Synthesis and radioligand binding studies of bis-isoquinolinium derivatives as small conductance Ca2+-activated K+ channel blockers

Article abstract of DOI:10.1021/jm070412j

Starting from the scaffold of N-methyllaudanosine and N-methylnoscapine, which are known small conductance Ca²⁺-activated K⁺ channel blockers, original bis-isoquinolinium derivatives were synthezised and evaluated using binding studies, electrophysiology, and molecular modeling. These quaternary compounds are powerful blockers, and the most active ones have 10 times more affinity for the channels than dequalinium. The unsubstituted compounds possess a weaker affinity than the analogues having a 6,7-dimethoxy- or a 6,7,8-trimethoxy substitution. The length of the linker has no influence in the alkane derivatives. In relation to the xylene derivatives, the affinities are higher for the ortho and meta isomers. These results are well corroborated by a molecular modeling study. Finally, the most effective compounds have been tested in electrophysiological experiments on midbrain dopaminergic neurons and demonstrate the blocking potential of the apamin-sensitive after- hyperpolarization.

Full text of DOI:10.1021/jm070412j

5070
J. Med. Chem. 2007, 50, 5070-5075
Synthesis and Radioligand Binding Studies of Bis-isoquinolinium Derivatives as Small
Conductance Ca2+-Activated K+ Channel Blockers
Amaury Graulich,*,† Se´bastien Dilly,‡ Amaury Farce,‡ Jacqueline Scuve´e-Moreau,§ Olivier Waroux,§ Ce´dric Lamy,§
Philippe Chavatte,‡ Vincent Seutin,§ and Jean-Franc¸ois Lie´geois†
Laboratory of Medicinal Chemistry, Drug Research Center, UniVersity of Lie`ge, AVenue de l’Hoˆpital, 1 (B36), B-4000 Lie`ge 1, Belgium,
Laboratory of Pharmacology, Research Center for Cellular and Molecular Neurobiology, UniVersity of Lie`ge, AVenue de l’Hoˆpital,
1 (B36), B-4000 Lie`ge 1, Belgium, and Faculte´ des Sciences Pharmaceutiques et Biologiques, Laboratoire de Chimie The´rapeutique,
EA 1043, 3, Rue du Professeur Laguesse, BP 83, 59006 Lille Cedex, France
ReceiVed April 6, 2007
Starting from the scaffold of N-methyllaudanosine and N-methylnoscapine, which are known small
conductance Ca2+-activated K+ channel blockers, original bis-isoquinolinium derivatives were synthezised
and evaluated using binding studies, electrophysiology, and molecular modeling. These quaternary compounds
are powerful blockers, and the most active ones have 10 times more affinity for the channels than dequalinium.
The unsubstituted compounds possess a weaker affinity than the analogues having a 6,7-dimethoxy- or a
6,7,8-trimethoxy substitution. The length of the linker has no influence in the alkane derivatives. In relation
to the xylene derivatives, the affinities are higher for the ortho and meta isomers. These results are well
corroborated by a molecular modeling study. Finally, the most effective compounds have been tested in
electrophysiological experiments on midbrain dopaminergic neurons and demonstrate the blocking potential
of the apamin-sensitive after-hyperpolarization.
Introduction
Small conductance Ca2+-activated K+ (SK) channels underlie
the prolonged postspike after-hyperpolarization (AHP) that plays
an important role in modulating the firing rate and the firing
pattern of neurones.1-2 Three SK channel subunits have been
identified by DNA cloning, namely, SK1, SK2, and SK3.1 The
distribution of the subunits has been investigated in the rat by
using in situ hybridization and immunohistochemistry. SK1 and
SK2 subtypes are mostly expressed in the cortex and hippoc-
ampus,3 while SK3 channel expression is higher in subcortical
areas, especially in the monoamine cell regions, e.g., the
substantia nigra, dorsal raphe, and locus coeruleus. These
features attract great attention to SK channels as putative targets
for the treatment of cognitive dysfunction,4-8 neuronal hyper-
excitability,9 dopamine-related disorders,10-12 and depression.7
So far, the most potent SK channel blocker is apamin. Apamin
is an oligopeptide extracted from the honey bee Apis mellifera
venom. This octadecapeptide possesses two arginine residues
in contiguous positions and a rigid cyclic conformation due to
two disulfide bridges.13 Dequalinium, a nonpeptidic ligand, has
some SK channel blocking properties.14 Different chemical
modulation of the dequalinium structure led to the discovery
of highly potent bis-quinolinium compounds such as UCL1684.14
Furthermore N-methyllaudanosine (NML)15 and more recently
N-methylnoscapine (NMN)16 were reported to block SK chan-
nels (Figure 1). Unlike apamin, these two molecules possess
medium potency blocking properties and their effects are quickly
reversible.16 In order to improve the affinities of these isoquino-
linium blockers, several bis-isoquinolinium derivatives were
prepared and evaluated by using an in vitro binding assay and
electrophysiological experiments to find compounds with
Figure 1. Chemical structures of N-methyllaudanosine (NML) and
N-methylnoscapine (NMN).
increased affinity and to improve our knowledge of the
pharmacophore.
Chemistry
The syntheses of methoxyisoquinolines were classically
carried out by using a modification of the Pomeranz-Fritsch
synthesis.17 The ensuing bis-alkylation was performed by using
the Reissert compound pathway (Scheme 1).18 The Reissert
compounds (1a-c) were obtained by reaction of the corre-
sponding isoquinoline with benzoyl chloride in the presence of
trimethylsilyl cyanide.19 This reaction was carried out in CH2-
Cl2 and gave the Reissert compounds in good yield.19 These
derivatives were deprotonated by sodium hydride in DMF. The
resulting Reissert anions were alkylated by using a half
equivalent of the appropriate bis-electrophilic reagent.20 Then
the alkylated Reissert compounds were hydrolyzed to give the
bis-isoquinolines (2a-r). Finally, the bis-isoquinolines were
methylated by methyl iodide in DMF under mild warming
conditions to obtain the bis-isoquinolinium derivatives (3a-r).
Results
The in vitro binding data are summarized in Tables 1 and 2.
First, the compounds were screened at 10 µM. Because all
compounds displaced 125I-apamin by more than 58%, the Ki
values were determined for all drugs.
In our binding conditions, (()-NML and dequalinium (DQ+)
have an affinity (Ki) for the apamin-sensitive sites of ∼1300
and 220 nM, respectively.
* To whom correspondence should be addressed. Phone: ++32 (0) 43
66 43 77. Fax: ++32 (0) 43 66 43 62. E-mail: A.Graulich@student.ulg.ac.be.
† Drug Research Center, University of Lie`ge.
‡ Laboratoire de Chimie The´rapeutique.
§ Research Center for Cellular and Molecular Neurobiology, University
of Lie`ge.
10.1021/jm070412j CCC: $37.00 © 2007 American Chemical Society
Published on Web 09/15/2007

Bis-isoquinolinium DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 21 5071
Scheme 1. Synthesis of Bis-isoquinolinium Derivativesa
Table 2. Binding Affinities of N-Methyllaudanosine (NML),
Dequalinium (DQ+), and Bis-isoquinolinium Xylenes (3j-r) for Rat
Cortical Apamin Sensitive Sites
compd
R1
R2
isomer
Ki (nM)
NML
DQ+
3j
3k
3l
3m
3n
3o
3p
3q
1295 ( 15
221 ( 11
438 ( 83
132 ( 17
1103 ( 117
33 ( 1
106 ( 12
163 ( 30
35 ( 1
23 ( 5
280 ( 10
H
H
H
OMe
OMe
OMe
OMe
OMe
OMe
H
H
H
H
H
H
OMe
OMe
OMe
ortho
meta
para
ortho
meta
para
ortho
meta
para
a R1 and R2 ) H or R1 ) OMe and R2 ) H or R1 and R2 ) OMe, L )
(CH2)n with n ) 3-5, o-xylyl, m-xylyl, p-xylyl; (i) Me3SiCN, BzCl, AlCl3,
CH2Cl2, room temp; (ii) X-L-X (X ) Br or I), NaH, DMF, -10 °C; (iii)
NaOH, EtOH/ H2O, reflux; (iv) MeI, DMF, ∆t.
3r
Table 1. Binding Affinities of N-Methyllaudanosine (NML),
Dequalinium (DQ+), and Bis-isoquinolinium Alkanes (3a-h) for Rat
Cortical Apamin-Sensitive Sites
Table 3. Results of the Molecular Modeling Study for
Bis-isoquinolinium Alkane Derivatives
distance A a distance B b
∆E c
(kcal/mol)
compd
R1
R2
n
Ki (nM)
compd
R1
R2
n
(Å)
(Å)
NML
DQ+
3a
3b
3c
3d
3e
3f
3g
1295 ( 15
221 ( 11
1687 ( 203
1962 ( 315
1465 ( 35
196 ( 19
164 ( 18
150 ( 6
3a
3b
3c
3d
3e
3f
3g
3h
3i
H
H
H
H
1
2
3
1
2
3
1
2
3
5.59
5.76
5.60
5.65
5.59
5.61
5.64
5.65
5.72
6.90
5.93
9.11
6.09
6.58
8.10
6.35
7.51
9.41
2.32
5.83
7.26
4.47
4.50
1.04
4.52
2.79
6.65
H
H
H
H
H
H
H
H
OMe
OMe
OMe
OMe
OMe
H
H
H
H
H
H
OMe
OMe
1
2
3
1
2
3
1
2
OMe
OMe
OMe
OMe OMe
OMe OMe
OMe OMe
90 ( 13
3h
42 ( 8
a N+-N+ distance of the conformer closest to the pharmacophoric
distance (5.6 Å) previously described.21 b N+-N+ distance of the minimum
energy conformer. c ∆E between the conformer closest to the pharmacophore
and the minimum energy conformer.
The affinities of compounds without substitution on the
isoquinoline nucleus and with an alkane linker (3a-c) are equal
to 1687 ( 203, 1962 ( 315, and 1465 ( 35 nM, respectively,
for propane (3a), butane (3b), and pentane (3c) derivatives. With
a 6,7-dimethoxy substitution (3d-f), the affinities of the
compounds are 196 ( 19, 164 ( 18, and 150 ( 6 nM,
respectively, for propane (3d), butane (3e), and pentane (3f)
derivatives. Finally, the compounds with the 6,7,8-trimethoxy
substitution (3g,h) have an affinity of 90 ( 13 and 42 ( 8 nM,
respectively, for propane (3g) and butane (3h) derivatives.
In the xylene series, the unsubstituted compounds (3j-l) have
an affinity of 438 ( 83, 132 ( 17, and 1103 ( 117 nM for the
ortho (3j), meta (3k), and para (3l) isomers, respectively. The
affinities of the compounds possessing a 6,7-dimethoxy sub-
stitution (3 m-o) are equal to 33 ( 1, 106 ( 12, and 163 ( 30
nM for the ortho (3m), meta (3n), and para (3o) isomers,
respectively. In the presence of three methoxy groups (3p-r),
the affinity is 35 ( 1, 23 ( 5, and 280 ( 10 nM for the ortho
(3p), meta (3q), and para (3r) isomers, respectively.
dopaminergic neurons with an IC50 of 4.4 ( 1.2 µM (n ) 3),
3.8 ( 0.7 µM (n ) 3), 4.1 ( 1.4 µM (n ) 3), and 2.4 ( 0.3
µM (n ) 3), respectively.
A molecular modeling study was performed in order to
determine the N+-N+ distance of the conformers closest to the
pharmacophore previously described,21 the N+-N+ distance of
the minimum energy conformers, and the ∆E between the two
conformers. The results of the molecular modeling study are
summarized in Tables 3 and 4.
Discussion
Pharmacologically, N-methyllaudanosine and N-methylnos-
capine are two interesting reversible blockers of SK channels,
but their affinities for apamin-sensitive sites are relatively
weak.16 The main difference between these N-methyl alkaloids
and the more potent SK channel blockers14 is the presence of
two positively charged centers in the latter.14,22 In order to obtain
more potent blockers, bis-isoquinoliniums were therefore syn-
In electrophysiological experiments, compounds 3h, 3m, 3p,
and 3q block the apamin-sensitive after-hyperpolarization in

5072 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 21
Graulich et al.
Table 4. Results of the Molecular Modeling Study for
Bis-isoquinolinium Xylene Derivatives
distance A a distance B b
∆E c
(kcal/mol)
compd
R1
R2
isomer
(Å)
(Å)
3j
3k
3l
3m
3n
3o
3p
3q
3r
H
H
H
OMe
OMe
OMe
H
ortho
meta
para
ortho
meta
para
5.67
5.12
7.62
5.64
5.53
7.88
5.77
5.63
8.42
8.14
5.11
8.44
8.15
5.48
9.25
8.04
8.21
8.49
4.39
0.41
4.87
4.34
5.65
11.46
6.36
3.25
4.53
H
H
H
H
H
Figure 2. Superimposition of the pharmacophore (colors by atom type),
the lowest energy conformer of compound 3a (green), and the best
fitting conformer (green-blue): (A) N+-N+ distance of the conformer
closest to the pharmacophoric distance (5.6 Å); (B) N+-N+ distance
of the minimum energy conformer.
OMe OMe ortho
OMe OMe meta
OMe OMe para
N+ are close to the pharmacophoric distance (5.6 Å between
both N+) with the exception of three compounds that probably
adopt the lowest energy conformation (3b, ∆E ) 5.8 kcal/mol;
3c, ∆E ) 7.3 kcal/mol; 3i, ∆E ) 6.7 kcal/mol).
a N+-N+ distance of the conformer closest to the pharmacophoric
distance (5.6 Å) previously described.21 b N+-N+ distance of the minimum
energy conformer. c ∆E between the conformer closest to the pharmacophore
and the minimum energy conformer.
In the xylene series, the presence of methoxy substituents
on the isoquinolinium also appears to be essential, but the gap
of affinity between unsubstituted and methoxylated compounds
is less important than that in some of the alkane derivatives.
Indeed, the affinities of methoxylated derivatives (3m, Ki ) 33
nM; 3p, Ki ) 35 nM) in the ortho series are only 10 times
higher than that of the unsubstituted compound (3j, Ki ) 438
nM) rather than 50 times as found for alkane derivatives 3b
and 3h, for example. The para-linker series also displays a
significant difference between unsubstituted (3p, Ki ) 1103 nM)
and substituted (3o, Ki ) 163 nM; 3q, Ki ) 280 nM)
compounds. With the m-xylene linker, the difference of affinity
is not well marked between unsubstituted and methoxylated
analogs (3k, Ki ) 132 nM; 3n, Ki ) 106 nM; 3q, Ki ) 23 nM)
because the trimethoxylated analogue is 4-5 times more potent
than the two other analogues. Both binding and molecular
modeling results show clearly that the ortho and meta linkers
lead to more favorable conformations than the para linker. In
fact, ortho and meta derivatives have always higher affinities
than the corresponding para derivatives. This is particularly clear
with the 6,7,8-trimethoxy compounds (o-3p, Ki ) 35 nM; m-3q,
Ki ) 23 nM; p-3r, Ki ) 280 nM) for which the affinity of the
para isomer is 10 times lower than that of the meta and ortho
isomers. The molecular modeling study explains this observation
by the fact that the para derivative is unable to adopt a
conformation in accordance with the determined pharmacophore.
Indeed, for the para derivatives, the conformers close to the
pharmacophore have a N+-N+ distance larger than 7.6 Å (3l,
7.6 Å; 3o, 7.9 Å; 3r, 8.4 Å). Moreover, for compound 3o, the
∆E between the minimum energy conformer and the pharma-
cophoric conformer is particularly high (∆E ) 11.5 kcal/mol).
thesized. Three types of substitutions of the isoquinoliniums
were used, namely, unsubstituted, 6,7-dimethoxyisoquinoline,
and 6,7,8-trimethoxyisoquinoline. The trimethoxylated deriva-
tives were prepared because the tetrahydroisoquinolinium,
8-methoxy-NML, was shown to have an affinity superior to that
of NML (Ki ) 730 and 1295 nM, respectively).23 In a previous
study, it had also been shown that 8-isopropyl substitution
influenced the affinity as the 6,7-dimethoxy substitution, but
the cost and difficulties to synthesize the compound are serious
drawbacks for further development.24 A methyl group was used
for the N-alkylation because it is the most effective substituent
among ethyl, benzyl, and butyl groups as previously reported.25
These bis-isoquinoliniums were then tested by in vitro binding
in order to determine their affinities for apamin-sensitive sites.
A pharmacophoric study examining only UCL compounds (bis-
quinoliniums) has shown that the distance between the centroids
of the pyridinium rings of the quinolinium groups must be close
to 5.8 Å.26 Another study taking into account apamin and UCL
structures shows that a distance of ∼5.6 Å is the optimal distance
between the two ionic centers.21 So for all compounds tested,
the N+-N+ distance of the minimum energy conformers and
the ∆E between the minimum energy conformer and the
conformer closest to the pharmacophore (5.6 Å) were deter-
mined (Figure 2).
In the alkane series (3a-i), the affinity increased clearly with
the presence of methoxy substituents on the isoquinolinium
nucleus. Thus, in the propane series, 3d (6,7-dimethoxy-
substituted compound, Ki ) 196 nM) and 3g (6,7,8-trimethoxy-
substituted compound, Ki ) 90 nM) have higher affinities than
the unsubstituted compound 3a (Ki ) 1687 nM). Moreover, the
same observations could be made for the butane analogues
(1962, 164, and 42 nM, respectively, for unsubstituted (3b),
6,7-dimethoxy-substituted (3e), and 6,7,8-trimethoxy-substituted
(3h) compounds) and for the pentane analogues (1465 and 150
nM, respectively, for unsubstituted (3c) and 6,7-dimethoxy-
substituted (3f) compounds). The methoxy substitution therefore
appears to be very important in this series. Interestingly, the
length of the linker is not an essential parameter, and this finding
is well correlated with the results of the molecular modeling
study. Indeed, almost all compounds could adopt with good
probability (∆E < 5 kcal/mol) a configuration in which both
In order to confirm a channel blockade, the four most effective
compounds (3h, 3m, 3p, and 3q) were evaluated in electro-
physiological experiments performed in rat brain slices. In these
electrophysiological experiences, all compounds block the
apamin-sensitive AHP recorded in midbrain dopaminergic
neurons with an IC50 between 2.4 and 4.4 µM. Thus, these
compounds are 3-7 times more potent than NML (IC50 ) 15
µM, n ) 3) in these experiments.
In conclusion, this study shows that bis-isoquinolinium
derivatives are potential pharmacological tools and reveals four

Bis-isoquinolinium DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 21 5073
2H, J ) 7.5 Hz), 8.44 (d, 2H, J ) 5.7 Hz). Anal. (C21H18N2) C, H,
N.
derivatives having an affinity nearly 50 times higher than the
affinity of NML and shows the importance of the pharmacoph-
oric distance between the two nitrogens. This distance factor is
not the only obvious condition for an optimal interaction with
the SK channels because the substitution of the aromatic nucleus
could also greatly influence the affinity for the apamin-sensitive
sites, as shown with methoxy substituents.
1,1′-(Propane-1,3-diyl)bis(2-methylisoquinolinium) Diiodide
3a. Compound 2a (1.3 g, 4.4 mmol) in DMF (10 mL) was heated
until dissolution with an excess of methyl iodide (1.0 mL, 16 mmol).
After 2 h, Et2O (100 mL) was added, resulting in a rapid
crystallization to give a yellow solid. The precipitate was filtered
off, washed with Et2O (2 × 10 mL), and dried (2.3 g): yield, 89%;
1
mp 237-238 °C, dec. IR (KBr) 1634, 1399, 821 cm-1. H NMR
Experimental Section
(DMSO-d6) δ 2.11 (br m, 2H), 4.08 (t, 4H, J ) 8.3 Hz), 4.54 (s,
6H), 8.12 (t, 2H, J ) 7.7 Hz), 8.23 (t, 2H, J ) 7.7 Hz), 8.31 (d,
2H, J ) 8.1 Hz), 8.43 (d, 2H, J ) 6.9 Hz), 8.71 (d, 2H, J ) 6.9
Hz), 9.00 (d, 2H, J ) 8.6 Hz). 13C NMR (DMSO-d6) δ 26.6 (CH2),
28.9 (CH2), 46.5 (CH3), 124.1 (CH), 128.1 (CH), 128.2 (CH), 131.3
(CH), 135.9 (CH), 137.1 (CH). Anal. (C23H24N2I2‚1/4H2O) C, H,
N.
1. Chemistry. Melting points were determined on a Bu¨chi-Tottoli
capillary melting point apparatus in open capillary and are uncor-
rected. 1H NMR and 13C NMR spectra were recorded on a Bruker
Avance 500 (500 and 125 MHz) spectrometer. IR spectra were
performed on a Perkin-Elmer FTIR-1750 spectrometer. IR spectra
were measured using KBr discs. Only significant bands from the
IR are reported. Elemental analyses were determined using a Carlo-
Erba elemental analyzer CHNSO model EA1108, and the results
are within 0.4% of the theoretical values. All starting materials and
reagents were obtained from Aldrich Chemical Co. and Acros
Chemical Co. and were used without further purification. Separa-
tions by column chromatography were carried out using Merck
Kieselgel 60 (230-400 mesh). Concentration and evaporation refer
to removal of volatile materials under reduced pressure (10-15
mmHg at 30-50 °C) on a Bu¨chi rotavapor.
2. Radioligand Binding Studies and Data Analysis. 2.1.
Synaptosomes Preparation. Rats (male Wistar, (250 g) were
killed by decapitation, and the brains were quickly removed and
kept on ice during dissection. Crude cortex was dispersed in
0.32 M sucrose by using a Potter homogenizer. After a first
centrifugation at 1500g for 10 min, the supernatant was centrifuged
at 25000g for 10 min. The resulting pellet was dispersed in 5 mL
of 0.32 M sucrose to be aliquoted. Protein concentration was
determined by the method of Hartree with bovine serum albumin
as a standard.28
2.2. Binding Experiments. The buffer consisted of a 10 mM
Tris-HCl (pH 7.5) solution containing 5.4 mM KCl and 0.1% bovine
serum albumin. The radioligand was 125I-apamin (Perkin-Elmer,
specific activity of 81.4 TBq mmol-1). Glass fiber filters (Whatman
GF/C) used in these experiments were coated for 1 h in 0.5%
polyethylenimine and then washed with 2.5 mL of the ice-cold
buffer just before use. Binding experiments were always terminated
as follows. Aliquots were filtered under reduced pressure through
Whatman filters. Filters were rapidly washed twice with 2.5 mL
of buffer. The radioactivity remaining on the filter was evaluated
with a Packard Tri-Carb 1600TR liquid scintillation analyzer with
an efficacy of 69%. Curve fitting was carried out using GraphPad
Prism.
The methoxylated isoquinolines were prepared by the Jackson
modification of the Pomeranz-Fritsch synthesis.17 Compound 1a
was obtained from isoquinoline by the procedure described by Uff
et al.27
2-Benzoyl-1-cyano-6,7-dimethoxy-1,2-dihydroisoquinoline 1b.
Anhydrous aluminum chloride (10 mg) was added to a stirred
solution of 6,7-dimethoxyisoquinoline (2.97 g, 15.7 mmol) and
trimethylsilyl cyanide (3.9 mL, 31.4 mmol) in anhydrous CH2Cl2
(50 mL) at room temperature. Then benzoyl chloride (3.6 mL, 31.4
mmol) was added dropwise to the stirred solution over the course
of 5 min. The mixture was warmed to 30 °C if no exothermic
reaction has begun after the addition of benzoyl chloride. After
the mixture was stirred for a further 3 h, water (50 mL) was added
and stirring continued for 30 min. The organic layer was collected
and washed successively with 1 N aqueous HCl (2 × 50 mL), water
(50 mL), 1 N aqueous NaOH (2 × 50 mL), and finally water (50
mL). The organic solution was dried over anhydrous MgSO4 and
evapored under reduced pressure. The resulting oil crystallized by
trituration with Et2O (20 mL). The crystals were collected, washed
with small volumes of Et2O, and dried (3.5 g): yield, 70%; mp
165-166 °C. IR (KBr) 1661, 1631, 1345 cm-1. 1H NMR (CDCl3)
δ 3.92 (s, 3H), 3.94 (s, 3H), 5.99 (br d, 1H, J ) 6.6 Hz), 6.51 (br
s, 2H), 6.72 (s, 1H), 6.85 (br s, 1H), 7.47 (t, 2H, J ) 7.4 Hz), 7.55
(t, 1H, J ) 7.4 Hz) 7.60 (d, 2H, J ) 7.4 Hz). Anal. (C19H16N2O3)
C, H, N.
2.2.1. Saturation Binding Experiments. Synaptosomes (0.2 mg
of protein/mL) were incubated with increasing concentrations of
125I-apamin (25 µL) with 975 µL of incubation buffer for 1 h at
0 °C. Samples were then filtered on Whatman GF/C filter, and the
radioactivity was measured as described above. Nonspecific binding
was determined in parallel experiments in the presence of an excess
of unlabeled apamin (0.1 µM) and substracted from the total binding
to obtain the specific binding.
2.2.2. Competition Experiments between 125I-Apamin and
Drugs. Synaptosomes (0.2 mg of protein/mL) were incubated for
1 h at 0 °C with (10 pM of 125I-apamin (25 µL) and eight
concentrations of drugs (10-4.5-10-8 M). Nonspecific binding was
determined in the presence of an excess of unlabeled apamin (0.1
µM). Samples were then filtered on Whatman filter, and the
radioactivity was measured as described above.
2.3. Electrophysiological Experiments. The procedure is largely
described in a previous paper.15 Briefly, male Wistar rats (150-
200 g) were anesthetized with chloral hydrate (400 mg/kg ip) and
decapitated. The brain was excised quickly and placed in cold
(∼4 °C) artificial cerebrospinal fluid (ACSF) of the following
composition (in mM): NaCl 126, KCl 2.5, NaH2PO4 1.2, MgCl2
1.2, CaCl2 2.4, glucose 11, NaHCO3 18, saturated with 95% O2
and 5% CO2 (pH 7.4). A block of tissue containing the midbrain
was cut into horizontal slices (thickness of 350 µm) in a Vibratome
(Lancer). The slice containing the region of interest was placed on
a nylon mesh in a recording chamber (volume of 500 µL). The
tissue was completely immersed in a continuously flowing (∼2 mL/
min) ACSF, heated at 35 °C. Most recordings were made from
dopaminergic neurons located in the substantia nigra pars compacta.
Identification of dopaminergic cells was performed as described
previously.15 Intracellular recordings were performed using glass
microelectrodes filled with 2 M KCl (resistance of 70-150 MΩ).
1,1′-(Propane-1,3-diyl)bis-isoquinoline 2a. A solution of 2-ben-
zoyl-1-cyano-1,2-dihydroisoquinoline 1a (5.0 g, 19.2 mmol) and
1,3-diiodopropane (1.1 mL, 9.6 mmol) in DMF (15 mL) was added
dropwise to a stirred suspension of sodium hydride (0.46 g, 19.2
mmol) in DMF (30 mL) at -10 °C. The mixture was stirred for 4
h and poured into ice-cold water (200 mL). The creamy solid was
filtered off. After the mixture was dried, the solid was dissolved in
EtOH (50 mL) and was then hydrolyzed by treatment with a 50%
aqueous NaOH (20 mL) at reflux. After removal of EtOH, the crude
residue was dissolved in toluene (50 mL) and water (50 mL). The
organic layer was collected, washed with water (50 mL), and then
extracted with 1 N aqueous HCl (2 × 50 mL). The acidic layers
were basified with concentrated NH4OH and finally extracted with
CH2Cl2 (3 × 30 mL). The organic layers were dried over anhydrous
MgSO4 and evaporated under reduced pressure to afford a white
solid, which recrystallized from petroleum ether 100-140 °C (2.5
g): yield, 40%; mp 96-97 °C. IR (KBr) 1560, 1386, 826, 820
1
cm-1. H NMR (CDCl3) δ 2.49 (pentuplet, 2H, J ) 7.7 Hz), 3.50
(t, 4H, J ) 7.7 Hz), 7.51 (d, 2H, J ) 5.7 Hz), 7.56 (t, 2H, J ) 7.5
Hz), 7.65 (t, 2H, J ) 7.5 Hz), 7.80 (d, 2H, J ) 7.5 Hz), 8.18 (d,

5074 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 21
Graulich et al.
(2) Stocker, M.; Krause, M.; Pedarzani, P. An apamin-sensitive Ca2+
-
All recordings were made in the bridge balance mode, using a npi
SEC1L amplifier (Tamm, Germany). The accuracy of the bridge
was checked throughout the experiment. Membrane potentials and
injected currents were recorded on a Gould TA240 chart recorder
and on a Fluke Combiscope oscilloscope. The Flukeview software
was used for off-line analysis in most cases. Drug effects on the
prominent apamin-sensitive AHP in dopaminergic neurones were
quantified as the percent reduction of the surface area of the AHP
(in mV s), which was blocked by a maximally active concentration
of apamin (300 nM).15 Averages of four sweeps were considered
in all cases. The spontaneous firing of the neurons was usually
reduced by constant current injection (-20 to -100 pA) in order
to increase the amplitude of the AHP. Because the amplitude of
the AHP is very sensitive to the firing rate, care was taken to
compare all AHPs of one cell at the same firing rate. All drugs
were applied by superfusion; complete exchange of the bath
solution occurred within 2-3 min. Curve fitting was carried out
using GraphPad Prism and the standard equation E ) Emax/[1 +
(IC50/x)h], where x is the concentration of the drug and h the
Hill coefficient. Numerical values are expressed as the mean (
SD. Apamin (Sigma) was dissolved in water. Compounds 3h,
3m, 3p, and 3q were first dissolved in dimethyl sulfoxide (2 ×
10-2 M) and then in water to reach the appropriate concen-
tration.
3. Molecular Modeling. Models of the compounds were built
under the Sybyl 6.91 molecular modeling package (SYBYL 6.9,
2001, Tripos Inc., 1699 South Hanley Road, St. Louis, MO 63144-
2913) running on Silicon Graphics Octane 2 workstations using
standard fragments library. Their geometry was then optimized
roughly by the method of Powell available in the Maximin2
procedure before generating conformational spaces by the CONFEX
method29 based on the distance geometry program DGEOM.30
These databases were minimized with the Maximin2 method and
reduced to their most representative conformations. The heads of
the various conformational families were compared to a pharma-
cophoric model of the SK channels blockers developed previously
at the laboratory. This comparison was carried out by an in-house
macrocommand written in Sybyl Programming Language (SPL),
which also calculated the energy of the conformers using the Tripos
force field31 including the electrostatic term calculated from Del
Re atomic charges. Later, the energy of the entire conformational
space for each compound was calculated following the same method
to find out the lowest energy conformation. The lowest energy
conformer and the best fitting conformation of compound 3a were
aligned manually on the pharmacophore to superimpose the
nitrogens without modifying their geometry.
activated K+ current in hippocampal pyramidal neurons. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 4662-4667.
(3) Stocker, M.; Pedarzani, P. Differential distribution of three Ca2+
-
activated K+ channel subunits, SK1, SK2 and SK3, in the adult rat
central nervous system. Mol. Cell. Neurosci. 2000, 15, 476-493.
(4) Messier, C.; Mourre, C.; Bontempi, B.; Sif, J.; Lazdunski, M.;
Destrade, C. Effect of apamin, a toxin that inhibits calcium-dependent
potassium channels, on learning and memory processes. Brain Res.
1991, 551, 322-326.
(5) Deschaux, O.; Bizot, J. C.; Goyffon, M. Apamin improves learning
in an object recognition task in rats. Neurosci. Lett. 1997, 222, 159-
162.
(6) Ikonen, S.; Riekkinen, P., Jr. Effect of apamin on memory processing
of hippocampal-lesioned mice. Eur. J. Pharmacol. 1999, 382, 151-
156.
(7) Van der Staay, F. J.; Fanelli, R. J.; Blokland, A.; Schmidt, B. H.
Behavioral effects of apamin, a selective inhibitor of the SKCa
channel, in mice and rats. Neurosci. BiobehaV. ReV. 1999, 23, 1087-
1110.
(8) Fournier, C.; Kourrich, S.; Soumireu-Mourat, B.; Mourre, C. Apamin
improves reference memory but not procedural memory in rats by
blocking small conductance Ca2+-activated K+ channels in an
olfactory discrimination task. BehaV. Brain Res. 2001, 121, 81-93.
(9) Pedarzani, P.; Mosbacher, J.; Rivard, A.; Cingolani, L. A.; Oliver,
D.; Stocker, M.; Adelman, J. P.; Fakler, B. Control of electrical
activity in central neurons by modulating the gating of small
conductance Ca2+-activated K+ channels. J. Biol. Chem. 2001, 276,
9762-9769.
(10) Steketee, J. D.; Kalivas, P. W. Effect of microinjections of apamin
into the A10 dopamine region of rats: a behavioural and neuro-
chemical analysis. J. Pharmacol. Exp. Ther. 1990, 254, 711-719.
(11) Shepard, P. D.; Bunney, B. S. Repetitive firing properties of putative
dopamine-containing neurons in vitro: regulation by an apamin-
sensitive Ca2+-activated K+ conductance. Exp. Brain Res. 1991, 86,
141-150.
(12) Seutin, V.; Johnson, S. V.; North, R. A. Apamin increases NMDA-
induced burst firing of rat mesencephalic dopamine neurons. Brain
Res. 1993, 630, 341-344.
(13) Vincent, J. P.; Schweitz, H.; Lazdunski, M. Structure-function
relationships and site of action of apamin, a neurotoxic polypeptide
of bee venom with an action on the central nervous system.
Biochemistry 1975, 14, 2521-2525.
(14) Campos Rosa, J.; Galanakis, D.; Piergentili, A.; Bhandari, K.;
Ganellin, C. R.; Dunn, P. M.; Jenkinson, D. H. Synthesis, molecular
modelling, and pharmacological testing of bis-quinolinium cyclo-
phanes: potent, non-peptidic blockers of the apamin-sensitive Ca2+
-
activated K+ channel. J. Med. Chem. 2000, 43, 420-431.
(15) Scuve´e-Moreau, J.; Lie´geois, J.-F.; Massotte, L.; Seutin, V. Methyl-
laudanosine: a new pharmacological tool to investigate the function
of small-conductance Ca2+-activated K+ channels. J. Pharmacol. Exp.
Ther. 2002, 302, 1176-1183.
(16) Scuve´e-Moreau, J.; Boland, A.; Graulich, A.; Van Overmeire, L.;
D’hoedt, D.; Graulich-Lorge, F.; Thomas, E.; Abras, A.; Stocker,
M.; Lie´geois, J.-F.; Seutin, V. Electrophysiological characterization
of the SK channel blockers methyl-laudanosine and methyl-noscapine
in cell lines and rat brain slices. Br. J. Pharmacol. 2004, 143, 753-
764.
(17) Birch, A. J.; Jackson, A. H.; Shannon, V. R. A new modification of
the Pomeranz-Fritsch isoquinoline synthesis. J. Chem. Soc., Perkin
Trans. 1. 1974, 2185-2190.
(18) Popp, F. D. Developments in the chemistry of Reissert compounds
(1968-1978). AdV. Heterocycl. Chem. 1979, 24, 187-214.
(19) Ruchirawat, S.; Phadungkul, N.; Chuankamnerdkarn, M.; Thebtara-
nonth, C. A versatile synthesis of Reissert compounds. Heterocycles
1977, 6, 43-46.
(20) Popp, F.; Klinowski, C.; Piccirilli, R.; Purcell, D.; Watts, R. Reissert
compound studies. XXIII. Reaction of the anion with some difunc-
tional compounds. J. Heterocycl. Chem. 1971, 8, 313-315.
(21) Dilly, S.; Graulich, A.; Farce, A.; Seutin, V.; Lie´geois, J.-F.; Chavatte,
P. Identification of a pharmacophore of SKCa channel blockers. J.
Enzyme Inhib. Med. Chem. 2005, 6, 517-523.
(22) Galanakis, D.; Davis, C.; Del Rey Herrero, B.; Ganellin, R.; Dunn,
P.; Jenkinson, D. Synthesis and structure-activity relationships of
dequalinium analogues as K+ channel blockers. Investigations on
the role of the charged heterocycle. J. Med. Chem. 1995, 38, 595-
606.
Acknowledgment. The technical assistance of Y. Abrassart,
S. Counerotte, L. Massotte, and J.-C. Van Heugen is gratefully
acknowledged. A.G. is a Research Fellow of the “Fonds pour
la Formation a` la Recherche Industrielle et Agricole (F.R.I.A.)”,
and J.-F.L. is a Senior Research Associate of the “Fonds
National de la Recherche Scientifique” (F.N.R.S.) of Belgium.
This work was financially supported in part by F.N.R.S. Grant
Nos. 3.4525.98 and 9.4560.03 and by grants from Fonds
Spe´ciaux pour la Recherche 2002 and 2003 of the University
of Lie`ge and the Fondation Le´on Fre´de´ricq of the Faculty of
Medicine of the University of Lie`ge.
Supporting Information Available: Routine experimental
procedures (compounds 1c, 2b-r, and 3b-r), spectroscopic data
(compounds 1c, 2b-r, and 3b-r), and elemental analysis results.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(23) Graulich, A.; Mercier, F.; Scuve´e-Moreau, J.; Seutin, V.; Lie´geois,
J.-F. Synthesis and biological evaluation of N-methyl-laudanosine
iodide analogues as potential SK channel blockers. Bioorg. Med.
Chem. 2005, 13, 1201-1209.
(1) Koehler, M.; Hirschberg, B.; Bond, C. T.; Kinzie, J. M.; Marrion,
N. V.; Maylie, J.; Adelman, J. P. Small-conductance, calcium-
activated potassium channels from mammalian brain. Science 1996,
273, 1709-1714.

Bis-isoquinolinium DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 21 5075
(24) Graulich, A.; Scuve´e-Moreau, J.; Alleva, L.; Lamy, C.; Waroux, O.;
Seutin, V.; Lie´geois, J.-F. Synthesis and radioligand binding studies
of methoxylated-1,2,3,4-tetrahydroisoquinolinium derivatives as ligands
of the apamine-sensitive Ca2+-activated K+ channels. J. Med. Chem.
2006, 49, 7208-7214.
(25) Graulich, A.; Scuve´e-Moreau, J.; Seutin, V.; Lie´geois, J.-F. Synthesis
and radioligand binding studies of C-5 and C-8-substituted 1-(3,4-
dimethoxybenzyl)-2,2-dimethyl-1,2,3,4-tetrahydroisoquinoliniums as
SK channel blockers related to N-methyl-laudanosine and N-methyl-
noscapine. J. Med. Chem. 2005, 48, 4972-4982.
(26) Galanakis, D.; Ganellin, C. R.; Chen, J.-Q.; Gunasekera, D.; Dunn,
P. M. Bis-quinolinium cyclophanes: toward a pharmacophore model
for the blockade of apamin-sensitive SKCa channels in sympathetic
neurons. Bioorg. Med. Chem. 2004, 14, 4231-4235.
(27) Uff, B.; Kershaw, J.; Neumeyer, J. Alkylation of isoquinolines Via
2-benzoyl-1,2-dihydroisoquinaldonitriles: 1-benzylisoquinoline. Org.
Synth. 1977, 56, 19-25.
(28) Hartree, E. F. Determination of protein: a modification of the Lowry
method that gives a linear photometric response. Anal. Biochem. 1972,
48, 422-427.
(29) Goldstein, S.; Neuwels, M.; Moureau, F.; Berkmans, D.; Lassoie,
M. A.; Differding, E.; Houssin, R.; He´nichart, J. P. Bioactive
conformations of peptides and mimetics as milestones in drug
design: investigation of NK1 receptor antagonists. Lett. Pept. Sci.
1995, 2, 125-134.
(30) Crippen G. M. Distance Geometry and Conformational Calculations;
Research Studies Press (Wiley): New York, 1981.
(31) Clark, M.; Cramer, R. D., III; Van Opdenbosch, N. Validation of
the general purpose Tripos 5.2 force field. J. Comput. Chem. 1989,
10, 982-1012.
JM070412J