Characterization of a new class of selective nonsteroidal progesterone receptor agonists

Article abstract of DOI:10.1016/j.steroids.2003.12.007

The identification of a new series of selective nonsteroidal progesterone receptor (PR) agonists is reported. Using a high-throughput screening assay based on the measurement of transactivation of a mouse mammary tumor virus promoter-driven luciferase reporter (MMTV-Luc) in human breast cancer T47D cells, a benzimidazole-2-thione analog was identified. Compound 1 showed an apparent EC₅₀ of 53nM and efficacy of 93% with respect to progesterone. It binds to PR with high affinity (K_i=89nM), but had no or very low affinity for other steroid hormone receptors. Structure-activity relationship studies of a series of benzimidazole-2-thione analogs revealed critical positions for high PR binding affinity and transactivation potency as well as receptor selectivity, as exemplified by 25. Compound 25 binds to human PR with high affinity (K_i=28nM) and had at least >1000-fold selectivity for PR versus other steroid receptors. Molecular modeling studies suggested that these agonists overlap favorably with progesterone in the ligand-binding domain of PR. In T47D cells, compound 25 acted as a full agonist in the MMTV-Luc reporter assay, as well as in the induction of endogenous alkaline phosphatase activity with apparent EC₅₀ values of 4 and 9nM, respectively. In the immature rat model, compound 25 provided a significant suppression of estrogen-induced endometrium hypertrophy as measured by luminal epithelial height. In contrast, compound 25 was inactive in the luteinizing hormone release assay in young ovariectomized rats. These benzimidazole-2-thione analogs constitute a new series of nonsteroidal PR agonists with an excellent steroid receptor selectivity profile. The differential activities observed in the in vivo progestogenic assays in rat models suggest that these analogs can act as selective PR modulators.

Full text of DOI:10.1016/j.steroids.2003.12.007

Steroids 69 (2004) 201–217
Characterization of a new class of selective nonsteroidal
progesterone receptor agonists
Yu Donga,1, Jacques Y. Robergeb,1, Zhaoqing Wanga, Xiaodong Wanga, Joseph Tamasia,
Vanessa Della, Rajasree Gollaa, James R. Cortec, Yalei Liub, Tianan Fangc,
Monique N. Anthonyd, Dora M. Schnure, Michele L. Aglerd, John K. Dickson Jr.c,
R. Michael Lawrenceb, Margaret M. Prackd, Ramakrishna Seethalaa, Jean H.M. Feyena,∗
a
Department of Metabolic and Cardiovascular Drug Discovery, Bristol-Myers Squibb Company, Pennington, NJ 08543, USA
b
Department of Early Discovery Chemistry, Bristol-Myers Squibb Company, Pennington, NJ 08543, USA
c
Department of Discovery Chemistry, Bristol-Myers Squibb Company, Pennington, NJ 08543, USA
d
Department of Lead Discovery, Bristol-Myers Squibb Company, Pennington, NJ 08543, USA
e
Department of Computer Assisted Drug Design, Wallingford, CT 06492, USA
Received 8 September 2003; received in revised form 21 November 2003; accepted 4 December 2003
Abstract
The identification of a new series of selective nonsteroidal progesterone receptor (PR) agonists is reported. Using a high-throughput
screening assay based on the measurement of transactivation of a mouse mammary tumor virus promoter-driven luciferase reporter
(MMTV-Luc) in human breast cancer T47D cells, a benzimidazole-2-thione analog was identified. Compound 1 showed an apparent
EC50 of 53 nM and efficacy of 93% with respect to progesterone. It binds to PR with high affinity (Ki = 89 nM), but had no or very
low affinity for other steroid hormone receptors. Structure–activity relationship studies of a series of benzimidazole-2-thione analogs
revealed critical positions for high PR binding affinity and transactivation potency as well as receptor selectivity, as exemplified by 25.
Compound 25 binds to human PR with high affinity (Ki = 28 nM) and had at least >1000-fold selectivity for PR versus other steroid
receptors. Molecular modeling studies suggested that these agonists overlap favorably with progesterone in the ligand-binding domain
of PR. In T47D cells, compound 25 acted as a full agonist in the MMTV-Luc reporter assay, as well as in the induction of endogenous
alkaline phosphatase activity with apparent EC50 values of 4 and 9 nM, respectively. In the immature rat model, compound 25 provided
a significant suppression of estrogen-induced endometrium hypertrophy as measured by luminal epithelial height. In contrast, compound
25 was inactive in the luteinizing hormone release assay in young ovariectomized rats. These benzimidazole-2-thione analogs constitute
a new series of nonsteroidal PR agonists with an excellent steroid receptor selectivity profile. The differential activities observed in the in
vivo progestogenic assays in rat models suggest that these analogs can act as selective PR modulators.
© 2004 Elsevier Inc. All rights reserved.
Keywords: Benzimidazole-2-thione; Nonsteroidal progesterone receptor agonists; Steroid receptor selectivity
1. Introduction
lack of selectivity with other steroid hormone receptors
such as the glucocorticoid receptor (GR) and androgen
receptor (AR) [13]. Selective progestins are needed to
improve the safety profile for clinical use. Steroidal pro-
gestins generally have higher cross-reactivity with other
steroid hormone receptors due to the four-ring structural
motif shared by natural steroid receptor ligands. In ad-
dition, the rich metabolic pathways of these steroids po-
tentially lead to active metabolites that cross-react with
several steroid hormone receptors. In recent years, novel
classes of nonsteroidal progesterone receptor (PR) modu-
lators with improved receptor and tissue selectivity have
been reported [14–22]. These new compounds offer the
Progesterone (P4) plays an important role in reproduc-
tive physiology, as well as bone metabolism and neu-
rotropic functions [1–6]. Synthetic steroidal progestins
are widely used as therapeutic agents in the control of
fertility, combination hormone replacement therapy, and
a variety of other endocrinologic diseases [7–12]. Cur-
rently, available progestins have side effects due to their
∗
Corresponding author. Tel.: +1-609-818-3277; fax: +1-609-818-3600.
E-mail address: jean.feyen@bms.com (J.H.M. Feyen).
1
Equal contributions to this work.
0039-128X/$ – see front matter © 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2003.12.007

202
Y. Dong et al. / Steroids 69 (2004) 201–217
potential for safer and broader pharmacologic use as pro-
gestins.
2.4. PR transactivation assay in T47D-MMTV-Luc cells
In this paper, structure–activity relationship (SAR) studies
within a series of benzimidazole-2-thione analogs acting as
selective, nonsteroidal progestin agonists are reported. These
compounds exhibited high affinity binding and full agonist
activity for PR in vitro. In addition, progestogenic in vivo
activity was demonstrated for the most active compound in
the series.
For each experiment, T47D-MMTV-Luc cells were
seeded at 10,000 cells per well in a 96-well plate and incu-
bated at 37 ◦C in 5% (v/v) CO2 incubator overnight. On the
next day, test compounds were added to the cells and incu-
bated for 16–24 h. The luciferase activities were measured
by using a Steady-Glo Luciferase kit (Promega, Madison,
WI) and quantified using a Topcount (Packard Instrument,
Downers Grove, IL). For the high-throughput screen, test
compounds from the Bristol-Myers Squibb (BMS) com-
pound collection were diluted in water/dimethyl sulfoxide
(DMSO) to give a final assay concentration of 10 ␮M in
1% (v/v) DMSO and transferred to 384-well plates. Cells
were then plated into 384-well plates at 5000 cells per well
in RPMI containing 5% charcoal-treated FBS (Cocalico
Biologicals, Reamstown, PA) and incubated overnight at
37 ◦C in a 5% (v/v) CO2 incubator (Forma Scientific). Pro-
gesterone (10 nM) was used as a control on every plate, and
the data were normalized with reference to it. Luciferase
activity was measured as above.
2. Materials and methods
2.1. Materials
Medroxyprogesterone acetate (MPA), P4, dexametha-
sone (DEX), 5␣-dihydrotestosterone (DHT), 17␤-estradiol
(E2), and aldosterone were purchased from Sigma Chem-
ical Co. (St. Louis, MO). All radiolabeled steroids were
obtained from NEN Life Science Products (Boston, MA).
All cell culture reagents were obtained from Life Tech-
nologies (Rockville, MD). Assay plates were obtained from
BD Biosciences (Franklin Lakes, NJ). From compound
synthesis, most chemicals and solvents were of analytical
grade and used without further purification. Commercial
reagents were purchased from Aldrich Chemical Company
(Milwaukee, WI).
2.5. Alkaline phosphatase (AP) assay using T47D cells
The assay was performed according to Di Lorenzo et al.
[23] with minor modifications. In short, T47D cells were
seeded into a 96-well plate at a density of 10,000 cells/well
in RPMI1640 medium supplemented with 2% charcoal-
stripped FBS and allowed to attach to the plate overnight.
On the next day, the cells were treated with compounds
for 16–24 h. After treatment, cells were washed with phos-
phate buffered saline (PBS) and lysed in 50 ␮l/well lysis
buffer (0.1 M Tris–HCl, pH 9.8, and 0.2% Triton X-100)
followed by freeze and thaw. For the measurement of AP
activity, 150 ␮l of substrate (4 mM p-nitrophenyl phosphate
in 0.1 M Tris–HCl, pH 9.8) were added to each well and
incubated in the dark at room temperature overnight. Op-
tical density measurements were taken at a wavelength of
405 nM.
2.2. Cell culture
Human breast cancer T47D cells (American Type Cul-
ture Collection, Manassas, VA) and their sub-line (T47D-
MMTV-Luc) containing stably transfected pMAM-neo-
Luc plasmid (Clontech, Palo Alto, CA) were maintained
in RPMI1640 media fortified with 10% fetal bovine
serum (FBS), penicillin (50 units/ml), and streptomycin
(50 ␮g/ml). In T47D-MMTV-Luc culture, geneticin at
300 ␮g/ml were routinely added to the culture to maintain
the selection pressure. Human breast cancer MDA.MB 453
cells (ATCC) were cultured in Dulbecco’s Modification of
Eagle’s Medium (DMEM) supplemented with 10% (v/v)
FCS, penicillin (50 units/ml), and streptomycin (50 ␮g/ml).
All cells were grown in tissue culture incubator at 37 ◦C in
a 5% (v/v) CO2 atmosphere.
2.6. Steroid receptor binding assays
2.6.1. PR binding assays
Exponentially growing T47D cells were harvested and
collected in RPMI-1640 medium (serum-free). Binding re-
actions were carried out in a 96-well plate, and each well
contained a mixture of 50,000 cells and 2.5 nM of [3H]-
progesterone (57 Ci/mmol) in the presence of increasing
amounts of test compounds. Incubations were continued for
2 h at room temperature. Unbound steroids were removed
by washing the cells on filter plates using a harvester. Bound
radioactivity was determined by liquid scintillation count-
ing. In the PR binding assay using recombinant full-length
human PRb (PanVera, Madison, WI), the procedure was the
same except that each well contained PR protein (5 nM) in-
stead of cells.
2.3. Animals
Female Sprague-Dawley rats were obtained from Harlan
(Indianapolis, IA). Adrenalectomized male Sprague-Dawley
rats were also obtained from Harlan and were given wa-
ter containing 1% sodium chloride during shipping and af-
ter arrival to help maintain sodium balance. Animals were
maintained under conditions of a 12-h light/dark cycle with
soy-free rodent chow provided to female rats and standard
chow to male rats (BioServ, Frenchtown, NJ) and water ad
libitum. The Institutional Animal Care and Use group ap-
proved all treatments and procedures.

Y. Dong et al. / Steroids 69 (2004) 201–217
203
2.6.2. AR whole-cell binding assay
with vehicle. The other groups of estrogen-primed animals
were given subcutaneous injections of either P4 at 0.1, 1,
and 10 mg/kg, MPA at 0.1, 1, and 10 mg/kg or 25 at 0.5 and
10 mg/kg. Blood was collected 5 h after dosing by cardiac
puncture under isoflurane anesthesia. LH concentrations in
the serum samples were analyzed by a double-antibody RIA
method (Amersham, Piscataway, NJ).
MDA.MB 453 cells growing in serum-free DMEM at 80%
confluence were harvested and counted. For the AR binding
assay, 400,000 cells were added to each well in a 96-well
plate and mixed with 2 nM [3H]-dihydrotestosterone (DHT,
60 Ci/mmol) in the presence of various concentrations of
test compound. Binding reactions were continued for 2 h
at room temperature. Unbound steroids were removed by
washing the cells on filter plates, and bound radioactivity
was determined by liquid scintillation counting.
2.8. Uterotropic activity
Immature rats, 19 days of age, were divided into groups
of six animals. The animals were given daily subcuta-
neous injections for 3 days of either vehicle (corn oil); E2
(2 ␮g/rat); E2 + P4; E2 + MPA or E2 + 25. All progestins
were tested at 0.1, 0.3, 1, 3, and 10 mg/kg. An additional
group of 25 at 30 mg/kg was also tested. The rats were
euthanized 24 h after the last dose by carbon dioxide inhala-
tion, and the uteri were collected and fixed in 10% neutral
buffered formalin. The tissues were processed in a Leica
TP1050 tissue processor (Bannockburn, IL) and embed-
ded in paraffin. Sections (5 ␮m cross-sections) were cut on
a Microm HM335E microtome (Walldorf, Germany) and
stained with hematoxylin and eosin. Epithelial cell height
in the endometrium was measured on an image analysis
system (Bioquant, Nashville, TN).
2.6.3. Estrogen receptor (ER) binding assay
Estrogen receptor binding affinity was determined us-
ing a baculovirus-expressed recombinant full-length human
ER␣ and a fluorescence polarization (FP) assay kit based on
manufacturer’s specifications (PanVera).
2.6.4. GR binding assay
Glucocorticoid binding affinity was determined using a
baculovirus-expressed recombinant full-length human GR
with glutathione S-transferase (GST) fusion and fluores-
cence polarization (FP) assay kit based on manufacturer’s
specifications (PanVera).
2.6.5. Mineralocorticoid receptor (MR) binding assay
The MR binding assay using rat kidney cytosols has
been described by De Gasparo et al. [24]. Briefly, bilat-
erally adrenalectomized male Sprague-Dawley rats were
euthanized by CO2 inhalation. Kidneys were collected and
kept in cold PBS for cytosol preparation. Kidneys were
minced into small pieces and homogenized by polytron in
binding buffer (50 mM Tris–HCl, pH 7.5, 1 mM EDTA,
10% glycerol, 1 mg/ml BSA, 2 mM DTT, and protease
inhibitor mixture). The cytosol fraction was collected af-
ter ultra-centrifugation of the homogenate at 80,000 × g
(Ti60, Beckman Instruments, Fullerton, CA) at 4 ◦C for
1 h. MR binding assays were performed in 96-well plates
in the presence of protein of the cytosol fraction (1 mg),
[3H]-aldosterone (3 nM, 49 Ci/mmol), and test compound.
Anti-glucocorticoid RU486 (150 nM) was added to each
binding reaction to block GR binding sites. After incuba-
tion at 4 ◦C for 2 h, unbound steroids were removed by
washing on filter plates. Radioactivity was determined by
liquid scintillation counting.
2.9. Chemistry
Compound 1 was prepared according to published pro-
cedures [26,27]. Analogs of 1 were prepared by parallel
synthesis using an adaptation of literature procedures as
shown in Fig. 1A. Compound 8, the oxo analog of 1, was
prepared by treatment of the requisite dianiline with triphos-
gene (Fig. 1B). Compound 9 was prepared by alkylation of
1 with methyl iodide in the presence of cesium carbonate.
Compounds 10 and 11 were prepared by the oxidation of
9 with meta-chloroperoxybenzoic acid (Fig. 1C) accord-
ing to the procedure of Bru-Magniez [26]. Compound 29
was prepared using the commercially available 3-fluoro-4-
nitrobenzylbromide as illustrated in Fig. 1D.
2.10. Experimental chemistry
2.10.1. General methods
2.7. Luteinizing hormone (LH) release assay
Analogs 2–7 and 12–28 were prepared as described below
for 1 and 5. Analogs in Table 1, except for 29 and 37, were
prepared as part of focused libraries and were characterized
The animal model used in this study has been described
previously [25]. Peripubertal rats, 26 days of age, were
ovariectomized under 5% isoflurane anesthesia via a dorsal
approach. After a 48-h recovery period, the animals were di-
vided into groups of six animals. One group of animals was
vehicle treated (corn oil), and the remaining animals were
estrogen-primed with E2 (2 ␮g/rat) by subcutaneous injec-
tion for 2 days. On day 3, the vehicle group and one group
of estrogen-primed animals were injected subcutaneously
1
by LCMS and H NMR. Reverse-phase analytical HPLC
was carried out on a Shimadzu Analytical HPLC system run-
ning DiscoveryVP software using a Phenominex Luna C18
column (4.6 mm × 50 mm) eluted at 4 ml/min with a 4 min
gradient from 100% A to 100% B (A: 10% methanol, 89.9%
water, 0.1% trifluoroacetic acid (TFA); B: 10% water, 89.9%
methanol, 0.1% TFA, UV 220 nm). Reverse-phase prepar-
ative HPLC was carried out using a Shimadzu Preparative

204
Y. Dong et al. / Steroids 69 (2004) 201–217
Fig. 1. Chemistry A: reaction sequence for the preparation of the benzimidazole-2-thiones; B: preparation of compound 8; C: alkylation and oxidation
of compound 1 to prepare compounds 9–11; D: preparation of compound 29.
HPLC system running DiscoveryVP software on a Shim-
PackVP-ODS column (50L × 20 mm) at 20 ml/min, 6 min
gradient 100% A to 100% B with the solvent system used for
the analytical. LCMS were obtained on a Shimadzu HPLC
system running DiscoveryVP software, coupled with a Wa-
ters Model PlatformLC mass spectrometer running MassL-
ynx version 3.5 software using the same column and con-
ditions as utilized for analytical described above. Melting
points were obtained on a MEL-TEMP II from Laboratory
Devices, USA, and are uncorrected. HRMS was obtained on
a Micromass LCT TOF with LockSpray.
septuplet, J = 7 Hz), 1.32 (6H, d, J = 7 Hz), 13C NMR
(125 MHz, CDCl3) δ: 167.6 (d, JC–F = 256 Hz), 146.7 (d,
JC–F = 13 Hz), 130.2 (d, JC–F = 13 Hz), 128.8, 103.7 (d,
JC–F = 25 Hz), 99.5 (d, JC–F = 28 Hz), 44.3, 22.5; 19F
NMR (470 MHz, CDCl3) δ: −99.56.
To 1a (17 g, 86 mmol) in DMSO (150 ml) were added
imidazole (9.7 g, 190 mmol) and cesium carbonate (61.9 g,
190 mmol), and the resulting mixture was heated at 90 ◦C
overnight. The reaction was allowed to cool to room temper-
ature and was poured into water (1 l). The resulting yellow
solid was collected and partitioned between EtOAc (200 ml)
and a 10% aqueous citric acid solution (w/v). The organic
layer was washed with 10% citric acid solution (3×100 ml).
The combined aqueous layers were neutralized to pH 7 with
aqueous NaHCO3 and stirred for 1 h to selectively precipitate
the mono-imidazole product from the bis-imidazole byprod-
uct derived from incomplete consumption of the difluoro
starting material by isopropylamine in the first step. The pre-
cipitate was collected and dried to afford 10.5 g (50%) of
N-(5-imidazol-1-yl-2-nitrophenyl)isopropylamine (1b) as a
yellow solid, which was used in the next step without further
purification.
2.10.2. 6-Imidazol-1-yl-1-isopropyl-1H-benzoimidazole-
2-thiol (1)
To 2,4-difluoronitrobenzene (25 g, 157 mmol) in ethanol
(100 ml) was slowly added isopropylamine (46.4 g,
786 mmol) over 10 min producing an exothermic reaction.
The reaction was refluxed for 40 min, cooled to room tem-
perature, and concentrated under reduced pressure. The
resulting orange slush was diluted with water (100 ml),
extracted with chloroform (3 × 75 ml), washed with brine
(50 ml), dried over sodium sulfate, filtered, and concentrated
to give 31.2 g of orange needles. This crude product was pu-
rified by flash chromatography (300 g SiO2) by dryloading
on 60 g of SiO2 and eluting with ethyl acetate:heptane (3:2)
to afford 23.8 g (77%, HPLC purity: 99%) of (5-fluoro-
2-nitrophenyl)isopropylamine (1a). 1H NMR (500 MHz,
CDCl3) δ: 8.21 (1H, dd, J = 6 and 9 Hz), 8.10 (1H, br. s),
6.48 (1H, dd, J = 2 and 12 Hz), 6.32 (1H, m), 3.72 (1H,
To 1b (4.1 g, 16.5 mmol) in ethanol (100 ml) were added
zinc dust (21.7 g, 33 mmol) and ammonium chloride (11.5 g,
83 mmol). The resulting mixture was stirred at room temper-
ature overnight. The mixture was filtered, and the filtrate was
concentrated. The residue was suspended in EtOAc (80 ml)
and methanol (10 ml). The resulting solid was isolated by
filtration, washed with EtOAc, and dried to afford 4.5 g of

Y. Dong et al. / Steroids 69 (2004) 201–217
205
4-imidazol-1-yl-N2-isopropyl-benzene-1,2-diamine (1c) as a
gray solid, which was used without further purification.
To 1c (3.36 g, 15.5 mmol) in ethanol (100 ml) was added
carbon disulfide (7.1 g, 93 mmol), and the resulting mixture
was heated at 80 ◦C overnight. The reaction was allowed to
cool and poured into water (1 l). The precipitate was iso-
lated by filtration, washed with water, and dried to provide
2.98 g (70% from 1b, HPLC purity: 97%) of 1 as a tan
solid.
102.4, 44.6, 20.7, 11.0; HRMS m/z Calc’d for C13H15N4S
[M + H]+: 259.1017; Found 259.1018.
2.10.4. 6-Imidazol-1-yl-1-isopropyl-1H-benzoimidazol-
2-ol (8)
Methanol (5 ml) was added to a mixture of 1b (124 mg,
0.5 mmol), zinc dust (660 mg, 10 mmol), and ammonium
chloride (350 mg, 5 mmol). After an additional 2 h of stir-
ring, the mixture was heated at 60 ◦C for 1 h. The mixture
was cooled to room temperature, filtered through CeliteTM
,
1H NMR (500 MHz, CD3OD) δ: 8.21 (s, 1 H), 7.71 (s, 1
H), 7.62 (s, 1 H), 7.40 (m, 1 H), 7.35 (d, J = 8.3 Hz, 1 H),
7.20 (s, 1 H), 5.57 (m, 1 H), 1.62 (s, 3 H), 1.60 (s, 3 H);
13C NMR (125 MHz, CD3OD) δ: 171.1, 137.6, 133.6, 133.1,
132.2, 129.7, 120.9, 118.4, 111.8, 105.9, 49.9, 19.7, 11.0;
HRMS m/z Calc’d for C13H15N4S [M + H]+: 259.1018;
Found 259.1014.
and washed with methanol. The methanol was concentrated
under reduced pressure to give 172 mg of 4-imidazol-1-
yl-N2-isopropylbenzene-1,2-diamine (>100% yield) as a
black solid, which was used without further purification.
The black solid was dissolved in tetrahydrofuran (THF)
(5 ml), and morpholine (225 ␮l, 2 mmol) and triphosgene
(75 mg, 0.25 mmol) were added. The exothermic reaction
was stirred overnight forming a black suspension. The re-
action was diluted with water (30 ml), and the gray solid
was filtered and washed successively with water and ether.
The solid was suspended in a mixture of methanol and
ethanol (1:1) and was converted to the hydrochloride salt
by adding a few drops of concentrated hydrochloric acid
followed by solvent removal. Ethanol was added to the
solid, and the product was concentrated again. The solid
was triturated with ether and kept under vacuum overnight
to give 61 mg (50% yield, HPLC purity: 90%) of 8 as a
2.10.3. 6-Imidazol-1-yl-1-propyl-1H-benzoimidazole-2-thiol
(5)
To a solution of 2,4-difluoronitrobenzene (0.291 g,
1.82 mmol) in DMSO (3.6 ml) was added cesium carbonate
(1.48 g, 4.55 mmol) and propyl amine (0.113 g, 1.91 mmol).
The resulting mixture was stirred at room temperature for
2 h and then warmed to 95 ◦C for 1 h. It was cooled, and im-
idazole (0.161 g, 2.37 mmol) was added. The mixture was
heated at 95 ◦C for 15 h. The resulting dark brown product
was cooled to room temperature and added to vigorously
stirred cold water (20 ml) to give a yellow suspension.
The solid was collected by filtration, washed with water
(two times), air-dried, and dried under vacuum to afford
316 mg (70%, HPLC purity: 88%) of N-(5-imidazol-1-yl-2-
nitrophenyl)propylamine (5a) as a yellow solid, which was
used without further purification.
To a yellow suspension of 5a (0.300 g, 1.22 mmol) in
ethanol (24 ml) was added 10% palladium on carbon (Pd/C)
(0.065 g, 0.061 mmol). Hydrogen gas was bubbled through
the reaction for several minutes, and then the reaction was
vigorously stirred under an atmosphere of hydrogen for
2 h. The reaction was filtered to remove Pd/C to give a
clear, yellow solution, to which were added carbon disulfide
(0.30 ml, 4.88 mmol) and triethylamine (1.7 ml, 12.2 mmol).
The reaction was placed in a preheated oil bath (95 ◦C)
for 3 h. The resulting orange reaction mixture was cooled
to room temperature and concentrated to give an orange-
brown solid. Trituration from methanol (2 ml) gave a gray
solid, which was collected and rinsed with methanol (1 ml).
Trituration from dichloromethane (2 ml) gave a gray solid,
which was collected and rinsed with dichloromethane (1 ml),
air-dried, and dried under vacuum to give 0.118 g (37%,
HPLC purity: 99%) of 5 as a gray solid. m.p. 242–245 ◦C;
1H NMR (500 MHz, DMSO-d6) δ: 12.90 (s, 1H), 8.23 (s,
1H), 7.77 (d, J = 2.2 Hz, 1H), 7.74 (s, 1H), 7.42 (dd,
J = 8.2, 2.2 Hz, 1H), 7.25 (d, J = 8.2 Hz, 1H), 7.10 (s,
1H), 4.20 (t, J = 7.2 Hz, 2H), 1.79-1.72 (m, 2H), 0.91
(t, J = 7.2 Hz, 3H); 13C NMR (125 MHz, DMSO-d6) δ:
169.3, 135.9, 133.4, 132.3, 129.6, 129.4, 118.7, 115.6, 110.2,
1
reddish brown solid; H NMR (500 MHz, CD3OD) δ: 9.43
(m, 1 H) 8.05 (d, J = 1.7 Hz, 1 H), 7.77 (dd, J = 2.02,
1.3 Hz, 1 H), 7.66 (d, J = 2.0 Hz, 1 H), 7.38 (dd, J = 8.06,
2.0 Hz, 1 H), 7.26 (d, J = 8.1 Hz, 1 H), 4.70 (ddd, J =
21.0, 14.0, 7.4 Hz, 1 H), 1.57 (d, J = 6.7 Hz, 6 H); 13C
NMR (125 MHz, CD3OD) δ: 156.5, 135.9, 131.2, 123.5,
121.5, 117.2, 111.0, 105.4, 66.9, 46.8, 20.2, 15.4; HRMS
m/z Calc’d for C13H15N4O [M + H]+: 243.1246; Found
243.1250.
2.10.5. 6-Imidazol-1-yl-1-isopropyl-2-methylsulfanyl-1H-
benzoimidazole (9)
Methyl iodide (46 ␮l, 0.74 mmol) was added to a so-
lution of 1 (126 mg, 0.49 mmol) and cesium carbonate
(320 mg, 0.98 mmol) in DMF (2.5 ml). The mixture was
heated at 60 ◦C for 72 h. The mixture was diluted with
water (10 ml) and extracted with ethyl acetate/heptane (5/2,
3 × 15 ml), washed with water (10 ml), brine (10 ml), dried
with sodium sulfate, filtered and, concentrated to give 97 mg
of a dark oil. The oil was dissolved in a minimum amount
of dichloromethane and was loaded onto a SiO2 cartridge
(Jones Chromatography) eluted with a gradient of 100%
dichloromethane to 70%acetone/30% dichloromethane to
afford 63 mg (48%, HPLC purity: 83%) of 9 as a brown
1
oil; H NMR (500 MHz, CDCl3) δ: 7.84 (s, 1 H), 7.72 (d,
J = 8.80 Hz, 1 H), 7.39 (m, 1 H), 7.29 (s, 1 H), 7.21 (m, 2
H), 4.69 (m, 1 H), 2.82 (s, 3 H), 1.64 (d, J = 6.6 Hz, 6 H);
13C NMR (125 MHz, CDCl3) δ: 154.3, 143.6, 136.4, 135.4,
131.6, 130.1, 119.4, 119.0, 116.4, 104.6, 48.9, 21.0, 14.9;

206
Y. Dong et al. / Steroids 69 (2004) 201–217
HRMS m/z Calc’d for C14H17N4S [M + H]+: 273.1174;
The yellow precipitate was isolated by filtration and washed
with water. The solid was dried under reduced pressure
overnight to give 37 g (107%, HPLC purity: 85%) of 25a
as a yellow solid, which was used without purification.
Ammonium chloride (53 g, 1000 mmol) was added to a
suspension of the 25a (20.7 g, 76 mmol) and zinc dust (100 g,
1500 mmol) in ethanol (300 ml). The resulting mixture was
stirred for 2 h at room temperature and for 15 min in a
Found 273.1161.
2.10.6. 6-Imidazol-1-yl-1-isopropyl-2-methanesulfinyl-1H-
benzoimidazole (10)
A stock solution of meta-chloroperbenzoic acid (230 ␮l,
0.139 mmol, 104 mg/ml) in chloroform was slowly added
to a cold (0 ◦C) solution of 9 (25 mg, 0.092 mmol) in chlo-
roform/methanol (1.2 ml, 1/1). The mixture was stirred at
0 ◦C for 20 min and warmed to room temperature, and the
solvents were removed under reduced pressure. The residue
was partitioned between dichloromethane (5 ml) and satu-
rated sodium bicarbonate (5 ml), and the aqueous layer was
back extracted with dichloromethane (2×5 ml). The organic
layer were dried with sodium sulfate, filtered, and concen-
trated. The residue was purified by preparative HPLC to af-
ford 16 mg (44%, HPLC purity: 96%) of 10 as a yellow
oil (TFA salt, 1:1 mixture of atropisomers; 1.01, 1.10 min);
1H NMR (500 MHz, CDCl3) δ: 12.84 (s, 1 H), 9.24 (s, 1
H), 8.04 (d, J = 8.7 Hz, 1 H), 7.91 (s, 1 H), 7.59 (d, J =
25.5 Hz, 2 H), 7.45 (d, J = 8.7 Hz, 1 H), 5.46 (m, J = 20.7,
13.95, 7.1 Hz, 1 H), 3.31 (s, 3 H), 1.77 (d, J = 6.1 Hz, 6
H); 13C NMR (125 MHz, CDCl3) δ: 155.3, 143.3, 134.9,
134.8, 131.1, 123.6, 121.7, 121.2, 118.0, 107.7, 50.2, 38.9,
21.7, 21.6; HRMS m/z Calc’d for C14H17N4OS [M + H]+:
289.1123; Found 289.1117.
steam bath. The suspension was filtered through CeliteTM
,
which was washed with ethanol, to provide a solution of
the diamine 25b, which quickly turns dark upon exposure to
air. Carbon disulfide (13.7 ml, 228 mmol) and triethylamine
(31.9 ml, 228 mmol) were added to the ethanolic solution
of 25b, and the mixture was refluxed under nitrogen for
5 h. The mixture was cooled to room temperature and con-
centrated to ∼200 ml and was refrigerated (4 ◦C) overnight.
The solid was filtered, washed with ethanol, and dried un-
der reduced pressure to give 9.1 g of an off-white powder
(HPLC purity: 95%). The solid was dissolved in a hot 1,2-
dichloroethane:methanol mixture (1:1, 1.6 l). The colloidal
ZnS was removed by filtration through a 0.45-␮m nylon fil-
ter. A solution of hydrogen chloride in dioxane (4N, 30 ml)
was added, and the solvents were removed under reduced
pressure. The solid was dissolved in refluxing methanol
(∼40–60 ml/g), treated with activated charcoal, filtered, and
cooled to 4 ◦C. The resulting crystals were filtered, washed
with a small amount of cold ethanol, and dried in a vac-
uum oven at 115–120 ◦C overnight to give 6.4 g (25% from
2.10.7. 6-Imidazol-1-yl-1-isopropyl-2-methanesulfonyl-1H-
benzoimidazole (11)
1
25a, HPLC purity 100%) of the 25; H NMR (500 MHz,
A stock solution of metachloroperbenzoic acid (250 ␮l,
0.15 mmol, 104 mg/ml) in chloroform was slowly added to
a solution of 9 (25 mg, 0.092 mmol) in chloroform/methanol
(1.2 ml, 1/1). The mixture was stirred at room temperature
overnight. The solvents were removed under reduced pres-
sure, the residue was partitioned between dichloromethane
(5 ml) and saturated sodium bicarbonate (5 ml), and the
aqueous layer was back extracted with dichloromethane
(2 × 5 ml). The organic layers were dried with sodium sul-
fate, filtered, and concentrated to give 30 mg of dark oil. The
residue was purified by preparative HPLC to give 1.8 mg
DMSO-d6) δ: 13.15 (1H, s), 9.60 (1H, s), 8.21 (1H, s), 7.85
(1H, s), 7.69 (1H, s), 7.50 (1H, d, J = 8 Hz), 7.34 (1H, d,
J = 8 Hz), 5.47 (1H, quintuplet, J = 7 Hz), 2.16 (2H, m),
1.90 (4H, m), 1.66 (2H, m); 13C NMR (100 MHz, CDCl3) δ:
170.2, 135.2, 131.7, 131.0, 129.8, 121.8, 120.7, 117.6, 110.4,
105.3, 56.0, 27.6, 24.3; HRMS m/z Calc’d for C15H16N4S
[M+H]+: 285.1174; Found 285.1181; ICP-MS: Zn 0.02%;
Anal. Calc’d for C15H16N4S + 1.04 eq HCl: C, 55.90; H,
5.34; N, 17.36; S, 9.95; Cl, 11.44. Found: C, 55.92; H, 5.34;
N, 17.36; S, 10.26; Cl, 11.43.
1
(5%, HPLC purity: 95%) of 11 as a yellow film; H NMR
2.10.9. 6-Imidazol-1-ylmethyl-1-isopropyl-1H-
(500 MHz, CDCl3) δ: 8.97 (s, 1 H), 8.02 (d, J = 8.8 Hz, 1
H), 7.79 (s, 1 H), 7.46 (m, 3 H), 5.47 (septuplet, J = 7.1 Hz,
1 H), 3.63 (s, 3 H), 1.73 (d, J = 6.1 Hz, 6 H); [M + H]+
= 305.14.
benzoimidazole-2-thiol (29)
To a clear, pale yellow solution of 3-fluoro-4-nitrobenzyl
bromide [28] (0.104 g, 0.44 mmol) in DMF (1.48 ml) were
added imidazole (0.030 g, 0.44 mmol) and cesium carbonate
(0.145 g, 0.44 mmol). The resulting pink solution was stirred
at room temperature for 1.5 h. The reaction was cooled
to 0 ◦C, and sat. ammonium chloride (1.5 ml) and water
(1.5 ml) were added. This mixture was extracted with methy-
lene chloride (3 × 1.5 ml). The combined organic layers
were washed with water (two times), sat. NaHCO3 (once),
brine (two times), dried over Na2SO4, filtered, and con-
centrated to give 58 mg of a red residue. The crude aryl
fluoride was dissolved in DMSO (1.5 ml), and isopropy-
lamine (0.11 ml, 1.33 mmol) was added. The reaction was
stirred at room temperature overnight. Water (3.0 ml) was
2.10.8. 1-Cyclopentyl-6-imidazol-1-yl-1H-benzoimidazole-
2-thiol hydrochloride (25)
Cesium carbonate (103.3 g, 317 mmol) was slowly added
to a cold (ice bath) DMSO (250 ml) solution of 2,4-
difluoronitrobenzene (20.2 g, 127 mmol) and cyclopenty-
lamine (11.1 g, 131 mmol). The orange-colored mixture
was stirred at room temperature for 2 h. Imidazole (13 g,
190 mmol) was added, and the mixture was stirred at 100 ◦C
overnight. The reaction mixture was cooled to room tem-
perature, poured into water (1.8 l), and stirred for 20 min.

Y. Dong et al. / Steroids 69 (2004) 201–217
207
added, and the reaction was extracted with methylene chlo-
ride (3 × 2 ml). The combined organic layers were washed
with water (two times), brine (once), dried over Na2SO4,
filtered, and concentrated to give 46 mg of an orange-red
residue. Column chromatography (5% methanol in methy-
lene chloride) gave 32 mg (28%) of (5-imidazol-1-ylmethyl-
2-nitrophenyl)isopropylamine (29a), as an orange solid; 1H
NMR (500 MHz, CDCl3) δ: 8.15 (d, J = 8.8 Hz, 1H), 8.06-
7.98 (m, 1H), 7.58 (s, 1H), 7.15 (s, 1H), 6.93 (s, 1H), 6.40
(s, 1H), 6.37 (dd, J = 8.8, 1.7 Hz, 1H), 5.12 (s, 2H), 3.70-
3.64 (m, 1H), 1.27 (d, J = 6.0 Hz, 6H); HRMS m/z Calc’d
for C13H17N4O2 [M + H]+: 261.1352. Found 261.1346.
To a clear yellow solution of 29a (30 mg, 0.115 mmol) in
ethanol (3 ml) was added 10% Pd/C (6.4 mg, 0.006 mmol).
Hydrogen gas was bubbled through the reaction for sev-
eral minutes, and then the reaction was stirred vigorously
under a hydrogen atmosphere for 2 h. The reaction was fil-
tered through a nylon filter to give a clear, yellow solu-
tion. Triethylamine (80 ul, 0.58 mmol) and carbon disulfide
(14 ul, 0.23 mmol) were added. The reaction was stirred at
room temperature for 30 min, and then at 80 ◦C for 2 h. The
reaction was concentrated to give an orange residue. Col-
umn chromatography (5% methanol in methylene chloride)
gave 25 mg (80%, HPLC purity: 99%) 29 as a brown solid;
0.60 mmol) in ethanol (12 ml) was added carbon disulfide
(73 ␮l, 1.2 mmol), and the solution was warmed to 88 ◦C.
After 3 h, additional carbon disulfide (73 ␮l, 1.2 mmol) was
added and the reaction was stirred at 95 ◦C for 10 h. The
reaction was cooled to room temperature and concentrated.
Column chromatography (20% ethyl acetate in hexane) gave
60 mg (38%, HPLC purity: 99%) of 37 as an off-white
solid; m.p. 242–246 ◦C (decomposed); 1H NMR (400 MHz,
CDCl3) δ: 11.75 (s, 1 H), 7.33 (s, 1H), 7.26 (d, J = 8.6 Hz,
1H), 7.18 (d, J = 8.6 Hz, 1H), 6.98 (s, 2 H), 6.30 (s, 2 H),
5.49-5.57 (m, 1 H), 1.55 (d, J = 7.0 Hz, 6 H); 13C NMR
(100 MHz, CDCl3) δ: 168.2, 136.6, 131.6, 129.1, 120.1,
116.9, 110.8, 110.5, 104.1, 48.7, 20.0; HRMS m/z Calc’d
for C14H16N3S [M + H]+: 258.1065; Found 258.1060.
2.10.11. Library synthesis—exploration of the imidazolyl
position of 1
Reaction tubes (polypropylene with frits) in a Bohdan
MiniBlockTM equipped with a heating block were charged
with cesium carbonate (∼200 mg/tube, 2.5 eq.). An aliquot
(440 ␮l, 0.252 mmol) of a solution of 1a (3.45 g) in DMF
(28 ml) was added via a Tecan liquid handler into 1-dram
vials containing preweighed secondary amines and hetero-
cycles (1.1 eq.). Any resulting heterogeneous mixtures were
transferred into the Bohdan MiniBlockTM manually while
the homogeneous solutions were transferred via a liquid han-
dler. The reactions were heated at 90 ◦C for 20 h with orbital
shaking, cooled to 60 ◦C, and drained into tall microtubes
(2.5 ml) using DMF (1 ml) to rinse the reaction tubes. The
filtrates were transferred into 2-dram vials by liquid handler
and were concentrated under reduced pressure (SpeedVac).
The samples were analyzed by LCMS and were used di-
rectly for the next step.
1
m.p. = 247–251 ◦C; H NMR (500 MHz, CD3OD) δ: 7.79
(s, 1H), 7.47 (s, 1H), 7.22 (d, J = 8.2 Hz, 1H), 7.15 (s, 1H),
7.13 (d, J = 8.2 Hz, 1H), 6.99 (s, 1H), 5.56-5.50 (m, 1H),
5.31 (s, 2H), 1.54 (d, J = 7.2 Hz, 6H); HRMS m/z Calc’d
for C14H17N4S [M + H]+: 273.1174; Found 273.1172.
2.10.10. 1-Isopropyl-6-pyrrol-1-yl-1H-benzoimidazole-2-
thiol (37)
To a brown solution of 1a (300 mg, 1.5 mmol) in DMSO
(1.5 ml) were added cesium carbonate (0.986 g, 3.0 mmol)
and pyrrole (0.13 ml, 1.8 mmol). The resulting suspension
was heated in an oil bath at 100 ◦C for 20 h, and then at
140 ◦C for 6 h. The reaction was cooled to room tempera-
ture, and water (10 ml) was added, resulting in a yellow pre-
cipitate. The mixture was filtered, and the solid was rinsed
with water (2 × 10 ml), air-dried, and dried under vacuum
to give 0.343 g (93%, HPLC purity: 94%) of (2-nitro-5-
pyrrol-1-yl-phenyl)isopropylamine (37a) as a yellow solid;
1H NMR (400 MHz, CDCl3) δ: 8.24-8.14 (m, 1 H), 8.23 (d,
J = 9.7 Hz, 1 H), 7.11 (t, J = 2.1 Hz, 2 H), 6.72 (d, J =
2.1 Hz, 1 H), 6.64 (dd, J = 9.7, 2.1 Hz, 1 H), 6.38 (d, J =
2.1 Hz, 2 H), 3.89-3.80 (m, 1 H), 1.35 (d, J = 6.4 Hz, 6 H);
13C NMR (100 MHz, CDCl3) δ: 146.5, 145.8, 129.3, 129.2,
119.1, 111.9, 107.2, 103.0, 44.2, 22.6; [M + H]+: 246.32.
To a yellow suspension of 37a (0.343 g, 1.4 mmol) in
methanol (14 ml) were added ammonium chloride (0.374 g,
7.0 mmol) and zinc dust (1.83 g, 28.0 mmol). The suspension
was stirred at room temperature for 1 h and then warmed
to 60 ◦C for 1 h. The colorless reaction was cooled to room
temperature, and the mixture was filtered through a plug of
cotton and concentrated to give 0.340 g of the crude diamine
37b, as a blue solid. To a blue solution of crude 37b (0.130 g,
Ethanol (1.5 ml) was added to each vial (repeater pipette)
using an ultrasonic bath to help suspend/dissolve the less
soluble samples. Ammonium chloride (240 mg, 7 eq.) and
zinc dust (330 mg, 20 eq.) were added to each using a pre-
calibrated Argoscoop, and the reactions were shaken with
an orbital shaker at room temperature for 3 h, followed by
heating at 60 ◦C for 3 h. The samples were cooled to room
temperature and shaken overnight. The reaction mixtures
were filtered using fritted polypropylene reaction tubes into
reaction tubes in a Bohdan MiniBlockTM equipped with a
heating block using ethanol (1 ml) to complete the transfer.
Aliquots of the diamine solutions were removed for LCMS
analysis, and the samples were rapidly used for the next step.
Carbon disulfide (50 ␮l, 3 eq.) was added to each well
(multichannel repeater pipette). The reactions were shaken
on an orbital shaker at 90 ◦C overnight, then allowed to cool
to room temperature followed by precipitation of the prod-
ucts by cooling in a freezer (−40 ◦C) for 1 h. The suspen-
sions were quickly drained through a cartridge containing
CeliteTM prewetted with ethanol, and the filtrate was col-
lected into microtubes. The remaining crude benzimidazole-
2-thiols were dissolved/suspended in methanol (0.8 ml)
and filtered through the original CeliteTM cartridge into

208
Y. Dong et al. / Steroids 69 (2004) 201–217
pretared microtubes, and each well was washed with addi-
tional methanol (0.2 ml × 2). Although significant product
remained in the filtrate, the precipitation provided suffi-
cient material for testing. Crude products were purified
by preparative HPLC using the standard solvent system
but not containing TFA in order to eliminate any acid cat-
alyzed degradation due to extended exposure in the library
synthesis format.
the MMTV-Luc reporter in T47D cells, the benzimidazole-
2-thione 1 was identified as a lead compound (Tables 1
and 2) from the Bristol-Myers Squibb compound collection.
Compound 1 showed good PR transactivation potency with
an apparent EC50 value of 53 nM and efficacy of 93% (effi-
cacy of progesterone = 100% by definition). Compound 1
also showed good PR binding affinity (Ki = 89 nM). More
importantly, this compound was very selective for PR as
demonstrated by poor affinity for AR, GR, and MR, and
negligible affinity for ER (Table 3).
Representative examples:
2.10.11.1. 1-Isopropyl-6-(2-methyl-imidazol-1-yl)-1H-
benzoimidazole-2-thiol (30). 1H NMR (400 MHz, CDCl3)
δ: 8.29 (d, J = 9 Hz, 1 H), 7.27 (d, J = 2 Hz, 1 H), 7.01 (d,
J = 9 Hz, 1 H), 7.00 (d, J = 9 Hz, 1 H), 6.68 (dd, J = 9,
2 Hz, 1 H), 3.95 (m, J = 6 Hz, 1 H), 1.33 (d, J = 6 Hz, 1
H); [M + H]+: 273.12; HRMS m/z Calc’d for C14H17N4S
[M + H]+: 273.1174; Found 273.1165.
Synthetic efforts directed at optimizing the potency and
selectivity of 1 focused on exploring modifications at three
positions, two extensively and one to a lesser extent. SAR’s
at the isopropyl position and the imidazole position were
developed by using a four-step solution-phase parallel syn-
thetic approach to rapidly survey a variety of replacements
(Tables 1 and 2). This four-step solution phase route pro-
vided the final products in acceptable yields (1–35%; aver-
age yield 12%) and with good purities (70–100%; average
purity 81%). The sulfur atom was found to be critical for
activity and, as a result, was explored in a more limited way.
2.10.11.2. 6-(5-Fluoro-indol-1-yl)-1-isopropyl-1H-
benzoimidazole-2-thiol (39). 1H NMR (400 MHz, DMSO-
d6) δ: 12.96 (s, 1 H), 7.75 (d, J = 3.08 Hz, 1H), 7.72 (s,
1 H), 7.43 (m, 2 H), 7.35 (m, 2 H), 7.03 (td, J = 9.23,
2.6 Hz, 1 H), 6.68 (d, J = 3.5 Hz, 1 H), 5.43 (ddd, J =
21.2, 14.2, 7.3 Hz, 1 H), 1.52 (d, J = 7.03 Hz, 6 H); 19F
NMR (376 MHz, DMSO-d6) δ: −123.8; [M + H]+: 326.12;
HRMS m/z Calc’d for C18H17FN3S [M + H]+: 326.1127;
Found 326.1128.
3.2. Docking studies
Docking studies yielded two reasonable hypotheses for
the binding of 1 in the PR binding domain (Figs. 2 and 3 ).
Both overlaid well with progesterone in the active site. For
hypothesis I (Fig. 2), the imidazole moiety, presumably,
can form a hydrogen bond with Gln-725 and Arg-766. If
these hydrogen bonding interactions of the imidazole were
2.11. Docking studies
Ligands were docked in the active site of the pub-
lished X-ray crystal structure of the progesterone receptor
[29] using ICM Version 2.8212 (Molsoft L.L.C., 3366
North Torrey Pines Court, Suite 300, La Jolla, CA 92037,
http://www.molsoft.com/info/technology.htm). The active
site was treated as rigid, and twenty ligand conforma-
tions/poses were obtained and scored using the VLS scoring
function. These were also evaluated visually, and the best
scoring conformation/pose was chosen for the binding hy-
pothesis. Several other hypotheses were also considered
and will be discussed elsewhere (Schnur et al., manuscript
in preparation).
2.12. Statistical analysis
In vivo data were expressed as the mean S.E. Data
were analyzed by Dunnett’s one-way analysis of vari-
ance (ANOVA) using SigmaStat software (SPSS Sciences,
Chicago, IL).
3. Results
3.1. Identification of compound 1 and synthesis of analogs
Fig. 2. Hypothesis I: ICM docked thiol form (in orange) in the PR active
site compared with P4 (in white). For detailed explanation, see Results,
Section 3.2.
Using a high-throughput screening assay based on the
measurement of luciferase activity after transactivation of

Y. Dong et al. / Steroids 69 (2004) 201–217
209
important, the Ki of binding for 37 would be expected to be
weaker than that of 1. As the Ki’s were roughly comparable,
an alternate binding mode (hypothesis II, Fig. 3a), which al-
lows the imidazole of 1 to form a weak hydrogen bond with
Asn-719, seemed more probable. Although the pyrrole of
37 cannot form a hydrogen bond with Asn-719, this binding
mode appeared to be the preferred. The orientation of the
pyrrole (or imidazole) may be stabilized by nearby aromatic
side chains such as Phe-905 and Tyr-890. Both hypotheses
allow sufficient room in the binding site to predict that re-
placement of the isopropyl group with a cyclopentyl group
would retain and/or improve binding activity by filling a
steric pocket. In fact, for hypothesis II, it seemed probable
that the cyclopentyl moiety shifted the molecule slightly
in the binding site to favorably increase the interaction
between the imidazole moiety and Asn-719 (Fig. 3b). For
compound 5, docking scores for the two hypotheses support
hypothesis II with the fluoroindole moiety being too large
to occupy the same region as the imidazole in hypothesis I.
3.2.1. Comparison of compound 25 with MPA and P4 in
PR competition binding
As T47D cells contain low levels of other steroid hor-
mone receptors that may affect the binding of compounds to
PR, binding experiments using recombinant full-length hu-
man PRb were performed. Fig. 4 shows the dose-dependent
Table 1
Structure–activity relations at the imidazole position
#
R1
EC50 (nM)
53
Emax (%)
PR Ki (nM) or %a
AR Ki (nM) or %a
GR Ki (nM) or %a
Yield (%)
14
Purity (%)
97
1
93
89
38%
28%
29
–
6
4058
6%
4%
22
99
30
31
25
105
91
104
470
1176
30%
20%
4%
28
30
89
96
3060
32
110
78
414
885
23%
7
86
33
34
35
–
5790
–
50
84
97
159
443
273
872
24%
5%
17%
23%
7%
22
15
7
80
97
98
36
37
38
–
41
15
12
89
96
300
62
1158
40%
50%
13%
32%
77
23
27
31
72
99
83
144
39
3
108
600
46%
210
29
83
40
41
6340
155
78
393
428
10%
41
16
99
93
MeNH
108
1723
47%
−2%
a % inhibition at 30 ␮M.

210
Y. Dong et al. / Steroids 69 (2004) 201–217
Table 2
Structure–activity relations at the isopropyl and thiol positionsa
#
R1
–
R2
EC50 (nM) Emax (%) PR Ki (nM) or % AR Ki (nM) or % GR Ki (nM) or % Yieldb (%) Purity (%)
P4
MPA –
1
2
3
4
5
6
–
–
0.5
0.1
53
–
–
100
120
93
14
12
7
10
89
26%
52%
19
15
38%
–
–
3
0.8
28%
iPr
H
Me
Et
Pr
Bu
2-MeOEt
iPr
iPr
SH
SH
SH
SH
SH
SH
SH
OH
SMe
SOMe
SO2Me
SH
SH
SH
SH
SH
SH
SH
SH
SH
SH
SH
SH
SH
SH
SH
14
–
–
8
26
4
23
50
48
44
5
97
89
92
99
99
99
87
82
83
96
95
97
99
99
99
99
99
99
99
99
96
98
98
97
99
94
93
99
c
c
–
–
–
70
98
125
44
5
10
17
34
109
117
115
103
55
104
12
101
92
659
214
328
45%
–
34%
26%
–
14%
9%
8%
29%
35%
35%
24%
41%
21%
17%
32%
43%
29%
−0.2%
11%
25%
17%
12%
–
5648
17099
–
7
8
9
226
–
–
–
–
14934
4864
9852
8129
292
150
32
118
131
70
401
216
360
4034
56%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
iPr
iPr
iBu
2%
–
23189
15217
11
120
–
130
–
2262
9548
–
24%
17%
989
25%
22%
220
19%
6%
28%
25%
NT
13
8
(S)-2-Bu
(R)-2-Bu
(R,S)-2-Pr-3-Ome
3-Pent
(R,S)-2-Pent
(2,4-Me2)-3-pent
tBu
26
33
32
25
30
2
10
27
–
17
6
6
27
40
21
Bn
CH2CF3
(N,N-Et2N)Et
CycloPr
CycloBu
CycloPent
CycloHex
88
7
c
–
NT
–
10
4
121
211
29
20
2429
58
28
215
182
112
41%
12%
15%
NT
4%
9%
4%
108
100
128
159
109
4731
25%
15%
22%
21%
(R,S)-Exo-bicyclo[2.2.1]heptyl SH
(R,S)-Endo-bicyclo[2.2.1]heptyl SH
a These compounds are represented as the thiol tautomers to simplify the table.
b Yields are for the four step synthetic sequence.
c Compounds were part of corporate screening deck and were not resynthesized.
Table 3
Nuclear hormone selectivity profile of selective series of PR analogs
#
R1
R2
EC50
(nM)
Emax
(%)
PR Ki (nM)
or %
AR Ki (nM)
or %
GR Ki (nM)
or %
MR Ki (nM)
or %
ERA IC50
(nM) or %
P4
MPA
–
–
–
–
0.5
0.1
100
120
7
10
19
15
3
0.8
NT
166
1145
2910
1
iPr
53
93
89
58
38%
28%
4731
>17540
5247
3%
24
cBu
10
108
12%
NT
25
37
cPent
iPr
4
100
89
28
62
15%
40%
25%
32%
>17540
>17540
13%
34%
41
39
iPr
3
108
600
46%
210
3633
16378

Y. Dong et al. / Steroids 69 (2004) 201–217
211
Fig. 3. Hypothesis II: A: compound 1 with progesterone; B: compound 25 and 37 showing H bond with asparagine-719. For detailed explanation, see
Results, Section 3.2.
Fig. 4. PR competition binding curves of compound 25, P4, and MPA. Each binding experiment was done with 8 points in duplicates with 2.5 nM
[3H]-progesterone, and increasing concentrations of test compounds. Panel A: intact T47D cell binding; panel B: purified recombinant full-length PRb.
The assay procedures are described in Methods. The same experiment has been repeated at least three times, and a representative graph is presented.
competition of compounds with [3H]-progesterone bind-
ing to PR in T47D whole cell (Fig. 4A) and recombinant
full-length PRb (Fig. 4B) binding assays. Table 4 summa-
rizes the binding data from the two assays. In the T47D
whole cell binding assay, the average Ki for 25 was 28.6
5.8 nM, which was about three- to fourfold higher than the
Ki for MPA and P4. The Ki for 25, determined by the re-
combinant PR binding assay, was 17.2 4.8, which was
slightly lower than that obtained in the whole cell extract
assay.
3.2.2. Comparison of compound 25 with MPA and P4 in
gene transactivation
Table 4
The effects of 25, MPA, and P4 on the transfected MMTV
promoter-driven luciferase activity were measured in T47D
cells, and the results are shown in Fig. 5A. All three com-
pounds induced luciferase activity dose dependently, albeit
with very different potency. Compound 25 was the least po-
tent compound with an average EC50 of 3.9 0.3 nM (Table
5). P4 and MPA were 10- and 40-fold more potent than
25, respectively. The compounds were then compared for
Summary of competitive human PR binding affinities
Compound
T47D whole
cell Ki (nM)
Recombinant
PRb Ki (nM)
P4
MPA
Compound 25
6.70
10.0
28.6
3.0
3.0
5.8
11.2
3.20
17.2
5.0
1.5
4.8
Values represent mean S.D. from at least three experiments.

212
Y. Dong et al. / Steroids 69 (2004) 201–217
Fig. 5. Gene transactivation activity of compound 25, P4, and MPA. Panel A: stimulation of luciferase activity in T47D cells stably transfected with
pMAM-neo-luciferase plasmid; panel B: alkaline phosphatase activities by test compounds were measured in T47D cells. Cells were treated with various
concentrations of compounds for 16 h, and the assays were performed according to the methods in Section 2.4. Each curve has 8 points. Results shown
are from 8-point titration experiments, and each point is the mean of duplicate samples. The same experiment has been repeated at least three times, and
a representative graph is presented.
their ability to induce the endogenous AP activity in T47D
cells (Fig. 5B). All three compounds were able to induce
AP activity dose dependently with slightly higher EC50
values than the luciferase assay. The rank order of potency
remained the same as in the luciferase assay, with 25 as
the least potent (EC50 = 8.7 3 nM) but being sufficiently
potent to warrant in vivo characterization (Table 5).
3.2.4. Comparison of compound 25 with MPA and P4 in
LH release in rats
As expected, MPA and progesterone significantly reversed
the estrogen-induced inhibition of serum LH in a dose de-
pendent manner (Fig. 8). It was evident from the data that
MPA was 10-fold more potent than progesterone. Maximal
efficacy of MPA was observed at 1 mg/kg. Progesterone was
maximally effective at 10 mg/kg. In contrast, 25 had no ef-
fect on LH levels at the tested concentrations.
3.2.3. Comparison of compound 25 with MPA and P4 in a
uterotropic assay in rats
Histologic studies (Fig. 6) revealed that the endometrium
of E2-treated rats showed clear signs of hypertrophy, as
indicated by the thickening of the luminal epithelial layer
(compare A and B). Both P4 and MPA effectively prevented
the E2-induced epithelial growth (C and D). Compound 25
also significantly reduced the epithelial thickness (E and
F). Quantitative measurements of the luminal epithelial cell
height are summarized in Fig. 7. E2 alone increased epithe-
lial cell height in the endometrium 3 to 4 fold compared
to vehicle treated animals. At efficacious doses of MPA
(1 mg/kg) and progesterone (10 mg/kg), there was a 50% in-
hibition in the E2-induced increase in epithelial cell height.
Compound 25 significantly decreased cell height at the 10
and 30 mg/kg doses, reducing cell height by 16 and 25%,
respectively.
4. Discussion
This report describes the identification of a new class
of selective nonsteroidal progestins belonging to the
benzimidazole-2-thione chemotype. These compounds were
characterized in several in vitro assays to demonstrate their
potency and steroid receptor selectivity. T47D cells express
very high levels of PR and are extremely responsive to
the actions of progestins [31]. As the MMTV promoter is
modulated by progestins [32], the MMTV-driven luciferase
reporter assay in T47D cells has become a very useful tool
to screen for progestogenic compounds in high-throughput
mode. Compound 1 was discovered using this assay format
and its action through PR was confirmed by competitive
PR binding using a T47D whole cell binding assay. Al-
though 1 was not a very potent progestin (EC50 = 53 nM),
it had superior selectivity against other steroid receptors
compared to P4 and MPA, as determined by competitive
receptor binding assays. Significantly, the binding of com-
pound 1 to AR, GR, MR, and ER was too minimal for the
calculation of Ki values. Instead, percent inhibition at max-
imal compound concentration used in each binding assay
was reported to demonstrate the quantitative difference in
binding for each compound (Table 3).
Table 5
Summary of in vitro PR agonist activity
Compound
MMTV-Luciferase
Alkaline phosphatase
EC50 (nM) Efficacy
EC50 (nM) Efficacy
(%)
(%)
P4
MPA
Compound 25
0.50
0.11
3.90
0.06
0.04
0.30
100
120
107
1.12
0.45
8.70
0.03
0.05
3
100
110
113
4
6
8
10
Benzimidazole-2-thiones with aryl groups at the 5- or 6-
positions have previously been reported as antagonists of
Values represent mean S.D. from at least three experiments. Efficacy
expressed as percent relative to P4 (=100%, by definition).

Y. Dong et al. / Steroids 69 (2004) 201–217
213
Fig. 6. Histologic analysis of uterine sections. Endometrium sections of the uterus of immature rats are shown in the photomicrographs (H&E stained, 10×
magnification) A: treated with vehicle; B: E2, 2 ␮g/rat; C: E2 + P4, 10 mg/kg; D: E2 + MPA, 1 mg/kg; E: E2 + 25, 10 mg/kg; and F: E2 + 25, 30 mg/kg.
PR [30]. In 1, the presence of a heteroaryl substituent at the
6-position provided PR agonist activity. Structure–activity
relationships demonstrated that it was possible to increase
PR potency while maintaining receptor selectivity. Table 2
summarizes the 6-imidazolyl-benzimidazole-2-thione SAR
at the R1 and R2 positions. Small-branched alkyl groups
were clearly preferred for functional activity over straight-
chain alkyl, as evidenced by replacement of isopropyl (1)
with propyl (5), leading to a 100-fold loss in agonist activ-
ity while having essentially equivalent binding affinity. In-
terestingly, within this branched alkyl series there appeared
to be a strong stereochemical bias for PR agonist activity.
Compound 14, with an (R)-2-butyl group was >1000-fold
more potent as a PR agonist than its S-enantiomer 13, al-
though the binding affinity was only fivefold greater. This
stereochemical preference appeared to be present for GR
as well. Replacement of the isopropyl group with small cy-
cloalkyl groups, as suggested from the modeling experi-
ments, provided an increase in both functional potency and
binding affinity with cyclopentyl leading to the most po-
tent analog in the series. A small stereochemical bias was
also observed in this cycloalkyl series for the exo- (27) and
endo-bicycloheptyl (28) replacements, with the endo being
favored. The strict steric requirements for functional activity
were highlighted when comparing the isopropyl (1), (R)-2-
butyl (14) and cyclopentyl (25) analogs to the 3-pentyl (16)
analog. This analog, which can be viewed as a ring-opened
cyclopentyl (25) or a methyl homolog of the 2-butyl group
(13 and 14), was devoid of functional activity even though
good binding is maintained (Schnur et al., manuscript in
preparation). The sulfur atom at R2 appears to be critical for
the PR transactivation activity and PR binding affinity. Such
a switch from agonist to antagonist activity for the PR has
been described previously [21]. Replacement of the sulfur
by oxygen essentially eliminated agonist activity (8). Methy-
lation and subsequent oxidation of the resulting thiomethyl
ether (9) to methyl sulfoxide (10) and methyl sulfone (11)
resulted in a dramatic reduction in both transactivation and
binding activities.
The SAR at the 6-position within the benzimidazole-2-
thione is outlined in Table 1. Insertion of a methylene group
(29) between the imidazole group and the aryl ring led to a

214
Y. Dong et al. / Steroids 69 (2004) 201–217
Fig. 7. Inhibition of estradiol-induced epithelial cell height. Mean luminal epithelial height in the endometrium of the immature rat uterus was determined.
Rats were treated with vehicle, E2 (2 ␮g/rat), E2 + P4, E2 + MPA, or E2 + 25 at the indicated concentrations for 3 days. Data shown are average S.E..
Statistically significant (ANOVA) points are denoted ∗P < 0.05 vs. E2, n = 6.
>40-fold drop in PR binding and elimination of functional
activity. SAR around the imidazole was found to be fairly
restrictive. Substitution on the imidazole ring, either with a
methyl or a fused phenyl, reduced activity for PR or selectiv-
ity versus AR. Of the nitrogen heterocycles surveyed, only
the pyrrole (37) was found to be comparable to compound
1 in terms of both potency and selectivity. Interestingly, in-
doles (38 and 39) displayed improved PR agonist potency
but at the expense of GR binding selectivity. Moreover, com-
parison of indole (38) with benzimidazole (33) highlights
the subtle nuances among nuclear hormone receptors where
small structural changes can lead to dramatic changes in se-

Y. Dong et al. / Steroids 69 (2004) 201–217
215
of 1 and 25 to AR and GR was not enough to elicit any AR-
or GR-dependent reporter gene transactivation in cultured
cells, while both P4 and MPA showed significant activities
in the same assays (data not shown).
While the MMTV-Luc reporter assay in T47D cells was
very suitable for testing large numbers of compounds, the
assay did not offer a very high degree of specificity due to
the low levels of expression of other steroid hormone re-
ceptors (including GR and AR) in T47D cells to which the
MMTV promoter responds as well [33]. As it was observed
that MMTV-Luc reporter activity in T47D cells could indeed
be induced by high concentrations of dexamethasone and
dihydrotestosterone (data not shown), an additional func-
tional in vitro assay to further validate the progestational
activity of 25 was utilized. In T47D cells, the expression
of a membrane-associated tissue-unspecific alkaline phos-
phatase is specifically induced by progestins, but not altered
by other steroids [23]. Compound 25 behaved like a full
agonist and induced maximal alkaline phosphatase activity
with an EC50 of 9 nM (Fig. 5A and Table 5). This value was
slightly higher than that obtained with the MMTV-Luc as-
say. Compared to P4 and MPA, 25 was approximately 10-
to 40-fold less potent in both the MMTV-Luc reporter assay
and the AP assay. While none of the compounds prepared
were found to be as potent against PR in vitro as P4 or MPA,
25 was considered to possess acceptable characteristics as a
selective progestin agonist to further profile for in vitro and
in vivo progestational activities.
Inhibition of estrogen-induced uterine growth in the im-
mature rat is regarded as a classical endpoint for progestin
response. Both the morphology and the morphometric anal-
ysis of the endometrium demonstrated that 25, although less
potent than P4 and MPA, was effective in protecting the
uterus from E2-induced hypertrophic transformation (Fig.
6). Luminal epithelial cell height was significant reduced in
a dose-dependent manner (Fig. 7).
The preovulatory surge of gonadotropins leading to ovu-
lation is actively regulated by estrogen and progesterone
[34]. The ovariectomized, estrogen-primed rat model is well
validated and extensively used for the study of the effects of
progesterone on the modulation of gonadotropin secretion
[25]. In this model, progesterone could either facilitate or
block luteinizing hormone (LH) release depending on the
dose and sample time after administration. The treatment
schedule adopted in this study was expected to show a surge
in serum LH levels by progestin. Both P4 and MPA increased
serum LH levels, while 25 showed no effect (Fig. 8). Based
on the positive results obtained in the uterotropic assay and
the absence of an effect in the LH release assay, it is pos-
sible that 25 may act as a tissue-selective progestin, active
on the uterus, but inactive on the hypothalamus–pituitary
axis. The lack of an effect on LH release by 25 may be
explained by limited sensitivity of the LH release assay as
the compound has lower in vitro potency than P4 and MPA.
However, other nonsteroidal compounds with equal in vitro
potency as MPA, which also fail to induce LH release in
Fig. 8. Effects of compound 25, P4, and MPA on E2-suppressed luteiniz-
ing hormone release. Serum LH levels of estrogen-primed rats 5 h af-
ter dosing with vehicle, P4, MPA, or 25 were determined. Data shown
are average S.E.. Statistically significant (ANOVA) points are denoted
∗P < 0.05 vs. E2, n = 6.
lectivity. Replacement of the pyrrole (37) with the saturated
pyrrolidine (40) led to a >150-fold loss in functional activity
and a sixfold loss in binding affinity. The NHR selectivity
of the benzimidazole-2-thione series is more thoroughly il-
lustrated in Table 3. With the exception of 39, the analogs
shown demonstrated better binding against PR than AR, GR,
MR, and ER␣, and a significantly better selectivity profile
than P4 and MPA. Interestingly, the residual binding affinity

216
Y. Dong et al. / Steroids 69 (2004) 201–217
this model, indicate that the lack of potency is not a likely
explanation (unpublished data). Nonetheless, further im-
provement of the potency of compound 25 is necessary to
further validate the progesterone tissue-selective concept.
In summary, it has been shown that benzimidazole-2-
thione analogs are a new class of selective nonsteroidal pro-
gestin agonists. Structure–activity relation studies revealed
critical positions for improving PR potency and receptor se-
lectivity. Compound 25, the lead analog in this series, has
demonstrated excellent PR-specific binding properties and
good progestational activity in vitro. The results from the
in vivo rat models may indicate that this compound exhibits
a tissue-selective progestogenic modulatory profile. Further
studies to more fully explore the SAR for potency and se-
lectivity in this series, as well as to evaluate the in vivo ef-
fects of a selective progesterone receptor modulator will be
required for a selective PR agonist.
[13] Editorial Progestins: present and future. J Steroid Biochem Mol Biol
1996;59:357–63.
[14] Pathirana C, Stein RB, Berger TS, Fenical W, Ianiro T, Mais DE, et
al. Nonsteroidal human progesterone receptor modulators from the
marine alga Cymopolia barbata. Mol Pharmacol 1995;47:630–5.
[15] Kurihara KI, Tanabe K, Yamamoto Y, Shinei R, Ajito K, Okonogi
T. Synthesis and structure-activity relationships of new non-steroidal
progesterone receptor ligands. Bioorg Med Chem Lett 1999;9:1837–
42.
[16] Dukes M, Furr BJ, Hughes LR, Tucker H, Woodburn JR.
Nonsteroidal progestins and antiprogestins related to flutamide.
Steroids 2000;65:725–31.
[17] Zhi L, Tegley CM, Pio B, West SJ, Marschke KB, Mais DE,
et al. Nonsteroidal progesterone receptor antagonists based on 6-
thiophenehydroquinolines. Bioorg Med Chem Lett 2000;10:415–8.
[18] Zhang P, Terefenko EA, Fensome A, Zhang Z, Zhu Y, Cohen J,
et al. Potent nonsteroidal progesterone receptor agonists. Synthesis
and SAR study of 6-aryl benzoxazines. Bioorg Med Chem Lett
2002;12:787–90.
[19] Palmer S, Campen CA, Allan GF, Rybczynski P, Haynes-Johnson
D, Hutchins A, et al. Nonsteroidal progesterone receptor ligands
with unprecedented receptor selectivity. J Steroid Biochem Mol Biol
2000;75:33–42.
[20] Tabata Y, Iizuka Y, Kashiwa J, Masuda NT, Shinei R, Kurihara K,
et al. Fungal metabolites, PF1092 compounds and their derivatives,
are nonsteroidal and selective progesterone receptor modulators. Eur
J Pharmacol 2001;430:159–65.
[21] Zhang P, Terefenko EA, Fensome A, Wrobel J, Winneker R,
Zhang Z. Novel 6-aryl-1,4-dihydrobenzo[d][1,3]oxazine-2-thiones as
potent, selective, and orally active nonsteroidal progesterone receptor
agonists. Bioorg Med Chem Lett 2003;13:1313–6.
[22] Fensome A, Koko M, Wrobel J, Zhang P, Zhang Z, Cohen J,
et al. Novel 5-aryl-1,3-dihydroxy-indole-2-thiones: potent, orally
active progesterone receptor agonists. Bioorg Med Chem Lett
2003;13:1317–20.
[23] Di Lorenzo D, Albertini A, Zava D. Progestin regulation of alkaline
phosphatase in the human breast cancer cell line T47D. Cancer Res
1991;51:4470–5.
Acknowledgements
The authors would like to thank all the progesterone re-
ceptor working group members at Bristol-Myers Squibb. We
also want to thank Dr. Brian Arey for his contribution to
the LH release assay, Barbara Bujak, Kimberly Glover and
Christopher Hasson for their work in processing, sectioning,
and staining the uterine tissue samples and Lauren Barber
for cell culture support during the high-throughput screen-
ing campaign.
References
[24] De Gasparo M, Joss U, Ramjoue SE, Whitebread SE, Haenni
H, Schenkel L, et al. Three new epoxy-spirolactone derivatives:
[1] Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery
Jr CA, et al. Mice lacking progesterone receptor exhibit pleiotropic
reproductive abnormalities. Genes Dev 1995;9:2266–78.
[2] Mandel FP, Davidson BJ, Erlik Y, Judd HL, Meldrum DR. Effects of
progestins on bone metabolism in postmenopausal women. J Reprod
Med 1982;8:511–4.
characterization in vivo and in vitro.
1987;240:650–6.
[25] Brann DW, Putnam CD, Mahesh VB. Validation of the mechanisms
proposed for the stimulatory effects of progesterone on gonadotropin
secretion in the estrogen-primed rat: a possible role for adrenal
steroids. Steroids 1991;56:103–10.
J Pharm Exp Therap
[3] DeCherney A. Physiologic and pharmacologic effects of estrogen
and progestins on bone. J Reprod Med 1993;38:1007–14.
[4] Ramirez VD, Dluzen DE, Ke FC. Effects of progesterone
and its metabolites on neuronal membranes. Ciba Found Symp
1990;153:125–41.
[5] McEwen BS. Non-genomic and genomic effects of steroids on neural
activity. TIPS 1991;12:141–7.
[6] Baulieu EE, Schumacher M. Neurosteroids with special reference to
the effect of progesterone on myelination in peripheral nerves. Mult
Scler 1997;3:105–12.
[26] Bru-Magniez N, Güngör T, Lacrampe J, Launay M, Teulon J-M.
Nouveaux dérivés de benzimidazoles et azabenzimidazoles, leurs
procédés de preparation, intermédiaires de synthèse, compositions
pharmaceutiques les contenant, utiles notamment pour le traitment
des maladies cardiovasculaires, et les ulcères duodénaux. European
Patent Office. Issue 90400537.8, France, Laboratoire UPSA. 1990;70.
[27] Güngör T, Fouquet A, Teulon J-M, Provost D, Cazes M, Cloarec
A. Cardiotonic agents: synthesis and cardiovascular properties of
novel 2-arylbenzimidazoles and azabenzimidazoles. J Med Chem
1992;35:4455–63.
[7] Collins DC. Sex hormone receptor binding, progestin selectivity, and
new oral contraceptives. Am J Obstet Gynecol 1994;170:1508–13.
[8] Mishell Jr DR. Pharmacokinetics of depot medroxyprogesterone
acetate contraception. J Reprod Med [Suppl] 1996;41:381–90.
[9] De Ziegler D, Fanchin R. Progesterone and progestins: applications
in gynecology. Steroids 2000;65:671–9.
[10] Schiff I. The effects of progestins on vasomotor flushes. J Reprod
Med [Suppl] 1982;27:498–502.
[11] Lundgren S. Progestins in breast cancer treatment. Acta Oncol
1992;31:709–22.
[28] Boger DL, Zhou J, Borzilleri RM, Nukui S, Castle SL. Synthesis
of (9R,12S)- and (9S,12S)-Cycloisodityrosine and their N-methyl
derivatives. J Org Chem 1997;62(7):2054.
[29] Williams SP, Sigler PB. Atomic structure of progesterone complexed
with its receptor. Nature 1998;393:392–6.
[30] Zhang P, Terefenko E A, Wrobel J, Zhang Z, Zhu Y, Cohen J, et
al. Synthesis and progesterone receptor antagonist activities of 6-
aryl benzimidazolones and benzothiazolones. Bioorg Med Chem Lett
2001;11:2747–50.
[31] Horwitz KB, Zava DT, Thilagar AK, Jensen EM, McGuire WL.
Steroid receptor analysis of nine human breast cancer cell lines.
Cancer Res 1978;38:2434–7.
[12] Schneider HPG, Jackisch C. Potential benefits of estrogens and
progestogens on breast cancer. Int J Fertil 1998;43:278–85.

Y. Dong et al. / Steroids 69 (2004) 201–217
217
[32] Cato AC, Miksicek R, Schutz G, Arnemann J, Beato M. The hormone
regulatory element of mouse mammary tumor virus mediates
progesterone induction. EMBO J 1986;5:2237–40.
[33] List HJ, Lozano C, Lu J, Danielsen M, Wellstein A, Riegel AT.
Comparison of chromatin remodeling and transcriptional activation
of the mouse mammary tumor virus promoter by the androgen and
glucocorticoid receptor. Exp Cell Res 1999;250:414–22.
[34] Mahesh VB, Brann DW. Regulation of the preovulatory
gonadotropin surge by endogenous steroids. Steroids 1998;63:616–
29.