin derivatives. The synthesis of the iso-
coumarin core structure was accomplished as reported
in Scheme 1. We were able to obtain the diarylacetylene
10 via a palladium-catalyzed Sonogashira coupling reac-
tion between derivative 8 and 4-ethynylanisole (9). The
original literature procedure17 worked satisfactorily on
a 1 mmol scale, giving the desired compound > 80%
yield, but it was unsatisfactory on a larger (10 mmol)
scale. By making several changes (using 5 mol % of
Pd(PPh3)4, 15 mol % of CuI, and 2 mmol of Bu4NI),
we were able to obtain a 75% yield at the larger scale.
According to a literature precedent,14 the acetylenic
ester 10 can be converted easily to the isocoumarin

M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
OTf OCH3
OCH3
6531
+
a
H3CO
O
8
9
OR
OCH3
X
b
O
RO
OCH3
O
H3CO
O
11 X= I, R=CH3
12a X= Br, R= CH3
12b X= Br, R=H
10
d
OR
c from 11
O
RO
d
O
13a R= CH3
13b R=H
Scheme 1. Reagents: (a) Pd(PPh3)4, CuI, Bu4NI; Et3N–CH3CN; (b) I2, CH3CN (for 11), Br2, CH3CN (for 12a); (c) HCO2H, Et3N, PPh3, Pd(OAc)2,
DMF; (d) BBr3, CH2Cl2.
13a by an iodo-lactonization reaction, giving the iodo
enol lactone 11, from which the iodine group can be re-
moved readily by hydrogenolysis. Although there are
only few examples of bromo-lactonization reported in
the literature,18 we found that the corresponding bromo
enol lactone 12a could also be easily obtained following
the reaction conditions used for the synthesis of 11, with
bromine replacing iodine.
resents an inaccuracy in melting point determination
or an error in attribution of structure. In any case the
melting point that we have measured for derivative
13b (>300 ꢁC) is consistent with that reported by Rose
et al.23
As shown in Scheme 2, the corresponding chloro com-
pound 15b could be obtained by reaction of the acety-
lenic acid 10 with mercuric acetate to give the
corresponding isocoumarin mercurial 14.25 Direct
replacement of the mercury group by chlorine was
accomplished using copper(II) chloride, and the chlo-
ro enol lactone obtained (15a) was then deprotected to
give 15b.
Compounds 11 and 12a were obtained not only in >70%
yield but also regioselectively (>90%). Their structure
was confirmed as the six-membered-ring enol lactone
structure—as opposed to the potential alternative lact-
onization product, the five-membered alkylidine ben-
zobutyrolactone—by the carbonyl stretch in the IR
spectra: 1726 cmꢀ1 for derivative 11, and 1734 cmꢀ1
for 12a; the butyrolactone is expected to have a much
higher stretching frequency (>1750 cmꢀ1).19–22 Deriva-
tives 12a and 13a were deprotected using BBr3 to give
the final free phenolic compounds 12b and 13b. At-
tempts to deprotect the iodo derivative 11 failed because
this compound was unstable under the hydrolytic condi-
tions employed.
Starting from compound 11, several different polar sub-
stituents were introduced at the C-4 position by dis-
placement of the iodine group (Scheme 3). Through a
Stille coupling reaction,26 the methyl (16a) and vinyl
(17) compounds were synthesized, and from the reduc-
tion of vinyl compound 17, ethyl derivative 18a could
be readily obtained. A Suzuki coupling reaction between
compound 11 and phenyl boronic acid was employed
for the synthesis of phenyl-substituted 19a. For the syn-
thesis of the CF3 analog 20a, we followed a literature
procedure described for the replacement of an aromatic
bromine group with a trifluoromethyl moiety using an
unusual fluorosulfonylfluoroacetate reagent as a precur-
sor for the in situ generation of the trifluoromethide spe-
cies.27 Although examples of this last reaction have
never been reported for our particular heterocyclic sys-
tem, we were able to obtain the desired compound
in good yield, within a few hours. With all of these
Isocoumarin 13b is a known compound (though synthe-
sized by different methodology).23,24 There is, however,
a large discrepancy in the melting points reported for
compound 13b in the two previous papers in which it
was described: Rose et al.23 report a mp of > 300 ꢁC,
whereas Mahto et al.24 report a mp of 226–227 ꢁC. Be-
cause other analytical data for the compounds prepared
in these two earlier reports are not available for compar-
ison, we cannot ascertain whether this discrepancy rep-

6532
M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
OCH3
OCH3
HgCl
O
a
OCH3
H3CO
H3CO
O
O
14
10
OR
Cl
b
O
RO
O
15a R= CH3
15b R=H
c
Scheme 2. Reagents: (a) Hg(OAc)2, AcOH, NaCl; (b) CuCl2, THF; (c) BBr3, CH2Cl2.
OR
OCH3
c
O
O
RO
H3CO
O
O
18a R= CH3
18b R=H
17
f
b
OR
CH3
OR
OCH3
I
d
a
O
O
RO
O
O
RO
H3CO
O
O
16a R= CH3
16b R=H
11
19a R= CH3
19b R=H
f
f
e
OR
CF3
O
RO
O
20a R= CH3
20b R=H
f
Scheme 3. Reagents: (a) (CH3)4Sn, Pd2(dba)3, PtBu3, CsF, dioxane; (b) VinylSn(Bu)3, Pd2(dba)3, PtBu3, CsF, dioxane; (c) Pd/C, H2, EtOH; (d)
PhB(OH)2, Pd(PPh3)4, Cs2CO3, DMF; (e) FSO2CF2CO2CH3, CuI, DMF; (f) BBr3, CH2Cl2.
compounds, the methyl ether protecting groups were
cleaved with boron tribromide.
ted to give product 22b, which was in turn reduced with
LiAlH4 to obtain isoequol 23 as a racemate.
2.1.2. Reduced analogs. Starting from the basic enol lac-
tone 13a, the keto-acid intermediate 21 could be obtained
by an alkaline hydrolysis (Scheme 4). Reduction of this
compound to the racemic hydroxyl acid and subsequent
cyclodehydration using acetic anhydride gave the desired
reduced lactone 22a.28 Derivative 22a was then deprotec-
The synthesis of dehydroisoequol 25 and its analogs 26
and 27 (Scheme 5) was accomplished by a simple reduc-
tion of the enol lactones 13b, 16b, and 2419 with LiAlH4.
In general, the reduction of the carboxyl group gave us
reasonable yields even if different reaction conditions
were employed.

M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
6533
OCH3
a
OCH3
O
H3CO
O
CO2H
H3CO
O
13a
21
OR
OH
d from 22b
b
O
O
RO
HO
O
23
22a R=CH3
22b R=H
c
Scheme 4. Reagents: (a) KOH, EtOH; (b) NaBH4, EtOH, Ac2O; (c) BBr3, CH2Cl2; (d) LiAlH4, THF.
OH
for the phytoestrogens, daidzein (4a), genistein (4b),
and ( )-equol (6).12
OH
a
R
R
X
X
In general, the other compounds of this series show ERb
binding affinities near or higher than that reported for
estradiol, as in the case of derivative 20b, 15b, and 12b
with RBA values of 55%, 63%, and 129%, respectively,
and ERb-selectivities between 6- and 8-fold. The only
exception is represented by derivative 19b, which is slight-
ly ERa-selective; the phenyl substituent is probably too
big for the ligand binding pocket of ERb.6,10 In fact, if
we compare the selectivity of derivatives 13b, 16b, 18b,
and 19b, it is evident that an increase of the dimension
of the C-4 substituent leads to a dramatic reduction in
selectivity. However, when the C-4 position remains
unsubstituted (13b), the receptor binding affinity is low,
especially for ERa. This result is consistent with the
earlier observations of Rose et al.23 who found 13b and
its analogs to be inactive in animal assays. In general,
from this set of data we can conclude that the introduc-
tion in position C-4 of different polar groups is well toler-
ated by both receptors, but that small substituents seem
to increase selectivity toward ERb (13b vs 16b vs 18b).
O
O
O
25 R=H, X=7-OH
26 R=Me, X= 7-OH
27 R=H, X= 6-OH
13b R=H, X=7-OH
16b R=Me, X= 7-OH
24 R=H, X= 6-OH
Scheme 5. Reagents: (a) LiAlH4, THF.
2.2. Biological results
2.2.1. Estrogen receptor binding affinity. The compounds
synthesized were evaluated in competitive radiometric
binding assays to determine their binding affinities for
human ERa and ERb. The results of these assays are
summarized in Table 1. In this table the binding-affinity
values of daidzein (4a), genistein (4b), ( ) equol (6), and
dehydroequol (7) are also reported.
Binding affinities are expressed relative to that of estradiol
(100%), as relative binding affinity (RBA) values (Table 1,
left), from which the Ki values have also been calculated
(Table 1, center). We have used RBA values when discuss-
ing affinity comparisons. One should note, however, that
estradiol has a 2.5-fold higher affinity for ERa than for
ERb (Kd [ERa] = 0.2 nM vs [ERb] = 0.5 nM). Thus, the
Ki values, which represent the inherent affinity of the com-
peting ligand itself for the two ER subtypes, are the appro-
priate ones to use in comparing binding affinity with
transcriptional potency (see next section).
The reduced analogs (22b, 23, 25–27) show lower affini-
ties than those of the family of isocoumarin derivatives,
but some of these compounds are the most ERb selec-
tive. In fact, derivatives 25 and 27, with a selectivity of
62- and 102-fold, respectively, and an RBA around
2.5%, are the best compounds of this series in terms of
their ERb affinity selectivity. It is particularly interesting
to note that an increase in selectivity is obtained when
the hydroxyl group is moved from the C-7 to C-6
position (25 vs 27). Curiously, the same change in the
isocoumarin derivatives leads to a dramatic reduction
both in activity and selectivity (13b vs 24). Some other
trends in the binding data should be noted: (1) When
the double bond of the heterocyclic system is reduced,
a decrease in ERb binding affinity is observed (13b vs
22b, 25 vs 23 and 7 vs 6). (2) When the oxygen in the
reduced isomeric analogs 23 and 25 is moved from the
benzylic position to the aryl position (6 and 7), an
Among the family of isocoumarin derivatives, the best
results in terms of binding affinity selectivity are
observed for compounds 13b and 16b, these being
82- and 40-fold ERb selective, respectively, in terms of
RBA values, with the methyl derivative 16b having an
ERb RBA of 24%, and compound 13b an ERb RBA
of 2.1%. It is also important to point out that these
compounds show selectivity greater than that reported

6534
M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
Table 1. ERa and ERb binding affinities (relative binding affinity [RBA] and Ki values), transcriptional potencies (EC50 values), and affinity and
potency selectivities for isocoumarin and related ligands
Compound
Ligand binding
Transcription potency
RBAa (%)
ERb
b/a ratiod
Kib (nM)
ERb
b/a ratiod
b/a ratiod
c
EC50 (nM)
ERa
ERa
ERa
ERb
OH
[100]
[100]
[1]
0.2
0.5
0.4
0.021 0.11
0.19
HO
Estradiol
OH
Br
16.2 4.8
129 24
7.9
1.23
0.38
3.2
27
0.06
450
O
HO
O
12b
OH
OH
OH
OH
0.026 0.004
2.14 0.08
82
770
23
33
NTe
110
600
300
NT
NT
0.06
1.4
NT
O
O
13b
Cl
9.8 1.3
63 13
6.4
2
34
4
0.8
2.1
3.1
2.5
1830
O
HO
HO
HO
O
15b
CH3
0.59 0.01
23.9 4.5
40
16
430
O
O
16b
CH3
5.0 1.1
16.0 4.3
3
1.3
0.11
2700
O
O
18b
OH
Ph
32.2 9.3
6.4 1.4
0.2
0.62
7.8
0.9
0.08
NT
NT
O
HO
O
19b
OH
CF3
9.1 1.4
55 17
6.0
2.2
2.4
7.1
0.04
180
O
HO
O
20b
OH
0.015 0.002 0.026 0.005
1.7
1300
1900
0.68
NT
NT
NT
O
HO
O
22b

M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
6535
Table 1 (continued)
Compound
Ligand binding
Transcription potency
c
EC50 (nM)
RBAa (%)
ERb
b/a ratiod
Kib (nM)
b/a ratiod
b/a ratiod
ERa
ERa
ERb
ERa
ERb
OH
HO
0.012 0.004
0.34 0.05
28
1700
570
150
11
NT
NT
NT
NT
NT
O
O
24
OH
0.035 0.007
0.27 0.01
7.7
190
3
NT
O
HO
23
OH
OH
0.042 0.01
2.6 0.7
62
36
480
19
25
NDe
12
ND
O
HO
25
CH3
0.23 0.06
8.3 1.8
87
6
14.5
600
1.5 400
O
HO
26
OH
OH
HO
0.024 0.001
2.45 0.02
102
833
100
20
29
37
ND
170
ND
O
27
0.20 0.02
1.66 0.5
8
3.2
200
74
2.7
HO
HO
O
(±) Equol (6)f
OH
0.046 0.01
4.3 1.2
93
430
12
36
3200
7.2 440
O
Dehydroequol (7)
OH
OH
O
0.010 0.006 0.040 0.001
4
2000
1300
1.5
7.6
250 100
2.5
HO
O
Daidzein (4a)f
OH
O
0.7
13
19
29
3.8
20
6.0
3.3
HO
O
Genistein (4b)f
½compoundꢁ
a Relative binding affinity (RBA) values are determined by competitive radiometric binding assays and are expressed as IC5½e0stradiolꢁ=IC50
(RBA, estradiol = 100). In these assays, the Kd for estradiol is 0.2 nM on ERa and 0.5 nM on ERb. For details, see Section 4.
b Ki = equilibrium binding competition constant, calculated from the Kd of estradiol on ERa (0.2 nM) or ERb (0.5 nM): (Kd/RBA) · 100.
· 100
c Transcriptional activity measured using a cotransfection assay in HEC-1 cells. Transcriptional potency = EC50
.
d For each index, the b/a ratio is calculated such that the ratio is > 1 for compounds having higher affinity or greater potency on ERb than on ERa.
e NT = not tested; ND = not determined.
f Data are from Meyers et al.9 As noted in Muthyala, et al.12 large variations and some lower values have been reported for the ERa affinity of
genistein, which would indicate a somewhat higher ERb/ERa affinity selectivity.

6536
M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
increase in affinity and selectivity is noted. (3) Dehydroe-
quol (7), with an ERb RBA of 4.3% and a 93-fold ERb
binding selectivity, has ERb selectivity greater than that
of genistein, daidzein, and (S)-equol (32-fold beta selec-
tive) with activity comparable with that reported for (S)-
equol (3.2% ERb RBA).12
affinity values (Table 1, center) with transcriptional
potency EC50 values (Table 1, right).
In general, all compounds tested were agonists on both
ER subtypes, having full (7, 12b, 15b, 16b, 25, and 27) or
nearly full (18b, 20b, and 26) efficacy, and they show
very pronounced ERb potency selectivity that, in all
cases but one (27), is greater—often far greater—than
their ERb binding affinity selectivities. In particular,
derivatives 12b, 15b, and 20b (bromo, chloro, and CF3
isocoumarin derivatives, respectively) show good activi-
ty on ERb at 10ꢀ10 M, and thus have a potency on this
ER subtype that is close to that of estradiol; by contrast,
their activity on ERa is evident only at much higher con-
centrations. Similar behavior is also observed for com-
pounds 16b, 18b, and 26 (methyl and ethyl
isocoumarins, and methyl dehydroisoequol derivatives,
respectively), although overall they are somewhat less
potent. Of these six compounds, the greatest ERb poten-
cy selectivity is found with compounds 15b, 16b, 18b,
and 26; at 10ꢀ8 M, these compounds fully activate
ERb, yet have negligible activity on ERa (Fig. 4). Thus,
their ERb potency selectivity is 400 to >1000 (Table 1,
right). The simple dehydroisoequol derivative 25 and
its isomer 27 show interesting activity: they have low
2.2.2. Activity and potency in transient transfection
reporter gene assays. The transcriptional activities of
some derivatives were assayed in human endometrial
cancer (HEC-1) cells transfected with expression vectors
for ERa and ERb, and an estrogen-responsive reporter
gene construct. The activities were normalized to that
obtained with 10ꢀ8 M estradiol, which is set to 100.
Nine derivatives were selected for testing in this tran-
scription assay (12b, 15b, 16b, 18b, 20b, 25, 26, 27 and
dehydroequol 7), based on their ERb affinity and selec-
tivity. The dose–response curves for these compounds
are shown in Figure 4, and the EC50 values (which are
comparable to the Ki values for binding affinity) are
summarized in Table 1 (right). In this table, the EC50
and Ki values of daidzein (4a), genistein (4b), and ( )-
equol (6), determined in our previous work,9,12 are also
reported for the purpose of comparison. As noted
above, it is most appropriate to compare the Ki binding
A
B
E
H
C
F
I
D
G
L
Figure 4. Transcriptional activation by ERa and ERb in response to estradiol (E2), 12b, 15b, 16b, 18b, 20b, 25, 26, 27, and 7. Human endometrial
cancer (HEC-1) cells were transfected with expression vectors for ERa or ERb, and the estrogen responsive gene 2EREp-pS2-Luc and were incubated
with the indicated concentrations of estradiol or ligands for 24 h. Luciferase activity was measured as noted in the Section 4.

M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
6537
to no activity for ERa at concentrations of 10ꢀ8 M and
below, but at 10ꢀ6 M, they are as active as estradiol on
ERb. Curiously, these compounds are some of the most
selective ones in the binding assay (RBA values show 60-
fold beta selectivity for derivative 25, 102-fold for com-
pound 27), although their affinities and potencies on
ERb are rather low.
daidzein and genistein, these ERb potency selectivities
also being greater than their binding selectivities.
Two sets of compounds stand out among those we have
studied. The halogen-substituted isocoumarins, 12b and
15b, have transcriptional potencies on ERb that equal or
exceed that of estradiol, combined with very favorable
ERb transcriptional potency selectivities. The two al-
kyl-substituted isocoumarins, 16b and 18b, also have
very high ERb potency selectivities while retaining good
ERb potency. These two sets of compounds should
facilitate investigations of the biological and physiolog-
ical roles of ERb, and they might also be useful for
examining the structural conformation of ERb agonist
complexes.
If we compare the results of the transcription assays on
these isocoumarin analogs with those reported for the
isoflavonoids daidzein (4a), genistein (4b), and ( ) equol
(6) (Table 1, right), we can clearly see that these new iso-
meric derivatives (isocoumarins) are much cleaner phar-
maceutical modulators of the ERb subtype than are the
isoflavonoid soy phytoestrogens; at an appropriate dose,
the isocoumarins are capable of fully activating ERb,
with no activating effect on ERa.
4. Experimental
One aspect of our findings is of note: we find that the
ERb potency selectivity (the ratio of EC50 values) of sev-
eral compounds in the isocoumarin series greatly ex-
ceeds their affinity selectivity (the ratio of Ki values).
This is evident, especially, with 12b, 15b, and 18b, where
ERb potency selectivities exceed the affinity selectivities
by more than 100. We previously noted a similar discor-
dance in ERb potency and affinity selectivities in the
indazole compounds on which we reported earlier.10
While we do not have an explanation of this difference
between affinity and potency selectivities for these com-
pounds, it is well appreciated that whereas affinity repre-
sents binding, potency can also depend on cell and
promoter-specific interactions.29 Thus, differences in
the conformation of ERb-ligand and ERa-ligand com-
plexes could have significant effects on their ability to
interact with coactivator proteins that support their
activity; these differences could then lead to the en-
hanced selectivity in terms of potency as opposed to
affinity that we observe. However, the experimental pro-
tocol of the two experiments may account for some of
the differences. The binding affinity is measured in a
competitive binding assay that involves two ligands
(the test compound and estradiol), and is thus depen-
dent upon their respective interaction (association and
dissociation) rates with the receptor protein. The poten-
cy experiment is assayed with the test compound alone
and, in this way, might be a more direct measure of its
potency.
4.1. Materials and methods
All reagents and solvents were obtained from Aldrich or
Fisher. Compound 7 (dehydroequol) was prepared as de-
scribed in the literature.30 Tetrahydrofuran, diethyl ether,
toluene, and dichloromethane were dried by the solvent
delivery system (SDS) (neutral alumina columns) de-
signed by J. C. Meyer.31 Glassware was oven dried,
assembled while hot, and cooled under an inert atmo-
sphere. Unless otherwise stated, all reactions were con-
ducted in an inert atmosphere. Reactions using
moisture- or air-sensitive reagents were performed in
anhydrous solvents. Reaction progress was monitored
using analytical thin-layer chromatography (TLC) on
0.25 mm Merck F-254 silica gel glass plates. Visualization
was achieved by either UV light (254 nm) or potassium
permanganate indicator. Flash chromatography was per-
formed with Woelm silica gel (0.040–0.063 mm) packing.
1H NMR and 13C NMR spectra were obtained on a 400
or 500 MHz instrument. The chemical shifts are report-
ed in parts per million and are referenced to either tetra-
methylsilane or the solvent. Mass spectra were recorded
under electron impact conditions at 70 eV. Melting
points were obtained on a Thomas–Hoover MelTemp
apparatus and are uncorrected. Elemental analyses were
performed by the Microanalytical Service Laboratory of
the University of Illinois. Those final components that
did not give satisfactory combustion analysis gave satis-
factory exact mass determinations and were found to be
at least 96% pure by HPLC analysis.
3. Conclusion
We have synthesized isocoumarin compounds and their
derivatives bearing the same core functionalities as the
isoflavonoid phytoestrogens daidzein (4a), genistein
(4b), coumestrol (5), ( ) equol (6), and dehydroequol
(7), but in an isomeric arrangement. Compared to the
phytoestrogens, some of the isocoumarins have compa-
rable to better ERb affinity selectivities; those with the
highest ERb affinity selectivities are 13b, 25, and 27.
What is most striking, however, is the very high ERb
potency selectivity shown by some of the isocoumarins
in the transcription assays: some show up to two orders
of magnitude higher ERb potency selectivities than
4.1.1. 5-Methoxy-2-trifluoromethansulfonyloxy benzoic
acid methyl ester (8). To a solution of 2-hydroxy-5-meth-
oxy-benzoic acid methyl ester (910 mg, 5 mmol), and
trifluoromenthanesulfonic anhydride (5 mmol, 1 mL)
dissolved in toluene (5 mL), Et3N (3 mmol, 2.09 mL)
was added dropwise at 0 ꢁC. The reaction mixture was
stirred at this temperature for 30 min and then allowed
to warm to rt for 1 h. The reaction mixture was quenched
with NaHCO3 (5% soln) and extracted with EtOAc (3·
20 mL). The organic extracts were dried over Na2SO4
and the solvent removed under vacuum. The crude prod-
uct was purified by flash chromatography (10% diethyl

6538
M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
ether/hexanes) to give a colorless oil (1.4 g, 90% yield).
1H NMR (500 MHz, CDCl3) d 7.53 (d, J = 3.2, 1H),
7.20 (d, J = 9.0, 1H), 7.09 (dd, J = 9.1, 3.3, 1H), 3.95 (s,
3H), 3.85 (s, 3H); MS (EI) m/z 314 (M+, 18).
was stirred at rt for 24 h. If at the end of this time peri-
od, the starting material was not totally consumed, an
additional amount of BBr3 (1.5 mmol) was added, and
the reaction mixture was stirred for other 24 h. The reac-
tion mixture was quenched with water (5 mL) and
extracted with EtOAc (3· 10 mL). The combined organ-
ic layers were dried over Na2SO4, and concentrated
under vacuum. The crude product was purified by flash
chromatography or by re-crystallization.
4.1.2. 5-Methoxy-2-(4-methoxyphenylethynyl)benzoic acid
methyl ester (10). Compound 8 (1.2 g, 3.8 mmol) was
dissolved in Et3N–CH3CN (1:5, 6 mL) and Pd(PPh3)4
(220 mg, 5 mol %), CuI (108 mg, 15 mol %), Bu4NI
(2.8 g, 4.5 mmol), and 4-ethynylanisole (9) (591 lL,
4.5 mmol) were added. The reaction mixture was heated
at reflux for 4 h and then allowed to cool to rt and
quenched with water (10 mL). The mixture was extract-
ed with EtOAc (3· 20 mL), dried over Na2SO4 and the
solvent removed under vacuum. The crude product was
purified by flash chromatography (20%, diethyl ether/
hexanes) to give a pale yellow oil (1.0 g, 75%) that crys-
tallized on standing (mp 65–67 ꢁC). 1H NMR
(400 MHz, CDCl3) d 7.52–7.45 (m, 4H), 7.00 (dd,
J = 8.8, 3.0, 1H), 6.85 (dd, J = 8.7, 2H), 3.94 (s, 3H),
3.80 (s, 3H), 3.77 (s, 3H); 13C NMR (500 MHz, CDCl3)
d 166.06, 159.58, 158.77, 135.11, 132.99, 132.84, 118.28,
116.21, 115.73, 115.01, 113.97, 92.71, 86.96, 55.47,
55.23, 52.21; MS (EI) m/z 296 (M+, 100).
4.1.6. 4-Bromo-7-hydroxy-3-(4-hydroxyphenyl)isocouma-
rin (12b). Compound 12b was purified by flash chroma-
tography (50% EtOAc/hexanes) to give a pale yellow
solid (60% yield) that was re-crystallized in acetone
1
(mp > 300 ꢁC dec.). H NMR (500 MHz, acetone-d6) d
10.56 (s, OH), 10.01 (s, OH), 7.75 (d, J = 8.7, 1H),
7.57 (dd, J = 6.9, 2.1, 2H), 7.50 (d, J = 2.5, 1H), 7.40
(dd, J = 9, 2.8, 1H), 6.86 (dd, J = 6.8, 2, 2H); 13C
NMR (500 MHz, acetone-d6) d 160.52, 158.87, 158.32,
148.90, 131.12, 128.28, 128.12, 124.58, 123.29, 121.18,
114.96, 113.13, 99.84; MS (EI) m/z 333 (M+, 8). HRMS
(EI) calcd for C15H9BrO4, 331.9684; found, 332.9765.
Anal. (C15H9BrO4) C, H.
4.1.7. 7- Methoxy-3-(4-methoxyphenyl)isocoumarin (13a).
Derivative 11 (200 mg, 0.5 mmol) was dissolved in DMF
(6 mL) and formic acid (37 lL, 1 mmol), Et3N (110 lL,
1.5 mmol), PPh3 (10 mg, 8 mol %), and Pd(OAc)2 (4 mg,
4 mol %) were added. The reaction mixture was heated
at 60 ꢁC overnight, then quenched with water and
extracted with diethyl ether (3· 5 mL). The organic ex-
tracts were dried over MgSO4, the solvent concentrated
under vacuum, and the crude product purified by flash
chromatography (50% diethyl ether/hexanes) to give a
pale yellow solid (120 mg, 0.42 mmol, 86% yield) crystal-
4.1.3. 4-Iodo-7-methoxy-3-(4-methoxyphenyl)isocoumarin
(11). Derivative 10 (800 mg, 2.7 mmol) was dissolved
in CH3CN (15 mL), and I2 (2.0 g, 8.1 mmol) was added.
The reaction mixture was stirred at rt for 2 h, then
quenched with
a saturated solution of Na2S2O3
(15 mL), and extracted with EtOAc (3· 20 mL). The
organic extracts were dried over Na2SO4 and the solvent
removed under vacuum. The crude product was purified
by crystallization in EtOAc to give a white solid
(805 mg, 73%, mp 181–182 ꢁC). 1H NMR (400 MHz,
CDCl3) d 7.80 (d, J = 8.8, 1H), 7.71 (d, J = 2.7, 1H),
7.66 (dd, J = 6.8, 2, 2H), 7.35 (dd, J = 9.2, 3.2, 1H), 6.98
(dd, J = 6.8, 2, 2H), 3.94 (s, 3H), 3.87 (s, 3H); 13C NMR
(400 MHz, CDCl3) d 162.08, 160.84, 160.22, 152.70,
133.33, 132.20, 131.79, 127.57, 124.94, 121.09, 113.46,
110.17, 75.59, 56.10, 55.50; MS (EI) m/z 408 (M+, 100).
1
lized in EtOAc (mp 147–149 ꢁC). H NMR (400 MHz,
CDCl3) d 7.77 (dd, J = 8.5,1.5, 2H), 7.67 (s, 1H), 7.37
(d, J = 8.8, 1H), 7.28 (dd, J = 7.5,1.7, 2H), 6.94 (d, J =
7.9, 2.0, 1H), 6.77 (s, 1H), 3.91 (s, 3H), 3.85 (s, 3H);
13C NMR (500 MHz, CDCl3) d 162.78, 160.85, 159.32,
151.92, 131.74, 127.42, 126.56, 124.87, 124.80, 121.26,
114.29, 109.96, 100.13, 55.83, 55.48. MS (EI) m/z 282
(M+, 100).
4.1.4. 4-Bromo-7-methoxy-3-(4-methoxyphenyl)isocoum-
arin (12a). Derivative 10 (700 mg, 2.36 mmol) was dis-
solved in CH3CN (16 mL) and Br2 was added (181 lL,
3.54 mmol). The reaction mixture was stirred at rt for
1 h, then quenched with water (10 mL) and extracted
with EtOAc (3· 20 mL). The organic extracts were dried
over Na2SO4 and the solvent removed under vacuum.
The crude product was purified by crystallization in
EtOAc to give a white solid (700 mg, 82%, mp 151–
4.1.8. 7-Hydroxy-3-(4-hydroxyphenyl)isocoumarin (13b).
Compound 13b was purified by flash chromatography
(50% EtOAc/hexanes) to give a pale yellow solid (65%
yield), that was re-crystallized in ethyl acetate.
Mp > 300 ꢁC dec. (lit.23 mp > 300 ꢁC), 1H NMR
(500 MHz, methanol-d4) d 7.70 (dd, J = 8.8, 2.2, 2H),
7.54 (d, J = 2.6, 1H), 7.47 (d, J = 8.6, 1H), 7.25 (dd,
J = 8.5, 2.5, 1H), 7.00 (s, 1H), 6.86 (dd, J = 8.8, 2.1,
2H); 13C NMR (500 MHz, methanol-d4) d 160.25,
158.84, 152.74, 132.16, 128.98, 127.44, 125.65, 124.91,
122.05, 116.70, 113.78, 101.01. MS (EI) m/z 254 (M+,
25). HRMS (EI) calcd for C15H10O4, 254.0579; found,
254.0581. Anal. (C15H10O4Æ0.2H20) C, H.
1
153 ꢁC). H NMR (500 MHz, CDCl3) d 7.85 (d, J = 9,
1H), 7.76 (dd, J = 6.9, 2.1, 2H), 7.72 (d, J = 2.8, 1H),
7.38 (dd, J = 9, 2.8, 1H), 6.97 (dd, J = 6.9 ,2.1, 2H),
3.93 (s, 3H), 3.87 (s, 3H); 13C NMR (500 MHz, CDCl3)
d 161.63, 160.85, 160.15, 149.79, 131.37, 130.61, 128.45,
125.18, 124.81, 121.62, 113.57, 110.38, 100.55, 56.04,
55.49; MS (EI) m/z 361 (M+, 100).
4.1.9. 4-Chloro-7-methoxy-3-(4-methoxyphenyl)isocoum-
arin (15a). To a suspension of Hg(OAc)2 (318 mg,
1 mmol) in CH3CO2H (3 mL), derivative 10 (300 mg,
1 mmol) was added. The reaction mixture was stirred
at rt for 1 h, and then at 0 ꢁC a saturated solution of
4.1.5. General method for deprotection of methoxy groups
with BBr3. The methyl ether protected compound
(0.5 mmol) was dissolved in CH2Cl2 (1 mL), and BBr3
(1.5 mmol) was added at 0 ꢁC. The reaction mixture

M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
6539
1
NaCl (3 mL) was added. The reaction mixture was al-
lowed reach to rt and stirred for 30 min. The separated
organomercurial was removed by filtration, washed well
with hexane, and dissolved in chloroform. The chloro-
form solution was filtered through silica gel, and the
organic solvent was evaporated. The crude product
(14) was used in the next step without any further
purification. Product 14 was dissolved in THF (3 mL),
and anhydrous CuCl2 (270 mg, 2 mmol) was added.
The reaction mixture was heated at reflux for 4 h, and
then the precipitate was removed by filtration. The
crude product was purified by flash chromatography
(30% EtOAc/hexane) to give a white solid (90 mg,
0.28 mmol, 28% yield), re-crystallized in EtOAc/hexane
144–145 ꢁC). H NMR (500 MHz, CDCl3) d 7.77–7.75
(m, 2H), 7.64 (dd, J = 6.9, 2.1, 2H), 7.31 (dd, J = 8.8,
2.8, 1H), 6.91 (dd, J = 6.8, 2.1, 2H), 6.57 (dd, J = 18,
11.4, 1H), 5.63 (dd, J = 11.4, 1.5, 1H), 5.47 (dd, J = 1.5,
18, 1H), 3.91 (s, 3H), 3.84 (s, 3H); 13C NMR
(500 MHz, CDCl3) d 162.41, 160.30, 159.26, 149.16,
131.22, 131.13, 130.41, 126.26, 125.63, 124.29, 122.63,
121.68, 113.55, 112.65, 110.21, 55.86, 55.41. MS (EI)
m/z 308 (M+, 100).
4.1.13. 4-Ethyl-7-methoxy-3-(4-methoxyphenyl)isoumarin
(18a). Derivative 17 (40 mg, 0.13 mmol), was dissolved
in EtOH (10 mL), and Pd/C (10%, 15 mg) was added.
The reaction mixture was stirred at rt under H2
(1 atm) for 6 h. The catalyst was filtered off and the sol-
vent removed under vacuum to give a white solid
(quant. yield) re-crystallized in EtOAc (mp 142–
144 ꢁC). 1H NMR (400 MHz, CDCl3) d 7.77 (d,
J = 3.2, 1H), 7.57 (d, J = 9.2, 1H), 7.49 (dd, J = 6.8,
2.0, 2H), 7.35 (dd, J = 8.8, 3.2, 1H), 6.96 (dd, J = 6.8,
2.0, 2H), 3.91 (s, 3H), 3.86 (s, 3H), 2.70 (q, J = 7.6,
2H). 1.27 (t, J = 7.6, 3H); 13C NMR (400 MHz, CDCl3)
d 162.88, 160.26, 159.11, 149.42, 131.66, 130.43, 126.01,
125.15, 124.43, 122.43, 114.71, 113.78, 110.36, 55.88,
55.46, 20.37, 14.97. MS (EI) m/z 310 (M+, 65).
1
(mp 155–157 ꢁC). H NMR (500 MHz, CDCl3) d 7.87–
7.83 (m, 3H), 7.74 (d, J = 2.6, 1H), 7.42 (dd, J = 9.0,
2.8, 1H), 6.69 (dd, J = 6.8, 2.1, 2H), 3.94 (s, 3H), 3.87
(s,3H); 13C NMR (500 MHz, CDCl3) d 161.44, 160.87,
160.17, 148.51, 130.96, 130.03, 125.86, 124.85, 123.97,
121.68, 113.75, 113.49, 110.51, 56.05, 55.52. MS (EI)
m/z 316 (M+, 100).
4.1.10. 4-Chloro-7-hydroxy-3-(4-hydroxyphenyl)isocouma-
rin (15b). Compound 15a (20 mg, 0.06 mmol) was depro-
tected according to the procedure described above with
BBr3 in CH2Cl2. Derivative 15b was purified by flash
chromatography (50% EtOAc/hexane) to give a pale
yellow solid essentially pure (11 mg, 60% yield). Mp >
4.1.14. 7-Methoxy-3-(4-methoxyphenyl)-4-phenylisocoum-
arin (19a). Derivative 11 (150 mg, 0.36 mmol) was dis-
solved in DMF (1.5 mL), and Pd(PPh3)4 (5 mol %),
Cs2CO3 (164 mg, 0.5 mmol) and PhB(OH)2 (53.7 mg,
0.44 mmol) were added. The reaction mixture was heated
at 80 ꢁC for 4 h, then quenched with water (10 mL), and
extracted with diethyl ether (3· 10 mL). The organic
extracts were dried over MgSO4 and the solvent removed
under vacuum. The crude product was purified by flash
chromatography (50% diethyl ether/hexane) to give a
white solid (110 mg, 0.30 mmol, 85% yield) re-crystallized
in diethyl ether/hexane (mp 135–137 ꢁC). 1H NMR
(400 MHz, CDCl3) d 7.80 (d, J = 2.6, 1H), 7.43–7.40
(m, 3H), 7.27–7.19 (m, 5H), 7.10 (d, J = 8.8, 1H), 6.70
(dd, J = 6.8, 2.2, 2H), 3.93 (s, 3H), 3.75 (s, 3H); 13C
NMR (500 MHz, CDCl3) d 162.70, 159.33, 149.08,
134.96, 132.96, 131.33, 130.61, 129.25, 128.14, 127.00,
125.49, 124.38, 121.43, 115.81, 113.40, 109.90, 55.94,
55.30. MS (EI) m/z 358 (M+, 100).
1
300 ꢁC dec. H NMR (500 MHz, acetone-d6) d 9.5 (br
s, OH), 9.0 (br s, OH), 7.85 (d, J = 8.8, 1H), 7.72 (dd,
J = 6.7, 2.1, 2H), 7.66 (d, J = 2.6, 1H), 7.50 (dd,
J = 9.0, 2.8, 1H), 6.68 (dd, J = 6.8, 2.1, 2H). MS (EI)
m/z 288 (M+, 100); HRMS (EI) calcd for C15H9ClO4,
288.0189; found, 288.0192. Purity >95% (HPLC).
4.1.11. 7-Methoxy-3-(4-methoxypheny)-4-methylisocoum-
arin (16a). Derivative 11 (150 mg, 0.36 mmol) was
dissolved in dioxane (2 mL), and Pd2(dba)3 (4.5 mol %),
PtBu (18 mol %), CsF (120 mg, 0.79 mmol), and
Sn(CH3)4 (100 lL, 0.72 mmol) were added. The reaction
mixture was heated in a sealed tube at 80 ꢁC for 12 h.
After that, the precipitate was removed by filtration
and the solvent removed under vacuum. The crude prod-
uct was purified by flash chromatography (30% EtOAc/
hexane) to give a white solid (95 mg, 0.32 mmol, 89%
yield) re-crystallized in EtOAc/hexane (mp 157–159 ꢁC).
1H NMR (400 MHz, CDCl3) d 7.76 (d, J = 2.9, 1H),
7.55–7.50 (m, 3H), 7.36 (dd, J = 8.8, 2.9, 1H), 6.95 (dd,
J = 6.8, 2.0, 2H), 3.91 (s, 3H), 3.85 (s, 3H), 2.28 (s,
3H);13C NMR (400 MHz, CDCl3) d 162.92, 160.19,
159.18, 149.33, 132.82, 130.96, 125.82, 125.08, 124.38,
121.86, 113.70, 110.15, 108.40, 55.88, 55.47, 13.81. MS
(EI) m/z 296 (M+, 96).
4.1.15. 7-Methoxy-3-(4-methoxyphenyl)-4-trifluorometh-
ylisocoumarin (20a). Derivative 11 (100 mg, 0.25 mmol)
was dissolved in DMF (1 mL), and FSO2CF2CO2CH3
(149 lL, 1.17 mmol), and CuI (46 mg, 0.25 mmol) were
added. The reaction mixture was heated at 80 ꢁC for
6 h, after which the solid obtained from the reaction
was filtered off, and the solvent removed under vacuum.
The crude product was purified by flash chromatogra-
phy (30% EtOAc/hexane) to give a colorless oil
(70 mg, 0.2 mmol, 81% yield) that solidified on standing
4.1.12. 7-Methoxy-3-(4-methoxyphenyl)-4-vinylisocouma-
rin (17). Derivative 11 (150 mg, 0.36 mmol) was dissolved
in dioxane (2 mL), and Pd2(dba)3 (4.5 mol %), PtBu3
(18 mol %), CsF (120 mg, 0.79 mmol), and vinylSn(Bu)3
(93 lL, 0.36 mmol) were added. The reaction was per-
formed as in the procedure given for 16a, and the crude
product was purified by flash chromatography (30%
EtOAc/hexane) to give a pale yellow solid (85 mg,
0.27 mmol, 76% yield) re-crystallized in EtOAc (mp
1
(mp 115–118 ꢁC). H NMR (500 MHz, CDCl3) d 7.76–
7.73 (m, 2H), 7.50 (dd, J = 6.9, 2.2, 2H), 7.38 (dd,
J = 9.0, 3.0, 1H), 6.98 (dd, J = 6.9, 2.2, 2H), 3.93 (s,
3H), 3.86 (s, 3H); MS (EI) m/z 350 (M+, 29).
4.1.16. 7-Hydroxy-3-(4-hydroxyphenyl)-4-methylisocoum-
arin (16b). Compound 16a (50 mg, 0.17 mmol) was

6540
M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
deprotected according to the procedure described above
with BBr3 in CH2Cl2. After the quenching with water,
the solid product obtained was removed by filtration,
and the filtrate washed with EtOAc, to give a compound
essentially pure (25 mg, 0.09 mmol, 55% yield).
Mp > 300 ꢁC dec. 1H NMR (500 MHz, DMSO-d6) d
10.29 (br s, OH), 9.88 (br s, OH), 7.62 (d, J = 8.8, 1H),
7.51 (d, J = 2.7, 1H), 7.38 (dd, J = 8.5, 1.9, 2H), 7.34
(dd, J = 8.8, 2.8, 1H), 6.86 (dd, J = 8.5, 2.0, 2H), 2.19
(s, 3H);13C NMR (500 MHz, DMSO-d6) d 161.62,
158.17, 157.24, 148.12, 130.72, 130.56, 125.79, 124.02,
123.75, 121.17, 115.10, 112.73, 107.80, 13.31. MS (EI)
m/z 254 (M+, 25). HRMS (EI) calcd for C16H12O4,
268.0736; found, 268.0736. Anal. (C16H12O4Æ0.2H2O) C,
H.
suspended in EtOH (4 mL) and KOH (5% soln, 4 mL)
was added. The reaction mixture was heated under reflux
for 3 h. The reaction mixture was allowed to reach rt and
EtOH removed under vacuum. The aqueous phase was
washed with diethyl ether, acidified with HCl (1M), and
extracted with EtOAc (3· 10 mL). The organic extracts
were dried over Na2SO4 and the solvent removed under
vacuum to give a pale yellow solid (21) pure enough to
be used for the next step without any further purification.
Derivative 21 (380 mg, 1.4 mmol) was dissolved in EtOH
(4 mL), and NaBH4 (310 mg, 8.4 mmol) was added. The
reaction mixture was heated under reflux for 4 h. EtOH
was then evaporated, the residue diluted with water
(10 mL), and acidified with HCl (1 M) to give a precipi-
tate, which was extracted with EtOAc (3· 10 mL). The
organic extracts were dried over Na2SO4 and the solvent
removed under vacuum. The racemic alcohol obtained
was dissolved in acetic anhydride (1 mL) and heated un-
der reflux for 2 h. The reaction mixture was then allowed
to reach rt, quenched with water, and extracted with
CH2Cl2 (3· 10 mL). The combined extracts were washed
with NaOH (5% soln) dried over Na2SO4, and the solvent
removed under vacuum. The crude product was purified
by flash chromatography (60% diethyl ether/hexane) to
give a white solid (160 mg, 0.56 mmol, 40% yield) re-crys-
tallized in EtOAc (mp 138–139 ꢁC). 1H NMR (500 MHz,
CDCl3) d 7.63 (d, J = 2.7, 1H), 7.40 (dd, J = 6.5, 1.7, 2H),
7.19 (d, J = 8.5, 1H), 7.13 (dd, J = 8.3, 2.7, 1H), 6.94 (dd,
J = 6.8, 2.2, 2H), 5.47 (dd, J = 11.9, 3.2, 1H), 3.86 (s, 3H),
3.82 (s, 3H), 3.28 (dd, J = 16.1, 12, 1H), 3.03 (dd,
J = 16.3, 3.1, 1H);13C NMR (500 MHz, CDCl3) d
165.56, 159.76, 159.05, 131.34, 130.65, 128.47, 127.62,
125.88, 121.79, 113.93, 112.88, 80.16, 55.60, 55.29,
34.61. MS (EI) m/z 284 (M+, 26).
4.1.17. 7-Hydroxy-3-(4-hydroxyphenyl)-4-ethylisocouma-
rin (18b). Compound 18b (40 mg, 0.13 mmol) was
deprotected according to the procedure described above
with BBr3 in CH2Cl2. Derivative 18a was purified by
flash chromatography (60% EtOAc/hexane) to give a
pale yellow solid (34 mg, 0.12 mmol, 92% yield) re-crys-
tallized in EtOAc (mp > 300 ꢁC dec.). 1H NMR
(500 MHz, acetone-d6) d 9.10 (br s, OH), 8.75 (br s,
OH), 7.72–7.69 (m, 2H), 7.45 (dd, J = 6.7, 2.1, 2H),
7.41 (dd, J = 8.8, 2.8, 1H), 6.98 (dd, J = 6.7, 2.1, 2H),
2.70 (q, J = 7.5, 2H), 1.25 (t, J = 7.3, 3H);13C NMR
(500 MHz, acetone-d6)
d 162.41, 159.09, 157.90,
149.92, 131.24, 131.04, 126.48, 126.01, 124.43, 123.45,
116.00, 115.04, 114.22, 20.75, 15.04. MS (EI) m/z 282
(M+, 11). HRMS (EI) calcd for C17H14O4, 282.0892;
found, 282.0894. Anal. (C17H14O40.5 H2O) C, H.
4.1.18. 7-Hydroxy-3-(4-hydroxyphenyl)-4-phenylisocoum-
arin (19b). Compound 19b (50 mg, 0.14 mmol) was
deprotected according to the procedure described above
with BBr3 in CH2Cl2. Derivative 19a was purified by flash
chromatography (50% EtOAc/hexane) to give a pale
yellow solid (40 mg, 0.12 mmol, 85% yield) re-crystallized
4.1.21. 7-Hydroxy-3-(4-hydroxyphenyl)-3,4-dehydroiso-
coumarin (22b). Compound 22a (50 mg, 0.17 mmol) was
deprotected according to the procedure described above
with BBr3 in CH2Cl2. Derivative 22b was purified by flash
chromatography (50% EtOAc/hexane) to give a white
solid (30 mg, 0.12 mmol, 70% yield) re-crystallized in
EtOAc (mp > 230 ꢁC dec.). 1H NMR (500 MHz, ace-
tone-d6) d 8.75 (br s, OH), 7.48 (d, J = 2.7, 1H), 7.37
(dd, J = 6.7, 2.2, 2H), 7.23 (d, J = 8.1, 1H), 7.10 (dd,
J = 8.2, 2.6, 1H), 6.87 (dd, J = 6.4, 1.9, 2H), 5.48 (dd,
J = 12, 3.2, 1H), 3.25 (dd, J = 16.3, 12.2, 1H), 3.06 (dd,
J = 16.3, 3.2, 1H);13C NMR (500 MHz, acetone-d6) d
165.56, 158.46, 157.53, 131.62, 131.26, 129.74, 128.80,
127.10, 122.06, 116.10, 116.03, 80.93, 34.87. MS (EI) m/z
256 (M+, 33). HRMS (EI) calcd for C15H12O4, 256.0736;
found, 256.0733. Anal. (C15H12O40.1 H2O) C, H.
1
in EtOAc. H NMR (500 MHz, acetone-d6) d 9.0 (br s,
OH), 7.70 (d, J = 2.6, 1H), 7.46–7.39 (m, 3H), 7.31–7.23
(m, 5H), 7.17 (dd, J = 8.6, 2.5, 1H), 6.68 (dd, J = 6.7,
2.1, 2H); 13C NMR (500 MHz, acetone-d6) d 162.21,
158.67, 158.09, 149.64, 136.06, 132.42, 132.12, 131.41,
129.81, 128.69, 127.84, 125.57, 124.39, 122.41, 116.20,
115.54, 113.83. MS (EI) m/z 330 (M+, 5). HRMS (EI)
calcd for C21H14O4, 330.0892; found, 330.0888. Anal.
(C21H14O40.5 H2O) C, H.
4.1.19. 7-Hydroxy-3-(4-hydroxyphenyl)-4-trifluoromethyl-
isocoumarin (20b). Compound 20a (70 mg, 0.2 mmol) was
deprotected according to the procedure described above
with BBr3 in CH2Cl2. Derivative 20b was purified by flash
chromatography (50% EtOAc/hexane) to give a white
solid (60 mg, 0.18 mmol, 94% yield) re-crystallized in
EtOAc (mp > 230 ꢁC dec.). 1H NMR (500 MHz, ace-
tone-d6) d 9.25 (br s, OH), 7.73–769 (m, 2H), 7.48–7.46
(m, 3H), 6.98 (dd, J = 6.7, 2.2, 2H); MS (ESI) m/z 323
(M+H, 100). HRMS (ESI) calcd for (M+H), 323.0531;
found, 323.0522. Anal. (C16H9F3O4) C, H.
4.1.22. 3-(4-Hydroxyphenyl)isochroman-7-ol (isoequol,
23). Compound 22b (30 mg, 0.12 mmol), was dissolved
in THF (2 mL), and LiAlH4 (6.6 mg, 0.18 mmol) was
added at rt. The reaction mixture was heated under re-
flux for 2 h and then at 0 ꢁC quenched with water
(2 mL). After the evaporation of organic solvent, the
aqueous phase was acidified with H2SO4 (5% soln) and
extracted with EtOAc (3· 5 mL). The organic extracts
were dried over Na2SO4 and the solvent removed under
vacuum. The crude product was purified by flash chro-
matography (40% EtOAc/hexane) to give a white solid
4.1.20. 7-Methoxy-3-(4-methoxyphenyl)-3,4-dehydroiso-
coumarin (22a). Compound 13a (400 mg, 1.4 mmol) was

M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
6541
(20 mg, 0.08 mmol, 71% yield) essentially pure (mp 228–
230 ꢁC dec.). 1H NMR (500 MHz, acetone-d6) d 8.28 (br
s, OH), 8.16 (br s, OH), 7.28 (dd, J = 6.7, 2.1, 2H), 6.98
(d, J = 8.4, 1H), 6.83 (dd, J = 6.8, 2.0, 2H), 6.68 (dd,
J = 8.2, 2.4, 1H), 6.55 (d, J = 1.9, 1H), 4.83 (s, 2H),
4.57 (t, J = 6.5, 1H), 2.82 (d, J = 6.7, 2H); 13C NMR
0.24 mmol) and a catalytic amount of AlCl3 (1 mg,
5 mol %) were added at rt. The reaction mixture was
stirred at rt for 2 h and then heated under reflux for
30 min. The reaction mixture was cooled down to 0 ꢁC
and then quenched with water (2 mL). After the evapo-
ration of organic solvent the aqueous phase was acidi-
fied with HCl (1M, 5 mL) and extracted with EtOAc
(3· 15 mL). The organic extracts were dried over
MgSO4 and the solvent removed under vacuum. The
crude product was purified by flash chromatography
(33% EtOAc/hexane) to give a pale yellow solid (26 mg,
0.11 mmol, 69% yield) essentially pure (mp 169 ꢁC
dec.). 1H NMR (500 MHz, acetone-d6) d 8.44 (br s,
OH), 7.61 (dd, J = 6.8, 2.0, 2H), 6.95 (d, J = 8.8, 1H),
6.85 (dd, J = 6.8, 2.0, 2H), 6.62–6.58 (m, 2H), 6.38 (s,
1H), 5.08 (s, 2H);13C NMR (500 MHz, acetone-d6) d
159.20, 158.27, 155.19, 134.66, 127.54, 126.71, 125.55,
120.11, 115.99, 113.28, 110.76, 99.74, 69.24. MS (EI)
m/z 240 (M+, 100). HRMS (EI) calcd for C15H12O3,
240.0786; found, 240.0792. Anal. (C15H12O3) C, H.
(500 MHz, acetone-d6)
d 157.56, 156.48, 136.72,
134.75, 128.03, 125.29, 115.74, 114.66, 111.20, 77.49,
69.10, 36.13. MS (EI) m/z 242 (M+, 3). HRMS (EI) calcd
for C15H14O3, 242.0943; found, 242.0948. Anal.
(C15H14O30.2 H2O) C, H.
4.1.23. 3-(4-Hydroxyphenyl)-1H-isochromen-7-ol (dehyd-
roisoequol, 25). Compound 13b (60 mg, 0.23 mmol) was
dissolved in THF (2 mL) and LiAlH4 (13.5 mg,
0.35 mmol) was added at rt. The reaction mixture was
stirred at this temperature for 45 min and then at 0 ꢁC
quenched with water (2 mL). After the evaporation of
organic solvent, the aqueous phase was acidified with
H2SO4 (5% soln) and extracted with EtOAc (3· 5 mL).
The organic extracts were dried over Na2SO4 and the
solvent removed under vacuum. The crude product
was purified by flash chromatography (50% EtOAc/hex-
ane) to give a white solid (20 mg, 0.08 mmol, 35% yield)
essentially pure (mp > 200 ꢁC dec.). 1H NMR
(500 MHz, acetone-d6) d 8.40 (br s, OH), 8.20 (br s,
OH), 7.57 (dd, J = 6.7, 2.1, 2H), 6.96 (d, J = 8.2, 1H),
6.84 (dd, J = 6.8, 2.3, 2H), 6.72 (dd, J = 8.2, 2.2, 1H),
6.64 (d, J = 1.9, 1H), 6.39 (s, 1H), 5.08 (s, 2H); 13C
NMR (500 MHz, acetone-d6) d 158.71, 157.02, 152.53,
130.72, 127.07, 127.02, 125.24, 125.02, 115.94, 115.45,
111.92, 99.67, 69.28. MS (EI) m/z 240 (M+, 40). HRMS
(EI) calcd for C15H12O3, 240.0786; found, 240.0789.
Anal. (C15H12O3Æ0.1 H2O) C, H.
4.1.26. Estrogen receptor binding affinity assays. Relative
binding affinities were determined by a competitive
radiometric binding assay as previously described,32
using 10 nM [3H]estradiol as tracer (Amersham BioSci-
ences, Piscataway, NJ), and purified full-length human
ERa and ERb (PanVera/InVitrogen, Carlsbad, CA).
Incubations were for 18–24 h at 0 ꢁC; the receptor–
ligand complexes were absorbed onto hydroxyapatite
(BioRad, Hercules, CA), and the unbound ligand was
washed away. The binding affinities are expressed as rel-
ative binding affinity (RBA) values, with the RBA of
estradiol being set at 100. The values given are the aver-
age range or SD for two or more independent deter-
minations. Estradiol binds to ERa with a Kd of
0.2 nM and to ERb with a Kd of 0.5 nM.
4.1.24. 3-(4-Hydroxyphenyl)-4-methyl-1H-isochromen-7-ol
(26). Compound 16b (60 mg, 0.22 mmol) was dissolved
in THF (2 mL) and LiAlH4 (13.5 mg, 0.35 mmol) was
added at RT. The reaction mixture was stirred at this
temperature for 1 h and then heated under reflux for
2 h. After that, the reaction mixture was cooled down
to 0 ꢁC and then quenched with water (2 mL). After
the evaporation of organic solvent, the aqueous phase
was acidified with H2SO4 (5% soln) and extracted with
EtOAc (3· 5 mL). The organic extracts were dried over
Na2SO4 and the solvent removed under vacuum. The
crude product was purified by flash chromatography
(50% EtOAc/hexane) to give a white solid (15 mg,
0.06 mmol, 27% yield) essentially pure (mp > 200 ꢁC
dec.). 1H NMR (500 MHz, acetone-d6) d 8.75 (br s,
OH), 8.25 (br s, OH), 7.36 (dd, J = 6.6, 1.9, 2H), 7.07
(d, J = 8.4, 1H), 6.85 (dd, J = 6.7, 1.9, 2H), 6.78 (dd,
J = 8.1, 2.3, 1H), 6.65 (d, J = 2.3, 1H), 4.98 (s, 2H),
2.09 (s, 3H); 13C NMR (500 MHz, acetone-d6) d
158.32, 156.99, 149.66, 132.00, 127.77, 127.36, 127.01,
123.29, 115.40, 115.21, 111.56, 107.38, 69.06, 14.42.
MS (ESI) m/z 255 (M+H, 100). HRMS (ESI) calcd
for C16H14O3, 254.0943; found, 254.0958. Anal.
(C16H14O4Æ0.2H2O) C, H.
4.1.27. Cell culture and transient transfections. Human
endometrial cancer (HEC-1) cells were maintained in
minimum essential medium (MEM) plus phenol-red
supplemented with 5% calf serum and 5% fetal calf ser-
um. Cells were plated in phenol-red-free Improved
MEM and 5% charcoal dextran-treated calf serum
(CDCS) and were given fresh medium 24 h before trans-
fection. Transfection assays were performed in 24-well
plates using 0.35 mL of serum-free OptiMEM medium
and 0.15 mL of Hankꢀs balanced salt solution containing
5 lL of lipofectin (Life Technologies, Inc., Gaithers-
burg, MD), 1.6 lg of transferrin (Sigma, St. Louis,
MO), 200 ng of pCMV b-galactosidase as internal con-
trol, 1 lg of 2xERE-pS2-Luc, and 100 ng of ER expres-
sion vector per well. The cells were incubated at 37 ꢁC in
a 5% CO2-containing incubator for 5 h. The medium
was then replaced with fresh Improved MEM supple-
mented with 5% CDCS plus the desired concentrations
of ligands. Cells were harvested 24 h later. Luciferase
and b-galactosidase activity were assayed as described.33
Acknowledgments
4.1.25. 3-(4-Hydroxyphenyl)-1H-isochromen-6-ol (27).
Compound 24 (40 mg, 0.16 mmol) was dissolved in
THF (3 mL), and LiAlH4 (240 lL of 1 M solution,
This work was supported through grants from the
National Institutes of Health (PHS 5R37 CA25836

6542
M. De Angelis et al. / Bioorg. Med. Chem. 13 (2005) 6529–6542
12. Muthyala, R. S.; Ju, Y. H.; Sheng, S.; Williams, L. D.;
Doerge, D. R.; Katzenellenbogen, B. S.; Helferich, W. G.;
Katzenellenbogen, J. A. Bioorg. Med. Chem. 2004, 12,
1559–1567.
13. Kelly, J. E.; Joannou, G. E. Novogen Rosearch Pty Ltd
2003, US Patent 6,649,648 B641.
and P01 AG024387A 4 to J.A.K., and PHS 5R01
CA19118 to B.S.K.). We are grateful to Kathryn Carl-
son for binding assays and for comments on the manu-
script. NMR spectra were obtained in the Varian
Oxford Instrument Center for Excellence in NMR Lab-
oratory. Funding for the instrumentation was provided
in part by the W.M. Keck Foundation, the National
Institutes of Health (PHS 1 S10 RR104444-01) and the
National Science Foundation (NSF CHE 96-10502).
Mass spectra were obtained on instruments supported
by grants from the National Institute of General Medi-
cal Sciences (GM 27029), the National Institutes of
Health (RR 01575), and the National Science Founda-
tion (PCM 8121494).
14. Rossi, R.; Carpita, A.; Bellina, F.; Stabile, P.; Mannina,
L. Tetrahedron 2003, 59, 2067–2081.
15. Sun, J.; Baudry, J.; Katzenellenbogen, J. A.; Katzenel-
lenbogen, B. S. Mol. Endocrinol. 2003, 17, 247–258.
16. Manas, E. S.; Unwalla, R. J.; Xu, Z. B.; Malamas, M. S.;
Miller, C. P.; Harris, H. A.; Hsiao, C.; Akopian, T.; Hum,
W. T.; Malakian, K.; Wolfrom, S.; Bapat, A.; Bhat, R. A.;
Stahl, M. L.; Somers, W. S.; Alvarez, J. C. J. Am. Chem.
Soc. 2004, 126, 15106–15119.
17. Dai, W. M.; Guo, D. S.; Sun, L. P. Tetrahedron Lett. 2001,
42, 5275–5278.
18. Oliver, M. A.; Gandour, R. D. J. Org. Chem. 1984, 49,
558–559.
References and notes
19. Biagetti, M.; Bellina, F.; Carpita, A.; Stabile, P.; Rossi, R.
Tetrahedron 2002, 58, 5023–5038.
20. Napolitano, E.; Spinelli, G.; Fiaschi, R.; Marsili, A.
Synthesis 1985, 38–40.
21. Napolitano, E.; Giannone, E.; Fiaschi, R.; Marsili, A.
J. Org. Chem. 1983, 48, 3653–3657.
22. Tuanli, Y.; Larock, R. C. J. Org. Chem. 2003, 68, 5936–
5942.
1. Turner, R. T.; Rickard, D.; Spelsberg, T. C.; Sibonga, J. D.
In Osteoporosis; Humana Press, 2002, pp 309–330.
2. Mendelsohn, M. E. Am. J. Cardiol. 2002, 89, 12E–17E.
3. Behl, C. Nat. Rev. Neurosci. 2002, 3, 433–442.
4. Mosselman, S.; Polman, J.; Dijkema, R. FEBS Lett. 1996,
392, 49–53.
5. Kuiper, G. G.; Enmark, E.; Pelto-Huikko, M.; Nilsson, S.;
Gustafsson, J. A. Proc. Natl. Acad. Sci. U.S.A. 1996, 93,
5925–5930.
23. Rose, A.; Buu-Hoi, P.; Jacquignon, P. J. Chem. Soc. 1965,
6100–6104.
24. Mahto, R. P.; Ghosh, P. K.; Sarkhel, B. K. Asian J. Chem.
1999, 11, 431–435.
25. Nagarajan, A.; Balasubramanian, T. R. Indian J. Chem. B
1987, 26B, 917–919.
6. Pike, A. C. W.; Brzozowski, A. M.; Hubbard, R. E.;
Bonn, T.; Thorsell, A.-G.; Engstrom, O.; Ljunggren, J.;
Gustafsson, J.-A.; Carlquist, M. EMBO J. 1999, 18, 4608–
4618.
26. Littke, A. F.; Schwarz, L.; Fu, G. C. J. Am. Chem. Soc.
2002, 124, 6343–6348.
27. Chen, Q.-Y. J. Fluorine Chem. 1995, 72, 241–246.
28. Zamani, K.; Rama, N.; Iqbal, R. J. Heterocycl. Chem.
2000, 37, 1651–1654.
29. Katzenellenbogen, J. A.; OꢀMalley, B. W.; Katzenellenb-
ogen, B. S. Mol. Endocrinol. 1996, 10, 119–131.
30. Heaton, A.; Kumar, N. In PCT/AU00/00103, WO 00/
49009; Ltd., N. R. P., Ed., 2000.
7. Weihua, Z.; Andersson, S.; Cheng, G.; Simpson Evan, R.;
Warner, M.; Gustafsson, J.-A. FEBS Lett. 2003, 546, 17–
24.
8. Paruthiyil, S.; Parmar, H.; Kerekatte, V.; Cunha, G. R.;
Firestone, G. L.; Leitman, D. C. Cancer Res. 2004, 64,
423–428.
9. Meyers, M. J.; Sun, J.; Carlson, K. E.; Marriner, G. A.;
Katzenellenbogen, B. S.; Katzenellenbogen, J. A. J. Med.
Chem. 2001, 44, 4230–4251.
31. Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen,
R. K.; Timmers, F. J. Organometallics 1996, 15, 1518–1520.
32. Carlson, K. E.; Choi, I.; Gee, A.; Katzenellenbogen, B. S.;
Katzenellenbogen, J. A. Biochemistry 1997, 36, 14897–
14905.
33. McInerney, E. M.; Tsai, M. J.; OꢀMalley, B. W.; Katzen-
ellenbogen, B. S. Proc. Natl. Acad. Sci. U.S.A. 1996, 93,
10069–10073.
10. De Angelis, M.; Stossi, F.; Carlson, K. E.; Katzenellenb-
ogen, B. S.; Katzenellenbogen, J. A. J. Med. Chem. 2005,
48, 1132–1144.
11. Harris, H. A.; Albert, L. M.; Leathurby, Y.; Malamas, M.
S.; Mewshaw, R. E.; Miller, C. P.; Kharode, Y. P.;
Marzolf, J.; Komm, B. S.; Winneker, R. C.; Frail, D. E.;
Henderson, R. A.; Zhu, Y.; Keith, J. C., Jr. Endocrinology
2003, 144, 4241–4249.