Design, syntheses, and characterization of piperazine based chemokine receptor CCR5 antagonists as anti prostate cancer agents

Article abstract of DOI:10.1016/j.bmcl.2014.03.073

Chemokine receptor CCR5 plays an important role in the pro-inflammatory environment that aids in the proliferation of prostate cancer cells. Previously, a series of CCR5 antagonists containing a piperidine ring core skeleton were designed based upon the proposed CCR5 antagonist pharmacophore from molecular modeling studies. The developed CCR5 antagonists were able to antagonize CCR5 at a micromolar level and inhibit the proliferation of metastatic prostate cancer cell lines. In order to further explore the structure-activity-relationship of the pharmacophore identified, the molecular scaffold was expanded to contain a piperazine ring as the core. A number of compounds that were synthesized showed promising anti prostate cancer activity and reasonable cytotoxicity profiles based on the biological characterization.

Full text of DOI:10.1016/j.bmcl.2014.03.073

Bioorganic & Medicinal Chemistry Letters 24 (2014) 2319–2323
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, syntheses, and characterization of piperazine based
chemokine receptor CCR5 antagonists as anti prostate cancer agents
Christopher K. Arnatt a, Joanna L. Adams a, Zhu Zhang b, Kendra M. Haney a, Guo Li a, Yan Zhang a,c,
⇑
a Department of Medicinal Chemistry, School of Pharmacy, Virginia Commonwealth University, 800 East Leigh Street, Richmond, VA 23298, USA
b Department of Chemistry, College of Pharmacy, Tianjin Medical University, Tianjin 300070, China
c Massey Cancer Center, Virginia Commonwealth University, Richmond, VA 23298, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Chemokine receptor CCR5 plays an important role in the pro-inflammatory environment that aids in the
proliferation of prostate cancer cells. Previously, a series of CCR5 antagonists containing a piperidine ring
core skeleton were designed based upon the proposed CCR5 antagonist pharmacophore from molecular
modeling studies. The developed CCR5 antagonists were able to antagonize CCR5 at a micromolar level
and inhibit the proliferation of metastatic prostate cancer cell lines. In order to further explore the struc-
ture–activity-relationship of the pharmacophore identified, the molecular scaffold was expanded to con-
tain a piperazine ring as the core. A number of compounds that were synthesized showed promising anti
prostate cancer activity and reasonable cytotoxicity profiles based on the biological characterization.
Published by Elsevier Ltd.
Received 13 January 2014
Revised 19 March 2014
Accepted 24 March 2014
Available online 3 April 2014
Keywords:
CCR5
Antagonists
Prostate cancer
Piperazine ring
Besides being a co-receptor for HIV-1 invasion to host cells, che-
mokine receptor CCR5 has been implicated in several types of can-
cer development due to its role in the inflammatory network of
cells.1–8 Prostate cancer is the most prevalent form of cancer found
in American men and is the second only to lung cancer as the most
lethal one in men.9 It has been reported that the expression of the
inflammatory chemokine RANTES (CCL5) correlates with the
growth and survival of prostate cancer cells and mediates these ac-
tions through activation of the G protein-coupled receptor (GPCR)
chemokine receptor CCR5 (CCR5).10–13 Among several CCR5 antag-
onists, TAK-779 (1), maraviroc (2), and our piperidine core contain-
ing lead compound (3), have been shown to inhibit prostate cancer
cell proliferation and/or invasion induced by RANTES.10,14
Previously, the rational design of novel CCR5 antagonists con-
taining the piperidine core skeleton was based on the homology
modeling study of the CCR5 and the conformation analysis of
several well-known CCR5 antagonists.14 It was found that these
well-known CCR5 antagonists all shared a scaffold with a second-
ary or tertiary amino group at the center, then connected to an
amide moiety with an aromatic moiety attached, and on the side
of the scaffold a hydrophobic moiety was linked to the secondary
or tertiary amine.14 From those observations, a novel molecular
skeleton was designed accordingly to fit this proposed pharmaco-
phore. The scaffold included a tri-substituted phenyl ring as the
spacer which connected the amino and amide groups (Fig. 1).
Among all the ligands synthesized compound 3 was identified as
a CCR5 antagonist and inhibited prostate cancer proliferation
in vitro and in vivo.14 However, the therapeutic index of this com-
pound against prostate cancer was hampered by its cytotoxicity.
Therefore, another series of compounds was designed as isoster-
es15,16 of the previously explored ligands, that is, the piperidine
ring was replaced with a piperazine ring system (Fig. 2), in order
to improve their biological activity profile. It was believed that
an additional nitrogen atom in the piperazine ring might allow an-
other possible interaction between the receptor and the ligand to
reinforce the binding of the ligand to the receptor.
The lead compound 3 had an ethyl group as its bulky group and
a phenyl ring as its aromatic group. However, the congruent com-
pound with an isopropyl group as its bulky group and a pyrazinyl
ring for its aromatic group had negligible cytotoxicity and still
inhibited prostate cancer growth.14 Therefore, both the isopropyl
group and pyrazinyl ring were adopted in the new molecular
skeleton and the substituents on the phenyl group were varied
based on the Topliss-tree scheme (Fig. 2).
Scheme 1 showed the 11-step synthetic route applied to
prepare the piperazine-based antagonists 13 through 32 (Fig. 3).
A Williamson ether synthesis was used to alkylate 4-nitrophenol
(4) with 2-bromopropane. This reaction was done in the presence
of K2CO3 in dimethylformamide (DMF). Temperature control
proved to be critical in this reaction; the reaction was kept
constant at 105 °C and after 1 h of reaction, 99% yields of 5 were
regularly achieved.
⇑
Corresponding author. Tel.: +1 804 828 0021; fax: +1 804 828 7625.
E-mail address: yzhang2@vcu.edu (Y. Zhang).
http://dx.doi.org/10.1016/j.bmcl.2014.03.073
0960-894X/Published by Elsevier Ltd.

2320
C. K. Arnatt et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2319–2323
F
F
H
N
O
HN
O
N+
N
N
N
N
H2N
O
Maraviroc
TAK-779
1
2
O
O
H2N
N
H
N
Piperidine based lead
3
Figure 1. Example CCR5 antagonists used as the basis of CCR5 antagonist pharmacophore analysis.
O
O
RBulky
NH
RAromatic
O
N
N
N
H
N
R
R
N
N
N
O
Figure 2. Molecular modeling based pharmacophore analysis, and designed CCR5 antagonist scaffold.
The catalytic hydrogenation of 5 to form the primary amine 6
K2CO3 and a catalytic amount of KI. The reaction yields ranged
from 35% to 94%. Both 31 and 32 had to be prepared via a reductive
amination and were synthesized with reasonable yields.
This series of newly synthesized compounds were first tested
for their CCR5 agonism and antagonism in a calcium mobilization
assay using MOLT-4 cells transfected with CCR5 (NIH AIDS Re-
search and Reference Reagent Program).18 Using the calcium sens-
ing dye Fluo-4, compounds 13 through 32 were first tested for their
was done with 10% palladium on carbon (Pd/C) under hydrogen
gas and 1.3 equiv of concentrated HCl, with yields up to 98% and
reaction times ranged from 1 to 2 h for up to 5 g of starting mate-
rial. The cyclization of 6 to 7 to form the piperazine ring proved to
be a very difficult and fastidious reaction. This reaction required
temperatures above 130 °C, but below 150 °C in order to get 7 in
appreciable yields. Immediately after workup, compound 7 was
protected with trifluoroacetic anhydride to form 8 with consistent
yields around 94%.
CCR5 agonism and none of them showed any agonism up to 30 lM.
The compounds were then tested for their antagonism of RANTES-
stimulated calcium release. Table 1 showed the results of three
independent assays each done in triplicate.
All of the compounds antagonized the RANTES stimulated cal-
cium flux on the CCR5 at micromolar levels. These compounds’
antagonism were relatively low compared to Maraviroc and
TAK-779,19 but comparable to our original series of designed
ligands.14 Some of them actually acted more potently than the lead
compound 3. Among them, compound 13 showed an IC50 value of
Mono-nitration of 8 to introduce the meta-nitro group was not
obtainable through normal aromatic nitration. Using several
different reaction conditions with acetic acid or acetic anhydride
and nitric acid, only di-meta nitration product was obtained. It
was speculated that due to the presence of both the 1-propyloxy
group and the 4-piperazine group with electron donating capabil-
ities, the aromatic ring was highly activated. This in turn made
mono-nitration unlikely with conventional synthetic routes.
Therefore, a nitrocyclohexadienone as a mild nitrating reagent
was adopted successfully. This reagent has previously been shown
to mono-nitrate several activated substrates without leading to the
usual oxidative byproducts of normal aromatic nitration.17
Subsequent reduction of the nitro group of 9 to the amine, 10,
was done using Pd/C catalyzed hydrogenation with yields around
98% without any complications. The amide-coupling of 10 with
pyrazine-2-carboxylic acid to form 11 was done using 1-ethyl-
3-(3-dimethylaminopropyl) carbodiimide (EDCI) with yields up
to 99%. Deprotection of the piperazine moiety was done by reflux-
ing 12 under basic conditions in a methanol/water mixture (1:1).
Yields of 12 for this reaction were quantitative. Compounds 13
through 32 all used the key intermediate 12, which allowed for fi-
nal compounds to be synthesized efficiently. For compounds 13
through 30 the benzyl chloride was coupled with the secondary
amine of the piperazine group. This reaction was done in DMF with
10.3 4.2 lM. Using this unsubstituted phenyl compound as a ref-
erence several observations can be summarized on the structure
activity relationship of these newly synthesized ligands. As seen
for compounds 13 through 25, there was a general trend that elec-
tron withdrawing substituents led to lower IC50 values than elec-
tron donating groups with some exceptions, for example,
compounds 19 and 20. On another note, the substitution position
seemed to play a key role in antagonism as seen for the series of
nitro-substituted compounds. para substitution (14) yields the
lowest IC50 value of 6.45 1.7
(16) is also well tolerated with a slightly higher IC50 value of
16.8 6.2 M. However, meta and di-meta substituted nitro-
groups (15 and 17) did not show any CCR5 antagonism below
50 M. Overall, CCR5 calcium signaling inhibition was optimized
lM, while as ortho substitution
l
l
with para substituted electron withdrawing groups. Further
exploration of the bulkiness of para-substitutions did not show

C. K. Arnatt et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2319–2323
2321
Cl
OH
O
O
O
N
Br
HN Cl
10% Pd/C
H2, HCL
MeOH
K2CO3
Chlorobenzene
Reflux
K2CO3
DMF
NO2
NH2
NO2
105 oC
.HCl
4
5
N
H
6
7
O
Br
Br
Br
O
O
O
Br
NO2
NH2
(CF3CO)2O
H3C NO2
Py
10% Pd/C
THF
R.T.
CH2Cl2
H2
N
N
N
N
N
N
M.S.
MeOH
0 oC to R.T.
F3C
O
F3C
O
F3C
O
8
9
10
Cl
R
13 - 30
O
KI, K2CO3
N
N
N
N
N
N
OH
O
O
DMF
R.T.
H
N
H
N
O
O
K2CO3
MeOH/H2O
Reflux
EDCI, HOBT
TEA, DMF, M.S.
0 oC to R.T.
N
N
N
R
O
31, 32
N
H
NaBH(OAc)3
THF, M.S.
R.T.
F3C
11
O
12
Figure 3. Synthetic route for CCR5 antagonists 13–32.
Table 1
any significant trend though it seemed that larger moiety seemed
to be less favorable for the CCR5 antagonism of these compounds.
An anti-proliferation assay using the colorimetric reagent, WST-
1, was applied to test compounds against two prostate cancer cell
lines, PC-3, and M12. They were chosen based upon their high
expression of CCR5 and RANTES.20 M12 cells were originally iso-
lated from the prostate gland and selected for more metastatic
cells from tumors in mice. PC-3 cells were obtained from a meta-
static prostate tumor obtained from a lumbar vertebra.21
CCR5 antagonism (calcium mobilization) of compounds 14 through 35 using RANTES
as the agonist
Compound #
Substitution
IC50 (lM)
3
N/A
H
4-NO2
3-NO2
2-NO2
3,5-NO2
4-Cl
4-CN
4-F
63.0 26.0
10.3 4.2
6.45 1.7
>50*
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
16.8 6.2
>50*
6.29 2.4
>50*
Table 2 showed the anti-proliferation data in both prostate can-
cer cell lines for compounds with reasonable CCR5 antagonism.
When tested against PC-3 cells, the unsubstituted compound 13
>50*
4-Br
23 18
had an IC50 value of 67 15 lM; further substitution has a drastic
4-CO2CH3
4-COOH
4-SO3CH3
4-SO2CH3
4-CH3
9.8 7.1
41.2 2.9
>50*
45.8 9.1
45.9 26.7
27 14
19.8 2.9
5.1 3.4
25.6 2.1
28.1 4.0
48.0 2.9
effect on the activity of the compounds. First, unlike what is seen
for calcium antagonism, there is no significant effect on anti-prolif-
eration activity for substitution position, as seen in the series of ni-
tro group substituted compounds. Second, electron donating
groups are generally more favored than electron withdrawing
groups, and for example, the diethylamino substituted compound
4-NH2
4-OCH3
4-SCH3
4-NH2COCH3
4-N(CH3)2
4-N(CH2CH3)2
32 had the lowest IC50 of 6.5 0.7
showed the lowest IC50 value of 11.4 0.2
l
M. For M12 cells, 32 also
M. Overall, the same
l
trend of electron donating groups having higher anti-proliferation
activity was seen for both cell lines. Compound 32 seemed to carry
improved anti-proliferation activity against PC-3 and M12 when
*
(>50) Denotes that the IC50 was above 50
l
M and was not tested at higher
concentrations due to limited solubility in the specific testing media.

2322
C. K. Arnatt et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2319–2323
Table 2
Table 3
Anti-proliferation assays for PC-3 and M12 prostate cancer cells using WST-1 to
measure cell proliferation
Basal cytotoxicity assays using NRU and WST-1 to test for exogenous toxicity of
compounds 14 through 32 in NIH 3T3 cells
Compound #
Substitution
PC-3 IC50
(
lM)
M12 IC50
(
lM)
Compound #
Substitution
NIH 3T3 (NRU) TC50 (lM)
TAK-779 (2)
3
N/A
N/A
H
4-NO2
2-NO2
4-Cl
4-Br
4-CO2CH3
4-COOH
4-SO2CH3
4-CH3
37.85 0.99
43.0 2.3
67 15
20.40 1.10
31.9 0.42
29.9 4.3
78.7 4.0
183 2.5
26.3 0.8
48.8 25
73.7 20
129 19
141 26
15.8 4.4
19.7 5.3
>100*
3
N/A
H
4-NO2
2-NO2
4-Cl
4-Br
4-NH2
4-NH2COCH3
4-N(CH3)2
4-N(CH2CH3)2
84.0 11.0
46.6 3.2
29.3 2.3
>50*
13
14
16
18
21
27
30
31
32
13
14
16
18
21
22
23
25
26
27
29
30
31
32
49.1 6.6
62.9 9.5
56.1 9.6
31.5 1.4
91.5 3.7
74 12
1.6 0.8
>50*
10.7 1.2
7.8 1.1
>50*
>100*
31.9 1.6
19.7 1.8
20.1 1.3
78.0 19
24.3 1.6
71.2 1.6
6.5 0.7
*
(>50) Denotes that the IC50 was above 50
centrations due to limited solubility in the specific testing media.
lM and was not tested at higher con-
4-NH2
4-SCH3
4-NH2COCH3
4-N(CH3)2
4-N(CH2CH3)2
39 12
65.9 2.9
11.4 0.2
31.9 1.6 lM. Further structural modification of this lead will be
conducted in order to improve its anti cancer activity profile.
*
(>100) Denotes that the IC50 was above 100 lM and was not tested at higher
concentrations due to limited solubility in the specific testing media.
Acknowledgments
We are grateful for the partial funding support from US Army
Prostate Cancer Research Program PC073739 and A. D. Williams
MultiSchool Research Funds at Virginia Commonwealth University.
The content of this report is solely the responsibility of the authors
and does not necessarily represent the official views of the US
Army Prostate Cancer Research Program. We thank the NIH AIDS
Research and Reference Reagent Program for providing the
MOLT-4/CCR5 cell line.
compared to TAK-779 (2) which was shown to have IC50’s of
20.40 1.1
(3) which was shown to have IC50’s of 31.9 0.42
43.0 2.3
M (PC-3), respectively.22 In the view of correlating the
lM (M12) and 37.85 0.99
l
M (PC-3), and to the lead
lM (M12) and
l
ligands’ shown anti prostate cancer activity and their CCR5 antag-
onism activity we did not observe a significant correlation, proba-
bly due to our limited size of compound pool. Another possible
explanation is that this could be related to the functional selectiv-
ity of our ligands on the target receptor CCR5.
To ensure the anti-proliferation results are not due to the toxic-
ity of the compounds, a basal cytotoxicity assay was conducted for
the more promising ligands from the anti-proliferation assays. The
NIH-3T3 cells used for the assay are mouse fibroblasts that have
been used extensively along with neutral red uptake (NRU) to as-
sess basal cytotoxicity levels of small molecules.23,24 In all, several
of the compounds (e.g., compound 27) that showed high anti-pro-
liferative activity in both M12 and PC-3 were also cytotoxic in NIH-
3T3 cells, which indicates their observed anti-proliferative activity
may be related to their cytotoxicity (Table 3). On the other hand,
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.03.
073.
References and notes
1. Chen, W.; Zhan, P.; DeClercq, E.; Liu, X. Curr. Pharm. Des. 2012, 18, 100.
2. Dorr, P.; Westby, M.; Dobbs, S.; Griffin, P.; Irvine, B.; Macartney, M.; Mori, G.;
Rickett, G.; Smith-Burchnell, C.; Napier, C.; Webster, R.; Armour, D.; Price, D.;
Stammen, B.; Wood, A.; Perros, M. Antimicrob. Agents Chemother. 2005, 49,
4721.
compound 32 showed a TC50 of 31.9 1.6 lM which indicated a
somehow promising toxicity profile for further exploration though
the TC50 value was relatively lower than the lead compound 3.
Accumulating evidence has shown the multiple roles that che-
mokine receptor CCR5 plays to promote the progression of several
types of cancer. The mechanism of action of this promotion is
thought to involve chronic inflammation, which creates a microen-
vironment that enhances tumor survival. Blocking CCR5 function
with an antagonist may provide a novel treatment of cancers such
as prostate cancer.25–28 Currently, several CCR5 antagonists are
available, but all have been optimized for their anti-HIV entry inhi-
bition rather than inhibition in endogenous signaling. Thus, there
is need to develop antagonists focused on blocking CCR5 signaling,
and inhibiting CCR5 related prostate cancer proliferation. Using a
combination of pharmacophore analysis and CCR5 docking studies,
a unique CCR5 antagonist skeleton was designed and functional-
ized at multiple positions to optimize its activity. A combination
of calcium inhibition, anti-proliferation, and basal cytotoxicity
assays were used to screen for active compounds. In the CCR5
calcium mobilization inhibition assays all of the synthesized com-
pounds acted as antagonists. From the anti-proliferation assays
and the basal cytotoxicity assays, compound 32 was identified as
3. Wood, A.; Armour, D. Prog. Med. Chem. 2005, 43, 239.
4. Tagat, J. R.; McCombie, S. W.; Nazareno, D.; Labroli, M. A.; Xiao, Y.; Steensma, R.
W.; Strizki, J. M.; Baroudy, B. M.; Cox, K.; Lachowicz, J.; Varty, G.; Watkins, R. J.
Med. Chem. 2004, 47, 2405.
5. Strizki, J. M.; Tremblay, C.; Xu, S.; Wojcik, L.; Wagner, N.; Gonsiorek, W.; Hipkin,
R. W.; Chou, C. C.; Pugliese-Sivo, C.; Xiao, Y.; Tagat, J. R.; Cox, K.; Priestley, T.;
Sorota, S.; Huang, W.; Hirsch, M.; Reyes, G. R.; Haroudy, B. M. Antimicrob. Agents
Chemother. 2005, 49, 4911.
6. Palani, A.; Tagat, J. J. Med. Chem. 2006, 49, 2851.
7. Baba, M.; Nishimura, O.; Kanzaki, N.; Okamoto, M.; Sawada, H.; Iizawa, Y.;
Shiraishi, M.; Aramaki, Y.; Okonogi, K.; Ogawa, Y.; Meguro, K.; Fujino, M. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 5698.
8. Maeda, K.; Nakata, H.; Koh, Y.; Miyakawa, T.; Ogata, H.; Takaoka, Y.; Shibayama,
S.; Sagawa, K.; Fukushima, D.; Moravek, J.; Koyanagi, Y.; Mitsuya, H. J. J. Virol.
2004, 78, 8654.
9. American Cancer Society. http://www.cancer.org/research/cancerfactsstatistics/
index (accessed Nov 9, 2012).
10. Vaday, G. G.; Peehl, D. M.; Kadam, P. A.; Lawrence, D. M. Prostate 2005, 66, 124.
11. Coussens, L. M.; Werb, Z. Nature 2002, 420, 860.
12. Robinson, S. C.; Scott, K. A.; Wilson, J. L.; Thompson, R. G.; Proudfoot, A. E. I.;
Balkwill, F. R. Cancer Res. 2003, 63, 8360.
13. Koenig, J. E.; Senge, T.; Allhoff, E. P.; Koenig, W. Prostate 2004, 58, 121.
14. Arnatt, C. K.; Zaidi, S. A.; Zhang, Z.; Li, G.; Richardson, A. C.; Ware, J. L.; Zhang, Y.
Eur. J. Med. Chem. 2013, 69, 647.
15. Langmuir, I. J. Am. Chem. Soc. 1919, 41, 1543.
16. Patani, G. A.; LaVoie, E. J. Chem. Rev. 1996, 96, 3147.
17. Arnatt, C. K.; Zhang, Y. Tetrahedron Lett. 2012, 53, 1592.
18. Baba, M.; Miyake, H.; Okamoto, M.; Iizawa, Y.; Okonogi, K. AIDS Res. Hum.
Retroviruses 2000, 16, 935.
19. Yuan, Y.; Arnatt, C. K.; Li, G.; Haney, K. M.; Ding, D.; Jacob, J. C.; Selley, D. E.;
Zhang, Y. Org. Biomol. Chem. 2012, 10, 2633.
a
favorable lead compound by carrying an IC50 value of
11.4 0.2
cells, respectively, and
l
M and 6.5 0.7
l
a
M in M12 and PC-3 prostate cancer
basal cytotoxicity of TC50 of

C. K. Arnatt et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2319–2323
2323
20. Zhang, X.; Haney, K. M.; Richardson, A. C.; Wilson, E.; Gewirtz, D. A.; Ware, J. L.;
Zehner, Z. E.; Zhang, Y. Bioorg. Med. Chem. Lett. 2010, 20, 4627.
BALB/c 3T3 neutral red uptake cytotoxicity test. A. T. f. B. C. F. a. I. V. V. S., phase III;
2003.
21. Webber, M. M.; Bello, D.; Quadar, S. Prostate 1997, 30, 58.
22. Haney, K. M.; Zhang, F.; Arnatt, C. K.; Yuan, Y.; Li, G.; Ware, J. L.; Gewirtz, D. A.;
Zhang, Y. Bioorg. Med. Chem. Lett. 2011, 21, 5159.
25. Dorff, T. B.; Glode, L. M. Curr. Opin. Urol. 2013, 23, 366.
26. Loblaw, D. A.; Walker-Dilks, C.; Winquist, E.; Hotte, S. J.; G.C.D.S.G. of C.C.O.P. in
E.-B Clin. Oncol. (R. Coll. Radiol.) 2013, 25, 406.
23. Borenfreund, E.; Puemer, J. A. Toxicol. Lett. 1985, 24, 119.
24. National Toxicology Program (NTP) Interagency Center for the Evaluation of
Alternative Toxicological Methods (NICEATM) Test method protocol for the
27. Jiaoyan, H.; Fangqiang, Z.; Xiang, X.; Hong, H.; Wenqiu, H.; Wenhui, C. J. Third
Mil. Med. Univ. 2013, 35, 105.
28. Zou, C.; Li, X.; Jiang, R. Chin. J. Androl. 2012, 26, 66.