Lead identification of acetylcholinesterase inhibitors-histamine H3 receptor antagonists from molecular modeling

Article abstract of DOI:10.1016/j.bmc.2007.12.048

Currently, the only clinically effective treatment for Alzheimer's disease (AD) is the use of acetylcholinesterase (AChE) inhibitors. These inhibitors have limited efficacy in that they only treat the symptoms and not the disease itself. Additionally, they often have unpleasant side effects. Here we consider the viability of a single molecule having the actions of both an AChE inhibitor and histamine H₃ receptor antagonist. Both histamine H₃ receptor antagonists and AChE inhibitors improve and augment cholinergic neurotransmission in the cortex. However, whereas an AChE inhibitor will impart its effect everywhere, a histamine H₃ antagonist will raise acetylcholine levels mostly in the brain as its mode of action will primarily be on the central nervous system. Therefore, the combination of both activities in a single molecule could be advantageous. Indeed, studies suggest an appropriate dual-acting compound may offer the desired therapeutic effect with fewer unpleasant side effects [CNS Drugs 2004, 18, 827]. Further, recent studies² indicate the peripheral anionic site (PAS) of AChE interacts with the β-amyloid (βA) peptide. Consequently, a molecule capable of disrupting this interaction may have a significant impact on the production of or the aggregation of βA. This may result in slowing down the progression of the disease rather than only treating the symptoms as current therapies do. Here, we detail how the use of the available crystal structure information, pharmacophore modeling and docking (automated, manual, classical, and QM/MM) lead to the identification of an AChE inhibitor-histamine H₃ receptor antagonist. Further, based on our models we speculate that this dual-acting compound may interact with the PAS. Such a dual-acting compound may be able to affect the pathology of AD in addition to providing symptomatic relief.

Full text of DOI:10.1016/j.bmc.2007.12.048

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 2968–2973
Lead identification of acetylcholinesterase inhibitors–histamine
H receptor antagonists from molecular modeling
3
a,
a
a
*
Scott D. Bembenek, John M. Keith, Michael A. Letavic,
b
c
a
a
Richard Apodaca, Ann J. Barbier, Lisa Dvorak, Leah Aluisio,
Kirsten L. Miller, Timothy W. Lovenberg and Nicholas I. Carruthers
a
a
a
a
Johnson & Johnson Pharmaceutical Research and Development, L.L.C. 3210 Merryfield Row, San Diego, CA 92121, USA
b
Metamolecular, LLC, PO Box 13091, La Jolla, CA 92039-3091, USA
Shire Human Genetic Therapies, 700 Main Street, Cambridge, MA 02139, USA
c
Received 26 September 2007; revised 19 December 2007; accepted 20 December 2007
Available online 25 December 2007
Abstract—Currently, the only clinically effective treatment for Alzheimer’s disease (AD) is the use of acetylcholinesterase (AChE)
inhibitors. These inhibitors have limited efficacy in that they only treat the symptoms and not the disease itself. Additionally, they
often have unpleasant side effects. Here we consider the viability of a single molecule having the actions of both an AChE inhibitor
and histamine H receptor antagonist. Both histamine H receptor antagonists and AChE inhibitors improve and augment cholin-
3
3
3
ergic neurotransmission in the cortex. However, whereas an AChE inhibitor will impart its effect everywhere, a histamine H antag-
onist will raise acetylcholine levels mostly in the brain as its mode of action will primarily be on the central nervous system.
Therefore, the combination of both activities in a single molecule could be advantageous. Indeed, studies suggest an appropriate
dual-acting compound may offer the desired therapeutic effect with fewer unpleasant side effects [CNS Drugs 2004, 18, 827]. Further,
2
recent studies indicate the peripheral anionic site (PAS) of AChE interacts with the b-amyloid (bA) peptide. Consequently, a mol-
ecule capable of disrupting this interaction may have a significant impact on the production of or the aggregation of bA. This may
result in slowing down the progression of the disease rather than only treating the symptoms as current therapies do. Here, we detail
how the use of the available crystal structure information, pharmacophore modeling and docking (automated, manual, classical,
3
and QM/MM) lead to the identification of an AChE inhibitor–histamine H receptor antagonist. Further, based on our models
we speculate that this dual-acting compound may interact with the PAS. Such a dual-acting compound may be able to affect the
pathology of AD in addition to providing symptomatic relief.
Published by Elsevier Ltd.
1
. Introduction
cognitive decline without altering the underlying pathol-
5
ogy. These compounds, however, leave much to be de-
sired, both in efficacy and tolerability. Their efficacy is
limited; the preferred target patient population suffers
Alzheimer’s disease is the most common form of neuro-
3
degenerative dementia. It accounts for approximately
5
5
0–60% of the overall cases of dementia among individ-
a mild to moderate cognitive impairment. Additionally,
4
uals over the age of 65 years. Unfortunately, the thera-
peutic options for Alzheimer’s disease are limited. No
true disease-modifying therapies are known and at best
physicians can only hope to alleviate the cognitive defi-
cits that characterize this disease with symptomatic
treatments. The most frequently prescribed anti-Alzhei-
mer’s drugs are the acetylcholinesterase (AChE) inhibi-
tors, which promote memory function and delay the
they are associated with unpleasant mechanism-based
side effects such as gastrointestinal discomfort and
risk of bradycardia, especially the nonselective
AChE inhibitors, which inhibit AChE as well as
6
butyrylcholinesterase.
Clinical studies suggest combination therapies can pro-
1
vide synergistic benefits in some instances. An alterna-
tive approach taken by several pharmaceutical
7
companies is to combine AChE inhibition with other
mechanisms targeted at symptoms of Alzheimer’s dis-
ease using a single molecular entity. For instance, inhibi-
tion of the serotonin reuptake transporter is expected to
Keywords: Acetylcholinesterase; Histamine; Modeling; Alzheimer’s.
*
Corresponding author. Tel.: +1 858 320 3375; e-mail:
sbembene@prdus.jnj.com
0968-0896/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.bmc.2007.12.048

S. D. Bembenek et al. / Bioorg. Med. Chem. 16 (2008) 2968–2973
2969
bestow antidepressant efficacy, which may alleviate
some of the psychiatric symptoms observed in Alzhei-
mer’s populations. Dual-acting compounds of AChE
and the serotonin transporter have been synthesized.
The argument in favor of the combination of an AChE
Further, there is evidence for some AChE inhibitors
exhibiting therapeutic benefits beyond influencing cho-
linergic transmission. AChE inhibitors able to interact
with the peripheral anionic site (PAS) are of particular
relevance. In addition to their ability to stimulate the
cholinergic system, these inhibitors may be able to dis-
rupt the interactions between AChE and the b-amyloid
(bA) peptide and thereby be capable of slowing the pro-
gression of the disease by inhibiting the production or
8
inhibitor and histamine H receptor antagonist can be
3
separated into a neurochemical and a procognitive argu-
ment. Neurochemically it is believed the decline in cog-
nitive function in normal aging, mild cognitive
impairment, Alzheimer’s disease, and related conditions
is caused by a decrease in cholinergic neurotransmission
in the cortex. Both AChE inhibitors and histamine H3
receptor antagonists improve and augment cholinergic
neurotransmission in the cortex, albeit by different
mechanisms. Therefore, the combination of both activi-
ties in a single molecule may lead to a potentiation of
cholinergic neurotransmission. Procognitive benefits
are expected from both AChE inhibitors and histamine
H3 receptor antagonists as both exhibit memory-
enhancing effects. The combination of both activities
in a single molecule might be expected to have signifi-
cant synergistic or additive effects on cognitive function
and memory.
3
aggregation of the bA promoted by AChE.
We describe herein the molecular modeling efforts that
lead to the identification of an AChE inhibitor–hista-
mine H receptor antagonist. The use of available crystal
3
structure information, pharmacophore modeling, classi-
cal, and quantum calculations lead us to the discovery.
Our models also predict this dual-acting compound
can interact with the PAS thus possibly having the po-
tential to directly impact the disease itself unlike current
therapies.
2. The acetylcholin esterase enzyme
Inhibition of AChE increases the synaptic levels of ace-
tylcholine and thereby promotes cholinergic neurotrans-
mission. The effect of AChE inhibitors is thus to
stabilize and prolong the biological half-life of acetyl-
The ability to study AChE by molecular modeling meth-
1
1
ods (in particular by structure-based design) is bene-
fited by the numerous crystal structures available both
of the apo and complexed forms. The three-dimensional
structure of Torpedo californica AChE was initially
choline. In contrast, histamine H receptor antagonists
3
˚
12
increase cholinergic neurotransmission via a distinct
mechanism. The presynaptic inhibitor–histamine H3
receptor, when activated, decreases the release of acetyl-
choline from cholinergic neurons. This has been shown
solved at a resolution of 2.8 A and later refined to
13
˚
2.5 A. For our modeling purposes we used the struc-
ture of T. californica (PDB ID: 1EVE) complexed with
Ò
14
the AChE inhibitor Aricept (donepezil). Reference
to residue numbers will follow from this structure. The
enzyme monomer is an a/b serine hydrolase consisting
of 537 residues with a 12-stranded mixed b sheet sur-
9
both in the gastrointestinal and central nervous sys-
1
0
tem. Consequently, when a histamine H3 receptor
antagonist is present, an increased release of acetylcho-
line is observed. Thus, both mechanisms contribute to
the same neurochemical end result (increased synaptic
levels of acetylcholine) in different ways. The histamine
1
2
rounded by 14 a helices. The active site is striking in
˚
that it is a deep, narrow tunnel approximately 20 A
long. This tunnel penetrates halfway into the enzyme
and widens out at the end. Fourteen highly conserved
residues cover a substantial portion of the tunnel. Crys-
tallographic and theoretical analyses suggest approxi-
H receptor antagonists increase the amount of acetyl-
3
choline molecules entering the synaptic space, and
AChE inhibitors prolong their survival time and hence
the likelihood of their interacting with a postsynaptic
cholinergic neurotransmitter and eliciting a biological
effect.
1
5,16
mately 20 water molecules occupy the tunnel
and
indeed, the presence of water is necessary for the deacyl-
˚
ation of acetylcholine. Approximately 4 A above the
bottom of the tunnel is the catalytic machinery: a cata-
lytic triad consisting of Ser 200, His 440, and Glu 327.
AChE is the first serine protease discovered to contain
a Glu rather than an Asp as a catalytic member. It is
in this area of the active site that deacylation of acetyl-
choline occurs. The electrophilic carbon on the ester
group of acetylcholine undergoes nucleophilic attack
from Ser 200 to form a tetrahedral intermediate fol-
lowed by acylation of the enzyme and displacement of
choline. Subsequent hydrolysis restores the enzyme with
the generation of acetic acid. Proximal to the catalytic
Ser 200 are the residues Gly 118, Gly 119, and Ala
201, which comprise the oxyanion hole. The amido pro-
tons of these residues aid in the deacylation process by
forming hydrogen bonds to the carbonyl oxygen of ace-
tylcholine. This orients the electrophilic carbon for
nucleophilic attack by Ser 200. Directly across from
the catalytic triad is Trp 84, which binds the quaternary
In order to achieve a specific, cognition-enhancing level
of cholinergic neurotransmission, a lower dose of an
AChE inhibitor–histamine H receptor antagonist com-
3
pound will likely be required than of either an AChE
inhibitor or histamine H receptor antagonist alone; a
3
combination molecule will have a higher potency for
increasing acetylcholine levels than either an AChE
inhibitor alone or histamine H3 receptor antagonist
alone. It is likely reduced compound requirements for
efficacy will then improve the side effect profile. At pres-
ent, not much is known about the side effects of H3
antagonists. For AChE inhibitors, on the other hand,
the major side effects are nausea and gastrointestinal dis-
6
comfort. By decreasing the amount of acetylcholine re-
leased in the periphery through augmentation of its
release centrally via histamine H receptor antagonism,
3
fewer side effects may result.

2970
S. D. Bembenek et al. / Bioorg. Med. Chem. 16 (2008) 2968–2973
amine of acetylcholine through a cation-p interaction.
This region is known by the unfortunate name of ‘anio-
reveals two sets of key interactions between the protein
and the ligand. One set of interactions is between the
Trp 84 (the AS), located at the base of the gorge, and
one of the quaternary amines of decamethonium. The
other important interaction occurs between the Trp
279 (the PAS), located at the opening of the tunnel,
(
nic’ site, AS, since it was initially believed that binding in
this region occurred due to a collection of negative
charges.) This finding was rather unexpected but is fur-
ther substantiated by theoretical calculations showing
1
7,18
the large stability of cation-p interactions.
There is
and the other quaternary amine on the ligand. Both of
17,18
1
9
also the peripheral ‘anionic’ site (PAS) corresponding
to Trp 279. The role of this site (with regard to the sub-
strate) appears to be to attract, via cation-p and/or
hydrophobic interactions with the quaternary amine
moiety, acetylcholine into the active site. One can imag-
ine Trp 279 attracting acetylcholine into the active site
and assisting in aligning it into the correct orientation
for subsequent deacylation. The PAS has gained much
interest in the design of newer drugs for AD. Recent
data suggest inhibitors interacting with the PAS could
prevent the bA peptide from interacting with AChE.
This could result in the elimination of AChE’s role in
these are cation-p interactions.
proposed three-dimensional pharmacophore for hista-
In Figure 2 is the
2
1
mine H based on our previous work along with a
3
compound (1 in Table 1) known to have good activity
for the histamine H3 receptor. The pharmacophore
2
5
modeling was carried out with Catalyst using the
Hip-Hop option. This procedure takes three-dimen-
sional conformations for each training molecule, identi-
2
6
fies common features, and overlays them in space. We
used the automatic conformer generation feature
employing the flexible option. Three active compounds
we felt captured all of the features required for activity,
while providing significant structural variation as well,
were used to define the three-dimensional pharmaco-
phore model. The three-dimensional pharmacophore
model consists of three features: two positive ionic cen-
ters and an aromatic ring forming the center of the mol-
ecule. The model well explains known experimental
2
0
the fibril formation process. Consequently, it is be-
lieved AChE inhibitors interacting with the PAS could
have an impact on the pathology of AD, rather than just
treating the symptoms as current therapies do.
2
7
3
. Molecular modeling
data.
Our goal was to design an AChE inhibitor–histamine H3
receptor antagonist capable of interacting with the PAS
of AChE. Our initial efforts focused on virtual screening
our in-house library. We sought to use the abundance of
crystal structure data on AChE, our knowledge of the
three-dimensional pharmacophore features of the hista-
For the most part, decamethonium fits this H pharma-
3
cophore rather well with its two quaternary amines and
hydrophobic alkyl chain acting as a replacement for the
aromatic center. However, we expected an aromatic cen-
ter would indeed be favored over a general hydrophobic
for two reasons: first, we concluded from our previous
2
1
21
mine H receptor and modeling to optimize this effort.
3
work the central aromatic ring can make p-stacking
interactions with the residues Tyr 115 and Trp 371 in
Our review of the crystal structure data focused on Ari-
23
Ò
14
22
cept (donepezil), tacrine, galantamine, rivastig-
2
the histamine H receptor; AChE contains several aro-
3
4
22
28
mine, and decamethonium. Decamethonium was of
particular interest to us.
matic resides in its active site (the gating residues)
capable of p-stacking interactions. We hypothesized this
pharmacophore model may well describe inhibitors of
AChE (capable of interactions with the PAS) and hista-
Shown in Figure 1 is the crystal structure of decametho-
2
nium in the active site of AChE. The crystal structure
2
mine H receptor antagonists. Starting from this premise
3
we performed similarity and substructure searches
around 1. In conjunction with this effort we also per-
2
9
formed manual docking of 1 in the active site of AChE
2
using Insight II version 2005 and the consistent va-
5
3
0
lence force field (CVFF), which is expected to be suffi-
cient for our system. All modeling requiring a protein
structure used T. californica (PDB ID: 1EVE) com-
Ò
14
plexed with Aricept (donepezil). While this effort pro-
vided useful information, the force field is not
parameterized to handle the important cation-p interac-
tions between the basic amines and the key tryptophans
(Trp 84 and Trp 279). To build off this original effort
and reconcile the issue with the cation-p interactions
we further performed automated docking with Glide
3
1
3
2–34
31
followed by QM/MM
calculations using QSite.
All Glide dockings were done in extra precision mode.
The docking grid was generated with the default size
using the aforementioned structure. All QM/MM were
performed with the protein fixed using B3LYP/6-
Figure 1. The crystal structure binding mode of decamethonium in the
active site of AChE. The most notable interactions are those of the
quaternary amines of decamethonium with the protein. One quater-
nary amine interacts with Trp 84 (the AS) and the other with Trp 279
*
*
3
1G . The QM region contained the following residues:
Ser 200, Glu 327, His 440 (the catalytic triad); Trp 84
(the AS); Trp 279 (the PAS); Tyr 121, Phe 290, Phe
(the PAS).

S. D. Bembenek et al. / Bioorg. Med. Chem. 16 (2008) 2968–2973
2971
Figure 2. The proposed three-dimensional pharmacophore for histamine H
3
with 1. The main features are two positive ionic centers and an aromatic
ring forming the center of the molecule.
Table 1. Data for hydroxyl linkers of various lengths in the four
position of the piperidine ring
4. Results and discussion
Figure 3 shows the QM/MM binding mode for 1 over-
laid with the crystal structure binding mode of decame-
thonium (the residues shown are of the crystal structure
N
O
2
2
of decamethonium ). There is excellent overlap of the
pharmacophore features with the most notable being be-
tween the basic amines of 1 and the quaternary nitro-
gens of decamethonium. The central aryl ring of 1 is
also coincident with the alkyl chain of decamethonium.
From the results of the QM/MM binding mode and
from the shown overlay we suggest 1 is making cation-
p interactions between the basic amines and the key try-
ptophans (Trp 84 and Trp 279). Also notable in Figure 3
are the residues Tyr 130 (green), Glu 199 (green), and
the catalytic serine Ser 200 (gray). It was speculated
these nearby residues could interact favorably with pen-
dant substituents on the proximal piperidine ring of 1
and thus improve the affinity of 1 for AChE. The QM/
MM binding mode for 6 is shown in Figure 4 (the resi-
dues shown are from the crystal structure of T. califor-
R
Compound
R
AChE IC50
H
3
(
nM)
(nM)
1
7
2
6
3200 ± 400
2300 ± 100
1100
0.53 ± 0.49
N
OH
0.72 ± 0.19
1.30 ± 0.60
0.98 ± 0.21
N
N
OH
OH
1
4
nica (PDB ID: 1EVE) complexed with donepezil). As
with 1, 6 also makes cation-p interactions with the key
tryptophans. Additionally the hydroxyl group is well
aligned to make hydrogen bond interactions with Glu
199 and Ser 200 and possibly Tyr 130. Table 1 shows
the data for hydroxyl linkers of various lengths in the
350 ± 10
N
Decamethonium
Donepezil
5900 ± 1400 820 ± 170
160 ± 30
350 ± 70
3
30, Phe 331, Tyr 334 (the gating residues) and Gly 117,
Glu 199, Tyr 130 (other residues showing the potential
for interaction). For these residues only the side chains
were chosen except for Gly 117 where we chose the
whole residue since the important interaction was with
the carbonyl backbone. All other settings for the QM/
MM were the defaults chosen by QSite. The automated
docking with Glide allowed us to get a more robust sam-
pling of the possible binding modes for 1 in AChE.
Clearly, it is not feasible to perform QM/MM on all
the binding modes returned by Glide. However, one
mode stood out among the rest and was followed up
with QM/MM. This mode scored high in Glide, was
similar to the initial binding mode from the manual
docking with Insight II and, upon overlaying the Glide
binding mode and the binding mode of decamethonium
from the crystal structure, showed excellent overlap of
their pharmacophore features. This was especially true
for the positively charged nitrogens.
Figure 3. The QM/MM binding mode for 1 overlaid with the crystal
structure binding mode of decamethonium in the active site of AChE.
The residues shown are from the crystal structure of decamethonium.
Note the near perfect overlay of the basic amines with the quaternary
nitrogens of decamethonium. Further, the central aryl ring of 1 is
coincident with the alkyl chain of decamethonium.

2972
S. D. Bembenek et al. / Bioorg. Med. Chem. 16 (2008) 2968–2973
References and notes
1
. Schmitt, B.; Bernhardt, T.; Moeller, H. J.; Heuser, I.;
Frolich, I. CNS Drugs 2004, 18, 827.
. Giacobini, E. Neurol. Res. 2000, 25, 1185.
2
3
. Dorronsoro, I.; Castro, A.; Martinez, A. Expert Opin.
Ther. Patents 2003, 13, 1725.
4
5
6
. Knopman, D. Clin. Neuropharmacol. 2003, 26, 314.
. Johannsen, P. CNS Drugs 2004, 18, 757.
. Jackson, S.; Ham, R. J.; Wilkinson, D. Br. J. Clin.
Pharmacol. 2004, 58, 1.
7
8
. Morphy, R.; Rankovic, Z. J. Med. Chem. 2005, 48, 6523.
. Kogen, H.; Toda, N.; Tago, K.; Marumoto, S.; Takami,
K.; Ori, M.; Yamada, N.; Koyama, K.; Naruto, S.; Abe,
K.; Yamazaki, R.; Hara, T.; Aoyagi, A.; Abe, Y.; Kaneko,
T. Org. Lett. 2004, 4, 3359.
9
. Yokatani, K.; Murakami, Y.; Okada, S.; Wang, M.;
Nakamura, K. Eur. J. Pharmacol. 2000, 392, 23.
1
0. Blandina, P.; Giorgetti, M.; Bartolini, L.; Cecchi, M.;
Timmerman, H.; Leurs, R.; Pepeu, G.; Giovannini, M. G.
Br. J. Pharmacol. 1996, 119, 1656.
Figure 4. The QM/MM binding mode of compound 6. The hydroxyl
group is positioned to make hydrogen bond interactions with Ser 200
and Glu 199 and possibly Tyr 130. Further, the central aromatic ring
appears well situated to make a p-stacking interaction with Tyr 334.
The residues shown are from the protein of Torpedo californica
1
1
1
1. Lushington, G. H.; Guo, J.; Hurley, M. M. Curr. Top.
Med. Chem. 2006, 6, 57.
2. Sussman, J. L.; Harel, M.; Frolow, F.; Oefner, C.;
Goldman, A.; Toker, L.; Silman, I. Science 1991, 253, 872.
3. Raves, M. L.; Harel, M.; Pang, Y. P.; Silman, I.;
Kozikowski, A. P.; Sussman, J. L. Nat. Struct. Biol.
(available from the PDB) complexed with donepezil.
1
997, 4, 57.
4. Kryger, G.; Silman, I.; Sussman, J. L. Struct. Fold. Des.
999, 7, 297.
four position of the piperidine ring. Docking calcula-
tions suggest the hydroxyethyl moiety provides the min-
imum linker length needed to have the opportunity to
make interactions with Glu 199, Ser 200, and Tyr 130
and indeed, 6 is the most potent compound of the three
analogs shown in Table 1. Finally, the central aromatic
ring of 6 appears to be well positioned for p-stacking
interactions with Tyr 334. We believe the central aro-
matic ring of 6 makes p-stacking interactions in the ac-
1
1
1
1
5. Koellner, G.; Kryger, G.; Millard, C. B.; Silman, I.;
Sussman, J. L.; Steiner, T. J. Mol. Biol. 2000, 296, 713.
6. Henchman, R. H.; Tai, K.; Shen, T.; McCammon, J. A.
Biophys. J. 2002, 82, 2671.
17. Cubero, E.; Luque, F. J.; Orozco, M. Proc. Natl. Acad.
Sci. 1998, 95, 5976.
1
8. Gallivan, J. P.; Dougherty, D. A. Proc. Natl. Acad. Sci.
999, 96, 9459.
1
tive sites of both AChE and histamine H receptor (in
3
1
9. Rosenberry, T. L.; Johnson, J. L.; Cusack, B.; Szegletes,
T.; Mallender, W. D.. In Recent Trends in the Acetylcho-
linesterase System; Parveen, M., Kumar, S., Eds.; Bio-
medical and Health Research; IOS Press: Amsterdam, The
Netherlands, 2005; Vol. 63, pp 53–62.
histamine H receptor the residues Tyr 115 and Trp
3
3
71 are most likely providing this interaction).
5. Conclusions
2
0. De Ferrari, G. V.; Canales, M. A.; Shin, I.; Weiner, L. M.;
Silman, I.; Inestrosa, N. C. Biochemistry 2001, 40, 10447.
Currently, AChE inhibitors represent the most common
therapeutic option for AD. Their procognitive and mem-
21. Axe, F. U.; Bembenek, S. D.; S a´ ndor, S. J. Mol. Graph.
Model. 2006, 24, 456.
3
5
2
2. Harel, M.; Schalk, I.; Ehret-Sabatier, L.; Bouet, F.;
Goeldner, M.; Hirth, C.; Axelsen, P. H.; Silman, I.;
Sussman, J. L. Proc. Natl. Acad. Sci. 1993, 90, 9031.
ory-enhancing effects are well known in the literature. In
addition, there is a considerable body of experimental
work suggesting antagonism of the histamine H receptor
3
36–38
23. Greenblatt, H. M.; Kryger, G.; Lewis, T.; Silman, I.;
Sussman, J. L. FEBS 1999, 463, 321.
translates into increased cognitive performance.
2
4. Bar-on, P.; Millard, C. B.; Harel, M.; Dvir, H.; Enz, A.;
Sussman, J. L.; Silman, I. Biochemistry 2002, 41, 3555.
5. Accelrys Software Inc., San Diego, CA.
Herein we demonstrated how the use of available crystal
structure information, pharmacophore modeling, dock-
ing (automated, manual, classical, and QM/MM) lead
us to a compound (6) (and potential series) with excellent
2
26. Clement, O. O.; Mehl, A. T. In Pharmacophore Perception,
Development and Use in Drug Design; G u¨ ner, O. F., Ed.;
International UniversityLine: LaJolla, CA, 2000; pp71–84.
activity at both AChE and the histamine H receptor. In
3
2
7. Apodaca, R.; Dvorak, C. A.; Xiao, W.; Barbier, A. J.;
Boggs, J. D.; Wilson, S. J.; Lovenberg, T. W.; Carruthers,
N. I. J. Med. Chem. 2003, 46, 3938.
8. Zhou, H. X.; Wlodek, S. T.; McCammon, J. A. Proc. Natl.
Acad. Sci. 1998, 95, 9280.
9. Toda, N.; Yoriko, I.; Tago, K.; Kogen, H.; Kaneko, T.;
Miyamoto, S. Bioinform. J. 2003, 3, 46.
0. Dauber-Osguthorpe, P.; Roberts, V. A.; Osguthorpe, D.
J.; Wolff, J.; Genest, M.; Hagler, A. T. Proteins Struct.
Funct. Genet. 1988, 4, 31.
addition, our models suggest such a dual-acting com-
pound may interact with the PAS and, based on current
thinking, could potentially slow the progression of AD.
2
2
3
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmc.2007.
1
2.048.
31. Schrodinger, LLC, San Diego, CA.

S. D. Bembenek et al. / Bioorg. Med. Chem. 16 (2008) 2968–2973
2973
3
2. Hurley, M. M.; Wright, J. B.; Lushington, G. H.; White,
W. E. Theor. Chem. Acc. 2003, 109, 160.
3. Cramer, C. J. In Essentials of Computational Chemistry,
Theories and Models; Wiley: West Sussex, England, 2002;
pp 411–438.
35. Simard, M.; Van Reekum, R. Drugs Aging 1999, 14, 197.
36. Passani, M. B.; Lin, J.-S.; Hancock, A.; Crochet, S.;
Blandina, P. Trends Pharmacol. Sci. 2004, 25, 618.
37. Witkin, J. M.; Nelson, D. L. Pharmacol. Ther. 2004, 103,
1.
3
3
4. Zhang, Y.; Kua, J.; McCammon, J. A. JACS 2002, 124,
1
38. Hancock, A. A.; Fox, G. B. Expert Opin. Investig. Drugs
2004, 13, 1237.
0572.