Synthesis and biological evaluation of pyrrolidine derivatives as novel and potent sodium channel blockers for the treatment of ischemic stroke

Article abstract of DOI:10.1016/j.bmcl.2013.05.009

A novel series of pyrrolidine derivatives as Na⁺ channel blockers was synthesized and evaluated for their inhibitory effects on neuronal Na⁺ channels. Structure-activity relationship (SAR) studies of a pyrrolidine analogue 2 led to the discovery of 5e as a potent Na⁺ channel blocker with a low inhibitory action against human ether-a-go-go-related gene (hERG) channels. Compound 5e showed remarkably neuroprotective activity in a rat transient middle cerebral artery occlusion (MCAO) model, suggesting that 5e would act as a neuroprotectant for ischemic stroke.

Full text of DOI:10.1016/j.bmcl.2013.05.009

Bioorganic & Medicinal Chemistry Letters 23 (2013) 4230–4234
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and biological evaluation of pyrrolidine derivatives
as novel and potent sodium channel blockers for the treatment
of ischemic stroke
⇑
Maki Seki , Osamu Tsuruta, Ryo Tatsumi, Aki Soejima
Research Division, Mitsubishi Tanabe Pharma Corporation, 1000 Kamoshida-cho, Aoba-ku, Yokohama 227-0033, Japan
a r t i c l e i n f o
a b s t r a c t
A novel series of pyrrolidine derivatives as Na+ channel blockers was synthesized and evaluated for their
inhibitory effects on neuronal Na+ channels. Structure–activity relationship (SAR) studies of a pyrrolidine
analogue 2 led to the discovery of 5e as a potent Na+ channel blocker with a low inhibitory action against
human ether-a-go-go-related gene (hERG) channels. Compound 5e showed remarkably neuroprotective
activity in a rat transient middle cerebral artery occlusion (MCAO) model, suggesting that 5e would act as
a neuroprotectant for ischemic stroke.
Article history:
Received 28 February 2013
Revised 2 May 2013
Accepted 5 May 2013
Available online 14 May 2013
Keywords:
Ó 2013 Elsevier Ltd. All rights reserved.
Sodium channel blocker
Ischemic stroke
Neuroprotection
hERG
Pyrrolidine
Ischemic stroke is a leading cause of death and long-lasting dis-
ability in developed countries. Neuroprotection is widely recog-
nized to be a potential strategy for the treatment of ischemic
stroke,1 and considerable efforts have been devoted to the develop-
Me
N
Me
N
OH
N
N
F
F
O
N
N
S
O
N
F
ment of neuroprotective agents, such as N-methyl-
D
-aspartate
lubeluzole
enecadin
(NMDA) receptor antagonists,
a-amino-5-hydroxy-3-methyl-4-
isoxazole propionic acid (AMPA) receptor antagonists, and Ca2+
channel blockers.2 However, none of them has been approved for
the treatment of ischemic stroke because of their limited efficacy
or unfavorable risk-benefit ratio in clinical trials.3
O
Me
Me
OH
N
F
OH
Me
N
O
Me
2CH3SO3H
As an alternative neuroprotectant, Na+ channel blockers have
been reported (Fig. 1).4 Although several Na+ channel blockers have
been tested in clinical trials, some of them were unsuccessful be-
cause of their limited efficacy5 and/or adverse effects such as
heart-conduction disorders, QT prolongation,6 which was caused
by blocking of hERG K+ channels.7 Indeed, enecadin,4c crobene-
tine,4d SUN N8075,4e and our previous lead 14f displayed the poten-
tial risk of QT prolongation.8 Therefore, potent Na+ channel
blockers with a low risk of QT prolongation are strongly desired.
In the course of our study to develop Na+ channel blockers, we be-
came interested in a novel pyrrolidine analogue 2 as a tractable
lead having low molecular weight and three sites available for
derivatization (Fig. 2). Here, we describe synthesis and biological
evaluation of a series of novel pyrrolidine derivatives.
H2N
Me
Me
SUN N8075
Me
crobenetine
N
N
H
2HCl
OH
O
N
1
Figure 1. Reported sodium channel blockers.
Scheme 1 illustrates the synthesis of 4-phenylpyrrolidines 5a–e
and 6. All compounds were prepared as racemates. Stereospecific
[3+2]-cycloaddition of trans-1-nitro-2-phenylethylene and N-
(methoxymethyl)-N-[(trimethylsilyl)methyl]benzylamine, an azo-
methine ylide equivalent, afforded trans-3-nitropyrrolidine 3 in a
⇑
Corresponding author. Tel.: +81 45 963 7239; fax: +81 45 963 7257.
E-mail address: Seki.Maki@ma.mt-pharma.co.jp (M. Seki).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.05.009

M. Seki et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4230–4234
4231
R1HN
R2
N
R3
2
Figure 2. Our new pyrrolidine analogue 2.
2HCl
H2N
O2N
NO2
a
b
N
N
MeO
N
TMS
Ph
3
4
c or d
e
R1:
5a:
∗
R1HN
nHCl
N
O
HCl
N
HN
b:
c:
∗
5a e
6
∗
O
d:
e:
∗
∗
Scheme 1. Reagents and conditions: (a) trifluoroacetic acid, CH2Cl2 (62%); (b) (i) Fe, acetic acid, 2-PrOH/H2O, 80 °C, (ii) 4 N HCl/EtOAc, EtOAc (90%); (c) (i) benzoyl chloride or
cyclohexanecarbonyl chloride, Et3N, CH2Cl2, (ii) BH3ÁTHF, THF, reflux, (iii) 4 N HCl/EtOAc, EtOAc (35–62%); (d) (i) cyclohexanone, NaBH(OAc)3, acetic acid, acetonitrile, (ii) 4 N
HCl/EtOAc, EtOAc (65%); (e) (i) iodobenzene, CuI, K3PO4, ethylene glycol, 80 °C, (ii) 4 N HCl/EtOAc, EtOAc (35%).
Table 1
Inhibition of Na+ and hERG channels of 4-phenylpyrrolidines 5a–e and 6a
H
R1 N
N
Compound
R1
Anti-veratridineb IC50
(
lM)
hERGc (% inhibition, 1
lM)
∗
5a
5b
5c
1.90
0.64
0.48
45
57
31
O
O
∗
∗
5d
5e
0.32
0.51
60
∗
8.4
∗
∗
6
0.48
43
a
b
c
All compounds are racemic and were tested using the hydrochlorides or dihydrochlorides except 5a and 5c.
Inhibitory activity on veratridine-induced depolarization in rat cerebrocortical synaptosomes using voltage-sensitive dye Rhodamine 6G.10
Inhibition rates were determined by a voltage patch clamp technique using HEK293 cells expressing hERG channels.
moderate yield.9 The nitro group of 3 was reduced with iron and
acetic acid to give 4, which was then converted into 3-amidopyrr-
olidines 5a and 5c by acylation with the corresponding acid chlo-
rides. Reduction of the amide group of 5a and 5c yielded 3-
aminopyrrolidines 5b and 5d, respectively, and reductive alkyl-
ation of 4 with cyclohexanone provided 5e. Compound 6 was ob-
tained by copper-catalyzed coupling of 4 and iodobenzene.
The effects of 5a–e and 6 on Na+ channels were evaluated by
inhibitory action on veratridine-induced depolarization in rat
cerebrocortical synaptosomes (Table 1).10 Compounds 5c and 5d

4232
M. Seki et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4230–4234
BocHN
EtO2C
CO2Et
a
b
N
N
MeO
N
TMS
Ph
7
8
9
R1:
11a:
R1HN
H2N
nHCl
2HCl
c
d or e
∗
∗
N
N
b:
10
11a d
O
c:
d:
∗
∗
Scheme 2. Reagents and conditions: (a) trifluoroacetic acid, CH2Cl2 (63%); (b) (i) 1 N NaOH, THF/EtOH, (ii) DPPA, Et3N, toluene, reflux, then tert-BuOH, 100 °C (38%); (c) 4 N
HCl/dioxane, CH2Cl2 (95%); (d) (i) benzoyl chloride or cyclohexanecarbonyl chloride, Et3N, CH2Cl2, (ii) BH3ÁTHF, THF, reflux, (iii) 4 N HCl/EtOAc, EtOAc (42–63%); (e) (i)
cyclohexanone, NaBH(OAc)3, Et3N, acetic acid, acetonitrile, (ii) 4 N HCl/EtOAc, EtOAc (76%).
R2:
∗
a
b
c
2HCl
N R2
2HCl
HN
BocN
BocN
HN
N
N
NH
14a:
b:
∗
5e
12
13
14a,b
Scheme 3. Reagents and conditions: (a) (Boc)2O, i-Pr2NEt, CHCl3, reflux (71%); (b) H2, Pd(OH)2, EtOH, 50 °C (88%); (c) (i) cyclohexanecarboxaldehyde or cyclohexanone,
NaBH(OAc)3, acetonitrile, (ii) 4 N HCl/EtOAc, EtOAc (79–89%).
Table 2
Inhibition of Na+ and hERG channels of 4-cyclohexylpyrrolidines 11a–d and 1-cyclohexylpyrrolidines 14a,ba
Structure
Compound
R1 or R2
Anti-veratridineb IC50
(lM)
hERGc (% inhibition, 1
79%
lM)
H
R1 N
11a
0.90
∗
∗
N
11b
0.74
45%
O
11c
11d
0.56
0.74
60%
35%
∗
∗
∗
14a
14b
0.38
3.35
45%
ndd
HN
N R2
∗
a
All compounds are racemic and were tested using the dihydrochlorides except 11b.
b,c
d
See footnotes in Table 1.
nd: not determined.
bearing the cyclohexyl group exhibited higher Na+ channel inhibi-
tory activity than the corresponding aromatized derivatives 5a and
5b, respectively. In contrast, 5e bearing the cyclohexyl group
showed Na+ channel inhibition comparable to that of the corre-
sponding aromatized compound 6. The inhibitory action against
hERG channels of 5a–e and 6 was also assessed. As shown in Ta-
ble 1, 5a–e and 6 showed less potent hERG channel inhibition
the basis of the significant differences of the hERG inhibitory activ-
ity among 5d, 5e, and 6, we speculated that a substituent at the 3-
position on the pyrrolidine ring would be critical for hERG–ligand
interactions and the cyclohexyl group would be a useful substitu-
ent to reduce hERG inhibition.11 Thus, we next synthesized 1- and
4-cyclohexylpyrrolidine analogues to identify the optimal substi-
tution site of the cyclohexyl group.
(8.4–60% at 1
l
M) than the reported Na+ channel blockers.8 Nota-
Synthesis of 4-cyclohexylpyrrolidines 11a–d is shown
in Scheme 2. Cycloaddition of azomethine ylide and ethyl
bly, 5e exhibited a drastic decrease in hERG-blocking activity. On

M. Seki et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4230–4234
4233
On the basis of these findings, 5e was selected for further phar-
macological evaluation. The neuroprotective effect of 5e was eval-
uated in rats with a transient MCAO model.14 Subcutaneous
administration of 5e (0.8, 2.5, and 8.2 mg/kg) 30 min after occlu-
sion produced neuroprotective action in a dose-dependent manner
(27%, 37%, and 56% at 0.8, 2.5, and 8.2 mg/kg, respectively). More-
over, 5e at 8.2 mg/kg exhibited significant reduction in the size of
cerebral infarction compared with the vehicle [166.7 42.9 mm3
(n = 5) versus 379.8 43.8 mm3 (n = 6) (mean SEM), ⁄⁄p <0.01
(Student’s t-test)]. Since the effect of a drug on action potential
duration at 90% repolarization (APD90) prolongation is considered
of primary relevance in the context of a proarrhythmic potential
accompanying QT prolongation,15 we evaluated the effects of 5e
O
a
b
2HCl
N
HN
HN
N
N
O
O
16
15
cis-17
Scheme 4. Reagents and conditions: (a) cyclohexylamine, p-TsOH, toluene, reflux
(94%); (b) (i) BH3ÁTHF, THF, reflux, (ii) 4 N HCl/EtOAc, EtOAc (44%).
Table 3
Effects of compound 5e and crobenetine on ADP90
(10, 30, and 100 lM) and crobenetine hydrochloride (3, 10, and
a
Compound
Concentration (
l
M)
APD90 (%)
30 lM) on APD90 prolongation in isolated guinea pig papillary
Crobenetine hydrochloride
3
10
30
96.6 0.6
89.1 1.6
72.9 10.2
muscles.16 As shown in Table 3, both compounds did not prolong
APD90. Compound 5e slightly shortened APD90 (9.2 2.7% at
100
l
M), whereas crobenetine substantially shortened APD90
5e
10
30
100
95.3 0.7
92.3 0.7
90.8 2.7
(27.1 10.2% at 30
tential for cardiac adverse effects.
lM), suggesting that 5e may have a lower po-
In conclusion, we discovered a novel and potent Na+ channel
blocker 5e, which showed remarkably in vivo neuroprotective
activity and reduced hERG-blocking activity. The effect of 5e on
APD90 prolongation was also investigated, and it was revealed that
5e had a low risk of drug-induced QT prolongation. Further studies
and results will be reported in due course.
a
APD90: action potential duration at 90% repolarization; %: percent change from
before treatment value. Each value represents the mean SEM of three cell
preparations.
(2E)-3-cyclohexylacrylate (7) readily prepared from cyclohexane-
carboxaldehyde by the Horner–Emmons reaction yielded trans-8
stereospecifically. Hydrolysis of 8, followed by the Curtius rear-
rangement with diphenyl phosphoryl azide (DPPA) afforded 9.
The tert-butyl carbamate (Boc) group of 9 was removed by treat-
ment with hydrochloric acid to give 10. Benzoylation of 10, fol-
lowed by reduction of the amide group afforded 11a. Acylation of
10 with cyclohexanecarbonyl chloride provided 11b, which was
then converted into 11c by reduction of the amide group. Reduc-
tive alkylation of 10 with cyclohexanone furnished 11d. Synthesis
of 1-cyclohexylpyrrolidine analogues 14a and 14b is depicted in
Scheme 3. Compound 12 was obtained from 5e by protection of
the cyclohexylamine group with a Boc group. Removal of the ben-
zyl group of 12 by catalytic hydrogenation yielded 13. Reductive
alkylation of 13 with cyclohexanecarboxaldehyde or cyclohexa-
none afforded 14a or 14b, respectively.
Biological evaluation results for 11a–d, 14a, and 14b are shown
in Table 2. Replacement of the 4-phenyl group on the pyrrolidine
ring with a 4-cyclohexyl group led to a decrease in Na+ channel
inhibitory activity (5b vs 11a, 5c vs 11b, 5d vs 11c, and 5e vs
11d). On the other hand, the hERG-blocking activity of the 4-cyclo-
hexyl analogues 11a, 11b, and 11d tended to increase compared
with that of the 4-phenyl analogues 5b, 5c, and 5e, respectively.
Although 1-cyclohexylmethylpyrrolidine 14a showed Na+ channel
inhibitory activity similar to that of 1-benzylpyrrolidine 5e, its
hERG-blocking activity was potentiated. Introduction of a 1-cyclo-
hexyl group (14b) substantially lowered Na+ channel inhibitory
activity. These results suggest that the cyclohexyl group at the 3-
position on the pyrrolidine ring would be crucial to reduce hERG
inhibition.
Acknowledgments
We are grateful to Dr. Yukimi Yoneyama, Hiroyuki Harada,
Makiko Asayama, and Hiroyuki Sato for performing pharmacolog-
ical and toxicological evaluation.
References and notes
1. Beresford, I. J. M.; Parsons, A. A.; Hunter, A. J. Expert Opin. Emerg. Drugs 2003, 8,
103.
2. (a) Simon, R. P.; Swan, J. H.; Griffiths, T.; Meldrum, B. S. Science 1984, 226, 850;
(b) Park, C. K.; Nehls, D. G.; Graham, D. I.; Teasdale, G. M.; McCulloch, J. Ann.
Neurol. 1988, 24, 543; (c) Carter, C. J.; Lloyd, K. G.; Zivkovic, B.; Scatton, B. J.
Pharmacol. Exp. Ther. 1990, 253, 475; (d) Sheardown, M. J.; Nielsen, E. O.;
Hansen, A. J.; Jacobsen, P.; Honore, T. Science 1990, 247, 571; (e) Takahashi, M.;
Kohara, A.; Shishikura, J.; Kawasaki-Yatsugi, S.; Ni, J. W.; Yatsugi, S.; Sakamoto,
S.; Okada, M.; Shimizu-Sasamata, M.; Yamaguchi, T. CNS Drug Rev. 2002, 8, 337;
(f) Zornow, M. H.; Prough, D. S. New Horiz. (Baltimore, MD) 1996, 4, 107.
3. Keyser, J. D.; Sulter, G.; Luiten, P. G. Trends Neurosci. 1999, 22, 535.
4. (a) Urenjak, J.; Obrenovitch, T. P. Pharmacol. Rev. 1996, 48, 21; (b) Grotta, J.
Stroke 1997, 28, 2338; (c) Shimdizu, T.; Itou, Y.; Tatsumi, S.; Hayashi, S.; Ukai,
Y.; Yoshikuni, Y.; Kimura, K. Naunyn-Schmiedebergs Arch. Pharmacol. 1997, 355,
601; (d) Carter, A. J.; Grauert, M.; Pschorn, U.; Bechtel, W. D.; Bartmann-
Lindholm, C.; Qu, Y.; Scheuer, T.; Catterall, W. A.; Weiser, T. Proc. Natl. Acad. Sci.
U.S.A. 2000, 97, 4944; (e) Annoura, H.; Nakanishi, K.; Toba, T.; Takemoto, N.;
Imajo, S.; Miyajima, A.; Tamura-Horikawa, Y.; Tamura, S. J. Med. Chem. 2000, 43,
3372; (f) Seki, M.; Tsuruta, O.; Aoyama, Y.; Soejima, A.; Shimada, H.; Nonaka, H.
Chem. Pharm. Bull. 2012, 60, 488.
5. Pulsinelli, W. A.; Mann, M. E.; Welch, K. M. A.; Zivin, J. A.; Biller, J.; Maisel, J.;
Rubin, J. J.; Verro, P.; Graham, G. D.; Pierce, M.; Kugler, A. R.; Mohberg, N.;
Knapp, L. E.; Poole, R. M. Neurology 1999, 52, A384.
6. Gandolfo, C.; Sandercock, P. A. G.; Conti, M. Cochrane Databese Syst. Rev. 2002,
CD001924.
7. Vandenberg, J. I.; Walker, B. D.; Campbell, T. J. Trends Pharmacol. Sci. 2001, 22,
240.
8. Enecadin hydrochloride, crobenetine hydrochloride, SUN N8075, and
The attractive biological profiles of 5e encouraged us to exam-
ine its stereoisomers. Thus, cis-pyrrolidine 17 was synthesized
(Scheme 4) and evaluated for its inhibitory action against Na+
and hERG channels.12 Compound 15 was prepared according to
the previously reported procedure.13 Amination of 15 with cyclo-
hexcylamine in the presence of a catalytic amount of p-toluenesul-
fonic acid (p-TsOH) yielded 16. Reduction of 16 with borane
tetrahydrofuran complex afforded cis-17 as a major product (cis/
trans = 81/19). Compound 17 maintained Na+ channel inhibitory
compound 1 exhibited hERG inhibition by 60%, 69%, 78%, and 99% at 1 lM,
respectively. The hERG inhibitory activity was assessed by a conventional
voltage patch-clamp technique.
9. Pandey, G.; Banerjee, P.; Gadre, S. R. Chem. Rev. 2006, 106, 4484.
10. The membrane potential of synaptosomes was measured according to the
previously reported procedure.4f For reference Aiuchi, T.; Daimatsu, T.; Nakaya,
K.; Nakamura, Y. Biochim. Biophys. Acta 1982, 685, 289.
11. (a) Mitcheson, J. S.; Chen, J.; Lin, M.; Culberson, C.; Sanguinetti, M. C. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 12329; (b) Sanguinetti, M. C.; Mitcheson, J. S. Trends
Pharmacol. Sci. 2005, 26, 119.
activity (IC50 = 0.56
more potent than that of 5e (28% for 17 vs 8.4% for 5e at 1
lM), whereas its hERG-blocking activity was
12. The enantiomers of ( )-trans-5e showed a low eudysmic ratio (the IC50 values
of (+)- and (À)-trans-5e were 0.54 and 0.57
lM, respectively). These
lM).

4234
M. Seki et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4230–4234
enantiomers were synthesized as follows: acylation of racemic amine 4 with
4H), 1.56–1.74 (m, 3H), 1.74–1.89 (m, 1H), 2.04–2.18 (m, 1H), 2.99–3.11 (m,
1H), 3.39–3.53 (m, 1H), 3.85–4.10 (m, 4H), 4.44–4.68 (m, 3H), 7.36–7.57 (m,
8H), 7.62 (m, 2H); 13C NMR (400 MHz, CD3OD) d ppm 25.2, 25.4, 25.8, 29.6,
30.7, 48.2, 56.5, 57.2, 58.6, 59.3, 60.7, 129.2, 130.1, 130.6, 130.8, 131.2, 131.4,
131.7, 136.6; ESI-MS: 335.2 [M+H]+; Anal. Calcd for C23H30N2Á2HCl: C, 67.80; H,
7.92; N, 6.88; Found: C, 67.57; H, 7.75; N, 6.86.
naproxen to give a mixture of diastereomers, resolution by crystallization
(P99.8% de), acid hydrolysis of each naproxen amide to provide chiral amines
(+)- and (À)-4, and reductive alkylation of the chiral amines with
cyclohexanone to yield (+)- and (À)-trans-5e.
13. King, J. A.; McMillan, F. H. J. Am. Chem. Soc. 1950, 72, 1236.
14. The method employed was performed as previously described.4f Wister rats
(female, 8 weeks old) were subjected to transient MCAO for 90 min, followed
by reperfusion. Compound 5e was subcutaneously administered 30 min after
the occlusion at doses of 0.8, 2.5, and 8.2 mg/kg. Control rats underwent MCAO
and received an equal volume of saline dosed at the corresponding time point.
The brains were dissected into 7 slices 24 h after the occlusion, and stained
with 2,3,5-triphenyltetrazolium chloride. The area of infarct tissue for each
slice was determined with a computerized image analysis system, and the
infarct volume was calculated. For reference: (a) Koizumi, J.; Yoshida, Y.;
Nakazawa, T.; Ooneda, G. Jpn. J. Stroke 1986, 8, 1; (b) Bederson, J. B.; Pitts, L. H.;
Germano, S. M.; Nishimura, M. C.; Davis, R. L.; Bartkowski, H. M. Stroke 1986,
17, 1304; (c) Physical and spectroscopic data for 5e: colorless solid; mp 287.8–
288.8 °C; 1H NMR (400 MHz, CD3OD) d ppm 0.87–1.03 (m, 1H), 1.08–1.30 (m,
15. Crumb, W.; Cavero, I. Pharm. Sci. Technol. Today 1999, 2, 270.
16. The electrophysiological evaluation was performed by the conventional
microelectrode technique using guinea pig papillary muscles. Cells were
prepared from tissue taken from the ventricle muscle of guinea pig hearts. In
brief, the heart of guinea pigs was excised, and the papillary muscles were
dissected in the Tyrode solution. The cell preparation was fixed on the paraffin
base of the bath and was usually driven at constant frequencies (1 Hz) of field
stimulation. Action potentials through microelectrode amplifier were recorded
using a conventional glass microelectrode technique (resistance. 10–50 MO).
The duration of stimulation was 1 ms, and 1.5–3 times the voltage threshold in
strength was used. The temperature of all solutions was maintained at 36 °C.
APD90 was measured before treatment and 30 min after treatment of the test
compounds.