Structure-Based Approach to Identify 5-[4-Hydroxyphenyl]pyrrole-2-carbonitrile Derivatives as Potent and Tissue Selective Androgen Receptor Modulators

Article abstract of DOI:10.1021/acs.jmedchem.7b00373

In an effort to find new and safer treatments for osteoporosis and frailty, we describe a novel series of selective androgen receptor modulators (SARMs). Using a structure-based approach, we identified compound 7, a potent AR (ARE EC₅₀ = 0.34 nM) and selective (N/C interaction EC₅₀ = 1206 nM) modulator. In vivo data, an AR LBD X-ray structure of 7, and further insights from modeling studies of ligand receptor interactions are also presented.

Full text of DOI:10.1021/acs.jmedchem.7b00373

Brief Article
pubs.acs.org/jmc
Structure-Based Approach To Identify 5‑[4-Hydroxyphenyl]pyrrole-2-
carbonitrile Derivatives as Potent and Tissue Selective Androgen
Receptor Modulators
Ray Unwalla,* James J. Mousseau, Olugbeminiyi O. Fadeyi, Chulho Choi, Kevin Parris, Baihua Hu,
Thomas Kenney, Susan Chippari, Christopher McNally, Karthick Vishwanathan, Edward Kilbourne,
Catherine Thompson, Sunil Nagpal, Jay Wrobel, Matthew Yudt, Carl A. Morris, Dennis Powell,
Adam M. Gilbert, and Eugene L. Piatnitski Chekler*
Pfizer Worldwide Research & Development, 610 Main Street, Cambridge, Massachusetts 02139, United States
S
* Supporting Information
ABSTRACT: In an effort to find new and safer treatments for osteoporosis and frailty, we describe a novel series of selective
androgen receptor modulators (SARMs). Using a structure-based approach, we identified compound 7, a potent AR (ARE EC50
= 0.34 nM) and selective (N/C interaction EC50 = 1206 nM) modulator. In vivo data, an AR LBD X-ray structure of 7, and
further insights from modeling studies of ligand receptor interactions are also presented.
proof-of-concept that nonsteroidal compounds have the
potential to function as an agonist in one tissue and an
antagonist in another.6 Researchers in the androgen field have
used the lessons learned from the SERM field to develop
SARMs for the treatment of male and female osteoporosis and
frailty. This new second generation of nonsteroidal SARMs
includes BMS-564929 (Bristol-Myers Squibb Co),7 ostarine
(GTx Inc.),8 and MK-0773 (Merck & Co)9 (Figure S1 in
Supporting Information). These SARMs exhibited tissue-
selective profiles in preclinical studies, indicating that they
may provide the required tissue selectivity for a muscle selective
agent.10 However, the clinical utility of oral SARMs appears to
be limited due to the unfavorable plasma lipoprotein changes
such as lowering of HDL-C11a which results in the increased
risk of cardiovascular diseases. It has been reported that
decreased plasma HDL-C levels are due to liver mediated
effects of AR activation. Current testosterone replacement
therapies have focused on use of topical formulations to
minimize these lipid changes and hepatic side effects. There are
also recent reports of transdermal SARMs for the treatment of
various muscles wasting disorders that potentially mitigate CV
and prostate risk.11b,c We have previously reported SARM
series that selectively affect the muscle tissue versus the
INTRODUCTION
■
The steroids testosterone and dihydrotestosterone (DHT) are
androgens that play an important role in the development and
maintenance of a variety of physiological responses such as
male sexual function, bone density, muscle mass, and strength.1
The androgen receptor is a nuclear hormone receptor that is
expressed in many tissues and is responsible for mediating the
actions of testosterone and DHT. Patients that have defects in
the androgen receptor or have androgen deficiencies can be
effectively treated with exogenous testosterone and other
steroidal androgens as a hormone replacement therapy.2−4
The anabolic effects of testosterone have shown benefit in age-
related decline of bone density and muscle mass.5 However, the
side-effect profile of testosterone and other currently available
anabolic steroids precludes their widespread use, and the
chronic administration of steroidal androgens is associated with
potential serious side effects such as hepatotoxicity, prostate
hypertrophy, and cancer. In addition, the oral bioavailability of
testosterone is poor and the route of its administration is
generally through topical formulations. As a result, several
companies have undertaken efforts to find nonsteroidal
selective androgen receptor modulators (SARMs) that exhibit
desired anabolic effects but are devoid of androgenic effects.
In the estrogen field, the development of tissue-selective
estrogen receptor modulators (SERMs) for the prevention and
treatment of postmenopausal osteoporosis has provided the
Received: March 15, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jmedchem.7b00373
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
a
Table 1. Structure−Activity Relationship of N-Me Pyrrole Derivatives 1−25
compd
R1
R2
R3
ARE-LUC agonist, EC50 (nM) (% efficacy) binding, IC50 (nM) N/C interaction, EC50 (nM) (% efficacy)
1
2,3-cyclopentyl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
17.8 (85)
573.2 (103)
37.3 (91)
94.8 (69)
391.0 (74)
1517.7 (69)
0.34 (91)
909.5 (80)
1.7 (80)
8.1
>10000
2
6.3
2014.9 (36)
1542.6 (42)
1550.9 (37)
>10000
3
2-Me
2.9
4
3-Me
3.7
5
2-Et
7.1
6
2-F
59.0
3.2
>10000
7
3-F
1205.8 (46)
>10000
8
2-Cl
47.6
1.4
9
3-Cl
1556.0 (31)
>10000
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
3-CF3
3-NO2
3-CN
3-COOMe
3-SO2Me
3,5-di-F
3,6-di-F
2,3-di-F
3-F, 5-CN
3,5-di-Cl
2,3-di-Me
H
15.2 (69)
233 (77)
14.8
65.0
6.8
>10000
29.7 (82)
1450.3 (36)
>10000
>10000
330.2
1465.9
1.6
>10000
>10000
0.14 (90)
229.1 (84)
980.0 (85)
5.3 (112)
0.17 (76)
34.6 (116)
1644.6 (31)
1341.7 (66)
1136.4 (64)
>10000
0.71 (61)
>10000
85.8
36.8
67.4
28.7
4.9
>10000
28.9 (49)
>10000
>10000
4′-Br
3′-CF3
H
166.9
11.1
127.1
6032.0
>10000
>10000
H
>10000
3-F
>10000
3-F
H
CH2CN
>10000
3-F
H
CH2COOH
>10000
>10000
a
Values are the geometric mean of at least three experiments. If % efficacy or % inhibition is <30%, EC50 and IC50 of >10000 nM are reported.
protstate.12 These compounds show diminished activity in
Scheme 1. Preparation of 5-(4-Hydroxyphenyl)-1-
promoting the intramolecular interaction between the AR
carboxyl (C) and amino (N) termini in an N/C-termini
interaction assay. This assay has been reported previously by
Ghali et al.13 and has been shown to be a good predictor for
undesired androgenic responses in vivo. Thus, a desired profile
for a muscle selective AR modulator would be to have minimal
activity in the N/C-interaction assay. The following report
outlines our strategy to identify and optimize a series of
functionally selective 5-(4-hydroxyphenyl)pyrrole-2-carboni-
triles using a structure-based design/docking approach that
maintains favorable in vitro and in vivo profile on muscle tissue
with minimum effects on the prostate.
methylpyrrole-2-carbonitriles
Structure−Activity Relationships. The in vitro anabolic
activity was assessed in androgen response element luciferase
(ARE-LUC) assay in CV-1 cells. Agonism in this assay is
predictive of in vivo muscle activity. The androgenic activity
was measured using the N/C-interaction in CV-1 cells. A
radiolabeled AR ligand binding assay was performed using
[3H]mibolerone in COS cells transfected with the AR. Table 1
shows the results of the ARE-LUC, AR binding, and N/C-
interaction functional assays. The project strategy was to select
compounds with minimal activity in N/C-interaction assay
relative to the ARE-LUC assay for advancement into in vivo
assessment. Thus, maintaining the agonism to support the
muscle growth and reducing the androgenic effect (N/C-
interaction) should result in tissue selective compounds.
Compound 1 was used as a starting point in a search for
tissue selective compounds. To delineate the critical structural
features responsible for the binding, we performed docking
studies on 1 with X-ray structure of published AR ligand
binding domain (LBD) (PDB code 2AXA).15 The top-scoring
pose16 from docking showed the nitrile group of the pyrrole
ring forming a hydrogen bond with Arg752 residue. The amino
group of Gln711 residue was in a nearby position to the nitrile
RESULTS AND DISCUSSION
■
In our effort to design novel AR inhibitors, we utilized a
cyanopyrrole scaffold discovered in the course of a high-
throughput screening campaign of our corporate library. The in
vitro profile of 1 (Table 1), 5-(7-hydroxy-2,3-dihydro-1H-
inden-4-yl)-1-methylpyrrole-2-carbonitrile, encouraged us to
conduct further structure−activity studies to improve potency
and physicochemical properties. 5-(4-Hydroxyphenyl)-1-meth-
ylpyrrole-2-carbonitrile SARMs (2−25 in Scheme 1) were
prepared via Suzuki cross-coupling of 4-OH-aryl bromides with
2-pyrroloboronates (Scheme 1).14 Boronic acids were
commercially available or prepared according to published
procedures described in the Supporting Information (exper-
imental and analytical information for 1−6, and 8−25 is
presented in the Supporting Information).
B
DOI: 10.1021/acs.jmedchem.7b00373
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
group to form a hydrogen bond, but the angle was found to be
not optimal for such an interaction. The hydroxyl group of the
phenol formed a hydrogen bond with the carbonyl group of the
Asn705 residue. The N-methyl group was oriented toward
helix-3 and was surrounded by residues Leu704 and Leu707,
whereas the fused cyclopentyl ring of 1 was oriented toward the
pocket surrounded by the hydrophobic residues Met780 and
Phe764, as shown in (Figure S2). Recognizing that the hydroxyl
and nitrile groups were critical for hydrogen bonding with the
AR receptor, we decided to keep those features and optimize
the rest of the core scaffold. Throughout our optimization
processes we noticed that subtle changes on the ligand scaffold
had a substantial effect on the functional activity. This type of
behavior, whereby small structural changes on the ligand
modulate functional activity has been previously observed with
other nuclear receptors such as progesterone.17 Removal of the
fused cyclopentyl ring (i.e., 2) surprisingly did not result in any
loss of binding (IC50 (binding) of ∼6.3 nM vs 8.1 nM), but it
did reduce agonist activity. This observation became the first
indication that the binding and the functional potencies do not
perfectly correlate (R2 ≈ 0.5). This phenomenon supports a
general hypothesis that the functional activity is an outcome of
a complex process of cofactors binding to the AR termini and
may not be directly proportional to binding affinity.12 As this
phenomenon deserves a further investigation, the team
proceeded to optimize in vitro profile based on functional
activity. However, the binding potency remained an important
indicator of the ligand−receptor interactions, and a general
trend was observed that a low nM binding potency is required
to achieve functional activity although not every potent binder
became a strong agonist (Table 1). The potency of 2 was
encouraging, and this template provided us further exploration
of the 3 and 5 positions of the phenol scaffold.
other regions of the phenylpyrrole scaffold in a search for
additional switch points of functional activity. Substitution on
the pyrrole in 21 and 22 led to the loss of potency compared to
2. That the N-methyl group of pyrrole is the optimal
substituent is shown by analogs 23−25. A number of previously
published derivatives14 where phenolic group was replaced with
an amine were also tested in the ARE-LUC and binding assays,
but no potent and selective compounds were discovered (see
Supporting Information).
Compound 7 showed a unique biological profile compared
to endogenous AR agonists such as testosterone and DHT.
While both testosterone and 7 are potent in ARE-LUC assay,
only testosterone shows potent N/C-interaction activity (EC50
= 1.1 nM, Figure S3). Also, the lead molecule 7 demonstrated
greater than 1000-fold selectivity against a panel of nuclear
receptors including progesterone receptor where only weak
agonist activity (54% at 10 μM) was observed in a PRE assay
which is equivalent to the ARE assay. In addition to the data
presented above, the 5-(4-hydroxyphenyl)-1-methylpyrrole-2-
carbonitriles showed no activity in hERG dofetilide binding
assay with IC50 > 100 μM for 7 and no effect on proliferation
(IC50 > 300 μM) in the human liver cell line (THLE). There
was also no significant inhibition of cytochrome P450 enzymes
CYP3A4, CYP2C9, CYP2D6, CYP1A2, and CYP2C8 at 3 μM
compound concentration. Additionally, no in-life toxicological
effects were observed for 7 when tested in vivo.
Molecular Modeling and Crystallography. We later
determined the X-ray crystal structure of 7 bound to the AR
LBD at a resolution of 2.3 Å, as shown in Figure 1. As expected
Compound 7 activity in the ARE-LUC assay approaches the
potency of testosterone in the same assay (EC50 ≈ 0.3 nM),
while the activity in N/C-interaction assay provided more than
a 4000-fold selectivity window between in vitro anabolic and
androgenic predictive assays (Figure S3). Compound 7 became
a lead molecule for proof-of-concept studies for both in vitro
and in vivo assays. The SAR around 7 first demonstrates that
groups in the 3-position (7 and 9) are preferred to analogs
substituted in the 2-position (6 and 8). Larger groups in the 3-
position (10−14) exhibited partial or complete loss of
functional activity and a loss in binding affinity. This outcome
can be rationalized by a relatively small size of the binding area
around the phenyl ring. The preferred 3-position inspired an
additional exploration of the chemical space where disub-
stituted analogs 15−20 were prepared. Substitution at the 3 or
at both 3 and 5 positions with small electron withdrawing
groups proved to be most fruitful for activity and pushed these
compounds into desirable lipophilic ligand efficiency (LipE)
space of >5 (for a discussion on the LipE parameter that
combines potency and lipohilicity parameters of ligands, see
Supporting Information Figure S4). The 3,5-difluoro com-
pound 15 demonstrated superior agonist potency compared to
7, but the activity in N/C-interaction assay increased
dramatically (Table 1). The reason for the switch in the
balance of ARE-LUC and N/C-interaction potencies remains
highly elusive with minor changes in structure leading to
significant functional activity profile changes. This phenomenon
has been discussed previously.12 A general trend exists that all
compounds potent in N/C-interaction assay also demonstrate
the agonist activity in the ARE-LUC assay. We also explored
Figure 1. X-ray structure of 7 (PDB code 5V04) in the AR ligand
binding domain. Ligand is colored in cyan, and the critical residues are
shown in green. Hydrogen bond distances are shown by red dotted
lines in Å. The distance between the fluoro group and the carbonyl
backbone of residues Leu704, i.e., 3.3 Å, is shown by dotted blue line.
from our docking studies, the observed binding mode was
similar to the docked pose of 1, the nitrile group formed critical
hydrogen bond with the Arg752 residue, whereas the phenolic
hydroxyl group had a hydrogen bond with the Asn705 side
chain. A highly ordered water molecule was present between
the Arg752 and Gln711 residues which is reminiscent of a
similar network observed in the LBD of the progesterone
receptor.18 The dihedral angle of 47° between the phenol and
the cyanopyrrole group provided an optimal orientation for the
ligand to ensure that key hydrogen bond interactions are
achieved within the binding pocket. The position of 3-fluoro
group of the ligand near the Leu704 residue of the binding site
was especially surprising since the atom closest to the fluorine
C
DOI: 10.1021/acs.jmedchem.7b00373
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
was the oxygen of the carbonyl backbone of the Leu704
residue, at a distance of 3.3 Å between the oxygen and fluorine
atoms. The bond dipoles (i.e., the C(Ar)−F and C(O)
groups) were oriented in parallel and the angle between dipoles
was ∼23° as shown in Figure 1. The orientation of the fluorine
atom close to the oxygen atom would indicate an unfavorable
electrostatic interaction for the ligand, so we examined the
unbiased electron density difference maps of the ligand in order
to confirm the fluorine atom orientation. The maps showed
that there was a clear density for the placement of the fluorine
atom near the carbonyl backbone of the Leu704 residue. There
was no other density within the binding site to suggest an
alternative location for the fluorine atom (Figure S5A). The
mining of the Cambridge Structural Database19 and the Protein
Data Bank showed several examples of a fluorine atom being in
close contact with a carbonyl group, i.e., distance (C−F···C(
O)) of approximately 3.0−3.8 Å) in small and large molecule
X-ray structures. However, in these cases the preferred
approach of the fluorine atom is toward the carbonyl carbon
in an orthogonal orientation (e.g., at an angle ∼170−180° in
C−F−C(O); see Figure S5B). A recent study demonstrated
that these interactions are favorable in protein−ligand X-ray
structures and may help to increase the potency of inhibitors.20
The parallel orientation of the two bond dipoles in our crystal
structure was surprising because this type of dipole−dipole
interaction leads to weak but still unfavorable electrostatic
interactions. The magnitude of such an interaction is unclear,
particularly given the specific protein−ligand contacts we
observed in the X-ray structure. To understand the energetics
of the ligand-bound conformation of 7, we calculated the local
and global ligand strain energies. The local strain energy is the
energy difference between the ligand-bound conformation and
its nearest local minimum conformation in the unbound state,
while the global strain energy is difference between the energies
of the ligand-bound state conformation and the global minima
conformation of the unbound ligand.
Both local and global strain energy calculations using two
widely used molecular mechanics force fields (MMFF94 and
OPLS2005) confirm that 7 pays only a small energy penalty for
the bound state: a mean local strain energy of 0.33 kcal/mol
and a mean global strain energy of 0.54 kcal/mol (Table S1).
These energies are well below the threshold established for the
local and global strain energies for a ligand with 1−3 rotatable
bonds found in a benchmark study of 150 X-ray structures of
protein−ligand complexes.21
The interaction of 3-fluoro group with the carbonyl
backbone of Leu704 was further examined by calculating the
intermolecular potential between these groups. To provide
valuable insights into the strength of such an interaction, we
used a model system between a 3-fluorophenol group
(representative of 7) and the carbonyl backbone of leucine
modeled as N-methylacetamide and performed ab initio
quantum chemical calculations. Figure S6A shows the potential
energy22a of 3-fluorophenol varies as a function of distance
from N-methylacetamide (a model for the backbone carbonyl
group). When fluorine approaches the oxygen of the carbonyl
group atom, the interaction is indeed repulsive as we had
expected; however, the magnitude of the repulsive interaction
at the X-ray observed distance of ∼3.3 Å is only ∼0.6 kcal/mol.
Thus, there are two factors at play when the ligand adopts the
bound state conformation. The low strain energy of the ligand
along with a favored torsion profile22b (Figure S6B) in the
bound state allows the ligand to bind in a similar conformation
as found in the unbound state and to be able to compensate for
the small unfavorable electrostatic interaction, i.e., ∼0.6 kcal/
mol of the 3-fluoro group with the carbonyl backbone of
Leu704 residue.
Pharmacokinetics. Compound 7 was selected based on its
potent agonist activity in the ARE-LUC assay, minimal activity
in the N/C-interaction assay, and good physicochemical and
ADME properties (cLogP = 2.6; solubility at pH 6.5, 146 μM;
human liver microsomal clearance, 40 μL min−1 mg−1; passive
permeability RRCK cell line Papp(AB) = 29 × 10−6 cm/s; Pgp
influenced efflux in MDR cells, ratio BA/AB = 2). The
pharmacokinetic (PK) studies of 7 in rats (Table 2) reveal that
Table 2. Pharmacokinetic Parameters for 7 in Rats
route
parameter
2 mg/kg, iv
10 mg/kg, ip
10 mg/kg, po
Clp (mL min−1 kg−1
)
35
Vss (L/kg)
0.5
970
0.4
AUC (ng·h/mL)
T1/2 (h)
2492
1
421
1
Cmax (ng/mL)
Tmax (h)
2617
1
110
2
F (%)
51
9
the intraperitoneal (ip) route achieved a robust maximum
concentration, a fast onset of Cmax, and a good availability when
compared to oral (po) or intravenous (iv) PK. Thus, the
decision was made to proceed with proof-of-concept in vivo
efficacy studies via ip route of administration.
In Vivo Studies. Compound 7 was evaluated in 4-day
immature orchidectomized rats that lack endogenous testoster-
one production and are expected to have an exaggerated effect
upon treatment with exogenous androgens. The levator ani
muscle is the most androgen-sensitive muscle in orchid-
ectomized rats. Testosterone propionate (TP) was expected
to have an effect on muscle and reproductive system organs.
After the compound administration the weights of prostate,
seminal vesicles, and levator ani muscle tissues were measured
and compared with the organ weights of vehicle-treated animals
and the testosterone-treated group. The effects of 1, 3, 10, and
30 mg/kg doses of 7 are shown in Figure 2. The ip
administration of 7 showed a statistically significant increase
in the levator ani muscle weight at 10 and 30 mg/kg.
Importantly, 7 did not show increase in weights for the ventral
prostate and seminal vesicles over the vehicle-treated group
except for a small but statistically significant increase in the 30
mg/kg group.
As a positive control, TP promoted anabolic and androgenic
responses in this model and induced significant tissue weight
changes in seminal vesicles, prostate, and levator ani muscle
when administered at 1 mg/kg sc once daily. This study
suggests that at least at 10 mg/kg dose 7 elicits an anabolic
effect on muscle mass while showing no androgenic effects.
Compound 7 demonstrated a therapeutic index with respect to
separation of anabolic and androgenic activities. We also found
that the concentrations of 7 in muscle, prostate, and seminal
vesicles were similar (data not shown), thereby indicating that
the observed tissue selectivity effects are not due to differences
in compound distribution between these tissues.
Hypothalamus−pituitary−gonadal axis, or HPG axis, is the
main regulating mechanism for testosterone level.23 The
anterior portion of the pituitary gland produces lutinizing
D
DOI: 10.1021/acs.jmedchem.7b00373
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
CONCLUSIONS
■
Herein, we have detailed the synthesis and biological activity of
a novel series of tissue-selective androgen receptor modulators
(SARMs), i.e., cyanopyrroles that selectively promote muscle
growth while showing reduced androgenic effects on the
prostate and seminal vesicles. Our SAR efforts using a
combined structure-based and docking approach led to a
putative binding mode of the ligand which allowed further
optimization of this series. We utilized the anabolic ARE-LUC
and the androgenic N/C-terminal interaction in vitro assays to
identify compounds that show tissue selectivity effects in vivo.
By using this strategy, we were able to identify lead compound
7, a potent SARM molecule that is in good physicochemical
property space (i.e., MW ≈ 216; cLogP ≤ 3) and shows a good
separation of anabolic (ARE EC50 = 0.34 nM) and androgenic
activities (N/C-interaction EC50 = 1206 nM). In a 4-day
immature orchidectomized rat in vivo model, 7 significantly
increased levator ani muscle growth at the two highest doses 10
and 30 mg/kg and demonstrated selectivity over seminal vesicle
and prostate tissues at least at one of the doses, i.e., 10 mg/kg.
These data suggest 7 would be a valuable tool molecule for
further evaluation and comparison to other SARMs in clinical
trials. Our ongoing efforts toward these objectives will be
described in future publications. An AR LBD X-ray structure of
7 offered useful insights on ligand strain energy and the nature
of ligand−receptor interactions.
Figure 2. Effects of 1, 3, 10, or 30 mg/kg on ip administration of 7 on
the tissue weights of seminal vesicles, prostate, and levator ani muscle
in orchidectomized male rats ((∗) p < 0.05 vs vehicle). TP was dosed
at 1 mg/kg sc.
hormone (LH) which is a key regulator for testosterone
production in the body. Once levels of testosterone become
elevated, additional testosterone production is suppressed via
LH regulation through a negative feedback effect on the
pituitary gland (Figure 3A). Unlike exogenous androgens that
EXPERIMENTAL SECTION
■
General Experiment. General Procedure for Suzuki Cou-
pling (Scheme 1). All solvents/water were degassed with inert
atmosphere for 30 min prior to use. To a 60 mL sealed tube with stir
bar was added the boronate (1.5 equiv), bromide (1 equiv), Pd(OAc)2
(6.25 mol %), and catacxium A (6.25 mol %). The vial was sealed and
sparged with argon. Degassed dioxane (0.05 M) was added under
argon flow followed by CsF solution (5.1 equiv, 1 M in degassed
water, solution prepared under argon atmosphere). The mixture was
stirred at 80 °C for 1 h at which time the reaction was determined to
be complete. The resulting solution was cooled to room temperature,
filtered through Celite, concentrated under reduced pressure, and
purified by automated preparatory SFC. All compounds used in this
study (Table 1) were chracterized with 1H and 13C NMR
spectroscopy, ESI-MS, and HPLC. The purities of all final compounds
were confirmed to be ≥95% by high-peformance liquid chromotog-
raphy (HPLC). All data for compound characterization are provided in
Supporting Information.
Figure 3. (A) Hypothalamus−pituitary−gonadal (HPG) axis is the
main regulating mechanism for testosterone level. (B) Effects of 1, 3,
10, or 30 mg/kg on ip administration of 7 on luteinizing hormone
concentrations in orchidectomized male rats ((∗) p < 0.05 vs vehicle).
TP was dosed at 1 mg/kg sc.
Preparation of 5-(2-Fluoro-4-hydroxyphenyl)-1-methyl-1H-
pyrrole-2-carbonitrile (7). The title compound was prepared
according to the general procedure using 0.578 mmol of 4-bromo-3-
fluorophenol and 0.607 mmol of 1-methyl-5-(4,4,5,5-tetramethyl-
1,3,2-dioxaborolan-2-yl)-1H-pyrrole-2-carbonitrile. The product was
purified by prep SFC (dC18 5 μm, 4.6 mm × 50 mm, EtOH, 2 mL/
min, retention time 2.7021 min) to give 65 mg of a yellow solid (51%
suppress LH release substantially, an ideal SARM would show a
reduced LH suppression and a minimum effect on the
testosterone production. Compound 7 effect on LH levels
was evaluated in the course of exploratory 14-day toxicology
study. At 10 mg/kg, a decrease in LH levels was observed which
was statistically significant and the 30 mg/kg dose led to a
substantially lower concentration of LH compared to a vehicle-
treated group (Figure 3B). As expected, TP at 1 mg/kg SC
produced significantly reduced LH levels. Therefore, 7 is
recognized as an androgen receptor binder in the brain that
leads to a decrease in endogenous testosterone/LH levels in
rats which may or may not translate to humans. However, our
preclinical models cannot be used to establish a window with
respect to testosterone lowering in humans. Testosterone levels
and regulation are highly species dependent, also with a huge
variability within the same species. Testosterone levels in
animals are highly dependent on social organization.
1
yield). H NMR (500 MHz, DMSO) δ 10.38 (br s, 1H), 7.22−7.27
(m, 1H), 7.01 (d, J = 4.16 Hz, 1H), 6.68−6.74 (m, 2H), 6.23 (d, J =
3.91 Hz, 1H), 3.56 (d, J = 1.22 Hz, 3H). 13C NMR (126 MHz,
DMSO) δ 159.4, 158.6, 158.5, 157.4, 132.4, 130.9, 130.9, 117.8, 112.4,
110.5, 110.5, 108.9, 107.1, 107.0, 102.5, 101.4, 101.2, 31.7, 31.7.
HRMS calcd for C12H10FN2O (M + H)+: 217.0772. Found: 217.0774.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.7b00373.
E
DOI: 10.1021/acs.jmedchem.7b00373
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
(9) Schmidt, A.; Kimmel, D. B.; Bai, C.; Scafonas, A.; Rutledge, S.;
Vogel, R. L.; McElwee-Witmer, S.; Chen, F.; Nantermet, P. V.;
Kasparcova, V.; Leu, C.-T.; Zhang, H.-Z.; Duggan, M. E.; Gentile, M.
A.; Hodor, P.; Pennypacker, B.; Masarachia, P.; Opas, E. E.; Adamski,
S. A.; Cusick, T. E.; Wang, J.; Mitchell, H. J.; Kim, Y.; Prueksaritanont,
T.; Perkins, J. J.; Meissner, R. S.; Hartman, G. D.; Freedman, L. P.;
Harada, S.-i.; Ray, W. J. Discovery of the selective androgen receptor
modulator MK-0773 using a rational development strategy based on
differential transcriptional requirements for androgenic anabolism
versus reproductive physiology. J. Biol. Chem. 2010, 285, 17054−
17064.
Additional information on synthesis procedure and
compound characterization; detailed biological assays;
figures and tables; X-ray coordinates (PDF)
Molecular formula strings and some data (CSV)
Accession Codes
Atomic coordinates for the X-ray structure of 7 can be accessed
using PDB code 5V04 in the RCSB Protein Data Bank (www.
rcsb.org). Authors will release coordinates upon article
publication.
(10) Kilbourne, E.; Moore, W.; Freedman, L.; Nagpal, S. Selective
androgen receptors modulators for frailty and osteoporosis. Curr. Opin.
Invest. Drugs 2007, 8, 821−829.
AUTHOR INFORMATION
Corresponding Authors
*R.U.: phone, +1-617-674-7398; e-mail, ray.unwalla@pfizer.
com.
*E.L.P.C.: phone, +1-718-473-5557; e-mail, piatnits@gmail.
com.
ORCID
Ray Unwalla: 0000-0002-5789-7336
Olugbeminiyi O. Fadeyi: 0000-0002-5525-1304
Notes
■
(11) (a) Schleich, F.; Legros, J. Effects of androgen substitution on
lipid profile in the adult and aging hypogonadal male. Eur. J.
Endocrinol. 2004, 151, 415−424. (b) Saeed, A.; Vaught, G.;
Gavardinas, K.; Matthews, D.; Green, J. E.; Losada, P.; Bullock, H.;
Calvert, N.; Patel, N.; Sweetana, V.; Krishnan, V.; Henck, J.; Luz, J.;
Wang, Y.; Jadhav, J. 2-Chloro-4[[(1R,2R)-2-hydroxy-2-methyl-
cyclopentyl]amino]-3-methyl-benzonitrile: A transdermal selective
androgen receptor modulator (SARM) for muscle atrophy. J. Med.
Chem. 2016, 59, 750−755. (c) Ullrich, T.; Sasmal, S.; Boorgu, V.;
Pasagadi, S.; Cheera, S.; Rajagopalan, S.; Bhumireddy, A.;
Shashikumar, D.; Chelur, S.; Belliappa, C.; Pandit, C.;
Krishnamurthy, N.; Mukherjee, S.; Ramanathan, A.; Ghadiyaram, C.;
Ramachandra, M.; Santos, P.; Lagu, B.; Bock, M.; Perrone, M.; Weiler,
S.; Keller, H. 3-Alkoxy-pyrrolo[1,2-b]pyrazolines as selective androgen
receptor modulators with ideal physiochemical properties for trans-
dermal administration. J. Med. Chem. 2014, 57, 7396−7411.
(12) Chekler, E. L. P.; Unwalla, R.; Khan, T. A.; Tangirala, R. S.;
Johnson, M.; St. Andre, M.; Anderson, J. T.; Kenney, T.; Chiparri, S.;
McNally, C.; Kilbourne, E.; Thompson, C.; Nagpal, S.; Weber, G. L.;
Schelling, S.; Owens, J.; Morris, C. A.; Powell, D.; Verhoest, P. R.;
Gilbert, A. M. 1-(2-Hydroxy-2-methyl-3-phenoxypropanoyl)indoline-
4-carbonitrile derivatives as Potent and tissue selective androgen
receptor modulators. J. Med. Chem. 2014, 57, 2462−2471.
The authors declare no competing financial interest.
ABBREVIATIONS USED
■
SARM, selective androgen receptor modulator; AR, androgen
receptor; LBD, ligand binding domain; SERM, tissue-selective
estrogen receptor modulator; HDL-C, high density lipoprotein
cholesterol; LipE, lipophilic ligand efficiency
REFERENCES
■
(1) Mangelsdorf, D. J.; Thummel, C.; Beato, M.; Herrlich, P.; Schutz,
G.; Umesono, K.; Blumberg, B.; Kastner, P.; Mark, M.; Chambon, P.;
Evans, R. M. The nuclear receptor superfamily: The second decade.
Cell 1995, 83, 835−839.
(2) 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 the skeletal
muscle. J. Clin. Endocrinol. Metab. 2005, 90, 678−688.
(3) Snyder, P.; Peachey, H.; Hannoush, P.; Berlin, J. A.; Loh, L.;
Holmes, J. H.; Dlewati, A.; Staley, J.; Santanna, J.; Kapoor, S. C.; Attie,
M. F.; Haddad, J. G., Jr.; Strom, B. L. Effect of testosterone treatment
on bone mineral density in men over 65 years of age. J. Clin.
Endocrinol. Metab. 1999, 84, 1966−1972.
(13) Ghali, S. A.; Gottlieb, B.; Lumbroso, R.; Beitel, L. K.; Elhaji, Y.;
Wu, J.; Pinsky, L.; Trifiro, M. A. The use of androgen receptor amino/
carboxyl-terminal interaction assays to investigate androgen receptor
gene mutations in subjects with varying degrees of androgen
insensitivity. J. Clin. Endocrinol. Metab. 2003, 88, 2185−2193.
(14) (a) McComas, C.; Cohen, J.; Huselton, C.; Marella, M.;
Melenski, E.; Mugford, C.; Slayden, O.; Winneker, R.; Wrobel, J.;
Yudt, M. R.; Fensome, A. Novel progesterone receptor modulators: 4-
Aryl-phenylsulfonamides. Bioorg. Med. Chem. Lett. 2012, 22, 7119−
7122. (b) McComas, C.; Fensome, A.; Melenski, E. Cyanopyrrole-
phenyl progesterone receptor modulators and uses thereof. U.S. Pat.
Appl. Publ. US 20070027125, 2007. (c) McComas, C.; Fensome, A.;
Melenski, E. Use of progesterone receptor modulators. U.S. Pat. Appl.
Publ. US 20070027201, 2007.
(4) 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.
(15) Bohl, C. E.; Miller, D. D.; Chen, J.; Bell, C. E.; Dalton, J. T.
Structural basis for accommodation of nonsteroidal ligands in the
androgen receptor. J. Biol. Chem. 2005, 280, 37747−37754.
(5) Bagatell, C. J.; Bremner, W. J. Androgens in menuses and
abuses. N. Engl. J. Med. 1996, 334, 707−715.
(6) Bhavnani, B. R.; Woolever, C. A. Interaction of ring B
unsaturated estrogens with estrogen receptors of human endometrium
and rat uterus. Steroids 1991, 56, 201−210.
(16) (a) Maestro; Schrodinger, LLC: New York, NY; http://www.
̈
schrodinger.com/. (b) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.;
Beard, H. S.; Frye, L. L.; Pollard, W. T.; Banks, J. L. Glide: A new
approach for rapid, accurate docking and scoring. 2. Enrichment
factors in database screening. J. Med. Chem. 2004, 47, 1750−1759.
(17) Winneker, R. C.; Fensome, A.; Zhang, P.; Yudt, M. R.;
McComas, C. C.; Unwalla, R. A new generation of progesterone
receptor modulators. Steroids 2008, 73, 689−701.
(7) Nirschl, A. A.; Zou, Y.; Krystek, S. R.; Sutton, J. C.; Simpkins, L.
M.; Lupisella, J. A.; Kuhns, J. E.; Seethala, R.; Golla, R.; Sleph, P. G.;
Beehler, B. C.; Grover, G. J.; Egan, D.; Fura, A.; Vyas, V. P.; Li, Y.-X.;
Sack, J. S.; Kish, K. F.; An, Y.; Bryson, J. A.; Gougoutas, J. Z.; DiMarco,
J.; Zahler, R.; Ostrowski, J.; Hamann, L. G. N-Aryl-oxazolidin-2-imine
muscle selective androgen receptor modulators enhance potency
through pharmacophore reorientation. J. Med. Chem. 2009, 52, 2794−
2798.
(18) Williams, S. P.; Sigler, P. B. Atomic structure of progesterone
complexed with its receptor. Nature 1998, 393, 392−396.
(19) (a) Allen, F. H.; Taylor, R. Research applications of the
Cambridge Structural Database (CSD). Chem. Soc. Rev. 2004, 33,
463−475. (b) Allen, F. H. The Cambridge Structural Data Base: a
quarter of a million crystal structures and rising. Acta Crystallogr., Sect.
(8) Jones, A.; Hwang, D. J.; Duke, C. B., III; He, Y.; Siddam, A.;
Miller, D. D.; Dalton, J. T. Nonsteroidal selective androgen receptor
modulators enhance female sexual motivation. J. Pharmacol. Exp. Ther.
2010, 334, 439−448.
F
DOI: 10.1021/acs.jmedchem.7b00373
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
B: Struct. Sci. 2002, 58, 380−388. CCDC Web site database: http://
www.ccdc.cam.ac.uk/free_services/website/.
(20) (a) Pollock, J.; Borkin, D.; Lund, G.; Purohit, T.; Dyguda-
Kazimierowicz, E.; Grembecka, J.; Cierpicki, T. Rational design of
orthogonal multipolar interactions with fluorine in protein-ligand
complexes. J. Med. Chem. 2015, 58, 7465−7474. (b) Paulini, R.; Miller,
K.; Diederich, F. Orthogonal multipolar interactions in structural
chemistry and biology. Angew. Chem., Int. Ed. 2005, 44, 1788−1805.
(21) Perola, E.; Charifson, P. Conformational analysis of drug-like
molecules bound to proteins: An Extensive Study of Ligand
Reorganization upon Binding. J. Med. Chem. 2004, 47, 2499−2510.
(22) (a) Jaguar, version 9.0; Schrodinger, LLC: New York, NY,
̈
2017. (b) PetaChem, LLC, 26040 Elena Road, Los Altos Hills, CA
94022.
(23) Matsumoto, A. M. Effects of chronic testosterone administration
in normal men: safety and efficacy of high dosage testosterone and
parallel dose-dependent suppression of luteinizing hormone, folli-
clestimulating hormone, and sperm production. J. Clin. Endocrinol.
Metab. 1990, 70, 282−287.
G
DOI: 10.1021/acs.jmedchem.7b00373
J. Med. Chem. XXXX, XXX, XXX−XXX