Discovery and biological characterization of a novel series of androgen receptor modulators

Article abstract of DOI:10.1038/bjp.2008.107

Background and purpose: Selective androgen receptor modulators are of great value in the treatment of prostate cancer. The purpose of this study was to provide a preliminary characterization of a new class of non-steroidal androgen receptor modulators discovered in a high-throughput screening campaign. Experimental approach: Competitive receptor binding, luciferase-based reporter methods, cell proliferation and in vivo assays were employed to evaluate an initial set of compounds from chemistry efforts. Key results: Forty-nine analogues from the chemistry efforts showed high affinity binding to androgen receptors, agonist and/or antagonist activities in both CV-1 and MDA-MB-453 transfection assays. A proliferation assay in LNCaP cells also exhibited this profile. A representative of these non-steroidal compounds (compound 21) was devoid of activity at other nuclear receptors (oestrogen, progesterone, glucocorticoid and mineralocorticoid receptors) in the CV-1 co-transfection assay. At the same time, in an immature castrated rat model, it behaved as an androgen receptor antagonist against the growth of prostate, seminal vesicles and levator ani induced by exogenous androgen. Separation of compound 21 into its enantiomers showed that nearly all the androgen receptor modulating activity and binding resided in the dextrorotatory compound (23) while the laevorotatory isomer (22) possessed weak or little effect depending on the cell type studied. Conclusions and implications: These non-steroidal compounds may represent a new class of androgen receptor modulators for the treatment of not only prostate cancer but other clinical conditions where androgens and androgen receptors are involved in the pathological processes.

Full text of DOI:10.1038/bjp.2008.107

British Journal of Pharmacology (2008) 154, 440–450
2008 Nature Publishing Group All rights reserved 0007–1188/08 30.00
&
$
www.brjpharmacol.org
RESEARCH PAPER
Discovery and biological characterization of a novel
series of androgen receptor modulators
C Zhou1,3, G Wu1,3, Y Feng1, Q Li1, H Su1, DE Mais1, Y Zhu1, N Li1, Y Deng2, D Yang2 and
M-W Wang1
1The National Center for Drug Screening and State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese
Academy of Sciences, Shanghai, China and 2Department of Medicinal Chemistry, West China School of Pharmacy, Sichuan
University, Chengdu, China
Background and purpose: Selective androgen receptor modulators are of great value in the treatment of prostate cancer. The
purpose of this study was to provide a preliminary characterization of a new class of non-steroidal androgen receptor
modulators discovered in a high-throughput screening campaign.
Experimental approach: Competitive receptor binding, luciferase-based reporter methods, cell proliferation and in vivo assays
were employed to evaluate an initial set of compounds from chemistry efforts.
Key results: Forty-nine analogues from the chemistry efforts showed high affinity binding to androgen receptors, agonist and/
or antagonist activities in both CV-1 and MDA-MB-453 transfection assays. A proliferation assay in LNCaP cells also exhibited
this profile. A representative of these non-steroidal compounds (compound 21) was devoid of activity at other nuclear
receptors (oestrogen, progesterone, glucocorticoid and mineralocorticoid receptors) in the CV-1 co-transfection assay. At the
same time, in an immature castrated rat model, it behaved as an androgen receptor antagonist against the growth of prostate,
seminal vesicles and levator ani induced by exogenous androgen. Separation of compound 21 into its enantiomers showed
that nearly all the androgen receptor modulating activity and binding resided in the dextrorotatory compound (23) while the
laevorotatory isomer (22) possessed weak or little effect depending on the cell type studied.
Conclusions and implications: These non-steroidal compounds may represent a new class of androgen receptor modulators
for the treatment of not only prostate cancer but other clinical conditions where androgens and androgen receptors are
involved in the pathological processes.
British Journal of Pharmacology (2008) 154, 440–450; doi:10.1038/bjp.2008.107; published online 14 April 2008
Keywords: nuclear receptor; androgen; prostate cancer
Abbreviations: compound 21, 1-(4-chlorophenyl)-3-(furan-2-yl)-3-(4-nitrophenylamino)-propan-1-one; DHT, 5a-dihydrotes-
tosterone; FBS, fetal bovine serum; MMTV, mouse mammary tumour virus; T, testosterone
Introduction
Prostate cancer, benign prostate hyperplasia, hirsutism,
alopaecia, anorexia nervosa, breast cancer and acne are well
known to be sensitive to androgens and to respond to
androgen receptor antagonist therapy (Liao, 1994; Shaw,
2002; Zouboulis and Degitz, 2004; Gao et al., 2006). Prostate
cancer is currently the most commonly diagnosed non-
dermatologic cancer among male subjects in the United
States. The American Cancer Society estimated that 218 890
men in the United States would be diagnosed with prostate
cancer and 27 050 men would die of it in 2007 (Jemal et al.,
2007). The androgen receptor is a ligand-regulated transcrip-
tion factor in the nuclear receptor superfamily and is a key
molecular target not only for the normal function of the
prostate but also in the growth and progression of prostate
cancer (Tsai and O’Malley, 1994; Zhou et al., 1994).
Testosterone and 5a-dihydrotestosterone (DHT) are the two
predominant naturally occurring androgens that bind to
androgen receptors and subsequently activate androgen-
responsive genes. This activation can be blocked by andro-
gen receptor antagonists through competitive inhibition of
androgen binding to the receptor (Tindall et al., 1984;
Moguilewsky and Bouton, 1988; Singh et al., 2000).
Correspondence: Dr M-W Wang, The Chinese National Center for Drug
Screening, 189 Guo Shou Jing Road, Zhangjiang High-tech Park, Pudong New
District, Shanghai 201203, China.
There are two types of androgen receptor antagonists,
steroidal and non-steroidal. Non-steroidal agents, such as
flutamide (Eulexin) and bicalutamide (Casodex) (Figure 1a;
1 and 2), are preferred, because steroidal antiandrogens, for
example, cyproterone acetate (Figure 1a; 3), in general,
E-mail: wangmw@mail.shcnc.ac.cn
3These authors contributed equally to this work.
Received 16 January 2008; accepted 6 February 2008; published online 14
April 2008

Novel modulators of androgen receptors
C Zhou et al
441
CF3
CF3
a
NC
O2N
O
O
O
O
S
F
Me
Me
N
H
N
H
OH
Me
2
1
Me
Me
O
OAc
Me
CH3
O
N
H
O
H3C
Cl
4
3
R3
R4
OH R2
b
R1
N
H
b
OH
R2
R3
R3
R4
HO
R1
R3
R4
R2
NH2
R3
N
N
O
c
O
R2
a
R2CHO
R4
R1
N
H
R1
N
H
R1
N
H
R4
R3
R4
d
R2
R1
N
H
c
NH2
R3
R4
R3
R4
R2
O
a
b
R2CHO
R2
R1
N
H
N
R4
R3
Figure 1 (a) Structures of various androgen receptor antagonists and an initial high-throughput screening hit. (b) Reagents and conditions: a,
conc. HCl, EtOH, 0ꢀ501C, overnight; b, NaBH4, MeOH, 01C to room temperature, 2 h; c, NH2OH.HCl, pyridine, room temperature, overnight;
d, H2NNH2, KOH, H2O, 140–1601C, 1 h. (c) Reagents and conditions: a, toluene, reflux, 32 h; b, 4-nitroacetophenone (3-nitroacetophenone),
conc. HCl, EtOH, 701C, 30 min.
induce adverse effects owing to crossreactivity at other
steroid hormone receptors (Hamann et al., 1995). Of the
androgen receptor antagonists currently marketed or under-
going clinical trials, none achieves effective therapeutic
results without efficacy-limiting side effects (Haddad,
2006). In addition, severe gynaecomastia, nausea, diarrhoea
and liver toxicity have been observed in many patients, and
the majority of prostate cancer patients receiving antiandro-
gen therapy ultimately become resistant. Indeed, this
resistance is characteristic of what is known as the ‘anti-
androgen withdrawal syndrome (AWS)’ where cessation of
treatment results in a dramatic drop in blood prostate-
specific antigen levels. In this case, resistance manifests itself
as antiandrogens acting as agonists on androgen receptors
(Kelly and Scher, 1993; Kelly et al., 1997; Paul and Breul,
2000). Clearly, there is an unmet medical need for the
treatment of prostate cancer. Thus, we set out to find novel,
non-steroidal androgen receptor antagonists that may lead
to more effective therapies.
androgen receptors will be different from the binding of the
natural hormone such that the structural changes occurring
within the receptor are somewhat different, leading to
altered cofactor recruitment patterns that will be dependent
upon the cell type (Rosen and Negro-Vilar, 2002). As a result,
certain desirable activities will be retained whereas undesir-
able ones (side effects) will be reduced or eliminated. This
approach has been well documented with oestrogen receptor
modulators where undesirable mammary gland effects have
been removed (Swaby et al., 2007). Fulvestrant (Faslodex)
that induces the degradation of oestrogen receptors (Reid
et al., 2003; Buzdar, 2004) is a typical example of such an
endeavour.
There have been a number of recent reports on the
synthesis of novel, non-steroidal androgen receptor agonists
that possessed reduced activities on the prostate and
associated tissues (for example, seminal vesicles) while
maintaining other desirable effects such as bone-building
properties (Higuchi et al., 1999; Kong et al., 2000; Rosen and
Negro-Vilar, 2002; Higuchi et al., 2007; Kearbey et al., 2007).
Thus the precedent to apply what has been learned in the
oestrogen receptor modulators area to androgen receptor
modulators seems secure.
Non-steroidal pure antiandrogens have been used in the
clinic for a number of years, with flutamide and bicaluta-
mide being the best-known examples. Recently, there have
been attempts to apply what has been discovered with
selective oestrogen receptor modulators to the area of
androgen-based therapeutics resulting in selective androgen
receptor modulators (Rosen and Negro-Vilar, 2002; Kearbey
et al., 2007). The idea here is that the binding of a ligand to
In this paper, we report the high-throughput screening
(HTS) of 16 000 synthetic or natural compounds from a
diverse chemical library, the discovery, preliminary structural
modifications and biological characterization, both in vitro
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
442
C Zhou et al
and in vivo, of a novel series of androgen receptor modulators
based on the core structure of the initial HTS hit,
3-(phenylamino)-propan-1-one pharmacaphore (Figure 1a; 4).
presence of Dulbecco’s modified Eagle’s medium supplemented
with 10% charcoal/dextran-treated fetal bovine serum for
24 h before being transfected with Fugene 6. Two micro-
grams of the reporter plasmid (MMTV-Luc) and 0.4 mg of
pSVAR0 were introduced simultaneously into cells and
incubated for 8 h. Cells were then harvested with 0.05%
trypsin and 0.02% EDTA prior to reseeding onto a 96-well
microtitre plate (8000 cells per well). They were incubated
for 24 h with or without various concentrations of control or
test compounds. For antagonist assay, test samples were
added 30 min ahead of DHT. Cell extracts were prepared and
the expressed luciferase activity was determined in a Wallac
1420 multilabel counter (VICTOR2, PerkinElmer) using a
Steady-Glo luciferase kit from Promega (Madison, WI, USA).
To detect potential cytotoxicity of compounds, treated cells
were reacted with AlamarBlue (United States Biological,
Swampscott, MA, USA) for 4 h and the fluorescence is
monitored at 540 nm excitation wavelength and 590 nm
emission wavelength on a FlexStation 384II (Molecular
Devices, Sunnyvale, CA, USA) prior to luciferase activity
measurement.
Methods
Receptor-binding assay
Baculovirus-derived androgen receptor extract protein
(B60 mg Lꢀ1) was loaded into each well of Isoplate (Perkin
Elmer, Boston, MA, USA) containing the assay buffer (10%
glycerol (v/v), NaH2PO4 25 mM, NaMoO4 10 mM, KF 10 mM,
dithiothrietol 1.6 mM, EDTA 1.5 mM, 3-[(3-Cholamidopropyl)
dimethylammonio]-1-propanesulfonate 2 mM, phenylmethane-
sulphonylfluoride 1 mM, aprotinin 5mgLꢀ1 and leupeptin
5 mg Lꢀ1), followed by [3H]DHT (110 Ci mMꢀ1, 5 nM) and
various concentrations of test compounds or DHT (2.5 mL), to
give a final volume of 100 mL per well. The plates were sealed
and incubated overnight at 41C. Hydroxyapatite (HA, 25%
(v/v), 25 mL) was added to each well the next morning and
the plates were gently agitated twice for 5 min each.
Following centrifugation at 2500 r.p.m. for 3 min at 41C,
the supernatant was decanted and 100 mL assay buffer was
added to each well. This washing procedure was repeated
twice before adding 150 mL scintillation liquid (PerkinElmer),
the plates were gently agitated to resuspend HA and radio-
activity was measured by a MicroBeta counter (PerkinElmer).
Nuclear receptor crossreactivity binding assays, including
oestrogen receptor-a, progesterone receptor, glucocorticoid
receptor and mineralocorticoid receptor, were performed as
described previously (Ning et al., 2007).
The human breast carcinoma cell line MDA-MB-453 was
purchased from American Type Culture Collection (ATCC,
Manassas, VA, USA). For the construction of reporter plasmid
stably transfected cell line, 2 mg of MMTV-Luc and 0.2 mg of
pSFFV.neo vector were transfected into 1 Â 105 MDA-MB-453
cells in a 6-cm cell-culture dish. On the second day, the cells
were split into three 100-mm cell-culture plates and
400 mg mLꢀ1 geneticin (Merck KGaA, Darmstadt, Germany)
was added to the cell culture on the third day. The selection
medium (Leibovitz’s L-15) containing 10% FBS in the
presence of 800 mg mLꢀ1 geneticin was changed every 3 days
until colonies were formed. The single colony was picked up
and the best responder selected based on its significant
reactions upon stimulation by different concentrations of
DHT. The stable MDA-MB-453 cell line expressing MMTV-
Luc reporter gene was used to evaluate the androgen receptor
transactivation activity with a similar procedure to that
described above except that cells were seeded onto 96-well
plates at a density of 2 Â 104 cells per well.
High-throughput screening procedure
To apply the HA binding assay to HTS under the optimal
androgen receptor concentration, both maximal binding
(MB; dimethyl sulphoxide (DMSO)) and nonspecific binding
(1 mM DHT) were assessed and the Z0 factor was calculated
according to the literature (Zhang et al., 1999).
The compound library used for screening consists of
16 000 pure synthetic compounds and natural products. A
10-compound pool per well mix was applied to the primary
screening (HA assay), with an average concentration of 8 mM
for each compound dissolved in 100% DMSO solution. This
X–Y matrix system maximizes the advantage of HTS and
allows duplicate screening of each compound for internal
confirmation of a potential hit. In each 96-well Isoplate, 16
wells of rows 1 and 12 were used as negative (DMSO) and
positive (DHT) controls, respectively. Samples showing
greater than 85% inhibition were considered ‘hits’. Positive
compounds were re-screened and confirmed ‘hits’ were
further studied by determining binding dose–response
curves.
In both the CV-1 and MDA-MB-453 cell-based assays, the
antagonist mode was run in the same manner except that
DHT (2 nM for CV-1 cells and 5 nM for MDA-MB-453 cells)
was incorporated into all the wells to establish an agonistic
baseline against which the compounds were tested. For these
assays, the intra-assay coefficient of variation was within
15% and that of inter-assay was below 30%.
Cell proliferation assay
Human prostate carcinoma cell line LNCaP was maintained
in RPMI-1640 medium supplemented with 10% FBS. In vitro
cell proliferation assay was performed as previously described
with minor modifications (Culig et al., 1999; Chen et al.,
2004). Briefly, the medium was changed to phenol red-free
RPMI-1640 with charcoal/dextran-treated FBS 48 h before
assaying. Medium (100 mL) containing 5000 LNCaP cells per
well in a 96-well microtitre plate was incubated at 371C in a
humidified atmosphere of 5% CO2 overnight. Various
concentrations of DHT were added to the medium with or
Reporter gene assay
Transient co-transfection assay was performed in monkey
kidney fibroblast CV-1 cells according to the method
described previously (Ning et al., 2007). Briefly, CV-1 cells
were seeded in 6-cm dish (6 Â 105 cells per dish) in the
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
C Zhou et al
443
without 10 mM test compounds and incubated for 72 h.
Medium containing the same amount of DHT and test
compounds was changed once and incubation was conti-
nued for an additional 72 h. Absorbance was detected with
Cell Counting Kit-8 (Dojindo Laboratories, Kumamoto,
Japan) according to the manufacturer’s instructions. Human
prostatic carcinoma cell line PC3 was maintained in RPMI-
1640 containing 10% FBS. Cells were plated at a density of
5000 per well onto 96-well plates and incubated overnight.
Different concentrations of test compounds were introduced
to the cells and incubated for 3 days before addition of
[3H]thymidine (0.15 mCi per well) followed by incubation for
2 h (Venkateswaran et al., 2002). They were then washed
with ice-cold PBS once, 5% ice-cold trichloroacetic acid
twice, lysed with 0.3 M NaOH (75 mL) for 5 min and
terminated by the addition of 0.3 M HCl (75 mL). Fifty
microlitres of cell lysate was transferred to a scintillation
counting tube and 0.5 mL of Ecoscint A was added to each
tube. Incorporation of [3H]thymidine into DNA was
measured with a MicroBeta counter (PerkinElmer). Mouse
mammary gland tumour cell line SC3 (Minesita and
Yamaguchi, 1965) was maintained in minimum essential
medium supplemented with 10 nM T and 2% FBS. The
medium was changed to serum-free Dulbecco’s modified
Eagle’s medium/F12 (Invitrogen, Carlsbad, CA, USA) med-
ium containing 0.1% BSA before plating onto 96-well plates
at a density of 2500 cells per well. After an overnight
incubation, different concentrations of test compounds were
added and incubated with the cells for 3 days. Thereafter,
cells were labelled with [3H]thymidine (0.15 mCi per well)
and incubated for 2 h. Cell proliferation was assayed as
above. IC50 values were calculated as the concentrations of
compounds required to give half-maximal inhibition on
the cell proliferation. Compound efficacy is expressed as the
compound 21 (200 mg kgꢀ1 dayꢀ1 Â 10). Twenty-four hour
after the last dosing, the rats were killed and the prostate,
seminal vesicles and levator ani were removed. The dry
weights of these tissues were determined and comparisons
made between the groups.
Chemical synthesis and analysis
The preparation of most 3-(phenylamino)-propan-1-one
analogues with the general structure listed in Table 2 is
shown in Figure 1b. The target compounds were conveni-
ently synthesized by Mannich reaction between various
substituted acetones, aldehydes and anilines via acid cata-
lysis (Lin et al., 1991). The total yield for these compounds
was between 61 and 92%. Reduction of the 3-(phenylamino)
-propan-1-one analogues using different conditions
afforded compounds 5, 11, 12, 30 and 49. Reactions between
3-(phenylamino)-propan-1-one analogues and hydroxyla-
mine hydrochloride in the presence of pyridine afforded
the compounds 29, 37 and 39. The designed 1-nitrophenyl-
3-(phenylamino)-propan-1-one analogues were prepared by
reacting the Schiff base derived from aniline with acetophe-
none derivatives under heating (Figure 1c).
Forty-nine analogues based on the 3-(phenylamino)-
propan-1-one pharmacophore (4) were designed and synthe-
sized as above. They were dissolved in DMSO and stored at
ꢀ201C before use. All target compounds possessed purity
superior to 90%. Nuclear magnetic resonance spectra were
obtained on a Varian Mercury-VX300 or 400 MHz apparatus
using CDCl3 or DMSO-d6 solutions as indicated and are
reported in p.p.m. relative to internal solvent signal, with
coupling constants (J) in Hertz (Hz). Splitting patterns are
designated as follows: s, singlet; d, doublet; t, triplet; q,
quartet; m, multiplet; br, broad. Mass spectrometry was
carried out on a PerSeptive Biosystems Mariner. Analytical
HPLC was performed on an Agilent 1100 series with DAD
detector using a Zorbax SB C18 column with an ACN/H2O
solvent. Chiral analysis and separation were performed on
the same Agilent system using a chiral CD-PH column, with
an ACN/H2O solvent. Microanalyses were performed with a
Elementor Vario-EL elemental analyser for C, H, N, and the
results were within 0.4% of the theoretical value (Supple-
mentary information 1).
maximum response produced by
a test compound in
comparison with that elicited by positive controls (100%;
hydroxyflutamide for SC3 cells and taxotere for PC3 cells).
Animals and treatment
Three-week-old male Sprague–Dawley rats were purchased
from Shanghai SLAC Laboratory Animal Co. Ltd (Shanghai,
China) and housed at 221C under a standard light/dark cycle
(12:12) without the restriction of food and water. They were
allowed to adapt to the new environment for at least 3 days.
Castration was carried out with a parallel sham group, which
underwent the surgical procedure without removal of testes.
Following a 6-week recovery period, rats were randomly
divided into six groups (a minimum of eight animals per
group): (1) sham-operated group received vehicle alone
(Sham); (2) castrated group received vehicle alone (control);
(3) castrated animals received s.c. injected testosterone
(0.25 mg kgꢀ1 dayꢀ1 Â 10); (4) castrated animals received s.c.
injected testosterone (0.25 mg kgꢀ1 dayꢀ1 Â 10) and an oral
dose of Casodex (25 mgkgꢀ1 dayꢀ1 Â 10); (5) castrated animals
received s.c. injected testosterone (0.25 mg kgꢀ1 dayꢀ1 Â 10)
and i.p. injected 1-(4-chlorophenyl)-3-(furan-2-yl)-3-(4-
nitrophenylamino)-propan-1-one (compound 21) (100 mg
kgꢀ1 dayꢀ1 Â 10); (6) castrated animals received s.c. injected
testosterone (0.25 mg kgꢀ1 dayꢀ1 Â 10) and i.p. injected
Chiral separation of compound 21
The isolation of two enantiomers of compound 21 was
carried out on an Agilent 1100 HPLC system with a chiral
CD-PH column (4.6 Â 250 mm, 5 mm; Shiseido Fine Chemi-
cals, Tokyo, Japan), using CH3CN-H2O (6:4) as solvents.
Isomer 22 (2 mg) and isomer 23 (2 mg) were obtained by
collecting the eluates at about 20.00 and 20.45 min in
repeated HPLC runs. Values of specific rotation of them
were measured with a PerkinElmer 342 polarimeter and were
ꢀ23 (c 0.1, CHCl3) for isomer 22 and þ 20 (c 0.1, CHCl3) for
isomer 23, respectively.
Statistical analysis
Statistical analysis with Student’s t-test was performed using
GraphPad Prism software (GraphPad, San Diego, CA, USA)
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
444
C Zhou et al
and data are presented as mean s.e.mean. The criterion for
significance was a probability of less than 0.05 or 0.01.
High-throughput screening
Of the 16 000 samples initially screened, 130 ‘hits’ (0.81%)
showed greater than 85% competitive inhibition on DHT
binding to androgen receptor (AR). Secondary (single
Reagents
Dihydrotestosterone was purchased from Sigma Chemical
Co. (St Louis, MO, USA). Casodex was provided by Professor
Lianzhi Chen of China Pharmaceutical University (Nanjing,
China). Dulbecco’s modified Eagle’s medium, RPMI-1640
medium and Leibovitz’s L-15 medium were procured from
Invitrogen Corporation. FBS and charcoal/dextran-treated
FBS were bought from Hyclone (Logan, UT, USA). Fugene 6
transfection reagent was the product of Roche (Indianapolis,
IN, USA) and Steady-Glo luciferase kit of Promega. Alamar-
Blue was obtained from United States Biological. [3H]DHT,
[3H]estradiol and [3H]thymidine were purchased from
PerkinElmer. [3H]progesterone, [3H]dexamethasone and
[3H]aldosterone were supplied by GE Healthcare UK Ltd
(Buckinghamshire, UK). Nuclear receptors of human origin
used in the binding assays were produced according to the
method reported previously (Wu et al., 2005). Human
androgen receptor (pSVAR0) plasmid was kindly provided
by Dr AO Brinnkmann of Erasmus University Medical Center
Rotterdam (Rotterdam, The Netherlands) and a luciferase
reporter plasmid (MMTV-Luc) by Dr Donald McDonnell of
Duke University (Durham, NC, USA). The pSFFV.neo vector
was a gift from Dr RD Ye of University of Illinois (Chicago, IL,
USA).
4000
a
3500
3000
2500
2000
1500
1000
Y = 3832X + 0.0002610
500
R2 = 0.9922
0
0
20
40
60
80
100
Receptor dilution (%)
4000
3500
3000
2500
2000
1500
1000
500
b
Results
Assay validation and HTS parameters
In the present study, we first used various concentrations of
androgen receptor protein preparation in the HA androgen
receptor-binding assay and the optimal protein concentra-
tion was determined as 60 mg Lꢀ1 (1:30 of the stock
solution), which allowed the assay to possess a specific
binding window of 2000 c.p.m. (Figure 2a). Under the above
optimized assay condition, the IC50 value for the natural
androgen receptor ligand, DHT, was measured to be
approximately 4 nM (Figure 2b). When maximal binding
(MB; DMSO) and nonspecific binding (1 mM DHT) were
assessed, the assay displayed a signal-to-background ratio of
5. Coefficient of variation (CV) values were 5.4% for MB and
7.9% for nonspecific binding, respectively. The Z0 factor,
which estimates the suitability to HTS, was calculated to be
0.75 (Figure 2c).
DHT: IC50 = 3.73 nmol/L
Zero
1
10
100
1000
Competing DHT concentration (nM)
4400
4000
3600
3200
2800
2400
2000
1600
1200
800
c
Z' = 0.75
Figure 2 High-throughput screening assay validation. (a) Specific-
binding (CPMMBꢀCPMNSB
) of dihydrotestosterone (DHT) with
MB
NSB
different amounts of androgen receptor. Serial titration of androgen
receptor protein was made to determine the optimal protein
concentration. (b) Androgen receptor-binding characteristics of
DHT measured by the hydroxyapatite (HA) assay with optimal assay
conditions, from which IC50 value was calculated (n ¼ 3, mean
s.e.mean). (c) Z0 factor determination. Assays were performed at the
optimized conditions and 56 replicates of specific and background
signals were studied. Lines indicate means and means 3 Â s.d. of
the 56 data points.
400
0
10
20
30
40
50
60
Number of wells
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
C Zhou et al
445
Table 1 The androgen receptor binding IC50 values of the confirmed
‘hits’
showed comparable potency to that of benchmark, Casodex.
Among these, compound 21 was chosen for further char-
acterization due to its weak agonist (12% efficacy) and
moderate antagonist activities (82% inhibition) in MDA-MB-
453 cells. As this class of analogues can exist as enantiomers,
compound 21 was separated into its D and L isomers and
evaluated independently. Nearly all the androgen receptor-
modulating activity and binding resided in the dextrorota-
tory compound (23) whereas the laevorotatory isomer (22)
possessed weak or little effect depending on the cell type
studied (Figures 3a and b; Tables 2 and 3).
Compound
IC50 (nM)
Compound
IC50 (nM)
SH00011659
SH00011635
SH00012452
SH00012299
SH00110817
SH00106147
SH00010288
1632
1540
2486
542.9
1414
1639
1717
SH00011867
NPLC0024
2959
809.8
889.0
1157
172.7
371.8
221.4
SH00014099
SH00017949
SH00000756
SH00000725
SH00000642
compound per well) screening confirmed that 14 of the
above ‘hits’ displayed consistent inhibitory effects with IC50
values between 150 and 3000 nM (Table 1). Of these,
compound SH00000725 (Figure 1a; 4) was chosen for further
chemistry development due to its consistent AR antagonist
activities in the CV-1 cell assay.
Inhibition of androgen receptor-mediated cell proliferation
SC3 cell is a mouse mammary cell line that responds well to
androgens with increased growth (Minesita and Yamaguchi,
1965). PC3 cell is a human prostate cancer cell line that lacks
measurable androgen receptors and can serve as a control for
specificity and potential cytotoxicity (Venkateswaran et al.,
2002). Table 4 demonstrates that compounds 21 and 26
behaved as androgen receptor antagonists with full efficacy
in SC3 cells whereas their effects on PC3 cell proliferation
were very weak with greatly reduced efficacy, indicating little
or no toxicity. Compound 26 was selected in this assay as,
although structurally similar to all the analogues, it lacked
biological activity in transfection assays and could serve as a
general cytotoxicity control. Figure 3c depicts the effects of
compound 21 and Casodex on androgen-induced prolifera-
tion in LNCaP cells. Like DHT, compound 21 significantly
increased proliferation of these cells but to a lesser extent
(data not shown). In the presence of varying concentrations
of DHT, compound 21, like Casodex, was able to markedly
block the effect of DHT (Po0.05 or 0.01) indicating that it is
acting as a partial AR antagonist in this assay. The effect of
compound 21 on the expression of prostate-specific antigen
in LNCaP cells was also investigated: the compound induced
prostate-specific antigen expression when DHT was absent
but it significantly reduced prostate-specific antigen levels in
the presence of DHT (Supplementary information 2).
AR-binding properties
Forty-nine analogues were designed and synthesized based
on the core structure of the initial hit SH00000725 (4).
Screening results of these compounds are illustrated in
Table 2 showing androgen receptor-binding affinities (IC50
values) and their agonist effects in transfected CV-1 and
MDA-MB-453 cells. Nine analogues demonstrated better
binding potencies than dihydrotestosterone. Clearly, the
best results so far are indicated for R1 ¼ haloaromatic,
R2 ¼ phenyl or one of its isosteres and between R3 and R4,
and R3 substitution with an electronegative group (NO2 for
example). In addition, as seen in Table 2, CNOH substitutes
for the keto moiety did not significantly alter activities.
Among the compounds in Table 2, a number of them were
selected for further biological evaluation based on the
general criteria described below.
Transcriptional activities on AR
Thirty analogues of compound 4 shown in Table 2 possessed
androgen receptor-binding activity (IC50) of less than 200 nM
and/or weak or no agonist activity (efficacy o40%) in CV-1
or MDA-MB-453 cell assays. This suggested that they may be
acting as antagonists. All of them, including compounds
that displayed special structural interest such as the en-
antiomeric pair 22 and 23 and the diastereomeric pair 11 and
12, were subsequently tested in the antagonist mode with
both cell lines and the results are presented in Table 3. DHT
was used as the standard steroidal androgen receptor agonist
and Casodex as a standard non-steroidal androgen receptor
antagonist. Some of these compounds, while exhibiting
weak agonist effects, demonstrated reasonable antagonistic
effects indicating that subtle changes in the structure could
switch between agonist and antagonist activities. For
example, changing the R3 group in compound 7 from NO2
to CN in compound 8 resulted in the total loss of agonist
activity in CV-1 cells. For compounds such as 13, 23, 24, 28,
45 and 46, there was considerable variation in their agonist
activities between these two cell lines (Table 2), but some
(compounds 13, 23 and 24) displayed noticeable antagonist
activities in MDA-MB-453 cells. Of the 30 analogues, eight
Crossreactivity with other nuclear receptors
The crossreactivity of compound 21 with related nuclear
receptors was assessed using human progesterone, oestrogen,
glucocorticoid and mineralocorticoid receptor-binding and
co-transfection assays. Although weak crossreactivity was
detected with progesterone (180.4 nM) and glucocorticoid
(531.7 nM) receptor-binding assays, at concentrations up to
10 mM, no agonist or antagonist activity was observed for
compound 21 except against androgen receptors in the CV-1
co-transfection assays. This result suggests that 3-(phenyla-
mino)-propan-1-one analogues were highly selective for
androgen receptors and were at least to 86-fold more potent
on androgen receptors than any other nuclear receptor in
receptor-binding assays.
Suppression of androgen actions in vivo
Compound 21 was subsequently tested in castrated imma-
ture male rats for its ability to inhibit prostate, seminal
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
446
C Zhou et al
Table 2 Structure–activity relationship at the substitutions of R1, R2, R3 and R4
R3
R2
X
R1
R4
N
H
Compound R1
R2
X
R3
R4
Bindinga
Agonist activity in CV-1b
Agonist activity in MDA-453c
IC50 (nM)
EC50 (nM)
Efficacy (%)
EC50 (nM)
Efficacy (%)
DHT
7.3
0.7
100
1.2
100
4
5
6
7
8
9
4-CH3-phenyl
4-Cl-phenyl
Phenyl
2-furanyl
Phenyl
Phenyl
Phenyl
C ¼ O
CH3
CHOH NO2
H
H
H
H
H
H
H
1640
20.4
32.6
2.6
15.8
14.3
21.6
262.3
334.8
10.8
2.8
NA
NA
1964
21.4
NA
28.8
1176
NA
2238
2857
309.9
59.6
976.2
303.2
14.2
NA
41.7
26.0
286.3
31.1
5132
NA
NA
NA
27
126
NA
9.0
26
NA
11
60
89
119
50
105
143
NA
39
96
39
164
84
NA
NA
38
49
39
17
87
62
76
71
112
43
38
28
91
NA
383.7
955.8
589.4
NA
547.2
NA
NA
NA
23
21
81
NA
10
NA
NA
NA
NA
61
46
28
47
82
4-NO2-phenyl
4-Cl-phenyl
4-Cl-phenyl
4-NO2-phenyl
4-NO2-phenyl
4-NO2-phenyl
4-NO2-phenyl
5-Br-2-thienyl
4-Br-phenyl
4-Cl-phenyl
4-Cl-phenyl
4-Cl-phenyl
4-Cl-phenyl
4-Cl-phenyl
4-Br-phenyl
4-Cl-phenyl
4-Cl-phenyl
4-Cl-phenyl
4-Br-phenyl
4-Cl-phenyl
4-NO2-phenyl
4-F-phenyl
C ¼ O
C ¼ O
C ¼ O
CN
NO2
CN
NO2
NO2
4-Cl-phenyl C ¼ O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
C ¼ O
CHOH OCH3 2-NO2
CHOH OCH3 2-NO2
NA
NA
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
Cl
NO2
H
CF3
H
CF3
NO2
H
CF3
H
H
H
H
H
NO2
NO2
Cl
Cl
Cl
2126
84.4
8335
9700
3812
NA
NA
2591
NA
6753
4043
NA
NA
NA
NA
61.2
NA
675.2
3197
2188
NA
1524
2107
NA
6511
38.5
479.1
NA
NA
323.4
NA
410000
NA
NA
271.8
NA
NA
9.5
19.0
13.9
8.1
71.0
10.3
2.9
56.4
2.1
3.6
173.8
407.9
8.4
3.2
8.8
416.6
15.1
4.1
8.8
2.6
4-Cl-phenyl C ¼ O
NO2
Br
NA
NA
12
NA
34
Phenyl
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
2-furanyl
2-furanyl
2-furanyl
2-furanyl
CH3
Phenyl
Phenyl
2-furanyl
Phenyl
Phenyl
NO2
NO2
NO2
NO2
NO2
F
17
NA
NA
NA
NA
39
NA
54
17
77
NA
80
58
NA
28
68
58
NA
NA
26
NA
70
NA
NA
90
NA
NA
22
H
NA
Cl
Cl
NO2
NO2
H
H
H
H
H
NO2
H
NO2
H
H
H
H
H
H
H
H
H
26.7
21.5
27.6
458.9
74.0
1014
166.4
245.8
101.5
1069
259.3
429.2
19.4
146.1
1800
3198
124.5
1023
120.1
5.2
4-F-phenyl
4-Br-phenyl
4-Cl-phenyl
5-Cl-2-thienyl
4-Cl-phenyl
5-Cl-2-thienyl
5-Br-2-thienyl
5-Cl-2-thienyl
5-Cl-2-thienyl
4-Cl-phenyl
5-Br-2-thienyl
4-Cl-phenyl
4-CH3-phenyl
3-NO2-phenyl
4-Cl-phenyl
4-Cl-phenyl
Me
CNOH NO2
CHOH NO2
Phenyl
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
NO2
NO2
NO2
Cl
NO2
Cl
2-thienyl
2-furanyl
2-thienyl
2-thienyl
2-thienyl
2-furanyl
2-thienyl
2-furanyl
Phenyl
Phenyl
H
Cyclohexyl
Phenyl
Phenyl
5.0
10.2
13.6
7.4
CNOH NO2
C ¼ O NO2
CNOH NO2
9.8
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
C ¼ O
CH2
C ¼ O
C ¼ O
C ¼ O
C ¼ O
NO2
NO2
NO2
NO2
NO2
NO2
Cl
NO2
NO2
NO2
H
14.0
154.9
169.3
73.1
146.9
14.8
7.5
96.6
44.8
NA
128
6
7
48
22
40
48
21
131
NA
NA
15
135
77
2-thienyl
5-Br-2-thienyl
2-furanyl
4-tert-butyl-phenyl Phenyl
NO2
H
H
H
Cl
H
254.6
23.8
NA
2-naphthyl
Phenyl
4-Cl-phenyl
4-NO2-phenyl
4-NO2-phenyl
4-Me-phenyl
4-Cl-phenyl
2-furanyl
2-thienyl
2-thienyl
2-thienyl
2-thienyl
197.1
16.0
12.2
56.5
NA
NO2
NO2
Cl
2168
507.5
4397
3014
4622
13352
H
H
85
28
Abbreviations: DHT, 5a-dihydrotestosterone; NA, not active.
Data shown are mean values from one to three independent experiments in triplicate measurements.
aAndrogen receptor-binding activities of compounds were tested with the hydroxyapatite assay.
bCV-1 cells were co-transfected with androgen receptor (pSVAR0) and a luciferase reporter plasmid (MMTV-Luc), and treated with various concentrations of
compounds. Efficacy is expressed relative to the response induced by 1 mM DHT (100%).
cMDA-MB-453 cells stably transfected with the luciferase reporter plasmid (MMTV-Luc) were treated with various concentrations of compounds. Efficacy is
expressed relative to the response induced by 1 mM DHT (100%).
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
C Zhou et al
447
Table 3 Antagonist activities of compounds in CV-1 co-transfection and
MDA-MB-453 reporter assays
200
175
150
125
100
75
DHT
22
a
Antagonist activity in CV-1a Antagonist activity in MDA-453b
23
Compound
IC50
(nM)
Max. inhibit.
(%)
IC50
(nM)
Max. inhibit.
(%)
50
2
4
5
6
8
9
96 15
1822 582
105 45
NA
1536 53
7819 925
NA
88
47
87
NA
89
59
NA
96
98
2
6
4
181 14
NA
NA
NA
NA
309 118
NA
3060 218
870 109
2481 689
NA
2104 798
716 459
NA
725 103
820 78
NA
572 181
NA
NA
87
NA
4
25
NA
NA
NA
46
NA
72
78 29
35
NA
32
82 11
NA
54
68
NA
59
NA
NA
NA
48
84
38 11
NA
23
34
0
-25
3
7
-6
-4
-2
0
2
4
6
8
Log concentration (nM)
10
11
12
13
19
20
21
22
23
24
26
27
28
29
30
34
37
38
39
41
42
44
45
46
47
51
504 76
177 35
2
1
9
175
150
125
100
75
Casodex
22
b
c
—
—
—
3
229 86
4231 944
—
97
77
2
6
3
NA
—
—
NA
NA
—
—
NA
89
—
77 22
4
7
50
2230 537
—
690 137
91 12
—
57 18
—
110 44
1268 285
566 173
5
5
25
0
86
9
NA
-2
0
2
4
6
—
—
7912 1220
4862 1207
420 72
NA
1340 72
3269 703
NA
8
6
Log concentration (nM)
88
7
Control
30
80
82
67
71
1
6
6
5
4
9
9
21 (10 µM)
Casodex (10 µM)
c
33
21
6
7
NA
150
125
100
75
NA
NA
NA
NA
NA
NA
NA
NA
—
454 174
5894 484
—
93
68
2
3
Abbreviations: DHT, 5a-dihydrotestosterone; Max. inhibit., maximum inhibi-
tion; NA, not active.
50
Data shown are mean s.e.mean from at least two independent experiments
in triplicate measurements.
aCV-1 cells were co-transfected with androgen receptor (pSVAR0) and a
luciferase reporter plasmid (MMTV-Luc), and treated with various concentra-
tions of compounds. DHT (2 nM) was added 30 min later to stimulate
luciferase activity. IC50 values represent concentrations of compounds that
inhibited 50% of the response elicited by 2 nM DHT. Percentage inhibition was
the maximum suppression produced by a test compound compared to the
control response detected in vehicle-treated cells.
0
0.1
1
10
100 1000
Log DHT concentration (nM)
Figure 3 Different bioactivities of compound 21 and its enantio-
mers. (a) CV-1 cells co-transfected with androgen receptor (pSVAR0)
and a luciferase reporter gene plasmid (MMTV-Luc) were treated
with various concentrations of dihydrotestosterone (DHT), com-
pounds 22 or 23 in the agonist mode. (b) Experiment was
performed as above in the antagonist mode, except that DHT
(2 nM) was added 30 min after the addition of compound 22 or
Casodex. (c) LNCaP cells were seeded onto 96-well plates and
incubated overnight. Different concentrations of DHT and test
compounds (10 mM) were added to the cells and incubated for 6
days. Data (mean s.e.mean) are representative of three indepen-
dent experiments.
bMDA-MB-453 cells stably transfected with a luciferase reporter gene plasmid
(MMTV-Luc) were treated with various concentrations of compounds and
incubated for 24 h. DHT (5 nM) was added 30 min after the addition of
compounds. IC50 values represent concentrations of compounds that
inhibited 50% of the response elicited by 5 nM DHT.
cNot determined because of the agonist activity in this cell type.
vesicles and levator ani growth induced by exogenous
testosterone. In the castrated control group, there was a
significant drop in dry weights of all test tissues compared to
the sham group. In the presence of exogenous testosterone,
there was a marked increase in the weight of these three
tissues as would be expected. Casodex (25 mg kgꢀ1) inhibited
this return towards normality as did the two doses (100 and
200 mg kgꢀ1) of compound 21. However, the compound was
not as potent as Casodex (Figure 4). In contrast, compound
26, which possesses neither AR agonist nor antagonist
properties in CV-1 and MDA-MB-453 cells, was inactive in
this model (data not shown).
Discussion
The initial chemical syntheses were primarily guided by a
[3H]DHT competitive receptor-binding assay for assessing the
androgen receptor-binding affinities of the synthetic ligands.
Our preliminary structure–activity relationship (SAR) studies
were focused on modifications at the R3 and R4 substituents
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
448
C Zhou et al
Table 4 Bioactivities in SC3 and PC3 cell proliferation assays
of the phenyl ring, as shown in Table 2. Most modifications,
including replacements with Cl, CN, NO2 or CF3, led to an
increased activity and as the electron-withdrawing capacity
diminished, so did activity (compare 7, 8 and 16). This
indicates that an electron-withdrawing group on the phenyl
ring was crucial for high-affinity binding with the receptor.
Moreover, strong electron-withdrawing groups at R3 position
seemed to promote such an interaction more than any other
groups. This is a common feature of androgen receptor
modulators where strong electron-withdrawing groups have
the greatest effect on activity as seen in the structures of
flutamide and Casodex (Figure 1a).
SC3 antagonist modea
PC3 cytotoxicityb
Compound
IC50 (mM) Inhibition (%) IC50 (mM) Inhibition (%)
Hydroxyflutamide
0.022
—
1.91
0.44
100
—
89.2
109
—
—
Taxotere
21
26
0.001
9.86
5.58
100
51.95
62.38
aSC3 cells were seeded onto 96-well plates and incubated for 3 days with
different concentrations of test compounds in the presence of 10 nM
testosterone. The cells were labelled with [3H]thymidine (0.15 mCi per well)
for 2 h thereafter and the incorporation of radioisotope into DNA was counted
with a MicroBeta counter.
bPC-3 cells were plated onto 96-well plates and incubated with different
concentrations of test compounds for 3 days. Cell growth was assessed with
[3H]thymidine incorporation as above. IC50 values were calculated as the
concentrations of compounds required to give half-maximal inhibition on the
cell proliferation. Efficacy is expressed as the maximum inhibition produced by
a test compound in comparison with that elicited by positive controls (100%;
hydroxyflutamide for SC3 cells and taxotere for PC3 cells). Data shown are
mean values from two independent experiments in triplicate measurements.
Analysis of the three-dimensional structures of compound
21 and Casodex points to a significant similarity between the
two (Figure 5). The only difference is the presence of the
phenyl ring at what one may consider the 11-position on a
steroid nucleus. The stereochemistry at this position is
important as revealed by compounds 22 and 23 where the
two stereoisomers have been separated. The cell-based assay
results suggest that compound 23 possessed both androgen
receptor agonistic and antagonistic activities.
We next turned our attention to modifications of R1, as
shown in Table 2. On the basis of the structure of compound
7, R1 was modified by various replacements based on steric
bulkiness or functionality to define the SAR at this site. The
aromatic or heterocyclic ring was an important requirement
for strong binding affinity, especially for a Cl at the para-
phenyl position. Attempts to reduce or increase the steric
bulkiness of R1 group resulted in diminished agonist
activities (compounds 44 and 47), whereas these two
analogues both exhibited strong androgen receptor antago-
nist profiles in the CV-1 assay. The results indicate that R1
group was important for the antagonistic effects observed.
With the SAR of the R1 group being better understood, we
then investigated the effect of the R2 group. As seen in
Table 2, good androgen receptor-binding activity was found
among the phenyl, furanyl and thiophene analogues.
However, compared with compound 7, the furanyl (21)
and thiophene (32) analogues exhibited less cytotoxicity in
HeLa cells (data not shown). Switching R2 from aromatic and
heterocyclic ring to H, CH3 or cyclohexyl (compounds 42, 25
and 43) resulted in partial loss of activity. These modifica-
Seminal vesicles
100
Prostate
Levator ani
80
60
40
20
0
*
**
*
**
**
*
**
**
**
Sham Ctrl
T
25 mg 100 mg200 mg
T+Casodex T+21
Figure 4 Effects of compound 21 on ventral prostate, seminal
vesicles and levator muscle (dry weight) in castrated immature rats.
Three-week-old male Sprague–Dawley rats were castrated and
allowed to recover for 6 weeks. Rats were randomly divided into
six groups (nX8 per group): sham-operated group received vehicle
alone (Sham); castrated group received vehicle alone (Ctrl);
castrated animals received s.c. testosterone alone (T); castrated
animals received s.c. testosterone and an oral dose of Casodex
(T þ Casodex); castrated animals received s.c. testosterone and i.p.
compound 21 (100 or 200 mg; T þ 21). The dry weights (corrected
for the body weight of each individual animal) of prostate, seminal
vesicles and levator ani were recorded following 10 days of
treatment. *Po0.05 and **Po0.01 compared to testosterone-
treated group (T). Data shown are mean s.e.mean.
tions often led to
a transformation from agonist to
antagonist. This could prove useful as the SAR develops
because the activity profile could be altered with relatively
easy structural changes.
Figure 5 Three-dimensional structures of the two enantiomers of compound 21 (left and middle panel) and Casodex (right panel).
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
C Zhou et al
449
The final group of compounds tested comprised modifica-
tions of carbonyl group (X), as shown in Table 2 along the
X column. The modifications in the X-linkage suggest that
oxime analogues also possess good activity in vitro (comparing
compound 21 to Z and E isomers 37 and 39). Reduction of
the carbonyl to an alcohol led to a drastic decrease in
androgen receptor binding (comparing compounds 7–30
and compounds 5–21). However, such modifications also
increased the antagonistic activities. At this early stage of
SAR, it is premature to guess as to the importance of this
linkage area to the overall androgen receptor-modulating
property.
discovery of analogues that possess greater pharmacological
efficacy and easy isolation or synthetic schemes to produce
enantiomerically pure compounds.
Acknowledgements
We are indebted to Mr Zhen Zhang, Ms Xiaoke Yang, Mr HQ
Zhang, Dr Song Zhang and Mr YY Shu for technical
assistance, and to Dr Guangxing Wang for valuable advice.
This work was partially supported by grants from the
Ministry of Science and Technology of China
(2004CB518902), National Natural Science Foundation
(30400555) and Shanghai Municipality government
(054307057 and 05ZR14031).
There are some discrepancies relating to the potency
values measured by receptor binding compared to those
obtained by transfection (Tables 2 and 3; Figures 3a and b)
and proliferation assays (Table 4; Figure 3c). Ideally, the
results from receptor-binding measurement should accord
with those from cell-based assays. Nevertheless, the former is
a molecular level assay that is more stable and less affected
by experimental conditions. In contrast, the outcome of cell-
based assays can be influenced by many factors. As a result,
discrepancies in potency measurements between these two
assay systems often exist (Gaido et al., 1999; Legler et al.,
1999) and, in the present study, the differences are in general
within the same log concentration range.
Conflict of interest
The authors state no conflict of interest.
References
Buzdar AU (2004). Fulvestrant: a new type of estrogen receptor
antagonist for the treatment of advanced breast cancer. Drugs
Today (Barc) 40: 751–764.
Chen F, Knecht K, Leu C, Rutledge SJ, Scafonas A, Gambone C et al.
(2004). Partial agonist/antagonist properties of androstenedione
and 4-androsten-3b,17b-diol. J Steroid Biochem Mol Biol 91: 247–
257.
Culig Z, Hoffmann J, Erdel M, Eder IE, Hobisch A, Hittmair A et al.
(1999). Switch from antagonist to agonist of the androgen
receptor bicalutamide is associated with prostate tumour progres-
sion in a new model system. Br J Cancer 81: 242–251.
Gaido KW, Leonard LS, Maness SC, Hall JM, McDonnell DP, Saville B
et al. (1999). Differential interaction of the methoxychlor
metabolite 2,2-bis-(p-hydroxyphenyl)-1,1,1-trichloroethane with
estrogen receptors a and b. Endocrinology 140: 5746–5753.
Gao W, Kim J, Dalton JT (2006). Pharmacokinetics and pharmaco-
dynamics of nonsteroidal androgen receptor ligands. Pharm Res
23: 1641–1658.
Towards the end of this preliminary SAR study, we
separated the enantiomers of one of the compounds, 21.
The first set of experiments with this pair of isomers showed
that most if not all of the agonist activity in CV-1 cells
resided in one of the isomers, compound 23, whereas the
other one, 22, possessed weak or little effect depending on
the cell type examined. This same agonist enantiomer also
had the greatest binding affinity of the pair. However, in the
MDA-MB-453 assays, 23 was the only isomer demonstrating
antagonist activity (Table 3). Our data point to the difficulty
inherent in comparing CV-1 results with those obtained
from MDA-MB-453, SC3 and LNCaP cells. The latter three
cell types are utilizing endogenous androgen receptor
present in the cells whereas the former lacks androgen
receptor and must be transfected into the cell. These
observations are important for this series of compounds
and clearly demonstrate that much (if not all) of the
discrepancies observed in the assay panels shown in this
paper resulted from potential mixtures of agonist plus
antagonist activities. At this early stage of the SAR, due to
the difficulty in purifying the two isomers, we did not have
enough quantity of the isolated isomers to re-evaluate in the
in vivo assay and clearly, the SAR is moving towards
optimization.
Haddad
E (2006). Management of gynecomastia induced by
bicalutamide. Ann Urol (Paris) 40 (Suppl 2): S49–S52.
Hamann LG, Farmer LJ, Johnson MG, Goldman ME, Mais DE,
Davtian A et al. (1995). Synthesis and biological activity of novel
nonsteroidal progesterone receptor antagonists. Ann NY Acad Sci
761: 383–387.
Higuchi RI, Arienti KL, Lopez FJ, Mani NS, Mais DE, Caferro TR et al.
(2007). Novel series of potent, nonsteroidal, selective androgen
receptor modulators based on 7H-[1,4]oxazino[3,2-g]quinolin-7-
ones. J Med Chem 50: 2486–2496.
Higuchi RI, Edwards JP, Caferro TR, Ringgenberg JD, Kong JW,
Hamann LG et al. (1999). 4-Alkyl- and 3,4-dialkyl-1,2,3,4-tetra-
hydro-8-pyridono[5,6-g]quinolines: potent, nonsteroidal andro-
gen receptor agonists. Bioorg Med Chem Lett 9: 1335–1340.
Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ (2007). Cancer
statistics, 2007. CA Cancer J Clin 57: 43–66.
Kearbey JD, Gao W, Narayanan R, Fisher SJ, Wu D, Miller DD et al.
(2007). Selective androgen receptor modulator (SARM) treatment
prevents bone loss and reduces body fat in ovariectomized rats.
Pharm Res 24: 328–335.
In summary, we have developed and characterized a novel
series of 3-(phenylamino)-propan-1-one analogues as andro-
gen receptor modulators. These compounds were evaluated
by androgen receptor binding, cell-based reporter and
proliferation assays, and the castrated immature rat experi-
ment, respectively. Preliminary SAR analysis revealed that
some common structural features are required for the in vitro
bioactivities. Several compounds were found to be potent
agonists and/or antagonists of androgen receptors. In
particular, compound 21 was moderately efficacious both
in vitro and in vivo. Further investigations will focus on the
Kelly WK, Scher HI (1993). Prostate specific antigen decline after
antiandrogen withdrawal: the flutamide withdrawal syndrome.
J Urol 149: 607–609.
Kelly WK, Slovin S, Scher HI (1997). Steroid hormone withdrawal
syndromes. Pathophysiology and clinical significance. Urol Clin
North Am 24: 421–431.
British Journal of Pharmacology (2008) 154 440–450

Novel modulators of androgen receptors
450
C Zhou et al
Kong JW, Hamann LG, Ruppar DA, Edwards JP, Marschke KB,
Jones TK (2000). Effects of isosteric pyridone replacements in
androgen receptor antagonists based on 1,2-dihydro- and 1,2,3,
4-tetrahydro-2,2-dimethyl-6-trifluoromethyl-8-pyridono[5,6-g]qui-
nolines. Bioorg Med Chem Lett 10: 411–414.
Legler J, van den Brink CE, Brouwer A, Murk AJ, van der Saag PT,
Vethaak AD et al. (1999). Development of a stably transfected
estrogen receptor-mediated luciferase reporter gene assay in the
human T47D breast cancer cell line. Toxicol Sci 48: 55–66.
Liao S (1994). Androgen action: molecular mechanism and medical
application. J Formos Med Assoc 93: 741–751.
Lin Y, Junhua Z, Huangshu L, Qilin H (1991). The Mannich reaction
of butanone, aromatic aldehydes and aromatic amines. Synthetic
Comm 21: 2109–2117.
Minesita T, Yamaguchi K (1965). An androgen-dependent mouse
mammary tumor. Cancer Res 25: 1168–1175.
Rosen J, Negro-Vilar A (2002). Novel, non-steroidal, selective
androgen receptor modulators (SARMs) with anabolic activity in
bone and muscle and improved safety profile. J Musculoskelet
Neuronal Interact 2: 222–224.
Shaw JC (2002). Hormonal therapies in acne. Expert Opin Pharmaco-
ther 3: 865–874.
Singh SM, Gauthier S, Labrie F (2000). Androgen receptor antagonists
(antiandrogens): structure–activity relationships. Curr Med Chem
7: 211–247.
Swaby RF, Sharma CG, Jordan VC (2007). SERMs for the treatment and
prevention of breast cancer. Rev Endocr Metab Disord 8: 229–239.
Tindall DJ, Chang CH, Lobl TJ, Cunningham GR (1984). Androgen
antagonists in androgen target tissues. Pharmacol Ther 24: 367–400.
Tsai MJ, O’Malley BW (1994). Molecular mechanisms of action of
steroid/thyroid receptor superfamily members. Annu Rev Biochem
63: 451–486.
Moguilewsky M, Bouton MM (1988). How the study of the biological
activities of antiandrogens can be oriented towards the clinic.
J Steroid Biochem 31: 699–710.
Venkateswaran V, Klotz LH, Fleshner NE (2002). Selenium modula-
tion of cell proliferation and cell cycle biomarkers in human
prostate carcinoma cell lines. Cancer Res 62: 2540–2545.
Ning M, Zhou C, Weng J, Zhang S, Chen D, Yang C et al. (2007).
Wu B, Gao J, Wang M-W (2005). Development of a complex
Biological activities of
modulator derived from raloxifene (Y134). Br J Pharmacol 150:
19–28.
Paul R, Breul J (2000). Antiandrogen withdrawal syndrome associated
with prostate cancer therapies: incidence and clinical significance.
Drug Saf 23: 381–390.
Reid G, Hubner MR, Metivier R, Brand H, Denger S, Manu D et al.
(2003). Cyclic, proteasome-mediated turnover of unliganded and
liganded ERa on responsive promoters is an integral feature of
estrogen signaling. Mol Cell 11: 695–707.
a
novel selective oestrogen receptor
scintillation proximity assay for high-throughput screening of
PPARg modulators. Acta Pharmacol Sin 26: 339–344.
Zhang JH, Chung TD, Oldenburg KR (1999). A simple statistical
parameter for use in evaluation and validation of high throughput
screening assays. J Biomol Screen 4: 67–73.
Zhou ZX, Wong CI, Sar M, Wilson EM (1994). The androgen receptor:
an overview. Recent Prog Horm Res 49: 249–274.
Zouboulis CC, Degitz K (2004). Androgen action on human skin -
from basic research to clinical significance. Exp Dermatol 13
(Suppl 4): 5–10.
Supplementary Information accompanies the paper on British Journal of Pharmacology website (http://www.nature.com/bjp)
British Journal of Pharmacology (2008) 154 440–450