7-Substituted 2-phenyl-benzofurans as ERβ selective ligands

Article abstract of DOI:10.1016/j.bmcl.2004.07.029

Substituted benzofuran 21 is a potent and selective ER ligand. A series of 2-(4-hydroxy-phenyl)-benzofuran-5-ols with relatively lipophilic groups in the 7-position of the benzofuran was prepared and the affinity and selectivity for ERβ was measured. Many

Full text of DOI:10.1016/j.bmcl.2004.07.029

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4925–4929
7-Substituted 2-phenyl-benzofurans as ERb selective ligands
a
a
c
Michael D. Collini,a, David H. Kaufman, Eric S. Manas, Heather A. Harris,
*
Ruth A. Henderson,c Zhang B. Xu,b Rayomand J. Unwallaa and Chris P. Millera,
aChemical and Screening Sciences, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
bChemical and Screening Sciences, Wyeth Research, 200 Cambridge Park Drive, Cambridge, MA 02140, USA
cWomenÕs Health Research Institute, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
Received 9 June 2004; accepted 13 July 2004
Available online 4 August 2004
Abstract—A series of 2-(4-hydroxy-phenyl)-benzofuran-5-ols with relatively lipophilic groups in the 7-position of the benzofuran
was prepared and the affinity and selectivity for ERb was measured. Many of the analogues were found to be potent and selective
ERb ligands. Additional modifications at the benzofuran 4-position as well as at the 30-position of the 2-phenyl group were found to
further increase selectivity. Such modifications led to compounds with <10nM potency and >100-fold selectivity for ERb.
Ó 2004 Elsevier Ltd. All rights reserved.
The estrogen receptor (ER) is a member of the nuclear
receptor family, which functions as a ligand activated
transcription factor. Binding of an ER agonist such as
17b-estradiol to the receptor leads to gene regulation
in various tissues.1 Many tissues throughout the body
are affected by estrogens, especially mammary gland,
bone and uterine tissues. It has been shown that ERa
mediates many of the well documented activities associ-
ated with estrogens on these tissues.2 Since the discovery
of ERb in 1996,3 there has been much interest in finding
potent and selective ligands for this receptor.4 While the
functional characterization of ERb is still under active
investigation, the recent report that an ERb selective ag-
onist demonstrated potent and efficacious action in two
rat models of inflammation confirmed the potential of
ERb as a viable drug target.5 In view of these recent
findings as well as the promise of yet more to come,
the desire for additional potent and selective ERb lig-
ands remains strong.
types equally well. Among the initial reported selective
ligands for ERb reported were isoflavones such as geni-
stein 1, which has a binding affinity (IC50) for ERb of
10nM and a selectivity for ERb over ERa of 41-fold
(Table 1). Many ER ligands like genistein have two
OH groups, which can overlay with the OH groups of
estradiol. In addition, genistein also contains an OH
group at the 5-position of the chromenone moiety. It
has been reported that an OH group at this position is
expected to form an intramolecular hydrogen bond with
the C-4 carbonyl group.7 This increases the combined
effective lipophilicity of the two groups,8 allowing these
polar substituents to occupy a relatively hydrophobic
region of the ER binding pocket.6
2-(4-Hydroxy-phenyl)-benzofuran-5-ol 2 is a compound
found to have a binding affinity (IC50) for ERb of 6nM
and a selectivity of 30-fold. Co-crystallization studies of
Although the ligand binding domains of ERa and ERb
share only 56% homology, the ligand binding cavities
differ by only two amino acids (ERa Leu384 ! ERb
Met336; ERa Met421 ! ERb Ile373).6 Many of the
known ER ligands such as 17b-estradiol bind both sub-
OH
OH
OH O
HO
O
HO
17β-Estradiol
1
7
3'
*
Corresponding author. Tel.: +1-484-865-3852; fax: +1-484-865-
9399; e-mail: collinm2@wyeth.com
O
OH
HO
4
Present address: Corp. Intellectual Property, Glaxo SmithKline, 709
Swedeland Road, King of Prussia, PA 19406, USA.
2
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.07.029

4926
M. D. Collini et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4925–4929
Table 1. Binding affinities (IC50) for human ERa and ERb ligand binding domain (mean+SD)
X
Y
HO
OH
O
R
Example
R
X
Y
ERb IC50 (nM)
ERa IC50 (nM)
Fold selectivity for ERb
17b-Estradiol
Genistein 1
3.6 ( 1.6) n = 144
10 ( 4) n = 79
6 ( 2) n = 32
2 n = 1
3.2 ( 1) n = 144
395 ( 181) n = 80
176 ( 76) n = 31
1.5 n = 1
1
41
30
9
2
H
Br
H
H
H
F
6
H
7
Br
H
0.35 ( 0.4) n = 2
10 ( 3) n = 5
10 ( 2) n = 6
0.5 ( 0.2) n = 10
3.3 ( 1.4) n = 5
2.5 ( 0.7) n = 3
32 ( 14) n = 2
1.6 n = 1
12 ( 3) n = 2
36
49
99
50
99
45
51
4
8
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
Cl
H
H
F
483 ( 152) n = 5
990 ( 380) n = 6
21 ( 5) n = 11
335 ( 151) n = 4
114 ( 32) n = 3
1650 ( 537) n = 2
6.8 ( 2.5) n = 2
114 ( 22) n = 3
109 ( 25) n = 2
1152 ( 558) n = 8
1056 ( 141) n = 4
209 ( 40) n = 5
625 ( 383) n = 4
103 ( 4) n = 2
49 ( 11) n = 2
263 ( 27) n = 4
46 ( 0.6) n = 7
23 ( 5) n = 2
9
H
10
11
12
13
14
19
20
21
22
23
24
27
28
29
30
31
Br
Br
Cl
CN
H
H
F
H
H
H
H
F
CH3
CH3
H
5.6 ( 1) n = 3
4.0 ( 1) n = 2
14 ( 3) n = 8
10 ( 3) n = 5
2.0 ( 1) n = 5
7.9 ( 6) n = 4
6.0 ( 1) n = 2
3.2 ( 0) n = 2
6.6 ( 2) n = 4
2.2 ( 1) n = 9
1.1 ( 0) n = 2
20
26
80
108
104
79
17
15
40
21
22
H
CH2CN
CH2CN
CH2CN
CH2CN
COCH3
COEt
CHO
CN
H
H
F
H
Br
Cl
H
H
H
H
H
H
H
F
H
H
H
CN
H
this ligand complexed to ERb indicated that the benzofu-
ran 7-position would overlay well with the genistein 5-
position when bound to ERb.9 However, 2 lacks any
group that could mimic the genistein 5-OH group, which
we felt was contributing to the modest selectivity of geni-
stein via interactions with ERa Met421/ERb Ile373.10
Therefore, to increase the potency and selectivity of com-
pound 2, a series of benzofurans was synthesized that
contained relatively lipophilic groups at the benzofuran
7-position including halogen, alkyl, cyano, methoxy,
CH2CN, or various carbonyl containing groups. The 7-
methoxy or 7-acetonitrile benzofurans were further sub-
stituted with halogens or cyano at the 4-position because
docking studies showed that these groups would help di-
rect the groups at the 7-position toward ERa Met421/
ERb Ile373 and reduce residual motion in the binding
pocket (cpds 10–13, 23–24). In addition, several com-
pounds with a fluoro group at the 30-position of the 2-
phenyl group were prepared and tested.
When pyridine hydrochloride was used instead of pyr-
idine hydrobromide, the bromo was displaced by chloro
to give 14. The methoxy compounds 8 and 9 were made
from a CuBr coupling of 6 or 7 with NaOCH3. Bromi-
nation with NBS or chlorination with NCS led to com-
pounds 10–12. Treatment of compound 8 with POCl3/
DMF gave the 4-formyl group, which was converted
to an oxime and dehydrated to give the nitrile com-
pound 13.
Benzofurans synthesized by the Sonogashira coupling
(Scheme 2) start with methyl 4-methoxysalicylate which
is iodinated to give methyl 2-iodo-4-methoxysalicylate
and then coupled by a Sonogashira type reaction
between either 4-methoxy-phenylacetylene or 3-fluoro-
4-methoxyphenylacetylene to give the benzofuran 15
or 16. Reduction followed by BBr3 demethylation led
to benzyl bromides 17 or 18. Hydrogenation of 17 or
18 led to the methyl analogues 19 or 20, while NaCN
displacement gave the CH2CN compounds 21 or 22.
Treatment of 21 with NBS or NCS gave the 4-halo
substituted compounds 23 or 24. Deprotection of 15
or 16 with pyridine hydrochloride converted the
methoxy groups to phenols and the ester to a carboxylic
acid. The acid was coupled with N,O-dimethylhydroxyl
amine using EDCI to give the Weinreb amide 25 or 26.
Addition of methyl or ethyl Grignard reagents gave
ketones 27 or 28. Reduction of 25 or 26 with LiAlH4
gave the formyl analogue 29 or the fluoro derivative,
which was converted to a cyano group by forming the
oxime followed by dehydration with pyridine hydro-
chloride at 200°C to give the nitriles 30 or 31.
The compounds were prepared either by the reaction se-
quences shown in Schemes 1 or 2. The synthesis in
Scheme 1 starts with 3-bromo-2,5-dimethoxy-benzyl
alcohol, which can be made by a three step procedure.11
The acid chloride 3 is prepared via a four step procedure
from the benzyl alcohol by chlorination with SOCl2,
conversion to the nitrile and hydrolysis to give the carb-
oxylic acid, which is then converted to the acid chloride
by SOCl2. A Friedel–Craft reaction between an anisole
and 3 with AlCl3 gave a 1,2-diphenyl-ethanone 4 or 5,
which was then deprotected with pyridine hydrobro-
mide at high temperature to give compounds 6 or 7.

M. D. Collini et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4925–4929
4927
OH
O
Table 1 presents the binding affinities (IC50) for this ser-
ies of benzofurans for the human ERb and ERa ligand
binding domains.12 It is clear that the groups incorpo-
rated at the 7-position of compound 2 are well tolerated
in ERb. The ERb binding affinities for most of the com-
pounds were at least as potent as 2 with some of the ana-
logues, for example, 7 and 10, having an IC50 < 1nM.
Interestingly, the ERb binding affinity appears to be rel-
atively insensitive to the identity of the 7-substituent for
this series of compounds. For example, the 7-bromo, 7-
chloro and 7-cyano compounds 6, 14, and 30 were all
potent, with an IC50 < 3nM each, while the 7-methoxy
8, and 7-acetonitrile 21 were slightly less potent.
O
Cl
O
a, b, c, d
O
3
O
e
Br
Br
R1
O
HO
OH
O
f
R1
O
O
Br
6 R1 = H
7 R1 = F
O
4 R1=H
5 R1=F
Br
g
k
R1
HO
HO
OH
OH
O
O
Much more noticeable SAR is revealed when consider-
ing the binding selectivity for ERb relative to ERa.
For example, the ERb selectivities of the 7-chloro com-
pound 14 and 7-bromo compound 6 are only 4-fold and
9-fold, respectively, so selectivity has actually been re-
duced compared to the parent benzofuran compound
2. In contrast, groups such as 7-methoxy and 7-acetonit-
rile exhibit significant improvements in ERb selectivity
relative to 2: the 7-methoxy compound 8 is 49-fold selec-
tive for ERb while the 7-acetonitrile compound 21 is 80-
fold selective. The 7-cyano compound 30 was more po-
tent, but not as selective as the 7-acetonitrile compound
21. Finally, the 7-carbonyl derivatives 27–29 were also
potent, but the ketones exhibited decreased ERb selec-
tivity, while the formyl group displayed selectivity simi-
lar to compound 2. The above SAR is consistent with
the results our docking calculations, which indicate that
groups at the 7-position are likely to occupy the pocket
near ERa Met421/ERb Ile373, as intended. Therefore,
we are probing a region of the binding site that is differ-
ent when comparing the two receptor isoforms, and thus
significant variations in selectivity are likely.
Cl
14
h
O
8 R1 = H
9 R1 = F
X
R1
i, j
HO
OH
O
CN
HO
O
10 R1=H, X= Br
11 R1=F, X=Br
12 R1=H, X=Cl
OH
O
13
O
Scheme 1. Reagents and conditions: (a) SOCl2, THF; (b) NaCN,
DMF; (c) H2SO4, AcOH, H2O; (d) SOCl2, CH2Cl2; (e) anisole or
2-fluoroanisole, AlCl3; (f) pyridine hydrobromide; (g) NaOCH3 CuBr;
(h) NBS or NCS, CH3CN; (i) POCl3, DMF; (j) NH2OH then Burgess
reagent; (k) pyridine hydrochloride.
O
R1
O
O
O
a, b
O
O
OH
15 R1=H
16 R1=F
O
O
h, i
R1
c, d
R1
HO
Since the 7-methoxy compound 8 and 7-acetonitrile
compound 21 were the most selective compounds in
the series so far, we decided to incorporate substituents
at the benzofuran 4-position, as described above. The 4-
bromo, 7-methoxy derivative 10 exhibited improved ER
binding affinity, but there was no increase in selectivity
relative to ERa, compared with 8. A similar result was
seen when the bromo was replaced by a chloro 12 or a
cyano 13. In contrast, 21 was halogenated at the 4-posi-
tion, and the 4-bromo, 7-acetonitrile compound 23
exhibited a modest gain in both ERb binding affinity
(from 14 to 2nM) and selectivity (from 80-fold to 104-
fold). The 4-chloro, 7-acetonitrile compound 24 did
not gain as much in ERb binding affinity, and had no
increase in ERb selectivity.
OH
O
O
HO
25 R1=H
26 R1=F
O
N
OH
O
17 R1=H
18 R1=F
Br
j or k
e or f
R1
HO
R1
OH
HO
O
OH
27 R1=H, R2=CH3
28 R1=H, R2=Et
29 R1=H, R2=H
O
O
R2
R2
19 R1=H,R2=CH3
20 R1=F,R2=CH3
21 R1=H,R2=CH2CN
22 R1=F,R2=CH2CN
l, m
g
R1
X
HO
HO
OH
OH
The final group of compounds tested were the fluoro
analogues 7, 9, 11, 20, 22, and 31. The addition of the
fluoro group ortho to the phenolic OH group led to
modest increases in selectivity for all the compounds ex-
cept for the CN compound 31. Selectivity enhancement
due to such ortho-fluoro substitution has been reported
previously by our laboratories.10 The most selective
fluoro compound was the 7-acetonitrile 22, which had
an ERb IC50 of 10nM and a selectivity of 108-fold rel-
ative to ERa. Compounds 22 and 23 were the most
O
O
30 R1=H
31 R1=F
23 X=Br
24 X=Cl
CN
CN
Scheme 2. Reagents and conditions: (a) I2, KOH, MeOH; (b)
Pd(PPh3)2Cl2, CuI, HNEt2/DMF, 4-methoxy-phenylacetylene or 3-
fluoro 4-methoxyphenylacetylene; (c) LiAlH4, THF; (d) BBr3, CH2Cl2;
(e) H2Pd/C, MeOH; (f) NaCN, DMF; (g) NBS or NCS, CH3CN; (h)
pyridine hydrochloride 200°C; (i) EDCI, CH3NHOCH3, DMF; (j)
CH3MgI or CH3CH2MgI, THF; (k) LiAlH4, THF; (l) NH2OH; (m)
pyridine hydrochloride.

4928
M. D. Collini et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4925–4929
selective compounds in this series. An interesting com-
pound would be a derivative of 23, which contained
the 30-fluoro group to ascertain whether the selectivity
gains seen in 22 and 23 compared with 21 would be
additive in the 7-acetonitrile series and would be a
promising basis for future work.
To assess whether the compounds prepared were agon-
ists or antagonists, the compounds were tested in a
cell-based transcriptional assay and their effect on IG-
FBP4 (insulin-like growth factor binding protein-4)
mRNA levels were assessed and compared to that of
17b-estradiol. Estradiol is able to upregulate IGFBP4
in SAOS-2 human osteosarcoma cells via ERb.2 Exam-
ples were tested at 1lM and 17b-estradiol was used at
10nM. As shown in Table 2, all compounds tested reg-
ulated IGFBP4 mRNA almost to the same extent as
17b-estradiol.
Figure 1. X-ray structure of ERb co-crystallized with 21 (colored by
atom type), overlaid with an X-ray of ERb/genistein, only key residues
shown for simplicity.
Examples 10, 11, and 21 were tested in vivo for their
ability to increase uterine wet weight, a standard estro-
genic bioassay. Sexually immature mice were dosed sub-
cutaneously for 3 or 4days with 50mg/kg of compound
in a vehicle of 5% DMSO/95% corn oil. In contrast to
the reference estrogen, which increased organ weight
4-fold, examples 10, 11, and 21 did not significantly
increase uterine weight. As the rodent uterus expresses
primarily ERa, these data suggest these compounds
are functionally selective for ERb in vivo.
ERb selectivity of 21 relative to 2. It is also instructive
to compare the selectivities of 21 and 30. Placing a meth-
ylene spacer between the benzofuran ring and the nitrile
leads to a compound with twice the ERb selectivity,
which is also consistent with the fact that the methylene
spacer allows the nitrile group to penetrate more deeply
into the ERb Ile373/ERa Met421 pocket. The exact
mechanism by which the 7-acetonitrile and other groups
enhance ERb selectivity via specific interactions with
these residues will be addressed in other venues.9
In order to further understand the mechanism of ERb
selectivity, compound 21 was co-crystallized with
ERb.9 Figure 1 shows the resulting structure overlaid
with that of ERb complexed with genistein, so that the
binding of these two compounds can be compared. It
is clear from these structures that the 7-acetonitrile does
indeed occupy a position analogous to that of the geni-
stein 5-OH group, confirming our earlier hypothesis and
the results of our docking studies. It can be seen from
Figure 1 that the 7-acetonitrile group of compound 21
and the 5-OH of genistein are both in close proximity
to ERb Ile373, which is substituted by Met421 in
ERa. This observation is consistent with the enhanced
In conclusion, placing various groups at the 7-position
of 2-(4-hydroxy-phenyl)-benzofuran-5-ol 2 to mimic
the 5-OH of genistein resulted in compounds that were
more potent as well as more selective than genistein. A
variety of groups led to potent binding affinities for
ERb, but the 7-methoxy and 7-acetonitrile groups re-
sulted in the greatest enhancement in selectivity for
ERb over ERa. Incorporating additional groups at the
4-position and/or a fluoro group ortho to the 2-phenyl
hydroxyl was found to further increase selectivity, and
led to the most selective compound of the series (22),
with an ERb IC50 of 10nM and 108-fold selectivity rel-
ative to ERs. Crystallography and molecular modeling
studies9 demonstrated that the benzofuran 7-position
provides access to ERa Met421/ERb Ile373, thus play-
ing a key role in the enhancement of ERb selectivity.
Table 2. Effect of examples on ERb mediated-IGFBP4 mRNA
regulation in SAOS-2 cells
Example (1 lM)
% Activity of 10nM
17b-estradiol
Acknowledgements
1
105
126
80
2
We thank the Wyeth Discovery Analytical Chemistry
department for the spectral data and Drs. Jay Wrobel
and Baihua Hu for their help in preparing this paper.
10
11
20
21
22
23
24
27
28
29
30
31
100
120
105
130
100
113
120
100
122
142
90
References and notes
1. (a) Kuiper, G. G. J. M.; VandenBemd, G. J. C. M.; Van
Leeuwen, J. P. T. M. J. Endocrinol. Invest. 1999, 22, 594;
(b) Ritzen, E. M.; Nilsson, O.; Grigelioniene, G.; Holst,
M.; Savendahl, L.; Wroblewski, J. J. Steroid Biochem.
Mol. Biol. 2000, 74, 383.

M. D. Collini et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4925–4929
4929
2. Harris, H. A.; Katzenellenbogen, J. A.; Katzenellenbogen,
B. S. Endocrinology 2002, 143, 4172.
3. Kuiper, G. G.; Enmark, E.; Pelto-Huikko, M.; Nilsson, S.;
Gustafsson, J. A. Proc. Nat. Acad. Sci. U.S.A. 1996, 93,
5925.
7. (a) Kozerski, L.; Kamienski, B.; Kawecki, R.; Urbanczyk-
Lipowska, Z.; Bocian, W.; Bednarek, E.; Sitkowski, J.;
Zakrzewska, K.; Nielsen, K. T.; Hansen, P. E. Org.
Biomol. Chem. 2003, 1, 3578; (b) Nishiyama, T.; Ogura,
K.; Nakano, H.; Kaku, T.; Takahashi, E.; Ohkubo, Y.;
Sekine, K.; Hiratsuka, A.; Kadato, S.; Watabe, T. Drug
Metabol. Pharmacokinet. 2002, 3, 221.
8. (a) Ferte, J.; Kuhnel, J. M.; Chapuis, G.; Rolland, Y.;
Lewin, G.; Scwaller, M. A. J. Med. Chem. 1999, 42, 478;
(b) Michalak, K.; Hendrich, A. B.; Wesolowska, O.; Pola,
A. Cell Biol. Mol. Lett. 2001, 6, 362.
9. Manas, E. S. et al., submitted for publication.
10. (a) Edsall, R. J.; Harris, H. A.; Manas, E. S.; Mewshaw,
R. E. Bioorg. Med. Chem. 2003, 11, 3457; (b) Yang, C.;
Edsall, R.; Harris, H. A.; Zhang, X.; Manas, E. S.;
Mewshaw, R. E. Bioorg. Med. Chem. 2004, 12, 2553.
11. Evano, G.; Schaus, J. V.; Panek, J. S. Org. Lett. 2004, 6,
525.
4. (a) Meyers, M. J.; Sun, J.; Carlson, K. E.; Katzenellen-
bogen, B. S.; Katzenellenbogen, J. A. J. Med. Chem. 1999,
42, 2456; (b) Meyers, M. J.; Sun, J.; Carlson, K. E.;
Gwendolyn, M. A.; Katzenellenbogen, B. S.; Katzenel-
lenbogen, J. A. J. Med. Chem. 2001, 44, 4230; (c)
Schopfer, U.; Schoeffter, P.; Bischoff, S. F.; Nozulak, J.;
Feuerbach, D.; Floersheim, P. J. J. Med. Chem. 2002, 45,
1399.
5. 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.
6. Pike, A. C. W.; Brzozowski, A. M.; Hubbard, R. E.;
Bonn, T.; Thorsell, A.; Engstrom, O.; Ljunggren, J.;
Gustafsson, J.; Carlquist, M. EMBO J. 1999, 18, 4608.
12. The binding assay is described in: Harris, H. A.; Bapat, A.
R.; Gonder, D. S.; Frail, D. F. Steroids 2002, 67,
379.