B-Ring unsaturated estrogens: Biological evaluation of 17α- Dihydroequilein and novel B-Nor-6-thiaequilenins as tissue selective estrogens

Article abstract of DOI:10.1016/j.bmcl.2003.09.060

The pharmacology and SAR of representative equine estogens is described. 17α-Dihydroequilenin was found to prevent bone loss after 5 weeks of oral administration to ovariectomized rats. The stereochemical significance of the D-ring and the C/D ring junctu

Full text of DOI:10.1016/j.bmcl.2003.09.060

Bioorganic & Medicinal Chemistry Letters 13 (2003) 4281–4284
B-Ring Unsaturated Estrogens: Biological Evaluation of 17ꢀ-
Dihydroequilein and Novel B-Nor-6-thiaequilenins as Tissue
Selective Estrogens
Charles W. Lugar, III, David Magee, Mary D. Adrian, Pamela Shetler, Henry U. Bryant
and Jeffrey A. Dodge*
Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN 46285, USA
Received 25 August 2003; revised 22 September 2003; accepted 27 September 2003
Abstract—The pharmacology and SAR of representative equine estogens is described. 17a-Dihydroequilenin was found to prevent
bone loss after 5 weeks of oral administration to ovariectomized rats. The stereochemical significance of the D-ring and the C/D
ring juncture was investigated with a series of benzothiophene-based equilenin analogues.
# 2003 Published by Elsevier Ltd.
The clinical use of conjugated equine estrogens (CEEs)
provides beneficial health outcomes for treating hot
flashes and preventing osteoporosis in post-menopausal
women. However, concerns regarding the risks1 for
women taking equine estrogens has prompted the
search for more selective therapeutic agents particularly
with respect totheir uterine safety prfiole. Because
CEEs are a complex mixture of estrogen-like steroids,
we have evaluated individual representative compounds
1–4 of this mixture and herein report the tissue selectiv-
ity profile on bone, lipids, and uterus for one such
component, 17a-dihydroequilenin (17a-DHEqn, 4). We
have also prepared a series of isosteric benzothiophene-
based equilenin analogues in order to investigate their
therapeutic potential as tissue selective estrogens.
hypertrophy in ovary-intact rats.2 All of the individual
equine components caused an increase in uterine wet
weight with the exception of 4 which lacked a significant
effect at the highest dose tested. More recently, the sul-
fate ester conjugate of 17a-DHEqn has been shown to
lower serum cholesterol,3 increase hippocampal den-
dritic spine density in rats,4 and improve arterial vaso-
motor function in macaques.5 Neuroprotective effects
for this compound have also been observed.6 The
cumulative tissue profile demonstrated in these studies,
in which estrogen-like pharmacology is observed on the
cardiovascular and central nervous system whereas
CEEs contain sulfate esters of two structural classes of
estrogen that include ring B saturated steroids such as
17b-estradiol as well as ring B unsaturated estrogens
such as equilin (1, Eq), equilenin (2, Eqn), 17b-dihy-
droequilenin (3, 17b-DHEqn), and 17a-dihy-
droequilenin (4, 17a-DHEqn) among others. In 1991,
Bhavnani and co-workers examined a number of ring B
unsaturated steroids in their unconjugated form in
order to determine their relative binding affinities for
the estrogen receptor and their in vivo effects on uterine
*Corresponding author. Tel.:+1-3172-768-252; fax:+1-3172-772-035;
e-mail: jd@jilly.com
0960-894X/$ - see front matter # 2003 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2003.09.060

4282
C.W.Lugar, III et al./ Bioorg.Med.Chem.Lett.13 (2003) 4281–4284
estrogen-neutral effects are observed on the uterus,
prompted us to explore the pharmacology of 17a-
DHEqn on uterus, bone, and lipids in order to char-
acterize its overall tissue selectivity profile.
with the sulfate ester conjugate of 4 in OVX rats have
shown that this compound reduces total cholesterol
without inducing uterine growth.3 However, the phar-
macology of unconjugated equine estrogens has not
been reported in OVX rats to our knowledge. As shown
in Figure 1, Eq (1) elicits the most pronounced agonist
response at 10 mg/kg (mpk), a 263% increase in uterine
wet weight relative to OVX control, followed by Eqn (2,
168%), 17b-DHEqn (3, 150%), and 17a-DHEqn (4,
100%). Thus, 4 is the least uterotrophic of the com-
pounds that we examined based on rank order of uterine
weight gain at the high dose (10 mpk). The trend we
observe for 1–4 is consistent with data previously repor-
ted in ovary-intact rats although in these studies 17a-
DHEqn did not induce uterine weight gain after daily
subcutaneous administration of 2 mg/kg for 3 days.2
Initially, representative equine estrogens (1–4) were
evaluated in rodents in order to determine their effects
on the uterus. While the pharmacology of CEEs on
uterine tissue has been well-characterized in ovary-
intact rodents,2 we elected to use ovariectomized (OVX)
rats as we have found this model to be a sensitive
method for evaluating the agonist characteristics of
estrogens and selective estrogen receptor modulators
(SERMs) on bone, lipids, and uterus.7 Previous studies
In order to assess whether 4 antagonizes the effects of
estrogen in the uterus it was evaluated in ovary-intact
rats that were co-administered with test compound and
ethynyl estradiol (EE2, 0.1 mpk) for three days. As
shown in Table 1, the known pure antagonist ICI
182,7808 effectively blocks the stimulatory response of
EE2 in a dose dependent manner. 17a-DHEqn (4)
shows no inhibitory activity and, in fact, causes a slight
but statistically significant increase in uterine weight at
the 10 mpk dose. Taken together, this data demon-
strates that 4 increases uterine weight in OVX rats and
does not antagonize the effects of estogen-induced uter-
ine weight gain in immature rats.
Figure 1. Uterine weight gain in OVX rats after oral exposure to var-
ious concentrations of equine estrogens. Points are mean (ÆSEM)
uterine weight (n=5–6 animals/group). *=p<0.05 versus OVX by
one way analysis of variance and Fisher’s posthoc analysis.
We subsequently examined the effects of longer term
administration of 4 on bone, uterus, and lipids in the
OVX rat.9 Oral treatment with 17a-DHEqn for five
weeks prevents bone loss in the proximal metaphysis of
excised femurs in a dose-dependent manner as shown in
Table 2. In addition, 4 efficaciously lowers total serum
cholesterol with an ED50 of <0.1 mpk. Uterine wet
weight is significantly increased by 100% and 142%
relative to OVX controls at doses of 1 and 10 mpk,
respectively. Thus, the efficacy demonstrated by 17a-
DHEqn as an estrogen agonist in preventing bone loss
and lowering cholesterol is observed only at doses at
which agonist effects are also observed in the uterus as
determined by induction of uterine weight gain.
Table 1. Effects on estrogen antagonism for 17a-DHEqn (4) and ICI
182780 expressed as the ability of test compound to block the stimu-
latory effects of EE2 in immature female rats
Compd
Dose (mpk)
% Inhibitiona
ICI 182780
0.10
1.00
10.0
40.3Æ7.2*
100Æ4.2*
117Æ1.1*
4
0.10
1.00
10.0
3.4Æ8.2
14.7Æ3.3
À22.8Æ8.1*
*=p<0.05 vs EE2 stimulated control. Statistical evaluations were
made by one way analysis of variance (ANOVA).
In order to identify compounds with an improved uter-
ine profile relative to 4 we chose to explore structure–
activity studies on the annulated cyclopentane ring of
a% Inhibition=[uterine weightEE2Àuterine weighttest/uterine weight-
weightcontrol]Â100. Negative signs indicate an increase rela-
EE2 Àuterine
tive to vehicle control.
Table 2. Effects of 17a-DHEqn (4) on uterine weight, serum cholesterol, and bone in 6-month old OVX rats after 5 weeks of oral dosing
Compd
Dose
(mpk)
Uterine weight
% increase versus OVX
Serum cholesterol
% decrease versus OVX
Bone (femur)
% protectiona
EE2b
0.10
213.5Æ12.6*
83.3Æ6.6*
90.9Æ17.8*
4
0.01
0.10
1.00
10.0
17.2Æ9.2
23.1Æ7.5
9.8Æ6.8
66.3Æ4.2*
82.6Æ4.3*
93.8Æ1.2*
6.1Æ17.2
26.5Æ20.1
59.5Æ14*
100.4Æ6.8*
142.3Æ14.4*
119.0Æ7.9*
*pꢀ0.05, statistical evaluations were made by one way analysis of variance (ANOVA).
aBone density was determined at the proximal metaphysis of excised femur. Percent protection was calculated for individual animals by the fol-
lowing formula: % protection=[BMDtestÀBMDovx/(BMDshamÀBMDovx)]Â100.
bEE2, ethynylestradiol.

C.W.Lugar, III et al./ Bioorg.Med.Chem.Lett.13 (2003) 4281–4284
4283
the equilenins. The biological effects of modifying the
C-17 hydroxy group of CEEs and other estrogens has
been well documented.2 We were specifically interested
in how alterations at the C/D ring juncture with respect
to the C-17 hydroxyl group might influence receptor
affinity and, ultimately, uterine pharmacology. How-
ever, due to the synthetic complexity involved in the
enantioselective preparation of such derivatives, we
chose to investigate the benzothiophene, or B-nor-5-
thiaequilenin, scaffold. The benzothiopene moiety is an
established A/B-ring surrogate for estrogen receptor
ligands10 and was readily accessed, albeit in racemic
form, via existing synthetic technology developed by
Crenshaw.11 The asymmetric preparation of the thiae-
quilenin scaffold was accomplished by modification of
Crenshaw’s procedure as shown in Schemes 1 and 2 in
which tetrahydro-7-methyl-indene-1,5-dione (5 or 15) is
used to establish the absolute stereochemistry of the
quaternary angular methyl group.12 For the analogues
maintaining this angular methyl group in the b con-
figuration (Scheme 1), epoxidation of 5 followed by
condensation with m-methoxythiophenol yielded 6 as a
mixture of diastereomers.10 Cyclization of the dione was
achieved under Friedel–Crafts conditions with AlCl3 to
yield 7 along with <10% of the 5-methoxy regioisomer
which was separated by crystallization. X-ray crystal-
lography of 7 confirmed the absolute stereochemistry.
Stereoselective hydride reduction of the ketone from the
a-face gave 8 which was subsequently reduced by
hydrogenation of the olefin using 5% palladium on
carbon to give a mixture of diastereomers, 9 and 10,
that were readily separated by silica gel chromato-
graphy. Catalyst selection was important in determining
the ratio of the diastereomers. Use of 5% Pd/C resulted
in nearly equivalent amounts of the trans and cis C/D
ring juncture while use of Pd/CaCO3 resulted in a 4:1
mix favoring the trans C/D ring juncture. Demethyl-
ation with sodium ethanethiolate in DMF at reflux
provided 11. Inversion of the 17b hydroxy alcohol was
accomplished using modified Mitsunobu conditions.13
Compounds 16 and 17 in which the angular methyl
group is in the a configuration were prepared in opti-
cally pure fashion as shown in Scheme 2 in analogous
fashion to 11 and 12 using the antipode of the starting
dione (15) and carrying it through an identical reaction
sequence.14
As shown in Table 3, the thiaequilenin ligands bind to
ERa with Ki’s ranging from 4 to 24 nM.15 Weaker
relative affinity toER b is observed in each case. The
ligand with the highest affinity toER a is 11 (Ki=4 nM)
while its enantiomer 16 has 3-fold weaker affinity
(Ki=13 nM). Inversion of the 17b-OH to17 a-OH
results in a>5-fold loss in potency (see 11 and 14), data
which is consistent with known trends for estradiol and
analogues.16 While the effects of modifying the C-17
position in the steroid backbone are well documented,
structure–activity studies examining the variations in
the C/D ring juncture are limited.16 In the thiaequilenin
series, the C/D trans 17b-OH analogue 11 binds with
higher affinity toER a than its cis counterpart 12. A
similar trend is observed for the analogous compounds
in the 17a-OH series (16 and 17). The orientation of the
angular methyl group has little influence on binding in
the 17a-OH series as shown by analogues 14 and 17 that
differ only by the orientation of the methyl group and
have identical affinity. When comparing ligands that are
Scheme 2.
Table 3. hERa and hERb binding affinity for 17a-DHEqn and ana-
logues
Compd
ERa
ERb
Ki, nM
Ki, nM
17a-DHEqn (4)
8.0
4.0
10
23
13
24
11
47
73
17
120
11
12
14
16
17
Scheme 1. Reagents and conditions: (a) H2O2, NaOH, (b) m-
methoxythiophenol, NaOH, 31%; (c) AlCl3, CH2Cl2, 68%; (d)
NaBH4, THF/MeOH, 99%: (e) H2, 5% Pd/C, THF/EtOH, 38% 9,
45% 10; (f) NaSEt, DMF, 30%; (g) p-nitrobenzoic acid, DEAD, PPh3,
70%; (h) potassium carbonate, 69%.
24

4284
C.W.Lugar, III et al./ Bioorg.Med.Chem.Lett.13 (2003) 4281–4284
thiaequilenin analogues of 4 indicate that the orienta-
tion of the C-17 hydroxy group plays a more significant
role than the C/D ring juncture with respect to deter-
mining affinity to the receptor. The uterine profile of the
thiaequilenins 11 and 14 is encouraging and suggests
that this structural platform may be useful for the
design of tissue selective estrogens or SERMs.
Table 4. Effects of 17a-DHEqn (4), 11, and 14 on uterine weight in
OVX rats after 4 days of oral dosing
Compd
Dose
(mpk)
Uterine weight/
body weight
Uterine weight
% increase versus OVX
OVX
EE2
4
—
0.10
0.01
0.10
1.00
10.0
0.01
0.10
1.00
5.00
0.01
0.10
1.00
10.0
0.32Æ0.03
1.03Æ0.03
0.33Æ0.13
0.29Æ0.13
0.45Æ0.02*
0.75Æ0.06*
0.37Æ0.03
0.33Æ0.01
0.32Æ0.01
0.48Æ0.01*
0.34Æ0.02
0.36Æ0.02
0.31Æ0.02
0.66Æ0.05*
—
219.6
1.5
À11.1
38.1
132.5
13.7
1.4
0.2
48.4
5.5
10.2
À5.3
103.1
Acknowledgements
11
14
We wish to thank the Lead Optimization Biology group
for conducting binding assays.
References and Notes
1. (a) Smith, D. C.; Prentice, R.; Thompson, D. J.; Herrman,
W. L. N.Eng.J.Med.
*p<0.05 verses OVX by one way analysis of variance and Fisher’s
posthoc analysis. n=5 animals/group.
1975, 293, 1164. (b) Rossouw, J. E.;
Anderson, G. L.; Prentice, R. L.; LaCroix, A. Z.; Kooperberg,
C.; Stefanick, M. L.; Jackson, R. D.; Beresford, S. A. A.;
Howard, B. V.; Johnson, K. C.; Kotchen, J. M.; Ockene
JAMA 2002, 288, 311.
2. Bhavnani, B. R.; Woolover, C. A. Steroids 1991, 56, 201.
3. Washburn, S. A.; Adams, M. R.; Clarkson, T. C.; Adel-
enantiomeric pairs, 11 and 16 or 12 and 17, the binding
affinity of the enantiomer with the 17b-OH is more
potent than its 17a counterpart. In general, compounds
with the 17b-OH functionality have higher affinity to
ERa regardless of whether the ring C/D ring juncture is
cis or trans. However, the overall differences in affinity
between ligands are relatively small indicating that the
receptor can readily accommodate changes in the spatial
arrangement of the C/D-ring juncture without dramatic
consequences to the binding affinity. Overall, this data
supports the hypothesis that there is a flexible binding
pocket in the estrogen receptor that allows for D-ring
modifications in the ligand.16
man, S. J.Am.J.Obstet.Gynecol
1993, 169, 251.
4. Washburn, S. A.; Lewis, C. E.; Johnson, J. E.; Voytko,
M. L.; Shively, C. A. Brain Res. 1997, 758, 241.
5. Washburn, S. A.; Honore, E. K.; Cline, J. M.; Helman, M.;
Wagner, J. D.; Adelman, S. J.; Clarkson, T. B. Am.J.Obstet.
Gynecol. 1996, 175, 341.
6. Berco, B. M.; Bhavnani, B. R. J.Soc.Obstet.Gynecol.
2001, 8, 245.
7. Black, L. J.; Sato, M.; Rowley, E. R.; Magee, D. E.; Bekele,
A.; Williams, D. C.; Cullinan, G. J.; Bendele, R.; Kauffman,
R. F.; Bensch, W. R.; Frolik, C. A.; Termine, J. D.; Bryant,
H. U. J.Clin.Invest. 1994, 93, 63.
In order to determine the effects of thiaequilenins on the
uterus, 11 and 14 were evaluated in the OVX rat. As
shown in Table 4, 17a-DHEqn (4) causes an increase in
uterine weight relative to OVX controls at both 1 and 10
mpk after 4 days of oral administration. Uterine hyper-
trophy is also observed with compounds 11 and 14 at
the 5 and 10 mpk dose, respectively, but not at 1 mpk.
Based on this data the thiaequilenins are less utero-
trophic than 4 at equivalent doses. Further studies will
be required in order to determine if this observation is
due todifferences in pharmacokinetics between these
compounds and whether the improvement in uterine
pharmacology translates to a better overall tissue selec-
tivity profile with respect to bone, lipids, and uterus.
8. Wakeling, A. E.; Bowler, J. J.Steroid Biochem.Mol.Biol.
1992, 43, 173.
9. Evans, G.; Bryant, H. U.; Magee, D.; Sato, M.; Turner,
R. T. Endocrinology 1994, 134, 2282.
10. Sato, M.; Grese, T. A.; Dodge, J. A.; Bryant, H. U.;
Turner, C. H. J.Med.Chem. 1999, 42, 1.
11. Crenshaw, R. R.; Luke, G. M. Tetrahedron Lett. 1969, 10,
4495.
12. Hajos, Z. G.;Parrish, D. R.Org.Syn.Coll.Vol.VII 1990, 363.
13. Dodge, J. A.; Lugar, C. W. Bioorg.Med.Chem.Lett.
1996, 6, 1.
14. All new compounds were characterized by MS and
1
300 MHz H NMR. For full experimental details for the pre-
paration of all compounds shown in Schemes 1 and 2, see:
Dodge, J. A. Lugar, C. W. US Patent 6218425. Compounds 1,
2, and 3 were purchased from Steraloids. Compound 4 was
prepared as described in ref 13.
15. For experimental protocols for the ERa and ERb binding
assays. see: Wallace, O. B.; Lauwers, K. S.; Jones, S. A.;
Dodge, J. A. Bioorg.Med.Chem.Lett. 2003, 13, 1907. Com-
pounds were generally tested in duplicate. The minimum sig-
nificant ratio for the binding assays is 2-fold.
In summary, we have evaluated a series of equine
estrogen analogues to determine whether these agents
demonstrate tissue selective pharmacology. 17a-DHEqn
(4) effectively prevents bone loss and lowers serum cho-
lesterol in the OVX rat at doses that also causes a sta-
tistically significant increase in uterine weight versus
OVX controls. Structure–activity studies with B-nor-6-
16. Anstead, G. M.; Carlson, K. A.; Katzenellenbogen, J. A.
Steroids 1997, 62, 268.