Synthesis and resolution of racemic eliprodil and evaluation of the enantiomers of eliprodil as NMDA receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(00)00246-8

A short synthesis of the NMDA receptor antagonist (rac)-Eliprodil (9) and its resolution into the enantiomers by chiral HPLC is described. The enantiomers (R)-9 and (S)-9 were found to exhibit markedly different affinities for NR2B subunit containing NMDA

Full text of DOI:10.1016/S0960-894X(00)00246-8

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1377±1380
Synthesis and Resolution of Racemic Eliprodil and Evaluation of
the Enantiomers of Eliprodil as NMDA Receptor Antagonists
È È
Jorg Pabel, Georg Hofner and Klaus Th. Wanner*
Institut fuÈr Pharmazie, Zentrum fuÈr Pharmaforschung, LMU MuÈnchen, Butenandtstr. 5±13, D-81377 MuÈnchen, Germany
Received 4 January 2000; accepted 25 April 2000
AbstractÐA short synthesis of the NMDA receptor antagonist (rac)-Eliprodil (9) and its resolution into the enantiomers by chiral
HPLC is described. The enantiomers (R)-9 and (S)-9 were found to exhibit markedly di€erent anities for NR2B subunit con-
taining NMDA receptors. # 2000 Elsevier Science Ltd. All rights reserved.
3
(rac)-Eliprodil ((rac)-9) is known as a ligand that is
is known. This was expected to be more ecient, as a
chromatographic procedure would provide both enantio-
mers in a single step and also would allow to easily check
for the enantiomeric purity of the separated isomers.
selective for NMDA receptors containing the NR2B
subunit. By acting as an antagonist at the ifenprodil or
polyamine sensitive recognition site present on this sub-
type of NMDA receptors it exhibits neuroprotective
properties as do other compounds with this mode of
action. As a promising drug candidate (rac)-Eliprodil
In the key step for the preparation of 6 the N-acylpyr-
idinium ion 3, in situ generated from pyridine and acetyl
1
,2
ꢀ
rate 4 to give the 4-substituted 1,4-dihydropyridine 5.
(
(rac)-9) was evaluated for stroke, until only recently
chloride at � 78 C in THF, was trapped with the cup-
the clinical trials with (rac)-Eliprodil ((rac)-9) were dis-
continued in phase III after side e€ects had emerged.2
For the enantiomers of Eliprodil in a rat ischemia model
it has been uncovered, that (R)-9 shows a distinctly
higher neuroprotective activity as compared to (S)-9.3
Though the reason for the di€erences found for the
enantiomers in their in vivo pharmacology might origi-
nate from di€erent receptor anities the in vitro poten-
cies of the enantiomers of Eliprodil ((R)-9 and (S)-9) as
NMDA receptor antagonists have not been investigated
up to now. Thus, it was the aim of the present study to
uncover whether the enantiomers of Eliprodil ((R)-9
and (S)-9) exhibit di€erent binding pro®les to NR2B
subunit containing NMDA receptors.
This reaction proceeded with high regioselectivity, no
5
other regioisomer was detected. Being aware, that 1,4-
dihydropyridines are easily oxidized, compound 5 was
subsequently hydrogenated over Pd/C and hydrolized
with acetic acid/HCl without prior puri®cation to give 6
in 44% overall yield. Finally, for the last two steps in
the synthesis of (rac)-9 the standard procedures dis-
played in the Scheme 1 were employed.
After some experimentation with various chiral columns
6
it was found that (rac)-9 may be eciently separated by
employing a CHIRALPAK1 AD HPLC column to
7
,8
give the pure enantiomers (R)-9 and (S)-9 (ee for both
enantiomers 599%, checked by analytical HPLC).
In an ongoing study directed toward the development of
more complex derivatives of Eliprodil with additional
substituents at the piperidine nucleus we utilized a syn-
thetic approach based on N-acylpyridinium ions to estab-
lish the piperidine core structure of these compounds. For
simplicity reasons this method was also applied to the
construction of 6. Furthermore it was envisaged to pre-
pare the pure enantiomers from the racemic material by
chiral chromatography, though an asymmetric synthesis
The anity of (rac)-9, (R)-9 and (S)-9 to the ifenprodil
recognition site at NMDA receptors was evaluated in a
3
[ H]ifenprodil binding assay employing a synaptosomal
9
fraction of porcine hippocampal brain membranes.
3
The displacement of [ H]ifenprodil binding by the
enantiomers was markedly di€erent.
4
The IC -values exhibited by (R)-9 (eutomer) and (S)-9
50
(
tively. This is in good accordance with the IC -value of
distomer) were 49.5Æ 0.7 nM and 545 Æ 120 nM, respec-
50
8
Fig. 1) and a value of 111 nM given in the literature.
2.6 Æ 10.0 nM obtained for (rac)-9 (see Table 1 and
*Corresponding author. Tel.: +49-89-2180-7249; fax: +49-2180-7247;
e-mail: klaus.wanner@cup.uni-muenchen.de
1
0
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00246-8

1
378
J. Pabel et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1377±1380
Scheme 1.
3
Table 1. Displacement of [ H]ifenprodil binding and inhibition of
[3
H]MK-801 binding
IC50 [3H]Ifenprodil
IC50 [3H]MK-801
High-anity
fraction
Low-anity
fraction
(
(
(
R)-9
S)-9
rac)-9
49.5Æ0.7 nM
545Æ120 nM
82.6Æ10.0 nM
213Æ76 nM
23.7Æ7.8 mM
a
5.10Æ0.77 mM
277Æ103 nM
11.3Æ2.7 mM
aNot assigned.
NR2B subunit selective NMDA receptor antagonists
such as ifenprodil or (rac)-9 have been shown to inhibit
3
3
[
H]MK-801 or [ H]TCP binding to brain membranes
under non-equilibrium conditions in a biphasic way, with
a high-anity phase re¯ecting allosteric inhibition at
NR2B subunit containing NMDA receptors and a low-
anity phase re¯ecting inhibition at NMDA receptors
3
Figure 1. Displacement of [ H]ifenprodil binding (means Æ SEM of
triplicates from one representative experiment out of three) to porcine
hippocampal brain membranes.9
10�12
lacking NR2B subunits.
3
A correlation between the
displacement of [ H]ifenprodil binding and the high-a-
3
3
nity phase of inhibition of [ H]MK-801 or [ H]TCP bind-
ing to brain membranes attributed to NR2B subunit
containing NMDA receptors has been demonstrated for
277 Æ 103 nM ((rac)-9), respectively. These values corre-
spond well to the potencies obtained for the displacement
of [ H]ifenprodil binding. For the low-anity fraction of
this type of NMDA receptor antagonists.1
0,11
Therefore it
3
3
3
appeared of interest to study the inhibition of [ H]MK-
01 binding e€ected by the enantiomers (R)-9 and (S)-9
inhibition of [ H]MK-801 binding IC -values of 23.7 Æ 7.8
50
8
mM ((R)-9) and 11.3 Æ 2.7 mM ((rac)-9), respectively, were
as well. These experiments were performed using the
synaptosomal fraction of porcine hippocampal brain
membranes mentioned above, however, this time in the
calculated (see Table 1 and Fig. 2).
3
In contrast to the biphasic inhibition of [ H]MK-801
binding caused by (R)-9 and (rac)-9, for the enantiomer
S)-9 only a monophasic inhibition of [ H]MK-801
9,11
presence of 100 mM l-glutamate and 30 mM glycine.
3
(
3
(R)-9 Inhibited [ H]MK-801 binding under these condi-
tions in a biphasic manner just as (rac)-9 with IC -
binding with an IC50 of 5.10 Æ 0.77 mM) (n =1.05) was
H
found (see Table 1 and Fig. 2). It may be assumed, that the
anities of (S)-9 for the sites attributed to allosteric inhi-
bition at NR2B subunit containing NMDA receptors and
50
values for the high-anity fraction of inhibition of
3
[
H]MK-801 binding being 213 Æ 76 nM ((R)-9) and

J. Pabel et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1377±1380
1379
CH MgCl (0.68 M) in 23 mL of THF was added. After
2
1
after warming to rt 10 mL of a saturated NH Cl-solu-
8 h ®rst 18 mL of phosphate bu€er (1M, pH 7) and
4
tion was added. The aqueous layer was extracted with
H CCl (4Â) and the combined organic phases were
2
2
dried (Na SO ) and concd iv. The residue was dissolved
2
4
in 180 mL of ethyl acetate and hydrogenated for 24 h at
atm in the presence of 970 mg of Pd/C (10%). Filtra-
1
tion and evaporation of the solvent iv yielded a colorless
solid, which was dissolved in 12 mL of acetic acid and 6
ꢀ
sealed tube. Then, at rt H O was added and the mixture
mL of concd HCl and heated to 120 C for 4 h in a
2
was washed with Et O. Finally the aqueous layer was
2
brought to pH 12±14 by the addition of KOH and
extracted with Et O. The combined organic layers were
2
dried (MgSO ), treated with HCl and concd iv to give
4
gas
a residue which upon recrystallisation from Et O/EtOH
2
yielded 6 as colorless crystals, 809 mg (44%); mp 103±
ꢀ
3
Figure 2. Inhibition of [ H]MK-801 binding (means Æ SEM of triplicates
� 1
1
1
04 C; IR (KBr) n~ =2922, 1599, 1583, 1511 cm ; H
NMR (CDCl ) d=1.50±1.75 (m, 3H, CH CHCH ), 1.82
from one representative experiment out of three) to porcine hippocampal
brain membranes in the presence of 100 mM l-glutamate and 30 mM
glycine after incubation for 15 min (non-equlibrium conditions).9
3
2
2
(d , J=11.0 Hz, 2H, CH CHCH ), 2.57 (d, J=6.2 Hz,
2
br
2
2
H, CH Ph), 2.79 (qu , Jꢁ12 Hz, 2H, NCH2ax), 3.46 (dbr,
2
br
to inhibition at NR2B subunit lacking NMDA receptors,
respectively, are close together leading to a monophasic
curve. Alternatively the anity of (S)-9 for the site
attributed to allosteric inhibition at NR2B subunit con-
taining receptors may be so low that the e€ect mediated
at this site could not be detected in the [ H]MK-801
binding assay. This could be due to a higher anity of
S)-9 for the radioligand binding site as compared to the
Jꢁ12 Hz, 2H, NCH ), 6.94±7.00 (m, 2H, Haromat.), 7.05±
2
eq
+
7.10 (m, 2H, H
), 9.36 (s , 1H, NH ), 9.65 (s ,
br 2 br
aromat.
+
1H, NH ); MS (CI, CH ): m/z (%): 194 (100)[M+
4
2
+
H � HCl]. Anal. calcd for C H NFCl: C 62.74; H
12
17
7.46; N 6.10. Found: C 62.78; H 7.49; N 6.03.
3
1-(4-Chlorophenyl)-2-[4-(4-¯uorobenzyl)-piperidin-1-yl]-
4
(
ethanon (8). To a solution of 800 mg (3.48 mmol) of 6
in 15 mL methanol and 45 mL acetone, 1.2 mL (7.31
site attributed to allosteric inhibition at NR2B subunit
containing receptors.
i
mmol) of NEt Pr and 894 mg (3.83 mmol) of 2-bromo-
2
0
was added until pH 1±2 was reached and extracted with
4
-chloroacetophenone was added. After 2 h 2N HCl
The results of the evaluation of both enantiomers of (rac)-
9
at NMDA receptors from porcine hippocampal brain
Et O. With KOH the aqueous layer was brought to pH
2
membranes suggest, that (R)-9 shows a substantially
higher anity for NR2B subunit containing versus NR2B
subunit lacking NMDA receptors. Although the anity
found for (S)-9 cannot be assigned to one of the two
binding sites that have been discussed above, it is obvious
that the potency of (S)-9 at the NR2B subunit contain-
ing NMDA receptors must be far lower. Assuming, that
the neuroprotective e€ects of (rac)-9 are mediated at
NMDA receptors containing NR2B subunits, our results
are in line with an investigation, which demonstrated,
that (R)-9 shows a distinctly higher neuroprotective
12±14 and again extracted with Et O. The combined
2
organic layers were dried (MgSO ) and evaporated to
4
dryness. The resulting residue was puri®ed by ¯ash
chromatography (heptane:ethyl acetate 6:4). Colorless
ꢀ
crystals (66%); mp 66 C; IR (KBr) n~ =2937, 2918, 1691,
�
1
1
1589, 1508, 1212 cm ; H NMR (CDCl ) d=1.38 (qd,
3
Jꢁ12/3.6 Hz, 2H, NCH CH2ax), 1.44±1.55 (m, 1H,
2
NCH CH CH), 1.61 (d , Jꢁ12 Hz, 2H, NCH CH2eq),
2
2
br
2
2.07 (td, Jꢁ12/2.2 Hz, 2H, NCH ), 2.50 (d, J=7.0 Hz,
2ax
2H, PhCH ), 2.94 (d , Jꢁ12 Hz, 2H, NCH2eq), 3.71 (s,
2
br
2H, NCH CO), 6.92±6.98 (m, 2H, H
2
), 7.05±7.11
aromat.
3
activity in a rat ischemia model as compared to (S)-9. At
(m, 2H, Haromat.), 7.40±7.44 (m, 2H, Haromat.), 7.95±8.00
(m, 2H, H ); MS (CI, CH ): m/z (%): 346
a ®rst view this seems to be in contrast to a conclusion of
the authors of the aforementioned study according to
which a discrepancy between their in vivo data and the
results of an in vitro study exists.13 Though the later
study describes that (R)-9 and (S)-9 are almost equally
potent, this refers to s and not to NMDA receptors.
aromat.
+
4
(100)[M+H ], 206 (18). Anal. calcd for C H NOClF:
21
20
C 69.46; H 6.12; N 4.05. Found: C 69.40; H 6.19; N 4.04.
(R,S)-(4-Fluorophenyl)-2-[4-(4-¯uorobenzyl)-piperidin-1-
4
yl]-ethanol (rac-9). To a stirred solution of 45 mg (0.13
mmol) of 8 in 1.5 mL Et O 142 mL (0.14 mmol) of LiAlH
in THF (1M) was added at 0 C. After 3 h the reaction was
2
4
ꢀ
quenched with H O. Addition of H CCl and 2N NaOH
2
2
2
Experimental and Analytical Data
gave a clear aqueous layer, which was extracted with
H CCl . The combined organic layers were dried over
MgSO and the solvent was evaporated. The resulting
2
2
4
4
-(4-Fluorobenzyl)-piperidine hydrochloride (6). 984 mg
4
(
8 mmol) of 1 and 632 mg (8 mmol) of 2 in 14 mL of
solid was puri®ed by ¯ash chromatography (heptane/
ethyl acetate 55:45+1% NEtMe ). Colorless crystals
ꢀ
suspension 4 generated from 786 mg (8.78 mmol) of
THF were stirred for 1 h at � 78 C. To the resulting
2
ꢀ
2747, 1600, 1510, 1488, 1447, 1220 cm ; H NMR
(76%); mp 129±130 C; IR (KBr) n~ =2919, 2803, 2769,
�
1
1
CuCN and 25.8 mL (17.54 mmol) of 4-F-C H4
6

1
380
J. Pabel et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1377±1380
(
1
CDCl ) d=1.21±1.38 (m, 2H, NCH CH ), 1.46±1.58 (m,
2. Parsons, C. G.; Danysz, W.; Quack, G. Drug News Per-
spect. 1998, 11, 523.
3
2
2
H, NCH CH CH), 1.60±1.70 (m, 2H, NCH CH ), 1.97
2 2 2 2
3
. Di Fabio, R.; Pietra, C.; Thomas, R. J.; Ziviani, L. Bioorg.
Med. Chem. Lett. 1995, 5, 551.
. Wick, A.; Frost, J.; Gaudilliere, B.; Bertin, J.; Dupont, R.;
Rousseau, J. EP Patent No. 0109317, 1984, Chem. Abstr. 1984,
(
td, J=11.5/2.6 Hz, 1H, NCH2ax), 2.25 (td, J=11.5/2.6
Hz, 1H, NCH2ax), 2.35 (dd, J=12.6/10.6 Hz, 1H,
CH CHOH), 2.45 (dd, J=12.6/3.8 Hz, 1H, CH CHOH),
2
4
2
2
.52 (d, J=7.3 Hz, 2H, PhCH ), 2.76 (d , J=11.5 Hz,
2 br
H, NCH2eq), 3.11 (d , J=11.5 Hz, 1H, NCH2eq), 4.17
br
1
01, 636.
. Comins, D. J.; Abdullah, A. H. J. Org. Chem. 1982, 47,
1
5
4315.
(
sbr, 1H, OH), 4.67 (dd, J=10.6/3.8 Hz, 1H, CHOH),
6
7
.94±6.99 (m, 2H, Haromat.), 7.06±7.12 (m, 2H, Haromat.),
6. For analysis of the enantiomers after derivatization see:
Malavasi, B.; Ripamonti, M.; Rouchouse, A.; Ascalone, V. J.
Chromatogr. 1996, 729, 323.
7. CHIRALPAK AD, Diacel Chemical Industries, Merck,
Darmstadt; heptane/ethanol=8/2, 1 ml/min; 265 nm; (R)-9:
.30 (s, 4H, H ); MS (CI, CH ): m/z (%): 348
aromat. 4
+
(100)[M+H ], 206 (40). Anal. calcd for C H NOClF:
20 23
C 69.06; H 6.66; N 4.03. Found: C 69.12; H 6.69; N 3.94.
6.2 min, (S)-9: 9.2 min.
8. Frost, J.; Bertin, J. FR Patent No. 2628740, 1989, Chem.
Acknowledgements
Abstr. 1989, 112, 612.
. Hofner, G.; Wanner, K. Th. Eur. J. Pharmacol. 2000, 394,
211.
10. Coughenour, L. L.; Cordon, J. J. J. Pharmacol. Exp. Ther.
1996, 280, 584.
Financial support of this work by the Deutsche
Forschungsgemeinschaft and the Fonds der Che-
mischen Industrie is gratefully acknowledged. We thank
Chiral Technologies-Europe for technical support.
9
È
1
Res. 1996, 16, 297.
È
1. Hofner, G.; Wanner, K. Th. J. Recept. Signal Transduct.
1
7
1
2. Reynolds, I. J.; Miller, R. J. Mol. Pharmacol. 1989, 36,
58.
3. Langer, S.; Frost, J.; Schoemaker, J.; Gaudilliere, B.;
References and Notes
1. Scatton, B.; Giroux, C.; Thenot, J. P.; Frost, J.; George, P.;
Carter, C.; Benavides, J. Drugs Future 1994, 19, 905.
Bertin, J.; Rousseau, J.; Dupont, R.; Wick, A. EP Patent No.
0428438A1, 1991, Chem. Abstr. 1991, 115, 79.