6H-Benzo[c]chromen-6-one derivatives as selective ERβ agonists

Article abstract of DOI:10.1016/j.bmcl.2005.12.057

A series of 6H-benzo[c]chromen-6-one and 6H-benzo[c]chromene derivatives were prepared, and the affinity and selectivity for ERα and ERβ was measured. Many of the analogs were found to be potent and selective ERβ agonists. Bis hydroxyl at positions 3 and 8 is essential for activity in a HTRF coactivator recruitment assay. Additional modifications at both phenyl rings led to compounds with ERβ < 10 nM potency and >100-fold selectivity over ERα.

Full text of DOI:10.1016/j.bmcl.2005.12.057

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1468–1472
6H-Benzo[c]chromen-6-one derivatives as selective ERb agonists
a,
a
b
b
b
*
Wanying Sun, Lovji D. Cama, Elizabeth T. Birzin, Sudha Warrier, Louis Locco,
a
a
b
Ralph Mosley, Milton L. Hammond and Susan P. Rohrer
a
Department of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Department of Atherosclerosis and Endocrinology, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
b
Received 10 November 2005; revised 5 December 2005; accepted 13 December 2005
Available online 18 January 2006
Abstract—A series of 6H-benzo[c]chromen-6-one and 6H-benzo[c]chromene derivatives were prepared, and the affinity and selectiv-
ity for ERa and ERb was measured. Many of the analogs were found to be potent and selective ERb agonists. Bis hydroxyl at
positions 3 and 8 is essential for activity in a HTRF coactivator recruitment assay. Additional modifications at both phenyl rings
led to compounds with ERb < 10 nM potency and >100-fold selectivity over ERa.
ꢀ
2006 Elsevier Ltd. All rights reserved.
Until 1996, estrogens were assumed to mediate their ef-
fects through a single nuclear receptor (now called
ERa). The discovery of a new estrogen receptor named
OH
1
2
ERb, in 1996, followed by tissue distribution studies
HO
HO
O
O
which showed that the expression of the two ERs is
not coincident in tissues, has led to renewed interest in
estrogen receptor modulators, especially compounds
that are selective for the two subtypes ERa and ERb.
X-ray crystal structures of receptor–ligand complexes
Effusol
3-Hydroxy-6H-
benzochromen-6-one
Figure 1.
3
of ERa and ERb show that the binding pockets of
the two receptors are very similar and differ by only
two amino acids. The leucine present at position 384
in ERa is replaced by a methionine in ERb (Met 336)
and the methionine in position 421 is replaced by an
isoleucine in ERb (Ile373). In light of this it is not
surprising that most previously known estrogens and
SERMs bind fairly non-selectively to the two receptors.
However, recent reports have described compounds that
considered unsuitable. Substituting the dihydrophe-
nanthrene by a 6H-benzo[c]chromen-6-one retained the
geometry of the dihydrophenanthrene without the possi-
bility of its oxidation to a phenanthrene. This publica-
tion describes the synthesis and SAR of such 6H-
benzo[c]chromen-6-one derivatives as selective ERb
agonists. Compounds with a similar core structure with
4
bind selectively to ERa and ERb.
6
SERM activity have been reported recently.
Our search for ERb agonists started with the natural
product Effusol which has a simple dihydrophenanth-
rene structure (Fig. 1). This compound bound to ERb
with an IC50 = 12 nM and was 20· selective. However,
the dihydrophenanthrene ring system, because of its
easy oxidation to the phenanthrene analog was
Most of the benzo[c]chromenone analogs were prepared
7
by a procedure of Bruggink and McKillop as shown in
8
Scheme 1.
5
Treatment of a resorcinol (2 equiv) with a substituted
bromobenzoic acid, catalyzed by CuSO in the presence
4
of 2 equiv of NaOH in water at 100 ꢁC, gave the desired
product in most cases as a solid, which was isolated by
filtration, washing with water, and drying to give the
clean product. The yields varied from 10% to 60%
depending on the substitution of the bromobenzoic acid.
Keywords: Selective estrogen receptor modulator; ERb agonist;
6
H-Benzo[c]chromen-6-one.
Corresponding author. Tel.: +1 7325944765; fax: +1 7325949556;
e-mail: wanying_sun@merck.com
*
0
960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.057

W. Sun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1468–1472
1469
OH
R9
R10
R10
OH
R9
10
R8
R1
R10
R
6
6'
1
c (R ,R =H)
R
6
6'
H
R6'
R8
H
d, e (R =R =alkyl)
+
Br
a
R7
HO
HO
O
O
HO
O
O
HO
HO
OH
R7
R6
HOOC
R4
HO
O
O
R4
R10
R10
OH
H
OH
H
R9
R9
1
0
10
R
R
f, g
1
OCH3
OH
R
1
R
R7
b
HO
OCH3
R7
O
O
R10
OH
H
O
O
HO
O
O
R4
R4
h
Scheme 1. Synthesis of 6H-benzo[c]chromen-6-one Reagents and
conditions: (a) 2 equiv, 5 N NaOH, 10% aq CuSO
, 100 ꢁC; (b)
BBr , CH Cl , rt.
R6
HO
O
4
R10
3
2
2
R10
OH
OR
i, j
8
-Hydroxy analogs were prepared by treatment of the
HO
O
R6
corresponding 8-methoxy compounds with BBr3.
RO
O
O
R=pivalate
Alternatively (Scheme 2) a suitably substituted aryl
boronic acid was coupled to a 2-bromo-benzoic acid
ester in the presence of (Ph P) Pd catalyst and Na CO
in EtOH/DME to give the corresponding biphenyl.
Removal of the methyl-protecting groups with BBr3
gave the desired products.
Scheme 3. Synthesis of 6H-benzo[c]chromenes. Reagents and condi-
6
3
4
2
3
tions: (c) BF
benzene, 80 ꢁC, 2 h; (e) BF –Et O, benzene; (f) 1.3 equiv DIBAL, THF
3 2 4
–Et O, NaBH , THF, Rfx, 1 h; (d) 10 equiv R MgX,
3
2
6
ꢀ78 ꢁC; (g) MeOH, 2 N HCl; (h) 4 equiv R MgX, benzene, 80 ꢁC, 2 h;
6
(i) 10 equiv R MgX, C
6
H
6
3
, 80 ꢁC, overnight; (j) 5 equiv Et SiH,
1
0 equiv TFA, CH Cl , rt, 1 h.
2
2
7
As shown in Scheme 3, when R was H, 6,6-unsubstitut-
ed 6H-benzo[c]chromenes were prepared by reduction of
H-benzo[c]chromene-6-ones with boron trifluoride
addition only. The resulting 6-carbinol was reduced to
the desired product with TFA and Et SiH.
6
9
etherate–sodium borohydride. 6-Monosubstituted 6H-
benzo[c]chromenes were synthesized by reducing the
lactones to lactols with DIBAL in THF followed by
treatment with MeOH and aq HCl to give the 6-meth-
oxy-6H-benzo[c]chromenes. Reaction of this intermedi-
ate with various Grignard reagents in benzene gave the
corresponding 6-monosubstituted 6H-benzo[c]-chrom-
3
The ERb affinities and selectivity over ERa of the ana-
logs were measured by a competitive binding assay.
The assay results are depicted in Table 1.
1
1
Compounds 23 and 25, close analogs of Effusol, which
have the methyl and the vinyl groups in positions 4
and 10, and a substituent on position 8, show binding
affinity and selectivity similar, or better than that of
Effusol, showing that the 6H-benzo[c]chromen-6-one is
an excellent substitute for the dihydrophenanthrene ring
system.
10
enes. 6,6-Disubstituted 6H-benzo[c]chromenes were
prepared by treatment of the lactone with excess of cor-
responding Grignard reagents followed by treatment
with boron trifluoride etherate to effect dehydrative
9
7
cyclization of the intermediate carbinol. When R was
CH , the dipivalate of the chromenone was reacted with
excess Grignard reagent and gave the product of mono
3
The importance of the 4-methyl group is shown by com-
parison of 6 versus 20 (H vs Me) and 3 versus 7 (Me vs
Et). The Me group at position 4 appears to be extremely
important for affinity and selectivity.
R1
R10
B(OH)2
Br
R9
a
+
We discovered that position 7 needs a substituent such
as Me, Et or bromo to improve the ERb-binding affini-
ties as well as the selectivity. Compounds 1 and 4; 11
and 2; 15 and 8; 28 and 27; 29 and 30 are five pairs of
compounds that show more potency and ERb
selectivity when Me or Br replaces H on position 7.
An ethyl group is only slightly poorer than methyl at
this position (30 vs 36).
H CO
OCH3
H COOC
OCH3
3
3
R4
R7
R9
R9
R10
OCH3
R7
10
R1
R
OH
R7
1
R
b
COOCH3
H CO
OCH3
3
HO
O
O
R4
R4
SAR study of position 8 revealed that along with the hy-
droxy at position 3, a hydroxy substituent here improves
activity (1 vs 29). However, a different polar group, ami-
no, lowers the binding affinity and selectivity (1 vs 12).
Scheme 2. Alternative synthesis of 6H-benzo[c]chromene-6-ones.
Reagents and conditions: (a) (Ph P) Pd, Na CO , EtOH, DME,
0 ꢁC, 2–24 h; (b) BBr , CH Cl , 0ꢁ–rt, 1–3 h.
3
4
2
3
8
3
2
2

1470
W. Sun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1468–1472
Table 1. Substituted 3-hydroxy-6H-benzo[c]chromene-6-ones
R9
R10
R8
R7
HO
O
O
R4
4
7
8
9
10
Compound
R
R
R
R
R
IC50 (nM)
ERa
ERa/ERb
ERb
1.2
Estradiol
Effusol
1.35
240
1.1
20
12
93
1
2
3
4
5
6
7
8
9
Me
Me
Me
Me
Me
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Br
H
H
Me
Br
H
>10,000
>10,000
>10,000
>10,000
4590
107
32
OMe
Br
H
308
128
1450
310
90
H
H
78
H
H
H
7
15
H
Vinyl
Me
Br
H
H
Me
H
585
5460
6.5
2.3
30
Et
H
2340
326
208
116
117
279
433
24
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
H
Me
vinyl
Propen-1-yl
H
>10,000
>10,000
3030
H
H
48
26
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
H
H
OMe
NH
Me
Me
Me
Br
Me
Me
H
1000
14,300
4258
85
51
2
H
NHSO
2
CH
3
H
9.8
25
H
H
H
597
3074
Me
H
80
38
10
NHCHO
Br
957
215
77
>10,000
6850
H
32
H
OMe
Br
Me
H
10,000
1090
3710
129
42
H
Buten-1-yl
H
H
26
20
H
Me
Br
184
71
H
Me
Me
H
44
35
3120
H
H
Allyl
Vinyl
Et
10,000
10,000
9830
285
1000
142
188
175
258
150
H
Me
H
10
69
H
OMe
OMe
Me
H
H
H
Vinyl
Me
Me
Me
53
57
>10,000
>10,000
4160
H
Br
OMe
OMe
H
Br
Me
Me
16
1.15
225
Increasing the acidity of the NH by substitution with
electron-withdrawing groups (13 and 16) gave poorly ac-
tive compounds because that position is unable to sus-
tain a group much larger than methyl (3, 5, and 11).
Compound 29 with a hydroxyl group showed very
potent affinity (IC50 = 4.1 nM) with good selectivity.
IC50 = 1276 nM, in the presence of serum, 142-fold-
increase in IC ), and were inactive in a HTRF
(homogeneous time-resolved fluorescence) coactivator
recruitment assay (23, EC50 > 1000 nM). In contrast,
29 with two hydroxyl groups showed much less serum
binding (IC50 = 78 nM in the presence of serum, 20-fold
increase) and has an EC50 of 7.2 nM in the coactivator
recruitment assay.
5
0
1
2
Positions 9 and 10 seem to have enough room for at
least a four-carbon chain: 19 and 22 have IC50 = 26
and 35 nM. Some of the most selective compounds in
Table 1 (21–28) are substituted at the 10-position. Com-
pound 23 with a vinyl at position 10 gave quite potent
binding affinity (IC50 = 10 nM) with very high selectivity
Table 2 presents the binding affinities for a series of
compounds with OH at positions 3 and 8. There is pseu-
do-symmetry in these molecules which could allow
either of the 2-hydroxy groups to bind at the same site
as the phenolic group of estradiol and the only difference
would be the way the lactone is oriented, Figure 2. Thus
the SAR of these compounds has to be considered with
this in mind. Three pairs (compounds 30 and 31, 33 and
42, 43 and 44), which show this pseudo-symmetry, have
comparable binding and selectivity. Compound 29 is an
internally symmetric compound with good activity,
while compounds 36 and 37 in which the positions of
Me and Et are interchanged have similar activity and
selectivity. Substitution at 4 and 7 is essential for good
(1000-fold). Compound 28 with five substituents on the
aromatic rings also has excellent binding and selectivity.
In general, it is noticeable that increasing the number of
small hydrophobic substituents on the phenyl rings gives
better ERb binding affinities with significant increase in
selectivity. Among the more substituted compounds, 23,
27, and 28 have outstanding binding and selectivity.
However, most of the compounds in Table 1 showed
large serum binding, exemplified by 23 (ERb

W. Sun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1468–1472
1471
Table 2. 3,8-Dihydroxy-6H-benzo[c]chromene-6-ones
R9
Table 3. Substituted 3-hydroxy-4-methyl 6H-benzo[c]chromenes
R10
OH
R10
R1
OH
7
R
R7
6R6'
HO
O
R
HO
O
O
R4
0
6
Compound R ,R
6
7
10
R
R
IC50 (nM)
ERb ERa
88 4610 52
1030 14
ERa/ERb
1
4
7
9
10
Compound
R
R
R
R
R
IC50 (nM)
ERb ERa
4.1 159 38.5
46 654 14
ERa/ERb
4
4
5
5
5
5
5
5
5
5
5
5
8
9
0
1
2
3
4
5
6
7
8
9
H, H
H, H
Me
H
H
Me 74
Me 24
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
H
H
H
Me Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
F
H
H
H
H
H
Me, H
Me, H
Et, H
Et, H
Pr, H
Me, H
Et, H
H
1210 50
335 42
101 32
435 54
626 29
237 26
129 57
627 13
468 10
157 12
Me
H
H
Me
Me
H
H
8
Me
H
49
27
67
1170 24
488 18
1860 28
2521 143
1280 16
484 35
229 29
1580 29
228 11
1920 29
1780 20
583 11
785 115
716 124
11,600 65
524 28
597 27
H
Me
3.2
8
Me Me
Me
H
H
H
H
H
H
H
H
H
H
Me
H Me 22
Me Me 9
Me Me 2.3
Cl Cl
Me OH
Me Et
Et Me
Me 18
H
H
H
H
H
H
Et
82
14
Isobutyl, H
Me, Me
Me, allyl
H
Me
Me 49
H 45
7.8
F
F
Me
Et
54
21
67
91
Me Me 13
Br Me
Me
Me
H
H
Me 54
Me Me Me
H
6.7
5.7
180
19
H
H
H
H
Me Me
Cl Cl
Me
H
Cl Me Cl
H
Cl Me Cl Me 22
A
B
9
2
OH
OH
1
0
1
1
A
4
C
10
C
A
O
2
7
9
4
B
O
B
O
HO
O
HO
7
Figure 2.
binding and selectivity (29, 36, and 37). Again additional
substitution at the 1 or 10 positions gives better selectiv-
ity (43 and 44). Analogs 43 and 44 possess high ERb
binding affinity and selectivity. Molecular modeling
studies described below are in accord with these obser-
vations and provide a rationale for the high selectivity
of 43 and 44.
Figure 3. Molecular modeling of compound 44 (white) against
estradiol (green). hERa is depicted in green and hERb in purple.
Residue numbering is hERa unless otherwise indicated.
(56) showed comparable binding activity and selectivity
to 6H-benzo[c]chromene-6-one (29).
1
3
Analogs with halogen substituents in this series have less
binding affinities than their corresponding methyl ana-
logs (44 vs 34, 29 vs 38, 36 vs 39, and 29 vs 40). Halogens
at the 9-position reduce activity (45 vs 34 and 46 vs 29).
A hydroxy group at position 7 as in 35 has poorer bind-
ing affinity its corresponding methyl analog compound
The docking and energy minimization approach ,
which will be described elsewhere, identified a binding
mode for compounds 29, 43, and 44 (Fig. 3) consistent
between the crystallographically determined ligand
1
4
binding domains of human ERa (1ERE) and ERb
1
4
(1QKM) in which the tricyclic core of the chromenone
derivatives spans the length of the steroid estradiol and
the lactone ring is oriented in the binding pocket to map
to the ‘B’ ring of the steroid away from helix 12. The
high selectivity of these compounds for hERb is presum-
ably due, in part, to the ability of the aromatic ring at
the ‘C’ ring position to better interact with the Met
336 side chain in hERb. In the case of compound 29
in which the phenols are symmetrically substituted and
either can mimic the estradiol ‘A’ ring position, energet-
ics between each receptor and the two different binding
2
9.
Table
benzo[c]chromenes. In general, 6-monosubstituted
3
presents the binding affinities for 6H-
analogs are more active than the 6-unsubstituted and
-di-substituted derivatives (compare 48, 51 and 58).
6
Among the 6-monosubstituted compounds 50–57, the
ethyl group gave the best binding affinities and selectiv-
ity. Compound 56, 53 and 52 are better than their close
methyl analogs 55, 50 and 51. The 6H-benzo[c]chromene

1472
W. Sun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1468–1472
orientations slightly favor (by ꢁ0.5 kcal/mol) that in
U.S.A. 1996, 93, 5925; (b) Mosselman, S.; Polman, J.;
Dijkema, R. FEBS Lett. 1996, 392, 49.
. Taylor, A. H.; Al-Azzawi, F. J. Mol. Endocrinol. 2000, 24,
which the 3-phenol is equivalent to the estradiol phenol
Structure A, Fig. 2) and the lactone carbonyl is directed
2
3
(
1
45.
toward His 524 (hERa numbering). For compounds 43
. (a) Brzozowski, A. M.; Pike, A. C. W.; Dauter, Z.;
Hubbard, R. E.; Bonn, T.; Engstrom, O.; Ohman, L.;
Greene, G. L.; Gustafsson, J.-A.; Carlquist, M. Nature
and 44, substituted by a CH at the 1 or the 10 position,
respectively, the energetics between each receptor and
the two different binding orientations clearly indicate
3
1
997, 389, 753; (b) Pike, A. C. W.; Brzozowski, A. M.;
(
by ꢁ15 kcal/mol) that the phenol which is least substi-
Hubbard, R. E.; Bonn, T.; Thorsell, A. G.; Engstrom, O.;
Ljunggren, J.; Gustafsson, J.-A.; Carlquist, M. EMBO J.
1999, 18, 4608.
tuted interacts with the Glu/Arg residue pair at the top
of the binding cavity. This appears to be primarily due
to a steric clash between the added CH and residues
4. Malamas, M. S.; Manas, E. S.; McDevitt, R. E.;
Gunawan, I.; Xu, Z. B.; Collini, M. D.; Miller, C. P.;
Dinh, T.; Henderson, R. A.; Keith, J. R., Jr.; Harris, H.
A. J. Med. Chem. 2004, 47, 5021, and references sited
herein..
. Birzin, E. T.; Witherup, K.; Mosley,R.; Locco, L.;
Warrier, S.; Yudkovitz, J.; Hayes, E.; Alves, S.; Ho, K.
L.; Dahllund, J.; Nilsson, S.; Zamora, N.; Tamayo-
Castillo, G.; Nanakorn, W.; Annable, C. R.; Beck, H.
T.; Steven, D. W.; Wilkinson, H. A.; Borris, R.; Goetz,
M.; Rohrer, S. P. J. Biol. Chem., submitted for
publication.;
3
in helix 3 particularly Ala 350.
The additional CH has another impact on the structure
3
as well. The energy minimized structures of compounds
not substituted at the 1 or 10 positions are essentially pla-
nar; the dihedral angle C1–C–C–C10 (where C–C is the
biphenyl bond) in these is 0.4ꢁ. Conversely, substitution
at positions 1 or 10 results in a twist about this same cen-
tral bond with a concomitant pucker of the lactone ring
5
presumably due to a steric clash between the CH and the
3
6
. Pandey, J.; Jha, A. K.; Hajela, K. Bioorg. Med. Chem.
proton on the adjacent aromatic. For CH substitution
3
2
. Bruggink, A.; McKillop, A. Tetrahedron 1975, 31, 2607.
004, 12, 2239.
at position 1 (compound 43), the dihedral angle about
C1–C–C–C10 is ꢁ11.5ꢁ and at position 10 (compound
7
8
. All new compounds were characterized by LC–MS and
5
. Devlin, J. P. Can. J. Chem. 1975, 53, 343.
4
4) ꢁ13.5ꢁ. In the models of the compounds 43 and 44
1
00 or 600 MHz H NMR.
docked into hERa and hERb, it appears that the twist
and pucker of the lactone results in a closer and presum-
ably repulsive interaction between the lactone and Leu
9
1
0. Zhi, L.; Tegley, C. M.; Edwards, J. P.; West, S. J.;
Marshke, K. B.; Gottardis, M. M.; Jones, T. K. Bioorg.
Med. Chem. Lett. 1998, 8, 3365.
3
84 in hERa than is seen with Met 336 in hERb.
1
1. The IC50 values were generated in an estrogen
receptor ligand binding assay. This scintillation assay
was conducted in NEN Basic Flashplates using triti-
ated estradiol and full-length recombinant human ERa
and ERb proteins. Most of the data reflect a 3 h
incubation period and were evaluated in duplicate in a
single assay. In our experience, this assay provides
IC50 values that are reproducible to within a factor of
We have shown that properly substituted 6H-
benzo[c]chromenenones and 6H-benzo[c]chromenes
yield very selective ERb ligands which are active in a
HTRF coactivator recruitment assay. Small hydropho-
bic groups, methyl and ethyl in positions 4, 7, and 10
are required for optimum binding and selectivity. Hy-
droxy groups at 3 and 8 positions are essential for activ-
ity in the HTRF coactivator recruitment assay.
2
–3.
1
2. Zhou, G.; Cummings, R.; Li, Y.; Mitra, S.; Wilkinson, H.;
Elbrecht, A.; Hermes, J. D.; Schaeffer, J. M.; Smith, R. G.;
Moller, D. E. Mol. Endocrinol. 1998, 12, 1594.
13. A distance-dependent dielectric model of 2r was employed
for all minimizations. For those within the context of a
receptor structure, the receptor was held fixed except for
6
H-Benzo[c]chromenes are also active if they bear a
methyl or ethyl group at the 6-position. The best com-
pounds (29, 37, 43, 44, 51, 52, 53 and 56) have an
IC50 < 10 nM and a selectivity over ERa ranging from
3
0 to 124·.
˚
side chains which fall within 5 A of the modeled ligand
which were allowed to minimize in conjunction with the
ligand. All minimizations were conducted using the
MMFFs force field: Halgren, T. A. J. Comput. Chem.
1999, 20, 730.
References and notes
1
. (a) Kuiper, G. G. J. M.; Enmark, E.; Pelto-Kuikko, M.;
Nilsson, S.; Gustafsson, J. A. Proc. Natl. Acad. Sci.
14. Protein Databank entry code.