Discovery of novel neuronal voltage-dependent calcium channel blockers based on emopamil left hand as a bioactive template

Article abstract of DOI:10.1016/S0960-894X(02)01070-3

A series of novel neuronal voltage-dependent calcium channel (VDCC) blockers, with inhibitory activity at low micromolar and moderate solubility in water, was discovered by constructing and screening a focused library based on emopamil (1) left hand (ELH)

Full text of DOI:10.1016/S0960-894X(02)01070-3

Bioorganic & Medicinal Chemistry Letters 13 (2003) 919–922
Discovery of Novel Neuronal Voltage-Dependent Calcium Channel
Blockers Based on Emopamil Left Hand as a Bioactive Template
Yuichi Suzuki,a,* Noboru Yamamoto,a Yoichi Iimura,a Koki Kawano,a Teiji Kimura,a
Satoshi Nagato,a Koichi Ito,a Makoto Komatsu,b Yoshihiko Norimine,a
Manami Kimura,a Tetsuyuki Teramoto,a Yoshihisa Kaneda,a Takeshi Hamano,a
Tetsuhiro Niidomea and Masahiro Yonagaa
aTsukuba Research Laboratories, Eisai Co., Ltd., 5-1-3 Tokodai, Tsukuba-shi, Ibaraki 300-2635, Japan
bEbara Research Co., Ltd, 2-1 Honfujisawa 4-chome, Fujisawa-shi, Kanagawa 251-8502, Japan
Received 9 October 2002; accepted 9 December 2002
Abstract—A series of novel neuronal voltage-dependent calcium channel (VDCC) blockers, with inhibitory activity at low micro-
molar and moderate solubility in water, was discovered by constructing and screening a focused library based on emopamil (1) left
hand (ELH) as a bioactive template.
# 2003 Elsevier Science Ltd. All rights reserved.
Cerebral ischemia causes a large increase in extracellular
glutamate.1,2 Release of glutamate from neurons
requires the entry of Ca2+ into presynaptic terminals,
which is mediated by neuronal voltage-dependent cal-
cium channels (VDCCs). These VDCCs are classified
into L, N, P, Q, R and T subtypes based on their elec-
trophysiological and pharmacological properties.3À5
The release of glutamate further allows Ca2+ to enter
the neurons by activation of postsynaptic glutamate
receptors.2 Ca2+ can also enter postsynaptic neurons
through neuronal VDCCs.6 These events cause an
excessive elevation of the intracellular free Ca2+ con-
centration ([Ca2+]i). The resulting Ca2+ overload leads
to activation of proteases, nucleases and phospholi-
pases, and finally to neuronal cell death.1,7 Thus, inhib-
ition of neuronal VDCCs is expected to reduce excessive
glutamate release at the presynaptic terminals and to
than various subtype-specific VDCC blockers, which
would produce only a marginal effect on the extra-
cellular glutamate level when added alone.4
Various attempts to discover neuronal VDCC blockers
or subtype-specific VDCC blockers have been made
during the past decade; however, few non-peptide neu-
ronal VDCC or subtype-specific VDCC blockers, which
possess potent activity, water-solubility, low molecular
weight and sufficient brain penetration, have been
obtained so far.8 Our strategy to search for such neuro-
nal VDCC blockers that inhibit a variety of VDCC
subtypes, including N, P and Q subtypes, does not
utilize high-throughput screening, but rather involves
constructing a focused library. Emopamil (1)3 left hand
(ELH) was chosen as a bioactive template for this
focused library design. In the course of constructing
and screening this focused library, we discovered a new
series of VDCC blockers, which exhibit inhibitory
activity for neuronal VDCCs at low micromolar in rat
cortical synaptosomes assay.9 Herein, we report the design,
synthesis and structure–activity relationships (SARs) of
neuronal VDCC blockers bearing ELH as a bioactive
template.
inhibit [Ca2+
]
elevation in postsynaptic neurons,
i
thereby affording protection against neuronal cell death.
Neuronal VDCC blockers are considered to be among
the best candidates for neuroprotective drugs. In addi-
tion, it is expected that neuronal VDCC blockers that
extensively inhibit a variety of neuronal VDCCs, such
as N, P and Q subtypes, would have a greater effect
The syntheses of piperazine analogues 7–9, 18–22 and
23a–23q are depicted in Schemes 1–5. Alkylation of
phenylacetonitrile 2 with isopropyl bromide provided 3.
*Corresponding author. Tel.: +81-298-47-7391; fax: +81-298-47-
4012; e-mail: y5-suzuki@hhc.eisai.co.jp
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(02)01070-3

920
Y. Suzuki et al. / Bioorg. Med. Chem. Lett. 13 (2003) 919–922
The racemic aldehyde 4 was prepared through alkyl-
ation of 3 with 2-(2-bromoethyl)-1,3-dioxolane, fol-
lowed by hydrolysis. The aldehyde 4 was used to
reductively alkylate a variety of piperazines, affording 7,
9, 23a, 23d–g and 23l (Scheme 1).10 The racemic alcohol
6 was generated via Michael reaction of 3, followed by
reduction of 5 with lithium aluminum hydride. The
alcohol 6 was coupled with a variety of piperazines to
yield 20–22 (Scheme 2). Compound 7 was debenzylated
by catalytic hydrogenation to furnish the piperazine 8
which was coupled with alkyl halides to give 18 and 19
(Scheme 3). Compounds 23b–c, 23h and 23o–q were
prepared by Mitsunobu reaction (Scheme 4), and 23j–k
were produced by O-alkylation of the chloride 10
(Scheme 4). Compound 23i was synthesized by aromatic
nucleophilic substitution (Scheme 4). Catalytic hydro-
genation of 23l produced 23m, which was converted to
23n by reductive amination as shown in Scheme 5.
Piperazine intermediates (12a–c, 14, 16 and 17) were
prepared as shown in Scheme 6. Piperazine inter-
mediates 12a–c were prepared through Mitsunobu
removal of the benzyl protecting group to provide 16 and
17. All analogues were prepared as racemic compounds.
Inhibitory effects of piperazine analogues on neuronal
VDCCs were evaluated by employing rat cortical
synaptosomes.12 The anticonvulsant effect of a moder-
ately active compound was evaluated in an audiogenic
seizure model using DBA/2 mice to confirm in vivo
activity and permeability into the brain.13
Nifedipine and amlodipine are dihydropyridine deriva-
tives.14 It is generally accepted that dihydropyridines
are specific blockers of the L-type Ca2+ channel.14
Nifedipine is also specific toward L-type Ca2+ chan-
nel.14 However, amlodipine acts on both L- and N-type
Ca2+ channels despite the minor structural difference
from nifedipine.14 This result suggests the possibility of
modifying an L-type Ca2+ channel blocker to shift its
activity towards neuronal VDCCs. Compound 1, which
is an L-type Ca2+ channel blocker,3 was selected as a
template to modify because 3-methyl-2-phenylbutane-
nitrile (ELH) was expected to be suitable template to
construct a focused library aimed at neuronal VDCC
blockers, due to its low molecular weight. We were also
encouraged to pursue analogues of 1, as this compound
already shows a slight activity against neuronal VDCCs
(1; IC50=28 mM). A piperazine moiety was attached to
reaction
between
formyl-protected
1-piper-
azineethanol11 11 and the requisite phenols followed by
hydrolysis. The 4-nitrophenyl intermediate 14 was gen-
erated by aromatic nucleophilic substitution of 4-
fluoronitrobenzene. 1-Benzylpiperazine 15 was coupled
with substituted phenoxyethyl bromide, followed by
Scheme 1. Reagent and reaction conditions: (a) 50% KOH aq, 2-bro-
mopropane, DMSO; (b) NaNH2, 2-(2-bromoethyl)-1,3-dioxolane,
THF reflux; (c) 2 N HCl–acetone; (d) NaB(OAc)3H, AcOH, 1,2-
dichloroethane, piperazine derivatives.
Scheme 4. Reagent and reaction conditions: (a) SOCl2, CH2Cl2;
(b) substituted phenols, Ph3P, diethyl azodicarboxylate or NaH, 2-
bromopyridine for 23i; (c) substituted phenols, NaH, DMSO.
Scheme 5. Reagent and reaction conditions: (a) H2, Pd/C, MeOH;
(b) aq HCHO, NaBH3CN, AcOH, MeCN.
Scheme 2. Reagent and reaction conditions: (a) potassium tert-but-
oxide, ethyl acrylate, THF; (b) lithium aluminum hydride, THF;
(c) MsCl, triethylamine, MeCN then NaI, H2O, piperazine derivatives.
Scheme 6. Reagent and reaction conditions: (a) (i) 2-, 3- or 4-cyano-
phenol, Ph3P, diethyl azodicarboxylate; (ii) 4 N HCl–MeOH;
(b) potassium tert-butoxide, 4-fluoronitrobenzene; (c) substituted phe-
noxyethyl bromide, triethylamine, MeCN; H2, Pd(OH)2/C, MeOH.
Scheme 3. Reagent and reaction conditions: (a) H2, Pd(OH)2/C,
MeOH; (b) alkyl halides, triethylamine, MeCN.

Y. Suzuki et al. / Bioorg. Med. Chem. Lett. 13 (2003) 919–922
921
ELH with a view to ensuring aqueous solubility by
adding another basic amine. This approach is summar-
ized in Figure 1.
lipophilic group into the phenyl moiety, either an elec-
tron-withdrawing group or an electron-donating group,
improved the blocking activity towards the neuronal
VDCCs. Compound 23h, which has a Cl substituent on
the phenyl moiety, displayed increased VDCC blocking
activity in the synaptosomes assay (23h, IC50=3.4 mM).
In contrast, introduction of a hydrophilic group was
detrimental. Compound 23m showed decreased VDCC
blocking activity in the synaptosomes assay (23m,
IC50=13 mM).
Our next course of action was to improve the in vitro
potency by way of adding a variety of optional moieties
to piperazine. An initial SAR study was started with the
replacement of R1 (Table 1). The results of these modi-
fications led us to identify compounds 21, 22 and 23a
with low micromolar inhibitory activities for the neuro-
nal VDCCs (Table 1). As regards introduction of lipo-
philic substituents (Table 2), less lipophilic compounds
were preferable and 23a (ClogP; 21: 4.37, 22: 4.75, 23a:
4.32) was selected for further SAR study. Modification
of substituents on the phenyl ring was first investigated.
The SARs are summarized in Table 2. Introduction of a
A fluorine atom and a cyano group were each used as a
probe to explore the activity of three regioisomers on
the phenyl ring (23b–d, 23e–g, Table 2). In the case of
the fluorine atom, the trend of neuronal VDCC block-
ing potency was: 23d (para)>23c (meta)>23b (ortho)
(Table 2: IC50=5.9, 6.2 and 9.3 mM, respectively) with
the para fluoro analogue (23d) being the most active of
the three. In the case of the cyano group, however, the
trend of potency was: 23e (ortho)<23g (para)<23f
(meta) (Table 2: IC50=16, 13 and 7.7 mM, respectively).
Though both fluorine and cyano groups are electron-
withdrawing, they exhibited significantly different
SARs. We reasoned that a lipophilic group could favor
potent activity for neuronal VDCCs blockade, while a
hydrophilic group could be less favorable, as exempli-
fied by 4-N,N-dimethylamino (23n) and 4-amino (23m)
substituents. To test this hypothesis, naphthyl (23q) and
pyridyl derivatives (23i–k) were prepared as a repre-
sentative lipophilic and hydrophilic groups, respectively.
In accordance with our hypothesis, 23q exhibited potent
activity (IC50=2.5 mM), while 23i–k had essentially lost
the activity (IC50 >16 mM). This result suggested that
polar functionality is poorly tolerated. In general, lipo-
philic compounds tended to show high potency with
relatively little dependence of the activity on substituent
(‘flat’ SAR; IC50 values ranged from 2.5 to 9.6 mM).
Figure 1. Library design applying ELH as a bioactive template for
neuronal VDCC blocker.
Table 1. Neuronal VDCC inhibitory activity of substituted piper-
azine analogues (1, 7–9, 18–22 and 23a)
Table 2. Neuronal VDCC inhibitory activity of substituted piper-
azine analogues (23a–q)
Compd
R1
Synaptosomesa
IC50 (mM)
Emopamil (1)
8
9
18
19
28
H
16<
16<
16<
16<
Compd
Ar
Synaptosomesa
IC50 (mM)
(CH2)2OH
H2C–
(H2C)4-ꢀ
23a
23b
23c
23d
23e
23f
23g
23h
23i
23j
23k
23l
23m
23n
23o
23p
23q
Ph
2-F-Ph
3-F-Ph
4-F-Ph
2-CN-Ph
3-CN-Ph
4-CN-Ph
4-Cl-Ph
2-Pyridyl
8.9
9.3
6.2
5.9
16
20
7
16<
7.7
13
12
3.4
16<
16<
16<
7.5
13
16<
9.6
5.2
2.5
3-Pyridyl
4-Pyridyl
21
22
7.7
4.5
4-NO2-Ph
4-NH2-Ph
4-N,N-Dimethylamino-Ph
4-MeO-Ph
3,4-di-F-Ph
1-Napthyl
23a
8.9
aSee ref 12.
aSee ref 12.

922
Y. Suzuki et al. / Bioorg. Med. Chem. Lett. 13 (2003) 919–922
Compound 23d, which has a good balance of lipophili-
city and aqueous solubility, was chosen as a representa-
tive potent compound. The anticonvulsant effect of 23d
was evaluated in the audiogenic seizure model, and clear
antiseizure activity was observed (100% protection from
the tonic phase at 10 mg/kg, iv).
Lotarski, S. M.; Rock, D. M.; Sinz, M.; Stoehr, S. J.; Taylor,
C. P.; Weber, M. L.; Bowersox, S. S.; Miljanich, G. P.; Mill-
erman, E.; Wang, Y.; Szoke, B. G. J. Med. Chem. 1999, 42,
4239. (b) Hu, L.; Todd, R. R.; Rafferty, M. F.; Taylor, C. P.;
Feng, M. R.; Kuo, B.; Lotarski, S. M.; Miljanich, G. P.;
Millerman, E.; Siebers, K. M.; Szoke, B. G. Bioorg. Med.
Chem. 2000, 8, 1203. (c) Seko, T.; Kato, M.; Kohno, H.; Ono, S.;
Hashimura, K.; Takimizu, H.; Nakai, K.; Maegawa, H.; Katsube,
N.; Toda, M. Bioorg. Med. Chem. Lett. 2001, 11, 2067. (d) Cox,
B.; Denyer, J. C. Expert Opin. Ther. Pat. 1998, 8, 1237.
9. Yamamoto, N.; Suzuki, Y.; Iimura, Y.; Komatsu, M.;
Kawano, K.; Kimura, T.; Norimine, Y.; Ito, K.; Nagato, S.;
Niidome, T.; Teramoto, T.; Kimura, M.; Kaneda, Y.; Amino,
H.; Ueno, M.; Hamano, T.; Nishizawa, Y.; Yonaga, M. Dis-
covery of a potent neuron-selective calcium channel blocker:
Structure–activity relationships and neuroprotective effects of
novel piperazine derivatives; 222th ACS Natl Meeting, Aug
26–30 2001, Chicago; MEDI 051.
10. Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.; Mar-
yanoff, C. A.; Shah, R. D. J. Org. Chem. 1996, 61, 3849.
11. Yazawa, H.; Goto, S. Tetrahedron Lett. 1985, 26, 3703.
12. (a) The aim of this study was to evaluate the effect of
piperazine analogues on the plateau phase of KCl-elicited
intracellular free Ca2+ ([Ca2+]i) elevation in rat cortical
synaptosomes. [Ca2+]i was determined with a fluorimetric
assay using Fura2. At least six types of VDCCs have been
defined in neuronal tissue: T-, L-, N-, P-, Q- and R-type. In rat
synaptosomes, it has been reported that nifedipine (an L-type
calcium channel blocker) at 10 mM, o-conotoxin GVIA (o-
CgTx GVIA, an N-type calcium channel blocker) at 1 mM and
o-agatoxin IVA (a P/Q-type calcium channel blocker) at 1 mM
inhibit the KCl-elicited [Ca2+]i elevation by 16(Æ)3%,
18(Æ)5% and 64(Æ)8%, respectively.12b (b) David, B.;
Samantha, A.; David, L. Neuropharmacology 1993, 32, 1195.
13. (a) The known N-type calcium channel blocker, o-con-
otoxin GVIA, blocked seizures in this model.13b (b) Jackson,
H. C.; Scheideler, M. A. Psychopharmacology 1996, 126, 85.
(c) Male DBA/2 mice13d (body weigh: 8–12 g) were used. Sei-
zure was tested at 5 min after test compounds ivadministra-
tion. Seizures were evoked by means of auditory stimulation
(100 dB, 11 kHz) in mice placed singly under a perspex dome.
Sounds were continued for 60 s. Test compounds at 10 mg/kg
completely prevented sound-induced tonic convulsion in
DBA/2 mice. (d) De Sarro, G. B.; Meldrum, B. S.; Nistico, G.
Br. J. Pharmacol. 1988, 93, 247.
In conclusion, ELH has been demonstrated to be a
viable bioactive template for the preparation of a
focused library aimed at neuronal VDCC blockers.
Several analogues displayed higher affinity than 1 for
neuronal VDCCs. Inhibitory activity for the neuronal
calcium channels turned out to be closely associated
with the balance between the lipophilicity and polarity.
The representative compound 23d exhibited an antic-
onvulsant effect in an audiogenic seizure model using
DBA/2 mice and is considered to be a candidate neuro-
protective agent. SARs for 23d will be discussed else-
where.
Acknowledgements
We wish to thank the staff of Eisai Analytical Chemistry
Section for the measurement of spectral data and our
colleagues for valuable advice and discussions. We are
grateful to Ms. Yang Xu and Ms. Reshma Jagasia for
their advice in preparing the manuscript.
References and Notes
1. Kristian, T.; Siesjo, B. K. Stroke 1998, 29, 705.
2. (a) Graham, S. H.; Shiraisi, K.; Panter, S. S.; Simon, R. P.;
Faden, A. I. Neurosci. Lett. 1990, 110, 124. (b) Rothman,
S. M.; Olney, J. W. Ann. Neurol. 1986, 19, 105. (c) Dunlop, K.;
Luebke, J. I.; Turner, T. J. Trends Neurosci. 1995, 18, 89.
3. Kobayashi, T.; Mori, Y. Eur. J. Pharmacol. 1998, 363, 1.
4. Nishizawa, Y. Life Sci. 2001, 69, 369.
5. Randall, A. D.; Tsien, R. W. J. Neurosci. 1995, 15, 2995.
6. (a) Mayer, M. L.; Miller, R. J. Trends Pharmcol. Sci. 1990,
11, 254. (b) Osuga, H.; Hakim, A. M. J. Cereb. Blood Flow
Metab. 1996, 16, 629.
14. Furukawa, T.; Nukada, T.; Suzuki, K.; Fujita, Y.; Mori,
Y.; Nishimura, M.; Yamanaka, M. Br. J. Pharmacol. 1997,
121, 1136.
7. Choi, D. W. Trends Neurosci. 1988, 11, 465.
8. (a) Hu, L.; Todd, R. R.; Rafferty, M. F.; Feng, M. R.;