Discovery of potent, orally-active, and muscle-selective androgen receptor modulators based on an N-aryl-hydroxybicyclohydantoin scaffold

Article abstract of DOI:10.1021/jm061101w

A novel, N-aryl-bicyclohydantoin selective androgen receptor modulator scaffold was discovered through structure-guided modifications of androgen receptor antagonists. A prototype compound (7R,7aS)-10b from this series is a potent and highly tissue-selective agonist of the androgen receptor. After oral dosing in a rat atrophied levator ani muscle model, (7R,7aS)-10b demonstrated efficacy at restoring levator ani muscle mass to that of intact controls and exhibited >50-fold selectivity for muscle over prostate.

Full text of DOI:10.1021/jm061101w

7596
J. Med. Chem. 2006, 49, 7596-7599
Discovery of Potent, Orally-Active, and
Muscle-Selective Androgen Receptor Modulators
Based on an N-Aryl-hydroxybicyclohydantoin
Scaffold†
Chongqing Sun,* Jeffrey A. Robl, Tammy C. Wang,
Yanting Huang, Joyce E. Kuhns, John A. Lupisella,
Blake C. Beehler, Rajasree Golla, Paul G. Sleph,
Ramakrishna Seethala, Aberra Fura, Stanley R. Krystek, Jr.,
Yongmi An, Mary F. Malley, John S. Sack, Mark E. Salvati,
Gary J. Grover, Jacek Ostrowski, and Lawrence G. Hamann
Bristol-Myers Squibb Pharmaceutical Research Institute, Post Office
Box 5400, Princeton, New Jersey 08543-5400
receptor modulators (SARMs), such as pyridinoquinoline 29 and
the bicalutamide-derived 310 have appeared in the literature, and
some of the most advanced compounds emerging from these
efforts have been reported to have reached the clinical stage.11
However, these compounds exhibit modest tissue selectivity in
preclinical models. Given the ever-growing aging population
due to the increase in life expectancy, there clearly exists an
unmet medical need for the prevention of functional decline in
aging men. Herein we report the discovery of a novel, potent,
orally active nonsteroidal SARM scaffold. Compounds based
on this core have the potential to provide the clinical benefits
of a classical anabolic steroid but with reduced risk of dose-
limiting side effects.
Previous in-house efforts directed at the identification of novel
AR antagonist scaffolds led to the discovery of a series of N-aryl
bicyclic hydantoins, exemplified by compound 4a.12 Compounds
of this chemotype bind tightly to the AR and function as weak
partial agonists in AR transactivation assays using the mouse
myoblast C2C12 cell line. The in vitro profile of partial agonist
4a (Ki ) 1 nM, EC50 ) 385 nM) encouraged us to conduct
structure-activity relationship (SAR) studies to improve potency
and intrinsic agonist activity.
The rationale to utilize the 4a template to develop bicyclic
azahydantoin-based AR antagonists was recently disclosed.12
Docking a 4a analogue into the X-ray crystal structure of the
DHT-bound wild-type AR ligand binding domain (LBD)13
suggests that 4a binds to the receptor via the nitro group, like
the C-3 carbonyl of DHT, forming H-bonds with both arginine
752 (R752) and glutamine 711 (Q711). The [2.2.1] bicyclic
portion of 4a occupies the same space as the D ring of DHT.
However, this bulky bicyclic moiety and the weak H-bonding
interactions with asparagine 705 (N705) and/or threonine 877
(T877) for this compound through the hydantoin backbone N
may underlie its weak agonist profile. Such an analysis prompted
us to design hydroxylated compounds 7b,c and 10a,b as possible
SARMs. We reasoned that improved binding interactions could
be achieved both by reducing the bulky [2.2.1] bicyclic
hydantoin portion through removal of the bridging ethano unit
and by installing a hydroxyl group at either C-6 or C-7 of the
outermost 5-membered ring of the resultant bicyclic scaffold
to capture stronger H-bonds with N705 and/or T877.
ReceiVed September 20, 2006
Abstract: A novel, N-aryl-bicyclohydantoin selective androgen recep-
tor modulator scaffold was discovered through structure-guided modi-
fications of androgen receptor antagonists. A prototype compound
(7R,7aS)-10b from this series is a potent and highly tissue-selective
agonist of the androgen receptor. After oral dosing in a rat atrophied
levator ani muscle model, (7R,7aS)-10b demonstrated efficacy at
restoring levator ani muscle mass to that of intact controls and exhibited
>50-fold selectivity for muscle over prostate.
The endogenous androgens testosterone (Ta) and 5R-dihy-
drotestosterone (DHT), play an essential role in normal male
development and the subsequent maintenance of secondary male
characteristics, such as muscle mass and strength, body com-
position, as well as sexual function.1 The physiological effects
of androgens are mediated through transcriptional activation of
the androgen receptor (AR), a member of the nuclear hormone
receptor superfamily of intracellular ligand-dependent transcrip-
tion factors.2 As men age, circulating serum T levels decline
progressively, which subsequently leads to age-related functional
decline (frailty).3 Several recent clinical studies have shown that
supplementation of T at physiological doses in elderly men
resulted in significant increases in lean body mass and muscle
strength and a decrease in adipose tissue.4 However, use of T
replacement therapy in elderly men is limited due to concerns
about potential side effects, for example, exacerbation of benign
prostatic hypertrophy (BPH) or occult prostate cancer.5 In
addition, T cannot be administered orally due to its rapid hepatic
elimination and, consequently, is given less conveniently by
intramuscular injection, surgical implantation, or transdermal
delivery using patches or gels.6 Alkylation of steroidal androgens
at C-17 as in fluoxymesterone (1) has allowed oral administra-
tion, but hepatotoxicity limits the use of such analogues in
chronic therapy.7
There has been growing interest in recent years in finding
orally active androgens that could offer greater separation
between desired anabolic and undesired androgenic effects.8
These efforts have been further stimulated by recent success in
developing and marketing selective estrogen receptor modulators
(SERMs). Several orally active, nonsteroidal selective androgen
A general synthesis of unsubstituted and 6-hydroxy com-
pounds 7a-c is outlined in Scheme 1. Treatment of proline or
4-hydroxyproline esters 5a-c with 4-nitro-1-naphthaleneiso-
cyanate at room temperature yielded urea 6a-c, which were
subsequently treated with DBU at elevated temperature to afford
compounds 7a-c as either racemic (7a) or diastereomeric (7b,c)
mixtures.
† Structure coordinates of (-)-10b bound to the rat AR-LBD have been
deposited in the Brookhaven protein database (PDB ID 2IHQ).
* To whom correspondence should be addressed. Phone: 609-818-5817.
Fax: 609-818-3460. E-mail: chongqing.sun@bms.com.
a Abbreviations: AR, androgen receptor; SARM, selective androgen
receptor modulator; T, testosterone; DHT, 5R-dihydrotestosterone; TP,
testosterone propionate; E2, estrodiol; LBD, ligand binding domain; LH,
luteinizing hormone.
10.1021/jm061101w CCC: $33.50 © 2006 American Chemical Society
Published on Web 11/24/2006

Letters
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26 7597
Scheme 3a
Scheme 1a
a Reagents and conditions: (a) Boc2O, NEt3, CH2Cl2, rt, 3 h, 92%; (b)
Ph3P, DEAD, PhCO2H, THF, rt, 2 h, 94%; (c) 1 M KOH/MeOH, rt, 1 h,
85%; (d) TFA, CH2Cl2, 0 °C, 2 h, 100%; (e) 4-cyano-1-naphthaleneiso-
cyanate, i-Pr2NEt, CH2Cl2, 4 Å mol. sieves, rt, 2 h; (f) DBU, rt, 12-40 h,
80% (2 steps).
a Reagents and conditions: (a) 4-nitro-1-naphthalene isocyanate, toluene,
4 Å mol. sieves, 25 °C, 16 h; (b) DBU, 80 °C, 2 h, 20-75% (2 steps).
Scheme 2a
Table 1. Receptor Binding and Transactivation Assay Data for 4a,b,
7a-c, and 10a,b
d
Kia
(nM)
EC50b (nM)
(I.A.)c
IC50
cmpd
DHT
4a
4b
7a
7b
7c
(()-10a
(-)-10a
(+)-10a
(()-10b
(-)-10b
(+)-10b
(nM)
0.3
1.0
24
2.8 (100%)
385 (85%)
>1000
260
1.7
1.5
4.3
3.5
2.1
1.5
5.5
3.2
5.3
270 (63%)
320 (65%)
265 (71%)
4.4 (96%)
1.5 (101%)
281 (63%)
6.3 (101%)
2.3 (103%)
1298 (49%)
60
a Binding determined through direct displacement of ligand with [3H]-
DHT in the MDA-453 cell line. b Functional agonist activity determined
through a stably transfected mouse myoblast C2C12 cell line utilizing a
luciferase reporter. c I.A. ) intrinsic activity relative to DHT. d Functional
antagonist activity determined in a stably transfected mouse myoblast C2C12
cell line utilizing a luciferase reporter in the presence of DHT.
a Reagents and conditions: (a) 4-R-1-naphthaleneisocyanate, toluene, 4
Å mol. sieves, rt, 16 h; (b) DBU, 80 °C, 2 h, 40-65% (2 steps); (c) chiral
HPLC.
When trans-3-hydroxy-L-proline methyl ester 8 was subjected
to the same reaction sequence using either 4-nitro- or 4-cyano-
1-naphthaleneisocyanate (Scheme 2), unexpectedly, a racemic
mixture of (()-10a (R ) NO2) or (()-10b (R ) CN) was
obtained. These mixtures were resolved by chiral HPLC to
provide individual enantiomers (7S,7aR)-10a or -10b [(+)-10a/-
10b] and (7R,7aS)-10a or -10b [(-)-10a/-10b], respectively.
The relative stereochemistry of (-)-10b was determined by
single-crystal X-ray crystallography,14 and the absolute stereo-
chemistry was confirmed by solving a cocrystal complex of (-)-
10b bound to the AR-LBD (vide infra). To our surprise, the
relative stereochemistry of the hydroxyl group and ring juncture
in (-)-10b was found to be cis rather than the trans-
stereochemistry of hydroxyester 8. Considering the presence of
a â-hydroxy amide in bicyclic hydantoin 10, we speculated that
formation of (()-10a or (()-10b occurred via retro-Aldol
reaction under the strongly basic and elevated-temperature
conditions to yield the more thermodynamically favored cis-
product. When urea intermediate 9b was treated with DBU at
room temperature for 4 days, (+)-10b was predominantly
formed, suggesting that at room temperature the ester group
epimerization precedes cyclization to yield (+)-10b. Based on
these findings, we designed a stereospecific synthesis of the
more active enantiomer (-)-10b, as outlined in Scheme 3. Boc
protection of 8, followed by a Mitsunobu inversion of the
hydroxyl group of the resultant 11 with benzoic acid, gave
diester 12. Treatment of 12 with KOH in MeOH, followed by
Boc deprotection with trifluoroacetic acid afforded cis-3-
hydroxy-L-proline 13. Reaction of 13 with 4-cyano-1-naphtha-
leneisocyanate at room temperature provided the intermediate
urea, which was subsequently treated with DBU at room
temperature to give, after crystallization, enantiomerically pure
(-)-10b.
Table 1 summarizes the AR binding affinities and transcrip-
tional activities of compounds 4a,b, 7a-c, and 10a,b. Com-
pound 7a (X ) H), a simplified bicyclic hydantoin, retained
comparable binding and functional activity to that of the bridged
bicyclic 4a. Both 7b (X ) (R)-OH) and 7c (X ) (S)-OH) also
showed agonist activity similar to 7a, indicating that a hydroxyl
group at the 6-postion of the bicyclic hydantoin core cannot
effectively approximate the H-bonding interactions between the
native hormones T and DHT with the AR. However, with the
hydroxyl group at the 7-position, as in (()-10a (R ) NO2), a
more pharmacologically significant result was observed. While
the binding affinity for (()-10a remained unchanged relative
to the unsubstituted parent 7a, the functional activity was
significantly enhanced (>60-fold). Compound (()-10b (R )
CN) exhibited essentially the same binding and agonist func-
tional activity as that of (()-10a. For cyano analogue (()-10b,
this represented a dramatic reversal in functional activity, as
the corresponding bridged bicyclic analogue 4b (R ) CN) is
functionally an AR antagonist, with an IC50 ) 260 nM. This
finding was also significant because compounds containing the
nitronaphthyl moiety such as 4a tested positive in Ames
mutagenicity assays, while the corresponding cyanonaphthyl
analogues (e.g., 4b) tested negative. Second, this demonstrated
that the specific molecular interactions of the 7-OH group in
(()-10 within the AR LBD were sufficient to convert an

7598 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26
Letters
Figure 1. Effects of testosterone propionate (TP; s.c., q.d.) or (-)-
10b (p.o., q.d.) on wet weight of levator ani and prostate in mature
castrated male rats.
Figure 3. X-ray cocrystal structures of (-)-10b (Panel A) and DHT
(Panel B) bound to the rat AR-LBD at 2.0 Å resolution. The helices
are colored magenta, loops are colored yellow, and strands are colored
green. The key ligand binding residues are labeled according to the
corresponding human AR amino acid sequence, and the dashed lines
represent potential hydrogen bonds. The crystallographic water is
rendered in space filling mode, Panel B.
Figure 2. Effects of testosterone propionate (TP; s.c., q.d.) or (-)-
10b (p.o., q.d.) on suppression of serum luteinizing hormone in mature
castrated male rats.
antagonist (e.g., 4b) to an agonist ((()-10b). On this basis, (()-
10a and (()-10b were resolved into their respective pure
enantiomers, which were assayed for AR activity. Compound
(-)-10b was found to be >500-fold more potent in functional
assays than its antipode (+)-10b. Compound (-)-10b was also
assayed against other steroid hormone receptors, and found to
be >1000-fold selective for AR versus estrogen receptor R and
â, glucocorticoid receptor, and mineralocorticoid receptor, and
approximately 270-fold selective versus progesterone receptor
(data not shown).
As the in vitro profile of (-)-10b closely mirrored that of
the native hormones, we proceeded to evaluate the potency and
efficacy of this AR agonist at stimulating skeletal muscle and
prostate growth in the mature castrated rat, a widely accepted
model for investigating anabolic and androgenic activity in
androgen responsive tissues.15 In this assay, mature male rats
(castrated two weeks prior to the start of the study) were treated
with (-)-10b (q.d., p.o.) or testosterone propionate (TP) (q.d.,
s.c.) for two weeks. At the termination of the study, tissue wet
weights were measured and potencies were calculated as ED50
values, defined as the dose at which tissue wet weight reaches
50% that of the corresponding intact sham-operated vehicle-
treated control animals (Figure 1). TP had 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 muscle versus prostate,
whereas compound (-)-10b exhibited an ED50 of 2 mg/kg in
the levator ani and an ED50 of >100 mg/kg in the prostate or
>50-fold selectivity for muscle versus prostate. In this same
animal model, (-)-10b also suppressed secretion of luteinizing
hormone (LH) with an ED50 of 60 mg/kg (Figure 2). This
corresponds to a 30-fold selectivity for the levator ani muscle
stimulation versus LH suppression compared to testosterone
propionate, which had an ED50 of 0.26 mg/kg for LH suppres-
sion and essentially no selectivity (1.2-fold) versus this endpoint.
In addition to the more obvious rationale for desired selectivity
for muscle versus prostate growth, this latter selectivity could
potentially be important for minimization of feedback inhibition
of LH secretion (and subsequent T secretion) through the
hypothalamic-pituitary-gonadal (HPG) axis. As T is the
biosynthetic precursor for estradiol (E2), the lipid balance
thought to be maintained in part through appropriate T/E2 ratio
in men could potentially be affected by a compound that acts
centrally to fully suppress new T secretion.
Comparison of the X-ray cocrystal structures of the AR LBD
complexed with either (-)-10b or DHT reveals that the
nonsteroidal compound has some of the same interactions as
those of the endogenous hormone (Figure 3). Residues R752
and Q711 form hydrogen bonds (2.89 Å and 3.36 Å, respec-
tively) with the A-ring carbonyl of DHT in a network that
includes a water molecule. For the complex of (-)-10b and
the AR, the hydrogen bonds from the cyano group to R752 and
Q711 are similar in nature (2.83 Å and 3.45 Å, respectively),
but the crystallographic water was not seen for this structure
(or other structures determined with analogous compounds),
despite the same 2.0 Å resolution for each. At the distal end of
the ligands, the DHT D-ring’s 17â-hydroxyl group forms strong
hydrogen bonds with T877 and N705 at 2.70 Å and 2.80 Å,
respectively, whereas (-)-10b forms a single hydrogen bond
with N705 at 2.50 Å. An additional significant binding element
unique to (-)-10b is seen with the π-edge to face interaction
between its naphthyl moiety and F764 of the AR. These two
key differences allow (-)-10b to have an overall unique binding
mode compared with DHT. This may result in altered dynamic

Letters
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26 7599
N. Engl. J. Med. 2004, 350, 482-492. (c) Liu, P. Y.; Swerdloff, R.
S.; Veldhuis, J. D. The rational, efficacy and safety of androgen
therapy in older men: Future research and current practice recom-
mendations. J. Clin. Endocrinol. Metab. 2004, 89, 4789-4796.
(6) Bhasin, S.; Bremner, W. J. Clinical review 85: Emerging issues in
androgen replacement therapy. J. Clin. Endocrinol. Metab. 1997, 82,
3-8.
(7) Gelfand, M. M.; Wiita, B. Androgen and estrogen-androgen hormone
replacement therapy: A review of the safety literature. Clin. Ther.
1997, 19, 383-404.
(8) Gao, W.; Bohl, C. E.; Dalton, J. T. Chemistry and structural biology
of androgen receptor. Chem. ReV. 2005, 105, 3352-3370.
(9) Hamann, L. G.; Mani, N. S.; Davis, R. L.; Wang, X.-N.; Marschke,
K. B.; Jones, T. K. Discovery of a potent, orally active, nonsteroidal
androgen receptor agonist: 4-Ethyl-1,2,3,4-tetrahydro-6-(trifluoro-
methyl)-8-pyridono-[5,6-g]-quinoline (LG121071). J. Med. Chem.
1999, 42, 210-212.
receptor conformations leading to differential affinity for/
recruitment of the co-activators and/or co-repressors, which
complete the transcriptional machinery leading to selective gene
activation. Studies are underway in these laboratories to further
elucidate the significance of these findings both at a molecular
and pharmacological level. The novel SARM scaffold described
herein represents a significant advance in achieving orally active,
highly tissue-selective, nonsteroid AR agonists in an animal
model relevant to human musculoskeletal health.
Supporting Information Available: Descriptions of AR bind-
ing and transactivation assays, in vivo pharmacology models, and
detailed procedures and characterization data for compounds
7a-c, 10a,b. This material is available free of charge via the
Internet at http://pubs.acs.org.
(10) Kim, J.; Wu, D.; Huang, D.; Miller, D.; Dalton, J. T. The para
substituent of S-3-(phenoxy)-2-hydroxyl-2-methyl-N-(4-nitro-3tri-
fluoromethyl-phenyl)-propionmides is a major structural determinant
of in vivo disposition and activity of selective androgen receptor
modulators. J. Pharmacol. Exp. Ther. 2005, 315, 230-239.
(11) Omwancha, J.; Brown, T. R. Selective androgen receptor modula-
tors: in pursuit of tissue-selective androgens. Curr. Opin. InVest.
Drugs 2006, 7, 873-881.
(12) Balog, A.; Salvati, M. E.; Shan, W.; Mathur, A.; Leith, L. W.; Wei,
D.; Attar, R. M.; Geng, J.; Rizzo, C. A.; Wang, C.; Krystek, S. R.;
Tokarski, J. S.; Hunt, J. T.; Gottardis, M.; Weinmann, R. The
synthesis and evaluation of [2.2.1]-bicycloazahydantoins as androgen
receptor antagonists Bioorg. Med. Chem. Lett. 2004, 14, 6107-6111.
(13) Sack, J. S.; Kish, K. F.; Wang, C; Attar, R. M.; Kiefer, S. E.; An,
Y.; Wu, G. Y.; Scheffler, J. E.; Salvati, M. E.; Krystek, S. R., Jr.;
Weinmann, R.; Einspahr, H. M. 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. 2001, 98, 4904-4909.
(14) Crystallographic data for (-)-10b have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publica-
tion nos. CCDC 618925.
(15) (a) Beyler, A. L.; Arnold, A.; Potts, G. O. Methods for evaluating
anabolic and catabolic agents in laboratory animals. J. Am. Med.
Women’s Assoc. 1968, 23, 708-712. (b) Chinoy, N. J.; Sheth, K.
M.; Shah, V. C. A comparative histophysiological study on the
normal, castrated and testosterone-treated levator ani muscle of
rodents. Acta Endocrinol. 1973, 74, 389-398.
References
(1) Mooradian, A. D.; Morley, J. E.; Korenman, S. G. Biological actions
of androgens. Endocr. ReV. 1987, 8, 1-28.
(2) Evans, R. M. The steroid and thyroid hormone receptor superfamily.
Science 1988, 240, 889-894.
(3) (a) Harman, S. M.; Metter, E. J.; Tobin, J. D.; Pearson, J.; Blackman,
M. R. Longitudinal effects of aging on serum total and free
testosterone levels in health men. J. Clin. Endocrinol. Metab. 2001,
86, 724-731. (b) Kaufman, J. M.; Vermeulen, A. The decline of
androgen levels in elderly men and its clinical and therapeutic
implications. Endocr. ReV. 2005, 26 (6), 833-876.
(4) (a) Snyder, P. J.; Peachey, H.; Hannoush, P.; Berlin, J. A.; Loh, L.;
Lenrow, D. A.; Holmes, J. H.; Dlewati, A.; Santanna, J.; Rosen, C.
J.; Strom, B. L. Effect of testosterone treatment on body composition
and muscle strength in men over 65 years of age. J. Clin. Endocrinol.
Metab. 1999, 84, 2647-2653. (b) Ferrando, A.; Sheffield-Moore,
M.; Yeckel. M.; Gilkison, C.; Jiang, J.; Achacosa, A.; Lieberman,
S. A.; Tipton, K.; Wolfe, R. R.; Urban, R. J. Testosterone administra-
tion to older men improves muscle function: Molecular and
physiological mechanisms. Am. J. Physiol. Endocrinol. Metab. 2002,
282, E601-E607. (c) Bhasin, S.; Woodhouse, L.; Casaburi, R.; Singh,
A. B.; Mac, R. P.; Lee, M.; Yarasheski, K. E.; Sinha-Hikim, I.;
Dzekov, C.; Dzekov, J.; Magliano, L.; Storer, T. W. Older men are
as responsive as young men to the anabolic effects of graded doses
of testosterone on skeletal muscle. J. Clin. Endocrinol. Metab. 2005,
90, 678-688.
(5) (a) Gruenwald, D. A.; Matsumoto, A. M. Testosterone supplementa-
tion therapy for older men: Potential benefits and risks. J. Am.
Geriatr. Soc. 2003, 51, 101-115. (b) Rhoden, E. L.; Morgentaler,
A. Ricks of testosterone-replacement therapy and recommendations.
JM061101W