Development of androgen receptor ligands by application of ten-vertex para-carborane as a novel hydrophobic core structure

Article abstract of DOI:10.1039/c2md00294a

We report here the design, synthesis, androgenic activity and anti-androgenic activity of novel derivatives of hexadecahedral 1,10-dicarba-closo-decaborane, which is generally called ten-vertex para-carborane. The synthesized compounds exhibited very high binding affinity for the androgen receptor and showed anti-androgenic activity toward the androgen-dependent SC-3 cell line. Two compounds also exhibited partial agonistic activity in SC-3 assay. The docking studies of carborane derivatives to AR demonstrate that ten-vertex carborane is useful for development of novel bioactive compounds as well as twelve-vertex carboranes that are well known to be versatile pharmacophores. Ten-vertex carborane, which has not previously been applied in the field of medicinal chemistry, appears to have potential as a hydrophobic pharmacophore with characteristic properties.

Full text of DOI:10.1039/c2md00294a

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
MedChemComm
Cite this: Med. Chem. Commun., 2012, 3, 680
www.rsc.org/medchemcomm
CONCISE ARTICLE
Development of androgen receptor ligands by application of ten-vertex
para-carborane as a novel hydrophobic core structure†
Shinya Fujii,a Kiminori Ohta,b Tokuhito Goto,b Akifumi Oda,b Hiroyuki Masuno,a Yasuyuki Endob
a
*
and Hiroyuki Kagechika
Received 28th November 2011, Accepted 1st April 2012
DOI: 10.1039/c2md00294a
We report here the design, synthesis, androgenic activity and anti-androgenic activity of novel
derivatives of hexadecahedral 1,10-dicarba-closo-decaborane, which is generally called ten-vertex para-
carborane. The synthesized compounds exhibited very high binding affinity for the androgen receptor
and showed anti-androgenic activity toward the androgen-dependent SC-3 cell line. Two compounds
also exhibited partial agonistic activity in SC-3 assay. The docking studies of carborane derivatives to
AR demonstrate that ten-vertex carborane is useful for development of novel bioactive compounds as
well as twelve-vertex carboranes that are well known to be versatile pharmacophores. Ten-vertex
carborane, which has not previously been applied in the field of medicinal chemistry, appears to have
potential as a hydrophobic pharmacophore with characteristic properties.
demonstrated that the hydrophobic twelve-vertex carboranes
Introduction
can interact effectively with the hydrophobic surface of the
receptor proteins.
Closo boranes ([BnHn]2ꢀ,where n is the number of vertices)1 have
characteristic properties, such as closed three-dimensional
structures and unusual stability, unlike other classes of boranes.
Dicarba-closo-boranes (C2Bnꢀ2Hn),1,2 a class of closo boranes
bearing two carbon atoms in place of boron atoms, are among
the most stable boron clusters because of their electrically neutral
nature and symmetrical, closed structure. Although there has
been extensive investigation of dicarba-closo-boranes as
components of functional molecules and materials, almost all
studies have focused on the twelve-vertex dicarba-closo-dodeca-
boranes (n ¼ 12, C2B10H12; 1a–c) (Fig. 1).3 Twelve-vertex
dicarba-closo-dodecaboranes, which are usually called twelve-
vertex carboranes or more simply carboranes, have a highly
symmetrical icosahedral geometry, and remarkable thermal and
chemical stability. Focusing on the hydrophobic character of
twelve-vertex carboranes, we have investigated their application
as hydrophobic building blocks of biological active molecules.
We have demonstrated that they are versatile hydrophobic core
structures of ligands for nuclear receptors, such as androgen
receptor,4,5 estrogen receptor,6 retinoid receptors,7 and vitamin D
receptor.8 Since hydrophobicity of the molecular surface is
essential for nuclear receptor ligand function, these results
Hundreds of twelve-vertex carborane derivatives have been
synthesized and investigated. However, there have been only
a few investigations of smaller dicarba-closo-boranes9 for liquid
crystal applications10 and electronic materials,11 and they have
not been utilized in the field of medicinal chemistry. We
considered that smaller dicarba-closo-boranes might have
different characteristics as hydrophobic pharmacophores from
twelve-vertex carboranes, and the present study was designed to
test this idea.
Here, we focused on 1,10-dicarba-closo-decaborane (n ¼ 10,
C2B8H10; 2), also called ten-vertex para-carborane. Like
aGraduate School of Biomedical Science, Institute of Biomaterials and
Bioengineering, Tokyo Medical and Dental University, 2-3-10,
Kanda-Surugadai, Chiyoda-ku, Tokyo, 101-0062 Japan. E-mail: kage.
omc@tmd.ac.jp; Fax: +81-3-5280-8127; Tel: +81-3-5280-8032
bFaculty of Pharmaceutical Sciences, Tohoku Pharmaceutical University,
4-4-1, Komatsushima, Aoba-ku, Sendai 981-8558, Japan
Fig. 1 Structures of carboranes. Hashed circles indicate the overall
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2md00294a
geometry of the boron clusters.
680 | Med. Chem. Commun., 2012, 3, 680–684
This journal is ª The Royal Society of Chemistry 2012

View Online
twelve-vertex carboranes, ten-vertex para-carborane has
a hydrophobic surface, and so is expected to have potential to
serve as the hydrophobic core of nuclear receptor ligands. On the
other hand, ten-vertex para-carborane has an overall ellipsoidal
geometry that is quite different from that of the highly
symmetrical twelve-vertex carboranes. Therefore, derivatives of
ten-vertex para-carborane may have activities distinct from those
of the corresponding twelve-vertex carborane derivatives. On the
basis of these considerations, we designed and synthesized
a series of ten-vertex para-carborane derivatives as candidate
nuclear receptor ligands.
Fig. 3 Structures of the designed ten-vertex carborane derivatives (right)
As the biological target, we selected the androgen receptor
(AR).12 AR is a member of the nuclear receptor superfamily of
ligand-regulated transcriptional factors13 specific for testosterone
(3) or dihydrotestosterone (DHT; 4), and regulates numerous
physiological functions, such as maintenance of the male
reproductive system, and bone and muscle homeostasis. AR is
also closely related to prostate cancer, and therefore AR antag-
onists such as flutamide (5) and bicalutamide (7) are used clini-
cally for treatment of prostate cancer (Fig. 2).14 Previously we
reported the development of two types of potent AR antagonists
bearing twelve-vertex carborane as a hydrophobic pharmaco-
phore, i.e., the phenylcarborane derivatives 8–11 (ref. 4) and the
carboranylcyclohexenone 12.5 In these compounds, the spatial
volume of the twelve-vertex carborane cage is essential for
antagonistic activity.15 Therefore, the difference in shape and
bulkiness of 2, compared to 1, may influence the biological
activities. To see whether this is indeed the case, we synthesized
ten-vertex para-carborane derivatives 13–17 (Fig. 3).
and the corresponding twelve-vertex carborane derivatives (left).
We first examined synthetic methods similar to those used for
the twelve-vertex carborane derivatives. The synthetic routes to
the phenyl-ten-vertex para-carborane derivatives 13–16 are
illustrated in Scheme 1. The coupling reaction between C-lithi-
ated 2 and iodobenzene using copper(I) chloride17 gave 1-phenyl-
ten-vertex para-carborane 18, and introduction of a methoxy-
carbonyl group afforded ester 19. Nitration of 19 yielded the
4-nitrophenyl compound 20 as the main product. However, 3-
nitrated derivative 22 was not obtained, and 2-nitrated derivative
21 was obtained as a minor product. In the case of twelve-vertex
carborane, nitration of phenylcarborane yielded the 4-nitro-
phenylcarborane derivative as the major product (ca. 60%) and
the 3-nitrated derivative as a minor product (ca. 20%), while the
2-nitrated compound was not obtained. The difference in
orientation of nitration could be attributed to the difference of
bulkiness between ten-vertex carborane and twelve-vertex car-
borane. Reduction of ester 20 with lithium borohydride gave 14.
The 3-nitrophenyl and cyanophenyl derivatives were synthesized
by another method. Reaction of 2 and paraformaldehyde gave
alcohol 23, and the hydroxyl group of 23 was protected with TBS
to afford 24. A copper(I)-mediated coupling reaction between
C-lithiated 24 and iodobenzenes afforded the corresponding
phenylcarborane derivatives 25–27. Finally, removal of the TBS
group afforded 13, 15 and 16 (Scheme 1).
Results and discussion
The crystal structure of AR bound to testosterone shows that the
3-carbonyl group of steroidal ligands interacts with Gln711 and
Arg752 of AR.16 The cyclohexenone ring of twelve-vertex car-
borane (12) corresponds to the hydrophilic pharmacophore
of testosterone (3). Further, a docking study revealed that the
nitro- and cyano groups of 8–11 interact effectively with the
same residues through hydrogen bonding.15 Thus, we designed
ten-vertex para-carborane derivatives bearing a nitro- or cya-
nophenyl group (13–16) and a cyclohexenone group (17) as
candidate AR ligands based on our previously developed twelve-
vertex carborane-based AR ligands.
Scheme 2 illustrates the synthesis of cyclohexenone 17. Reac-
tion of 2 with 1,4-cyclohexanedione-mono-ethyleneketal gave
tertiary alcohol 28. Deprotection of the ketal, followed by
dehydration and isomerization of the olefinic double bond with
concentrated sulfuric acid, afforded the a,b-unsaturated ketone
29 in one pot. The unsaturated ketone 29 was converted to allyl
alcohol 30 by reduction, and protection of the alcohol gave 31.
Reaction of 31 with n-butyl lithium and paraformaldehyde
afforded alcohol 32, and deprotection of the MOM group with
hydrochloric acid gave diol 33. Finally, selective oxidation of
allyl alcohol using manganese oxide afforded the desired enone
17 (Scheme 2).
The androgenic and anti-androgenic activities of the synthe-
sized molecules were assessed by assay of growth promotion/
inhibition potency toward androgen-dependent SC-3 cells
(Fig. 4).18 Fig. 4A shows the cell growth-promoting activity of
the carborane derivatives. As we had previously reported, none
of the twelve-vertex carborane derivatives 8–12 showed growth-
promoting activity, which means these twelve-vertex carborane
derivatives do not have AR agonistic activity.4,5 On the other
hand, the cyclohexanone derivative of ten-vertex para-carborane
Fig. 2 Structures of endogenous and synthetic AR ligands.
This journal is ª The Royal Society of Chemistry 2012
Med. Chem. Commun., 2012, 3, 680–684 | 681

View Online
Scheme 1 Synthesis of phenyl-ten-vertex para-carborane derivatives.
17 showed significant SC-3 cell growth-promoting activity, and
the 3-cyanophenyl derivative 15 also promoted SC-3 cell growth.
The other compounds did not exhibit agonistic activity. Fig. 4B
shows the antagonistic activity toward SC-3 cell growth
promoted by 1 nM testosterone (3). All the synthesized molecules
exhibited cell growth-inhibitory activity toward SC-3 cells, that
is, these compounds have AR antagonistic activity. The nitro-
phenyl and cyanophenyl derivatives exhibited significant antag-
onistic potency, and the 3-nitro derivative 13 and the 3-cyano
derivative 15 showed the greatest activity, being more potent
than hydroxyflutamide 6, the metabolically activated form of
flutamide 5. The cyclohexenone derivative 17 also exhibited
androgen antagonistic activity. There was no significant differ-
ence in antagonistic potency between ten-vertex para-carborane
derivatives and twelve-vertex carborane derivatives, except
for the agonistic activity of 15 and 17 at the concentration of
1 ꢁ 10ꢀ5 M. These results suggest that ten-vertex para-carborane
derivatives 15 and 17 are partial agonists for AR.
Next, the AR-binding affinity of the ten-vertex para-carborane
derivatives was examined by a competitive binding assay using
3H-labeled dihydrotestosterone ([3H]DHT) and human AR
ligand-binding domain (hAR-LBD),19 and compared with that
of the twelve-vertex derivatives. Fig. 5 shows the dose–response
relationship of cyanophenylcarborane derivatives and cyclo-
hexanone derivatives in competitive binding over the concen-
tration range 1 ꢁ 10ꢀ7 to 1 ꢁ 10ꢀ4 M. The binding potencies of
these ten-vertex derivatives were comparable to those of the
twelve-vertex derivatives, and the most potent ten-vertex deriv-
Fig. 4 SC-3 cell proliferation promotion/inhibition assay. (A) SC-3 cell
proliferation-promoting activity of test compound only. (B) Inhibitory
activity towards testosterone-induced SC-3 cell proliferation. The
concentration of testosterone was 1 nM. Cell number was normalized to
that in the absence of test compounds, taken as 1.
derivatives, the ten-vertex derivative 17 also exhibited higher
affinity for hAR (IC50 ¼ 6.1 ꢁ 10ꢀ6 M) than the twelve-vertex
derivative 12 (IC50 ¼ 9.7 ꢁ 10ꢀ6 M). On the other hand, the
ten-vertex derivative 16 with the 4-cyano group exhibited lower
affinity (IC50 ¼ 1.6 ꢁ 10ꢀ5 M) than the corresponding twelve-
vertex derivative 11 (IC50 ¼ 1.0 ꢁ 10ꢀ5 M). It is interesting that
ative 15 with the 3-cyano group exhibited higher affinity (IC50
7.6 ꢁ 10ꢀ7 M) than the corresponding twelve-vertex derivative 10
¼
(IC50 ¼ 1.9 ꢁ 10ꢀ6 M). With regard to cyclohexanone
Scheme 2 Synthesis of cyclohexenone derivative 17.
This journal is ª The Royal Society of Chemistry 2012
682 | Med. Chem. Commun., 2012, 3, 680–684

View Online
group are essential for full-antagonistic activity of 10.15 The
difference of cluster geometry can cause the different binding
modes of these two compounds, and also induce partial agonistic
activity of compound 15.
Conclusions
We have designed and synthesized derivatives of ten-vertex car-
borane, which has never previously been used in medicinal
chemical applications, and studied their activity as novel AR
ligands in comparison with the corresponding twelve-vertex
carborane derivatives. The synthesized molecules exhibited
potent androgen antagonistic activity, as did the corresponding
twelve-vertex carborane derivatives, but the ten-vertex carbor-
anes showed higher AR binding affinity. Among the synthesized
molecules, compounds 15 and 17 exhibited partial agonistic
activity, in marked contrast to the corresponding twelve-vertex
carborane derivatives. Introduction of the ten-vertex cage in
place of the twelve-vertex cage may reduce steric interference
with the ligand-binding pocket of AR, resulting in partial
agonistic activity. Docking simulation suggested that ten-vertex
carborane functions as a hydrophobic anchor, and that the
characteristic ellipsoidal structure of ten-vertex carborane
induces binding modes of the derivatives different from the
corresponding twelve-vertex carborane derivatives. Thus, we
have shown for the first time that ten-vertex para-carborane can
function as a hydrophobic core structure of biologically active
molecules as well as twelve-vertex carboranes, and may generate
a unique profile of biological activities. Structural development
studies with ten-vertex and twelve-vertex carboranes may be
fruitful approaches to uncover key hydrophobic interactions of
other ligand–receptor combinations.
Fig. 5 Competitive binding assay of synthesized ten-vertex carborane
derivatives and twelve-vertex carborane derivatives. Solid lines indicate
ten-vertex carborane derivatives, and hashed lines indicate twelve-vertex
carborane derivatives. The concentration of [3H]DHT was 4 nM.
compounds 15 and 17, which exhibited partial agonistic activity,
showed higher binding affinity than the corresponding twelve-
vertex carborane derivatives (Fig. 5). These results suggest that
the ten-vertex cage is more favorable as a hydrophobic core of
AR ligands than twelve-vertex carborane for exerting androgen-
dependent transcriptional activity.
We conducted docking studies of carborane derivatives to AR
in order to investigate the difference between the ten-vertex
carborane derivative 15 and the corresponding twelve-vertex
carborane derivative 10. Fig. 6 shows the docking simulation of
these compounds with AR-LBD by using the co-crystal structure
of hAR-LBD with metribolone (PDB:1E3G), by a docking
program GOLD.20 In the calculated binding mode, the ten-vertex
carborane derivative 15 is embedded in the ligand-binding pocket
of AR in a manner similar to that of 10. However, the locations
of cyano groups, that form key hydrogen-bonds with Gln711 and
Arg752, are considerably different from each other. The differ-
ences can be attributed to the shape of the clusters. X-ray crys-
tallographic studies reported that ten-vertex para-carborane has
not only narrower structure than twelve-vertex para-carborane
Acknowledgements
This study was partially supported by Grants-in-Aid for Scien-
tific Research from the Ministry of Education, Culture, Sports,
Science and Technology, Japan (no. 21790110). We also thank
Sasagawa Scientific Research Grant by the Japan Science Society
(no. 20-605).
ꢀ
but also longer C–C distance (3.39 A) than twelve-vertex
21
(3.13 A). We also calculated the molecular volume difference
ꢀ
Notes and references
between compounds 10 and 15; and compound 15 occupies
1 The Borane, Carborane, Carbocation Continuum, ed. J. Casanova,
Wiley Interscience, New York, 1998.
3
3
ꢀ
ꢀ
a smaller volume (239 A ) than compound 10 (263 A ). These
results are consistent with our prediction that the spatial volume
of the twelve-vertex carborane cage and direction of the cyano
2 Carboranes, ed. R. N. Grimes, Academic Press, Elsevier, 2nd ed, 2011.
3 (a) V. I. Bregadze, Chem. Rev., 1992, 92, 209–223; (b)
M. F. Hawthorne, Angew. Chem., Int. Ed. Engl., 1993, 32, 950–984;
(c) A. H. Soloway, W. Tjarks and J. G. Barnum, Chem. Rev., 1998,
98, 1515–1562; (d) F. Issa, M. Kassiou and L. M. Rendina, Chem.
Rev., 2011, 111, 5701–5722.
4 (a) S. Fujii, T. Goto, K. Ohta, Y. Hashimoto, T. Suzuki, S. Ohta and
Y. Endo, J. Med. Chem., 2005, 48, 4654–4662; (b) S. Fujii,
A. Yamada, K. Tomita, M. Nagano, T. Goto, K. Ohta,
T. Harayama, Y. Endo and H. Kagechika, Med. Chem. Commun.,
2011, 2, 877–880.
5 S. Fujii, Y. Hashimoto, T. Suzuki, S. Ohta and Y. Endo, Bioorg. Med.
Chem. Lett., 2005, 15, 227–230.
6 (a) Y. Endo, T. Iijima, Y. Yamakoshi, M. Yamaguchi, H. Fukasawa
and K. Shudo, J. Med. Chem., 1999, 42, 1501–1504; (b) Y. Endo,
T. Iijima, Y. Yamakoshi, H. Fukasawa, C. Miyaura, M. Inada,
A. Kubo and A. Itai, Chem. Biol., 2001, 8, 341–355.
7 (a) T. Iijima, Y. Endo, M. Tsuji, E. Kawachi, H. Kagechika and
K. Shudo, Chem. Pharm. Bull., 1999, 47, 398–404; (b) Y. Endo,
Fig. 6 Docking models of 10 and 15 to hAR-LBD. Compounds 10 and
15 are colored in green and brown, respectively. (A) Overall structure of
AR-LBD bound to ligands. (B) Amino acid residue of hAR-LBD around
the ligands.
This journal is ª The Royal Society of Chemistry 2012
Med. Chem. Commun., 2012, 3, 680–684 | 683

View Online
K. Yaguchi, E. Kawachi and H. Kagechika, Bioorg. Med. Chem.
Lett., 2000, 10, 1733–1736; (c) K. Ohta, T. Iijima, E. Kawachi,
H. Kagechika and Y. Endo, Bioorg. Med. Chem. Lett., 2004, 14,
5913–5918.
8 (a) S. Fujii, H. Masuno, Y. Taoda, A. Kano, A. Wongmayura,
M. Nakabayashi, N. Ito, M. Shimizu, E. Kawachi, T. Hirano,
Y. Endo, A. Tanatani and H. Kagechika, J. Am. Chem. Soc., 2011,
133, 20933–20941; (b) A. Wongmayura, S. Fujii, S. Ito, A. Kano,
Y. Taoda, E. Kawachi, H. Kagechika and A. Tanatani, Bioorg.
Med. Chem. Lett., 2012, 22, 1756–1760.
14 R. K. Ross, M. C. Pike, G. A. Coetzee, J. K. Reichardt, M. C. Yu,
H. Feigelson, F. Z. Stanczyk, L. N. Kolonel and B. E. Henderson,
Cancer Res., 1998, 58, 4497–4504.
15 K. Ohta, T. Goto, S. Fujii, M. Kawahata, A. Oda, S. Ohta,
K. Yamaguchi, S. Hirono and Y. Endo, Bioorg. Med. Chem., 2011,
19, 3540–3548.
16 (a) P. M. Matias, P. Donner, R. Coelho, M. Thomaz, C. Peixoto,
€
S. Macedo, N. Otto, S. Joschko, P. Scholz, A. Wegg, S. Basler,
M. Schafer, U. Egner and M. A. Carrondo, J. Biol. Chem., 2000,
€
275, 26164–26171; (b) J. S. Sack, K. F. Kish, C. Wang,
R. M. Attar, S. E. Kiefer, Y. An, G. Y. Wu, J. E. Scheffler,
M. E. Salvati, S. R. Krystek, Jr, R. Weinmann and
H. M. Einspahr, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 4904–
4909.
9 B. Stibr, Chem. Rev., 1992, 92, 225–250.
10 (a) B. Ringstrand and P. Kaszynski, J. Mater. Chem., 2010, 20, 9613–
9615; (b) P. Kaszynski, S. Pakhomov, K. F. Tesh and V. G. Yong, Jr,
Inorg. Chem., 2001, 40, 6622–6631.
11 S. Pakhomov, P. Kaszynski and V. G. Yong, Jr, Inorg. Chem., 2000,
39, 2243–2245.
12 (a) A. D. Mooradian, J. E. Morley and S. G. Korenman, Endocr.
Rev., 1987, 8, 1–28; (b) C. J. Bagatell and W. J. Bremner, N. Engl.
J. Med., 1996, 334, 707–714; (c) J. M. Kaufman and A. Vermeulen,
Endocr. Rev., 2005, 26, 833–876.
13 (a) R. M. Evans, Science, 1988, 240, 889–894; (b) H. E. MacLean,
G. L. Warne and J. D. Zajac, J. Steroid Biochem. Mol. Biol., 1997, 62,
233–242; (c) I. J. McEwan, Biochem. Soc. Trans., 2000, 28, 369–373; (d)
M. E. Taplin, Expert Rev. Anti-Infect. Ther., 2008, 8, 1495–1508.
17 R. Coult, M. A. Fox, W. R. Gill, P. L. Herbertson, J. A. H. MacBride
and K. Wade, J. Organomet. Chem., 1993, 462, 19–29.
18 D. Hiraoka, N. Nakamura, Y. Nishizawa, N. Uchida, S. Noguchi,
K. Matsumoto and B. Sato, Cancer Res., 1987, 47, 6560–6564.
19 T. Ishioka, A. Kubo, Y. Koiso, K. Nagasawa, A. Itai and
Y. Hashimoto, Bioorg. Med. Chem., 2002, 10, 1555–1566.
20 G. Jones, P. Willett and R. C. Glen, J. Mol. Biol., 1995, 245,
43–53.
21 M. A. Fox, T. E. Baines, D. Albesa-Jove, J. A. K. Howard and
P. J. Low, J. Organomet. Chem., 2006, 691, 3889–3894.
684 | Med. Chem. Commun., 2012, 3, 680–684
This journal is ª The Royal Society of Chemistry 2012