Synthesis and SAR of potent and selective androgen receptor antagonists: 5,6-Dichloro-benzimidazole derivatives

Article abstract of DOI:10.1016/j.bmcl.2006.10.071

The synthesis and in vivo SAR of 5,6-dichloro-benzimidazole derivatives as novel selective androgen receptor antagonists are described. During screening of 2-alkyl benzimidazoles, it was found that a trifluoromethyl group greatly enhances antagonist activity in the prostate. Benzimidazole 1 is a potent AR antagonist in the rat prostate (ID₅₀ = 0.15 mg/day).

Full text of DOI:10.1016/j.bmcl.2006.10.071

Bioorganic & Medicinal Chemistry Letters 17 (2007) 784–788
Synthesis and SAR of potent and selective androgen receptor
antagonists: 5,6-Dichloro-benzimidazole derivatives
Raymond A. Ng, Jihua Guan, Vernon C. Alford,, Jr., James C. Lanter,
George F. Allan, Tifanie Sbriscia, Olivia Linton, Scott G. Lundeen and Zhihua Sui*
Johnson & Johnson Pharmaceutical Research and Development, LLC, Drug Discovery, 665 Stockton Drive, Exton, PA 19341, USA
Received 25 September 2006; revised 23 October 2006; accepted 24 October 2006
Available online 28 October 2006
Abstract—The synthesis and in vivo SAR of 5,6-dichloro-benzimidazole derivatives as novel selective androgen receptor antagonists
are described. During screening of 2-alkyl benzimidazoles, it was found that a trifluoromethyl group greatly enhances antagonist
activity in the prostate. Benzimidazole 1 is a potent AR antagonist in the rat prostate (ID50 = 0.15 mg/day).
Ó 2006 Elsevier Ltd. All rights reserved.
Testosterone, the predominant endogenous ligand to the
androgen receptor1 (AR), is synthesized in the testes and
the adrenal cortex. In some tissues, testosterone is rapid-
ly converted to a higher affinity ligand, dihydrotestoster-
one (DHT). During the early stages of prostate cancer,
androgen activation of AR exacerbates the disease and
stimulates hyperplasia of the prostate.2 The American
Cancer Society estimates that 232,090 men in the U.S.
will be diagnosed with prostate cancer and 30,350 men
will die of prostate cancer in 2005.3 Early treatment of
the disease during the hormone responsive stage with a
non-steroidal androgen receptor (AR) prostate-selective
antagonist benefits patients with a 99.8% 5-year survival
rate.3 Bicalutamide, a non-steroidal antiandrogen, is
currently the predominant antiandrogen used for treat-
ment of androgen-dependent prostate cancer.4
of our studies on the synthesis and pharmacology of
5,6-dichloro-benzimidazole derivatives.
Chart 1
NC
F
O2N
F3C
O
O
O
S
F3C
N
H
N
H
OH
O
OH
hydroxy-flutamide
bicalutamide
Cl
Cl
CF3
OH
N
N
H
1
Initially benzimidazoles 2–5 were prepared by condensa-
tion of phenylene 1,2-diamines with 2-hydroxy-2-meth-
yl-propionic acid under Phillips’ conditions [4 N HCl
(aq), 100 °C].7 The 5-fluoro-6-chloro benzimidazole
showed modest activity (Table 1). The 5,6-dichloro sub-
stitution (3) was superior to 2 and either the 5-chloro-6-
trifluoromethyl (4) or the 5-cyano-6-trifluoromethyl (5)
substitution, reducing prostate weight by 77% in the
orchidectomized immature rat model.8 The 5,6-dichloro
substitution was chosen for further derivatization.
In conjunction with our ongoing interest in androgen
receptor ligands,5 a program was initiated to explore
the use of a benzimidazole scaffold in place of the propi-
onanilide motif found in the bicalutamide and flutamide
structures. Unfortunately, as noted in the literature,
in vitro AR binding does not correlate well with in vivo
efficacy.6 Therefore, compounds were screened in vivo in
immature three-week-old orchidectomized male Spra-
gue–Dawley rats. In this paper, we describe the results
Since flutamide is metabolically oxidized to pharmaco-
logically active 2-hydroxyflutamide,9 des-hydroxy deriv-
atives were explored. Compound 6 showed little effect
on the prostate weight (Table 2), indicating that it
may not be hydroxylated to compound 3 in vivo or
Keywords: Androgen receptor antagonists; Benzimidazoles; Prostate
weight.
*
Corresponding author. Tel.: +1 610 458 6985; fax: +1 610 458
8249; e-mail: zsui@prdus.jnj.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.071

R. A. Ng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 784–788
Table 3. Effect of side chain in 1H-benzimidazoles
785
Table 1. Effect of benzene substitution
X
Y
Cl
Cl
N
N
R
N
H
N
H
OH
OH
P.W. inh. %a
Compound
R
P.W. inh. %a
Compound
X
Y
16
H
Me
na
77
74
67
64
na
29
11
20
16
70
2
F
Cl
Cl
57
77
42
63
70
3
17
3
Cl
Cl
CN
CF3
CH2OH
CH2CN
4
CF3
CF3
18
19
5
Bicalutamide
20
21
CH2S(O)2Me
CH2S(O)2Ph(4-F)
a Prostate weight inhibition
immature Sprague–Dawley rats. Dose = 2 mg/day. Normalized to
control group administered with vehicle (N = 3/group).
% in testosterone treated castrated
22
23
CH2S(O)2Ph(4-Me)
2-Furanyl
2-Thiophene
24
Bicalutamide
a Prostate weight inhibition
%
in testosterone treated castrated
Table 2. Effect of 2-alkyl substitution
immature Sprague–Dawley rats. Dose = 2 mg/day. Normalized to
control group administered with vehicle (N = 3/group) (na, not
active).
Cl
Cl
Z
N
Y
N
R
H
divergent SAR from that of bicalutamide. Substitution
with a heteroaromatic ring, such as 2-furan (23) and
2-thiophene (24), also diminished activity.
Compound
R
Y
Z
P.W. inh. %a
6
H
Me
Et
Me
Me
Me
Et
11
88
32
80
93
na
23
77
82
3
7
H
8
H
n-Pr
Et
Installation of the side chain was carried out according to
Scheme 1. Phillip’s condensation of 4,5-dichloro-1,2-
phenylene diamine with lactic acid afforded alcohol 16,
which was subsequently oxidized with potassium dichro-
mate in dilute sulfuric acid. Addition to the methyl ke-
tone with 3 equivalents of the lithium reagent derived
from the metallation of trimethyl-(2-tributylstannanyl-
methoxy-methoxy-ethyl)-silane afforded the SEM
ether.10 Initial attempt to desilylate with TBAF in THF
failed to remove the SEM group. Deprotection with
LiBF4 in wet CH3CN gave diol 18.11 Compounds 19–
24 were prepared similarly by addition of a lithium or
Grignard organometallic to the methyl ketone.
9
10
H
Me
Me
Me
Me
Et
Me
Et
Me
Me
Me
Et
11
12
n-Pr
Et
13
14
Me
Et
Me
Et
15
Bicalutamide
Et
70
a Prostate weight inhibition
% in testosterone treated castrated
immature Sprague–Dawley rats. Dose = 2 mg/day. Normalized to
control group administered with vehicle (N = 3/group) (na, not
active).
The effect of N-substitution on compound 3 was briefly
explored (Table 4). N-Methyl analog, 25, showed
enhanced activity over 3. Sequential lengthening from
N-methyl to N-butyl decreased activity. Compounds 3
and 25 were 100% orally bioavailable and had a po
half-life of 4 and 3 h, respectively. Interestingly, inser-
tion of an oxygen (29) or sulfur (30) into the alkyl group
gave potent compounds, while methylene nitrile (31)
substitution knocked out activity. Since 29 and 30 may
be sensitive to acid hydrolysis, introduction of sulfide
and sulfone side chains off C2 of the benzimidazole
was carried out on the N-methyl benzimidazole scaffold.
may possess poor pharmacokinetic properties. Exten-
sion of the alkyl substitution by one or two carbons
restored activity as seen with compounds 7 and 9. Inter-
estingly, N-methylation of 6 provided compound 10,
which showed potent activity (93% pros. inh.). Com-
pound 10 was found to reduce prostate weight in a
dose-dependent manner and the ID50 was 0.26 mg/day.
Bicalutamide has similar potency (ID50 = 0.23 mg/d) in
this model. N-Methylation of 7 knocked out activity
completely, while there was no significant difference in
activity between compounds 8 and 9 versus 12 and 13,
respectively. N-Ethylation did not improve activity over
the N-methylated analogs. Only small alkyl chains are
tolerated.
Benzimidazole 3 was prepared similarly as 16, using the
Phillips’ procedure. Selective N-methylation was carried
out with dimethylsulfate in the presence of a phase
transfer catalyst (Scheme 2). The sulfide or sulfone side
chain was introduced by dehydration of alcohol 25 with
Burgess’ inner salt. Hydroxychlorination12 with trichlo-
roisocyanuric acid was followed by displacement with
ethanethiol to afford sulfide 33. Hydroxychlorination
failed on 1H-benzimidazole substrates. Subsequent
oxidation with m-CPBA afforded sulfone 40. With the
Varying the side chain in 1H-benzimidazoles did not
lead to improvements in activity (Table 3). Replacement
of one methyl group in compound 3 with a trifluoro-
methyl group (17) gave comparable activity. However,
17 was found to be toxic at the 3 mg/d dose. Installing
a methylene hydroxyl (18) or a methylene nitrile (19) re-
duced activity. Sulfone analogs, including 21 (bicaluta-
mide side chain), showed little activity, indicating

786
R. A. Ng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 784–788
Cl
Cl
Cl
NH2
NH2
N
b
a
N
H
OH
N
Cl
16
Cl
Cl
Cl
Cl
N
OSEM
c
N
H
O
N
H
OH
Cl
Cl
OH
N
d
N
H
OH
18
Scheme 1. Synthesis of 16 and 18. Reagents and conditions: (a) lactic acid, HCl (aq), 100 °C, 35%; (b) K2CrO7, 25% (v/v) H2SO4, 49% (aq);
(c) Bu4SnCH2OSEM, n-BuLi, THF, 26%; (d) LiBF4, CH3CN, H2O, 29%.
(Table 5). Interestingly, trifluoroethyl thioether 38 abol-
ished activity. Benzyl thioether 37 showed moderate
activity, weaker than 33. In contrast to 33, ethyl ether
32 had low activity. Installation of a 4-fluoro phenyl thi-
oether (39) also did not show significant activity. Since
sulfides can be oxidized in vivo to the corresponding
sulfones (vide supra), sulfone derivatives were synthe-
sized and compared. In general, the sulfones had weaker
activity than the sulfides, with the exception of the sul-
fone derived from oxidation of trifluorothioether, which
showed slightly better activity. N-Ethyl analogs, 34 and
41, had similar activity as 33 and 40. Compound 34 had
low oral bioavailability (15%) and a short half-life (1 h).
Thioether 34 was quickly converted in vivo to the corre-
sponding sulfone 41, which has high bioavailability
(99%) and a longer half-life of 15 h.
Table 4. Effect of N-substitution on 3
Cl
Cl
N
N
R
OH
Compound
R
P.W. inh. %a
3
25
H
77
91
82
53
na
90
96
5
Me
Et
26
27
n-Pr
n-Bu
28
29
CH2OMe
CH2SMe
CH2CN
30
31
Bicalutamide
70
a Prostate weight inhibition
% in testosterone treated castrated
immature Sprague–Dawley rats. Dose = 2 mg/day. Normalized to
control group administered with vehicle (N = 3/group) (na, not
active).
Although compound 17 showed some toxicity, further
investigation revealed that the trifluoromethyl group im-
parts greater activity on prostate weight inhibition (Ta-
ble 6). 2-Ethyl benzimidazole 46, as expected, is inactive.
However, 2-(2,2,2-trifluoro-ethyl)-benzimidazole 47 is a
potent prostate antagonist with an ID50 of 0.097 mg/d
(more potent than bicalutamide) (Table 7). Moreover,
compound 47 has high oral bioavailability with a half-
appropriate sulfide, compounds 33–45 were prepared
similarly.
Ethyl thioether 33 inhibited prostate weight by 86%,
while bulkier alkyl thioethers 35–37 decreased activity
Cl
N
Cl
Cl
N
a
N
H
N
OH
OH
Cl
25
3
Cl
Cl
Cl
Cl
Cl
N
N
N
N
b
d
c
OH
Cl
Cl
Cl
Cl
SEt
S(O)2Et
N
N
N
e
N
OH
OH
33
40
Scheme 2. Synthesis of 25, 33, and 40. Reagents and conditions: (a) Me2SO4, NaOH, CH3CN, benzyltriethylammonium chloride, 52%; (b) Burgess’
inner salt, THF, 68%; (c) trichloroisocyanuric acid, acetone, H2O, 35%; (d) EtSH, NaOMe, MeOH, 97%; (e) m-CPBA, CH2Cl2, 47%.

R. A. Ng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 784–788
787
Table 5. Sulfide and sulfone derivatives
Table 7. ID50s of select compounds
a
ID50 (immature rat)
b
ID50 (mature rat)
Compound
Cl
Cl
N
Y
47
1
(+)-1
0.097
0.15
0.032
0.12
0.17
0.23
18
30
N
R
OH
NT
NT
23
Compound
R
Y
P.W. inh. %a
(À)-1
50
Bicalutamide
32
33
Me
Me
Et
CH2OEt
CH2SEt
CH2SEt
CH2Sn-Pr
CH2Si-Bu
CH2SBn
10
86
79
40
9
30
a In mg/day; tested in testosterone propionate treated (0.1 mg/day sc)
castrated immature Sprague–Dawley rats. Five-day doses = 3, 1, 0.3,
0.1, 0.03 mg/day. Normalized to control group administered with
vehicle (N = 3/group).
b In mg/kg; tested in testosterone propionate treated (0.4 mg/kg) cas-
trated mature Sprague–Dawley rats. Six-week doses = 30, 10, 3, 1,
and 0.3 mg/kg. Normalized to control group administered with
vehicle. (N = 3/group) (NT, not tested).
34
35
Me
Me
Me
Me
Me
Me
Et
36
37
63
6
38
39
CH2SCH2CF3
CH2SPh(4-F)
CH2S(O)2Et
6
40
41
79
74
26
7
CH2S(O)2Et
42
43
Me
Me
Me
Me
CH2S(O)2n-Pr
CH2S(O)2i-Bu
CH2S(O)2Bn
CH2S(O)2CH2CF3
44
45
16
38
70
50 had an ID50 of 18 and 23 mg/kg, respectively, which
are more efficacious than bicalutamide (ID50 = 30 mg/
kg). Racemate 1 had an ID50 of 30 mg/kg, similar to
bicalutamide.
Bicalutamide
a Prostate weight inhibition
immature Sprague–Dawley rats. Dose = 2 mg/day. Normalized to
control group administered with vehicle (N = 3/group).
% in testosterone treated castrated
Selected compounds were tested in a COS AR whole-cell
binding assay. Compound 3 showed 90% inhibition of
binding at 3000 nM (ID50 = 80 nM). Sulfone 41, which
had similar in vivo activity (74% P.W. inh.) as 3 (77%
P.W. inh.), only showed 20% inhibition of binding at
3000 nM. One of the most effective antagonists discov-
ered in this series, 47, also showed weak binding (30%
inhibition of binding at 3000 nM). Similar to other
investigators’ findings, we also do not see a correlation
between in vitro binding and in vivo efficacy.6
Table 6. Trifluoromethyl group increases activity
Cl
X
N
Y
N
R
Z
Cl
Compound
R
X
Y
Z
P.W. inh. %a
46
16
H
H
H
H
H
H
H
H
H
Me
Me
na
na
90
96
81b
80b
60
60
94
74
70
OH
H
47
1
(+)-1
CF3
CF3
We have designed and synthesized AR ligands contain-
ing a benzimidazole core in place of the propionanilide
found in the bicalutamide structure. The 5,6-dichloro
substitution was found to be superior to the 5-cyano-
6-trifluoromethyl substitution pattern found in bicaluta-
mide. Small alkyl groups at N1, such as methyl and eth-
yl, were tolerated, while larger alkyl groups eroded
activity. Installation of sulfide or sulfone side chains
off C2 did not lead to more potent antagonists. The tri-
fluoromethyl group in compounds 1, 47, and 50 impart-
ed good efficacy on the prostate. Moreover, compounds
47 and 50 were found to be more efficacious than bica-
lutamide in mature rats.
OH
(À)-1
48
49
Me
Me
H
H
H
CF3
CF3
CF3
CF3
H
OMe
OH
OH
50
17
OH
Me
H
Bicalutamide
a Prostate weight inhibition
immature Sprague–Dawley rats. Dose = 2 mg/day. Normalized to
% in testosterone treated castrated
control group administered with vehicle (N = 3/group) (na, not
active).
b Dose = 0.3 mg/day.
life of 5.3 h. Compound 47 is converted to compound 1
in vivo. The hydroxyl analog 1 is also very potent
(ID50 = 0.15 mg/d), compared to 16 (inactive). Com-
pound 1 has high po bioavailability (132%) and a half-
life of 14.7 h. N-Methylation and/or O-methylation de-
creased activity. The enantiomers of 1 were chemically
resolved by making the corresponding Mosher esters.
Both (+)-1 and (À)-1 were potent antagonists (80% pros.
wgt. inh. at 0.3 mg/day). It was speculated that the hy-
drate of the corresponding trifluoromethyl ketone may
be active. Oxidation of 1 with TEMPO and bleach gave
ketone hydrate (50), which is a potent AR antagonist on
the prostate (ID50 = 0.17 mg/d). In a 6-week-old intact
mature rat (2-month-old) model, compounds 47 and
Acknowledgments
The authors thank Earl Danser, Mary Evangelisto,
Martin Cousineau, Dr. Sean Peng, and Dr. Dave Ritch-
ie for pharmacokinetic studies. We also thank Drs. Xun
Li and Ron Russell for scale-up of 4-cyano-5-trifluoro-
methyl-1,2-phenylene diamine, and Dr. Xuqing Zhang
for helpful discussions.
References and notes
1. Evans, R. M. Science 1988, 240, 889.
2. Kokontis, J. M.; Liao, S. Vitam. Horm. 1999, 55, 219.

788
R. A. Ng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 784–788
3. Ries, L. A. G; Eisner, M. P.; Kosary, C. L; Hankey, B. F;
D.; Lundeen, S. G.; Sui, Z. Bioorg. Med. Chem. Lett. 2006,
doi:10.1016/j.bmcl.2006.09.086; (e) Lanter, J. C.; Fiordel-
iso, J. J.; Allan, G. F.; Musto, A.; Hahn, D.; Sui, Z.
Bioorg. Med. Chem. Lett. 2006, 16, 5646.
Miller, B. A.; Clegg, L.; Mariotto, A.; Feuer, E. J.;
Edwards, B. K., Eds; SEER Cancer Statistics Review,
1975–2002, National Cancer Institute. Bethesda, MD,
http://seer.cancer.gov/csr/1975_2002/.
6. (a) Yin, D.; He, Y.; Hong, S. S.; Marhefka, C.; Stourman,
N.; Kirkovsky, L.; Miller, D. D.; Dalton, J. T. Mol.
Pharmacol. 2003, 63, 211; (b) Yin, D.; Gao, W.; Kearbey,
J. D.; Xu, H.; Chung, K.; He, Y.; Marhefka, C. A.;
Veverka, K. A.; Miller, D. D.; Dalton, J. T. J. Pharm.
Exp. Ther. 2003, 304, 1334.
7. Phillips, M. A. J. Chem. Soc. Abstracts 1928, 2393.
8. Bicalutamide was used as a positive control, showing 70%
prostate weight inhibition.
9. Schulz, M.; Schmoldt, A.; Donn, F.; Becker, H. Eur. J.
Clin. Pharm. 1988, 34, 633.
10. (a) Kuehne, M. E.; Xu, F. J. Org. Chem. 1998, 63, 9427;
(b) Johnson, C. R.; Medich, J. R. J. Org. Chem. 1988, 53,
4131.
4. For recent reviews of selective androgen receptor modu-
lators (SARMs), see, (a) Mohler, M. L.; Nair, V. A.;
Hwang, D. J.; Rakov, I. M.; Patil, R.; Miller, D. D.
Expert Opin. Ther. Patents 2005, 15, 1565; (b) Zhi, L.;
Martinborough, E. Annu. Rep. Med. Chem. 2001, 36, 169;
(c) Negro-Vilar, A. J. Clin. Endocrinol. Metab. 1999, 84,
3459; (d) Dalton, J. T.; Mukherjee, A.; Zhu, Z.; Kirkov-
sky, L.; Miller, D. D. Biochem. Biophys. Res. Commun.
1998, 244, 1.
5. (a) Zhang, X.; Li, X.; Allan, G.; Musto, A.; Lundeen, S.
G.; Sui, Z. Bioorg. Med. Chem. Lett. 2006, 16, 3233; (b)
Zhang, X.; Li, X.; Allan, G. F.; Sbriscia, T.; Linton, O.;
Lundeen, S. G.; Sui, Z. Bioorg. Med. Chem. Lett. 2006, 16,
5763; (c) Zhang, X.; Li, X.; Allan, G. F.; Sbriscia, T.;
Linton, O.; Lundeen, S. G.; Sui, Z. Bioorg. Med. Chem.
Lett. 2006, doi:10.1016/j.bmcl.2006.10.035; (d) Lanter, J.;
Fiordeliso, J. J.; Jiang, W.; Allan, G. F.; Lai, M.; Hahn,
11. Lipshutz, B. H.; Pegram, J. J.; Morey, M. C. Tetrahedron
Lett. 1981, 22, 4603.
12. Mendonca, G. F.; Sanseverino, A. M.; De Mattos, M. C.
S. Synthesis 2003, 1, 45.