Docking studies of benzylidene anabaseine interactions with α7 nicotinic acetylcholine receptor (nAChR) and acetylcholine binding proteins (AChBPs): Application to the design of related α7 selective ligands

Article abstract of DOI:10.1016/j.ejmech.2011.09.033

AChBPs isolated from Lymnaea stagnalis (Ls), Aplysia californica (Ac) and Bulinus truncatus (Bt) have been extensively used as structural prototypes to understand the molecular mechanisms that underlie ligand-interactions with nAChRs [1]. Here, we describe docking studies on interactions of benzylidene anabaseine analogs with AChBPs and α7 nAChR. Results reveal that docking of these compounds using Glide software accurately reproduces experimentally-observed binding modes of DMXBA and of its active metabolite, in the binding pocket of Ac. In addition to the well-known nicotinic pharmacophore (positive charge, hydrogen-bond acceptor, and hydrophobic aromatic groups), a hydrogen-bond donor feature contributes to binding of these compounds to Ac, Bt, and the α7 nAChR. This is consistent with benzylidene anabaseine analogs with OH and NH₂ functional groups showing the highest binding affinity of these congeners, and the position of the ligand shown in previous X-ray crystallographic studies of ligand-Ac complexes. In the predicted ligand-Ls complex, by contrast, the ligand OH group acts as hydrogen-bond acceptor. We have applied our structural findings to optimizing the design of novel spirodiazepine and spiroimidazoline quinuclidine series. Binding and functional studies revealed that these hydrogen-bond donor containing compounds exhibit improved affinity and selectivity for the α7 nAChR subtype and demonstrate partial agonism. The gain in affinity is also due to conformational restriction, tighter hydrophobic enclosures, and stronger cation-π interactions. The use of AChBPs structure as a surrogate to predict binding affinity to α7 nAChR has also been investigated. On the whole, we found that molecular docking into Ls binding site generally scores better than when a α7 homology model, Bt or Ac crystal structure is used.

Full text of DOI:10.1016/j.ejmech.2011.09.033

European Journal of Medicinal Chemistry 46 (2011) 5625e5635
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Docking studies of benzylidene anabaseine interactions with
a7 nicotinic
acetylcholine receptor (nAChR) and acetylcholine binding proteins
(AChBPs): Application to the design of related
a7 selective ligands
,
a
a
a
a
David C. Komboa , Anatoly Mazurov , Kartik Tallapragada , Philip S. Hammond , Joseph Chewning ,
Terry A. Hausera, Montserrat Vasquez-Valdiviesob, Daniel Yohannesa, Todd T. Talleyc, Palmer Taylorc,
William S. Caldwella
*
a Targacept Inc., Molecular Design, 200 East First Street, Suite 300, Winston-Salem, NC 27101-4165, USA
b University of Edinburgh, Structural Biochemistry Group, Mayfield Rd, Edinburgh EH9 3JR, Scotland, UK
c Department of Pharmacology, Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, CA 92093-0657, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
AChBPs isolated from Lymnaea stagnalis (Ls), Aplysia californica (Ac) and Bulinus truncatus (Bt) have been
extensively used as structural prototypes to understand the molecular mechanisms that underlie
ligand-interactions with nAChRs [1]. Here, we describe docking studies on interactions of benzylidene
Received 10 August 2011
Received in revised form
17 September 2011
Accepted 20 September 2011
Available online 29 September 2011
anabaseine analogs with AChBPs and a7 nAChR. Results reveal that docking of these compounds using
Glide software accurately reproduces experimentally-observed binding modes of DMXBA and of its
active metabolite, in the binding pocket of Ac. In addition to the well-known nicotinic pharmacophore
(positive charge, hydrogen-bond acceptor, and hydrophobic aromatic groups), a hydrogen-bond donor
Keywords:
AChBPs
feature contributes to binding of these compounds to Ac, Bt, and the a7 nAChR. This is consistent with
benzylidene anabaseine analogs with OH and NH2 functional groups showing the highest binding affinity
of these congeners, and the position of the ligand shown in previous X-ray crystallographic studies of
ligand-Ac complexes. In the predicted ligand-Ls complex, by contrast, the ligand OH group acts as
hydrogen-bond acceptor. We have applied our structural findings to optimizing the design of novel
spirodiazepine and spiroimidazoline quinuclidine series. Binding and functional studies revealed that
Nicotinic acetylcholine receptor
Docking
Benzylidene-anabaseine analogs
Spirodiazepine and spiroimidazoline
quinuclidine series
a7
these hydrogen-bond donor containing compounds exhibit improved affinity and selectivity for the
nAChR subtype and demonstrate partial agonism. The gain in affinity is also due to conformational
restriction, tighter hydrophobic enclosures, and stronger cation- interactions. The use of AChBPs
structure as a surrogate to predict binding affinity to 7 nAChR has also been investigated. On the whole,
a7
p
a
we found that molecular docking into Ls binding site generally scores better than when a
model, Bt or Ac crystal structure is used.
a7 homology
Ó 2011 Elsevier Masson SAS. All rights reserved.
1. Introduction
nAChRs are members of the Cys-Loop ligand-gated ion channel
superfamily, located both in the peripheral and central nervous
systems. These receptors, existing as both homopentameric and
heteropentameric transmembrane ion channels, are validated
therapeutic targets for various CNS pathologies (for reviews of
nAChRs as targets for drug discovery, see Romanelli et al. [1],
Breining [2], Schmitt [3], Conejero-Goldberg et al. [4], Mazurov
et al. [5], Daly [6], and Taly et al. [7]). Examples of disease indica-
tions under active investigation include Alzheimer’s disease (AD)
and Parkinson’s disease (PD), cognitive dysfunction in Schizo-
phrenia (CDS), addiction disorders, attention deficit hyperactivity
Abbreviations: Ac, Aplysia californica; ACHBP, Acetylcholine-binding protein; AD,
Alzheimer’s disease; ADHD, Attention deficit hyperactivity disorder; BA, Benzyli-
dene anabaseine; Bt, Bulinus truncatus; CDS, Cognitive dysfunction in schizo-
phrenia; CNS, Central nervous system; DMXBA, 3-(2,4-dimethoxybenzilidene)-
anabaseine; HBD, Hydrogen-bond donor; Ls, Lymnaea stagnalis; nAChR, Nicotinic
acetylcholine receptor; PD, Parkinson’s disease; Pdb, Protein databank; QPLD, QM-
polarized ligand docking; rmsd, Root-mean-squared-deviation; ROC, Receiver
operating characteristic curve; SP, Standard precision; XP, Extra precision.
* Corresponding author. Tel.: þ1 336 480 2127; fax: þ1 336 480 2107.
E-mail address: david.kombo@targacept.com (D.C. Kombo).
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.09.033

5626
D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
disorder (ADHD), age-associated memory impairment (AAMI), pain
management, anxiety, depression and inflammation-mediated
processes. A number of compounds targeting nAChRs and repre-
senting a wide variety of pharmacologic actions are in advanced
clinical trials or on the market.
a proprietary compound collection [21]. Three-dimensional struc-
tures of AChBP have also been used as templates to rationalize
agonists binding to the homologous homopentameric
a7 nAChR
[17]. Thus, the homologous AChBPs and associated co-crystal
structures with nAChR ligands provide a rich starting point for
understanding interactions for the various nAChR subtypes.
Benzylidene derivatives of the alkaloid anabaseine, are known
ChantixÒ, a partial agonist at
a4b2 and a full agonist at a7 nAChR
from Pfizer, has recently been launched for smoking cessation [8,9].
ABT-089, a selective nAChR agonist from Abbot, completed Phase II
clinical trials for ADHD and AD but was not advanced further. ABT-
894, an nAChR agonist discovered by Abbot in collaboration with
NeuroSearch, completed three Phase II clinical trials in 2008 for
for their functional selectivity toward
a7 nicotinic receptors
[22,23]. A set of these compounds has been studied extensively
with regards to their dissociation constants (Table 1) for the AChBPs
from Ls, Ac, and Bt [24]. It was found that the 4-hydroxy benzyli-
dene anabaseine compounds are among the most potent ligands in
the series. Spectroscopic studies were able to ascertain the ioni-
zation state of 4-hydroxy substituted and un-substituted benzyli-
dene anabaseines. Binding studies of benzylidene anabaseine
ADHD and diabetic neuropathic pain. AZD-3480, an a4b2 selective
partial agonist, showed positive results in Phase 2 trials for AAMI
and ADHD. One of the anabaseine analogs under study in this work,
GTS-21 (DMXBA or compound 2 in Table 1) has been investigated in
clinical trials for ADHD (completed), AD (completed), and inflam-
analogs with rat
a7 and a4b2 have also been reported more
mation (ongoing) [10]. Finally, TC-5619, an
a
7 selective modulator,
recently [25]. Recent crystallographic studies of the 4-hydroxy
metabolite of 3-(2,4-dimethoxybenzylidene)-anabaseine in
complexation with Aplysia AChBP have shown that its OH group
donates a hydrogen-bond to a polar side-chain triad of Asp-164,
Ser-166 and Ser- 167 in loop F [17]. These findings provided us
has shown positive top-line results from a Phase 2 clinical proof of
concept trial to assess it as an augmentation therapy to improve
cognition in patients with schizophrenia [11]. Understanding the
molecular basis for the binding and selectivity of nAChR ligands
that interact with this protein superfamily would be an important
step toward designing better drugs against these targets.
AChBP is a homopentamer similar in structure to the extracel-
lular ligand-binding domain of nAChRs. The availability of several
crystal structures of AChBPs, both free and/or in complex with
various nicotinic ligands, has provided much needed information
regarding the protein-ligand molecular recognition process
[12e17]. These and additional studies have provided data indi-
O
O
HO
O
O
HN
N
NH
N
NH
NH
cating that ligand-nAChR interactions are characterized bycation-
p
N
N
interactions, hydrogen-bonding between the typical cationic center
of secondary and tertiary amine containing nAChR ligands and the
protein backbone, receptor loop C flexibility, water-mediated and
hydrophobic interactions, and capacity to accommodate ligands of
differing structure at the binding site at the interface of two
subunits [18e20]. Successful results have recently been reported
using an AChBP crystal structure as a template for molecular
N
3e
2i
3a
e
HO
b
Ar
NH2
HN
NH
2
HN
a
d
N
N
N
N
docking, to identify novel a7 nAChR ligands through screening of
N
g
f
1
2a-g
c
2h
Table 1
2D structure of benzylidene anabaseine analogs and their dissociation constants
with AChBPs isolated from Ls, Ac and Bt [24].
OH
Structures
Name
Ls Kd (nM)
Ac Kd (nM)
Bt Kd (nM)
N
N
NH
NH
H N
2
O
N
N
N
3c
3d
NH
1
0.8
3
33
N
3b
Scheme 1. Synthesis of
a7 nAChR ligands. Reagents and conditions: (a) ArCOCH2
-
2
3
19
330
3
59
22
CO2Me, i-BuOH, 100 ꢀC, overnight; (b) methyl 2-(cyanomethoxy)benzoate, CS2 (one
drop), 100e110 ꢀC, overnight; (c) (2-cyanophenoxy)acetonitrile, CS2 (one drop),
100e110 ꢀC, overnight; (d) [3-(dimethylamino)-2-phenylprop-2-enylidene]-dimethy-
lammonium hexafluorophosphate, MeOH, reflux, overnight; (e) diethyl 2-
phenylmalonate, 150 ꢀC, 5 min; (f) methyl benzimidate hydrochloride, methanol,
150 ꢀC (microwave), 5 min; (g) 4-hydroxycoumarin, t-butanol, 100 ꢀC, overnight.
Compounds containing hydroxyl group (2aeg, 3a) or amino group (3b) as a hydrogen
bond donor have been synthesized by coupling 3-amino-3-(aminomethyl)quinuclidine
(1) with series of aroylacetates [27], methyl 2-(cyanomethoxy)benzoate or (2-
cyanophenoxy)acetonitrile. Spirodiazepine 2h without hydroxyl moiety has been ob-
0.4
tained by condensation of diamine
enylidene]-dimethylammonium hexafluorophosphate. Heating in
diamine with diethyl 2-phenylmalonate or methyl benzimidate provided 1,4-
1
with [3-(dimethylamino)-2-phenylprop-2-
a
microwave of
4
0.8
6
93
1
diazepindione 2i and phenylimidazoline 3c. Hydroxyphenylimidazoline 3d has been
carried out by cyclization of 3-amino-3-(aminomethyl)quinuclidine (1) with 2-(2-
hydroxyphenyl)-1,3-benzoxazin-4-one.

D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
5627
Table 2
nAChR competitive binding: Percent inhibition (ꢁstandard error) of control radioligand binding to nAChRs of the designed spirodiazepine and spiroimidazoline quinuclidines
tested at 5 M in radioactive displacement assays. Values in parentheses represent Ki values in nM. The 7 binding studies
7 radioligand [3H]-methyllycaconitine was used for
on rat hippocampal membranes and the nicotinic radioligand [3H]-epibatidine was used for binding studies at
2 on rat cortical membranes, ganglion-type nicotinic
m
a
a
a4b
receptors on SH-SY5Y cellular membranes and muscle-type nicotinic receptors on TE-671 cellular membranes. See Experimental Section for details. NT ¼ not tested. For
comparison purpose, corresponding data for compounds 2 and 3 of Table 1 are also shown.
Compound
Structure
a
7
a
4
b
2
a
3
b4
a1bgd
O
HN
2a
95 ꢁ 3 (140 ꢁ 7)
15 ꢁ 1
NT
NT
NT
NT
N
N
O
HN
2b
2c
2d
87 ꢁ 3 (110)
93 ꢁ 7 (87)
98 ꢁ 3 (42)
7 ꢁ 1
NT
N
O
N
O
HN
12 ꢁ 1
42 ꢁ 14
NT
N
O
N
O
HN
O
63 ꢁ 0.4
89 ꢁ 0.45
NT
38 ꢁ 5
N
N
O
HN
2e
2f
95 ꢁ 3 (77)
98 ꢁ 2 (58)
84 ꢁ 4 (150)
62 ꢁ 15
20 ꢁ 1
14 ꢁ 3
28 ꢁ 6
NT
S
Br
N
N
N
N
O
HN
S
N
O
HN
2g
NT
NT
N
O
H
N
2h
2i
27 ꢁ 5
6 ꢁ 2
NT
NT
N
N
O
H
N
19 ꢁ 1
2 ꢁ 1
NT
NT
N
O
O
N
H
N
3a
3b
93 ꢁ 6 (7.2 ꢁ 3)
96 ꢁ 1 (120)
32 ꢁ 11
9 ꢁ 1
55 ꢁ 3
47 ꢁ 3
15 ꢁ 10
38 ꢁ 2
N
N
H
OH
H
O
N
N
H
N
N
3c
1 ꢁ 9
11 ꢁ 2
NT
NT
NT
N
NH
N
3d
3e
77 ꢁ 3 (290)
33 ꢁ 3
4 ꢁ 1
4 ꢁ 2
NT
N
N
OH
N
N
NH
13 ꢁ 1
26 ꢁ 5
(continued on next page)
O
N
H

5628
D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
Table 2 (continued )
Compound
Structure
a
7
a
4
b2
a
3
b4
a1bgd
N
O
2
3
(1300 ꢁ 550)
106 ꢁ 3 (420)
(51 ꢁ 10)
(2500 ꢁ 270)
(3900 ꢁ 770)
72 ꢁ 3
N
O
N
O H
97 ꢁ 1 (19 ꢁ 6)
75 ꢁ 1 (970 ꢁ 110)
N
O
with a starting point for modeling studies aimed at deciphering the
structural features that drive the binding affinity of these nAChR
ligands. Using molecular docking, and homology modeling studies,
we found that a similar hydrogen-bond donor feature contributes
restriction, hydrophobic enclosures, and cations-
p
interactions also
contribute to the binding free energy of benzylidene anabaseine, in
addition to the more common van der walls, coulombic, lipophilic
and hydrogen-bond interactions (data not shown). Compounds 1
and 3 exhibit the strongest binding affinities (Table 1) to the Ac
protein. While the crystal structure of compound 3 in complex with
Ac has been solved, no such data are available for compound 1.
Docking of 1 into the Ac binding site (Fig. 1), suggests that
compound 1 exhibits strikingly similar interactions to those
crystallographically-observed for the Ac-compound 3 complex,
including the bifurcated HB donation to the whole amino acid triad
Asp-164, Ser-166 and Ser-167 (see also Fig. S3 herein and Fig. 4 of
Hibbs et al. [17]).
As shown in Fig. 2, docking compound 3 in the Bt protein (pdb
code: 2BJ0, resolution 2.0 Å) also results in a similar hydrogen-
bonding pattern: the protonated nitrogen atom makes
a hydrogen bond with the backbone carbonyl oxygen atom of Trp-
142, the nitrogen atom of the pyridine ring makes a hydrogen-bond
with the backbone NH group of Val-113, while the ligand OH group
donates a hydrogen-bond to the hydroxyl group of Tyr-164,
a residue located in the F-loop region of the protein. Given the
presence of Glu-163 within hydrogen-bond forming distance of this
region, it is possible that an alternate conformation might include
to the interaction with Bulinus AChBP and the a7 nAChR as well. In
the case of Lymnaea, however, the F-loop region appears not to
participate to such hydrogen-bond interaction. Instead, the ligand
OH group accepts a hydrogen bond from an amino acid side-chain
located in a different region of the binding site. Targeting the a7
subtype of neuronal nicotinic acetylcholine receptor, we have
designed related hydrogen-bond donor containing spirodiazepines
and spiroimidazolines quinuclidine series (Scheme 1). We find that
these compounds selectively bind to the
low nanomolar range.
a7 receptor with Ki in the
2. Results
2.1. Molecular docking of benzylidene anabaseine analogs
Validation of the docking procedure for AChBP was carried out
by self-docking compounds 2 and compounds 3 co-crystallized
with Ac [17], as described in supporting information (S1, Table S1,
Figs. S1, S2 and S3). Table S1 illustrate that conformational
Fig. 1. Compound 1 docked into Ac: note the NH2 group hydrogen bond with Ser-166, Ser-167 & Asp-164 in the complementary face, in a manner reminiscent of compound 3 pose
in Ac crystal structure.

D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
5629
Fig. 2. Binding mode of compound 3 in the binding site of Bt protein. The OH group of the ligand donates a hydrogen-bond to Tyr-164 of the complementary face. The pyridine N
atom makes an H-bond with the NH group of Val-113 in the complementary face, while the cationic center hydrogen bonds to Trp-142 backbone carbonyl group. The F-loop is
shown in dark blue. Also shown is the distance (12.6 Å) between the backbone oxygen atom of Trp-142 and side-chain oxygen atom of Tyr-164.
the ligand hydroxyl group being involved in a bifurcated hydrogen
bond donation to Tyr-164, and Glu-163.
group accepts a hydrogen-bond from the side-chain of Tyr-192 of
the principal face, as shown in Fig. 3. Compared to its orientation in
Ac and Bt binding sites, it appears as though the ligand has kept its
quintessential hydrogen bond between the protonated ligand and
the carbonyl oxygen atom of the conserved Trp-143, but has made
In the case of Ls, however, the ligand hydroxyl group at position
4
accepts
a hydrogen-bond from the side-chain carbamoyl
hydrogen of Gln-73 of the complementary face, while the methoxy
Fig. 3. Compound 3 docked into Ls: Ligand OH group accepts an HB from Gln-73 of the complementary face, while its methoxy group accepts an HB with Tyr-192 of the principal
face. The cationic center donates to the backbone carbonyl group of Trp-143, in the principal face. The ligand makes a 90ꢀ rotation compared to its orientation in Ac and Bt binding
sites. The F-loop is shown in dark blue. The amino acid side-chains which directly interact with the ligand and those of the F-loop that could have interacted with the ligand like in
Ac and Bt are shown in ball and stick.

5630
D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
Fig. 4. Compound 3 (in green) docked into rat
a
7 homology model. The ligand OH group donates a Hydrogen-bond to Ser-36 (in the complementary face), which is located in the
neighborhood of Tyr-168, and Ser-167 in the F-loop (shown in dark blue). The cationic center donates a hydrogen bond to the backbone CO group of Trp-148, in the principal face.
Also shown are the distance between the backbone oxygen atom of Trp-148 and side-chain oxygen atom of Ser-36, and Gln-57 (in the complementary face).
a 90ꢀ rotation, thereby allowing its OH and methoxy group to
interact with Gln-73 and Tyr-192, respectively. Likewise, OH group
at position 4 of compound 4, which is among the most potent in
binding to Ls, also accepts a hydrogen-bond from Gln-73 of the
complementary face, while its OH group at position 2 donates
a hydrogen-bond to the OH group of Tyr-192 of the principal face,
as shown in Fig. S4.
lead to the identification of novel
recently, using dataset mainly comprised of quinuclidine-
containing compounds, de Kloe et al. have shown that a strong
correlation exists between binding affinities for 7 and AChBP-Ls
and to a significantly lesser extent, to AChBP-Ac [26]. We have
herein shown that a hydrogen bond donor feature observed to be
important for binding affinity of anabaseine analogs to AChBP by
interacting with this protein F-loop can also act in a similar manner
a7 nAChR ligands [21]. More
a
a
The three-dimensional structure of the
homology modeling, as described in the methods section.
Biochemical validation of the 7 homology model was carried out
by docking a diverse chemical library of 493 compounds of known
binding affinity to the 7 receptor. The procedure used and results
a7 nAchR was derived by
in the binding pocket of the a7 nAChR. Therefore, we reasoned that
a
quinuclidine analogs containing a hydrogen bond donor functional
group separated from the cationic center by a distance within the
distance constraint range shown in Fig. 4 (13.6e20.9 Å) could
a
obtained are described in Supplemental information S2. In
summary, a Roc score of 0.74 was achieved, which is much better
than 0.53, the corresponding value obtained when one uses the
experimentally-observed crystal structure of Ac as a surrogate, as
shown in Table 3. The hydrogen-bond donation pattern involving
the OH group of compound 3 was observed in the docking pose of
satisfy the pharmacophoric requirements for binding to the a7
protein. We accordingly designed a library of spirodiazepine and
spiroimidazoline quinuclidines and predicted their binding affinity
to rat
a7, using molecular docking as described herein. The
synthetic scheme of the 14 compounds made is shown in Scheme 1.
Experimentally-observed binding data of the designed
compounds are shown in Table 2. Results indicate that while all the
this ligand into the binding site of the homomeric rat a7 nAChR,
with an orientation analogous to the one observed in Ac and Bt, as
shown in Fig. 4. Ser-36 of the complementary face, a residue close
in space to Ala-163, Asp-164, and Ile-165, accepts the hydrogen-
bond donated by the ligand OH group.
designed compounds do not interact with the
compounds 3a, 3b and 3d, which contain hydrogen-bond donors
interact with 7 nAChR with Ki less than 300 nM, and so do
compounds 2a through 2g, which might tautomerize into lactim.
Substitution within the aromatic ring of compounds 2aeg slightly
affects interaction with the receptor, while potential conjugation of
a4b2 subtype,
a
2.2. Rational design of novel a7 nicotinic receptor ligands
the aromatic ring and azomethine moiety provide
Compound 3a ( 7 nAChR Ki 7.2 nM) exhibited better binding
affinity for 7 than compounds 2 ( 7 nAChR Ki 1500 nM) and
compound 3 ( 7 nAChR Ki 420 nM), and demonstrated partial
agonism with an Emax of 53.0 ꢁ 4.4% and an EC50 of 0.6 ꢁ 0.5 M, by
patch-clamp electrophysiology in rat 7 nAChR [28]. Under the
same experimental conditions, compound 3b ( 7 nAChR Ki 120 nM)
had an Emax of 22.0 ꢁ 4.7% and an EC50 of 1.2 ꢁ 0.7 M. Functional
data for compound 3 (
7 nAChR Emax of 77.00 ꢁ 0.02% and EC50 of
M) have been already reported in the literature [29]. For
pep interaction.
a
Ulens et al. have demonstrated that docking-based virtual
screening using AChBP crystal structure as target can successfully
a
a
a
m
Table 3
a
Glide XP terms for designed compounds.
a
m
Compound Rat
a
7
Glide
Rotable bond Hydroph enclosure p-cation
Ki (nM) score
penalty
(Kcal/mol)
rewards (Kcal/mol) rewards
(Kcal/mol)
a
1.6 ꢁ 0.02
m
2f
2e
2c
3b
2a
2g
3d
3
58
77
87
120
140
150
290
420
ꢂ11.87 0.09
ꢂ11.79 0.06
ꢂ2.55
ꢂ2.28
ꢂ2.38
ꢂ2.40
ꢂ2.05
ꢂ2.48
ꢂ2.40
ꢂ1.83
ꢂ3.07
ꢂ3.16
ꢂ2.91
ꢂ3.00
ꢂ3.01
ꢂ3.05
ꢂ3.45
ꢂ2.03
comparison, compounds 2h, 2i, 3c, 3e, lacking a hydrogen-bond
donor satisfying the required distance constraint, do not demon-
ꢂ11.31 0.08
ꢂ10.57 0.09
ꢂ11.79 0.10
ꢂ12.14 0.10
ꢂ11.25 0.11
ꢂ9.40 0.18
strate any interaction with
a7 nAChR. Fig. 5aeb illustrate the pre-
dicted binding mode of compounds 2b and 3b into the rat
a7
protein binding site, as derived from docking. We find that the OH
group of the lactim form of compound 2b donates a hydrogen-bond
to Gln-57, whereas its methoxy oxygen atom accepts a hydrogen-

D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
5631
Fig. 5. a. Designed compound 2b (in its lactim form) docked into rat a7 homology model, is shown in green. The ligand OH group donates a hydrogen-bond to the backbone carbonyl
oxygen atomofLeu-119 (inthecomplementaryface), while itsmethoxygroupacceptsa hydrogenbondfromGln-57(inthecomplementaryface). Alsoshown areneighboring amino-acid
sidechainssuchasSer-167, Ser-16, andAsp-164. ThecationiccenterdonatestothebackboneCOgroupofTrp-148, in theprincipalface. Allthethree hydrogen-bonddistances are shownin
brokenblack lines. b. Designedcompound 3b dockedinto rat a7 homology model. TheligandOH group donates a hydrogen-bondtoTyr-168 (in thecomplementaryface), whichis located
in the F-loop (shown in dark blue). Also shown is neighboring amino-acid side chain, Ser-167. The cationic center donates to the backbone CO group of Trp-148, in the principal face. Both
hydrogen-bond distances are shown in broken black lines.
bond from the backbone NH group of Leu-119, as shown in Fig. 5a.
The NH2 group of compound 3b donates a hydrogen-bond to the
OH group of Tyr-168, a residue located in the F-loop region, as
shown in Fig. 5b.
After docking in SP mode, the resulting best poses were
subsequently docked in XP mode. The results thus obtained, shown
in Table 3, indicate that the designed compounds are predicted to
bind tighter than compound 3, as Glide scores of the former are
more negative than the Glide score of the latter. The results also
illustrate that the best scoring conformations of the designed
compounds which were successfully docked exhibit tighter
compared to compound 3. This results in an increase in binding
affinity. Furthermore, the XP-derived rotatable bond penalty
appears to linearly correlate with rat a
7 Ki, with an r2 value of 0.85
(Fig. 6). This finding indicates that the reduced flexibility of
designed compounds, as compared to compound 3, also contrib-
utes to improve binding affinity [32].
The use of AChBP crystal structures as surrogates for predicting
binding of these novel spirodiazepines and spiroimidazolines to a7
protein has also been investigated, by comparing the trend in
docking score to the actually observed Ki data. With respect to the
ability of the molecular docking to top-rank active
a7 compounds,
hydrophobic enclosures [30,31] and cation-
p interactions, as
the ROC scores obtained were low in all cases, varying from 0.51 to

5632
D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
Table 5
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
RMSD table (expressed in Å) obtained from superimposing the crystal structures of
Ac, Bt, and Ls proteins. The pdb codes used are 2byq, 2bj0, and 2byq, respectively.
3
Protein system
Ac
0
Bt
Ls
Ac
Bt
Ls
2.2
0
1.9
1.6
0
3d
observed experimentally by Taylor and coworkers [17], our docking
studies highlight the importance of the D-SS (Asp-164, Ser-166, Ser-
167) amino acid residues triad from the complementary face of Ac,
which interacts via a hydrogen bonding network with the benzy-
lidene 4-position substituent. This was also found in docking poses
for the Bt protein, but in this case stabilization appears to occur via
a hydrogen bonding network with Glu-163 and Tyr-164, located in
the same complementary face region. While the same amino acid
residue motif is found in the Ls protein, i.e., Glu-163 and Tyr-164,
we have not observed a similar docking pose for compounds 3
and 4, in this protein. Instead, we have found a low-energy pose in
which the ligand hydroxyl group at the 4-position in this case
accepts a hydrogen-bond from Gln-73 of the complementary face,
while the oxygen atom of the methoxy group at the 2-position
accepts a hydrogen-bond from Tyr-192 of the principal face, as
shown in Fig. 3. Such an alternative binding mode results from
a rotation of approximately 90ꢀ about an axis passing through the
ligand basic nitrogen atom and perpendicular to the indole side-
chain of the conserved Trp-143. This pose, while reminiscent of
“conformation B” observed in the co-crystal structure of compound
2 complexed with Ac (Fig. S2), represents a greater displacement of
the benzylidine region of the ligand.
Docking observations, at least in part help to explain the
significant loss of affinities for DMXBA (compound 2) vs. that for
the 4-OH metabolite (compound 3). For Bt, loss of the hydrogen
bond donor capacity (OH to OCH3) leads to slight (2.5 fold) decrease
in Kd value. But for Ac, with a more hydrophilic amino acid residue
environment in the same complementary face region, the loss in
affinity is much more dramatic (110-fold). In the case of Ls, in order
to probe whether OH group at position 4 acts as a better hydrogen-
bond acceptor than OCH3, we have carried out a quantum
mechanical calculation of atom polarization in the binding site,
using Maestro QPLD protocol [33]. Results indicated that the
oxygen atom in the OH group at position 4 of benzylidene anaba-
seine was more negatively charged (q ¼ ꢂ0.55) than in the methoxy
group (q ¼ ꢂ0.37), suggesting that compound 3 might be a better
hydrogen bond acceptor than compound 2.
Docking results obtained for anabaseine analogs are consistent
with the historical nicotinic cholinergic pharmacophore models [1].
The pyridine nitrogen atom acts as hydrogen bond acceptor, while
the pyridine ring, itself, fits the required hydrophobic aromatic
feature. Anabaseine congeners containing OH and NH2 functional
groups additionally donate a hydrogen-bond, (or accept one as
shown by docking into Ls binding site), which may explain the fact
that such analogs bind the strongest in this series of compounds
[24]. The ligand-protonated nitrogen and the OH group at position-
4 and at position-2, are separated by a distance of 8.94 Å and 5.05 Å,
respectively. It is important to mention that these interfeature
distances are consistent with the distances of about 12 Å and 5.5 Å
between the carbonyl oxygen atom of the quintessential conserved
Trp and the side-chain oxygen atom of Asp-164 (or Tyr-164) in the
F-loop and of Tyr-192, respectively, in the docked poses for Bt
(Fig. 2), Ac, and Ls. This result suggests a lock and key model of
2g
2a
R2=0.85
0.18
3b
2c
2e
2f
0.06
0.08
0.10
0.12
0.14
0.16
XP-RotBond-Penalty
Fig. 6. Rat a7 Ki highly correlates with XP-derived rotatable bond penalty, indicating
that the more rigid the compound, the better the binding affinity.
0.62, as shown in Table 4. Of particular interest, however, is that
using a rat
(accuracy of 0.55) than the use of the original AChBP crystal
structure templates. On average, when the validation set is also
included, docking into Ls binding site scores the highest, followed
a7 homology model in docking did not score better
by docking into the a7 homology model.
The RMSD table derived from superimposing the crystal struc-
ture of Ac, Bt, and Ls is shown in Table 5. Examination of this table
indicate that Ls is structurally more similar to Bt (rmsd ¼ 1.6 Å) than
to Ac (rmsd ¼ 2.2 Å). This result is consistent with the finding that
the weighted average value of roc score obtained in Table 4 for Ls
protein (0.76) is closer to the one obtained with Bt protein (0.71) as
compared to the one derived from docking into Ac protein (0.53).
3. Discussion and conclusions
We have carried out docking studies for a homologous series of
benzylidene anabaseine analogs, for which crystallographic infor-
mation on two of the analogs (2 and 3, Table 1) are available. These
efforts are aimed at providing a more comprehensive under-
standing of the molecular basis of their interactions with nAChRs,
which may aid in designing better drug candidates. Docking of
benzylidene anabaseine analogs using Glide software accurately
reproduces both experimentally-observed binding modes of
DMXBA and of its metabolite, in the binding pocket of Ac. Similar
docking poses of anabaseine analogs bound to Bt, Ls and a7 nAChR,
respectively, have been derived. These results support the impor-
tance of the hydrogen bond donor feature in the ligand, as
described previously by Talley et al. [24] and Hibbs et al. [17]. As
Table 4
ROC accuracy of docking in ranking the binding of ligands to rat
crystal structures of AChBPs and
described in the methods section.
a
7 protein. X-ray
a7 homology model were used for docking, as
Protein system
Validation
Designed
set (N ¼ 14)
0.62
0.51
0.62
0.55
Weighted
set (N ¼ 493)
average Roc score
Ac
Bt
Ls
0.53
0.72
0.77
0.74
0.53
0.71
0.76
0.73
molecular recognition. Exception is found, however, with rat
model binding site where the side-chain oxygen atom of Asp-164 is
separated from the carbonyl oxygen atom of the conserved Trp by
a7
a
7

D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
5633
a distance of about 19.4 Å. Not surprisingly, Ser-36, which is sepa-
rated from the conserved Trp by a 12.9 Å distance, interacts with
the ligand OH group at 4-position, in the docked poses (Fig. 4).
Docking results obtained for the designed compounds, shown in
Fig. 5a, suggest that the lactim form of spirodiazepine analogs
whole, the results obtained show that models derived from Ls
exhibits the highest accuracy in predicting ligand binding to rat
a7
protein. These findings may have implications in selecting the best
template to use in building an homology model for nAChR or
a chimera AChBP/nAChR. It is important to notice that in this
specific case of docking, the use of a validated homology model
didn’t add significant value, as compared to the simple use of the
template protein. Experimental data have shown that the designed
might interact with the rat a7 protein by adopting a “conformation
B-like” binding mode of DMXBA into the Ac protein, as also shown
in Fig. S2. The ligand OH group donates a hydrogen bond to the
backbone carbonyl oxygen atom of Leu-119, while its methoxy
group accepts a hydrogen bond from the amide side-chain group of
Gln-57, located in the complementary face. By contrast, a designed
spiroimidazoline containing a hydrogen-bond donor group at the
appropriate location, shown in Fig. 5b, might interact with the rat
novel compounds selectively bind to rat
a7 nAChR. In particular,
compound 3a binds 7 nAChR with a much better affinity than both
a
DMXBA (compound 2) and its active 4-OH metabolite (compound
3), and demonstrates partial agonism.
a
7 protein by adopting a “conformation A-like” binding mode of
4. Experimental section
DMXBA into the Ac protein, as also illustrated in Fig. S1.
In addition to hydrogen-bond interactions that have been
widely discussed herein, and the more commonly observed inter-
actions such as van der Walls, columbic and pairwise lipophilic,
Glide XP has revealed the importance of hydrophobic enclosures,
4.1. Molecular docking of ligands into AChBPs
Docking studies were carried out using Glide 5.5 [34e36].
Protein structures were prepared using Maestro protein prepara-
tion wizard. Following hydrogen-bonding assignment optimiza-
tion, water molecule orientations were exhaustively sampled, and
the protein-ligand complexes were energy-minimized until an
rmsd of 0.30 Å was reached. To validate the docking procedure,
a self-docking of the crystallized ligands into their cognate AChBP
protein was carried out. Validation studies were carried out using
co-crystal structures with the following pdb codes: 1uw6 (Ls in
complex with nicotine) [15], 2bj0 (Bt in complex with CAPS, i.e. 3-
(cyclohexylamino)-1-propanesulfonic acid) [16], and 2wn9 (Ac in
complex with 2-Methoxy-4-OH benzylidene anabaseine, called
compound 3 herein) [17]. A docking grid was constructed for each
protein by using the centroid of the bound ligand and a maximum
size of 20 Å. Flexibility of OH groups for amino acid side chains at
the binding site was allowed. Crystallographic water molecules at
each binding site were included. To mimic the quintessential
hydrogen bond involving Trp-147 (Ac) carbonyl oxygen and the
ligand-protonated nitrogen, experimentally observed in the crystal
structures of anabaseine DMXBA (compound 2 in Table 1), and 2-
MeO, 4-OHBA (compound 3 in Table 1), bound to Ac [17], an H-
bond constraint was accordingly specified. To mimic flexibility of
the protein structure, a scaling of van der Waals radii for non-polar
parts in the binding site and the ligands was carried out. Best
docked poses derived in SP mode were subsequently re-docked in
XP mode. The Glide XP visualizer module [34], was used to visualize
and analyze the derived docking poses.
cation-p interactions, and loss of configurational entropy in
contributing to the gain in binding affinity observed for the novel
spirodiazepine and spiroimidazoline quinuclidine analogs (Table 3
and Fig. 6). These additional and important interactions also
contribute to the binding of benzylidene anabaseine analogs to
AChBPs, as shown in Table S1. However, the fact that active
compounds such as 3a, 2b and 2d were not well docked in XP mode
after being successfully docked in SP mode, suggests that signifi-
cant work to account for receptor flexibility is still needed in order
to improve the accuracy of available docking packages.
As shown in Table 4, a comparison of Roc scores derived from
docking ligands into AChBPs crystal structures, in order to predict
ligand binding to rat
obtained with Ls, followed by the rat
a
7, suggests that the best results have been
7 homology model. On
a
average, poorest results were obtained when using Ac model,
which appear to be the most structurally dissimilar to both its
counterparts Bt and Ls, as can be seen from the rmsd values shown
in Table 5. These findings are in agreement with the work of de Kloe
et al. which has recently shown that a strong correlation exists
between binding affinities for a7 and AChBP-Ls and to a signifi-
cantly lesser extent, to AChBP-Ac [26]. In the light of the results
discussed above, we suggest that Ls crystal structure may be
a better template to use in developing an homology model for rat
a
7, as compared to Bt and Ac crystal structures. Moreover, Ls or Bt
protein may also be a better choice for designing an AChBP/rat
a
7
chimera protein. It will also be of interest to ascertain whether
differences in ring orientation are evident in the fluorescence
quantum yield, emission shifts and difference absorption spectra of
the substituted anabaseines in a complex with the Bulinus and
Lymnaea species.
4.2. Homology modeling of the
a7 nAChR
Three-dimensional models of the extracellular domain of the rat
7 protein were obtained from comparative modeling using spatial
a
In conclusion, we have used molecular docking to study the
interactions of benzylidene anabaseine congeners with AChBPs and
restraints, as implemented within Discovery Studio, using MOD-
ELLER [37e39]. The three-dimensional structure of AChBP from Ls
complexed with nicotine (pdb code 1uw6) was used as template.
The derived protein structures were validated with regard to
protein structure architecture and stereochemistry using the
protein analysis programs Procheck [40], and Verify_3D [41], as
implemented within the UCLA DOE web server [42]. Procheck
results showed that 86% of residues were found in the most-favored
region of the Ramachandran plot, and Verify_3D results indicated
that 77.4% of the residues had an averaged 3D-1D score >0.2.
Further refinement of the models was carried out within Maestro
(Schrodinger, Inc.) protein preparation module, prior to any dock-
rat
mediated via electrostatic interactions (hydrogen-bond,
and ), pairwise lipohilic, hydrophobic enclosures and loss of
a7 nAChR. We have found that binding affinity is primarily
p-cation
pep
configurational entropy as well. In particular, a hydrogen bond
donor feature appears to contribute to ligand binding to Ac, Bt and
a
7 nAChR. Docking of compounds 3 and 4 into the ligand binding
pocket of Ls shows the ligand OH group acting as hydrogen-bond
acceptor, instead. The use of AChBP as surrogate to design novel
spirodiazepines and spiroimidazolines quinuclidines which bind a7
nAChR has been investigated. Here the aza nitrogen in the bicycle
ring system takes the place of the cyclic imine nitrogen in
protonating, to be the hydrogen bond donor. A restriction in ligand
motion upon binding also appears to contribute to the increase in
binding affinity observed for the designed compounds. On the
ing simulation. In addition, biochemical validation of the
homology model was carried out by docking a diverse chemical
library of 493 compounds of known binding affinity to the
nAChR. The dataset, extracted from an in-house compound
a7
a7

5634
D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
collection, was comprised of 127 actives (
a
7 Ki ꢃ 500 nM) and 366
12 Hz), 3.53e3.15 (m, 4H), 2.23e1.98 (m, 5H). High resolution LCMS
m/e 300.1723, C17H22N3O2, calc. 300.1712.
decoys ( 7 Ki > 500 nM). The performance of the docking exercise
a
to classify compounds with respect to their binding affinity using
this homology model was evaluated by means of the area under the
receiver operating characteristic (ROC) curve. The accuracy ob-
tained was 0.74. Note that an area of 0.50 indicates random
performance, while an area of 1.00 indicates a perfect model.
4.3.7. 2-(3-Hydroxybenzofuran-2-yl)spiro[1,5-dihydroimidazole-
4,30-quinuclidine] (3a)
A mixture of 3-amino-3-(aminomethyl)quinuclidine (1) (75 mg,
0.5 mmol), methyl 2-(cyanomethoxy)benzoate (96 mg, 0.5 mmol)
and one drop of carbon disulfide was heated in a sealed vial at
100e110 ꢀC overnight. The reaction mixture was cooled to r.t. and
purified by preparative HPLC to yield 3a trifluoroacetate (98 mg,
4.3. Chemistry
Unless otherwise stated, all reactions were carried out under
a nitrogen atmosphere, using commercially available anhydrous
solvents. NMR spectra were carried out on a Varian 300 in the
solvents specified. HPLC purity determinations were carried out
48%). 1H NMR (CD3OD, 300 MHz)
d 7.98 (d, 1H), 7.65 (m, 2H), 7.43
(m, 2H), 4.26 (dd, 2H, 76, 12), 3.77 (dd, 2H, J ¼ 35,14), 3.60e3.38 (m,
4H), 2.50e2.33 (m, 2H), 2.15 (m, 3H).
using a Phenomenex C18 100 mm ꢄ 4.6 mm column with 5
m
m
4.3.8. 2-(3-Aminobenzofuran-2-yl)spiro[1,5-dihydroimidazole-4,30-
quinuclidine] (3b)
particle size; gradient: 97.5e2.5% A, 1 mL/min flow rate (A, 0.05%
trifluoroacetic acid in H2O; B, 0.05% trifluoroacetic acid in aceto-
nitrile). All test compounds were confirmed to be >95% pure by
HPLC and high resolution LCMS.
A mixture of 3-amino-3-(aminomethyl)quinuclidine (1) (75 mg,
0.5 mmol), (2-cyanophenoxy)acetonitrile (79 mg, 0.5 mmol) and
one drop of carbon disulfide was heated in a sealed vial at
100e110 ꢀC overnight. The reaction mixture was cooled to r.t. and
purified by preparative HPLC to yield 3b trifluoroacetate (145 mg,
4.3.1. 5-Phenylspiro[1,3-dihydro-1,4-diazepine-2,30-quinuclidine]-
7-ol (2a)
68%). 1H NMR (CD3OD, 300 MHz)
d 7.96 (d, 1H), 7.64 (t, 1H), 7.50 (d,
This procedure illustrates the general method for preparation of
2aeg. 3-Amino-3-(aminomethyl)quinuclidine (1) (75 mg,
0.5 mmol) and ethyl benzoylacetate (100 mg, 0.5 mmol) were
dissolved in i-butanol (3 ml). The reaction mixture was heated at
100 ꢀC overnight, cooled to ambient temperature and concentrated.
The residue was purified by preparative HPLC to yield 2a tri-
1H), 7.36 (t, 1H), 4.20 (dd, 2H, J ¼ 60, 12 Hz, 2H), 3.79 (dd, 2H, J ¼ 22,
12 Hz, 2H), 3.63e3.38 (m, 4H), 2.42 (m, 2H), 2.17 (m, 3H). High
resolution LCMS m/e 290.1707, C17H21N4O, calc. 290.1715.
4.3.9. 2-Phenylspiro[1,5-dihydroimidazole-4,30-quinuclidine (3c)
A
mixture of 3-amino-3-(aminomethyl)quinuclidine (1)
fluoroacetate (105 mg, 53%). 1H NMR (CD3OD, 300 MHz)
d
8.07 (d,
(475 mg, 3.0 mmol) and methyl benzimidate hydrochloride
(602 mg,3.5 mmol) in methanol (3 ml) was heated in microwave at
150 ꢀC for 5 min. The reaction mixture was concentrated in vacuo.
The residue was purified by preparative HPLC to yield 3c (0.3 g,
2H), 7.72 (m, 1H), 7.60 (m, 2H), 4.23 (dd, 2H, J ¼ 65, 12 Hz), 3.94 (dd,
2H, J ¼ 28, 10 Hz), 3.59e3.38 (m, 6H), 2.49 (s, 1H), 2.35e2.10 (m,
4H). High resolution LSMS, m/e 284.1767, C17H22N3O, calc. 284.1763.
28%). 1H NMR (CD3OD, 300 MHz)
d 8.00 (d, 2H), 7.85 (m, 1H), 7.69
4.3.2. 5-(2-Methoxyphenyl)spiro[1,3-dihydro-1,4-diazepine-2,30-
quinuclidine]-7-ol (2b)
(m, 2H), 4.33 (dd, 2H, J ¼ 75, 11), 3.80 (dd, 2H), 3.59e3.37 (m, 4H),
2.59e2.40 (m, 2H), 2.09 (m, 3H). High resolution LCMS m/e
242.1649, C15H20N3, calc. 242.1657.
High resolution LCMS m/e 314.1869, C18H24N3O2, calc. 314.1869.
4.3.3. 5-(4-Methoxyphenyl)spiro[1,3-dihydro-1,4-diazepine-2,30-
quinuclidine]-7-ol (2d)
4.3.10. 2-Spiro[1,5-dihydroimidazole-4,30-quinuclidine]-2-ylphenol
(3d)
High resolution LCMS m/e 314.1882, C18H24N3O2, calc. 314.1869.
3-Amino-3-(aminomethyl)quinuclidine (1) (75 mg, 0.5 mmol)
and 4-hydroxycoumarin (81 mg, 0.5 mmol) were dissolved in i-
butanol (3 ml). The reaction mixture was heated at 100 ꢀC over-
night, cooled to ambient temperature and concentrated. The
residue was purified by preparative HPLC to yield 3d tri-
4.3.4. 5-(2-Furyl)spiro[1,3-dihydro-1,4-diazepine-2,3’-
quinuclidine]-7-ol (2g)
High resolution LCMS m/e 274.1553, C15H20N3O2, calc. 274.1556.
fluoroacetate (15 mg, 8%). 1H NMR (CD3OD, 300 MHz)
d 7.88 (d, 1H),
4.3.5. 6-phenylspiro[1,3-dihydro-1,4-diazepine-2,30-quinuclidine]
(2h)
7.66 (m, 1H), 7.12 (m, 2H), 4.29 (dd, 2H, J ¼ 79, 10), 4.78 (dd, 2H),
3.56e3.37 (m, 4H), 2.50e2.33 (m, 2H), 2.16 (m, 3H).
3-Amino-3-(aminomethyl)quinuclidine (1) (75 mg, 0.5 mmol)
and
[3-(dimethylamino)-2-phenylprop-2-enylidene]-dimethy-
4.3.11. 2-Spiro[1,5-dihydroimidazole-4,30-quinuclidine]-2-ylbenzo
[b]furan (3e)
lammonium hexafluorophosphate (100 mg, 0.4 mmol) were dis-
solved in methanol (5 ml). The reaction mixture was refluxed
overnight and concentrated. The residue was purified by prepara-
tive HPLC to yield 2h trifluoroacetate (75 mg, 20%). 1H NMR
Compound (3e) was prepared from 3-amino-3-(aminomethyl)
quinuclidine (1) and methyl benzofurancarboximidate hydrochloride
according to procedure for 3c.1HNMR(CD3OD,300MHz)
d8.12(s,1H),
(CD3OD, 300 MHz)
d
0.15 (s, 1H), 8.01 (s, 1H), 7.52e7.38 (m, 5H),
7.88(d,1H), 7.74(d,1H), 7.66(t,1H), 7.48(t,1H), 4.36(dd, 2H, J¼ 75,10),
3.81(dd,2H), 3.62e3.38 (m, 4H), 2.58e2.40 (m, 2H), 2.18 (m, 3H). High
resolution LCMS m/e 351.1821, C20H23N4O2, calc. 351.1821.
4.50 (d, 1H, J ¼ 12 Hz), 3.73 (dd, 2H, J ¼ 27, 12 Hz), 3.60e3.38 (m,
5H), 2.22e1.98 (m, 5H).
4.3.6. 6-Phenylspiro[1,4-diazepane-2,30-quinuclidine]-5,7-dione
(2i)
4.4. Competition binding to receptors in membrane preparations
A mixture of 3-amino-3-(aminomethyl)quinuclidine (1) (75 mg,
0.5 mmol) and diethyl 2-phenylmalonate (118 mg, 0.5 mmol) were
heated at 150 ꢀC in microwave for 5 min. The reaction mixture was
cooled to r.t. and purified by preparative HPLC to yield 2i tri-
Binding assays to membrane bound nicotinic receptors were
carried out using standard methods adapted from published
procedures [43,44]. In brief, membranes were reconstituted from
frozen stocks and incubated for 2 h in 150
the presence of competitor compound (0.001 nMe100
radioligand. [3H]-methyllycaconitine ([3H]-MLA, PerkineElmer Life
m
l assay buffer (PBS) in
fluoroacetate (98 mg, 24%). 1H NMR (CD3OD, 300 MHz)
d
7.46e7.22
mM) and
(m, 5H), 4.30 (s, 1H), 4.02 (dd, 2H, J ¼ 50, 13 Hz), 3.53 (dd, 2H, J ¼ 53,

D.C. Kombo et al. / European Journal of Medicinal Chemistry 46 (2011) 5625e5635
5635
[12] K. Brejc, W.J. van Dijk, R.V. Klaassen, M. Schhurmans, J. van de Oost, A.B. Smit,
T.K. Sixma, Nature 411 (2001) 269e276.
[13] N. Unwin, J. Mol. Biol. 346 (2005) 967e989.
[14] Y. Bourne, T.T. Talley, S.B. Hansen, P. Taylor, P. Marchot, EMBO J. 24 (2005)
1512e1522.
[15] P.H. Celie, S.E. Rossum-Fikkert, W.J. van Dijk, K. Brejc, A.B. Smit, T.K. Sixma,
Neuron 41 (2004) 907e914.
[16] P.H. Celie, R.V. Klaassen, S.E. van Rossum-Fikkert, R. van Elk, P. van Nierop,
A.B. Smit, T.K. Sixma, J. Biol. Chem. 280 (2005) 26457e26466.
[17] R.E. Hibbs, G. Sulzenbacher, J. Shi, T.T. Talley, S. Conrod, W.R. Kem, P. Taylor,
P. Marchot, Y. Bourne, EMBO J. 28 (2009) 3040e3051.
[18] S. Mecozzi, A.P. West, D.A. Dougherty, Proc. Natl. Acad. Sci. USA 93 (1996)
10566e10571.
Sciences) was used for
membranes and [3H]-epibatidine (PerkineElmer Life Sciences) was
used for binding studies at the 2 on rat cortical membranes,
ganglion-type nicotinic receptors on SH-SY5Y cellular membranes
and muscle-type nicotinic receptors on TE-671 cellular membranes.
Incubation was terminated by rapid filtration on a multimanifold
tissue harvester (Brandel, Gaithersburg, MD) using GF/B filters
presoaked in 0.33% polyethyleneimine (w/v) to reduce non-specific
binding. Filters were washed 3 times with ice-cold PBS and retained
radioactivity was determined by liquid scintillation counting.
a7 binding studies on rat hippocampal
a4b
[19] J.D. Schmitt, C.G. Sharples, W.S. Caldwell, J. Med. Chem. 42 (1999) 3066e3074.
[20] D.A. Dougherty, J. Org. Chem. 16 (2008) 3667e3673.
[21] C. Ulens, A. Akdemir, A. Jongejan, R. van Elk, S. Bertrand, A. Perrakis, R. Leurs,
A.B. Smit, T.K. Sixma, D. Bertrand, I.J.P. de Esch, J. Med. Chem. 52 (2009)
2372e2383.
4.5. Binding data analysis
Single-point binding data was determined at a competitor
[22] R.L. Papke, E.M. Meyer, S. Lavieri, S.R. Bollompally, T.A. Papke, N.A. Horenstein,
Y. Itoh, J.K. Porter Parpke, Neuropharmacol. 46 (2004) 1023e1038.
[23] C. Stokes, J.K. Papke, N.A. Horenstein, W.R. Kem, T.J. McCormack, R.L. Papke,
Mol. Pharmacol. 66 (2004) 14e24.
[24] T.T. Talley, S. Yalda, K.-Y. Ho, Y. .Tor, F.S. Soti, W.R. Kem, P. Taylor, Biochem-
istry 45 (2006) 8894e8902.
[25] S.H. Slavov, M. Radzvilovits, S. LeFrancois, I. Stoyanova-Slavova, F. Sot,
W.R. Kem, A.R. Katritzky, Eur. J. Med. Chem 45 (2010) 2433e2446.
[26] G.E. De Kloe, K. Retra, M. Geitmann, P. Kallblad, T. Nahar, R. van Elk, A.B. Smit,
J.E. van Muijlwijk-Koezen, R. Leurs, R.H. . Irth, U.H. Danielson, I.J.P. . de Esch,
J. Med. Chem. 53 (2010) 7192e7201.
concentration of 5 mM and are expressed as the percent inhibition
of control radioligand binding. For IC50 determinations, replicates
for each point of a seven-point doseeresponse curve were averaged
and plotted against the log of drug concentration. IC50 values
(concentration of the compound that produces 50% inhibition of
binding) were determined by least squares non-linear regression
using GraphPad Prism software (GraphPAD, San Diego, CA). Ki
values were calculated using the ChengePrusoff equation [45].
[27] G. Mullen, J. Napier, M. Balestra, T. DeCory, G. Hale, J. Macor, R. Mack, J. Loch,
E. Wu, A. Kover, P. Verhoest, A. Sampognaro, E. Phillips, Y. Zhu, R. Murray,
R. Griffith, J. Blosser, D. Gurley, A. Machulskis, J. Zongrone, A. Rosen, J. Gordon,
J. Med. Chem. 43 (2000) 4045e4050.
[28] A.N. Placzek, F. Grassi, E.M. Meyer, R.L. Papke, J. Med. Chem. 68 (2005)
1863e1876.
[29] W.R. Kem, V.M. Mahnir, L. Prokai, R.L. Papke, X. Cao, S. LeFrancois,
K. Wildeboer, K. Prokai-Tatrai, J. Porter-Papke, F. Soti, Mol. Pharmacol. 65
(2004) 56e67.
Acknowledgment
We sincerely thank Dr Gary Byrd for high resolution LCMS and
Christopher Helper for binding affinity studies. We also thank
Merouane Bencherif, Craig Miller, and Heather Savelle for helpful
discussions, suggestions and comments.
[30] T., R. Young, B. Abel, B.J. Kim, R.A. Berne Friesner, Proc. Natl. Acad. Sci. USA 104
(2007) 808e813.
Appendix. Supplementary material
[31] R.A. Friesner, R.B. Murphy, M.P. Repasky, L.L. Frye, J.R. Greenwood,
T.A. Halgren, P.C. Sanschagrin, D.T. Mainz, J. Med. Chem. 49 (2006)
6177e6196.
[32] C.A. Chang, W. Chen, M.K. Gilson, Proc. Natl. Acad. Sci. USA 104 (2007)
1534e1539.
[33] A.E. Cho, V. Guallar, B.J. Berne, R. Friesner, J. Comp. Chem. 26 (2005) 915e931.
[34] Schrodinger, Inc., 101 SW Main Street, Suite 1300, Portland, OR 97204.
[35] R.A. Friesner, J.L. Banks, R.B. Murphy, T.A. Halgren, J.J. Klicic, D.T. Mainz,
M.P. Repasky, E.H. Knoll, M. Shelley, J.K. Perry, D.E. Shaw, J. Med. Chem. 47
(2004) 1739e1749.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ejmech.2011.09.033.
References
[1] M.N. Romanelli, P. Gratteri, L. Guangalini, E. Martini, C. Bonaccini, F. Gualtieri,
Chem. Med. Chem. 2 (2007) 2e24.
[2] S. Breining, Curr. Top. Med. Chem. 4 (2004) 609e629.
[3] J.D. Schmitt, Curr. Med. Chem. 7 (2000) 749e800.
[4] C. Conejero-Goldberg, P. Davies, L. Ulloa, Neurosci. Biobehav. Rev. 32 (2008)
693e706.
[36] T.A. Halgren, R.B. Murphy, R.A. Friesner, H.S. Beard, L.L. Frye, W.T. Pollard,
J.L. Banks, J. Med. Chem. 47 (2004) 1750e1759.
[37] Accelrys Inc., San Diego, CA, 2006.
[38] N. Eswar, M.A. Marti-Renom, B. Webb, M.S. Madhusudhan, D. Eramian,
M. Shen, U. Pieper, A. Sali, Comparative protein structure modeling with
MODELLER, in: Current Protocols in Bioinformatics. John Wiley & Sons, Inc.,
2006 Supplement 15, 5.6.1e5.6.30, 200.
[39] M.A. Marti-Renom, A. Stuart, A. Fiser, R. Sánchez, F. Melo, A. Sali, Annu. Rev.
Biophys. Biomol. Struct. 29 (2000) 291e325.
[40] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, App. Cryst 26
(1993) 283e291.
[41] D. Eisenberg, R. Luthy, J.U. Bowie, Methods Enzymol. 277 (1997) 396e404.
[42] http://www.doe-mbi.ucla.edu/SAVES/.
[5] A. Mazurov, T. Hauser, C.H. Miller, Curr. Med. Chem. 13 (2006) 1567e1584;
Erratum in, Curr. Med. Chem. 14 (2007) 1593.
[6] J.W. Daly, Cell. Mol. Neurobiol. 25 (2005) 513e552.
[7] A. Taly, P.J. Corringer, D. Guedin, P. Lestage, J.P. Changeux, Nat. Rev. Drug
Discov. 8 (2009) 733e750.
[8] K.B. Mihalak, F.I. Caroll, C.W. Luetje, Mol. Pharmacol. 70 (2006) 801e805.
[9] J.W. Coe, P.R. Brooks, M.G. Vetelino, M.C. Wirtz, E.P. Arnold, J. Huang,
S.B. Sands, T.I. Davis, L.A. Lebel, C.B. Fox, A. Shrikhande, J.H. Heym, E. Schaeffer,
H. Rollema, Y. Lu, R.S. Mansbach, L.K. Chambers, C.C. Rovetti, D.W. Schulz,
F.D. Tingley, B.T. O0Neill, J. Med. Chem. 48 (2005) 3474e3477.
[10] www.Clinicaltrials.gov.
[11] Targacept announces positive top-line results from phase 2 trial of TC-5619 in
cognitive dysfunction in schizophrenia. Press release, 01/21/2011, Winston-
Salem, NC, USA.
[43] P.M. Lippiello, K.G. Fernandes, Mol. Pharmacol. 29 (1986) 448e454.
[44] A.R. Davies, D.J. Hardick, I.S. Blagbrough, B.V. Potter, A.J. Wolstenholme,
S. Wonnacott, Neuropharmacol. 38 (1999) 679e690.
[45] Y. Cheng, W.H. Prusoff, Biochem. Pharmacol. 22 (1973) 3099e3108.