Interaction of (benzylidene-hydrazono)-1,4-dihydropyridines with β-amyloid, acetylcholine, and butyrylcholine esterases

Article abstract of DOI:10.1016/j.bmc.2010.01.002

Approved drugs for the treatment of Alzheimer's disease belong to the group of inhibitors of the acetylcholinesterase (AChE) and NMDA receptor inhibitors. However none of the drugs is able to combat or reverse the progression of the disease. Thus, the recently reported promising multitarget-directed molecule approach was applied here. Using the lead compound DUO3, which was found to be a potent inhibitor of the AChE and butyrylcholinesterase (BuChE) as well as an inhibitor of the formation of the amyloid (Aβ) plaque, new non-permanently positively charged derivatives were synthesized and biologically characterized. In contrast to DUO3 the new bisphenyl-substituted pyridinylidene hydrazones 5 are appropriate to cross the blood-brain barrier due to their pK_a values and lipophilicity, and to inhibit both the AChE and BuChE. More important some of the pyridinylidene hydrazones inhibit the Aβ fibril formation completely and destruct the already formed fibrils significantly.

Full text of DOI:10.1016/j.bmc.2010.01.002

Bioorganic & Medicinal Chemistry 18 (2010) 2049–2059
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Interaction of (benzylidene-hydrazono)-1,4-dihydropyridines
with b-amyloid, acetylcholine, and butyrylcholine esterases
Vildan Alptüzün a, Michaela Prinz b, Verena Hörr b, Josef Scheiber b, , Krzysztof Radacki c, Adyary Fallarero d,
Pia Vuorela d, Bernd Engels e, Holger Braunschweig c, Ercin Erciyas a, Ulrike Holzgrabe b,
*
a Faculty of Pharmacy, Ege-University, Bornova-Izmir, Turkey
b Institute of Pharmacy and Food Chemistry, University of Würzburg, Am Hubland, 97074 Würzburg, Germany
c Institute of Inorganic Chemistry, University of Würzburg, Am Hubland, 97074 Würzburg, Germany
d Division of Pharmacy, Faculty of Mathematics and Natural Sciences Åbo Akademi University BioCity, Artillerigatan 6A, FI-20520 Turku, Finland
e Institute of Organic Chemistry, University of Würzburg, Am Hubland, Würzburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Approved drugs for the treatment of Alzheimer’s disease belong to the group of inhibitors of the acetyl-
cholinesterase (AChE) and NMDA receptor inhibitors. However none of the drugs is able to combat or
reverse the progression of the disease. Thus, the recently reported promising multitarget-directed mole-
cule approach was applied here. Using the lead compound DUO3, which was found to be a potent inhib-
itor of the AChE and butyrylcholinesterase (BuChE) as well as an inhibitor of the formation of the amyloid
(Ab) plaque, new non-permanently positively charged derivatives were synthesized and biologically
characterized. In contrast to DUO3 the new bisphenyl-substituted pyridinylidene hydrazones 5 are
appropriate to cross the blood–brain barrier due to their pKa values and lipophilicity, and to inhibit both
the AChE and BuChE. More important some of the pyridinylidene hydrazones inhibit the Ab fibril forma-
tion completely and destruct the already formed fibrils significantly.
Received 26 October 2009
Revised 22 December 2009
Accepted 2 January 2010
Available online 15 January 2010
Keywords:
AChE inhibitors
b-Amyloid plaques
(Benzylidene-hydrazono)-1,4-
dihydropyridines
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
inhibitors, antioxidants, c-secretase modulators, peroxisome pro-
liferator-activated receptors (PPAR)
c agonists, H3 antagonists,
Alzheimer’s disease (AD) is a progressive neurodegenerative
disease characterized by a loss of cognitive function and behavioral
abnormalities. Immense efforts have been made to develop effi-
cient strategies for the treatment of AD because of its greatly
increased prevalence. According to predictions the number of AD
cases in the western world will double every 20 years and even tri-
ple in India and China with 29 million people in 2020, mostly owed
to increased human longevity.1
Due to the multifactorial nature of involved processes, recently
nicely summarized in Ref. 2, for example, protein misfolding and
aggregation, intracellular oxidative stress and free radical forma-
tion, elevated levels of trace metals, mitochondrial dysfunction,
and phosphorylation impairment, several attempts have been
conducted to overcome the development of the progressive and
eventually fatal neurodegenerative disease. Among others the
most important, current treatment strategies are the approved
and HMG-CoA reductase inhibitors.3
Among different neurotransmitter deficits occurring in AD, loss
of cholinergic neurons and resulting deficit of neurotransmitter
acetylcholine (ACh), and presynaptic M2 muscarinic and nicotinic
receptors have been found.4 These findings led to the cholinergic
hypothesis postulating that many of the cognitive, functional, and
behavioral symptoms experienced by AD patients are a conse-
quence of deficient cholinergic neurotransmission.5 Hence for the
most part, inhibitors of acetylcholinesterase, enhancing the acetyl-
choline concentration in the brain, have been introduced to the
market for treating mild-to-moderate AD. Well-known examples
are tacrine, galanthamine, rivastigmine, and donepezil.6 Another
promising approach is the development of dual inhibitors for AChE
and butyrylcholinesterase (BuChE),7 as BuChE activity seems to
correlate with AChE activity in AD and a cognitive improvement
could be reached.8,9 Recently, dimebon, originally approved as a
non-selective antihistamine, was discovered to weakly inhibit the
BuChE and AChE, weakly block the NMDA receptor signaling path-
way and inhibit mitochondrial permeability transition pore open-
ing. Due to neuroprotective effects observed in models for AD’s
and Huntington’s disease, a clinical trial phase III revealed a signif-
icant improvement of the cognition of AD patients.10 However,
none of these disease-modifying anti-Alzheimer’s drugs is able to
cholinesterase inhibitors (see below) and NMDA (N-methyl-D-
aspartate) receptor antagonists, that is, memantine, and sub-
stances in clinical trials, for example, amyloid-b (Ab) aggregation
* Corresponding author. Tel.: +49 931 888 5460; fax: +49 931 888 5494.
E-mail address: u.holzgrabe@pzlc.uni-wuerzburg.de (U. Holzgrabe).
Present address: Roche, Nonnenwald 2, Penzberg, Germany.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.01.002

2050
V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
combat or reverse the progression of the disease; they are only able
to delay the process, emphasizing the need for more potent AD
drugs to modify the disease rather than treating its symptoms only.
The deposition of Ab in the brain is the pathological hallmark of
AD and one of the potential reasons of the neuronal damage.11 Ab
is a fragment of 37–42 amino acids excised from the transmem-
brane region of the amyloid precursor protein (APP) by the sequen-
tial action of b-site APP cleaving enzyme (BACE = b-secretase) and
Cl
Cl
Br-
Br-
O
O
N
N
N+
N+
Cl
Cl
DUO 3
Cl
Br-
c-secretase, eventually resulting in neurotoxicity, tau hyper-
O
N
phosphorylation and Ab oligomerization, aggregation, and finally
N+
plaque formation. Consequently, inhibitors of BACE and
tase are prime objectives for AD drug development.12
c-secre-
Cl
1
Several substances of peptidic13 and non-peptidic origin with
anti-amyloid potency were reported. Among the latter are endoge-
nous compounds such as melatonin, in addition to rifampicin,
benzofuran, benzothiazole, hydroxyindole, curcumin-derived pyra-
zoles and isoxazoles, and naphthylazo derivatives, phenols, classical
antibiotics, for example, tetracycline as well as cyclodextrins.14 In
addition, the role of AChE is not only the hydrolysis of the neuro-
transmitter ACh, but also the acceleration of the aggregation of Ab
into amyloid fibrils. Responsible for this activity seems to be the
peripheral anionic site (PAS) of AChE gorge.15 Within this context,
it is expectedthat even a peripheral site blocker prevents the Ab pep-
tidetointeractwithAChE, and, thus, inhibitsthefibrilformation pro-
cess. Studies regarding the development of such AChE inhibitors
were already reported, for example, for coumarine derivatives.16
In order to achieve the highest efficiency the new approach of
multitarget-directed molecules introduced by Melchiorre and co-
workers17 seems to be promising. In this study we aimed to address
the active site as well as the peripheral site of the AChE, the active
site of the BuChE, and the Ab in order to inhibit the formation of
the amyloid fibril plaque or reverse the fibril formation. The multi-
target approach might open the door to overcome the clinical prob-
lems with drugs targeting the Ab deposition only.18 Parts of the
concept has already worked, for example, for phenserine,19 benzofu-
ran-based hybrid compounds,20 and tacrine–donezepil hybrids.21
The group of DUO compounds (see Chart 1), derived from the
reactivator of the poisoned AChE obidoxim, were recently found
to be potent ditopic inhibitors of AChE (Torpedo californica) occupy-
ing both the active site and the peripheral binding site.22 The inter-
Cl
N
O
N
Cl
2
N
N
HO
N
3
Cl
O
N
4
Cl
R
Cl
N
N
N
Cl
5
Chart 1. The lead compound DUO3 and the derived compounds.
actions of, for example, DUO3 found after docking include
p
ꢀ
p
stacking and cation– contacts with amino acid residues of the an-
p
tives (compound 5, see Chart 1) were synthesized. The
quaternary salt formation is also important for the rich retention
of active compound within the brain to ensure therapeutically sig-
nificant concentrations. The pKa values as well as the lipophilicity
(log P) were determined for an estimate of the possible BBB pas-
sage. The structure of the (benzylidene-hydrazono)-1,4-dihydro-
pyridine 5a was elucidated by means of an X-ray analysis in
order to find out, whether these compounds are principally able
ionic substrate binding site (Trp84, Phe331, and Tyr334) and the
peripheral anionic binding site (Trp279). However, molecular mod-
eling studies also revealed the compounds being too long and thus,
occupying areas of the enzyme outside the catalytic gorge. Hence,
DUO3 was systematically shortened to give the most active com-
pound 1 in this series inhibiting the AChE in the nanomolar range
of concentration.23 Both DUO3 and correspondingly shortened
molecules were able to inhibit the amyloid fibril formation24 and
thus are appropriate lead compounds for the development of mul-
titarget-directed molecules.
Since 1 is permanently positively charged and therefore not
able to pass the blood–brain barrier (BBB), the first aim of this
study was to replace the pyridinium ring with a piperidine ring
(compounds 2 and 4) whose nitrogen is under physiological condi-
tions partially protonated (necessary for AChE interaction) and
partially non-protonated (transport form) due to the expected pKa
value of approximately 7 (see Chart 1).
to form
p–p stacking or p–cation driven structures which would
be an indication that a corresponding interaction with the amyloid
is possible.26 All compounds were subjected to Ellman’s test in or-
der to measure the inhibitory potency towards the AChE, BuChE
and to the thioflavin T staining assay in order to determine the
inhibition of the amyloid fibril formation as well as the fibril
destruction. Additionally, the core peptide of the Ab42 was synthe-
sized in order to check whether this short peptide can be used for
pre-screening of the inhibition of the amyloid fibril formation.
In order to open the possibility to pass the BBB as a prodrug we
followed the chemical delivery systems developed by Bodor and
co-workers25 which is based on the dihydropyridine-pyridinium-
type redox conversion of a lipophilic dihydropyridine. The latter
is able to pass the BBB and the cationic, lipid-insoluble pyridinium
salt can perfectly interact with the AChE. Thus, as a second aim cor-
responding (benzylidene-hydrazono)-1,4-dihydropyridine deriva-
2. Results and discussion
2.1. Chemistry
2.1.1. Synthesis
The compounds 2–4 were synthesized according to Ref. 27.

V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
2051
Condensation of 4-hydrazinopyridine with 2,6-dichlorobenzal-
dehyde gave the hydrazone 6 in analogy to Douglas et al.28
Compound 6 can be alkylated with the corresponding alkylhaloge-
nides to give the series of (benzylidene-hydrazono)-1,4-dihydro-
pyridine compound 5 (see Chart 2).
2.1.3. Log P determination
The partition coefficients were determined by means of RP
HPLC using the correlation between the capacity factor and the
log P values of reference compounds.30 As can be seen in Table
1a, the log P values of compounds 3–5 ranged between 2.8 and 4
suggesting the possibility of the BBB passage. The log P of com-
pound 4 was found to be the highest which can be explained by
the low pKa indicating a high degree of deprotonation (i.e., the neu-
tral form) at pH 7.4. Additionally, the partition coefficients of the
(benzylidene-hydrazono)-1,4-dihydropyridine 5d and 5e were
representatively determined by the classical shake flask method
in phosphate buffer at pH 5, 6, 7, and 7.4 using UV spectroscopy.
The higher the pH value the more lipophilic are both compounds
because according to their pKa value the less they are protonated
(see Table 1b). With a log P of ꢁ3 at a pH value of 7.4 compounds
5 are sufficiently lipophilic to pass the BBB.
2.1.2. pKa determination
In order to find out whether the newly synthesized compounds
are protonated in physiological conditions (pH 7.4), and along with
that positively charged, the pKa values for compounds 2, 4, and 5d
were determined by means of titration using Sirius Analytical
Instrument. Since the compounds were insufficiently water soluble
the pKa values were determined in water/dioxane and water/meth-
anol mixtures.29 Compound 2 was found to have a pKa of approx-
imately 7 and will be thus 50% protonated under physiological
conditions which is ideal for the passage through the BBB and
the
p–cation interaction with Phe331 of the AChE. In comparison
the corresponding piperidone oxime compound 4 is less basic
(pKa ꢁ 6) which will be advantageous for the BBB transport and
disadvantageous for the AChE interaction. The (benzylidene-
hydrazono)-1,4-dihydropyridine 5d is more basic with a pKa value
of about 9 and, thus, 95% protonated at the AChE. However, the
small neutral portion of 5 in equilibrium provides the possibility
to pass the BBB (sink conditions).
2.1.4. X-ray analysis
In the series of the compound 5 only the methylated compound
5a gave crystals from methanol–water suitable for X-ray diffrac-
tion analysis. The compound 5a crystallizes as yellow needles in
the monoclinic P21/n space group. The unit cell contains half a mol-
ecule of water. A
p–p stacking between the dichlorobenzyl moie-
ties and dihydropyridine skeleton can be seen in Figure
1
indicating that the compound may form a similar stacking with
Ab. This will be likely the same for all compound 5.
Cl
2.2. In vitro bioassays
N
NH2
O
HN
N
HN
N
2.2.1. AChE/BuChE inhibition
Cl
Cl
Cl
All compounds were subjected to Ellman’s test in order to eval-
uate their potency to inhibit the AChE (E.C. 3.1.1.7 from Electric
Eel). The data are displayed in Table 1a. Interestingly, the replace-
ment of the pyridinium ring in compound 1 with a piperidine ring
resulted in a loss of activity by a factor of approximately 100 which
holds true for both compounds 2 and 4. Compound 3 missing the
dichlorobenzyl group is almost inactive indicating the importance
of this moiety. The potency of the hydrazone derivatives is sensi-
tive to the substitution on the 1,4-dihydropyridine nitrogen. Com-
pound 5a having only a methyl group is almost inactive whereas
all other compounds having an additional phenyl ring are active
in the micromolar range of concentration indicating that the phe-
nyl ring is pivotal for an interaction with the AChE. Since 5c having
a phenylethyl group has a rather low activity the phenyl ring
seems not to be able to interact properly. However it has to be kept
+
6
R
Cl
CH3
5a
5b
N
N
RHal
Cl
5
N
R
5c
5d
5e
Table 1b
pH dependent partition coefficients (log P) of compounds 5d and 5e, determined by
means of the shake flask method
Cl
No.
pH
5
6
7
7.4
Cl
Chart 2. Synthesis pathway of compound 5.
5d
5e
0.7
0.8
0.9
1.5
1.3
2.2
3.2
3.4
Table 1a
IC50 values, evaluated by means of Ellman’s test (tacrine in this test 0.044 0.004
pharmacological data are the mean SEM, n = 3–6 experiments; IC50 = concentration inhibiting the AChE activity by 50%)
l
M) and partition coefficients of compounds 2–5, determined by the RP HPLC method (the
No.
DUO3
1
2
3
4
5a
5b
5c
5d
5e
Log P
0.650
3.6
2.8
4.3
3.4
3.5
3.9
3.7
4.0
IC50
IC50
(l
M) AChE 0.34 0.2247 0.18 0.00748 19.13 1.14 277.50 10.57 12.05 0.08 739 139.8 17.40 2.56 56.70 9.48 5.66 0.17 6.40 1.62
(
lM)
1.83 1.13
2.78 1.06
Not tested
Not tested
>100
11*
8*
9*
10*
Not tested
BuChE
*
Due to the low activity only two experiments were performed and, therefore, no SEM are given.

2052
V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
Figure 1. View along cell a axis showing p–p stacking between molecules of 5a (water molecules omitted for clarity).
in mind, that the prodrug compound 5 might be oxidized in vivo
resulting a flat, permanently positively charged pyridinium ring
system as present in the highly active compound 1.
The monitoring of the BuChE (EC 3.1.1.8, from horse serum)
activity using the kinetic assay based also on Ellman’s reagent
and has been previously developed and validated.11 The BuChE
inhibition of the most interesting AChE inhibitors was also evalu-
ated (cf. Table 1a). Best results showed again DUO3 and 1. How-
Fibril formation
120
100
80
60
40
20
0
control
DUO 3
1
5d
ever, all tested derivatives of compound
5 showed BuChE
inhibitory activity in the low micromolar range of concentration.
The activity of compound 4 was out of range, again pointing out,
that an aromatic or flat unsaturated moiety in the core of the mol-
ecule is of importance.
0
20
40
60
80
100
t [h]
2.2.2. Thioflavin T test
Thioflavin has been used for several decades to stain Ab. It can
visualize the block of fibril formation. The corresponding data are
displayed in Figure 2. Initially, compounds 1 and 5d having the
highest AChE inhibitory activity in Ellman’s test were subjected
to the thioflavin T test using Ab40 and compared to the activity
of DUO3 which was found to retard the nucleation phase of the
amyloidogenesis.49 Both compounds 1 and 5d show similar fibril
formation inhibition behavior as DUO3 indicating that all (benzyl-
idene-hydrazono)-1,4-dihydropyridines 5 deserve a closer inspec-
tion (see below).
Figure 2a. Fibril inhibition of DUO3 and its derivatives tested by ThT test with
amyloid b1-40.
83 lM which is in good accordance with the aforementioned find-
ing. Additionally, as can be seen in Figure 2b the short oligopeptide
showed almost identical results concerning the fibril formation
and their inhibition by the compounds DUO3, 1, and 5d. A similar
approach was recently reported where a corresponding core pep-
tide flanked with a b-sheet-promoting sequence 8(Leu-Ser)n) was
used.33
Due to the encouraging results with DUO3, 1, and 5d the test
was extended to all compounds 5a–e (see Fig. 2c). With exception
of compound 5a having no phenyl moiety attached to the pyridine
nitrogen all compound show an inhibition of the fibril aggregation
between 60% and 92% at a concentration of 0.5 mM. Since 5c and
5d having an ethyl and propyl linker between the pyridone and
the phenyl ring (at the dihydropyridine nitrogen) show the highest
activity, the inhibition of the fibril formation of the latter com-
Due to the high price of Ab40 peptide, a smaller peptide featur-
ing the same fibril formatting properties was aimed to develop. Re-
cently a core peptide of only five highly conserved amino acids
(KLVFF) was described to be responsible for fibril formation.31
Therefore, a peptide composed of eleven amino acids and including
the core sequence (HHQKLVFFAED) was synthesized and used in
the thioflavin T test. The test was validated using 4-aminophenol
whose IC50 value of blocking the Ab oligomerization was reported
to be ꢁ90
l
M.32 Applying the core peptide revealed an IC50 value of

V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
2053
Formation of fibrils
Formation of fibrils
8000%
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
control
DUO3 (0.5 mM)
7000%
6000%
5000%
4000%
3000%
2000%
1000%
0%
5d (0.5 mM)
5d (0.05 mM)
5d (0.005 mM)
5c (0.5 mM)
5c (0.05 mM)
5c (0.005 mM)
control
DUO3
1
5d
0
10
20
30
40
50
60
70
80
0
-10.0
5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
0.
-1000%
t [h]
t [h]
Figure 2b. Fibril inhibition of DUO3 and its derivatives 1 and 5d tested by ThT test
with shortened peptide.
Figure 2d. Concentration-dependent fibril inhibition of 4(1H)-pyridinylidene
hydrazones 5c and 5d tested by ThT test with shortened peptide.
pounds was measured concentration dependently (see Fig. 2d).
Even at a concentration of 0.05 mM an inhibition of about 64%
and 58% was observed for 5d and 5c, respectively.
Destruction of preformed fibrils
700
Prompted by the good results, the compounds were subjected
to a modified test system, where fibril destruction can be mea-
sured. Figure 2e representatively displays the concentration-
dependent effect of 5d on the fibril destruction. The destruction oc-
curs very quickly within the first minutes as can be seen from the
low starting fluorescence emission in comparison to the control.
After a further small decrease within the next 5 min the emission
stays constant at least until 15 h (data not shown). With exception
of 5a all other compounds show concentration-dependent similar
effects (data not shown).
600
500
control
5d (0.005mM)
400
5d (0.05mM)
5d (0.5mM)
300
200
100
0
2.3. Molecular modeling of the AChE interaction
0
10
20
30
t [min]
40
50
60
The possible interactions between the compounds and AChE
were determined by calculating a score that applies a measure
for the strength of the interaction between the binding pocket
and the structure of the ligands. To investigate possible binding
modes of the compounds synthesized we performed qualitative
docking experiments. This means that we are trying to elucidate
possible binding modes, but we are not using the calculated dock-
ing score for the bonding strength as the latter one is known to
deliver imprecise results.34 Consequently, we used the following
workflow.
Figure 2e. Concentration-dependent fibril destruction of 5d tested by ThT test with
shortened peptide-amyloid.
From the pH value one can assume that the compounds are pro-
tonated only at one position. To determine the most basic proton-
ation sites the various possibilities were computed and compared.
As an initial step a set of conformers for each compound was gen-
erated using the MMFF94 force field as implemented in Macro-
model.35 The resulting structures served as starting point of
subsequent DFT calculations. In these computations an aqueous
solution was mimicked employing the COSMO36 approach imple-
mented in Turbomole.37 For these computations the BP86 func-
tional38 was used in combination with basis sets of TZVP quality
and the RI approximation.39 The calculations predict that com-
pounds 2 and 4 are firstly protonated at the amine group of the
piperidine moiety. For compound 5d the azine group was com-
puted to be the most basic one. Based on these DFT/COSMO calcu-
lations the standard COSMO-RS40 method implemented in
COSMOTHERM program was applied to predict the acidity of the com-
pounds. The results are in good concordance with the experimental
pKa values indicating the reliability of the predicted protonated
structures (pKa: 2 exp. = 7.1; calcd = 7.1; 4 exp. = 5.7; calcd = 6.1;
5d exp. = 8.7; calcd = 10.6).
Formation of fibrils
250
200
control
Duo3
150
5a
5b
5c
100
5d
5e
50
0
0
20
40
60
t [h]
80
100
120
For docking purposes, the connection table with calculated pro-
tonation states for each compound was used as an input for a dock-
ing with Openeye FRED. Interpretation of the poses generated
during docking straightforwardly represents the activities seen in
Figure 2c. Fibril inhibition of 4(1H)-pyridinylidene hydrazones 5a–e tested by ThT
test with shortened peptide.

2054
V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
experiment (for details please refer to Section 4). We always used
the top-ranking pose determined by consensus scoring as input for
a detailed visual investigation which is described in the following
as this is expected to provide unbiased results. As a target structure
we used AChE from T. californica (PDB: 1EVE) as it was successfully
used in previous investigations.47 Furthermore it contains Donepe-
zil as co-crystallized ligand, which has a structure that is some-
what related to our compounds.
basic it is less positively charged at pH 7.4 which results in a less
strong –cation interaction with Phe331 and, thus, explains the
p
decrease in activity. Compound 5d makes nice interactions with
Trp84 and Tyr334, instead of Phe331, but does not reach the
peripheral Trp279, because the phenyl ring points to the wrong
direction (see Fig. 3d). This explains the fact that the compound
does not have the high inhibitory activity of compound 1.
Whereas the highly active compound 1 shows an excellent
interaction between the dichlorobenzyl ring and Trp84, the pyrid-
inium ring and Phe331, and the phenyl ring and Trp 279 (see
Fig. 3a), the ‘saturated’ compound 2 cannot reach the amino acid
Trp84 of the anionic center which easily explains the decrease in
inhibitory potency (see Fig. 3b). In addition the interaction be-
tween the piperidine ring and the Phe331 is not strong. Interest-
ingly, compound 4 which is shorter can find all interaction
partner of the AChE mentioned for 1 (see Fig. 3c). Since 4 is less
2.4. Structure–activity relationships
In contrast to the starting compounds DUO3 and 1 which are
permanently positively charged, the compounds 2–5 can be non-
protonated according to the pKa values resulting in a higher lipo-
philicity (see log P values), and, thus, being able to pass the BBB.
The neutral form can be regarded as a (membrane) transport form.
Additionally, the compounds perfectly fulfill the Lipinski’s rule of
five41 and do have less than three hydrogen-bond donors and six
Figure 3. (a) Binding mode of compound 1, (b) of compound 2, (c) of compound 4 and (d) of compound 5d.

V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
2055
hydrogen-bond acceptors which is a special requirement for cross-
ing the BBB.42 Since they are also ionizable at physiological pH they
are theoretically able to form a favorable p–cation interaction with
the phenylalanine 331 of the AChE. Taken together, compounds 2–
5 fulfill the requirement of being active in the brain which is a step
forward compared with DUO3 and compound 1.
interesting result: due to its two dichlorobenzyl moieties this com-
pound fits into the binding pocket exactly the other way round
(analogous to the non-5 compounds). The moiety that interacts
with Trp84 in all other cases now interacts with Trp279, whereas
the newly introduced dichlorobenzyl is directed towards the inner
part of the binding pocket.
With exception of compounds 3 and 5a all compounds are
reasonable able to inhibit the AChE activity indicating that the
dichlorobenzyl moiety at the oxime end of the molecule and
the phenylalkyl skeleton at the 1,4-dihydropyridine nitrogen are
belonging to the pharmacophore for AChE inhibition. A series of
(benzylidene-hydrazono)-1,4-dihydropyridines compounds hav-
ing a p-chlorobenzyl instead of a 2,6-dichlorobenzyl ring shows
analogous inhibitory activities (data not shown).
In summary, the docking gives a very good interpretation of the
activity seen in the experiments and will guide future research. In
order to enhance the inhibitory properties of the pyridinylidene
hydrazones 5 the future developments have to focus at the optimi-
zation of the linkers between the dihydropyridine and both lateral
phenyl rings.
Compounds 5c and 5d, carrying a phenylpropyl and a phenyl-
ethyl residue, were found to be the best inhibitors of the fibril for-
mation even better than DUO3 whereas compounds 5b and 5e,
having both a benzyl moiety attached to the 1,4-dihydropyridine
nitrogen, are less active. The already formed fibrils can be de-
stroyed by compounds 5b–e to the same extent. Since 5a did not
show any activity, the second phenyl rings seems to be essential
for the inhibition of the fibril formation. Additionally the distance
between the dihydropyridine nitrogen and the phenyl rings has
to be at least two methylene groups which goes in parallel with
the SAR of the AChE inhibitory activity.
In order to occupy the active site of the AChE properly, that is, to
form a
p
–
p
interaction with Trp84 at the bottom of the active site
interaction with Trp279, a
and to reach the peripheral site for
p–p
phenylpropyl substituent attached to the piperidine nitrogen,
being present in 4 and 5d, is necessary at least. However, the ‘sat-
urated’ pyridinium ring, that is, a piperidine ring, cannot properly
interact with Phe331 which explains the loss in inhibitory activity
of compounds 2 and 4 (Fig. 3b and c). The dihydropyridine moiety
in 5d forms a perfect
could be expected from the
ring and the 1,4-dihydropyridine observed in crystals (see Fig. 1).
p
–p
interaction with Tyr334 (Fig. 3d) which
p–p
interaction of the dichlorophenyl
3. Conclusion
In the case of protonation (pKa 9) at physiological pH a
p–cation
interaction with Trp334 can be additionally formed.
Starting from the permanently positively charged lead com-
pound DUO3 which is able to inhibit the activity of the AChE and
BuChE as well as the fibril formation of Ab40, but cannot pass
the BBB, we have developed a concept to combine the ability to
cross the BBB, to interact with the AChE and BuChE in positively
charged form, and even more important to inhibit the plaque for-
mation responsible for the appearance of Alzheimer’s disease,
and additionally to destroy the already formed plaques. The most
active (benzylidene-hydrazono)-1,4-dihydropyridines 5d merges
the inhibition of both the AChE/BuChE and the fibril formation as
well as the destruction. In the next step, the inhibitory activities
have to be enhanced by variations of the linkers and the substitu-
tion pattern. This work is in progress.
The docking poses for the 5-series (shown in Fig. 4) reveal a
good correlation between the alkyl chain length (between the
dihydropyridine and the phenyl moiety) and the inhibition of AChE
with exception of 5e: the terminal dichlorobenzyl moiety strongly
favors an p–p interaction with Trp84, and the constant part of the
structure fits reasonably well along the binding pocket. Compound
5d is the only one that reaches the Trp279 and through its phenyl
terminus is able to have a
p–p interaction with this tryptophane.
The alkyl chain length of 5d can be seen as almost perfect for the
binding pocket of AChE. Strikingly, compound 5e, which is consid-
erably shorter still exhibits almost the same strength of inhibition.
A detailed investigation of its preferred docking pose shows a very
Figure 4. Here the overlay of all 5-series compounds is shown. This figure outlines the interpretation which nicely fits the chain length of the different molecules to its
inhibitory activity against AChE with 5e being an exception (colors: blue 5e, yellow 5d, green 5b, orange 5c, 5a is hidden within the orange).

2056
V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
4.3.2. 1-Benzyl-4-[(2,6-dichlorobenzylidene)hydrazono)]-1,4-
dihydropyridine 5b
4. Experimental section
4.1. Material and methods
Yield 62%; yellowish oil; IR (KBr)
m 1644, 1510, 1393, 1174, 825,
774 cmꢀ1 1H NMR (CH3OH-d4): d ppm 5.01 (2H, s, CH2-Ph), 6.39
;
(1H, dd, J = 2.5/7.5 Hz,), 7.19–7.46 (11H, m), 8.50 (1H, s, N@CH);
13C NMR (CH3OH-d4): d ppm 60.33 (t), 108.54 (d), 112.30 (d),
128.73 (d), 129.60 (d), 130.20 (d), 130.24 (d), 130.56 (d), 132.96
(s), 135.86 (s), 137.61 (s), 139.82 (d), 140.31 (d), 145.73 (d),
163.12 (s). EI-MS m/z (% relative intensity): 357 (M+2), 355 (M+),
154 (34), 152 (100), 125 (26), 91 (40), 92 (40), 63 (28). Anal. Calcd
for (C19H15N3Cl2; 356.25 g/mol): C, 64.1; H, 4.23; N, 11.8. Found: C,
63.9; H, 4.56; N, 11.5.
Melting points were determined with a Büchi 510 melting point
apparatus (Büchi, Switzerland) and are not corrected. 1H and 13C
NMR spectra were recorded on a Bruker AV 400 instrument (1H
400.132 MHz; 13C 100.613 MHz). Abbreviations for data quoted
are: s, singlet; d, doublet; t, triplet; quin, quintet; dd, doublet of
doublets; m, multiplet; br s, broad signal. The centers of the peaks
of CDCl3 and CH3OH-d4 were used as internal references (1H NMR
CDCl3: 7.26 ppm; CH3OH-d4: 3.31; 13C NMR CDCl3: 77.00 ppm;
CH3OH-d4: 49.00 ppm). IR spectra of compounds were recorded
as potassium bromide pellets on a Jasco FT/IR-400 spectrometer.
Dry solvents were used throughout. The electron impact (EI) mass
spectra were measured on Finnigan Mat 8200. Reagents used for
synthesis were purchased from Aldrich, Fluka, and Merck compa-
nies. Organic solvents were purchased from Merck Company.
Thin-layer chromatographies were done on pre-coated Silica Gel
60 F254 plates (Merck). The spots were visualized with UV light
or iodine. Microanalyses (C, H, N) of new compounds agreed with
the theoretical value within 0.4%.
4.3.3. 1-(Phenethyl)-4-[(2,6-dichlorobenzylidene)hydrazono)]-
1,4-dihydropyridine 5c
Yield 55%; mp 145 °C; IR (KBr)
m 1636, 1484, 1390, 1182, 826,
753 cmꢀ1 1H NMR (CH3OH-d4): d ppm 3.04 (2H, t, J = 6.8 Hz,
;
CH2), 4.06 (2H, t, J = 6.9 Hz, CH2), 6.29 (1H, dd, J = 2.7/7.7 Hz),
7.11 (1H, dd, J = 2.7/7.4 Hz), 7.17 (1H, dd, J = 1.5/8.3 Hz), 7.22–
7.30 (7H, m), 7.42 (2H, d, J = 8.3 Hz), 8.51 (1H, s, N@CH); 13C
NMR (CH3OH-d4): d ppm 38.04 (t), 58.73 (t), 108.21 (d), 111.84
(d), 128.00 (d), 129.81 (d), 130.04 (d), 130.25 (d), 130.49 (d),
132.96 (s), 135.81 (s), 138.55 (s), 139.75 (d), 140.18 (d), 145.33
(d), 163.16 (s); EI-MS m/z (% relative intensity): 371 (M+2), 369
(M+), 334 (24), 105 (100), 79 (20). Anal. Calcd for (C20H17N3Cl2;
370.28 g/mol): C, 64.9; H, 4.63; N, 11.4. Found: C, 64.7; H, 4.78;
N, 11.2.
The compounds 4-[(2,6-dichlorobenzyloxyimino)methyl]-1-(3-
phenylpropyl)-piperidinium hydrochloride (2), 1-(3-phenylpro-
pyl)-4-piperidone oxime (3), and 1-(2-phenylpropyl)piperidin-4-
one O-(2,6-dichlorobenzyl)-oxime (4) were prepared as described
in Ref. 27.
4.3.4. 1-(3-Phenylpropyl)-4-[(2,6-dichlorobenzylidene)hydraz-
ono)]-1,4-dihydropyridine 5d
4.2. Synthesis of 4-(2-(2,6-dichlorobenzylidene)hydrazinyl)-
pyridine 6
Yield 52%; yellowish oil; IR (KBr)
m 1644, 1509, 1393, 1186, 825,
774 cmꢀ1 1H NMR (CH3OH-d4): d ppm 2.12 (2H, quin, J = 7.3 Hz,
;
To a solution of 2 g (13.7 mmol) 4-hydrazinylpyridine hydro-
chloride in EtOH/H2O (30 ml; 1:1) and 2% triethylamine a solution
of 2.5 g (14.3 mmol) 2,6-dichlorbenzaldehyde in EtOH (10 ml) was
added. The reaction mixture was refluxed for 12 h, evaporated and
the residues purified by chromatography (silica gel, EtOH/H2O 1:1
with 2% triethylamine). Yield: 2.7 g (76%) of a light yellow solid
CH2), 2.68 (2H, t, J = 7.4 Hz, CH2), 3.86 (2H, t, J = 7.3 Hz, CH2),
6.42 (1H, d, J = 6.3 Hz,), 7.18–7.33 (7H, m), 7.39–7.48 (4H, m),
8.52 (1H, s, N@CH); 13C NMR (CH3OH-d4): d ppm 33.38 (t), 54.90
(t), 56.86 (t), 108.45 (d), 112.15 (d), 127.27 (d), 129.41 (d),
129.64 (d), 130.25 (d), 130.51 (d), 132.99 (s), 135.83 (s), 139.69
(d), 140.17 (d), 141.96 (s), 145.33 (d), 163.21 (s); EI-MS m/z (% rel-
ative intensity): 385 (M+2), 383 (M+), 348 (24), 230 (11), 91 (100),
79 (11), 51 (11), 41 (11). Anal. Calcd for (C21H19N3Cl2; 384.31): C,
65.6; H, 4.98; N, 11.9. Found: C, 65.6; H, 5.22; N, 10.8.
(mp 242 °C); IR (ATR)
m 769, 991, 1120, 1209, 1315, 1421, 1550,
2028, 2159, 2832; 1H NMR (DMSO-d6): d ppm 6.96 (2H, d, J
= 6.9 Hz, pyr N–CH–CH), 7.37 (1H, dd, J = 8.5/7.7, Ph), 7.5 (2H, d,
J = 8.1, Ph), 8.21 (1H, s, N@CH), 8.25 (2H, d, J = 6.3, pyr N–CH),
11.12 (1H, s, NH).
13C NMR (DMSO-d6): 107.25 (pyr N–C–C), 129.52 (Ph), 130.13
(Ph), 130.20 (Ph Cl–C–C), 133.50 (Ph C–Cl), 134.97 (N@C), 150.15
(pyr N–C), 150.36 (pyr N–C–C–C).
4.3.5. 1-(2,6-Dichlorobenzyl)-4-[(2,6-dichlorobenzylidene)hydr-
azono)]-1,4-dihydropyridine 5e
Yield 74%; yellowish oil; mp 204 °C; IR (KBr)
m 1640, 1486,
1436, 1170, 825, 780 cmꢀ1 1H NMR (CDCl3): d ppm 5.14 (2H, s,
;
CH2-Ph), 6.53 (1H, br s), 7.08–7.33 (8H, m), 7.42 (2H, d,
J = 8.6 Hz), 8.69 (1H, s, N@CH); 13C NMR (CDCl3): d ppm 53.82
(t), 107.85 (d), 112.31 (d), 128.81 (d), 128.95 (d), 129.17 (d),
130.16 (s), 131.02 (d), 135.03 (s), 136.71 (s), 137.56 (s), 138.62
(d), 141.15 (d), 146.49 (d), 163.05 (s); EI-MS m/z (% relative inten-
sity): 425 (M+2), 421 (M+1), 423 (M+), 159 (100), 123 (15), 51 (15),
49 (35). Anal. Calcd for (C19H13N3Cl4; 425.14): C, 53.7; H, 3.08; N,
9.9. Found: C, 53.9; H, 3.23; N, 9.7.
4.3. General procedure for the synthesis of the 1-substituted
(benzylidene-hydrazono)-1,4-dihydropyridines derivatives
1-Substituted 4(1H)-pyridinone hydrazones derivatives were
synthesized in analogy to Douglas et al.28 All compounds were
recrystallized from methanol, ethanol–water or methanol–water
mixture.
4.3.1. 1-Methyl-4-[(2,6-dichlorobenzylidene)hydrazono]-1,4-
dihydropyridine 5a
4.4. pKa determination
Yield 88%; mp 123 °C; IR (KBr)
m 1644, 1493, 1381, 1197, 833,
774 cmꢀ1 1H NMR (CH3OH-d4): d ppm 3.64 (3H, s, N–CH3), 6.40
;
The experimental determination of the pKa values was per-
formed on a Sirius PCA-101 (Sirius Analytical Instruments Ltd, For-
est Row, East Sussex, United Kingdom). The compounds were
dissolved in 0.15 M KCl to keep the ionic strength constant during
titration. 30% Dioxane (compounds 2 and 4) and 80% methanol
(5d) were used as modifier for complete dissolution. The pH was
adjusted to 1.8 by adding 0.5 M HCl. With 0.5 M KOH titration
was performed. The dissociation constant calculated by the pro-
gram PKALOGP5.1 (Sirius Analytical Instruments Ltd) was determined
(1H, dd, J = 3.0/7.7 Hz), 7.22 (1H, dd, J = 3.0/8.0 Hz), 7.28 (1H, dd,
J = 7.8/8.6 Hz), 7.35–7.39 (2H, m), 7.45 (2H, d, J = 8.0 Hz), 8.50 (1H,
s, N@CH); 13C NMR (CH3OH-d4): d ppm 43.21 (q), 108.28 (d),
112.01 (d), 130.16 (d), 130.39 (d), 132.96 (s), 135.74 (s), 140.37 (d),
140.84 (d), 145.15 (d), 163.07 (s). EI-MS m/z (% relative intensity):
281 (M+2), 279 (M+, 10), 244 (100), 181 (25), 93 (60), 92 (27), 66
(32), 42 (24). Anal. Calcd for (C13H11N3Cl2 ꢂ H2O; 298.17 g/mol): C,
52.4; H, 4.39; N, 14.1. Found: C, 52.6; H, 3.99; N, 14.1.

V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
2057
for different additions of the modifier. The pKa values were ob-
tained by extrapolation of the calibration line. In order to calibrate
the electrode, blank titrations were performed, titrating the 0.15 M
KCl solution and determining the corresponding ‘Four Plus™’-
parameters.
Crystal data for 5a: C13H11Cl2N3 ꢂ H2O, Mr = 280.15644
(C13H11Cl2N3) + 18.01528 (H2O) = 298.17172, yellow needle, 0.20ꢂ
0.10ꢂ0.06, monoclinic space group P21/n, a = 13.9396(3), b =
7.0876(2), c = 14.3087(3) Å, b = 107.4800(10)°, Vc = 1348.39(6) Å3,
Z = 4,
q , l , F(0 0 0) = 616, T =
calcd = 1.469 g cmꢀ3 = 0.476 mmꢀ1
100(2) K, R1 = 0.0389, wR2 = 0.0953, 4124 independent reflections
[2h 6 64°] and 178 parameters.
4.5. Log P determination
Crystallographic data have been deposited with the Cambridge
Crystallographic Data Center as supplementary publication no.
CCDC-709897. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.a-
c.uk/data_request/cif.
4.5.1. Log P determination by means of RP HPLC
The partition coefficients of all compounds were determined by
RP-chromatography using methanol/phosphate buffer 70/30 as an
eluent. The chromatographic systems were calibrated with solutes
for which an experimental octanol/water partition coefficient was
available.43 The capacity factors of the reference substances were
correlated against with experimental octanol/water log P values,
and the obtained correlation equation was used to calculate log P
values of the tested compounds. The experiments were performed
on a LiChroCARTÒ 125–4.6 HPLC-Cartridge filled with LiChrospherÒ
4.7. Inhibition of AChE
The inhibitory potency against AChE was investigated by
slightly modified colorimetric method of Ellman et al.44 as reported
in Ref. 48. Acetylcholinesterase (AChE) –E.C. 3.1.1.7 from Electric
Eel was purchased from Sigma–Aldrich (Steinheim, Germany).
5,50-Dithiobis-(2-nitrobenzoic acid), potassium dihydrogen phos-
phate, potassium hydroxide, sodium hydrogen carbonate, and
acetylthiocholine iodide were obtained from Fluka (Buchs, Swit-
zerland). Spectrophotometric measurements were performed on
a Varian Cary 50 UV–vis spectrophotometer. Details of the inhibi-
tion study are given in Kapková et al.22
100, RP18, 5 lm, endcapped (Merck, Darmstadt, Germany). The
methanol (LiChrosolvÒ (Merck))/buffer (70/30) mobile phase was
used at a flow-rate of 1.0 ml. The phosphate buffer (pH 7.4, DAB
1999) was obtained by mixing a 0.2 M potassium dihydrogen
phosphate solution (1000 ml) and 0.1 M sodium hydroxide
(1573 ml). 0.02% N,N-Dimethylhexylamine was added to minimize
the peak tailing.
The following aromatic compounds were applied for calibra-
tion: 2-phenylethanol, benzene, dimethylaniline, toluene, biphe-
nyl, anthracene.
4.8. Inhibition of BuChE
The analytes were dissolved in methanol at a concentration of
4 lg/ml. Experiments were run in duplicate, and peak maxima
BuChE inhibitory activity was investigated by a modified Ell-
man’s test, conducted as in Ingkaninan et al.45 Details of the meth-
od using the BuChE (EC 3.1.1.8, from horse serum) are given in
Fallarero et al.7
were determined at the respective absorption maximum. Capacity
ratios were determined as k0 = (tr ꢀ to)/to, where tr is the average
retention time of the analyte and to is the average retention time
of the solvent. A linear regression was performed for the log k0/
log P data of the reference compounds (y = 2.5935 + 1.8353;
r2 = 0.9782), and the regression equation was used to calculate
the log P of the compounds.
4.9. Thioflavin T Fluorescence Assay (all carried out in analogy
to46,47
)
4.9.1. Ab
Amyloid b-protein (1–40) trifluoroacetate salt was obtained
from Bachem (Switzerland, Lot 1012093; Lot 9004754). The pep-
tide was stored at ꢀ20 °C, as recommended by the manufacturer.
It was dissolved in hexafluoroisopropanol (HFIP) at 20 mg/ml. This
solution was kept at room temperature until the peptide was com-
pletely dissolved (30 min to 1 h). The HFIP was removed under a
stream of nitrogen until a clear film remained in the test tube.
The residue was then dissolved in DMSO to obtain a 2 mM stock
solution, which was subsequently stored frozen at ꢀ20 °C for max-
imum three days.
4.5.2. Log P determination by means of shake flask method
Partition coefficients of 5d and 5e were determined by shake
flask method using UV spectroscopy and were studied in phos-
phate buffer at pH 5, 6, 7, and 7.4 for each compound. The phos-
phate buffer was saturated with octanol prior to partitioning by
adding octanol, mixing, and allowing the phases to separate over-
night; in the same manner portions of octanol were saturated with
the buffer. A stock solution for each compound was prepared at
10ꢀ5 mol/l concentration in octanol saturated with phosphate buf-
fer and the absorbance (Ao) was measured between 314
and360 nm. A 1.0 ml portion of this solution was shaken at 37 °C
for 4 h with 1.0 ml phosphate buffer saturated with octanol. The
emulsion was allowed to stand overnight, centrifuged for 5 min
(3000 U/min), the water layer separated and again the absorption
(A1) measured and the log P value was calculated as log P = log
((Ao ꢀ A1)/A1).
4.9.2. ThT assay Ab
A 2 mM stock solution of Ab in DMSO was diluted in buffer
(25 mM NaH2PO4, 120 mM NaCl, 3
and a final pH of 7.4) to reach a final peptide concentration of
100 M. Inhibitors were added to reach concentrations between
lM thioflavin T, 0.02% NaN3
l
0.005 and 0.5 mM. Incubations were performed at 37 °C on 96-well
fluorescence microtiter plates (Nunc GmbH, Wiesbaden, Ger-
many). The fluorescence was measured (excitation wavelength
450 nm, emission wavelength 482 nm) on a Cary Eclipse fluores-
cence spectrophotometer (Varian, Darmstadt, Germany).
4.6. Crystal structure determination
The crystal data of 5a were collected at Bruker X8Apex diffrac-
tometer with CCD area detector and multi-layer mirror monochro-
4.9.3. ThT assay fibril formation with peptide HHQKLVFFAED
Inhibitors were diluted in buffer (25 mM NaH2PO4, 120 mM
NaCl, 3 lM thioflavin T, 0.02% NaN3 and a final pH of 7.4) to reach
concentrations between 0.005 and 0.5 mM. A 2 mM stock solution
of the peptide HHQKLVFFAED in DMSO was added to reach a final
mated Mo K
a radiation. The structure was solved using direct
methods, refine with SHELX software package (G. Sheldrick, Univer-
sity of Göttingen 1997) and expanded using Fourier techniques. All
non-hydrogen atoms were refined anisotropically. Hydrogen
atoms were assigned idealized positions and were included in
structure factors calculations.
peptide concentration of 100
lM. Incubations were performed on
96-well fluorescence microtiter plates (Nunc GmbH, Wiesbaden,

2058
V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
Germany). The fluorescence was measured (excitation wavelength
450 nm, emission wavelength 482 nm) on a Cary Eclipse fluores-
cence spectrophotometer (Varian, Darmstadt, Germany).
The protein structure was prepared as follows: The co-crystallized
ligand (1-benzyl-4-[(5,6-dimethoxy-1-indanon-2-yl)methyl]piper-
idine) in EVE was used as a starting point. All amino acids where at
least one atom has a distance of not more than 6 Å to the original
ligand were selected for the binding pocket definition. This re-
duced protein structure information was then exported in the na-
tive Openeye format. Next, the binding pocket was read into
Openeye Fred. Every conformer of every ligand was docked into
this pre-defined binding pocket by using Fred’s standard parame-
ters. Beside the binding pocket no constraints were applied. The
best pose for each compound was then used by applying all scoring
functions available in FRED to obtain a consensus score for a rank-
ing of the docked conformers. This scoring was only applied within
conformer sets, but not for the qualitative analysis of binding
modes. Only the analysis of this top-ranking result in the crystal
structure was then analyzed in depth and the result was
interpreted.
4.9.4. ThT assay fibril destruction with peptide HHQKLVFFAED
A 2 mM stock solution of the peptide HHQKLVFFAED in DMSO
was diluted in buffer (25 mM NaH2PO4, 120 mM NaCl, 3
vin T, 0.02% NaN3 and a final pH of 7.4) to reach a final peptide con-
centration of 100 M. Incubations were performed on 96-well
lM thiofla-
l
fluorescence microtiter plates (Nunc GmbH, Wiesbaden, Germany)
for 10 h. Inhibitors were added to reach concentrations between
0.005 and 0.5 mM. The fluorescence was measured (excitation
wavelength450 nm, emissionwavelength482 nm) ona CaryEclipse
fluorescence spectrophotometer (Varian, Darmstadt, Germany).
4.10. Solid phase synthesis of HHQKLVFFAED
The shortened peptide was synthesized by solid-phase peptide
synthesis on a Milligen 9050 PepSynthesiser using standard Fmoc
protocol.48 Synthesis was performed in DMF/DCM (60:40) with
DIC/HOBt activation starting from Fmoc-Asp(OtBu)-Wang-resin.
The peptide was cleaved from the resin by treatment with TFA/
H2O/TIS (95/2.5/2.5). The following side chain protecting groups
were used during the automated synthesis: tert-butyl ester (Glu),
tert-butyloxycarbonyl (Lys), and trityl (Gln, His).
Acknowledgements
The authors thank the TUBITAK-DFG for a grant given to V.A.
and the Fonds der Chemischen Industrie for financial support.
U.H. and B.E. thank the SFB630 for financial support. Thanks are
due to Professor Schirmeister for helping with the peptide synthe-
sis and providing the synthesizer. J.S. thanks Markus Kossner/TU
Braunschweig for fruitful discussions about the use of Openeye
FRED.
The product substrate as TFA salt was purified by HPLC
(H2O + 0.1% TFA; acetonitrile (Merck); LiChroCARTÒ 125–4.6
HPLC-Cartridge (LiChrospherÒ 100, RP18 (5
lm, endcapped),
(Merck); pressure: 143 bar; flow-rate: 1.0 ml; temperature:
23 °C; eluent gradient: acetonitrile/water + 0.1% TFA 20–80% in
40 min. The peptide was stored at ꢀ20 °C. For each test a fresh
solution of 2 mM in DMSO was prepared.
References and notes
1. Rajkumar, S.; Tanyakitpisal, P.; Chandra, V. WHO, 2001.
2. Jakob-Roetne, R.; Jacobsen, H. Angew. Chem. 2009, 121, 3074.
3. Roberson, E. D.; Mucke, L. Science 2006, 314, 781.
4. Davies, P.; Maloney, A. J. Lancet 1976, 2, 1403.
5. Bartus, R. T.; Dean, R. L., III; Beer, B.; Lippa, A. S. Science 1982, 217, 408.
6. Holzgrabe, U.; Kapkova, P.; Alptüzün, V.; Scheiber, J.; Kugelmann, E. Expert
Opin. Ther. Targets 2007, 11, 161.
7. Fallarero, A.; Oinonen, P.; Gupta, S.; Blom, P.; Galkin, A. Pharmacol. Res. 2008,
58, 215.
8. Decker, M.; Kraus, B.; Heilmann, J. Bioorg. Med. Chem. 2008, 16, 4252.
9. Kamal, M. A.; Klein, P.; Luo, W.; Li, Y.; Holloway, H. W.; Tweedie, D.; Greig, N. H.
Neurochem. Res. 2008, 33, 745.
4.11. Determination of the protonated sites
4.11.1. Conformational search
The conformational search was performed using the MMFF94
force field as implemented in MacroModel 8.0.49
4.11.2. Geometry optimization
10. Doody, R. S.; Gavrilova, S. I.; Thomas, R. G.; Aisen, P. S.; Bachurin, S. O.; Seely, L.;
Hung, D. Lancet 2008, 372, 207.
11. Hamley, I. W. Angew. Chem., Int. Ed. 2007, 46, 8128.
The DFT calculations were carried out with TURBOMOL 5.6 pro-
gram package.37 Full geometry optimizations were performed at
the BP86/TZVP level of theory using the RI approximation. Solva-
tion effects were taken into account via the COSMO approach as
implemented in TURBOMOL employing an dielectric constant of
12. Olsen, R. E.; Marcin, L. R. Annu. Rep. Med. Chem. 2007, 42, 27.
13. Austen, B. M.; Paleologou, K. E.; Ali, S. A. E.; Qureshi, M. M.; Allsop, D.; El-Agnaf
Biochemistry 2008, 47, 1984.
14. Török, B.; Dasgupta, S.; Török, M. Curr. Bioactive Comp. 2008, 4, 159.
15. Recanatini, M.; Valenti, P. Curr. Pharm. Des. 2004, 10, 3157.
16. Piazzi, L.; Rampa, A.; Bisi, A.; Gobbi, S.; Belutti, F.; Cavalli, A.; Bartolini, M.;
Andrisano, V.; Valenti, P.; Recanatini, M. J. Med. Chem. 2003, 46, 2279.
17. Cavalli, A.; Bolognesi, M. L.; Minarini, A.; Rosini, M.; Tumiatti, V.; Recanatini,
M.; Melchiorre, C. J. Med. Chem. 2008, 51, 347.
e
= 78 to simulate a water-like solvent.
4.11.3. pKa calculation
pKa determination was performed by COSMO-RS approach
18. Hardy, J. J. Neurochem. 2009, 110, 1129.
19. Greig, N. H.; Sambamurti, K.; Yu, Q. S.; Brossi, A.; Bruinsma, G. B.; Lahiri, D. K.
implemented in COSMOthermX. Calculating
ated molecule by TZVP the corresponding pKa values were
determined.
DGdiss of the proton-
Curr. Alzheimer Res. 2005, 2, 281.
20. Rizzo, S.; Rivière, C.; Piazzi, L.; Bisi, A.; Gobbi, S.; Bartolini, M.; Andrisano, V.;
Morroni, F.; Tarozzi, A.; Monti, J. P.; Rampa, A. J. Med. Chem. 2008, 51, 2883.
21. Camps, P.; Formosa, X.; Galdeano, C.; Gómez, T.; Muñoz-Torrero, D.;
Scarpellini, M.; Viayna, E.; Badia, A.; Clos, M. V.; Camins, A.; Pallàs, M.;
Bartolini, M.; Mancini, F.; Andrisano, V.; Estelrich, J.; Lizondo, M.; Bidon-
Chanal, A.; Luque, F. J. J. Med. Chem. 2008, 51, 3588.
4.12. Docking
The structures carrying protonation state information deter-
mined in the first step where exported with their connection tables
to a format readable by Openeye Omega. Precalculated coordinates
were not used for the subsequent steps as the applied docking
algorithm requires a set of conformers for each molecule that is
docked. This set of possible conformers for each compound was
generated using Omega’s standard procedure with the option
‘-fromCT’. Further constraints were not used in this step. The here-
by generated conformers were then docked into the crystal struc-
ture of T. californica AChE (PDB: 1EVE). This structure was chosen
because of its previous successful use in similar experiments.22
22. Kapková, P.; Stiefl, N.; Sürig, U.; Engels, B.; Baumann, K.; Holzgrabe, U. Arch.
Pharm. (Weinheim) 2003, 336, 523.
23. Alptüzün, V.; Kapková, P.; Baumann, K.; Erciyas, E.; Holzgrabe, U. J. Pharm.
Pharmacol. 2003, 10, 1397.
24. Kapková, P.; Alptüzün, V.; Frey, P.; Erciyas, E.; Holzgrabe, U. Bioorg. Med. Chem.
2006, 14, 472.
25. Bodor, N.; Buchwald, P. Drug Discovery Today 2002, 7, 766.
26. Shoval, H.; Lichtenberg, D.; Gazit, E. Amyloid 2007, 14, 73.
27. Goebel, T.; Ulmer, D.; Projahn, H.; Ponte-Sucre, A.; Saeftel, M.; Hoerauf, A.;
Kaiser, A.; Hauber, I.; Hauber, J.; Holzgrabe, U. J. Med. Chem. 2008, 51, 238.
28. Douglas, W. A.; Fishinger, J. J.; Gund, P.; Haris, E. E.; Olson, G.; Patchett, A. A.;
Ruyle, W. V. J. Med. Chem. 1977, 20, 939.
29. Avdeef, A.; Box, K. J.; Comer, J. E. A.; Gilges, M.; Hadley, M.; Hibbert, C.;
Patterson, W.; Tam, K. Y. J. Pharm. Biomed. Anal. 1999, 20, 631.

V. Alptüzün et al. / Bioorg. Med. Chem. 18 (2010) 2049–2059
2059
30. Van der Waterbeemd, H.; Kansy, M.; Wagner, B.; Fischer, H. In Lipophilicity in
Drug Action and Toxicology; Pliska, V., Testa, B., Van der Waterbeemd, H., Eds.;
Series: Methods and Principles in Medicinal Chemistry; VCH, 1996; Vol. 4, p 89.
31. Gazit, E. FEBS J. 2005, 272, 5971.
41. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Adv. Drug Delivery Rev.
2001, 46, 3.
42. Silverman, R. B. The Organic Chemistry of Drug Design and Drug Action, 2nd ed.;
Elsevier: Amsterdam, 2004. p 66.
32. De Felice, F. G.; Vieira, M. N. N.; Saraiva, L. M.; Figueroa-Villar, J. D.; Garcia-
Abreu, J.; Liu, R.; Chang, L.; Klein, W. L.; Ferreira, S. T. FASEB J. 2004, 18,
1366.
33. Camus, M. S.; Dos Santos, S.; Chandravarkar, A.; Mandal, B.; Schmid, A. W.;
Tuchscherer, G.; Mutter, M.; Lashuel, H. A. ChemBioChem 2008, 9, 2104.
34. Coupez, B.; Lewis, R. A. Curr. Med. Chem. 2006, 13, 2995.
35. Halgren, T. A. J. Comput. Chem. 1996, 17, 490–641.
36. Klamt, A.; Schüürmann, G. J. Chem. Soc., Perkin Trans. 2 1993, 799.
37. Schäfer, A.; Klamt, A.; Sattel, D.; Lohrenz, J. C. W.; Eckert, F. Phys. Chem. Chem.
Phys. 2000, 2, 2187.
43. Hansch, C.; Sammer, P. G.; Taylor, J. B. Comprehensive Medicinal Chemistry, 1st
ed., Pergamon Press: Oxford; 1990; Vol. 6.
44. Ellman, G. L.; Courtney, D. K.; Andres, V.; Feather-Stone, R. M. Biochem. Pharmacol.
1961, 7, 88.
45. Ingkaninan, K.; de Best, C. M.; van der Heijden, R.; Hofte, H. J.; Karabatak, B.; Irth,
H.; Tjaden, U. R.; van der Greef, J.; Verpoorte, R. J. Chromatogr., A 2000, 872, 61.
46. Goldsbury, C. S.; Wirtz, S.; Müller, S. A.; Sunderji, S.; Wicki, P.; Aebi, U.; Frey, P.
J. Struct. Biol. 2000, 130, 217.
47. LeVine, H. Protein Sci. 1993, 2, 404.
48. Bodanszky, M. Principles of Peptide Synthesis; Springer: Berlin, Heidelberg, 1993.
49. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.; Lipton, M.; Caufield,
C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput. Chem. 1990, 11, 440.
50. Botero Cid, M. H.; Holzgrabe, U.; Kostenis, E.; Mohr, K.; Tränkle, C. J. Med. Chem.
1994, 37, 1439.
38. Becke, A. D. Phys. Rev. A 1988, 38, 3098.
39. Eichkorn, K.; Treutler, O.; Öhm, H.; Häser, M.; Ahlrichs, R. Chem. Phys. Lett.
1995, 240, 283.
40. Klamt, A. J. Phys. Chem. 1995, 99, 2224.