Synthesis, Biological Activity, and Docking Studies of New Acetylcholinesterase Inhibitors of the Bispyridinium Type

Article abstract of DOI:10.1002/ardp.200300795

A novel series of acetylcholinesterase (AChE) inhibitors of the bispyridinium type was synthesized and the inhibitory activity against AChE and butyrylcholinesterase (BChE) measured. In essence, the substitution pattern influenced the inhibitory potency against AChE, where the most active bispyridiniumoxime (TMB-4) was bisbenzyl substituted followed by monobenzyl substituted, bismethyl substituted, and unsubstituted derivatives of TMB-4. Hence, the bisbenzyl ether of TMB-4 was further investigated. In order to obtain diverse lipophilic and electronic properties for these bisbenzyl bispyridinium derivatives (so-called DUO series), the lateral ring substitution was systematically varied. The lowest IC₅₀ value against AChE found thus far in the DUO series was 0.34 μM. Docking studies were carried out to elucidate the differences in biological activity. A general binding mode for nearly all compounds could be identified by these investigations. In this binding mode, the docked ligands span the narrow, deeply buried active-site gorge, interacting with Trp84 at the bottom of the gorge, Tyr334 or Phe331 halfway down the gorge, and Trp279 at the peripheral anionic site at the mouth of the gorge. For specific ligands, additional interactions were found which helped to explain their deviating activity. Based on the promising characteristics of the novel acetylcholinesterase inhibitors presented, a series of structurally related, optimized candidates will be developed.

Full text of DOI:10.1002/ardp.200300795

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 523
Petra Kapkováa,
Synthesis, Biological Activity, and Docking Studies
of New Acetylcholinesterase Inhibitors of the
Bispyridinium Type
Nikolaus Stiefla,
Ulf Süriga,
Bernd Engelsb,
Knut Baumanna,
Ulrike Holzgrabea
A novel series of acetylcholinesterase (AChE) inhibitors of the bispyridinium type
was synthesized and the inhibitory activity against AChE and butyrylcholinesterase
(BChE) measured. In essence, the substitution pattern influenced the inhibitory po-
tency against AChE, where the most active bispyridiniumoxime (TMB-4) was bis-
benzyl substituted followed by monobenzyl substituted, bismethyl substituted, and
unsubstituted derivatives of TMB-4. Hence, the bisbenzyl ether of TMB-4 was fur-
ther investigated. In order to obtain diverse lipophilic and electronic properties for
these bisbenzyl bispyridinium derivatives (so-called DUO series), the lateral ring
substitution was systematically varied. The lowest IC50 value against AChE found
thus far in the DUO series was 0.34 µM. Docking studies were carried out to eluci-
date the differences in biological activity. A general binding mode for nearly all com-
pounds could be identified by these investigations. In this binding mode, the docked
ligands span the narrow, deeply buried active-site gorge, interacting with Trp84 at
the bottom of the gorge, Tyr334 or Phe331 halfway down the gorge, and Trp279 at
the peripheral anionic site at the mouth of the gorge. For specific ligands, additional
interactions were found which helped to explain their deviating activity.Based on the
promising characteristics of the novel acetylcholinesterase inhibitors presented, a
series of structurally related, optimized candidates will be developed.
a
Institute of Pharmacy and
Food Chemistry,
University of Würzburg,
Würzburg, Germany
Institute of Organic
Chemistry,
University of Würzburg,
Würzburg, Germany
b
Keywords: Acetylcholinesterase; Inhibitors; Bispyridinium oxime ethers
Received: March 5, 2003; Accepted May 9, 2003 [FP795]
DOI 10.1002/ardp.200300795
hibitors, such as tacrine and donepezil, and the enzyme
is characterized by cation-π interactions between the
Introduction
While the physiological role of the acetylcholinesterase
(AChE) in the nervous transmission has been known for
some 80 years, this enzyme is still in the focus of phar-
maceutical research. It has been targeted in treatments
of myasthenia gravis, glaucoma, and Alzheimer’s dis-
ease as well as in the recovery of victims of organophos-
phate-type insecticide and nerve agent intoxication, e.g.
by soman, sarin, or tabun.
protonated nitrogens and tryptophan (Trp84) and
phenylalanine (Phe330) residues. Moreover, π-π stack-
ing between the aromatic moieties of the inhibitors and
the conserved aromatic amino acids (Trp84 and
Phe330), as well as ion-ion-interactions between the
protonated nitrogens of the inhibitors and anionic
asparaginic acid (Asp72) play a crucial role (Figure 1) [7].
The aim of this study was to synthesize bisbenzyl ethers
of the bispyridinium-type compound TMB-4, which is
analogous to the reactivator obidoxime, and to evaluate
their potency to inhibit the acetylcholinesterase activity.
The compounds are a kind of ditopic ligands, which were
designed to allow binding to both the active site and the
peripheral site. These sites are characterized by two
tryptophan residues, Trp84 at the active site and Trp279
at the mouth of the gorge. Bis-galanthamine [8], bis-ta-
crine [9, 10], a huperzine A/tacrine hybrid [11] and 1,n-
bispyridiniumalkanes [12] show better inhibitory potency
than their respective monomers.However, in these stud-
ies the inhibitory potency of ditopic ligands was found to
sensitively depend on the distance between the two pos-
itive charges. An equivalent of 8 to 10 methylene groups
X-ray structures of the enzyme co-crystallized with vari-
ous ligands [1–5] provided insights into the essential
structural elements and motifs central to its catalytic
mechanism and mode of acetylcholine (ACh) process-
ing. One of the striking structural features of the AChE
revealed from the X-ray analysis is the presence of a nar-
row, long, hydrophobic gorge which is approximately
20 Å deep.Aromatic residues lining the gorge constitute
a series of binding sites for quaternary ammonium
moieties [6]. The interaction between highly potent in-
Correspondence: Ulrike Holzgrabe, Institute of Pharmacy
and Food Chemistry, University of Würzburg, Am Hubland,
D-97074 Würzburg, Germany. Phone: +49 931 888-5460,
Fax: +49 931 888-5494;
e-mail: U.Holzgrabe@pharmazie.uni-wuerzburg.de
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

524 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
Figure 1. Schematic representation of the active-site of the AChE modified after [7].
produced the lowest IC50 values in a nanomolar range of
concentration. In previous studies, a representative of
the bisbenzylether of the bispyridinium-type com-
pounds, the so-called DUO3 (4.5) (Figure 2, Table 1),
was characterized stereochemically and electrostatical-
ly by means of semi-empirical methods [13]. In an ex-
tended conformation the two positive charges were
found to be delocalized over the entire pyridinium-oxime
moiety. Hence, distances corresponding to the afore-
mentioned 8–10 methylene groups can be easily identi-
fied. Owing to their high flexibility [14] the molecules are
expected to find a number of interaction points (e.g.
π-cation interactions) within the gorge of the active site.
In addition, the bisbenzyl feature may form aromatic in-
teractions (π-π stacking, face-to-face as well as edge-to-
face) with the aromatic residues of the enzyme. Since
substituents attached to the aromatic rings of the benzyl
moieties gradually change the electronic and lipophilic
properties of these rings, they are supposed to influence
the interaction of the entire molecule with the enzyme.
Therefore, it was attempted to achieve a wide variation in
lipophilic and electronic properties according to the
Craig plot (plot σ vs. π) [15]. As a result, halogen atoms,
methyl, nitro, methoxy, and cyano groups were intro-
duced in ortho, meta and para position of the aromatic
ring, respectively. Additionally, the phenyl rings were re-
placed by naphthyl rings. In order to reveal insights into
possible binding modes to AChE and to explain the dif-
ferences in affinity docking studies were performed (cf.
[16]).
Chemistry
The 4,4Ј-bis-[(benzyloxyimino)-methyl)]-1,1Ј-propane-
1,3-diyl-bispyridinium dibromides 4.1–4.26 were syn-
thesized in analogy to the pathway given in Gasteiger
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 525
Figure 2.TMB-4, compound 5, and the oxime mono- (UNO3) and bisbenzylethers 4 studied.For substituents attached
to the lateral phenyl rings, see Table 1a.
et al. and Botero et al. [13, 17].The oxime ethers 3 were
formed via a phase transfer reaction between the pyrid-
ine-4-carbaldehyde oxime 2 and the corresponding ben-
zyl halogenide 1, which can be obtained from the corre-
sponding benzylalcohol with PBr3 or SOCl2, in presence
of phase-transfer catalyst tetramethylammonium chlo-
ride, aqueous sodium hydroxide solution, and dichlo-
romethane, or in presence of sodium methylate/metha-
nol.The bispyridinium compounds 4 can be synthesized
by conversion of two moles of the oxime ethers 3 with
1,3-dibromopropane when refluxing in acetonitrile for
20 h.All compounds were obtained in acceptable yields,
the spectral data were found to be in agreement with the
structures assigned to them (Tables 3 and 4). Very low
amounts (< 2 %) of the corresponding Z-isomer of some
compounds could be identified in the aromatic section of
the NMR spectra by comparison with data reported for
the E and Z isomers of pyridine-4-carbaldehyde oxime
and the corresponding methyl and ethyl ethers [18, 19].
However, applying e.g. TRIS-buffered media of pH 7.4,
bispyridinium oxime ethers were found to be stable [20].
The corresponding compound UNO3 was synthesized
from 3 and bromopropylpyridine-4-carbaldehyde oxime
according to [17]. The 4,4Ј-bis[(2,6-dichlorobenzyloxy-
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

526 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
Scheme 1. Synthesis pathway of compounds 4.
Table 1a. Pharmacological data are the mean SEM.
The measurements were performed as described in the
experimental section.
Table 1a. (continued)
Compd.
Substitution
pattern (R)
IC50 SEM
(µM)
Compd.
Substitution
pattern (R)
IC50 SEM
(µM)
4.20
4.21
4.22
4.23
4.24
4.25
DUO3N
Tacrine
5
2,6-OCH3
2,5-OCH3
2-Br
2,6-F
1-naphthyl
2-naphthyl
11.37 0.73
11.72 2.25
1.15 0.16
2.68 0.36
0.88 0.09
1.36 0.06
0.68 0.20
0.044 0.004
144.50 0.18*
4.1
4.2
4.4
4.5
4.6
4.7
4.8
4.9
4.11
4.12
4.13
4.14
4.15
4.16
4.17
4.18
4.19
H
2-Cl
4-Cl
2,6-Cl*
3,5-Cl
2-CH3
3-CH3
4-CH3
2-CN
3.51 0.70
0.58 0.18
4.46 1.04
0.34 0.05
0.81 0.09
1.70 0.24
2.07 0.31
7.96 1.83
1.32 0.21
1.42 0.34
9.00 0.44
1.37 0.17
0.92 0.08
0.72 0.18
2.81 0.39
1.95 0.25
20.66 1.69
* DUO3 according to ref. [13].
3-CN
4-CN
2-NO2
3-NO2
4-NO2
2-OCH3
3-OCH3
4-OCH3
hydrazono)-methyl)]-1,1Ј-propane-1,3-diyl-bispyridinium
dibromide compound DUO3N was synthesized from di-
bromopropane and 4-pyridinecarbaldehyde-(2,6-di-
chlorobenzyl)-hydrazone, which can be obtain by con-
version of the 2,6-dichlorobenzylhydrazine with 4-pyri-
dinecarbaldehyde, according to [13].
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 527
Table 1b. Selectivity for AChE: IC50 for BChE divided by
IC50 for AchE.
served, respectively (Table 1a), indicating a decrease in
potency towards AChE and, along with this, the impor-
tance of the phenyl rings.
IC50 SEM (µM)
AChE BChE
tacrine 0.044 0.004 0.0044 0.0004
Selectivity
BChE/AChE
When comparing the activity of the corresponding mo-
no- and bis-2,6-dichlorobenzyloxime ethers of TMB-4,
referred to as UNO3 [17] and DUO3 (4.5), UNO3 was
found to be three times less active than 4.5 (Table 1b).
However, UNO3 showed an approximately threefold
higher activity than the unsubstituted bispyridinium com-
pound 4.1. When determining the IC50 value for UNO3
the solution turned yellow during the course of the meas-
urement (see above).Thus, the IC50 value is associated
with a larger experimental error. Nonetheless, these re-
sults show the importance of a substitution of the bispy-
ridinium oximes with regard to the inhibition of the AChE.
0.1
0.52
2.64
DUO3
UNO3
0.34 0.05
0.95 0.13
0.18 0.002
2.51 0.13
Results and discussion
Inhibition of AChE
The inhibitory potency against AChE was evaluated by
means of Ellman’s test [21].The IC50 values are shown in
Tables 1a and 1b. For sake of comparison, the activity of
tacrine was also determined. The IC50 values of tacrine
reported in the literature [22, 23, 24, 25] vary between
0.0098 and 0.05 µM. The value determined here falls in
the same range (see Table 1b). Three compounds, 4.3,
4.10, and 4.26, could not be dissolved at suitable con-
centrations neither in water nor in DMSO aqueous solu-
tions. Hence, no complete dose-effect curve could be
obtained and consequently no IC50 values are given.
Exchanging the oxime oxygen for a nitrogen (DUO3 ver-
sus DUO3N, Figure 2, Table 1a) did not improve the in-
hibitory potency; the activity of the N-derivative is two
times lower than that of the O-derivative.
The IC50 values of the bisbenzyl substituted compounds
4.1-4.26 are compiled in Table 1a. A qualitative inspec-
tion of the IC50 values revealed the halogenated com-
pounds to be the most active ones and the chlorine sub-
stitution to be superior to fluorine and bromine substitu-
tion (4.2/4.22 and 4.5/4.23). In contrast, compounds
bearing one or two methoxy groups are less active, es-
pecially in the case of a para substitution.With the excep-
tion of nitro substituents, substituents in para position
seem to be disadvantageous.
In a first step, it was tried to find out whether the benzyl
residues are essential for the inhibition of the AChE.
Therefore, the IC50 value of the bisoxime TMB-4 lacking
the benzyl residues was determined.Sikder and cowork-
ers [26] already reported a 30 percent inhibition of AChE
when applying 2 × 10–3 MTMB-4.Here, a 31 percent inhi-
bition of the AChE was obtained when applying 1 × 10–2
M TMB-4, and a 14 percent inhibition when applying 1 ×
10–3 M TMB-4. For higher concentrations of TMB-4 in al-
kaline solution (pH 8) a troublesome absorption oc-
curred with a maximum in the range 330-360 nm (“betain
band”), giving a yellow colored solution. This effect has
already been reported by Inkmann et al. [20] and
Schoene [27]. It was assumed that the oxime function is
deprotonated in basic medium and a formation of an
intermolecular charge-transfer complex takes place.
Since the color caused by this complexation interferes
with Ellman’s test, no IC50 value for TMB-4 could be de-
termined. However, a preliminary analysis of the data in-
dicates a very low inhibitory potency of the bispyridinium
skeleton. Vice versa, they indirectly pinpoint the impor-
tance of substituents attached to the oxime groups.
Inhibition of butyrylcholinesterase
The butyrylcholinesterase (BChE), also called nonspe-
cific cholinesterase or pseudocholinesterase, is able to
act on hydrophilic and hydrophobic choline esters. Even
though BChE is closely related to AChE, owing to a larg-
er acti2ve site gorge of BChE a broader variety of sub-
strates and inhibitors is accepted as compared to AChE
[28].
Currently, the exact physiological role of BChE is not well
understood, but it is already known that this enzyme hy-
drolyses a variety of xenobiotics [29]. In order to avoid
the potential for side effects selective AChE inhibitors
are wanted. Therefore, the second objective addressed
was the examination of the activity against BChE. This
was investigated with three representative compounds
(TMB-4, 4.5, and UNO3).
In order to check whether the phenyl rings of the benzyl
ethers are necessary for a high inhibitory potency, the
bismethyloxime ether ofTMB-4 (5) [20], was additionally
studied. In comparison to DUO1 (4.1), DUO3 (4.5) and
UNO3 a 41-fold, 425- and 152-fold higher IC50 was ob-
As expected, our investigation showed that no inhibition
of BChE could be measured for the unsubstituted bisoxi-
me derivative TMB-4. The second compound tested for
inhibition of BChE was 4.5, since its AChE inhibitory ac-
tivity was the highest of the series investigated. Com-
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

528 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
pared to its AChE activity, 4.5 proved to be slightly more
potent on BChE. In contrast, the monosubstitution of
bisoxime (i.e.UNO3) greatly reduced the activity against
BChE, but not against AChE (Table 1b).
The second postulate refers to the binding position of the
compounds under scrutiny. Unfortunately, when using a
generic docking protocol, usually not all of the docked
conformations end up inside the active site cavity [31].
Structural analysis of the available crystallized complex-
es from PDB led to the assumption that the bisbenzyl
substituted TMB-4 compounds (4) interact at the bottom
of the gorge with additional contacts along the gorge and
the gorge entrance. Consequently, the top scoring con-
formation accepted for interpretation purposes needed
to be in an extended shape inside the active site cavity.
Folded conformations or structures docked to the out-
side of the gorge were discarded.
Docking studies
At a first glance, the IC50 values measured span a rea-
sonable activity space. However, when investigating the
data more closely, it can be seen that they are clustered
in two subsets. The smaller cluster comprises the less
active compounds 4.19, 4.20, and 4.21. Even though, i it
would be possible to build a QSAR model with this type
of data, this was not done here.Statistical figures of merit
(Q2, R2) tend to be misleadingly overoptimistic with clus-
tered data, and the resulting models often simply dis-
criminate between the clusters. Consequently, this may
reduce the understanding of the factors determining bio-
logical activity. Hence, docking studies investigating the
possible binding modes of the compounds in the binding
pocket were carried out.These experiments should help
to explain the differences in activity of the compounds
tested.
The third postulate refers to the conformational flexibility
of the docked compounds. Since the available degrees
of freedom of the docked compounds have a huge im-
pact on the docking results, usually all single bonds are
defined to be rotatable and all others as rigid. For the
bispyridinium compounds this definition is not applicable
without modification. The question arises whether the
pyridinium-C4 oxime-C single bond can freely rotate, or
if the rotation is hindered due to the delocalization be-
tween the pyridinium ring and the oxime double bond.
Computations employing the B3LYP functional [33, 34]
in combination with a 6-31G* basis set [35] using the
package GAUSSIAN98 [36] for one half of compound 5
gave an energy difference of about 10 kcal/mol between
the planar conformation and the geometry obtained by a
90° clockwise rotation around the single bond where the
planar conformation was energetically favorable. In the
90° rotated non-planar conformation the oxime double
bond stays perpendicular with respect to the pyridinium
ring, i.e.delocalization cannot take place.If the rotation is
performed anti-clockwise the energy difference is even
higher.Consequently, the single bond connecting the py-
ridinum ring and the oxime was fixed in the planar confor-
mation.
In order to achieve the stated goals, a highly generic
docking protocol was employed to extract the most com-
mon features of the wide variety of possible binding
modes. The Brookhaven Protein Data Bank (PDB) [30]
contains several small molecule-AChE complexes.
Some of them were already used for docking experi-
ments, including the Torpedo californica AChE
(TcAChE) complex with decamethonium (1ACL), edro-
phonium (2ACK), tacrine (1ACJ), and huperzine A
(1VOT), respectively [31, 32].Prior to the docking studies
three assumptions were made.The first postulate refers
to the orientation of Phe330.This side chain controls the
access to the bottom of the gorge and was identified to
adopt three major conformations, an open, a closed, and
an intermediate access position, respectively [31]. The
closed position was identified for the native enzyme as
well as for the complexes with huperzine A, edrophoni-
um, and the transition state analog m-(N,N,N-trimethyl-
ammonio)trifluoro-acetophenone (1AMN) [3]. All these
compounds bind close to the bottom of the gorge. A dif-
ferent behavior can be found for tacrine. Here, Phe330
adopts an intermediate access conformation which al-
lows a ‘sandwich-like’ arrangement with tacrine and
Trp84. For gorge-spanning ligands such as donepezil
(1EVE) [5] and decamethonium Phe330 adopts an open
access position pointing towards the gorge wall. Owing
to their size and shape a gorge-spanning binding mode
for the tested compounds was assumed. Therefore, the
crystal structure of co-crystallized donepezil (1EVE)
was used during this study.
Currently several docking tools are available (e.g.DOCK
[35], GOLD [36], AUTODOCK [37], FlexX [37]) of which
DOCK, FlexX, and AUTODOCK were selected for pre-
liminary studies.Results of these studies were compara-
ble for all three programs. Since AUTODOCK was al-
ready successfully applied for published docking studies
with AChE [31, 32], it was selected here.
Investigation of the active site of AChE allows the identifi-
cation of one peripheral anionic site (PAS) and at least
five major binding sites: the oxyanion hole (OH), the es-
teratic site (ES), the anionic substrate binding site (AS),
the active site-selective aromatic binding site (AACS),
and the acyl binding site (ACS) [38, 39]. Due to this wide
variety of possible binding sites, docking results were ex-
pected to be highly diverse. Since no general statement
on conserved water molecules are available [40], all wa-
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 529
ter molecules were deleted.This adds a further degree of
freedom in terms of variability in the docking results. For
these reasons, the ligand-enzyme complexes (in the fol-
lowing referred to as docking configurations) obtained
using AUTODOCK were surprisingly homogenous. In-
spection showed that one overall binding mode could be
identified for nearly all docked compounds among the
three best scoring docking configurations. This binding
mode incorporates mainly a π-π stacking interaction of
the benzyl substituent and Trp84, and a face-to-face in-
teraction (π-π and cation-π, respectively) of the pyridin-
ium ring with either Tyr334 or Phe331 which is found for
nearly all compounds (Figure 3).
binding modes is the location of one of the benzyl rings
outside the active site cavity. In order to interact with the
protein surface this part of the molecule needs to be
moved in a sterically unfavorable manner.Therefore, it is
expected that interactions with this substructure will
have little or no effect on the binding of the ligand to the
enzyme, since the intramolecular penalty will overcom-
pensate the gain in interaction energy. This observation
is experimentally supported by the activity measured for
UNO3 (IC50 = 0.95 0.13 µM) compared to the IC50
=
0.34 0.22 µM for DUO3 (4.5) which is not significantly
different (Student’s t-test, 5 % error of the first kind, two-
sided).
Single compounds allowed identification of interesting
additional binding patterns. These are displayed in Fig-
ures 5 to 7.The nitro substituted compounds (4.14–4.16)
are able to build hydrogen bonds to different amino acid
residues at the bottom of the gorge. The latter include
Ser200, Tyr334, Ser122, andTyr130, of which especially
Tyr130 or Ser200 show up quite frequently (Figure 5).
An additional feature which can be identified in about
60 % of the docked structures is another face-to-face in-
teraction of the second pyridinium ring and Trp279 (Fig-
ure 4).
For the latter two interactions, it can be seen that, the
face-to-face orientation is not as perfect as for Trp84.
This finding can be explained in two different ways.First,
the shape of the gorge in the vicinity ofTrp279 is far wider
than halfway down the gorge where Phe331 and Tyr334
are located.Consequently, more space is available to the
ligand, which in turn allows more conformational flexibili-
ty. Second, possible interactions for the pyridinium rings
of the ligands can be split into π-electron interactions and
cationic interactions of which the latter are much less
spatially directed. Owing to this, parallel orientation of
the ring systems is less pronounced when compared to
pure π-π interactions. Another common feature of these
Keeping in mind that activity was measured against
Electrophorus electricus AChE (EeAChE) where
Phe330 is exchanged for Tyr337 one can think of the for-
mation of additional hydrogen bonded binding modes.
These hydrogen bonds might be one reason for the en-
hanced activity of nitro substituted compounds. This is
supported by the fact that these bonds are mainly
formed by the 3-nitro (4.15) and the 4-nitro (4.16) substi-
tuted bisbenzylethers, whereas for 4.14, which is the 2-
nitro substituted compound, hydrogen bonds show up
less frequently in the docking configurations.
Figure 3.General binding mode found for nearly all compounds of the data set.Ligands shown are 4.22 (right) and 4.2
(left), respectively. For reasons of clarity only the part of the ligand which is found inside the active side gorge is dis-
played.
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

530 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
Figure 4. Binding mode of 4.23: the additional face-to-
face interaction with the second pyridinium ring and
Trp279 was found for roughly 60 % of the docked struc-
tures (here 4.23).
Figure 5. Binding mode of the meta nitro substituted
compound 4.18: hydrogen bonds are displayed as
dashed lines.
Moreover, the bismethoxy substituted compounds (4.20,
4.21) do not seem to bind deep enough in the active site
gorge (Figure 6).In nearly all top scoring docking config-
urations, the bismethoxybenzyl structures form a π-π
stacking with Phe330, which is not located at the bottom
of the gorge, as opposed to Trp84. Consequently, half of
the molecule lies outside the active side gorge. This
might be an explanation for their reduced activity. This
possible different binding mode of the latter structure
supports the decision not to build a 3D-QSAR model for
the data set, since 3D-QSAR studies are based on the
assumption of a similar binding mode for all compounds.
tended shape with the bischlorobenzyl ring forming a π-π
stacking interaction with Trp84 (Figure 7).
Here, the location of the bischlorobenzyl ring is found to
be closely located to the phenyl ring of donepezil in the
crystallized complex.Even though a chloro substituent is
expected to withdraw electrons from an aromatic ring
system and hence should affect adversely the stacking
of bischloro benzyl ring with Trp84, a similar finding was
published for the crystal structure of huprine X [41].Ana-
logs of huprine X showed a strong decrease in affinity to
AChE when lacking a chloro substituent. The crystal
structure allowed interpretation of this behavior. The
chloro atom of huprine X analogs either has close spatial
contact to neighboring amino acid residues or modu-
lates the π-π stacking interaction of the ligand's aromatic
ring and the aromatic residues of the anionic site at the
bottom of the gorge.This π-π stacking modulation seems
In contrast to the docking configurations found for com-
pounds 4.20 and 4.21, chloro substitution of the tested
bispyridinium compounds seems to further enhance the
orientation of the benzyl rings. This becomes obvious
when inspecting the docking configurations of 4.5 and
4.6. All three top scoring configurations display an ex-
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 531
Figure 6. Binding mode of the bismethoxybenzyl ether 4.20.
Figure 7. Bishalogenation of the benzyl substituents (here, 4.5 is shown) seems to force this ring to orient parallel to
Trp84.This results in an extended shape of the ligands which may explain their enhanced activity.(a) Extended shape of
the ligand. (b) Top view and (c) side view of the bishalogenated benzyl ring system (purple) and the phenyl ring of
donepezil as found in the crystal structure (green).
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

532 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
to be very likely here since the chloro atoms arrange di-
rectly parallel to the ring system of Trp84 (Figure 7).
may, in addition, explain the low activity of 4.20 and 4.21
which ess an extremely bulky bismethoxybenzyl residue
on both ends of the molecule.The results obtained so far
will be used to develop novel second generation AChE
inhibitor candidates structurally related to the tested
compounds but being shorter than the series of com-
pounds evaluated in this study. Corresponding studies
are under way.
As a validation of the docking procedure the structure of
donepezil was redocked back into the active site of the
enzyme with all water molecules deleted. Six out of the
ten top scoring docking configurations resembled almost
exactly the crystallized donepezil-enzyme complex.This
allowed the conclusion that the results obtained with the
applied docking procedure are meaningful.
Acknowledgements
Thanks are due to the KAAD for grant given to P. K. and
the Fonds der Chemischen Industrie for financial sup-
port, to Dr. Eberhard Heller and Mrs. Liana Pogorelaja,
Institute of Pharmacy, University Würzburg, for the syn-
thesis of tacrine, 4.1, 4.5, and DUO3N compounds.
Moreover, we would like to thank Julian Henn for help
with the quantum chemical calculations, Dr. Imran M.
Siddiqi (Department of Physiological Chemistry II, Uni-
versity Würzburg, Germany) for making possible the
FlexX calculations, and Dr. Fabio Zuccotto (Inpharmati-
ca, London, UK) for discussions of the docking proto-
cols.
Conclusion
The order of the AChE inhibitory potency between un-
substituted, bismethylsubstituted, monobenzylsubsti-
tuted and bisbenzylsubstituted bispyridinium com-
pounds was as follows:TMB-4 < 5 UNO3 < 4.5.In gener-
al, the inhibitory activity of the DUO compounds is in the
lower micromolar range with halogenated compounds
being the most active ones. The inhibitory activity de-
pends sensitively on the substitution pattern on the ben-
zyl rings.This experimental finding can be explained by
docking studies whose goal was to thoroughly explore
possible binding modes of the bispyridinium-type com-
pounds. One major observation is that almost all of the
compounds docked display a general binding mode.
However, not all compounds are likely to interact in the
same manner with the protein. In particular, the less ac-
tive compounds 4.20 and 4.21 exhibit a different binding
mode in the docking studies, which supports the earlier
finding that the data are not suitable for meaningful 3D-
QSAR studies.
Experimental
General methods
Melting points were determined with Dr.Tottoli melting point ap-
paratus (Büchi, Switzerland) and were not corrected. 1H- and
13C-NMR spectra were recorded on aVarian XL 300 instrument
(Varian, Palo Alto, CA, USA) (1H 299.956, 13C 75.433 MHz).
The centers of the peaks of CDCl3 and DMSO-d6 were used as
internal references. Abbreviations for data quoted are: d, dou-
blet; dd, double doublet, etc; dt, doublet of triplet; t, triplet; qu,
quintet;m, multiplet;br, broad signal;φ, phenyl.Infrared spectra
were obtained using a Perkin-Elmer 298 spectrometer (Perkin-
Elmer, Norwalk, CT, USA). Dry solvents were used throughout.
Chemicals were of analytical grade, purchased from Aldrich
(Steinheim, Germany) or Acros (Schwerte, Germany).DUO3N
was synthesized according to ref. [13], UNO3 according to ref.
[17], TMB-4 and the 4,4Ј-bis-[(methyloxyimino)-methyl)]-1,1Ј-
propane-1,3-diyl-bispyridinium 5 according to ref. [20], and ta-
crine (1,2,3,4-tetrahydro-9-aminoacridine hydrochloride) ac-
cording to ref. [42].
The interactions found after docking include π-π stacking
and cation-π contacts with amino acid residues of the
anionic substrate binding site (Trp84, Phe331, and
Tyr334) and the peripheral anionic binding site (Trp279).
Additionally, hydrogen bonds to amino acid residues at
the bottom of the gorge are observed.These types of in-
teraction are already known for other crystallized inhibi-
tors of AChE (e.g. donepezil, galanthamine) and there-
fore the docking results are likely to be meaningful for the
tested compounds. Moreover, all compounds are poten-
tially able to bind inside the active side gorge, although,
not the whole molecule is able to interact with amino acid
residues of the enzyme.This‘size problem’of the ligands
may be one reason for their reduced activity as com-
pared to other AChE inhibitors.McCammon and cowork-
ers addressed this problem with extensive molecular dy-
namics studies [40]. These authors were able to show
that the primary entrance to the gorge of AChE is work-
ing as a bottleneck which is not likely to hinder the bind-
ing of ACh significantly.Larger ligands, however, may es-
cape by diffusion before fluctuations open the bottleneck
wide enough to allow binding. This ‘dynamic selectivity’
General procedure for synthesis of 4-[(benzyloxyimino)me-
thyl]-pyridine
The oxime ether 3 were synthesized using two general proce-
dures:
Synthesis via phase-transfer catalysis (variant A)
The corresponding benzylhalogenide 1 (50 mmol) was dis-
solved in CH2Cl2 (50 mL) and pyridine-4-aldoxime 2 (7.63 g,
62.5 mmol), NaOH (9.0 g, 225 mmol), and tetrabutylammoni-
um chloride (0.69 g, 2.5 mmol) in H20 were added.The solution
was refluxed stirring vigorously at room temperature in dark-
ness for 15 h.Afterwards the phases were separated, the aque-
ous layer was extracted three times with CH2Cl2. The organic
layers were dried over Na2SO4 and evaporated in vacuo. After
purification by means of column chromatography (column:
1000 g silica gel, length = 30 cm, ID = 10 cm; mobile phase
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 533
Acetylcholinesterase (AChE)/butyrylcholinesterase (BChE)
solution
CH2Cl2/ethylacetate 5:2) crystals of each compound were ob-
tained.
Synthesis with NaOCH3/MeOH (variant B)
The enzyme (500 units) was dissolved in gelatin solution (1 mL,
1%) and taken into a 100 mL volumetric flask, made up with wa-
ter to c = 5 units/mL.This stock solution was stored in frozen al-
iquots (0.7 mL) at –30 °C and, prior to use, was thawed and di-
luted to its final concentration c = 2.5 units/mL.
Pyridin-4-carbaldoxime 2 (6.11 g, 50 mmol) and NaOCH3
(2.7 g, 50 mmol) were dissolved in boiling non-aqueous MeOH
(150 mL), then the solution of benzylhalogenide (50 mmol) in
non-aqueous MeOH (30 mL) was added. After refluxing for 3 h
the solution was cooled to room temperature and evaporated in
vacuo. A small amount of CH2Cl2/ethyl acetate (5:2) was add-
ed, the crystals of sodium halogenide salt were filtered. For the
purification see variant A.For analytical and spectroscopic data
of the compounds 3, see Table 2 and Table 4.
Acetylcholinesterase/butyrylcholinesterase activity assay
Enzyme activity was investigated using a slightly modified
colorimetric method of Ellman et al. [21]. As the product of the
enzymatic hydrolysis, the thiocholine, does not possess a sig-
nificant chromophore for UV detection, the evaluation of en-
zyme activity was performed using a specific chromogenic rea-
gent, DTNB. The measurements were carried out as follows:
Stock solutions of the inhibitor compounds were prepared in
2 % DMSO. The enzyme activity was determined in the pres-
ence of at least five different concentrations of an inhibitor, gen-
erally between 10–4 and 10–9, in order to obtain inhibition of
AChE or BChE activity between 0 and 100 %. Each concentra-
tion was assayed in triplicate. The samples were investigated
immediately after preparation.
Synthesis of 4,4Ј-bis-[(benzyloxyimino)-methyl)]-1,1Ј-pro-
pane-1,3-diyl-bispyridinium dibromides 4
1,3-Dibromopropane (1.01 g, 5 mmol) and the corresponding
oxime ether 3 (10 mmol) were dissolved in non-aqueous ace-
tonitrile (30 mL) and refluxed for 20 h.After cooling to room tem-
perature a precipitate was formed and filtered; in the case no
precipitate appears a small amount of diethyl ether was added.
The so obtained crystals were recrystallized from MeOH/
diethyl ether. For analytical and spectroscopic data for com-
pounds 4, see Table 3 and Table 5.
Inhibition study
Biological assay: Ellman’s test
Chemicals and materials
Prior to use, all solutions were adjusted to 20 °C. Enzyme solu-
tion (100 mL) and inhibitor solution (100 µL) were added into a
cuvette containing the phosphate buffer (3.0 mL, 0.1 M; pH =
8.0).After 5 min incubation, required aliquots of the DTNB solu-
tion (100 mL) and of the acetylthiocholine/butyrylthiocholine io-
dide (20 µL) were added. After rapid and immediate mixing the
absorption was measured at 412 nm. As a reference, an identi-
cal solution of the enzyme without the inhibitor is processed fol-
lowing the same protocol.The blank reading contained 3.0 mL
buffer, 200 µL water, 100 µL DTNB, and 20 µL substrate.
Acetylcholinesterase (AChE) – E.C.3.1.1.7., Type VI-S, from
Electric Eel, 500 units – was purchased from Sigma-Aldrich
(Steinheim, Germany). 5,5Ј-Dithiobis(2-nitrobenzoic acid)
DTNB – Ellman’s reagent, buffer compounds (potassium dihy-
drogen phosphate, potassium hydroxide), sodium hydrogen
carbonate and acetylthiocholine iodide (ATC) used as a sub-
strate were obtained from Fluka (Buchs, Switzerland). Deion-
ized water prepared by means of a Milli-Q purification appara-
tus (Millipore®, Eschborn, Germany) was used throughout.
Spectrophotometric measurements were performed on a
Varian Cary 50 UV-Vis spectrophotometer (Varian-equipped
with a thermostated cell holder).
The relative enzyme activity was calculated according to the
equation:
activity [%] = Mcorr × 100 / Rcorr
where Mcorr is the enzyme activity in the presence of the inhibi-
tor, Rcorr is the enzyme activity of the reference solution (without
the inhibitor). Both of the values are corrected with blank-read-
ing value. The results were analyzed with the program Graph-
Pad PrismTM (GraphPad Software, San Diego, CA, USA) using
non-linear regression analysis toward a sigmoid dose-re-
sponse (variable slope) model. Inhibition curves were obtained
for each compound by plotting the % enzyme activity in the
presence of the inhibitor versus the logarithm of inhibitor con-
centration in the assay solution. For each inhibition curve four
variables were determined: IC50 (the log[Dose] when response
= 50 %), hill slope, the bottom plateau, and the top plateau.
Solutions: phosphate buffer pH 8.0
Potassium dihydrogen phosphate (13,61 g) was dissolved in
water (1L) and adjusted with potassium hydroxide to a pH of 8.0
0.1. The buffer was filtered through 0.22 µm (pore size) dis-
posable filter units (Schleicher and Schuell, Dassel, Germany)
prior to use. The buffer was freshly prepared every week and
kept at 4 °C.
5,5Ј-Dithiobis(2-nitrobenzoic acid) DTNB solution (0.01M
The solution (100 mL) was prepared in water, containing DTNB
(0.396 g) and sodium hydrogen carbonate (0.15 g). The solu-
tion was directly used or placed in the –30 °C freezer until need-
ed.
Molecular modeling studies
Enzyme preparation
Crystal structures were obtained from the PDB. Since no lig-
and-enzyme complexes for Electrophorus electricus AChE
(EeAChE) are available at the PDB, the correspondingTorpedo
californica AChE (TcAChE) complexes were utilized.The most
important difference between these two enzymes is the re-
placement of Phe330 in TcAChE with Tyr337 in EeAChE. Both
residues belong to the group of gorge lining hydrophobic amino
acids which probably interact with the tested compounds.Even
though this might result in different absolute values for the ac-
tivity measured, docking results should be nearly equivalent,
since the hydrophobic character of the gorge is characterizing
this area of the enzyme.
Acetylthiocholine iodide (ATC) 0.075M
0.217 g ATC were weight and dissolved in 10 mL water.This so-
lution was stored in frozen aliquots (0.4 mL) in Eppendorf-caps
at –30 °C.
Butyrylthiocholine iodide (BTC) 0.075M
0.237g BTC were weight and dissolved in 10 mL water.The so-
lution was stored in frozen aliquots (0.4mL) in Eppendorf-caps
at –30 °C.
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

534 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
box is chosen too large, ligands may be placed outside the ac-
tive site. Initially the center of the active site box was positioned
at the center of mass of the co-crystallized ligand (donepezil).
In order to enclose the whole active site, including the catalytic
pocket, the gorge and the peripheral binding site, the center
was shifted slightly toward the periphery of the enzyme.The di-
mensions of the active site box were set to 24 Å × 24.6 Å ×
26.4 Å with a gridding of 0.3 Å.Atomic affinity maps for all atom
types included in the set of ligands as well as electrostatic maps
were calculated. Initial translation (tran0) of the ligand was set
to the center of the precalculated grids. The cluster tolerance
(rmstol) was set to 0.5 Å and the number of genetic algorithms
runs (ga_run) was set to 30.The local search was done with the
implemented pseudo-Solis and Wets algorithm. All other set-
tings were used as set by the mkdpf3 program.
Enzyme preparation for all docking studies was similar.Starting
conformation of the enzyme was the donepezil-enzyme com-
plex (1EVE).Various, already successfully applied docking pro-
tocols [31, 32] showed that docking of the crystallized ligand
back into its enzyme with co-crystallized water molecules de-
leted reproduced the originally crystallized geometry well.
Therefore, all water molecules in the crystal structure were de-
leted.The ligand was extracted in order to redock it in the follow-
ing studies as a validation compound.Using the BIOPOLYMER
module of SYBYL [44], polar atoms only were added and AM-
BER charges were loaded for AUTODOCK enzyme prepara-
tion.Hydrogens at histidine residues were positioned at the ep-
silon nitrogen. Serine residues were considered to be neutral,
whereas all basic and acidic residues were considered proto-
nated and deprotonated, respectively.This structure was mini-
mized for 200 steps with the steepest descent minimizer as im-
plemented in SYBYL with the backbone atoms kept at fixed po-
sitions. Since AUTODOCK is using a grid-based approach for
the calculation of atomic affinity maps and electrostatic maps
for the reduction of computational burden, the placement of the
grid is of great importance for the results obtained. On the one
hand, this grid, which is also called the active site box, should
comprise all amino acid residues relevant for binding. On the
other hand, it should be as small as possible under the con-
straint that all ligands can be accommodated. If the active site
Ligand preparation
The compounds were sketched in 2D and then converted to 3D-
structures using CONCORD as implemented in Alchemy 2000
[43]. Next, a full geometry optimization using the MMFF94 force
field of SYBYL 6.8 [44] (max. iterations: 500, dielectric constant:
3.0) was carried out. Since all employed docking programs allow
flexible docking of the compounds, no conformational search
was employed to the ligand structures. AM1 [45] charges were
assigned to the set of compounds prior to docking.
Table 2. The analytical data of compounds 3.
Compound Amount of starting
Yield
mp
(°C)
Formula
(mol wt)
IR (ν, cm–1)
substituent
comp. 1/Variant
3.1
H
6.33 g (50 mmol) 7.43 g (70 %) yellow-brown oil C13H12N2O
3080; 3030; 2940; 1600; 155;
1420; 1365; 820; 760; 700 [Film]
1.1-Cl/A
(212.3)
3.2
2-Cl
8.05 (50 mmol)
1.2-Cl/A
11.40 g (92 %) 39–40 °C
(aceton)
C13H11ClN2O 3040; 3000; 2940; 1600; 155;
(246.7) 815; 755 [KBr]
3.3
3-Cl
10.27 g (50 mmol) 9.78 g (79 %) 38–40 °C
1.3-Cl/A (aceton)
C13H11ClN2O 3070; 2930; 1600, 1410; 135;
(246.7) 1210; 790; 710 [KBr]
3.4
4-Cl
8.05 g (50 mmol) 10,30 g (84 %) 79–80 °C
1.4-Cl/A (aceton)
C13H11ClN2O 2920; 1605; 1490; 1400; 135;
(246.7) 1010; 920; 800 [KBr]
3.5
2,6-Cl
9.77 g (50 mmol) 12.37 g (88 %) 78–82 °C
1.5-Cl/A (Et2O)
C13H10Cl2N2O 3030; 2930; 1600; 1440; 940;
(281.2) 810; 790; 730 [KBr]
3.6
3,5-Cl
12.00 g (50 mmol) 2.36 g (17 %) 46 °C
C13H10Cl2N2O 3030; 2940; 1590; 1570; 143;
1.6-Br/A
(aceton)
(281.2)
820; 795; 700 [KBr]
3.7
2-CH3
7.03 g (50 mmol) 6.31 g (56 %) 36 °C
C14H14N2O
(226.3)
3030; 2940; 2880; 1595; 141;
815; 770 [KBr]
1.7-Cl/A
(aceton)
3.8
3-CH3
4.22 g (30 mmol) 5.05 g (74 %) yellow-brown oil C14H14N2O
3020; 2920; 1600; 1410; 810;
780; 700 [Film]
1.8-Cl/A
(226.3)
3.9
4-CH3
9.25 g (50 mmol) 3.63 g (32 %) 74–75 °C
C14H14N2O
(226.3)
3020; 2920; 2880; 1600; 141;
1230; 810; 760 [KBr]
1.9-Br/A
(aceton)
3.10
5.06 g (30 mmol) 4.83 (63 %)
79–81 °C
(Et2O)
C16H18N2O
(254.4)
3000; 2970; 2920; 1600; 142;
1350; 940; 810 [KBr]
2,4,6-CH3 1.10-Cl/A
3.11
2-CN
9.80 g (50 mmol) 6.98 g (59 %) 84–86 °C
1.11-Br/B (Et2O)
C14H11N3O
(237.3)
3040; 2960; 2890; 2220; 160;
1220; 810; 770 [KBr]
3.12
3-CN
9.80 g (50 mmol) 8.18 g (69 %) 85–87 °C
1.12-Br/A (aceton)
C14H11N3O
(237.3)
3040; 2890; 2220; 1600; 143;
1230; 810; 795 [KBr]
3.13
4-CN
9.80 g (50 mmol) 8.16 g (69 %) 106–108 °C
1.13-Br/A (aceton)
C14H11N3O
(237.3)
3070; 2940; 2220; 1600; 159;
1415; 1235; 960; 815 [KBr]
3.14
2-NO2
8.58 g (50 mmol) 5.94 g (46 %) 85–87 °C
1.14-Cl/A (aceton)
C13H11N3O3 3100; 2930; 1600:1520; 134;
(257.3) 810; 795; 725 [KBr]
3.15
3-NO2
8.58 g (50 mmol) 10.94 g (85 %) 86–88 °C
1.15-Cl/A (aceton)
C13H11N3O3 2870:1600; 1525; 1345; 815;
(257.3) 810; 725 [KBr]
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 535
Table 2. (continued)
Compound Amount of starting
Yield
mp
(°C)
Formula
(mol wt)
IR (ν, cm–1)
substituent
comp. 1/Variant
3.16
4-NO2
8.58 g (50 mmol) 3.45 g (27 %) 92–94 °C
1.16-Cl/A (aceton)
7.83 g (50 mmol) 3.50 g (29 %) 42–44 °C
1.17-Cl/A (aceton)
7.83 g (50 mmol) 9.98 g (82 %) 36 °C
C13H11N3O3
(257.3)
2930; 1600; 1520; 1345; 104;
810; 800; 735 [KBr]
3.17
2-OCH3
C14H14N2O2
(242.3)
2940; 2840; 1600; 1490; 146;
1245; 815; 755 [KBr]
3.18
3-OCH3
C14H14N2O2
(242.3)
3020; 2930; 2830; 1600; 145
1265; 815; 780; 745; 695 [KB]
1.18-Cl/A
(aceton)
3.19
4-OCH3
7.83 g (50 mmol) 7.90 g (65 %) 66–67 °C
C14H14N2O2
(242.3)
3030; 3000; 2950; 1610; 160;
1510; 1250; 820 [KBr]
1.19-Cl/A
8.40 g (45 mmol) 5.05 g (41 %) 108 °C
1.20-Cl/A (aceton)
9.33 g (50 mmol) 5.49 g (40 %) 55–56 °C
(aceton)
3.20
2,6-OCH3
C15H16N2O3
(272.3)
2940; 2840; 1600; 1475; 125;
815; 780; 740 [KBr]
3.21
2,5-OCH3
C15H16N2O3
(272.3)
2920; 2840; 1600; 1500; 122;
815; 800; 740 [KBr]
1.21-Cl/A
(aceton)
6.54 g (75 %) 63 °C
(Et2O)
3.22
2-Br
7.5 g (30 mmol)
1.22-Br/A
C13H11BrN2O 2940; 1600; 1440; 1360; 124;
(291.2) 810; 750 [KBr]
C13H10F2N2O 3030; 2940; 1590; 1570; 143;
3.23
2,6-F
8.13 g (50 mmol) 10.18 g (82 %) 65 °C
1.23-Cl/A (Et2O)
8.83 g (50 mmol) 8.39 g (64 %) 90 °C
1.24-Cl/A (Et2O)
6.63 g (30 mmol) 5.20 g (66 %) 88–90 °C
1.25-Br/A (CH2Cl2)
11.34 g (50 mmol) 10.33 g (64 %) 134°–136 °C
1.26-Cl/A (312.4)
(248.3)
1230; 810; 790; 730 [KBr]
3.24
1-naphthyl
C17H14N2O
(262.3)
3030; 2930; 1600; 1460; 140;
1215; 800; 770 [KBr]
3.25
2-naphthyl
C17H14N2O
(262.3)
3050; 2940; 2880; 1600; 142;
1235; 830; 820; 755 [KBr]
3.26
9-anthranyl
C21H16N2O
(312.4)
3060; 1600; 1400; 1330; 815;
725 [KBr]
Table 3. Reaction and physical data, IR and UV/Vis spectroscopic data of DUO-compounds 4.
Compound
substituent starting comp. 3
Amount of
Yield
mp Formula
(°C) (mol wt)
IR (ν, cm–1)
λmax (nm)
(1-oct.-sat.H2O)
(KBr)
4.1
H
2.12 g (10 mmol) 1.11 g
3.1 (36 %)
2.47 g (10 mmol) 1.11 g
3.2 (32 %)
2.47 g (10 mmol) 1.14 g
3.3 (33 %)
2.47 g (10 mmol) 1.11 g
3.4 (32 %)
2.81 g (10 mmol) 1.45 g
212 C29H30Br2N4O2
(626.4)
3030; 2920; 1640; 1595;
1450; 1050; 845; 740
206; 298
206; 296
206; 296
206; 220; 296
208; 296
204; 296
206; 300
206; 298
4.2
2-Cl
211 C29H28Br2Cl2N4O2 3030; 2950; 1640; 1595;
(695.3) 1460; 1350; 1240; 840; 750
195 C29H28Br2Cl2N4O2 3020; 1640; 1595; 1460;
(695.3) 121; 845; 790
203 C29H28Br2Cl2N4O2 3030; 1640; 1595; 1460;
(695.3) 134; 930; 845; 800
192 C29H26Br2Cl4N4O2 3040; 1640; 1600; 1580;
(764.2) 144; 1200; 830; 770
207 C29H26Br2Cl4N4O2 3040; 1640; 1600; 1580;
4.3
3-Cl
4.4
4-Cl
4.5
2,6-Cl
3.5
(38 %)
4.6
3,5-Cl
2.25 g (8 mmol)
3.6
1.45 g
(38 %)
(764.2)
144; 1200; 925; 845; 710
4.7
2-CH
32.26 g (10 mmol) 1.40 g
3.7 (43 %)
2.26 g (10 mmol) 1.44 g
3.8 (43 %)
189 C31H34Br2N4O2
(654.5)
3030; 2940; 1640; 1595;
146; 1250; 1180; 845; 745
4.8
3-CH3
185 C31H34Br2N4O2
(654.5)
3020; 2940; 1640; 1595;
146; 1240; 845; 790
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

536 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
Table 3. (continued)
Compound
substituent starting comp. 3
Amount of
Yield
mp Formula
(°C) (mol wt)
IR (ν, cm–1)
λmax (nm)
(1-oct.-sat.H2O)
(KBr)
4.9
4-CH3
2.26 g (10 mmol) 1.47 g
3.9 (45 %)
207 C31H34B42N4O2
(654.5)
3040; 2920; 1640; 1590;
144; 1270; 850
206; 298
4.10
2,4,6-CH3 3.10
2.54 g (10 mmol) 1.25 g
(18 %)
193 C35H24Br2N4O2
(710.6)
3030; 2970; 2910; 1640;
159; 1460; 1340; 925; 845
204; 302
4.11
2-CN
2.37 g (10 mmol) 1.51 g
3.11 (45 %)
170 C31H28Br2N6O2
(676.5)
3030; 2220; 1640; 1600;
145; 1245; 840; 765
206; 226; 294
206; 224; 294
206; 230; 296
206; 292
4.12
3-CN
2.37 g (10 mmol) 0.83 g
3.12 (25 %)
185 C31H28Br2N6O2
(676.5)
3030; 2220; 1640; 1595;
146; 1345; 1250; 850; 800
4.13
4-CN
2.37 g (10 mmol) 1.15 g
175 C31H28Br2N6O2
(676.5)
3030; 2220; 1640; 1595;
146; 1345; 1170; 840; 750
3.13
(34 %)
4.14
2-NO2
2.06 g (8 mmol)
3.14
0.69 g
(24 %)
191 C29H28Br2N6O6
(716.4)
3020; 1640; 1595; 1520;
144; 1010; 930; 855; 730
4.15
3-NO2
2.57 g (10 mmol) 1.41 g
3.15 (39 %)
120 C29H28Br2N6O6
(716.4)
3040; 2940; 1640; 1595;
152; 1460; 1350; 1240;
850; 800
206; 292
4.16
4-NO2
2.57 g (10 mmol) 0.99 g
3.16 (28 %)
183 C29H28Br2N6O6
(716.4)
3040; 2940; 1640; 1595;
152; 1460; 1350; 1240;
850; 800
206; 294
4.17
2-OCH3
2.42 g (10 mmol) 0.75 g
3.17 (22 %)
172 C31H34Br2N4O4
(686.5)
3020; 2830; 1640; 1590;
149; 1460; 1250; 840; 750
206; 298
4.18
3-OCH3
2.42 g (10 mmol) 1.37 g
3.18 (40 %)
181 C31H34B42N4O4
(686.5)
3040; 2940; 1640; 1590;
145; 1345; 1270; 840; 740
206; 296
4.19
4-OCH3
2.42 g (10 mmol) 0.81 g
3.19 (24 %)
222 C31H34Br2N4O4
(686.5)
3040; 1640; 1595; 1505;
146; 1250; 845; 830
206; 224; 296
208; 300
4.20
2,6-OCH3 3.20
2.72 g (10 mmol) 1.31 g
(35 %)
216 C33H38Br2N4O6
(746.6)
3000; 2940; 2840; 1640;
159; 1475; 1250; 845; 780;
730
4.21
2,5-OCH3 3.21
2.72 g (10 mmol) 1.25 g
(33 %)
181 C33H38Br2N4O6
(746.6)
3030; 2840; 1640; 1590;
150; 1460; 1220; 800
206; 298
4.22
2-Br
2.91 g (10 mmol) 1.15 g
3.22 (29 %)
178 C29H28Br2N4O4
(784.2)
3020; 1640; 1600; 1460;
134; 1240; 845; 750
204; 296
4.23
2,6-F
2.48 g (10 mmol) 1.80 g
3.23 (52 %)
175 C29H26Br2F4N4O2
(698.4)
3030; 1640; 1590; 1445;
133; 1230; 810; 790
206; 294
4.24
1-naphthyl 3.24
2.62 g (10 mmol) 1.66 g
200 C37H34Br2N4O2
(726.6)
3040; 1640; 1596; 1460;
135; 850; 800; 780
206; 222; 298
204; 224; 288
204; 254
(46 %)
4.25 2.62 g (10 mmol) 1.44 g
2-naphthyl 3.25 (40 %)
180 C37H34Br2N4O2
(726.6)
3040; 1640; 1600; 1460;
138; 1335; 965; 815
4.26
3.12 g (10 mmol) 1.75 g
241 C45H38Br2N4O2
(826.7)
3030; 1640; 1600; 1450;
117; 865; 845; 730
9-anthranyl 3.26 (42 %)
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 537
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

538 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
New acetylcholinesterase inhibitors of the bispyridinium type 539
Table 5. 1H–NMR–spectroscopic data of oxime ether 3 in CDCl3 (#: 60 MHz; *: 80 MHz).
Compound
substituent
2,6-H-Pyr
3,5-H-Pyr
CH=N Φ-CH2-X-N=
H-aromat.
other signals
3.1 H#
8.55 (d; 6.0) 7.30 (m)
7.98 (s) 5.23 (s)
7.30 (m)
–
–
–
–
–
–
3.2 2-Cl*
3.3 3-Cl*
3.4 4-Cl*
3.5 2,6-Cl#
3.6 3,5-Cl#
3.7 2-CH3*
8.55 (d; 6.3) 7.35 (d; 6.3) 8.03 (s) 5.30 (s)
8.55 (d; 6.3) 7.40 (d; 6.3) 8.00 (s) 5.18 (s)
8.55 (d; 6.3) 7.35 (d; 6.3) 8.05 (s) 5.18 (s)
8.60 (d; 6.0) 7.40 (d; 6.0) 7.95 (s) 5.50 (s)
8.60 (d; 6.0) 7.40 (d; 6.0) 8.05 (s) 5.13 (s)
8.55 (d: 6.3) 7.40 (d; 6.3) 8.03 (s) 5.25 (s)
8.60 (d; 6.0) 7.35 (d; 6.0) 8.00 (s) 5.20 (s)
8.58 (d; 6.3) 7.40 (d; 6.3) 8.03 (s) 5.18 (s)
7.27 (m)
7.30 (m)
7.25 (s)
7.35 (m)
7.26 (s)
7.23 (m)
7.23 (s)
2.40 (s; 2-CH3)
2.40 (s; 3-CH3)
#
3.8 3-CH3
3.9 4-CH3*
7.25 (d; 8.8; 2,6-H); 2.35 (s; 4-CH3)
7.10 (d; 8.8; 3,5-H)
3.10 2,4,6-CH3*
8.55 (d; 6.3) 7.35 (d; 6.3) 7.95 (s) 5.25 (s)
6.85 (s)
2.40 (s; 2,6-CH3)
2.25 (s; 4-CH3)
3.11 2-CN*
3.12 3-CN*
3.13 4-CN*
8.60 (d; 6.3) 7.50 (m)
8.05 (s) 5.20 (s)
7.50 (m)
7.60 (m)
–
–
–
8.60 (d; 6.3) 7.40 (d; 6.3) 8.03 (s) 5.17 (s)
8.60 (d; 6.3) 7.35 (d; 6.3) 8.05 (s) 5.20 (s)
7.63 (d; 7.6; 2,6-H);
7.45 (d; 7.6; 3,5-H)
8.05 (d; 8.8; 3-H);
7.52 (m; 4,5,6-H)
3.14 2-NO2*
3.15 3-NO2*
8.57 (d; 6.3) 7.38 (d; 6.3) 8.10 (s) 5.60 (s)
8.60 (d; 6.3) 7.43 (d; 6.3) 8.10 (s) 5.33 (s)
–
–
8.25 (s; 2-H); 8.17
(d; 8.8; 4-H); 7.65
(m; 5,6-H)
3.16 4-NO2*
8.60 (d; 6.3) 7.40 (d; 6.3) 8.07 (s) 5.30 (s)
8.20 (d; 8.8; 3,5-H);
7.50 (d; 8.8; 2,6-H)
–
3.17 2-OCH3*
3.18 3-OCH3*
3.19 4-OCH3*
8.55 (d; 6.3) 7.35 (d; 6.3) 8.00 (s) 5.25 (s)
8.57 (d; 6.3) 7.35 (d; 6.3) 8.03 (s) 5.17 (s)
8.55 (d; 6.3) 7.40 (d; 6.3) 8.00 (s) 5.15 (s)
7.40–6.75 (m)
7.10 (m)
3.80 (s; 2-OCH3)
3.80 (s; 3-OCH3)
7.33 (d; 8.8; 2,6-H); 3.80 (s; 4-OCH3)
6.87 (d; 8.8; 3,5-H)
#
3.20 2,6-OCH3
8.60 (d; 6.0) 7.45 (d; 6.0) 8.05 (s) 5.35 (s)
8.55 (d; 6.0) 7.40 (d; 6.0) 8.05 (s) 5.30 (s)
7.30 (m; 4-H) 6.55
(d; 8.0; 3,5-H)
3.80 (s; 2,6-OCH3)
#
3.21 2,5-OCH3
6.83 (m)
3.80 (s; 2-OCH3)
3.75 (s; 5-OCH3)
3.22 2-Br*
8.57 (d; 6.3) 7.40 (d; 6.3) 8.07 (s) 5.35 (s)
8.62 (d; 6.3) 7.44 (d; 6.3) 8.00 (s) 5.18 (s)
8.57 (d; 6.3) 7.39 (d; 6.3) 7.95 (s) 5.65 (s)
8.53 (d; 6.3) 7.45 (d; 6.3) 7.03 (s) 5.35 (s)
7.60–7.00 (m)
7.06 (m)
–
–
–
–
–
3.23 2,6-F*
3.24 1–naphthyl*
3.25 2–naphthyl*
8.20–7.25 (m)
7.75–7.25 (m)
3.26 9–anthranyl#* 8.50 (m)
7.45 (m)
7.95 (s) 6.20 (s)
8.50 (m; 4,5,10-H);
790 (m; 1,8-H);
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

540 Kapková et al.
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 523–540
[25] A. Mary, D. Z. Renko, C. Guillou, C. Thal, Bioorg. Med.
Chem. 1998, 6, 1835–1850.
References
[1[1] M. Harel, I. Schalk, L. Ehret-Sabatier, F. Bouet, M. Goeld-
ner, C. Hirth, P. H. Axelsen, I. Silman, J. L. Sussman, Proc.
Natl. Acad. Sci. USA 1993, 90, 9031–9035.
[26] A. K. Sikder, A. K. Gosh, D. K. Jaiswal, J. Pharm. Sci. 1993,
82, 258–261.
[27] K. Schoene, Arzneimittelforsch. 1968, 18, 1350–1351.
[2] I. Silman, M. Harel, P. Axelsen, M. Raves, J. L. Sussman,
Biochem. Soc. Transaction 1994, 22, 745–749.
[28] Z. Radic, A. N. Pickering, D. C. Vellom, S. Camp, P. Taylor,
Biochemistry 1993, 32, 12074–12084.
[3] M. Harel, D. M. Quinn, H. K. Nair, I. Silman, J. L. Sussman,
J. Am. Chem. Soc. 1996, 118, 2340–2346.
[29] D. H. Cheng, X. C. Tang, Pharmacol. Biochem. Behav.
1998, 60, 377–386.
[4] H. M. Greenblatt, G. Kryger, T. Lewis, I. Silman, J. L. Suss-
man, FEBS Letters 1999, 463, 321–326.
[30] http://www.rcsb.org.
[5] G. Kryger, I. Silman, J. L. Sussman, J. Physiol. Paris 1998,
92, 191–194.
[31] C. Pilger, C. Bartoluci, D. Lamba, A. Tropsha, G. Fels, J.
Mol. Graphics Modell. 2001, 19, 288–296.
[6] N. Zacharias, D. A. Dougherty, Trends Pharmacol. Sci.
2002, 23, 281–287.
[32] W. Sippl, J. M. Contreras, I. Parrot, Y. M. Rival, C. G. Wer-
muth, J. Comput.-Aided Mol. Design, 2001, 15, 395–410.
[7] S.A.Botti, Felder, C.E.;Lifson, S.;Sussman, J.L.;Silman,
I. A Modular Treatment of Molecular Traffic Through the
Active Site of Cholinestaerase. Biophys. J. 1999, 77,
2430–2450.
[33] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–52.
[34] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–
789.
[35] W.J.Hehre, R.Ditchfield, J.A.Pople, J.Chem.Phys.1972,
56, 2257–2261.
[8] C. Guillou, A. Mary, D. Z. Renko, E. Gras, C. Thal, Bioorg.
Med. Chem. Lett. 2000, 10, 637–639.
[36] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrevski, J. A. Mont-
gomery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J.Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, A. Adamo, S. Clifford, J. Ochterski, G. A. Peters-
son, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D.
Rabuck, K. Raghavachari, J. B. Foersman, J. Cioslowski,
J. V. Ortiz, B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, M. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez,
M. Challacombe, P. M.W. Gill, B. G. Johnson, W. Chen, M.
W.Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle, J.
A.Pople, Gaussian98, Revision A.7;Gaussian, Inc., Pitts-
burgh, PA, 1998.
[9] P.R.Carlier,Y.F.Han, E.S.-H.Chow, C.P.-L.Li, H.Wang,T.
X. Lieu, H. S.Wong,Y.-P. Pang, Bioorg. Med. Chem. 1999,
7, 351–357.
[10] M.-L. Hu, L.-J. Wu, G. Hsiao, M.-H. Yen, J. Med. Chem.
2002, 45, 2277–2282.
[11] P. R. Carlier, D.-M. Du, Y. Han, J. Liu, Y.-P. Pang, Bioorg.
Med. Chem. Lett. 1999, 9, 2335–2338.
[12] T. Bunyapaiboonsri, O. Ramström, S. Lohmann, J.-M.
Lehn, L. Peng, M. Goeldner, ChemBioChem. 2001, 2,
438–444.
[13] J. Gasteiger, U. Holzgrabe, E. Kostenis, K. Mohr, U. Sürig,
M. Wagener, Pharmazie, 1995, 50, 99–105.
[14] U. Holzgrabe, A. J. Hopfinger, J. Chem. Inf. Comp. Sci.
1996, 36, 1018–1024.
[37] M. Rarey, B. Kramer, T. Lengauer, G. Klebe, J. Mol. Biol.
1996, 261, 470–489.
[15] P. N. Craig, J. Med. Chem. 1971, 14, 680–684.
[38] G. Lin, C.Y. Lai, W. C. Liao, Bioorg. Med. Chem. 1999, 7,
2683–2689.
[16] M. L. Amzel, Curr. Opin. Biotechnol. 1998, 9(4), 366–369.
[39] J. Eichler, A. Anselmet, L. J. Sussman, J. Massouilie, I. Sil-
[17] M. H. Botero Cid, U. Holzgrabe, E. Kostenis, K. Mohr, C.
man, Mol. Pharmacol. 1994, 45, 335–340.
Tränkle, J. Med. Chem. 1994, 37, 1439–1445.
[40] T. Shen, K. Tai, R. H. Henchman, J. A. McCammon, Acc.
Chem. Res. 2002, 35, 332–340.
[18] E.J.Poziomek, D.N.Kramer, W.A.Mosher, H.O.Michel, J.
Am. Chem. Soc. 1961, 83, 3916–3917.
[41] H. Dvir, D. M.Wong, M. Harel, X. Barril, M. Orozco, F. J. Lu-
que, D. Munoz-Torrero, P. Camps, T. L. Rosenberg, I. Sil-
man, J. L. Sussman, Biochemistry 2002, 41, 2970–2981.
[19] M. N. Dinia, A. Hassikou, A. Lattes, Bull. Soc. Chim. Belg.
1993, 102, 623–624.
[20] E. Inkmann, U. Holzgrabe, K.-F. Hesse, Pharmazie, 1997,
52, 764–774.
[42] J.A.Moore, L.D.Kornreich,Tetrahedron Letters.1963, 20,
1277–1281.
[21] G.L.Ellman, K.D.Coutney, V.Andres, R.M.Featherstone,
[43] Alchemy 2000® 2.05 Tripos Inc., 1699 South Hanley Rd.,
St. Louis, Missouri, 63144, USA, 1998.
Biochem. Pharmacol. 1961, 7, 88–95.
[22] V.E.Gregor, M.R.Emmerling, C.Lee, C.J.Moore, Bioorg.
Med. Chem. Lett. 1992, 2, 861–864.
[44] SYBYL® 6.7.1 Tripos Inc., 1699 South Hanley Rd.,
St. Louis, Missouri, 63144, USA, 2000.
[23] A. S. Hodge, D. R. Humphrey, T. L. Rosenberry, Mol. Phar-
macol. 1992, 41, 937–942.
[45] J. J. P. Stewart, J. Comput.-Aided Mol. Deisgn, 1990, 4,
1–105.
[24] M. T. McKenna, G. R. Proctor, L. C.Young, A. L. Harvey, J.
Med. Chem. 1997, 40, 3616–3523.
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim