Please Scroll Down to See Forums Below 
.jpg "Research Chemical Sciences")
.png "Research Chemical Sciences")
lanky
Well-known member
Pharmacological and X-ray Structural Characterization of a Novel
Selective Androgen Receptor Modulator: Potent Hyperanabolic
Stimulation of Skeletal Muscle with Hypostimulation of Prostate in Rats
Abbreviated Title: Muscle Selective Androgen Receptor Agonist
Jacek Ostrowski*,1, Joyce E. Kuhns1, John A. Lupisella1, Mark C. Manfredi2, Blake C.
Beehler1, Stanley R. Krystek, Jr.3, Yingzhi Bi2,†, Chongqing Sun2, Ramakrishna
Seethala1, Rajasree Golla1, Paul G. Sleph1, Aberra Fura4, Yongmi An3, Kevin F. Kish3,
John S. Sack3, Kasim A. Mookhtiar1,§, Gary J. Grover1,¶ and Lawrence G. Hamann*,2
Departments of
1
Metabolic Diseases,
2
Discovery Chemistry,
3
Macromolecular Structure,
4
Pharmaceutical Candidate Optimization, Bristol-Myers Squibb Pharmaceutical
Research Institute, P.O. Box 5400, Princeton, NJ 08543, USA.
Key Words: Androgen Receptor, Tissue Selectivity, SARM, Anabolic Agent
*Correspondence should be addressed to Lawrence G. Hamann at: Department of
Discovery Chemistry, Bristol-Myers Squibb, Pharmaceutical Research Institute, 5
Research Parkway, Wallingford CT 06492; telephone: 203-677-6948. fax: 203-677-
7884. e-mail: [email protected]
†Present address: Lexicon Pharmaceuticals, 350 Carter Road, Princeton, NJ 08540.
email: [email protected]
§Present address: Advinus Therapeutics, Pune, India. email:
[email protected]
¶Present address: Product Safety Laboratories, 2394 Route 130, Dayton, NJ 08810.
email: [email protected]
DISCLOSURE STATEMENT: J.O., J.K., J.L., M.M., B.B., S.K., C.S., R.S., R.G., P.S.,
A.F., Y.A., K.K., J.S., and L.H. are employed by Bristol-Myers Squibb. Y.B., K.M., and
G.G. were previously employed by Bristol-Myers Squibb. M.M., A.F., J.S., K.M., G.G.,
and L.H. have equity interests in Bristol-Myers Squibb.
Endocrinology. First published ahead of print September 28, 2006 as doi:10.1210/en.2006-0843
Copyright (C) 2006 by The Endocrine Society
2
Abstract
A novel, highly potent, orally active, non-steroidal tissue selective androgen receptor
modulator (BMS-564929) has been identified, and this compound has been
advanced to clinical trials for the treatment of age-related functional decline. BMS-
564929 is a subnanomolar androgen receptor agonist in vitro, is highly selective for
the androgen receptor versus other steroid hormone receptors, and exhibits no
significant interactions with sex hormone binding globulin or aromatase. Dose
response studies in castrated male rats show that BMS-564929 is substantially more
potent than testosterone in stimulating the growth of the levator ani muscle, and
unlike testosterone, highly selective for muscle versus prostate. Key differences in
the binding interactions of BMS-564929 with the androgen receptor relative to the
native hormones were revealed through x-ray crystallography, including several
unique contacts located in specific helices of the ligand binding domain important
for coregulatory protein recruitment. Results from additional pharmacological
studies effectively exclude alternative mechanistic contributions to the observed
tissue selectivity of this unique, orally active androgen. As concerns regarding the
potential hyperstimulatory effects on prostate and an inconvenient route of
administration are major drawbacks which limit testosterone’s clinical use, the
potent oral activity and tissue selectivity exhibited by BMS-564929 is expected to
yield a clinical profile which provides the demonstrated beneficial effects of
testosterone in muscle and other tissues with a more favorable safety window.
3
Serum androgen (testosterone, T) levels decline progressively with aging in men,
beginning in the third decade of life at the rate of approximately one percent per year (1-
3). Furthermore, ninety-eight percent of T in men is not freely available to most tissues
because it is bound with high affinity to sex hormone binding globulin (SHBG) (4), a
serum protein whose concentration increases with age. This decline in available T is
associated with alterations in body composition, diminished energy, diminished muscle
strength and physical function, reduced sexual function, and depressed mood (5-9). This
androgen deficient state in aging males, often called andropause or androgen decline in
the aging male (ADAM), can subsequently lead to age-related functional decline (frailty).
Several clinical studies have shown that supplementation of T at physiologic doses in
elderly men results in a significant increase in lean body mass, a decrease in adipose
tissue, and an increase in muscle strength (10-15). However, use of T replacement
therapy (TRT) in elderly males by primary care physicians is limited due to concerns
about potential side effects (16-20). Among the more serious side effects are
hyperstimulation of the prostate, which may be a preamble to or exacerbate occult
subclinical benign prostatic hypertrophy (BPH) and/or prostate cancer, and increases in
hematocrit.
T cannot be administered orally due to its rapid and extensive metabolism, and
consequently, is given by injection, transdermal patch or in a recently introduced gel
form (21, 22). Each of these delivery routes has certain drawbacks: the need for needles
and physician administration for the former, skin irritation and potential for contact
transfer to others with the latter. Several T analogues with either 7α- or 17α-alkyl
substitution, including 7α-methyl-19-nortestosterone (MENT) (23) and oxandrolone (24),
4
have been used clinically in place of T, as these are shown to circumvent metabolism and
improve oral bioavailability and half-life. However, 17α-alkylated androgens have
frequently been associated with hepatotoxicity (25), although, these effects are
hypothesized to be structure– and not mechanism–related. There are currently no
approved drugs indicated for the prevention of functional decline in the aging male, but
the growing desire among the aging population to remain independent has raised
awareness within the medical community regarding the potential for therapeutic
intervention. Consequently, there exists an unmet medical need for safe and effective
oral therapies which might offer greater separation between the desired anabolic and
undesired androgenic effects. In theory, a selective androgen receptor modulator
(SARM) (26) has the potential to maintain or improve muscle strength and function,
thereby enabling or facilitating the performance of everyday activities, and to improve
body composition, mental health, libido and sexual function in elderly men without the
concomitant deleterious effects on prostate, liver, or erythrocytes associated with steroid
treatment regimens.
Since the first identification of orally-active non-steroidal androgens (27), several
groups have actively pursued the identification and development of SARMs (28-30), and
some of the most advanced compounds to emerge from these efforts have recently
reached the clinic (31). However, these compounds exhibit limited efficacy and modest
tissue selectivity in preclinical models. As part of our program to discover and develop
novel treatments for age-related functional decline, we have discovered a novel, potent,
orally bioavailable SARM, BMS-564929 (Fig. 1). which, analagous to the natural
hormone agonists T and 5-dihydrotestosterone (DHT), exerts its effects via the regulation
5
of androgen receptor (AR) mediated gene transcription in tissues which express the AR
(32). This highly potent compound exhibits remarkable muscle vs. prostate selectivity in
vitro and in chronic in vivo models of androgen action in rodents. The mechanism by
which BMS-564929 tissue selectivity is achieved is anticipated to be analogous to that of
the more well-characterized selective estrogen receptor modulators (SERMs) (33-36),
which involves ligand-based selective activation of target genes through differential
recruitment of co-factors present in various tissues expressing the receptor (37, 38).
Recent efforts directed at beginning to understand the structural basis for ligand induced
differential gene expression through the AR are making advances toward characterizing
these relationships in greater detail at the molecular level (39). X-ray data of the AR
ligand binding domain (AR-LBD) bound to BMS-564929 and DHT shows that BMS-
564929 has unique binding interactions compared to DHT. Therefore, binding of BMS-
564929 to the AR-LBD results in a receptor conformation disparate to that induced by
DHT, which may thereby allow the BMS-564929-AR complex to interact with a different
set of coactivators/corepressors than the DHT-AR complex. Through additional
investigations, we have been able to rule out contributions from a number of potential
alternative selectivity mechanisms. Based on the present results, we believe that BMS-
564929 is the most potent and muscle-selective, orally-available AR agonist reported to
date. Consequently this compound has advanced into human clinical trials for the
treatment of age-related functional decline.
Materials and Methods
Reagents
[3H]-DHT, [3H]-progesterone and [3H]-estradiol were obtained from Perkin-Elmer.
SARM compound BMS-564929 [(7R,7aS)-2-(3-chloro-4-cyano-2-methylphenyl)-7-
hydroxytetrahydro-2H-pyrrolo[1,2-e]imidazole-1,3-dione] was synthesized in our
laboratories by the following procedures as outlined schematically (Fig. 2).
(3S)-N-tert-Butoxycarbonyl-3-hydroxy-L-proline methyl ester. Hydrogen chloride
gas was bubbled through a suspension of trans-3-hydroxy-L-proline (50 g, 0.38 mol) in
MeOH (600 mL) at 0 °C for 10 min. The resulting clear solution was stirred at room
temperature for 4 h, then concentrated under reduced pressure. The resulting white solid
was dried in vacuo overnight to afford 68.3 g of the title compound. The hydroxyproline
ester thus obtained was suspended in CH2Cl2 (1.0 L) cooled to 0 °C, and Et3N (105.3 mL,
0.755 mol) was added, followed by portionwise addition of di-tert-butyl dicarbonate
(82.96 g, 0.380 mol). The resulting mixture was stirred at room temperature for 4 h, then
partitioned between water and CH2Cl2. The CH2Cl2 layer was washed twice with water,
and once each with 20% aqueous citric acid, water and brine, then dried over Na2SO4 and
concentrated under reduced pressure to give an oily residue. The crude product was
chromatographed (silica gel) eluting with 15%-50% EtOAc/hexane to afford the
diprotected compound (73.3 g) as a pale yellow viscous oil. 1H NMR (CDCl3, 400 MHz,
3:2 mixture of rotamers, data for major rotamer) δ 1.41 (s, 9H), 1.93 (m, 1H), 2.11 (m,
1H), 2.45 (d, J = 4.8 Hz, 1H), 3.60 (m, 2H), 3.75 (s, 3H), 4.18 (d, J = 1.3 Hz, 1H), 4.44
(m, 1H). (3R)-N-tert-Butyloxycarbonyl-3-benzoyloxy-L-proline methyl ester. To a
7
stirred solution of (3S)-N-tert-butoxycarbonyl-3-hydroxy-L-proline methyl ester (69.1 g,
0.282 mol), Ph3P (88.7 g, 0.338 mol) and benzoic acid (41.3 g, 0.338 mol) in anhydrous
THF (1.35 L) cooled to 0 °C was added dropwise over 1 h (through an addition funnel) a
solution of diethylazodicarboxylate (62 mL, 0.33 mol) in anhydrous THF (50 mL). After
addition, the resulting light yellow solution was stirred at room temperature for 8 h. The
reaction mixture was then partitioned between EtOAc and aqueous NaHCO3. The
organic layer was washed with saturated aqueous NaHCO3, water (2×), brine, dried over
Na2SO4 and concentrated under reduced pressure to yield a crude product as a semi-solid.
The crude benzoate product was suspended in 25% EtOAc/hexane and stirred vigorously
for 3 h. The resulting suspension was filtered and the collected white solid (Ph3PO)
rinsed twice with 20% EtOAc/hexane. The combined filtrates were concentrated under
reduced pressure to yield an oily residue, which was triturated twice with 20%
EtOAc/hexane as described above to yield approximately 150 g of partially purified
product as a yellow oil, which was further purified by flash chromatography (silica gel)
eluting with 10-20% EtOAc/hexane to furnish the pure title compound (88.4 g) as a light
yellow viscous oil. 1H NMR (400 MHz, CDCl3, 2:1 mixture of rotamers, data for major
rotamer) δ 1.39 (s, 9H), 2.26 (m, 2H), 3.57-3.75 (m, 2H), 3.62 (s, 3H), 4.65 (d, J = 4.0
Hz), 5.71 (q, J = 4.0 Hz), 7.45 (t, J = 4.0 Hz, 2H), 7.58 (t, J = 4.0 Hz, 1H), 7.98 (d, J =
4.0 Hz, 2H). (3R)-3-Hydroxy-L-proline methyl ester. To a solution of (3R)-N-tertbutyloxycarbonyl-
3-benzoyloxy-L-proline methyl ester (88.4 g, 0.253 mol) in anhydrous
MeOH (700 mL) at 0 °C was slowly added through an addition funnel a freshly prepared
1N solution of KOH in anhydrous MeOH (367 mL, 0.367 mol) over 25 min. After the
addition, the resultant light yellow solution was stirred at 0 °C for 2 h, then the reaction
8
quenched by slow addition (over 25 min) of a solution of 1N HCl in dioxane/EtOAc (380
mL) through an addition funnel. The resulting white suspension was concentrated under
reduced pressure to remove most of the solvent, and the remaining mixture was
partitioned between water and EtOAc. The separated organic phase was washed with
water (2×), saturated aqueous NaHCO3 (2×), water, brine, and dried over Na2SO4. The
filtrate was concentrated under reduced pressure to give a light yellow oily residue, which
was chromatographed (silica gel) eluting first with 25-30% EtOAc/hexane, then 5%
MeOH in 30% EtOAc/hexane to furnish the N-Boc-protected cis-hydroxy-L-proline
methyl ester (44.6 g) as a pale yellow oil. To a solution of the free hydroxy compound
thus obtained (44.6 g, 0.182 mol) in CH2Cl2 (450 mL) at 0 °C was slowly added TFA
(275 mL) through an addition funnel over 40 min. After the addition, the reaction
mixture was stirred at 0 °C for 2 h, then concentrated under reduced pressure to give a
viscous oily residue which was evaporated with ether (2×), toluene (1×), ether (2×) and
dried in vacuo overnight to yield the Boc-deprotected TFA salt of cis-hydroxy-L-proline
methyl ester (59.5 g) as a light yellow solid. A solution of this TFA salt (12.64 g, 48.8
mmol) in MeOH (150 mL) was free based by treatment with WA21J resin (Diaion®
WA21J, polyamine resin from Supelco, Bellefonte, PA, 60 g). The resulting suspension
was stirred at room temperature for 1 h, and then filtered. The collected resin was rinsed
with MeOH (2×) and combined filtrate concentrated carefully under reduced pressure to
give the desired cis-hydroxy-L-proline methyl ester (7.7 g) as a colorless oil. [α]D =
+14.9° (c. 1.0, MeOH); 1H NMR (400 MHz, CD3OD) δ 1.86 (m, 1H), 2.02 (m, 1H), 2.87
(m, 1H), 3.24 (m, 1H), 3.69 (d, J = 4.0 Hz, 1H), 3.75 (s, 3H), 4.48 (t, J = 4.0 Hz, 1H);
HPLC: 100% at 0.157 min (retention time) (Conditions: Phenominex Luna C18 (4.6 x 50
9
mm); eluted with 0% to 100% B; 4 min gradient (A = 90% H2O/10% MeOH/0.1%
H3PO4 and B = 10% H2O/90% MeOH/0.1% H3PO4 ), flow rate at 4 mL/min., UV
detection at 220 nm); MS (ES): m/z 146 [M+H]+
N-(4-Bromo-3-chloro-2-methylphenyl)acetamide. To a solution of commercially
obtained 3-chloro-2-methylaniline (10.0 g, 70.6 mmol) in EtOH (85 mL) at room
temperature was slowly added acetic anhydride (8.00 mL, 84.7 mmol) and the reaction
mixture was stirred for 15 min. The resulting suspension was concentrated and dried in
vacuo to provide the desired acetamide (17.0 g) as a red-brown solid. To a solution of
this acetamide (13.0 g, 70.6 mmol) in AcOH (100 mL) at 15 °C was added bromine (10.9
mL, 212 mmol) over 20 min. The solution was allowed to warm to room temperature
and stir for 1 h. The solution was then poured into ice water while stirring, and the
precipitate which formed was filtered, washed with water until the filtrate was at neutral
pH, and dried to provide the desired bromide (17.6 g) as a pale orange solid. 1H NMR
(400 MHz, DMSO-d6) δ 2.05 (s, 3H), 2.28 (s, 3H), 7.29 (d, J = 8.25 Hz, 1H), 7.56 (d, J =
8.80 Hz, 1H), 9.60 (s, 1H); 13C NMR (100 MHz, DMSO-d6) δ 16.71, 23.14, 118.05,
125.48, 130.36, 132.68, 133.39, 137.14, 168.44; HPLC a) column: Phenominex ODS
C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90% H2O/0.1% TFA to 90% MeOH/10%
H2O/0.1% TFA; 1 min hold, 4 mL/min UV detection at 220 nm, 2.95 min retention time;
HPLC b) column: Shimadzu Shim-Pack VP-ODS C18 4.6 x 50 mm, 4 min gradient, 10%
MeOH/90% H2O/0.1% TFA to 90% MeOH/10% H2O/0.1% TFA, 1 min hold; 4 mL/min,
UV detection at 220 nm, 2.87 min retention time (98%); MS (ES) m/z 263 [M+H]+. 4-
Amino-2-chloro-3-methylbenzonitrile. A suspension of the bromide thus obtained
(17.5 g, 66.7 mmol) and copper cyanide (7.16 g, 80.0 mmol) in DMF (200 mL) was
10
heated to 150 °C for 5 h. The solution was then cooled and poured into ice water while
stirring. The precipitate which formed was filtered, washed with water and dried. The
solid was triturated in refluxing methanol, filtered and the filtrate concentrated to provide
the nitrile (9.94 g) as a brown solid. A suspension of the nitrile thus obtained (9.90 g,
47.4 mmol) in conc. HCl (50 mL) and EtOH (50 mL) was heated at reflux for 30 min.
The solution was concentrated and dried to provide the hydrochloride salt of the desired
aniline (9.41 g) as a brown solid. 1H NMR (400 MHz, DMSO-d6) δ 2.12 (s, 3H), 6.30 (s,
2H), 6.61 (d, J = 8.25 Hz, 1H), 7.36 (d, J = 8.25 Hz, 1H); 13C NMR (100 MHz, DMSOd6)
δ 13.83, 96.86, 112.11, 118.31, 118.85, 132.16, 135.56, 152.52; HPLC a) column:
Phenominex ODS C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90% H2O/0.1% TFA to
90% MeOH/10% H2O/0.1% TFA; 1 min hold, 4 mL/min UV detection at 220 nm, 2.42
min retention time; HPLC b) column: Shimadzu Shim-Pack VP-ODS C18 4.6 x 50 mm,
4 min gradient, 10% MeOH/90% H2O/0.1% TFA to 90% MeOH/10% H2O/0.1% TFA, 1
min hold; 4 mL/min, UV detection at 220 nm, 2.31 min retention time (99%); MS (ES):
m/z 167 [M+H]+. 2-Chloro-4-isocyanato-3-methylbenzonitrile. To a mixture of the
hydrochloride salt of 4-amino-2-chloro-3-methylbenzonitrile (5.78 g, 28.4 mmol) in
CH2Cl2 (230 mL) at 0 °C was added NaHCO3 (23.9 g, 284 mmol) followed by 20%
phosgene in toluene (60.2 mL, 114 mmol) over 30 min. The mixture was allowed to
warm to room temperature and stir for 1.5 h. The mixture was filtered and the filtrate
concentrated. The residue was azeotroped with toluene (3 × 100 mL) and concentrated to
provide the isocyanate (5.45 g) as a pale, orange solid which was used immediately in the
next step without further purification.
11
(7R,7aS)-2-Chloro-4-(7-hydroxy-1,3-dioxotetrahydropyrrolo[1,2-c]imidazol-2-yl)-3-
methylbenzonitrile (BMS-564929). A suspension of 2-chloro-4-isocyanato-3-
methylbenzonitrile (5.45 g, 28.3 mmol), diisopropylethylamine (5.95 mL, 34.1 mmol), 4
Å molecular sieves (5 g) and (3R)-3-hydroxy-L-proline methyl ester (7.37 g, 28.4 mmol)
in CH2Cl2 (300 mL) was stirred at room temperature for 1.5 h. Diazabicycloundecane
(5.10 mL) was then added, and after 18 h at room temperature, the suspension was
filtered and the solid triturated with acetone, refiltered and the combined filtrates were
concentrated in vacuo. The residue was chromatographed on silica gel (CH2Cl2/CH3OH;
49:1 to 9:1 gradient) to provide BMS-564929 (4.08 g) as a white solid. 1H NMR (400
MHz, DMSO-d6) δ 2.05-2.11 (m, 1H), 2.15-2.22 (m, 1H), 2.20, 2.24 (s, 3H), 3.29-3.33
(m, 1H), 3.59-3.68 (m, 1H), 4.42-4.50 (m, 2H), 5.64, 5.72 (d, J = 3.85, 3.30 Hz, 1H),
7.22, 7.51 (d, J = 8.25 Hz, 1H), 7.96 (d, J = 8.25, 1H); 13C NMR (100 MHz, DMSO-d6) δ
15.44, 15.63, 35.49, 35.62, 43.30, 43.41, 68.76, 69.25, 69.84, 112.86, 113.08, 115.79,
128.11, 128.73, 132.07, 136.27, 136.42, 136.85, 137.12, 158.63, 169.09, 169.60; HPLC
a) column: Phenominex ODS C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90%
H2O/0.1% TFA to 90% MeOH/10% H2O/0.1% TFA; 1 min hold, 4 mL/min UV detection
at 220 nm, 2.07, 2.32 min retention time; HPLC b) column: Shimadzu Shim-Pack VPODS
C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90% H2O/0.1% TFA to 90%
MeOH/10% H2O/0.1% TFA, 1 min hold; 4 mL/min, UV detection at 220 nm, 1.93, 2.23
min retention time (97%); HPLC c) column: Daicel Chiralcel OD 4.6 x 250 mm, isocratic
25% isopropanol/hexanes, 30 min, 1 mL/min, UV detection at 220 nm, 10.99 min
retention time (98%); MS (ES): m/z 306 [M+H]+.
12
Receptor and binding protein binding assays
The human cancer epithelial breast cell lines MDA MB-453 and T47D, which
endogenously express AR and PR, respectively, were used for radioligand competition
binding assays. Binding assays were conducted by incubating BMS-564929 at various
concentrations with either [3H]-DHT or [3H]-progesterone with the cells for 2 hours at
room temperature. For ERα and ERβ, fusion proteins expressed in Escherichia coli,
consisting of maltose binding protein (MBP), a specific biotinylation sequence (BioP), an
enterokinase cleavage site (EK) and either the ERα or ERβ LBD was used. Binding
reactions were conducted by incubating ERα and ERβ LBD with BMS-564929 and [3H]-
estradiol for 2 hours at room temperature. Specific binding activity to the
mineralocorticoid receptor (MR) by BMS-564929 was evaluated by competition binding
assay using kidney cytosolic preparations and [3H]-aldosterone. The kidneys were
obtained from adrenalectomized (ADX) rats in order to remove the endogenous source of
aldosterone and to increase the MR concentration in the cytosol of kidney cells. Binding
reactions were incubated for 2 hours on ice in the presence of excess mifepristone
(RU486, Sigma) to block non-specific glucocorticoid receptor (GR) binding. A
fluorescence polarization based assay (Glucocorticoid Receptor Competitor Assay Kit
Red, PANVERA) was used for GR binding, as per manufacturer recommendations.
Inhibitory constants (Ki, app) defining apparent binding affinity of test compounds to
intracellular receptors were calculated from the observed inhibition of natural ligand
binding at multiple concentrations of test compound. Sex hormone binding globulin
(SHBG, Research Diagnostics) binding was performed using a standard charcoal assay.
13
Reagents: 1 mg lyophylized SHGB powder (Tris), 3H DHT (NEN), 3% charcoal and
0.4% Dextran in PBS; binding buffer: 50 mM Tris, pH 7.6, 100 mM NaCl, 1 mM EDTA,
1 mM DTT and mock lysate (3.5 µg /100 µL buffer); stock solutions: stock SHBG
protein: 1 mg/mL in water = 20 µM, stock 3H-DHT ligand: 9 µM, DHT: 10 mM in
DMSO, BMS 564929: 10 mM in DMSO. Compounds diluted in binding buffer were
added to 40 nM 3H-DHT and 20 nM SHBG protein in 200 µL volume and incubated for
1 hour at room temperature. Total Binding: 40 nM 3H-DHT and 20 nM SHBG protein in
200 µL volume, Non Specific Binding: 40 nM 3H-DHT and 20 nM SHBG protein and
1mM cold DHT in 200 µL volume. At the end of the incubation period, 200 µL of the
charcoal solution (3% containing 0.04% dextran) was added to 200 µL of the reactions
and shaken for 15 minutes before centrifugation. Supernatent (200 µL) was then
transferred to the wells of 24-well white Optiplate, 200 µL of scintillant were added with
mixing. Radioactivity counts were read in Topcount.
Stable androgen receptor-luciferase reporter transactivation assays
Functional transactivation assays were conducted to assess the potency and efficacy of
androgen agonists in muscle and prostate cell backgrounds. We stably transfected
C2C12 mouse myoblast (ATCC) and rat PEC prostate epithelial cell lines (obtained from
Dr. Douglas Hixson, Rhode Island Hospital) with both full length rat AR and the
enhancer/luciferase reporter. The enhancer/reporter construct, containing the 2XDR-1
ARE was developed by random mutagenesis of an AR/GR consensus response element
sequence and has been used as an AR specific response element based on its AR
14
selectivity demonstrated in CV-1 cells (40). We plated both cell lines in 96 well format
at 6,000 cells/well in high glucose DMEM without phenol red (Gibco BRL, Cat. No.:
21063-029) containing 10% charcoal and dextran treated FBS (HyClone Cat. No.:
SH30068.02), 50 mM HEPES Buffer (Gibco BRL, Cat. No.: 15630-080), 1X MEM Na
Pyruvate (Gibco BRL, Cat. No.: 11360-070), 0.5X Antibiotic-Antimycotic, 800 µg/mL
Geneticin (Gibco BRL, Cat. No.: 10131-035) and 800 µg/mL Hygromycin β(Gibco BRL,
Cat. No.: 10687-010). Forty eight hours later, we added ligands (10 µL/well) in fresh
media. After an additional 24 hours, we used the Steady-Glo™ Luciferase Assay System
to detect activity according to the manufacturer’s instructions (Promega, Cat. No.:
E2520).
Enzymatic assays
A CYP19 aromatase inhibition assay using human CYP19+P450 Reductase
Supersomes™ GENTEST P260 was used to measure inhibition of aromatase by BMS-
564929. IC50’s were determined utilizing assay conditions recommended by the
manufacturer (BD Gentest Corporation).
Rodent muscle and prostate tissue growth assays
Matched sets of castrated, sexually mature Harlan Sprague-Dawley rats (42-56 days old,
200-250 g), were dosed once daily by oral gavage with BMS-564929 (0.00001 mg/kg –
10 mg/kg) in solution/suspension of 80% PEG 400 and 20% Tween 20 (PEG/TW, both
15
obtained from Sigma Chemical) for 14 days. Two control groups, one sham operated
intact and one castrated, were dosed orally with the PEG/TW vehicle only, beginning on
day 15 following surgery. Animals were dosed (v/w) at 1 mL/kg body weight. TP was
dosed once daily sub-cutaneously in a 10% ethanol/90% peanut oil vehicle as a reference
compound (0.03 mg/kg – 10 mg/kg). After 14 days of treatment, the animals were
sacrificed by carbon dioxide asphyxiation and the levator ani and the ventral prostate
were surgically removed and weighed, and serum was collected for LH measurements.
All experiments were conducted in accordance with regulations of the Animal Care and
Use Committee of the Bristol-Myers Squibb Company, in facilities fully accredited by
the Association for Assessment and Accreditation of Laboratory Animal Care.
Rat luteinizing hormone (rLH) measurements
rLH was quantitatively determined by radioimmunoassay using a Biotrak [125I] kit
(Amersham Pharmacia Biotech) under the manufacturer’s protocol.
Statistical analysis
Data are presented as means ± standard deviation (s.d.) for each group. Statistical
analysis of the in vivo data was performed by a one-way ANOVA. When a significant F
ratio was identified, groups were compared using Fisher’s PLSD post hoc test. P < 0.05
was considered to be significant.
16
X-ray crystallography
The rat AR LBD cDNA, from amino acids 646-901 (rat numbering, human numbering
664-919) was cloned as previously described (41) and included an N-terminal
polyhistidine tag and a thrombin cleavage site. Since the rat and human AR LBD are
identical at the protein level the human numbering scheme was used for discussion and
structural analysis. The rat AR LBD–BMS-564949 complex was crystallized at 20 °C by
vapor diffusion in the hanging drop mode using a 3 mg/mL concentration of AR LBDBMS-
564929 complex as described previously (41). The crystals have the symmetry of
the space group P212121 with unit cell dimensions a = 55.6 Å, b = 65.6 Å, c = 70.8 Å.
Data to 3.0 Å resolution were collected on a Rigaku R-AXIS IV++ Imaging Plate system
mounted on a Rigaku RU-300 rotating-anode X-ray generator (Rigaku-MSC, Woodlands,
TX). The data were processed and scaled using the HKL package (HKL Research, Inc.
Charlottesville, VA). The crystals are iso-structural to the AR LBD-DHT structures
previously determined (41, 42). The initial model was subjected to rigid-body refinement
followed by positional and simulated-annealing refinements using CNX (Accelrys,
Inc.,San Diego, CA). At the end of refinement, the structure has an R factor of 24.9%
with a total of 2032 atoms (2003 protein atoms, 21 ligand atoms and 8 solvent atoms).
Results
AR binding affinity of BMS-564929
17
Competition binding assays with the appropriate radioligands were used to determine the
binding affinity of BMS-564929 for AR as well as for progesterone receptor (PR),
estrogen receptors α and β (ERα, β), the glucocorticoid receptor (GR) and the
mineralocorticoid receptor (MR). BMS-564929 is a high affinity ligand for AR with a Ki
of 2.11 ± 0.16 nM (Table 1). This compound is >1000-fold selective for AR versus ERα
and β, GR, and MR, and approximately 400-fold selective vs. PR (data not shown).
Transcriptional activation of AR in rodent skeletal muscle and prostate epithelial stable
cell lines by BMS-564929
The potency, efficacy and in vitro selectivity of BMS-564929 were measured in
functional transactivation assays in muscle and prostate cell backgrounds in order to
assess activity in cellular contexts where appropriate levels of endogenous cofactors for
each tissue type are expressed. Varying concentrations of T and BMS-564929 were
tested using the mouse myoblast (C2C12) and rat prostate epithelial (PEC) cell lines,
each stably transfected with rat AR (rAR) and an androgen–specific response element
(ARE)-driven luciferase reporter (Fig. 3) (40). BMS-564929 exhibited a potency (EC50,
calculated as the concentration at which 50% of the maximum stimulatory effect of DHT
is achieved) of 0.44 ± 0.03 nM compared with 2.81 ± 0.48 nM measured for T in the
C2C12 myoblast cell line (Table 1). In the PEC cell line, the EC50 for BMS-564929 was
8.66 ± 0.22 nM as compared to 2.17 ± 0.49 nM for T. Both compounds achieved
equivalent maximal stimulatory efficacy in each cell line. These data suggest that BMS-
564929 is approximately 20-fold more potent at induction of luciferase reporter gene
18
expression in muscle cells versus prostate cells in vitro. This is in contrast to that for T,
which showed similar potency in both cell backgrounds, suggesting that endogenous cofactors
complementing the trancriptional machinery in each cell line are differentially
recruited by the two ligands. BMS-564929 shows no measurable activity in functional
transactivation assays with ERα/β, GR, MR or PR at concentrations up to 30 µM (data
not shown).
BMS-564929 does not bind to sex hormone binding globulin or inhibit CYP19 aromatase
In addition to selective binding to intracellular hormone receptors, androgens and other
sex steroid hormones can interact with SHBG to trigger a signal transduction pathway via
a membrane receptor for SHBG (RSHBG), using cyclic adenosine monophosphate (cAMP)
as a second messenger (43, 44). This mechanism is believed to be partly responsible for
non-AR-mediated effects of T on gene activation in target cells. In order to evaluate
BMS-564929 binding to the SHBG we used increasing concentrations of BMS-564929 to
compete [3H]-DHT from purified SHBG. In addition to its androgenic and anabolic
activities, T undergoes metabolic conversion to estradiol (E2) by CYP19 (aromatase),
and this process plays an important role in the regulation of gonadotropin feedback,
several brain functions, bone remodelling and lipid metabolism. We measured the ability
of BMS-564929 and T to inhibit aromatase to assess the potential for interference with
E2 production. The results (Table 1) show that BMS-564929 does not interact
appreciably with SHBG, and does not significantly inhibit aromatase activity at
19
concentrations up to 30 µM, indicating that BMS-564929 is unlikely to exhibit or
interfere with these T-mediated physiological effects in humans.
Effects of BMS-564929 on tissue growth and serum luteinizing hormone concentrations
after two weeks of treatment in mature castrated rats
We evaluated the effects of BMS-564929 in sexually mature, castrated male rats, a wellcharacterized
animal model for studying the anabolic or androgenic effects of AR
modulators on skeletal muscle and sex accessory tissues (45, 46). Castration of mature
male rats results in the rapid involution and atrophy of both levator ani muscle and
prostate, reaching a steady state approximately ten days following castration. Increases
in prostate and levator ani wet weights after drug treatment serve as indicators of
androgenic and anabolic activities, respectively. We used T propionate (TP) as the
positive control due to its superior pharmacokinetic properties and enhanced efficacy
both in humans and in preclinical species. In the recovery mode, we dosed rats for
fourteen days beginning fourteen days after surgical castration. Wet weights of prostate
and levator ani (normalized by body weight) remained unchanged during the two week
treatment period for both sham-operated rats and castrated controls treated with vehicle,
relative to pre-treatment values. TP treatment (s.c.) increased both prostate and levator
ani weights in a dose-dependent manner beginning with the 0.03 mg/kg dose and
reaching normal weight (100% of sham-operated vehicle controls) at 1 mg/kg for the
levator ani and 3 mg/kg for the prostate respectively (Fig. 4A). The potency of TP was
calculated as an ED50 value, (defined as the dose at which tissue wet weight reached 50%
20
of the weight of that of sham operated vehicle control animals). TP exhibited an ED50 of
0.21 mg/kg in the levator ani muscle and an ED50 of 0.42 mg/kg in the prostate,
exhibiting 2-fold selectivity for the levator ani vs. prostate. In the same model BMS-
564929 (p.o.) showed substantially more potent activity in the levator ani, exhibiting an
ED50 of 0.0009 mg/kg in the levator ani and an ED50 of 0.14 mg/kg in the prostate; a net
160-fold selectivity for muscle vs. prostate (Fig. 4B). Approximately 100% muscle
stimulation was achieved at 0.1 mg/kg, reaching greater than 125% stimulation at 0.3 and
1 mg/kg. Compared with TP in the same model, BMS-564929 is more than 200 times
more potent in stimulation of muscle and 80 times more selective for muscle vs. prostate.
In man and rodents, luteinizing hormone (LH) directly stimulates the production of T
in the Leydig cells of the testes. In response to this stimulation, increased levels of T
inhibit further production and secretion of LH through feedback inhibition of the
hypothalmic-pituitary axis, leading to the suppression of T production in normal healthy
individuals. In elderly men treated for age-related sarcopenia, complete inhibition of LH
production might possibly lead to a near total depletion of endogenous T, which could
subsequently result in E2 deficiency and a consequent hormonal imbalance. To evaluate
the potential for BMS-564929 to suppress LH secretion, we measured serum LH levels at
the termination of the castrated rat assay. BMS-564929 suppressed secretion of LH with
an ED50 of 0.008 mg/kg and showed 9-fold selectivity for levator ani muscle stimulation
vs. LH suppression (Fig. 4C). Compared to TP, which has an ED50 of 0.26 mg/kg for LH
suppression and essentially no selectivity (1.2-fold) for levator ani stimulation vs. LH
suppression, BMS-564929 is approximately 33 times more potent in LH suppression, but
9-fold more selective for levator ani.
21
The selectivity of BMS-564929 on tissue growth is maintained after eight weeks of
treatment in mature castrated rats
To determine the effects of longer–term dosing with BMS-564929 on androgen sensitive
tissues in castrated rats, the compound was administered across a wide dose range (1 µg –
1 mg/kg) for eight weeks in the same recovery schedule of administration as the two–
week experiments. A dose-dependent increase in levator ani muscle and prostate wet
weight was observed with BMS-564929 treatment, with ED50 values of 0.001 mg/kg and
0.09 mg/kg for the levator ani and prostate, respectively (Fig. 5). In these same
experiments, TP exhibited somewhat less potent and minimally selective activity in each
tissue (ED50 of 0.19 mg/kg and 0.39 mg/kg for the levator ani and prostate, respectively).
The results for both compounds are in close agreement with those observed in the twoweek
study, indicating a lack of temporal effect on the relative and absolute potency and
selectivity of BMS-564929.
In addition to the previously obtained endpoints, whole body adiposity was measured
by dual-energy x-ray absorptiometry (DEXA) scan in this experiment. Interestingly, we
observed similar dose-dependent decreases in adiposity in both the BMS-564929 and TP
treatment groups (4% decrease in fat mass at 1 mg/kg BMS-564929 or 3 mg/kg TP, data
not shown). The significant decrease in adiposity suggests the potential for clinical
improvements in lean body mass in humans in addition to strictly anabolic effects.
22
Structure of BMS-564929 in complex with the ligand binding domain of Androgen
Receptor
The AR-LBD structure is analogous to our previously published structures (41, 42), and
consists of the nuclear receptor ligand binding domain which has been shown to have
three layers of orthogonally packed α-helices and four β-strands arranged in two β-
hairpin motifs. For the current structure, AR-LBD was initially refined in the absence of
ligand. The difference electron density clearly showed the position of BMS-564929 so
the ligand was then fit to the difference electron density and the protein-ligand complex
was subjected to two additional cycles of CNX refinement. The structure was analyzed
following refinement. The His-Tag with the first seven residues of the amino-terminus
and the carboxy-terminal residue of AR-LBD were not visible in the electron density and
were excluded from the final model. No solvent atoms were found in or near the ligand
binding site. The ligand binding site is well defined, and residues L704, N705, M745,
M749, R752, F764, M780, T877 have direct interactions with BMS-564929. There are
two possible hydrogen bonds between AR-LDB and BMS-564929 involving residues
N705 and R752 (these interactions are described in detail below). As seen previously for
DHT (41), the majority of the interactions between ligand and protein involve van der
Waals interactions. In our previous work it was seen that for DHT hydrophobic residues
were shown to interact with the steroid nucleus, and for BMS-564929 the same residues
contact the phenyl and bi-cyclic rings while hydrogen bonding interactions anchor each
end of the ligand providing affinity and selectivity.
Discussion
23
The results described herein demonstrate that BMS-564929 is a highly potent and
selective AR agonist in vitro and in vivo, and it exhibits unprecedented muscle selectivity
in an established rodent tissue growth model. In the broadest context, certain
components of pharmacological selectivity are inherent to non-steroidal SARM agents
such as BMS-564929, since compounds of this class are not reduced by 5α-reductase or
aromatized by aromatase to other hormonally active agents which act to modulate other
pathways. The lack of potential nongenomic effects associated with SHBG binding adds
a further means of selectivity to BMS-564929 relative to T and DHT.
The 20-fold selectivity observed in relevant cell contexts in vitro correlates with the
tissue selectivity observed in vivo under varied dosing regimens. This strongly suggests
that selectivity has a fundamental molecular basis whereby differential utilization of
endogenous coregulators in each respective cell or tissue type affect the induction of
genes responsible for each tissue’s growth and differentiation. In a similar manner,
various diverse ER ligands differentially induced unique receptor conformations and
subsequently recruited specific coactivators, ultimately leading to tissue selective action
(47, 48). Ongoing studies target identification and characterization of the specific ARregulated
genes involved in the growth of each respective tissue, facilitated by the
availability of selective tool molecules. We are also conducting further studies to
evaluate the potential for additional benefits on bone, body composition, cognition, and
sexual function. It is particularly noteworthy that in both the two–week and eight–week
in vivo experiments in castrated rats, BMS-564929 is able to achieve significant and
sustained hyperanabolic activity on skeletal muscle at doses which incompletely restore
24
prostate size to that of normal animals. Should this level of separation of effects extend
to the human clinical setting, it would potentially allow more aggressive therapy for
abbrogation of functional decline than can be presently pursued with T and its analogues.
Insight into the molecular basis for tissue selectivity may be gleaned from
comparative x-ray crystallographic analysis of various agonists bound to the ligand
binding domain (LBD) of the AR (49), as the molecular basis for differential co-factor
recruitment is a ligand-mediated process. Comparison of the specific residues in the AR
LBD involved in binding to DHT (41) versus those interacting with BMS-564929 (Fig.
6B) suggest intriguing opportunities for further mechanistic study. Since the x-ray
crystal structures of T and DHT with the AR LBD are essentially identical, we chose that
with DHT for purposes of discussion, as this is the active androgen in prostate due to
local conversion of T by 5α-reductase. Figure 6A shows the alignment of AR-LBD from
the DHT (blue) and BMS-654929 (red) complexes. Structural alignment shows that the
two proteins, including amino acid side-chains, are very similar with a RMSD (root mean
square deviation) of 0.58 Å (backbone) and 0.97 Å (all heavy atoms). Both DHT and
BMS-564929 engage R752 via the 3-keto group of DHT (2.89 Å) or the CN group of
BMS-564929 (3.22 Å). A significant difference for these respective ligands is the
interaction of the side chain of F764 with the phenyl ring of BMS-564929 via a π-edge to
face interaction (5.20 Å, ring centroids), which is only weakly captured by DHT through
its single A-ring π-bond. This interaction for BMS-564929 is likely to provide up to 5
kcal/mol binding energy, and may result in steric compression of the LBD in a manner
distinct from that induced by DHT. Additionally, the 17β-hydroxyl group of DHT is
within hydrogen bonding distance to both N705 (2.80 Å) and T877 (2.70 Å), whereas the
25
hydroxyl group of BMS-564929 engages only N705 (2.73 Å). As T877 has been shown
to be a critical residue for recognition of agonism / antagonism of AR ligands in the
common A877T AR (50) mutant seen in prostate cancer patients, these findings may
have significant implications regarding differential activation of AR in prostate. It should
also be noted that this particular point mutation allows antagonists to co-crystallize with
the AR, which to date has not been possible with the wild-type sequence (42). These key
contact differences for the two AR ligands are located in specific helices which
cooperatively form a hydrophobic groove critical for recognition of the LxxLL motif of
coactivator proteins (51), and ligand binding alters the positioning of these helices,
thereby affecting coactivator binding. As x-ray structural analysis provides only a static
picture of ligand-receptor interactions, additional studies are ongoing in these laboratories
using dynamic methods to investigate AR helical mobility as it relates to these unique
interactions identified through crystallography (52). Several AR-regulated genes which
are differentially induced in specific tissues by the present SARM’s relative to the natural
ligands have been identified (unpublished results). We are also investigating the
structure-function relationship of additional ligands structurally related to BMS-564929
which induce different phenotypes in rodents to more specifically characterize the
importance of selected interactions as it relates to tissue selectivity.
Based on the present results, as well as extensive preclinical safety testing, BMS-
564929 has advanced to clinical trials. We expect that this compound will be suitable as
a once daily oral drug for the treatment of age-related musculoskeletal decline in men
with an improved side-effect profile relative to T (e.g. prostate growth, hematopoiesis,
liver toxicity). The marked muscle selectivity of BMS-564929 has the potential to
26
provide a greater safety window for the beneficial effects of androgens while reducing
the likelihood of deleterious side effects observed with T treatment, most notably the
potential for hyperstimulatory effects on the prostate, thereby providing a valuable means
for addressing a growing unmet medical need.
Acknowledgements
We gratefully acknowledge Celia Darienzo and Lifei Wang, and the Department of
Pharmaceutical Candidate Optimization at BMS for extensive characterization of BMS-
564929. We thank Dr. Michael Blanar and Dr. Robert Zahler for their helpful comments.
27
References
1. Morley JE, Kaiser, FE, Perry HM 3rd, Patrick P, Morley PM, Stauber PM,
Vellas B, Baumgartner RN, Garry PJ 1997 Longitudinal changes in testosterone,
luteinizing hormone, and follicle-stimulating hormone in healthy older men.
Metabolism 46:410-413
2. Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR 2001 Longitudinal
effects of aging on serum total and free testosterone levels in healthy men. J Clin
Endocrinol Metab 86:724-731
3. Feldman HA, Longcope C, Derby CA, Johannes CB, Araujo AB, Coviello AD,
Bremner WJ, McKinlay JB 2002 Age trends in the level of serum testosterone and
other hormones in middle-aged men: Longitudinal results from the Massachusetts
male aging study. J Clin Endocrinol Metab 87:589-598
4. Dunn JF, Nisula BC, Rodbard D 1981 Transport of steroid hormones – binding of
21 endogenous steroids to both testosterone-binding globulin and corticosteroidbinding
globulin in human plasma. J Clin Endocrinol Metab 53:58-68
5. Matsumoto AM 2002 Andropause: Clinical implications of the decline in serum
testosterone levels with aging in men. J Gerontol 57A:M76-M99
6. Kaufman JM, Vermeulen A 2005 The decline of androgen levels in elderly men
and its clinical and therapeutic implications. Endocrine Rev 26:833-876
7. Tan RS, Pu SJ, Culberson JW 2003 Role of androgens in mild cognitive
impairment and possible interventions during andropause. Med Hypoth 60:448-452
28
8. Gillett MJ, Martins RN, Clarnette RM, Chubb SA, Bruce DG, Yeap BB 2003
Relationship between testosterone, sex hormone binding globulin and plasma amyloid
beta peptide 40 in older men with subjective memory loss or dementia. J. Alzheimers
Dis 5:267-269
9. Bates KA, Harvey AR, Carruthers M, Martins RN 2005 Androgens, andropause
and neurodegeneration: exploring the link between steroidogenesis, androgens and
Alzheimer’s disease. Cell Mol Life Sci 62:281-292
10. Snyder PJ, Peachey H, Hannoush P, Berlin JA, Loh L, Lenrow DA, Holmes JH,
Dlewati A, Santanna J, Rosen CJ, Strom BL 1999 Effect of testosterone treatment
on body composition and muscle strength in men over 65 years of age. J Clin
Endocrinol Metab 84:2647-2653
11. Ferrando AA, Sheffield-Moore M, Yeckel CW, Gilkison C, Jiang J, Achacosa A,
Lieberman SA, Tipton K, Wolfe RR, Urban RJ 2002 Testosterone administration
to older men improves muscle function: Molecular and physiological mechanisms.
Am J Physiol Endocrinol Metab 282:E601-E607
12. Blackman MR, Sorkin JD, Münzer T, Bellantoni MF, Busby-Whitehead J,
Stevens TE, Jayme J, O’Connor KG, Christmas C, Tobin JD, Stewart KJ,
Cottrell E, St. Clair C, Pabst KM, Harman SM 2002 Growth hormone and sex
steroid administration in healthy aged women and men. A randomized controlled
trial. J Am Med Assoc 288:2282-2292
13. Schroeder ET, Singh A, Bhasin S, Storer TW, Azen C, Davidson T, Martinez C,
Sinha-Hikim I, Jaque SV, Terk M, Sattler FR 2003 Effects of an oral androgen on
29
muscle and metabolism in older, community-dwelling men. Am J Physiol Endocrinol
Metab 284:E120-E128
14. Bhasin S, Woodhouse L, Casaburi R, Singh AB, Mac, RP, Lee M, Yarasheski
KE, Sinha-Hikim I, Dzekov C, Dzekov J, Magliano L, Storer TW 2005 Older
men are as responsive as young men to the anabolic effects of graded doses of
testosterone on skeletal muscle. J Clin Endocrinol Metab 90:678-688
15. Page S, Amory JK, Bowman FD, Anawalt BD, Matsumoto AM, Bremner WJ
Tenover JL 2005 Exogenous testosterone (T) alone or in combination with
finasteride increases physical performance, grip strength, and lean body mass in older
men with low serum T. J Clin Endocrinol Metab 90:1502-1510
16. Basaria S, Dobs AS 2001 Hypogonadism and androgen replacement therapy in
elderly men. Am J Med 110:563-572
17. Gruenwald DA, Matsumoto AM 2003 Testosterone supplementation therapy for
older men: Potential benefits and risks. J Am Geriat Soc 51:101-115
18. Rhoden EL, Morgentaler A 2004 Risks of testosterone-replacement therapy and
recommendations for monitoring. N Engl J Med 350:482-492
19. Liu PY, Swerdloff RS, Veldhuis JD 2004 The rationale, efficacy and safety of
androgen therapy in older men: Future research and current practice
recommendations. J Clin Endocrinol Metab 89:4789-4796
20. Hijazi RA, Cunningham GR 2005 Andropause: Is androgen replacement therapy
indicated for the aging male? Ann Rev Med 56:117-137
21. Ly LP, Jimenez M, Zhuang TN, Celermajer DS, Conway AJ, Handelsman DJ
2001 A double-blind, placebo-controlled, randomized clinical trial of transdermal
30
dihydrotestosterone gel on muscular strength, mobility, and quality of life in older
men with partial androgen deficiency. J Clin Endocrinol Metab 86:4078-4088
22. Wang C, Swedloff RS, Iranmanesh A, Dobs A, Snyder PJ, Cunningham G,
Matsumoto AM, Weber T, Berman N 2000 Transdermal testosterone gel improves
sexual function, mood, muscle strength, and body composition parameters in
hypogonadal men. Testosterone Gel Study Group. J Clin Endocrinol Metab 85:2839-
2853
23. Anderson RA, Wallace AM, Sattar M, Kumar N, Sundaram K 2003 Evidence
for tissue selectivity of the synthetic androgen 7-alpha-methyl-19-nortestosterone in
hypogonadal men. J Clin Endocrinol Metab 88:2784-2793
24. Schroeder ET, Zheng L, Yarasheski KE, Qian D, Stewart Y, Flores C, Martinez
C, Terk M, Sattler FR 2004 Treatment with oxandrolone and the durability of
effects in older men. J Appl Physiol 96:1055-102
25. Gelfand MM, Wiita B 1997 Androgen and estrogen-androgen hormone replacement
therapy: A review of the safety literature, 1941-1996. Clin Therapeut 19:383-404
26. Negro-Vilar A 1999 Selective androgen receptor modulators (SARMs): A novel
approach to androgen therapy for the new millennium. J Clin Endocrinol Metab
84:3459-3462
27. Hamann LG, Mani NS, Davis RL, Wang XN, Marschke KB, Jones TK 1999
Discovery of a potent orally active nonsteroidal androgen receptor agonist: 4-ethyl-
1,2,3,4-tetrahydro-6-(trifluoromethyl)-8-pyridono-[5,6-g]-quinoline (LG121071). J
Med Chem 42:210-212
31
28. Zhi L, Martinborough E 2001 Selective androgen receptor modulators (SARMs).
Ann Rep Med Chem 36:169-180
29. Chengalvla M, Oh T, Roy AK 2003 Selective androgen receptor modulators.
Expert Opin Ther Patents 13:59-66
30. Gao W, Bohl CE, Dalton JT 2005 Chemistry and structural biology of androgen
receptor. Chem Rev 105:3352-3370
31. Chen J, Hwang JD, Bohl CE, Miller DD, Dalton JT 2005 A selective androgen
receptor modulator for hormonal male contraception. J Pharmacol Exp Therap
312:546-553
32. Shang Y, Myers M, Brown M 2002 Formation of the androgen receptor
transcription complex. Mol Cell 9:601-610
33. Katzenellenbogen BS, Choi I, Delage-Mourroux R, Ediger TR, Martini PG,
Montano M, Sun J, Weis K, Katzenellenbogen JA 2000 Molecular mechanisms of
estrogen action: selective ligands and receptor pharmacology. J Steroid Biochem Mol
Biol 74:279-285
34. Dutertre M, Smith CL 2000 Molecular mechanisms of selective estrogen receptor
modulator (SERM) action. J Pharmacol Exp Ther 295:431-437
35. McDonnell DP, Connor CE, Wijayaratne A, Chang CY, Norris JD 2002
Definition of the molecular and cellular mechanisms underlying the tissue-selective
agonist/antagonist activities of selective estrogen receptor modulators. Recent Prog
Horm Res 57:295-316
36. Shang Y, Brown M 2002 Molecular determinants for the tissue specificity of
SERMs. Science 295:2465-2468
32
37. Smith CL, O’Malley BW 2004 Coregulator function: A key to understanding tissue
specificity of selective receptor modulators. Endocrine Rev 25:45-71
38. Heinlein, CA, Chang C 2002 Androgen receptor (AR) coregulators: An overview.
Endocrine Rev 23:175-200
39. Kazmin D, Prytkova T, Cook CE, Wolfinger R, Chu TM, Beratan D, Norris JD,
Chang CY, McDonnell DP 2006 Linking ligand induced alterations in androgen
receptor structure to differential gene expression; a first step in the rational design of
selective androgen receptor modulators (SARMs). Mol Endo 20:1201-1217
40. Zhou Z, Corden JL, Brown TR 1997 Identification and characterization of a novel
androgen response element composed of direct repeat. J Biol Chem 272:8227-8235
41. Sack JS, Kish KF, Wang C, Attar RM, Kiefer SE, An Y, Wu GY, Scheffler JE,
Salvati ME, Krystek SR Jr, Weinmann R, Einspahr HM 2001 Crystallographic
structures of the ligand-binding domains of the androgen receptor and its T877A
mutant complexed with the natural agonist dihydrotestosterone. Proc Natl Acad Sci
USA 98:4904-4909
42. Salvati ME, Balog A, Shan W, Wei DD, Pickering D, Attar RM, Geng J, Rizzo
CA, Gottardis MM, Weinmann R, Krystek SR, Sack J, An Y, Kish K 2005
Structure based approach to the design of bicyclic-1H-isoindole-1,3(2H)-dione based
androgen receptor antagonists. Bioorg Med Chem Lett 15:271-276
43. Rosner W, Hryb DJ, Khan MS, Nakhla AM, Romas NA 1999 Androgen and
estrogen signaling at the cell membrane via G-proteins and cyclic adenosine
monophosphate. Steroids 64:100-106
33
44. Heinlein CA, Chang C 2002 The roles of androgen receptors and androgen-binding
proteins in nongenomic androgen actions. Mol Endocrinol 16:2181-2187
45. Beyler AL, Arnold A, Potts GO 1968 Methods for evaluating anabolic and catabolic
agents in laboratory animals. J Am Med Women’s Assoc 23:708-712
46. Chinoy NJ, Sheth KM, Shah VC 1973 A comparative histophysiological study on
the normal, castrated and testosterone-treated levator ani muscle of rodents. Acta
Endocrinol 74:389-398
47. Kraichely DM, Sun J, Katzenellenbogen JA, Katzenellenbogen BS 2000
Conformational changes and coactivator recruitment by novel ligands for estrogen
receptor-β: Correlations with biological character and distinct differences among
SRC coactivator family members. Endocrinol 141:3534-3545
48. Egner U, Heinrich N, Ruff M, Gangloff M, Mueller-Fahrnow A, Wurtz JM 2001
Different ligands–different receptor conformations: Modeling of the hERα LBD in
complex with agonists and antagonists. Med Res Rev 21:523-539
49. Poujol N, Wurtz JM, Tahiri B, Lumbroso S, Nicolas JC, Moras D, Sultan C
2000 Specific recognition of androgens by their nuclear receptor. J Biol Chem
275:24022-24031
50. Bohl CE, Gao W, Miller DD, Bell CE, Dalton JT 2005 Structural basis for
antagonism and resistance of bicalutamide in prostate cancer. Proc Natl Acad Sci
USA 102:6201-6206
51. He B, Gampe RT Jr, Kole AJ, Hnat AT, Stanley TB, An G, Stewart EL, Kalman
RI, Minges JT, Wilson EM 2004 Structural basis for androgen receptor interdomain
34
and coactivator interactions suggests a transition in nuclear receptor activation
function dominance. Mol Cell 16:425-438
52. Togashi M, Borngraeber S, Sandler B, Fletterick RJ, Webb P, Baxter JD 2005
Conformational adaptation of nuclear receptor ligand binding domains to agonists:
potential for novel approaches to ligand design. J Steroid Biochem Mol Biol 93:127-
137
35
FIGURE LEGENDS
FIG. 1. Structures of endogenous androgen receptor ligands testosterone (T) and 5α-
dihydrotestosterone (DHT), and selective androgen receptor modulator BMS-564929.
FIG. 2. Synthesis route for the preparation of BMS-564929.
FIG. 3. In vitro activation of AR-mediated luciferase reporter gene induction in C2C12
and PEC cells with (A) testosterone and (B) selective androgen receptor modulator BMS-
564929 relative to 5α-dihydrotestosterone.
FIG. 4. Effects of (A) subcutaneous testosterone propionate or (B) oral BMS-564929 on
wet weight of levator ani muscle and prostate, and (C) on suppression of serum
luteinizing hormone in mature castrated male rats after 2 weeks of once daily treatment.
FIG. 5. Effects of (A) subcutaneous testosterone propionate or (B) oral BMS-564929 on
wet weight of levator ani muscle and prostate in mature castrated male rats after 8 weeks
of once daily treatment in recovery mode.
FIG. 6. (A) Protein Structural alignment (Maestro, Schrodinger LLC. Portland, OR) of
AR ligand binding domains from the dihydrotestosterone (DHT) (blue/cyan) and BMS-
564929 (red/yellow) complex. The protein backbone is rendered in ribbon diagram and
the active site residues are displayed in stick rendering. (B) Superimposition of x-ray cocrystal
structures of DHT (green) and BMS-564929 (grey) bound to the androgen
36
receptor ligand binding domain at 2.0 Å and 3.0 Å resolution respectively. Key ligand
binding domain residues are labeled in yellow. DHT makes binding contacts with R-752,
Q-711, and a bifurcated H-bond to N-705 and T-877. BMS-564929 makes contacts with
R-752 and N-705, as well as a π-edge-face interaction with F-764.
TABLE
Table 1. In vitro activities of BMS-564929 (BMS) and testosterone (T)
BMS T
AR binding
Ki (nM)
2.11 ± 0.16
0.25 ± 0.03
C2C12
EC50 (nM)
0.44 ± 0.03
2.81 ± 0.48
PEC
EC50 (nM)
8.66 ± 0.22
2.17 ± 0.49
SHBG
IC50 (nM)
> 30,000
7 ± 1
Aromatase
IC50 (nM)
> 30,000
740 ± 2
N
N
HO H
O
O
CH3 Cl
CN
OH OH
H
H H
O O
BMS-564929
T DHT
FIG. 1
OCN CN
N
N
H O
O Cl
CN
NH
HO
CO2Me
N
HO
CO2Me
HN
O
CN
1. i-Pr2NEt
2. DBU
85%
HO
Cl
Cl
H2N
Cl
1. Ac2O / AcOH, NaOH
2. Br2
3. Zn(CN)2/ Pd(dba)2
4. HCl / EtOH
5. phosgene, NaHCO3
82%
BMS-564929
HCl
NH
HO
CO2H
1. HCl / MeOH
2. BOC2O / NEt3
3. Ph3P / DEAD / PhCO2H
4. KOH / MeOH
5. HCl
80%
FIG. 2
FIG. 3
20
20
0
20
40
60
80
100
120
140
C2C12
PEC
Testosterone
( nM)
Percent Control
-
20
40
60
80
100
120
140
BMS-564929
(nM)
C2C12
PEC
Percent Control
)
-
0
-
10-3 10-2 10-1 1 10 102 103
10-3 10-2 10-1 1 10 102 103
A
B
FIG. 4
Levator ani
Prostate
BMS-564929
Percent Intact Control
-25
0
25
50
75
100
125
150
B
Dose (mg/kg)
Testosterone propionate
0
20
40
60
80
100
120
Dose (mg/kg)
Percent Intact Control
10-3 10-2 10-1 1 10 102 10-4 10-5
Levator ani
Prostate
10-3 10-2 10-1 1 10 102 10-4 10-5
Dose (mg/kg)
-100
-80
-60
-40
-20
0
20
Testosterone
propionate
BMS-564929 Percent Control
C
10-3 10-2 10-1 1 10 10
2
10-4 10-5
A
FIG. 5
-50
0
50
100
150
200
250
Dose (mg/kg)
Percent Intact Control
Levator ani
Prostate
BMS-564929 B
-50
0
50
100
150
200
250
Dose (mg/kg)
Percent Intact Control
Levator ani
Prostate
Testosterone propionate A
10-3 10-2 10-1 1 10 102 10-4 10-5
10-3 10-2 10-1 1 10 102 10-4 10-5
Q-711
R-752
N-705
T-877
F-764
F-876
FIG. 6
Selective Androgen Receptor Modulator: Potent Hyperanabolic
Stimulation of Skeletal Muscle with Hypostimulation of Prostate in Rats
Abbreviated Title: Muscle Selective Androgen Receptor Agonist
Jacek Ostrowski*,1, Joyce E. Kuhns1, John A. Lupisella1, Mark C. Manfredi2, Blake C.
Beehler1, Stanley R. Krystek, Jr.3, Yingzhi Bi2,†, Chongqing Sun2, Ramakrishna
Seethala1, Rajasree Golla1, Paul G. Sleph1, Aberra Fura4, Yongmi An3, Kevin F. Kish3,
John S. Sack3, Kasim A. Mookhtiar1,§, Gary J. Grover1,¶ and Lawrence G. Hamann*,2
Departments of
1
Metabolic Diseases,
2
Discovery Chemistry,
3
Macromolecular Structure,
4
Pharmaceutical Candidate Optimization, Bristol-Myers Squibb Pharmaceutical
Research Institute, P.O. Box 5400, Princeton, NJ 08543, USA.
Key Words: Androgen Receptor, Tissue Selectivity, SARM, Anabolic Agent
*Correspondence should be addressed to Lawrence G. Hamann at: Department of
Discovery Chemistry, Bristol-Myers Squibb, Pharmaceutical Research Institute, 5
Research Parkway, Wallingford CT 06492; telephone: 203-677-6948. fax: 203-677-
7884. e-mail: [email protected]
†Present address: Lexicon Pharmaceuticals, 350 Carter Road, Princeton, NJ 08540.
email: [email protected]
§Present address: Advinus Therapeutics, Pune, India. email:
[email protected]
¶Present address: Product Safety Laboratories, 2394 Route 130, Dayton, NJ 08810.
email: [email protected]
DISCLOSURE STATEMENT: J.O., J.K., J.L., M.M., B.B., S.K., C.S., R.S., R.G., P.S.,
A.F., Y.A., K.K., J.S., and L.H. are employed by Bristol-Myers Squibb. Y.B., K.M., and
G.G. were previously employed by Bristol-Myers Squibb. M.M., A.F., J.S., K.M., G.G.,
and L.H. have equity interests in Bristol-Myers Squibb.
Endocrinology. First published ahead of print September 28, 2006 as doi:10.1210/en.2006-0843
Copyright (C) 2006 by The Endocrine Society
2
Abstract
A novel, highly potent, orally active, non-steroidal tissue selective androgen receptor
modulator (BMS-564929) has been identified, and this compound has been
advanced to clinical trials for the treatment of age-related functional decline. BMS-
564929 is a subnanomolar androgen receptor agonist in vitro, is highly selective for
the androgen receptor versus other steroid hormone receptors, and exhibits no
significant interactions with sex hormone binding globulin or aromatase. Dose
response studies in castrated male rats show that BMS-564929 is substantially more
potent than testosterone in stimulating the growth of the levator ani muscle, and
unlike testosterone, highly selective for muscle versus prostate. Key differences in
the binding interactions of BMS-564929 with the androgen receptor relative to the
native hormones were revealed through x-ray crystallography, including several
unique contacts located in specific helices of the ligand binding domain important
for coregulatory protein recruitment. Results from additional pharmacological
studies effectively exclude alternative mechanistic contributions to the observed
tissue selectivity of this unique, orally active androgen. As concerns regarding the
potential hyperstimulatory effects on prostate and an inconvenient route of
administration are major drawbacks which limit testosterone’s clinical use, the
potent oral activity and tissue selectivity exhibited by BMS-564929 is expected to
yield a clinical profile which provides the demonstrated beneficial effects of
testosterone in muscle and other tissues with a more favorable safety window.
3
Serum androgen (testosterone, T) levels decline progressively with aging in men,
beginning in the third decade of life at the rate of approximately one percent per year (1-
3). Furthermore, ninety-eight percent of T in men is not freely available to most tissues
because it is bound with high affinity to sex hormone binding globulin (SHBG) (4), a
serum protein whose concentration increases with age. This decline in available T is
associated with alterations in body composition, diminished energy, diminished muscle
strength and physical function, reduced sexual function, and depressed mood (5-9). This
androgen deficient state in aging males, often called andropause or androgen decline in
the aging male (ADAM), can subsequently lead to age-related functional decline (frailty).
Several clinical studies have shown that supplementation of T at physiologic doses in
elderly men results in a significant increase in lean body mass, a decrease in adipose
tissue, and an increase in muscle strength (10-15). However, use of T replacement
therapy (TRT) in elderly males by primary care physicians is limited due to concerns
about potential side effects (16-20). Among the more serious side effects are
hyperstimulation of the prostate, which may be a preamble to or exacerbate occult
subclinical benign prostatic hypertrophy (BPH) and/or prostate cancer, and increases in
hematocrit.
T cannot be administered orally due to its rapid and extensive metabolism, and
consequently, is given by injection, transdermal patch or in a recently introduced gel
form (21, 22). Each of these delivery routes has certain drawbacks: the need for needles
and physician administration for the former, skin irritation and potential for contact
transfer to others with the latter. Several T analogues with either 7α- or 17α-alkyl
substitution, including 7α-methyl-19-nortestosterone (MENT) (23) and oxandrolone (24),
4
have been used clinically in place of T, as these are shown to circumvent metabolism and
improve oral bioavailability and half-life. However, 17α-alkylated androgens have
frequently been associated with hepatotoxicity (25), although, these effects are
hypothesized to be structure– and not mechanism–related. There are currently no
approved drugs indicated for the prevention of functional decline in the aging male, but
the growing desire among the aging population to remain independent has raised
awareness within the medical community regarding the potential for therapeutic
intervention. Consequently, there exists an unmet medical need for safe and effective
oral therapies which might offer greater separation between the desired anabolic and
undesired androgenic effects. In theory, a selective androgen receptor modulator
(SARM) (26) has the potential to maintain or improve muscle strength and function,
thereby enabling or facilitating the performance of everyday activities, and to improve
body composition, mental health, libido and sexual function in elderly men without the
concomitant deleterious effects on prostate, liver, or erythrocytes associated with steroid
treatment regimens.
Since the first identification of orally-active non-steroidal androgens (27), several
groups have actively pursued the identification and development of SARMs (28-30), and
some of the most advanced compounds to emerge from these efforts have recently
reached the clinic (31). However, these compounds exhibit limited efficacy and modest
tissue selectivity in preclinical models. As part of our program to discover and develop
novel treatments for age-related functional decline, we have discovered a novel, potent,
orally bioavailable SARM, BMS-564929 (Fig. 1). which, analagous to the natural
hormone agonists T and 5-dihydrotestosterone (DHT), exerts its effects via the regulation
5
of androgen receptor (AR) mediated gene transcription in tissues which express the AR
(32). This highly potent compound exhibits remarkable muscle vs. prostate selectivity in
vitro and in chronic in vivo models of androgen action in rodents. The mechanism by
which BMS-564929 tissue selectivity is achieved is anticipated to be analogous to that of
the more well-characterized selective estrogen receptor modulators (SERMs) (33-36),
which involves ligand-based selective activation of target genes through differential
recruitment of co-factors present in various tissues expressing the receptor (37, 38).
Recent efforts directed at beginning to understand the structural basis for ligand induced
differential gene expression through the AR are making advances toward characterizing
these relationships in greater detail at the molecular level (39). X-ray data of the AR
ligand binding domain (AR-LBD) bound to BMS-564929 and DHT shows that BMS-
564929 has unique binding interactions compared to DHT. Therefore, binding of BMS-
564929 to the AR-LBD results in a receptor conformation disparate to that induced by
DHT, which may thereby allow the BMS-564929-AR complex to interact with a different
set of coactivators/corepressors than the DHT-AR complex. Through additional
investigations, we have been able to rule out contributions from a number of potential
alternative selectivity mechanisms. Based on the present results, we believe that BMS-
564929 is the most potent and muscle-selective, orally-available AR agonist reported to
date. Consequently this compound has advanced into human clinical trials for the
treatment of age-related functional decline.
Materials and Methods
Reagents
[3H]-DHT, [3H]-progesterone and [3H]-estradiol were obtained from Perkin-Elmer.
SARM compound BMS-564929 [(7R,7aS)-2-(3-chloro-4-cyano-2-methylphenyl)-7-
hydroxytetrahydro-2H-pyrrolo[1,2-e]imidazole-1,3-dione] was synthesized in our
laboratories by the following procedures as outlined schematically (Fig. 2).
(3S)-N-tert-Butoxycarbonyl-3-hydroxy-L-proline methyl ester. Hydrogen chloride
gas was bubbled through a suspension of trans-3-hydroxy-L-proline (50 g, 0.38 mol) in
MeOH (600 mL) at 0 °C for 10 min. The resulting clear solution was stirred at room
temperature for 4 h, then concentrated under reduced pressure. The resulting white solid
was dried in vacuo overnight to afford 68.3 g of the title compound. The hydroxyproline
ester thus obtained was suspended in CH2Cl2 (1.0 L) cooled to 0 °C, and Et3N (105.3 mL,
0.755 mol) was added, followed by portionwise addition of di-tert-butyl dicarbonate
(82.96 g, 0.380 mol). The resulting mixture was stirred at room temperature for 4 h, then
partitioned between water and CH2Cl2. The CH2Cl2 layer was washed twice with water,
and once each with 20% aqueous citric acid, water and brine, then dried over Na2SO4 and
concentrated under reduced pressure to give an oily residue. The crude product was
chromatographed (silica gel) eluting with 15%-50% EtOAc/hexane to afford the
diprotected compound (73.3 g) as a pale yellow viscous oil. 1H NMR (CDCl3, 400 MHz,
3:2 mixture of rotamers, data for major rotamer) δ 1.41 (s, 9H), 1.93 (m, 1H), 2.11 (m,
1H), 2.45 (d, J = 4.8 Hz, 1H), 3.60 (m, 2H), 3.75 (s, 3H), 4.18 (d, J = 1.3 Hz, 1H), 4.44
(m, 1H). (3R)-N-tert-Butyloxycarbonyl-3-benzoyloxy-L-proline methyl ester. To a
7
stirred solution of (3S)-N-tert-butoxycarbonyl-3-hydroxy-L-proline methyl ester (69.1 g,
0.282 mol), Ph3P (88.7 g, 0.338 mol) and benzoic acid (41.3 g, 0.338 mol) in anhydrous
THF (1.35 L) cooled to 0 °C was added dropwise over 1 h (through an addition funnel) a
solution of diethylazodicarboxylate (62 mL, 0.33 mol) in anhydrous THF (50 mL). After
addition, the resulting light yellow solution was stirred at room temperature for 8 h. The
reaction mixture was then partitioned between EtOAc and aqueous NaHCO3. The
organic layer was washed with saturated aqueous NaHCO3, water (2×), brine, dried over
Na2SO4 and concentrated under reduced pressure to yield a crude product as a semi-solid.
The crude benzoate product was suspended in 25% EtOAc/hexane and stirred vigorously
for 3 h. The resulting suspension was filtered and the collected white solid (Ph3PO)
rinsed twice with 20% EtOAc/hexane. The combined filtrates were concentrated under
reduced pressure to yield an oily residue, which was triturated twice with 20%
EtOAc/hexane as described above to yield approximately 150 g of partially purified
product as a yellow oil, which was further purified by flash chromatography (silica gel)
eluting with 10-20% EtOAc/hexane to furnish the pure title compound (88.4 g) as a light
yellow viscous oil. 1H NMR (400 MHz, CDCl3, 2:1 mixture of rotamers, data for major
rotamer) δ 1.39 (s, 9H), 2.26 (m, 2H), 3.57-3.75 (m, 2H), 3.62 (s, 3H), 4.65 (d, J = 4.0
Hz), 5.71 (q, J = 4.0 Hz), 7.45 (t, J = 4.0 Hz, 2H), 7.58 (t, J = 4.0 Hz, 1H), 7.98 (d, J =
4.0 Hz, 2H). (3R)-3-Hydroxy-L-proline methyl ester. To a solution of (3R)-N-tertbutyloxycarbonyl-
3-benzoyloxy-L-proline methyl ester (88.4 g, 0.253 mol) in anhydrous
MeOH (700 mL) at 0 °C was slowly added through an addition funnel a freshly prepared
1N solution of KOH in anhydrous MeOH (367 mL, 0.367 mol) over 25 min. After the
addition, the resultant light yellow solution was stirred at 0 °C for 2 h, then the reaction
8
quenched by slow addition (over 25 min) of a solution of 1N HCl in dioxane/EtOAc (380
mL) through an addition funnel. The resulting white suspension was concentrated under
reduced pressure to remove most of the solvent, and the remaining mixture was
partitioned between water and EtOAc. The separated organic phase was washed with
water (2×), saturated aqueous NaHCO3 (2×), water, brine, and dried over Na2SO4. The
filtrate was concentrated under reduced pressure to give a light yellow oily residue, which
was chromatographed (silica gel) eluting first with 25-30% EtOAc/hexane, then 5%
MeOH in 30% EtOAc/hexane to furnish the N-Boc-protected cis-hydroxy-L-proline
methyl ester (44.6 g) as a pale yellow oil. To a solution of the free hydroxy compound
thus obtained (44.6 g, 0.182 mol) in CH2Cl2 (450 mL) at 0 °C was slowly added TFA
(275 mL) through an addition funnel over 40 min. After the addition, the reaction
mixture was stirred at 0 °C for 2 h, then concentrated under reduced pressure to give a
viscous oily residue which was evaporated with ether (2×), toluene (1×), ether (2×) and
dried in vacuo overnight to yield the Boc-deprotected TFA salt of cis-hydroxy-L-proline
methyl ester (59.5 g) as a light yellow solid. A solution of this TFA salt (12.64 g, 48.8
mmol) in MeOH (150 mL) was free based by treatment with WA21J resin (Diaion®
WA21J, polyamine resin from Supelco, Bellefonte, PA, 60 g). The resulting suspension
was stirred at room temperature for 1 h, and then filtered. The collected resin was rinsed
with MeOH (2×) and combined filtrate concentrated carefully under reduced pressure to
give the desired cis-hydroxy-L-proline methyl ester (7.7 g) as a colorless oil. [α]D =
+14.9° (c. 1.0, MeOH); 1H NMR (400 MHz, CD3OD) δ 1.86 (m, 1H), 2.02 (m, 1H), 2.87
(m, 1H), 3.24 (m, 1H), 3.69 (d, J = 4.0 Hz, 1H), 3.75 (s, 3H), 4.48 (t, J = 4.0 Hz, 1H);
HPLC: 100% at 0.157 min (retention time) (Conditions: Phenominex Luna C18 (4.6 x 50
9
mm); eluted with 0% to 100% B; 4 min gradient (A = 90% H2O/10% MeOH/0.1%
H3PO4 and B = 10% H2O/90% MeOH/0.1% H3PO4 ), flow rate at 4 mL/min., UV
detection at 220 nm); MS (ES): m/z 146 [M+H]+
N-(4-Bromo-3-chloro-2-methylphenyl)acetamide. To a solution of commercially
obtained 3-chloro-2-methylaniline (10.0 g, 70.6 mmol) in EtOH (85 mL) at room
temperature was slowly added acetic anhydride (8.00 mL, 84.7 mmol) and the reaction
mixture was stirred for 15 min. The resulting suspension was concentrated and dried in
vacuo to provide the desired acetamide (17.0 g) as a red-brown solid. To a solution of
this acetamide (13.0 g, 70.6 mmol) in AcOH (100 mL) at 15 °C was added bromine (10.9
mL, 212 mmol) over 20 min. The solution was allowed to warm to room temperature
and stir for 1 h. The solution was then poured into ice water while stirring, and the
precipitate which formed was filtered, washed with water until the filtrate was at neutral
pH, and dried to provide the desired bromide (17.6 g) as a pale orange solid. 1H NMR
(400 MHz, DMSO-d6) δ 2.05 (s, 3H), 2.28 (s, 3H), 7.29 (d, J = 8.25 Hz, 1H), 7.56 (d, J =
8.80 Hz, 1H), 9.60 (s, 1H); 13C NMR (100 MHz, DMSO-d6) δ 16.71, 23.14, 118.05,
125.48, 130.36, 132.68, 133.39, 137.14, 168.44; HPLC a) column: Phenominex ODS
C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90% H2O/0.1% TFA to 90% MeOH/10%
H2O/0.1% TFA; 1 min hold, 4 mL/min UV detection at 220 nm, 2.95 min retention time;
HPLC b) column: Shimadzu Shim-Pack VP-ODS C18 4.6 x 50 mm, 4 min gradient, 10%
MeOH/90% H2O/0.1% TFA to 90% MeOH/10% H2O/0.1% TFA, 1 min hold; 4 mL/min,
UV detection at 220 nm, 2.87 min retention time (98%); MS (ES) m/z 263 [M+H]+. 4-
Amino-2-chloro-3-methylbenzonitrile. A suspension of the bromide thus obtained
(17.5 g, 66.7 mmol) and copper cyanide (7.16 g, 80.0 mmol) in DMF (200 mL) was
10
heated to 150 °C for 5 h. The solution was then cooled and poured into ice water while
stirring. The precipitate which formed was filtered, washed with water and dried. The
solid was triturated in refluxing methanol, filtered and the filtrate concentrated to provide
the nitrile (9.94 g) as a brown solid. A suspension of the nitrile thus obtained (9.90 g,
47.4 mmol) in conc. HCl (50 mL) and EtOH (50 mL) was heated at reflux for 30 min.
The solution was concentrated and dried to provide the hydrochloride salt of the desired
aniline (9.41 g) as a brown solid. 1H NMR (400 MHz, DMSO-d6) δ 2.12 (s, 3H), 6.30 (s,
2H), 6.61 (d, J = 8.25 Hz, 1H), 7.36 (d, J = 8.25 Hz, 1H); 13C NMR (100 MHz, DMSOd6)
δ 13.83, 96.86, 112.11, 118.31, 118.85, 132.16, 135.56, 152.52; HPLC a) column:
Phenominex ODS C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90% H2O/0.1% TFA to
90% MeOH/10% H2O/0.1% TFA; 1 min hold, 4 mL/min UV detection at 220 nm, 2.42
min retention time; HPLC b) column: Shimadzu Shim-Pack VP-ODS C18 4.6 x 50 mm,
4 min gradient, 10% MeOH/90% H2O/0.1% TFA to 90% MeOH/10% H2O/0.1% TFA, 1
min hold; 4 mL/min, UV detection at 220 nm, 2.31 min retention time (99%); MS (ES):
m/z 167 [M+H]+. 2-Chloro-4-isocyanato-3-methylbenzonitrile. To a mixture of the
hydrochloride salt of 4-amino-2-chloro-3-methylbenzonitrile (5.78 g, 28.4 mmol) in
CH2Cl2 (230 mL) at 0 °C was added NaHCO3 (23.9 g, 284 mmol) followed by 20%
phosgene in toluene (60.2 mL, 114 mmol) over 30 min. The mixture was allowed to
warm to room temperature and stir for 1.5 h. The mixture was filtered and the filtrate
concentrated. The residue was azeotroped with toluene (3 × 100 mL) and concentrated to
provide the isocyanate (5.45 g) as a pale, orange solid which was used immediately in the
next step without further purification.
11
(7R,7aS)-2-Chloro-4-(7-hydroxy-1,3-dioxotetrahydropyrrolo[1,2-c]imidazol-2-yl)-3-
methylbenzonitrile (BMS-564929). A suspension of 2-chloro-4-isocyanato-3-
methylbenzonitrile (5.45 g, 28.3 mmol), diisopropylethylamine (5.95 mL, 34.1 mmol), 4
Å molecular sieves (5 g) and (3R)-3-hydroxy-L-proline methyl ester (7.37 g, 28.4 mmol)
in CH2Cl2 (300 mL) was stirred at room temperature for 1.5 h. Diazabicycloundecane
(5.10 mL) was then added, and after 18 h at room temperature, the suspension was
filtered and the solid triturated with acetone, refiltered and the combined filtrates were
concentrated in vacuo. The residue was chromatographed on silica gel (CH2Cl2/CH3OH;
49:1 to 9:1 gradient) to provide BMS-564929 (4.08 g) as a white solid. 1H NMR (400
MHz, DMSO-d6) δ 2.05-2.11 (m, 1H), 2.15-2.22 (m, 1H), 2.20, 2.24 (s, 3H), 3.29-3.33
(m, 1H), 3.59-3.68 (m, 1H), 4.42-4.50 (m, 2H), 5.64, 5.72 (d, J = 3.85, 3.30 Hz, 1H),
7.22, 7.51 (d, J = 8.25 Hz, 1H), 7.96 (d, J = 8.25, 1H); 13C NMR (100 MHz, DMSO-d6) δ
15.44, 15.63, 35.49, 35.62, 43.30, 43.41, 68.76, 69.25, 69.84, 112.86, 113.08, 115.79,
128.11, 128.73, 132.07, 136.27, 136.42, 136.85, 137.12, 158.63, 169.09, 169.60; HPLC
a) column: Phenominex ODS C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90%
H2O/0.1% TFA to 90% MeOH/10% H2O/0.1% TFA; 1 min hold, 4 mL/min UV detection
at 220 nm, 2.07, 2.32 min retention time; HPLC b) column: Shimadzu Shim-Pack VPODS
C18 4.6 x 50 mm, 4 min gradient, 10% MeOH/90% H2O/0.1% TFA to 90%
MeOH/10% H2O/0.1% TFA, 1 min hold; 4 mL/min, UV detection at 220 nm, 1.93, 2.23
min retention time (97%); HPLC c) column: Daicel Chiralcel OD 4.6 x 250 mm, isocratic
25% isopropanol/hexanes, 30 min, 1 mL/min, UV detection at 220 nm, 10.99 min
retention time (98%); MS (ES): m/z 306 [M+H]+.
12
Receptor and binding protein binding assays
The human cancer epithelial breast cell lines MDA MB-453 and T47D, which
endogenously express AR and PR, respectively, were used for radioligand competition
binding assays. Binding assays were conducted by incubating BMS-564929 at various
concentrations with either [3H]-DHT or [3H]-progesterone with the cells for 2 hours at
room temperature. For ERα and ERβ, fusion proteins expressed in Escherichia coli,
consisting of maltose binding protein (MBP), a specific biotinylation sequence (BioP), an
enterokinase cleavage site (EK) and either the ERα or ERβ LBD was used. Binding
reactions were conducted by incubating ERα and ERβ LBD with BMS-564929 and [3H]-
estradiol for 2 hours at room temperature. Specific binding activity to the
mineralocorticoid receptor (MR) by BMS-564929 was evaluated by competition binding
assay using kidney cytosolic preparations and [3H]-aldosterone. The kidneys were
obtained from adrenalectomized (ADX) rats in order to remove the endogenous source of
aldosterone and to increase the MR concentration in the cytosol of kidney cells. Binding
reactions were incubated for 2 hours on ice in the presence of excess mifepristone
(RU486, Sigma) to block non-specific glucocorticoid receptor (GR) binding. A
fluorescence polarization based assay (Glucocorticoid Receptor Competitor Assay Kit
Red, PANVERA) was used for GR binding, as per manufacturer recommendations.
Inhibitory constants (Ki, app) defining apparent binding affinity of test compounds to
intracellular receptors were calculated from the observed inhibition of natural ligand
binding at multiple concentrations of test compound. Sex hormone binding globulin
(SHBG, Research Diagnostics) binding was performed using a standard charcoal assay.
13
Reagents: 1 mg lyophylized SHGB powder (Tris), 3H DHT (NEN), 3% charcoal and
0.4% Dextran in PBS; binding buffer: 50 mM Tris, pH 7.6, 100 mM NaCl, 1 mM EDTA,
1 mM DTT and mock lysate (3.5 µg /100 µL buffer); stock solutions: stock SHBG
protein: 1 mg/mL in water = 20 µM, stock 3H-DHT ligand: 9 µM, DHT: 10 mM in
DMSO, BMS 564929: 10 mM in DMSO. Compounds diluted in binding buffer were
added to 40 nM 3H-DHT and 20 nM SHBG protein in 200 µL volume and incubated for
1 hour at room temperature. Total Binding: 40 nM 3H-DHT and 20 nM SHBG protein in
200 µL volume, Non Specific Binding: 40 nM 3H-DHT and 20 nM SHBG protein and
1mM cold DHT in 200 µL volume. At the end of the incubation period, 200 µL of the
charcoal solution (3% containing 0.04% dextran) was added to 200 µL of the reactions
and shaken for 15 minutes before centrifugation. Supernatent (200 µL) was then
transferred to the wells of 24-well white Optiplate, 200 µL of scintillant were added with
mixing. Radioactivity counts were read in Topcount.
Stable androgen receptor-luciferase reporter transactivation assays
Functional transactivation assays were conducted to assess the potency and efficacy of
androgen agonists in muscle and prostate cell backgrounds. We stably transfected
C2C12 mouse myoblast (ATCC) and rat PEC prostate epithelial cell lines (obtained from
Dr. Douglas Hixson, Rhode Island Hospital) with both full length rat AR and the
enhancer/luciferase reporter. The enhancer/reporter construct, containing the 2XDR-1
ARE was developed by random mutagenesis of an AR/GR consensus response element
sequence and has been used as an AR specific response element based on its AR
14
selectivity demonstrated in CV-1 cells (40). We plated both cell lines in 96 well format
at 6,000 cells/well in high glucose DMEM without phenol red (Gibco BRL, Cat. No.:
21063-029) containing 10% charcoal and dextran treated FBS (HyClone Cat. No.:
SH30068.02), 50 mM HEPES Buffer (Gibco BRL, Cat. No.: 15630-080), 1X MEM Na
Pyruvate (Gibco BRL, Cat. No.: 11360-070), 0.5X Antibiotic-Antimycotic, 800 µg/mL
Geneticin (Gibco BRL, Cat. No.: 10131-035) and 800 µg/mL Hygromycin β(Gibco BRL,
Cat. No.: 10687-010). Forty eight hours later, we added ligands (10 µL/well) in fresh
media. After an additional 24 hours, we used the Steady-Glo™ Luciferase Assay System
to detect activity according to the manufacturer’s instructions (Promega, Cat. No.:
E2520).
Enzymatic assays
A CYP19 aromatase inhibition assay using human CYP19+P450 Reductase
Supersomes™ GENTEST P260 was used to measure inhibition of aromatase by BMS-
564929. IC50’s were determined utilizing assay conditions recommended by the
manufacturer (BD Gentest Corporation).
Rodent muscle and prostate tissue growth assays
Matched sets of castrated, sexually mature Harlan Sprague-Dawley rats (42-56 days old,
200-250 g), were dosed once daily by oral gavage with BMS-564929 (0.00001 mg/kg –
10 mg/kg) in solution/suspension of 80% PEG 400 and 20% Tween 20 (PEG/TW, both
15
obtained from Sigma Chemical) for 14 days. Two control groups, one sham operated
intact and one castrated, were dosed orally with the PEG/TW vehicle only, beginning on
day 15 following surgery. Animals were dosed (v/w) at 1 mL/kg body weight. TP was
dosed once daily sub-cutaneously in a 10% ethanol/90% peanut oil vehicle as a reference
compound (0.03 mg/kg – 10 mg/kg). After 14 days of treatment, the animals were
sacrificed by carbon dioxide asphyxiation and the levator ani and the ventral prostate
were surgically removed and weighed, and serum was collected for LH measurements.
All experiments were conducted in accordance with regulations of the Animal Care and
Use Committee of the Bristol-Myers Squibb Company, in facilities fully accredited by
the Association for Assessment and Accreditation of Laboratory Animal Care.
Rat luteinizing hormone (rLH) measurements
rLH was quantitatively determined by radioimmunoassay using a Biotrak [125I] kit
(Amersham Pharmacia Biotech) under the manufacturer’s protocol.
Statistical analysis
Data are presented as means ± standard deviation (s.d.) for each group. Statistical
analysis of the in vivo data was performed by a one-way ANOVA. When a significant F
ratio was identified, groups were compared using Fisher’s PLSD post hoc test. P < 0.05
was considered to be significant.
16
X-ray crystallography
The rat AR LBD cDNA, from amino acids 646-901 (rat numbering, human numbering
664-919) was cloned as previously described (41) and included an N-terminal
polyhistidine tag and a thrombin cleavage site. Since the rat and human AR LBD are
identical at the protein level the human numbering scheme was used for discussion and
structural analysis. The rat AR LBD–BMS-564949 complex was crystallized at 20 °C by
vapor diffusion in the hanging drop mode using a 3 mg/mL concentration of AR LBDBMS-
564929 complex as described previously (41). The crystals have the symmetry of
the space group P212121 with unit cell dimensions a = 55.6 Å, b = 65.6 Å, c = 70.8 Å.
Data to 3.0 Å resolution were collected on a Rigaku R-AXIS IV++ Imaging Plate system
mounted on a Rigaku RU-300 rotating-anode X-ray generator (Rigaku-MSC, Woodlands,
TX). The data were processed and scaled using the HKL package (HKL Research, Inc.
Charlottesville, VA). The crystals are iso-structural to the AR LBD-DHT structures
previously determined (41, 42). The initial model was subjected to rigid-body refinement
followed by positional and simulated-annealing refinements using CNX (Accelrys,
Inc.,San Diego, CA). At the end of refinement, the structure has an R factor of 24.9%
with a total of 2032 atoms (2003 protein atoms, 21 ligand atoms and 8 solvent atoms).
Results
AR binding affinity of BMS-564929
17
Competition binding assays with the appropriate radioligands were used to determine the
binding affinity of BMS-564929 for AR as well as for progesterone receptor (PR),
estrogen receptors α and β (ERα, β), the glucocorticoid receptor (GR) and the
mineralocorticoid receptor (MR). BMS-564929 is a high affinity ligand for AR with a Ki
of 2.11 ± 0.16 nM (Table 1). This compound is >1000-fold selective for AR versus ERα
and β, GR, and MR, and approximately 400-fold selective vs. PR (data not shown).
Transcriptional activation of AR in rodent skeletal muscle and prostate epithelial stable
cell lines by BMS-564929
The potency, efficacy and in vitro selectivity of BMS-564929 were measured in
functional transactivation assays in muscle and prostate cell backgrounds in order to
assess activity in cellular contexts where appropriate levels of endogenous cofactors for
each tissue type are expressed. Varying concentrations of T and BMS-564929 were
tested using the mouse myoblast (C2C12) and rat prostate epithelial (PEC) cell lines,
each stably transfected with rat AR (rAR) and an androgen–specific response element
(ARE)-driven luciferase reporter (Fig. 3) (40). BMS-564929 exhibited a potency (EC50,
calculated as the concentration at which 50% of the maximum stimulatory effect of DHT
is achieved) of 0.44 ± 0.03 nM compared with 2.81 ± 0.48 nM measured for T in the
C2C12 myoblast cell line (Table 1). In the PEC cell line, the EC50 for BMS-564929 was
8.66 ± 0.22 nM as compared to 2.17 ± 0.49 nM for T. Both compounds achieved
equivalent maximal stimulatory efficacy in each cell line. These data suggest that BMS-
564929 is approximately 20-fold more potent at induction of luciferase reporter gene
18
expression in muscle cells versus prostate cells in vitro. This is in contrast to that for T,
which showed similar potency in both cell backgrounds, suggesting that endogenous cofactors
complementing the trancriptional machinery in each cell line are differentially
recruited by the two ligands. BMS-564929 shows no measurable activity in functional
transactivation assays with ERα/β, GR, MR or PR at concentrations up to 30 µM (data
not shown).
BMS-564929 does not bind to sex hormone binding globulin or inhibit CYP19 aromatase
In addition to selective binding to intracellular hormone receptors, androgens and other
sex steroid hormones can interact with SHBG to trigger a signal transduction pathway via
a membrane receptor for SHBG (RSHBG), using cyclic adenosine monophosphate (cAMP)
as a second messenger (43, 44). This mechanism is believed to be partly responsible for
non-AR-mediated effects of T on gene activation in target cells. In order to evaluate
BMS-564929 binding to the SHBG we used increasing concentrations of BMS-564929 to
compete [3H]-DHT from purified SHBG. In addition to its androgenic and anabolic
activities, T undergoes metabolic conversion to estradiol (E2) by CYP19 (aromatase),
and this process plays an important role in the regulation of gonadotropin feedback,
several brain functions, bone remodelling and lipid metabolism. We measured the ability
of BMS-564929 and T to inhibit aromatase to assess the potential for interference with
E2 production. The results (Table 1) show that BMS-564929 does not interact
appreciably with SHBG, and does not significantly inhibit aromatase activity at
19
concentrations up to 30 µM, indicating that BMS-564929 is unlikely to exhibit or
interfere with these T-mediated physiological effects in humans.
Effects of BMS-564929 on tissue growth and serum luteinizing hormone concentrations
after two weeks of treatment in mature castrated rats
We evaluated the effects of BMS-564929 in sexually mature, castrated male rats, a wellcharacterized
animal model for studying the anabolic or androgenic effects of AR
modulators on skeletal muscle and sex accessory tissues (45, 46). Castration of mature
male rats results in the rapid involution and atrophy of both levator ani muscle and
prostate, reaching a steady state approximately ten days following castration. Increases
in prostate and levator ani wet weights after drug treatment serve as indicators of
androgenic and anabolic activities, respectively. We used T propionate (TP) as the
positive control due to its superior pharmacokinetic properties and enhanced efficacy
both in humans and in preclinical species. In the recovery mode, we dosed rats for
fourteen days beginning fourteen days after surgical castration. Wet weights of prostate
and levator ani (normalized by body weight) remained unchanged during the two week
treatment period for both sham-operated rats and castrated controls treated with vehicle,
relative to pre-treatment values. TP treatment (s.c.) increased both prostate and levator
ani weights in a dose-dependent manner beginning with the 0.03 mg/kg dose and
reaching normal weight (100% of sham-operated vehicle controls) at 1 mg/kg for the
levator ani and 3 mg/kg for the prostate respectively (Fig. 4A). The potency of TP was
calculated as an ED50 value, (defined as the dose at which tissue wet weight reached 50%
20
of the weight of that of sham operated vehicle control animals). TP exhibited an ED50 of
0.21 mg/kg in the levator ani muscle and an ED50 of 0.42 mg/kg in the prostate,
exhibiting 2-fold selectivity for the levator ani vs. prostate. In the same model BMS-
564929 (p.o.) showed substantially more potent activity in the levator ani, exhibiting an
ED50 of 0.0009 mg/kg in the levator ani and an ED50 of 0.14 mg/kg in the prostate; a net
160-fold selectivity for muscle vs. prostate (Fig. 4B). Approximately 100% muscle
stimulation was achieved at 0.1 mg/kg, reaching greater than 125% stimulation at 0.3 and
1 mg/kg. Compared with TP in the same model, BMS-564929 is more than 200 times
more potent in stimulation of muscle and 80 times more selective for muscle vs. prostate.
In man and rodents, luteinizing hormone (LH) directly stimulates the production of T
in the Leydig cells of the testes. In response to this stimulation, increased levels of T
inhibit further production and secretion of LH through feedback inhibition of the
hypothalmic-pituitary axis, leading to the suppression of T production in normal healthy
individuals. In elderly men treated for age-related sarcopenia, complete inhibition of LH
production might possibly lead to a near total depletion of endogenous T, which could
subsequently result in E2 deficiency and a consequent hormonal imbalance. To evaluate
the potential for BMS-564929 to suppress LH secretion, we measured serum LH levels at
the termination of the castrated rat assay. BMS-564929 suppressed secretion of LH with
an ED50 of 0.008 mg/kg and showed 9-fold selectivity for levator ani muscle stimulation
vs. LH suppression (Fig. 4C). Compared to TP, which has an ED50 of 0.26 mg/kg for LH
suppression and essentially no selectivity (1.2-fold) for levator ani stimulation vs. LH
suppression, BMS-564929 is approximately 33 times more potent in LH suppression, but
9-fold more selective for levator ani.
21
The selectivity of BMS-564929 on tissue growth is maintained after eight weeks of
treatment in mature castrated rats
To determine the effects of longer–term dosing with BMS-564929 on androgen sensitive
tissues in castrated rats, the compound was administered across a wide dose range (1 µg –
1 mg/kg) for eight weeks in the same recovery schedule of administration as the two–
week experiments. A dose-dependent increase in levator ani muscle and prostate wet
weight was observed with BMS-564929 treatment, with ED50 values of 0.001 mg/kg and
0.09 mg/kg for the levator ani and prostate, respectively (Fig. 5). In these same
experiments, TP exhibited somewhat less potent and minimally selective activity in each
tissue (ED50 of 0.19 mg/kg and 0.39 mg/kg for the levator ani and prostate, respectively).
The results for both compounds are in close agreement with those observed in the twoweek
study, indicating a lack of temporal effect on the relative and absolute potency and
selectivity of BMS-564929.
In addition to the previously obtained endpoints, whole body adiposity was measured
by dual-energy x-ray absorptiometry (DEXA) scan in this experiment. Interestingly, we
observed similar dose-dependent decreases in adiposity in both the BMS-564929 and TP
treatment groups (4% decrease in fat mass at 1 mg/kg BMS-564929 or 3 mg/kg TP, data
not shown). The significant decrease in adiposity suggests the potential for clinical
improvements in lean body mass in humans in addition to strictly anabolic effects.
22
Structure of BMS-564929 in complex with the ligand binding domain of Androgen
Receptor
The AR-LBD structure is analogous to our previously published structures (41, 42), and
consists of the nuclear receptor ligand binding domain which has been shown to have
three layers of orthogonally packed α-helices and four β-strands arranged in two β-
hairpin motifs. For the current structure, AR-LBD was initially refined in the absence of
ligand. The difference electron density clearly showed the position of BMS-564929 so
the ligand was then fit to the difference electron density and the protein-ligand complex
was subjected to two additional cycles of CNX refinement. The structure was analyzed
following refinement. The His-Tag with the first seven residues of the amino-terminus
and the carboxy-terminal residue of AR-LBD were not visible in the electron density and
were excluded from the final model. No solvent atoms were found in or near the ligand
binding site. The ligand binding site is well defined, and residues L704, N705, M745,
M749, R752, F764, M780, T877 have direct interactions with BMS-564929. There are
two possible hydrogen bonds between AR-LDB and BMS-564929 involving residues
N705 and R752 (these interactions are described in detail below). As seen previously for
DHT (41), the majority of the interactions between ligand and protein involve van der
Waals interactions. In our previous work it was seen that for DHT hydrophobic residues
were shown to interact with the steroid nucleus, and for BMS-564929 the same residues
contact the phenyl and bi-cyclic rings while hydrogen bonding interactions anchor each
end of the ligand providing affinity and selectivity.
Discussion
23
The results described herein demonstrate that BMS-564929 is a highly potent and
selective AR agonist in vitro and in vivo, and it exhibits unprecedented muscle selectivity
in an established rodent tissue growth model. In the broadest context, certain
components of pharmacological selectivity are inherent to non-steroidal SARM agents
such as BMS-564929, since compounds of this class are not reduced by 5α-reductase or
aromatized by aromatase to other hormonally active agents which act to modulate other
pathways. The lack of potential nongenomic effects associated with SHBG binding adds
a further means of selectivity to BMS-564929 relative to T and DHT.
The 20-fold selectivity observed in relevant cell contexts in vitro correlates with the
tissue selectivity observed in vivo under varied dosing regimens. This strongly suggests
that selectivity has a fundamental molecular basis whereby differential utilization of
endogenous coregulators in each respective cell or tissue type affect the induction of
genes responsible for each tissue’s growth and differentiation. In a similar manner,
various diverse ER ligands differentially induced unique receptor conformations and
subsequently recruited specific coactivators, ultimately leading to tissue selective action
(47, 48). Ongoing studies target identification and characterization of the specific ARregulated
genes involved in the growth of each respective tissue, facilitated by the
availability of selective tool molecules. We are also conducting further studies to
evaluate the potential for additional benefits on bone, body composition, cognition, and
sexual function. It is particularly noteworthy that in both the two–week and eight–week
in vivo experiments in castrated rats, BMS-564929 is able to achieve significant and
sustained hyperanabolic activity on skeletal muscle at doses which incompletely restore
24
prostate size to that of normal animals. Should this level of separation of effects extend
to the human clinical setting, it would potentially allow more aggressive therapy for
abbrogation of functional decline than can be presently pursued with T and its analogues.
Insight into the molecular basis for tissue selectivity may be gleaned from
comparative x-ray crystallographic analysis of various agonists bound to the ligand
binding domain (LBD) of the AR (49), as the molecular basis for differential co-factor
recruitment is a ligand-mediated process. Comparison of the specific residues in the AR
LBD involved in binding to DHT (41) versus those interacting with BMS-564929 (Fig.
6B) suggest intriguing opportunities for further mechanistic study. Since the x-ray
crystal structures of T and DHT with the AR LBD are essentially identical, we chose that
with DHT for purposes of discussion, as this is the active androgen in prostate due to
local conversion of T by 5α-reductase. Figure 6A shows the alignment of AR-LBD from
the DHT (blue) and BMS-654929 (red) complexes. Structural alignment shows that the
two proteins, including amino acid side-chains, are very similar with a RMSD (root mean
square deviation) of 0.58 Å (backbone) and 0.97 Å (all heavy atoms). Both DHT and
BMS-564929 engage R752 via the 3-keto group of DHT (2.89 Å) or the CN group of
BMS-564929 (3.22 Å). A significant difference for these respective ligands is the
interaction of the side chain of F764 with the phenyl ring of BMS-564929 via a π-edge to
face interaction (5.20 Å, ring centroids), which is only weakly captured by DHT through
its single A-ring π-bond. This interaction for BMS-564929 is likely to provide up to 5
kcal/mol binding energy, and may result in steric compression of the LBD in a manner
distinct from that induced by DHT. Additionally, the 17β-hydroxyl group of DHT is
within hydrogen bonding distance to both N705 (2.80 Å) and T877 (2.70 Å), whereas the
25
hydroxyl group of BMS-564929 engages only N705 (2.73 Å). As T877 has been shown
to be a critical residue for recognition of agonism / antagonism of AR ligands in the
common A877T AR (50) mutant seen in prostate cancer patients, these findings may
have significant implications regarding differential activation of AR in prostate. It should
also be noted that this particular point mutation allows antagonists to co-crystallize with
the AR, which to date has not been possible with the wild-type sequence (42). These key
contact differences for the two AR ligands are located in specific helices which
cooperatively form a hydrophobic groove critical for recognition of the LxxLL motif of
coactivator proteins (51), and ligand binding alters the positioning of these helices,
thereby affecting coactivator binding. As x-ray structural analysis provides only a static
picture of ligand-receptor interactions, additional studies are ongoing in these laboratories
using dynamic methods to investigate AR helical mobility as it relates to these unique
interactions identified through crystallography (52). Several AR-regulated genes which
are differentially induced in specific tissues by the present SARM’s relative to the natural
ligands have been identified (unpublished results). We are also investigating the
structure-function relationship of additional ligands structurally related to BMS-564929
which induce different phenotypes in rodents to more specifically characterize the
importance of selected interactions as it relates to tissue selectivity.
Based on the present results, as well as extensive preclinical safety testing, BMS-
564929 has advanced to clinical trials. We expect that this compound will be suitable as
a once daily oral drug for the treatment of age-related musculoskeletal decline in men
with an improved side-effect profile relative to T (e.g. prostate growth, hematopoiesis,
liver toxicity). The marked muscle selectivity of BMS-564929 has the potential to
26
provide a greater safety window for the beneficial effects of androgens while reducing
the likelihood of deleterious side effects observed with T treatment, most notably the
potential for hyperstimulatory effects on the prostate, thereby providing a valuable means
for addressing a growing unmet medical need.
Acknowledgements
We gratefully acknowledge Celia Darienzo and Lifei Wang, and the Department of
Pharmaceutical Candidate Optimization at BMS for extensive characterization of BMS-
564929. We thank Dr. Michael Blanar and Dr. Robert Zahler for their helpful comments.
27
References
1. Morley JE, Kaiser, FE, Perry HM 3rd, Patrick P, Morley PM, Stauber PM,
Vellas B, Baumgartner RN, Garry PJ 1997 Longitudinal changes in testosterone,
luteinizing hormone, and follicle-stimulating hormone in healthy older men.
Metabolism 46:410-413
2. Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR 2001 Longitudinal
effects of aging on serum total and free testosterone levels in healthy men. J Clin
Endocrinol Metab 86:724-731
3. Feldman HA, Longcope C, Derby CA, Johannes CB, Araujo AB, Coviello AD,
Bremner WJ, McKinlay JB 2002 Age trends in the level of serum testosterone and
other hormones in middle-aged men: Longitudinal results from the Massachusetts
male aging study. J Clin Endocrinol Metab 87:589-598
4. Dunn JF, Nisula BC, Rodbard D 1981 Transport of steroid hormones – binding of
21 endogenous steroids to both testosterone-binding globulin and corticosteroidbinding
globulin in human plasma. J Clin Endocrinol Metab 53:58-68
5. Matsumoto AM 2002 Andropause: Clinical implications of the decline in serum
testosterone levels with aging in men. J Gerontol 57A:M76-M99
6. Kaufman JM, Vermeulen A 2005 The decline of androgen levels in elderly men
and its clinical and therapeutic implications. Endocrine Rev 26:833-876
7. Tan RS, Pu SJ, Culberson JW 2003 Role of androgens in mild cognitive
impairment and possible interventions during andropause. Med Hypoth 60:448-452
28
8. Gillett MJ, Martins RN, Clarnette RM, Chubb SA, Bruce DG, Yeap BB 2003
Relationship between testosterone, sex hormone binding globulin and plasma amyloid
beta peptide 40 in older men with subjective memory loss or dementia. J. Alzheimers
Dis 5:267-269
9. Bates KA, Harvey AR, Carruthers M, Martins RN 2005 Androgens, andropause
and neurodegeneration: exploring the link between steroidogenesis, androgens and
Alzheimer’s disease. Cell Mol Life Sci 62:281-292
10. Snyder PJ, Peachey H, Hannoush P, Berlin JA, Loh L, Lenrow DA, Holmes JH,
Dlewati A, Santanna J, Rosen CJ, Strom BL 1999 Effect of testosterone treatment
on body composition and muscle strength in men over 65 years of age. J Clin
Endocrinol Metab 84:2647-2653
11. Ferrando AA, Sheffield-Moore M, Yeckel CW, Gilkison C, Jiang J, Achacosa A,
Lieberman SA, Tipton K, Wolfe RR, Urban RJ 2002 Testosterone administration
to older men improves muscle function: Molecular and physiological mechanisms.
Am J Physiol Endocrinol Metab 282:E601-E607
12. Blackman MR, Sorkin JD, Münzer T, Bellantoni MF, Busby-Whitehead J,
Stevens TE, Jayme J, O’Connor KG, Christmas C, Tobin JD, Stewart KJ,
Cottrell E, St. Clair C, Pabst KM, Harman SM 2002 Growth hormone and sex
steroid administration in healthy aged women and men. A randomized controlled
trial. J Am Med Assoc 288:2282-2292
13. Schroeder ET, Singh A, Bhasin S, Storer TW, Azen C, Davidson T, Martinez C,
Sinha-Hikim I, Jaque SV, Terk M, Sattler FR 2003 Effects of an oral androgen on
29
muscle and metabolism in older, community-dwelling men. Am J Physiol Endocrinol
Metab 284:E120-E128
14. Bhasin S, Woodhouse L, Casaburi R, Singh AB, Mac, RP, Lee M, Yarasheski
KE, Sinha-Hikim I, Dzekov C, Dzekov J, Magliano L, Storer TW 2005 Older
men are as responsive as young men to the anabolic effects of graded doses of
testosterone on skeletal muscle. J Clin Endocrinol Metab 90:678-688
15. Page S, Amory JK, Bowman FD, Anawalt BD, Matsumoto AM, Bremner WJ
Tenover JL 2005 Exogenous testosterone (T) alone or in combination with
finasteride increases physical performance, grip strength, and lean body mass in older
men with low serum T. J Clin Endocrinol Metab 90:1502-1510
16. Basaria S, Dobs AS 2001 Hypogonadism and androgen replacement therapy in
elderly men. Am J Med 110:563-572
17. Gruenwald DA, Matsumoto AM 2003 Testosterone supplementation therapy for
older men: Potential benefits and risks. J Am Geriat Soc 51:101-115
18. Rhoden EL, Morgentaler A 2004 Risks of testosterone-replacement therapy and
recommendations for monitoring. N Engl J Med 350:482-492
19. Liu PY, Swerdloff RS, Veldhuis JD 2004 The rationale, efficacy and safety of
androgen therapy in older men: Future research and current practice
recommendations. J Clin Endocrinol Metab 89:4789-4796
20. Hijazi RA, Cunningham GR 2005 Andropause: Is androgen replacement therapy
indicated for the aging male? Ann Rev Med 56:117-137
21. Ly LP, Jimenez M, Zhuang TN, Celermajer DS, Conway AJ, Handelsman DJ
2001 A double-blind, placebo-controlled, randomized clinical trial of transdermal
30
dihydrotestosterone gel on muscular strength, mobility, and quality of life in older
men with partial androgen deficiency. J Clin Endocrinol Metab 86:4078-4088
22. Wang C, Swedloff RS, Iranmanesh A, Dobs A, Snyder PJ, Cunningham G,
Matsumoto AM, Weber T, Berman N 2000 Transdermal testosterone gel improves
sexual function, mood, muscle strength, and body composition parameters in
hypogonadal men. Testosterone Gel Study Group. J Clin Endocrinol Metab 85:2839-
2853
23. Anderson RA, Wallace AM, Sattar M, Kumar N, Sundaram K 2003 Evidence
for tissue selectivity of the synthetic androgen 7-alpha-methyl-19-nortestosterone in
hypogonadal men. J Clin Endocrinol Metab 88:2784-2793
24. Schroeder ET, Zheng L, Yarasheski KE, Qian D, Stewart Y, Flores C, Martinez
C, Terk M, Sattler FR 2004 Treatment with oxandrolone and the durability of
effects in older men. J Appl Physiol 96:1055-102
25. Gelfand MM, Wiita B 1997 Androgen and estrogen-androgen hormone replacement
therapy: A review of the safety literature, 1941-1996. Clin Therapeut 19:383-404
26. Negro-Vilar A 1999 Selective androgen receptor modulators (SARMs): A novel
approach to androgen therapy for the new millennium. J Clin Endocrinol Metab
84:3459-3462
27. Hamann LG, Mani NS, Davis RL, Wang XN, Marschke KB, Jones TK 1999
Discovery of a potent orally active nonsteroidal androgen receptor agonist: 4-ethyl-
1,2,3,4-tetrahydro-6-(trifluoromethyl)-8-pyridono-[5,6-g]-quinoline (LG121071). J
Med Chem 42:210-212
31
28. Zhi L, Martinborough E 2001 Selective androgen receptor modulators (SARMs).
Ann Rep Med Chem 36:169-180
29. Chengalvla M, Oh T, Roy AK 2003 Selective androgen receptor modulators.
Expert Opin Ther Patents 13:59-66
30. Gao W, Bohl CE, Dalton JT 2005 Chemistry and structural biology of androgen
receptor. Chem Rev 105:3352-3370
31. Chen J, Hwang JD, Bohl CE, Miller DD, Dalton JT 2005 A selective androgen
receptor modulator for hormonal male contraception. J Pharmacol Exp Therap
312:546-553
32. Shang Y, Myers M, Brown M 2002 Formation of the androgen receptor
transcription complex. Mol Cell 9:601-610
33. Katzenellenbogen BS, Choi I, Delage-Mourroux R, Ediger TR, Martini PG,
Montano M, Sun J, Weis K, Katzenellenbogen JA 2000 Molecular mechanisms of
estrogen action: selective ligands and receptor pharmacology. J Steroid Biochem Mol
Biol 74:279-285
34. Dutertre M, Smith CL 2000 Molecular mechanisms of selective estrogen receptor
modulator (SERM) action. J Pharmacol Exp Ther 295:431-437
35. McDonnell DP, Connor CE, Wijayaratne A, Chang CY, Norris JD 2002
Definition of the molecular and cellular mechanisms underlying the tissue-selective
agonist/antagonist activities of selective estrogen receptor modulators. Recent Prog
Horm Res 57:295-316
36. Shang Y, Brown M 2002 Molecular determinants for the tissue specificity of
SERMs. Science 295:2465-2468
32
37. Smith CL, O’Malley BW 2004 Coregulator function: A key to understanding tissue
specificity of selective receptor modulators. Endocrine Rev 25:45-71
38. Heinlein, CA, Chang C 2002 Androgen receptor (AR) coregulators: An overview.
Endocrine Rev 23:175-200
39. Kazmin D, Prytkova T, Cook CE, Wolfinger R, Chu TM, Beratan D, Norris JD,
Chang CY, McDonnell DP 2006 Linking ligand induced alterations in androgen
receptor structure to differential gene expression; a first step in the rational design of
selective androgen receptor modulators (SARMs). Mol Endo 20:1201-1217
40. Zhou Z, Corden JL, Brown TR 1997 Identification and characterization of a novel
androgen response element composed of direct repeat. J Biol Chem 272:8227-8235
41. Sack JS, Kish KF, Wang C, Attar RM, Kiefer SE, An Y, Wu GY, Scheffler JE,
Salvati ME, Krystek SR Jr, Weinmann R, Einspahr HM 2001 Crystallographic
structures of the ligand-binding domains of the androgen receptor and its T877A
mutant complexed with the natural agonist dihydrotestosterone. Proc Natl Acad Sci
USA 98:4904-4909
42. Salvati ME, Balog A, Shan W, Wei DD, Pickering D, Attar RM, Geng J, Rizzo
CA, Gottardis MM, Weinmann R, Krystek SR, Sack J, An Y, Kish K 2005
Structure based approach to the design of bicyclic-1H-isoindole-1,3(2H)-dione based
androgen receptor antagonists. Bioorg Med Chem Lett 15:271-276
43. Rosner W, Hryb DJ, Khan MS, Nakhla AM, Romas NA 1999 Androgen and
estrogen signaling at the cell membrane via G-proteins and cyclic adenosine
monophosphate. Steroids 64:100-106
33
44. Heinlein CA, Chang C 2002 The roles of androgen receptors and androgen-binding
proteins in nongenomic androgen actions. Mol Endocrinol 16:2181-2187
45. Beyler AL, Arnold A, Potts GO 1968 Methods for evaluating anabolic and catabolic
agents in laboratory animals. J Am Med Women’s Assoc 23:708-712
46. Chinoy NJ, Sheth KM, Shah VC 1973 A comparative histophysiological study on
the normal, castrated and testosterone-treated levator ani muscle of rodents. Acta
Endocrinol 74:389-398
47. Kraichely DM, Sun J, Katzenellenbogen JA, Katzenellenbogen BS 2000
Conformational changes and coactivator recruitment by novel ligands for estrogen
receptor-β: Correlations with biological character and distinct differences among
SRC coactivator family members. Endocrinol 141:3534-3545
48. Egner U, Heinrich N, Ruff M, Gangloff M, Mueller-Fahrnow A, Wurtz JM 2001
Different ligands–different receptor conformations: Modeling of the hERα LBD in
complex with agonists and antagonists. Med Res Rev 21:523-539
49. Poujol N, Wurtz JM, Tahiri B, Lumbroso S, Nicolas JC, Moras D, Sultan C
2000 Specific recognition of androgens by their nuclear receptor. J Biol Chem
275:24022-24031
50. Bohl CE, Gao W, Miller DD, Bell CE, Dalton JT 2005 Structural basis for
antagonism and resistance of bicalutamide in prostate cancer. Proc Natl Acad Sci
USA 102:6201-6206
51. He B, Gampe RT Jr, Kole AJ, Hnat AT, Stanley TB, An G, Stewart EL, Kalman
RI, Minges JT, Wilson EM 2004 Structural basis for androgen receptor interdomain
34
and coactivator interactions suggests a transition in nuclear receptor activation
function dominance. Mol Cell 16:425-438
52. Togashi M, Borngraeber S, Sandler B, Fletterick RJ, Webb P, Baxter JD 2005
Conformational adaptation of nuclear receptor ligand binding domains to agonists:
potential for novel approaches to ligand design. J Steroid Biochem Mol Biol 93:127-
137
35
FIGURE LEGENDS
FIG. 1. Structures of endogenous androgen receptor ligands testosterone (T) and 5α-
dihydrotestosterone (DHT), and selective androgen receptor modulator BMS-564929.
FIG. 2. Synthesis route for the preparation of BMS-564929.
FIG. 3. In vitro activation of AR-mediated luciferase reporter gene induction in C2C12
and PEC cells with (A) testosterone and (B) selective androgen receptor modulator BMS-
564929 relative to 5α-dihydrotestosterone.
FIG. 4. Effects of (A) subcutaneous testosterone propionate or (B) oral BMS-564929 on
wet weight of levator ani muscle and prostate, and (C) on suppression of serum
luteinizing hormone in mature castrated male rats after 2 weeks of once daily treatment.
FIG. 5. Effects of (A) subcutaneous testosterone propionate or (B) oral BMS-564929 on
wet weight of levator ani muscle and prostate in mature castrated male rats after 8 weeks
of once daily treatment in recovery mode.
FIG. 6. (A) Protein Structural alignment (Maestro, Schrodinger LLC. Portland, OR) of
AR ligand binding domains from the dihydrotestosterone (DHT) (blue/cyan) and BMS-
564929 (red/yellow) complex. The protein backbone is rendered in ribbon diagram and
the active site residues are displayed in stick rendering. (B) Superimposition of x-ray cocrystal
structures of DHT (green) and BMS-564929 (grey) bound to the androgen
36
receptor ligand binding domain at 2.0 Å and 3.0 Å resolution respectively. Key ligand
binding domain residues are labeled in yellow. DHT makes binding contacts with R-752,
Q-711, and a bifurcated H-bond to N-705 and T-877. BMS-564929 makes contacts with
R-752 and N-705, as well as a π-edge-face interaction with F-764.
TABLE
Table 1. In vitro activities of BMS-564929 (BMS) and testosterone (T)
BMS T
AR binding
Ki (nM)
2.11 ± 0.16
0.25 ± 0.03
C2C12
EC50 (nM)
0.44 ± 0.03
2.81 ± 0.48
PEC
EC50 (nM)
8.66 ± 0.22
2.17 ± 0.49
SHBG
IC50 (nM)
> 30,000
7 ± 1
Aromatase
IC50 (nM)
> 30,000
740 ± 2
N
N
HO H
O
O
CH3 Cl
CN
OH OH
H
H H
O O
BMS-564929
T DHT
FIG. 1
OCN CN
N
N
H O
O Cl
CN
NH
HO
CO2Me
N
HO
CO2Me
HN
O
CN
1. i-Pr2NEt
2. DBU
85%
HO
Cl
Cl
H2N
Cl
1. Ac2O / AcOH, NaOH
2. Br2
3. Zn(CN)2/ Pd(dba)2
4. HCl / EtOH
5. phosgene, NaHCO3
82%
BMS-564929
HCl
NH
HO
CO2H
1. HCl / MeOH
2. BOC2O / NEt3
3. Ph3P / DEAD / PhCO2H
4. KOH / MeOH
5. HCl
80%
FIG. 2
FIG. 3
20
20
0
20
40
60
80
100
120
140
C2C12
PEC
Testosterone
( nM)
Percent Control
-
20
40
60
80
100
120
140
BMS-564929
(nM)
C2C12
PEC
Percent Control
)
-
0
-
10-3 10-2 10-1 1 10 102 103
10-3 10-2 10-1 1 10 102 103
A
B
FIG. 4
Levator ani
Prostate
BMS-564929
Percent Intact Control
-25
0
25
50
75
100
125
150
B
Dose (mg/kg)
Testosterone propionate
0
20
40
60
80
100
120
Dose (mg/kg)
Percent Intact Control
10-3 10-2 10-1 1 10 102 10-4 10-5
Levator ani
Prostate
10-3 10-2 10-1 1 10 102 10-4 10-5
Dose (mg/kg)
-100
-80
-60
-40
-20
0
20
Testosterone
propionate
BMS-564929 Percent Control
C
10-3 10-2 10-1 1 10 10
2
10-4 10-5
A
FIG. 5
-50
0
50
100
150
200
250
Dose (mg/kg)
Percent Intact Control
Levator ani
Prostate
BMS-564929 B
-50
0
50
100
150
200
250
Dose (mg/kg)
Percent Intact Control
Levator ani
Prostate
Testosterone propionate A
10-3 10-2 10-1 1 10 102 10-4 10-5
10-3 10-2 10-1 1 10 102 10-4 10-5
Q-711
R-752
N-705
T-877
F-764
F-876
FIG. 6
