Synthesis and nicotinic receptor activity of chemical space analogues of N -(3 R)-1-azabicyclo[2.2.2]oct-3-yl-4-chlorobenzamide (PNU-282,987) and 1,4-diazabicyclo[3.2.2]nonane-4-carboxylic acid 4-bromophenyl ester (SSR180711)

Article abstract of DOI:10.1021/jm300030r

The Chemical Universe Generated Databases up to 11 atoms of CNOF (GDB-11) and up to 13 atoms of CNOClS (GDB-13) were used to enumerate analogues of the diamine part of two known α7 nicotinic receptor agonists and construct libraries of virtual analogues of these drugs. The libraries were scored using structure-based (docking to the nicotine binding site of the acetylcholine binding protein 1uw6.pdb) or ligand-based (similarity to the parent drugs) methods, and the top-scoring virtual ligands were inspected for easily accessible synthetic targets. In total, 21 diamines were prepared and acylated with aromatic carboxylic or oxycarbonic acids to produce 85 analogues of the parent drugs. The compounds were profiled by electrophysiology in Xenopus oocytes expressing human nicotinic acetylcholine receptor (nAChR) subtypes α7, α3β2, α4β2, α3β4, or α4β4. Characterization of selected compounds revealed eight inhibitors of the α7 nicotinic receptor and three positive allosteric modulators of the α3β2 nAChR.

Full text of DOI:10.1021/jm300030r

Article
pubs.acs.org/jmc
Synthesis and Nicotinic Receptor Activity of Chemical Space
Analogues of N-(3R)-1-Azabicyclo[2.2.2]oct-3-yl-4-chlorobenzamide
(PNU-282,987) and 1,4-Diazabicyclo[3.2.2]nonane-4-carboxylic Acid
4-Bromophenyl Ester (SSR180711)
Lise Brethous,† Noemi Garcia-Delgado,† Julian Schwartz,† Sonia Bertrand,‡ Daniel Bertrand,‡
́
and Jean-Louis Reymond*,†
†Department of Chemistry and Biochemistry, University of Berne, Freiestrasse 3, 3012 Berne, Switzerland
‡HiQScreen, 15 rue de l'Athen
ee, 1206 Geneva, Switzerland
́ ́
S
* Supporting Information
ABSTRACT: The Chemical Universe Generated Databases up to 11 atoms of
CNOF (GDB-11) and up to 13 atoms of CNOClS (GDB-13) were used to
enumerate analogues of the diamine part of two known α7 nicotinic receptor
agonists and construct libraries of virtual analogues of these drugs. The libraries were
scored using structure-based (docking to the nicotine binding site of the
acetylcholine binding protein 1uw6.pdb) or ligand-based (similarity to the parent
drugs) methods, and the top-scoring virtual ligands were inspected for easily
accessible synthetic targets. In total, 21 diamines were prepared and acylated with
aromatic carboxylic or oxycarbonic acids to produce 85 analogues of the parent
drugs. The compounds were profiled by electrophysiology in Xenopus oocytes
expressing human nicotinic acetylcholine receptor (nAChR) subtypes α7, α3β2,
α4β2, α3β4, or α4β4. Characterization of selected compounds revealed eight
inhibitors of the α7 nicotinic receptor and three positive allosteric modulators of the α3β2 nAChR.
nAChR ligands, in particular with new subtype selectivity and
INTRODUCTION
■
activity profiles. Although nAChR ligands span a very broad
area of chemical space, we asked the question whether new
ligands and activities might be uncovered using a highly focused,
in depth exploration of chemical space around selected known
ligands. This approach would be somewhat similar to the drug
design concept of systematic analoging, whereby new drugs can
be derived from known drugs by optimization of side activities
through structural optimization.4 Herein, we report the
discovery of new nAChR modulators with original activity
profiles among analogues of the structurally closely related
ligands N-(3R)-1-azabicyclo[2.2.2]oct-3-yl-4-chlorobenzamide
(PNU-282,987) and 1,4-diazabicyclo[3.2.2]nonane-4-carboxylic
acid 4-bromophenyl ester (SSR180711) (Figure 2).5−7 The
ligands were identified by enumeration of over 500000 virtual
analogues by de novo drug design,8,9 virtual screening using either
docking to the nicotinic binding site of the acetylcholine binding
protein (AChBP) or ligand-based similarity to the parent drugs,
chemical synthesis of 85 selected analogues, and profiling by
electrophysiology in Xenopus oocytes expressing various human
nAChR subtypes (α7, α3β2, α4β2, α3β4, or α4β4).
Nicotinic acetylcholine receptors (nAChRs) are widely ex-
pressed in the central and peripheral nervous system as well as
on non-neuronal cells.1 nAChRs are composed by the assembly
of five subunits around an axis of pseudo symmetry, and
binding of the ligand occurs in the extracellular domain at the
interface between two subunits. Most nAChR are heteromeric
and composed by the assembly of two or more distinct protomers.
To date, 17 genes encoding for nAChR have been identified in
mammals, yielding a large repertoire of receptor subtypes. Many of
these are implicated in fundamental biological and physiological
processes and their associated pathologies, leading to intense
pharmocology research to discover nAChR modulators. Thou-
sands of ligands have been reported that act on various nAChR as
partial or full agonists, antagonists, or positive or negative allosteric
modulators. Annotated lists of nAChR ligands are available in
public databases such as ChEMBL2 or in the online compound
catalog of Tocris.3 Nicotinic ligands are structurally highly diverse
and comprise small molecule natural products (e.g., nicotine,
epibatidine) and synthetic ligands (e.g., varenicline, morantel),
most often acting as subtype selective agonists, as well as larger
natural products (e.g., ivermectin, methyllycaconitine, tubocur-
arine) and peptide toxins (e.g., α-bungarotoxin, α-conotoxins)
typically acting as antagonists or allosteric modulators (Figure 1).
Despite the large number and structural diversity of already
known nAChR ligands, there is a continued demand for new
The enumeration of virtual analogues of 1 (PNU-282,987)
and 2 (SSR180711) was based on our recently reported
Received: January 10, 2012
Published: May 16, 2012
© 2012 American Chemical Society
4605
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
Figure 2. Structures of 1 and 2 and related α7 and α3β2 nAChR
modulators identified by virtual screening, synthesis, and electro-
physiology assay in Xenopus oocytes. The IC50 values given are for
inhibition of human α7 nAChR measured by electrophysiology.
Compounds 6a, 6f, and rac-12k are positive modulators of the α3β2
nAChR.
Figure 1. Structures of selected synthetic and natural nAChR
modulators. PAM = positive allosteric modulator. Peptides are written
in one-letter code aminoacids with uppercase numbers at cysteines
indicating disulfide bridge connectivities using the SMILES formalism
for cycles.10,11 Data from ref 3.
modulators of the α3β2 nAChR might find applications in the
treatment of sarcopenia, the muscular atrophy that develops
with aging.
Chemical Universe Generated Databases up to 11 atoms
(GDB-11) and up to 13 atoms (GDB-13).12−14 These publicly
available databases list 26.4 million molecules up to 11 atoms of
C, N, O, and F, respectively, and 977 million molecules up to
13 atoms of C, N, O, S, and Cl that are virtually possible
following selected criteria for chemical stability and synthetic
feasibility and have been used previously to guide the synthesis
of new inhibitors for the N-methyl D-aspartate (NMDA)
receptor glycine site and the glutamate transporter GLT-1.15−17
While the starting ligands were partial agonists of the α7
nAChR, our study uncovered eight antagonists of the α7
nAChR and three positive allosteric modulators of the α3β2
nAChR. Thus, the relatively narrow chemical space defined by
the monoacylated bicyclic diamine chemotype of the α7
nAChR agonists 1 and 2 provided sufficient structural diversity
to significantly vary the activity profiles, illustrating the value of
an in depth exploration of a local chemical space for drug
discovery. α7 nAChR modulators might find application in
treating neurological disorders and cancer.18,19 The positive
allosteric modulators of the α3β2 nAChR reported here
seem to be significantly more potent than the effect reported
for the α7 nAChR agonist Morantel.20−23 Positive allosteric
RESULTS AND DISCUSSION
■
Virtual Screening. The drug candidates 1 and 2 are
representative of a family of α7 nAChR agonists consisting of a
bicyclic scaffold comprising a bridgehead tertiary amine and a
primary or secondary amine acylated with various aromatic
carboxylic acids or carbamates, providing a favorable situation
for structural diversification via enumeration of analogues of the
diamine part. We set out to explore the chemical space around
these ligands by enumerating analogues of the diamine parts
with the help of the databases GDB-11 or GDB-13 and com-
bining the resulting diamines with various aromatic acyl groups
to generate a large library of virtual analogs.
Enumeration from GDB-11 and Selection by Docking
(Autodock and Glide). A first library DiamA was extracted
from GDB-11 by selecting the 72740 nonaziridine diamines
with heavy atom formula C4−9N2 containing a tertiary aliphatic
amine and a primary or secondary amine separated by a two-
carbon spacer. The library was expanded by also selecting the
4606
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
103528 tertiary amines with heavy atom formula C10N in GDB-
11 and attaching a primary amino group at all β-carbons relative
to the tertiary amine or by inserting a secondary amino group
in all β−γ carbon−carbon bonds relative to the tertiary amine,
which provided an additional 160832 diamines with 12 heavy
atoms again excluding aziridines. The combined set of 233572
diamines was then derivatized by attaching the five different
aromatic acyl groups a−e (Figure 3) occurring frequently in
ligands (Figure S2 in the Supporting Information). Although
the docking scores of 1 and 2 were only average, the docking
scores were considered to guide the selection of analogues fitting
into the nicotinic binding site but with scaffolds significantly
different from those of the parent molecules. The 1000 top-scoring
compounds from each docking series were visually inspected to
select structures for synthesis and testing. Easily accessible
acyclic, monocyclic, or bicyclic diamines were chosen, focusing
on compounds not previously described in the α7 nAChR
literature. Five diamines were taken from the Autodock series
and four scaffolds from the Glide series, featuring both chiral
and achiral diamines (Figure 3). For the chiral diamines, the
higher scoring enantiomer was considered as the synthetic
target and is shown in the figure.
Enumeration from GDB-13 and Selection by Ligand
Similarity. A second round of enumeration was undertaken
aiming at ligand-based virtual screening to select analogues of 1
and 2. In this case, a library of virtual diamines was assembled
from the Chemical Universe Database GDB-13, which was at
that point the newest database available. We followed simple
criteria relating to the structural composition of the diamines
using restrictions in values of the MQN descriptors, because
this method produces high-scoring compounds for ligand-based
virtual screening of very large databases such as PubChem, the
fragment subset of Pubchem, and GDB-13.30−34 Capitalizing
on the results of the first round of enumeration and selection
from GDB-11 discussed above, we limited the size of the
diamine to a maximum of nine carbon atoms and two nitrogen
atoms, suppressed all unsaturations, and focused on bicyclic
diamines since these seemed the most relevant as analogues of
the parent diamines. Thus, all molecules with up to nine carbon
atoms and exactly two nitrogen atoms, two cycles of size 4−8,
and containing only single bonds, were selected from GDB-13,
which produced library DiamB with 35666 diamines. A further
selection was performed by allowing only N-methyl groups as
terminal carbon atoms, thus avoiding consideration of multiple
analogues of the parent drugs containing the same diamine
scaffold appended with methyl groups, yet allowing N-methyl
amines related to the tropane scaffold to be retained. The
selection was further limited by considering only 5-, 6-, or 7-
membered rings, one tertiary amine, a maximum of two acyclic
carbon atoms, and imposing at least two bonds shared by the
two rings. These additional criteria focused the selection on
“non-zero bridged” bicyclic systems having a globular shape
similar to those of the parent drugs. This tighter selection left
only 344 diamines for further consideration.
To estimate if the diamines in library DiamB led to analogues
that were similar to the parent drugs, the shape similarity of their
acylated derivatives was quantified using the ROCS Tanimoto-
Combo score. The ROCS (rapid overlay of chemical structures)
TanimotoCombo score measures the similarity between 3D
shapes of molecules by maximizing an overlap function bet-
ween molecular shapes, considering these shapes as continuous
functions constructed from atom-centered electrostatic and
volume Gaussians.35 The score is maximized by comparing
various conformers of both query and reference molecule. This
3D shape-based approach is well-validated for ligand-based
virtual screening.36 To perform the shape similarity scoring,
the 35666 diamines in DiamB were virtually monoacylated on
available primary or secondary amines either with the 4-
chlorobenzoyl group to obtain virtual analogues of 1 or with
the 4-bromophenoxycarbonyl group to obtain virtual analogues
of 2, providing two sets of 56788 structures. These were
Figure 3. Synthesized analogues of 1 and 2 derived from virtual
screening.
bioactive analogues of the parent drugs24 to the free amino
group to generate 1167860 analogues of 1 and 2.
We initially chose docking to the nicotinic binding site of the
AChBP 1UW6.pdb as a model because this approach has been
shown to provide useful criteria for selecting nicotinic
ligands.25,26 The library of virtual acylated diamines was ranked
by docking using Autodock 3.0.527 and Glide.28 Both programs
correctly positioned nicotine in its crystallographic position.
Because of the limited throughput of the method, docking was
performed with a randomly selected subset of 72745 structures
taken from DiamA (6.2% of the entire library). These
structures produced 507030 stereoisomers using CORINA29
(LigPrep was used for Glide) to be considered for docking. The
docking score showed a typical Gaussian curve distribution
(Autodock, −12.9 to −4.6 kcal/mol estimated binding energy;
Glide, −9.1 to −1.5 GScore; see Figure S1 in the Supporting
Information). The docking poses of high-scoring ligands were
generally very similar to the poses obtained for the reference
4607
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
converted to 374728 different 3D stereoisomers, which were
individually scored. The ROCS score of the highest scoring
stereoisomer was retained for each 2D structure.
Derivatives of Known Diamines. The amide derivatives 3a−
d, 4a−e, 15a,b,e, 22b, and 28a were prepared in one step by
acylation of the corresponding commercially available free
diamines. Compounds 6,37−39 7,40 13, and 2741 were obtained
by derivatization of the corresponding diamines, which were
synthesized according to the literature. Although the 4- and 5-
member ring diamine analogues of 7 were also high-scoring in
virtual screening, their synthesis failed in our hands, although
these diamines have been described in the literature.42
The shape similarity data showed that most of the virtual
ligands generated had a score higher than the threshold of 1.4
considered to be sufficient to predict bioactivity (Figure S3 in
the Supporting Information). Indeed, from the 56294
successfully generated diamines, 74.9% had a ROCS score
higher than 1.4 to one of the two reference drugs, and 19.2%
had a score higher than 1.4 to both reference drugs (in each
case with the corresponding acyl group). The percentages were
even higher with the more restricted subset of 344 analogues,
with 93.3% having a ROCS score above 1.4 for one reference
compound and 54.1% for both reference compounds, high-
lighting the focused design of the virtual library. The subset of
344 bicyclic diamines was inspected to identify targets for
synthesis and testing. Most of the diamines giving high-scoring
compounds comprised ring size analogues of the diamines
cores of 1 and 2 and as such were already described in the
patent literature. Nevertheless, the bicyclic diamines 12−14
were identified as potentially interesting structures, which had
not been previously described in the nicotinic receptor ligand
literature (Figure 3). In addition to the ligands derived from
virtual screening, analogues 15−28 were also prepared in the
course of the investigations, in part during activity optimization
attempts of selected active compounds (see below) and in part
due to synthetic opportunities that arised in the course of the
various syntheses (Figure 4).
2-Aminomethyl-pyrrolidine Derivatives (Scheme 1). Dia-
mine 31 was prepared starting with alkylation of D-proline
Scheme 1. Synthesis of 2-(Aminomethyl)pyrrolidine 8a−e,
a
16b, 17e, and 18k
a
Conditions: (a) 1-Bromo-2-pentyne, LiI, NaHCO3, acetonitrile,
60 °C, 46 h (65%, 29). (b) THF/NH3 aqueous, 60 °C, 14 h (20−
59%). (c) LiAlH4, Et2O, reflux, 12 h (13−31%) [32% (S)-31, 6% (S)-
32]. (d) ArCO2H, EDC, HOBt, N-methyl morpholine, CH2Cl2, 25 °C,
14 h, (10−61%) or ArOCOCl, NEt3, THF, 2 h (3−32%).
(e) Cyclohexanone, NEt3, NaBH3CN (quant., 33) or acetone,
NaHCO3, NaBH(OAc)3, MeOH, 25 °C, 48 h (34) or 1-bromopropane,
NaHCO3, CH3CN, 60 °C, 48h (78%, 35).
methyl ester hydrochloride with 1-bromo-2-pentyne43 in the
presence of NaHCO3 and LiI in ACN at 60 °C to give amino
ester 29 in 65% yield. Amide 30 was obtained in 54% yield by
dissolving intermediate 29 in THF/NH3 25% (1:9) and heating
up to 60 °C. Reduction with LiAlH4 in Et2O, quenched with
water and NaOH,44 and purification by column chromatog-
raphy finally afforded diamine 31 in 32% yield together with a
small amount of amino alcohol 32. Diamine 31 was then
acylated to afford ligands (R)-8a−e. Using the same procedure,
the enantiomer (S)-8b was prepared starting with L-proline
methyl ester hydrochloride. The corresponding ester (S)-19b
was obtained by acylation of alcohol (S)-32 obtained as a
byproduct of the reduction step. Amides 16b, 17e, and 18k
were synthesized starting from D-proline methyl ester hydro-
chloride by reductive amination with cyclohexanone (33) or
acetone (34) or alkylation with 1-bromopropane (78% 35),
followed by treatment with an aqueous solution of ammonia
Figure 4. Further analogues prepared in this study.
Synthesis. The synthesis of the various ligands is described
below starting with the acylation of known diamines, followed
by the synthesis and acylation of further diamines in the order
of structural complexity. The selection of carboxylic acids and
carbamates used for acylation was dictated by the availability of
the reagents and synthetic yields when preparing the five acyl
groups a−e known to lead to nAChR activity (Figure 3) and in
selected cases by efforts to optimize activities as detailed in the
electrophysiology section.
4608
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
a
(51% 36, 20% over two steps 37, and 59% 38), reduction of the
amide by LiAlH4 and acylation.
Scheme 3. Synthesis of Diazepine 25c and Diazocines 5a−d
3-Aminopyrrolidine and 4-Aminopiperidine Derivatives
(Scheme 2). (S)-(−)-1-Boc-3-aminopyrrolidine was acylated
Scheme 2. Synthesis of 3-Aminopyrrolidine and 4-
Aminopiperidine Derivatives (S)-20b, 21b, and 23b and (S)-
a
9a,b
a
(a) Acetone, NaBH3CN, HCl cat., MeOH (93%). (b) TFA, CH2Cl2.
(c) ArCO2H, EDC, HOBt, N-methyl morpholine, CH2Cl2, 25 °C, 14 h
(21−59%), or ArOCOCl, NaOH (10%). (d) 1,4-Dibromobutane,
Na2CO3, o-xylene, 130 °C, 18 h (42%).
TFA and acylation of the free secondary amine. Diazocines
5a−d were synthesized starting from a dialkylation reaction of
N-isopropylethylenediamine with 1,4-dibromobutane using
Na2CO3 in o-xylene, as previously described in a methodology
study of synthesis of cyclic diamines.45
Bicyclic Amine Derivatives (R,S)-10, (S)-11, and (R)-11
(Scheme 4). Bicyclic diamine 48, which has been reported in a
study of the reactivity of 2-azaquinolizinium oxides,46 was
prepared starting with peptide coupling between (R)-(+)-N-
Boc-2-piperidinecarboxylic acid and L-alanine ethyl ester. The
resulting dipeptide was deprotected and cyclized in a one-pot
reaction to give diketopiperazine 47 in 34% yield. Finally,
reduction with LiAlH4 and distillation gave diamine 48 in 22%
yield, which was converted to amides (R,S)-10a−c. Diamine 51
was prepared as described in a methodology study of reductive
cyclization with nitriles.7,47 Thus, alkylation of L-proline
ethylester with acrylonitrile gave 49 in 58% yield. The
reduction of the nitrile to the corresponding amine with Nickel
Raney gave the lactam 50 directly in 57% yield. Reduction of
the lactame with LiAlH4 in dry THF and purification by
distillation finally gave the diamine in 36% yield. Diamine 51
was then acylated to the corresponding amides (S)-11a−e,
which were purified by preparative RP-HPLC. Following the
same procedure, the enantiomer (R)-11c,e−g,k were obtained
starting from D-proline methyl ester hydrochloride.
2-Aminomethyl Quinuclidine Derivatives (Scheme 5).
Diamine 54, reported as an intermediate in the synthesis of a
serotonin-3 receptor ligand, was prepared by adapting the
literature.48 Starting with the deprotonation and reduction of
the commercial 2-methylene-3-quinuclidinone hydrochloride,
followed by a SN2′ substitution by treatment with thionyl
chloride (58%), a substitution of chloride by reaction with an
aqueous solution of ammonia gave the desired diamine 54 in a
quantitative yield. The corresponding amides 26a,c,e−g,k were
then prepared directly from 54. The saturated analogues (rac-
12c,e,f,k) were obtained by acylation of the hydrogenated
diamine 55 or by reduction under hydrogen conditions of the
amide compounds.
a
Conditions: (a) ArCO2H, EDC, HOBt, N-methyl morpholine,
CH2Cl2, 25 °C, 14 h, or ArOCOCl, NaOH, 3 h (9−89%). (b) TFA,
CH2Cl2. (c) Acetaldehyde, NaBH3CN, MeOH, 30 min [(S)-20b, 10%]
or cyclohexanone, NaBH3CN, MeOH [(S)-21b: 50%] or 1-bromo-2-
pentyne, K2CO3, DMF, 25 °C, 7 h [(S)-23b, 75%]. (d) 2-Methyl-
cyclopentanone, NaBH3CN, MeOH/AcOH, 25 °C, 14 h [(R)-40, 54%;
24b, 27%]. (e) MsCl, Et3N, THF, −10 to 25 °C, 14 h (90%).
(f) Potassium phthalimide, DMF, 90 °C, 38 h (96%). (g) NH2NH2·H2O,
Et2O, reflux, 3 h (94%).
with 1,3-benzothiazole-6-carboxylic acid providing compound
(S)-39b, which was deprotected by TFA before being
submitted to reductive amination or alkylation to give the
amides 20b, 21b, and 23b. Diamine (S)-43 was prepared
starting with reductive alkylation of (R)-3-pyrrolidinol with 2-
methylcyclopentanone using NaBH3CN to give amino alcohol
(R)-40 in 54% yield as a 1:1 diastereomeric mixture. The
alcohol was mesylated to 41, substituted with potassium
phthalimide to (S)-42, and deprotected with hydrazine in Et2O
to provide (S)-43, which was then used to prepare the
corresponding amides (S)-9a,b. The 4-aminopiperidineamide
24b was obtained from 4-amino-1-Boc-piperidine, which was
acylated to give 44b before being deprotected and submitted to
a reductive amination with 2-methylcyclopentanone.
Bicyclic Amine Derivatives rac-14 (Scheme 6). Bicyclic
diamine rac-62 was synthesized in eight steps starting from the
commercial 5-norbonen-2-yl-acetate, which was submitted to a
dihydroxylation reaction under osmium tetroxide condition to
give compound rac-56 (quant.). A Lemieux−Johnson reaction,
followed by a reductive amination with benzylamine, afforded
the bicyclic compound rac-57 (62%). After reduction of the
acetal and a swern oxidation, keto-compound rac-58 was trans-
formed to oxime rac-59, which was submitted to a Beckmann
Diazepine and Diazocine Derivatives (Scheme 3).
Diazepine 45 was prepared by reductive amination of 1-Boc-
hexahydro-1,4-diazepine with acetone. The corresponding
amide 25c was then obtained after Boc deprotection with
4609
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
Scheme 4. Synthesis of Bicyclic Amides (R,S)-10, (S)-11, and
(R)-11
Scheme 6. Synthesis of Bicyclic Amine Derivatives rac-
a
a
14a,e−g
a
Conditions: (a) N-Methyl morpholine N-oxide, OsO4, acetone, H2O,
40 °C, 18 h (quant.). (b) NaIO4, DCE, H2O, 0 °C, 18 h then
benzylamine, NaBH(OAc)3, DCE, 18 h (62%). (c) (i) KOH, MeOH,
H2O, 25 °C, 18 h (52%); (ii) oxalyl chloride, DMSO, CH2Cl2,
−78 °C, 15 min then NEt3, −78 °C, 15 min (89%). (d) Hydroxylamine
hydrochloride, NaOAc, MeOH, reflux, 3 h (65%). (e) H2SO4
concentrated, 100 °C, 4 h (46%). (f) NaH, THF, 0 °C, 30 min then
MeI, 25 °C, 18 h (82%). (g) BH3·THF, THF, reflux, 12 h then MeOH,
25 °C, 1 h (67%). (h) Pd(OH)2, H2, MeOH, reflux, 6 h (quant.).
(i) ArCO2H, PyBOP, DIEA, DCM, 14 h (13−22%) or ArOCOCl, NEt3,
THF, 2 h (13%).
a
Conditions: (a) Ala-OEt·HCl, EDC, HOBt, Et3N, 25 °C, 15 h, then
CF3CO2H, 25 °C, 1 h, then Et3N, MeOH, reflux, 4 h (34%).
(b) LiAlH4, THF, reflux, 14 h (22%). (c) ArCO2H, EDC, HOBt,
N-methyl morpholine, CH2Cl2, 25 °C, 14 h (8−52%). (d) Acrylonitrile,
aqueous KOH, 50−70°C, 2.5 h (58%). (e) H2 (48 bar), Raney-Ni,
MeOH, 80 °C, 6 h (57%). (f) LiAlH4, THF, reflux, 14 h (36%).
(g) ArCO2H, EDC, HOBt, N-methyl morpholine, CH2Cl2, 25 °C, 14 h,
or ArOCOCl, NaOH or NEt3 3 h (8−52%).
rearrangement by treatment with concentrated sulfuric acid to
give amide rac-60 (46%). A methylation of the amide (82%),
followed by reduction with borane (67%) and deprotection of
the benzylated amine under Pd(OH)2/H2 conditions, afforded
the desired bicylic diamine rac-62, which was directly used to
form the bicyclic amides compounds rac-14a,e−g.
Scheme 5. Synthesis of 2-Aminomethyl Quinuclidine
Derivatives 12c,f,k and 26a,c,e−g,k
a
Electrophysiology. Activity was measured by electro-
physiology in oocytes expressing various recombinant human
nAChR subtypes.49 Compound effects were assessed by first
recording the current evoked by a brief ACh test pulse (200 μM,
5 s) and then by recording the response to the same ACh test
pulse following a 30 s incubation in presence of the compound.
α7 nAChR Activity. The initial activity screening was
performed against the α7 nAChR subtype, for which 1 and 2
are partial agonists (Table S1 in the Supporting Information).
Compounds 2a−(S)-11e (36 compounds, Figure 3) synthe-
sized as examples of high-scoring analogues by docking were
investigated. All ligands behaved as antagonists of acetylcholine,
evoking no current during exposure to the compound alone but
causing a reduction of the subsequent ACh-evoked current at
the three concentrations tested. Out of the 36 ligands tested, 12
showed weak activity (less than 30% inhibition at 10 μM), 19
showed moderate inhibition (30−70% inhibition at 10 μM),
and 5 showed a strong inhibition of the acetylcholine response
(>70% at 10 μM, observed for 5c, 6d, 8a,b, and 9b).
The strongest hits inspired the preparation of further
analogues to see if activity could be improved (Figure 4).
Analogues of pyrrolidines 8b and 9b were prepared bearing
simplified alkyl residues at the tertiary amine in form of
analogues 15a−23b, including the enantiomer (S)-8b and an
ester analogue (S)-19b obtained as a side product. In addition,
the ring-expanded analogue of 9b was prepared as 24b derived
from the achiral 4-amino-piperidine. The diazepine compound
a
Conditions: (a) K2CO3, CH2Cl2, H2O, 25 °C, 3 days then NaBH4,
MeOH, 0 °C, 30 min (51%). (b) Thionyl chloride, CH2Cl2, 16 h
(58%). (c) NH4OH, MeOH, 25 °C, 4 h (quant.). (d) ArCO2H, EDC,
HOBt, N-methyl morpholine, CH2Cl2, 25 °C, 14 h or ArOCOCl,
NEt3, THF, 2 h (5−48%). (e) Pd/C, H2, MeOH, 4 h (3−76%).
4610
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
25c was also prepared with an intermediate ring size between
the weakly active piperazine 4c (19% inhibition) and its more
active 8-membered ring analogue 5c (75% inhibition). (R)-
Enantiomers of (S)-11 were also prepared because the
bromophenyl carbamate showed a much higher shape similarity
(ROCS score) to 2 than (S)-11, the enantiomer originally
selected by its docking score, to test if any (R)-11 derivative
might act as an agonist. However, none of these derivatives
showed any significant activity at the α7 nAChR, with no
detectable agonist effect and less than 50% inhibition of the
ACh signal at 10 μM ligand.
Inhibition of the α7 nAChR was characterized in more detail
by measuring IC50 values for compounds 5a,c, 6a, 7e, 8b, 9b,
10c, and (S)-11e, thus including one compound for each of the
acylated diamines investigated (except for 3 and 4, which
showed only very weak activities with all derivatives). Most
values were in the 5−10 μM range, which corresponds to the
affinity range for the parent ligand 2 (Table 1). Closer
concentration (Figure 5B). Ligands 6f, rac-12k, and rac-14e
were further characterized by measuring the concentration−
activation relationship for acetylcholine in the absence or
presence of a prepulse of the putative allosteric modulator (10 μM,
Figure 5C). This showed that only 6f induced a sustained
potentiation characterized by an increase of the amplitude of the
ACh-evoked currents over the entire range of the concentration−
activation curve and caused a shift of the agonist response toward
a higher sensitivity. These features are typical of an allosteric effect
and cannot be attributed to potentiation such as that observed
with a coagonist activity.
Allosteric modulators are compounds that, by definition,
bind on the receptor at a site that is distinct from the natural
ligand binding site, also called orthosteric binding site. The
physical interaction of the allosteric modulator with the
receptor complex yields either an increase of the receptor
function or a reduction of the overall activity.50 While most
allosteric modulators have a molecular structure quite different
from the natural ligand, molecules sharing some analogies with
the ligand have also been shown to influence receptor
functions. For example, tubocurarine that is known to inhibit
the nAChRs was found to potentiate the ACh-evoked currents
at the α3β4 receptors by a mechanism that is probably distinct
from the true allosteric modulation.51 Another example of
receptor modulation, including the α3β2 receptors, was
provided by studies of the anthelminthic compound morantel.
This molecule, which acts as an agonist at the invertebrate
nAChRs, was shown to enhance, albeit to a limited extend, the
ACh-evoked current at the α3β2. Moreover, it was recently
proposed that morantel binds at the interface between the
β- and the α-subunit.20−23 Altogether, the data presented above
in Figure 5 indicate that the mode of action of compound 6f is
distinct from that previously reported for morantel.
a
Table 1. Activities on the Human α7 nAchR
compd
IC50 or EC50 (μM)
activity type
mixed antagonist
5a
7.0 1.1
6.9 1.0
7.2 1.2
10 0.8
5c
ND
6a
mixed antagonist
7e
ND
(R)-8b
(S)-8b
9b
40
45
6
1
ND
ND
5.0 0.7
6.1 1.5
5.6 1.7
ND
10c
(S)-11e
2
noncompetitive antagonist
competitive antagonist to ACh
b
b
4.4
partial agonist
a
Electrophysiological assay in Xenopus oocytes. See the Supporting
b
Information for details. Data from ref 6, indicated as EC50 = 4.4 μM
(2.5−7.8 μM), which corresponds to a binding affinity of Ki = 14 nM.
ND, not determined. See Figure S4 in the Supporting Information for
electrophysiology traces.
As it is known that α3β2 nAChRs are expressed at the
presynaptic ending of the neuromuscular junction, it can be
postulated that a positive allosteric modulator should increase
the release of neurotransmitter from the motor nerve. This
effect could find immediate applications in diseases, or
conditions, in which neuromuscular transmission is impaired,
for example, in the treatment of sarcopenia, a progressive
neuromuscular atrophy that develops with aging.
investigation of the inhibition mechanism was carried out in
four cases. This showed that only (S)-11e behaved as a
competitive inhibitor of acetylcholine and, therefore, could be
assigned to binding to its binding pocket as anticipated from
the virtual screening procedure.
Similarity Analysis. Most of the 85 compounds tested in
this study showed measurable levels of activity in terms of an
inhibitory effect on the α7 nAChR or a positive modulation of
the α3β2 nAChR. However, none of the compounds showed
agonistic effects on the α7 nAChR, which was the main
characteristic of the two reference compounds 1 and 2. A
similarity analysis was performed to test if the differences in
observed activities could be traced back to inherent structural
features of the diamines. The diamine libraries DiamA and
DiamB and the synthesized examples were compared to 1 and
2 in terms of shape and substructure similarity. A random set of
small molecules of similar size from the public access database
of commercially available druglike ligands zinc is not
commercial (ZINC) was included in each case as a control.52
A shape similarity analysis based on the ROCS Tanimoto
Comboscore was performed on both diamine series using the
4-bromophenyl carbamates for comparison to 2 and the 4-
chlorobenzamides for comparison with 1, as used above for
ligand-based similarity selection. Both acylated diamine series
scored significantly higher than a control set of 39853
compounds from ZINC (with 16 ≤ hac ≤ 22 corresponding
to the range covered by the carbamates and amides) and
α3β2 nAChR Activity. Additional subtypes of the nAChR
were tested with the different compounds to assess the
functional diversity of the prepared chemical space analogues.
Activity profiling was performed in oocytes expressing the
α3β2, α4β2, α3β4, and α4β4 nAChR subtypes (Table S1 in the
Supporting Information). These data showed that a positive
deflection was specifically observed at α3β2, suggesting a
possible positive allosteric modulation of this receptor subtype
in compounds 6a, 6f, rac-12k, and rac-14e (Figure 5A). To
probe the effect of the aromatic acyl group on the α3β2 nAChR
modulatory effect, we focused our attention on analoging 6a
due to its simple synthesis and its elegant achiral diamine
scaffold, yielding analogues 6b−k. Activity profiling showed
that the positive modulatory effect on α3β2 was preserved in
several derivatives of 6a, in particular the unsubstituted
benzamide 6f.
Closer investigation of 6a, 6f, rac-12k, and rac-14e at the
α3β2 nAChR by testing the effect at 1, 10, and 100 μM ligand
on a 50 μM acetylcholine pulse showed that only 6f and rac-
12k preserved the positive modulatory effect at a high ligand
4611
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
Figure 5. Positive modulation at the human α3β2 nAChR. (A) Screening for the effects of compounds at the human α3β2. Cells expressing the
human α3β2 nAChR were challenged first with a reference ACh test pulse (50 μM, 5 s), and the response to the same test pulse was repeated after
30 s of exposure to a fixed concentration of compound. Plot of the ratio of the peak currents, recorded after compound exposure over peak current
recorded in control, yields a ratio where unity corresponds to the absence of effect and values superior to unity indicate a potentation of the ACh
response. See Table S1 in the Supporting Information for complete data set. (B) Effects of four positive allosteric modulators at three concentrations.
Data were collected and analyzed as in panel A. Note that exposure to a concentration of 100 μM inhibits the ACh-evoked current. (C) Effects of a
fixed concentration of the positive allosteric modulator (10 μM) on the ACh concentration activation curve. Note that compound 6f causes a shift of
the concentration activation curve toward a higher sensitivity and an increase of the maximal current amplitude.
Table 2. Similarity Values for Virtual Libraries
no. of
compds
ROCS to 1 (% >
ROCS to 2 (% >
% > 1.4 for 1 and 2
% > 0.7 for 1 and 2
a
b
c
a
b
c
data set
1.4)
1.4)
(1 or 2)
TSF to 1 (% > 0.7)
TSF to 2 (% > 0.7)
(1 or 2)
d
ZINC
39853
44410
56294
344
1.12 0.17 (6.0)
1.43 0.19 (57.0)
1.34 0.19 (41.7)
1.50 0.19 (77.0)
1.11 0.13 (2.2)
1.42 0.15 (53.0)
1.42 0.12 (52.4)
1.49 0.15 (70.4)
1.4 (6.7)
34.1 (75.9)
19.2 (74.9)
0.28 0.08 (0.1)
0.77 0.10 (73.6)
0.82 0.06 (99.9)
0.25 0.04 (0)
0.0 (0.1)
DiamA
DiamB
subs344
0.76 0.08 (73.1)
0.79 0.05 (94.3)
0.81 0.05 (99.7)
68.4 (78.3)
94.3 (100.0)
99.7 (100.0)
54.1 (93.3)
0.82 0.06 (100.0)
a
b
ROCS Tanimoto Comboscore and TSF for the 4-chlorobenzamide. ROCS Tanimoto Comboscore and TSF for the 4-bromophenylcarbamate of
the diamines from DiamA, DiamB. or subs344 or for the compounds from ZINC. The 1024 bits Daylight type substructure fingerprint from
c
d
ChemAxon was used. Refers to the scores of the corresponding acylated diamines or of the ZINC cpds. Random selection from ZINC with 16 ≤
hac ≤ 22 covering the molecular sizes of compounds derived from acylation of the diamines libraries DiamA and DiamB.
contained a much more significant fraction of compounds scoring
above 1.4 than the control set (Table 2). The diamine libraries
also showed a high degree of similarity to the reference drugs
when compared by substructure similarity using the Tanimoto
similarity coefficient TSF for a 1024 bit daylight type substructure
fingerprint. Both the average TSF and the fraction of high similarity
compounds (TSF > 0.7) were much higher in these libraries as
compared to the control set extracted from ZINC.
substructure similarity to the reference compounds when com-
pared to random compounds of similar size (Figure 6). The ligands
derived from DiamA showed substantially lower substructure simi-
larity to the reference ligands (lower TSF) than the ligands derived
from DiamB. This dissimilarity was caused by the presence of
multiple CC double bonds and triple bonds and small rings present
only in the former. Some ligands derived from DiamA combined
low TSF similarity with high ROCS score to the reference
compounds. These compounds were derived from polyunsaturated
diamines often containing 3- and 4-membered rings.
The ROCS and TSF data above showed that the diamine
library designs enforced a substantial degree of shape and
4612
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
Figure 6. Shape and substructure similarities. For each diamine, the corresponding 4-chlorobenzamide derivative is compared to 1, and the 4-
bromophenyl-carbamate is compared to 2. Compound sets: orange dots, DiamA, 44410 compounds; cyan dots, DiamB, 56294 compounds; black
dots, restricted subset of 344 diamines from DiamB; and blue dots, ZINC random subset 16 ≤ hac ≤ 22, 39853 compounds. Red crosses and
combined bar/line plot: synthesized compounds 3−28 and the two reference ligands 1 and 2.
The diamines selected for synthesis were quite representative
of the overall library design, with substantial but not extremely
high similarity scores to either reference drug (in form of the
derivatives with common acyl groups as discused above, Figure 6).
The correlation between similarity scores and the observed activity
was weak within the compound series. For example, while the
acyclic N,N-dimethyl-ethylenediamine used for 3a−d had a much
lower shape and substructure similarity to either drugs caused by
its smaller size and lacked activity, the 8-membered ring derived 5a
and 5c showed good inhibition of the α7 nAChR despite a low
shape similarity to either reference compounds. On the other
hand, the N-ethyl-3-aminopyrrolidine 20 returned high scores
toward both reference drugs, but compound 20b did not exhibit
any particularly high activity. As to the lack of α7 nAChR agonist
effect among the compounds tested, it should be noted that an
agonist effect requires the stabilization of the open channel state of
the receptor in transition to its desensitized state and that this
effect is only rarely observed when screening compound series.
Antagonistic effects as those observed here are generally more
common.
The similarity analysis showed that the exploration of
chemical space suggested by the GDB approach guided the
search somewhat further away from the parent drugs than a
classical analoging series might have achieved, for example, by
preparing ring size and methylated analogues of the parent
diamines. While this diversification was apparently too
significant to retain the agonistic effects of the parent drugs,
the compounds prepared clearly covered a useful diversity
range, as exemplified by the discovery of several competitive
inhibitors of the α7 nAChR and of positive modulators of the
α3β2 nAChR. In particular, 6f was derived from a relatively
simple diamine quite substantially different from the parent
4613
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
only chain C and D of 1UW6.pdb. The receptor was prepared by
Maestro program using default settings.
diamine that might not have been considered by classical
analoging strategies.
Shape Similarity Scoring with ROCS. Prior to the ROCS
calculation, all molecules of the input database were submitted to
“FLIPPER”, which identifies all unspecified stereocenters in the
molecule and enumerates both stereochemistry states of each
unspecified stereocenter, resulting in all possible stereoisomers.
“OMEGA” was applied on the stereoisomers to create a maximum
of 200 lowest energy 3D conformers. The conformers were used as
input for the ROCS calculation, in which the best “TanimotoCombo”
overlap score was retrieved per stereoisomer. Lastly, the 3D structures
were converted back to SMILES, and the best score per SMILES was
kept as final ROCS score. The example structures 3−28 and the two
reference structures 1 and 2 were prepared similarly prior to ROCS
calculation.
CONCLUSION
■
The experiments above demonstrate the potential of the
chemical universe databases GDB-11 and GDB-13 to inspire
the synthesis of new analogues of 1 and 2. Starting from these
α7 nAChR partial agonists, exploration of chemical space was
performed by enumeration of virtual analogues with the help of
the databases, selection based on docking or ligand-based
virtual screening, chemical synthesis of selected virtual hits and
analogues, and testing by electrophysiology. This workflow led
to the discovery of eight new α7 antagonists and three new
α3β2 nAChR positive modulators. Such α3β2 nAChR positive
modulators might have therapeutic potential for the treatment
of sarcopenia. Although the identified ligands possess clear
similarities to 1 and 2 when considering the overall structure of
the ligands, their diamine parts are still significantly different
and might not have been chosen by a pure intuition-driven
design exercise. These data illustrate the innovative contribu-
tion of systematic, in depth exploration of local chemical space
using the GDB databases as a support to ligand discovery.
Synthesis. General. All reagents used were purchased from
commercial sources without further purification. All solvents used in
reactions were distilled from technical quality. Dry solvents were
obtained directly from a drying solvent system. Chromatographic
purifications (flash) were performed with Silica Gel 60 from Fluka
(0.04−0.063 nm; 230−400 mesh ASTM). Preparative RP-HPLC was
performed with a Waters Delta Prep 4000 system with a Waters
Prepak Cartridge (500 g) as column and Waters 486 Tunable
Absorbance Detector or with a Waters Prep LC Controller sytem with
a Dr. Maisch GmbH Reprospher C18-DE, 100 mm × 30 mm, particle
size 5 μm, 100 Å pore size column, and a Waters 2489 UV/Visible
Controller. The compound purity was determined by analytical RP-
HPLC. All compounds whose IC50 values were determined had
purities ≥95%. Overall, 77 compounds had purities ≥95%, and 7
compounds had 85−95% purity; see the Supporting Information, pp
S138−S139, for details. Analytical RP-HPLC was performed on
Waters 600E systems with a Waters Atlantis (4.6 mm × 100 mm, dC18,
5 lm) column, UV detection with Waters 996 photodiode array
detector. Analytical RP-UHPLC was also performed on a Dionex
ULTIMATE 3000 RS chromatography system (ULTIMATE 3000 RS
Photo diode array detector) using a Dionex Acclaim RSLC 120 C18,
3.0 mm × 50 mm, particle size 2.2 μm, 120 Å pore size, flow rate
1.2 mL min−1 column. Compounds were detected by UV absorption
at 214 nm. Eluents for Waters Delta Prep 4000 system and Analytical
RP-HPLC were as follows: A, water with 0.1% TFA; C, acetonitrile/
water (90/10) with 0.1%TFA; and D, acetonitrile/water (40/60) with
0.1%TFA. Eluents for Waters Prep LC Controller sytem and
Analytical RP-UHPLC were as follows: A, water with 0.1% TFA;
and D, acetonitrile/water (90/10) with 0.1% TFA. NMR spectra were
EXPERIMENTAL SECTION
■
Virtual Screening. Ligand Enumeration. The diamine libraries
were enumerated from GDB-11 or GDB-13 by in-house Java scripts
using ChemAxon's academic JChem API package. In the case of GDB-
11, all diamines with one nonaziridine tertiary amine and a primary or
secondary amine separated by a two-carbon spacer were selected from
GDB-11. A second set of nonaziridine tertiary monoamines of general
formula C10N was extracted from GDB-11, and a primary amine was
added to all a β-carbon relative to the tertiary amine or inserted in a
β−γ single bond relative to the tertiary amine in all possible
combinations. All of the diamines from direct selection and
addition/insertion were pooled, yielding 233572 unique diamines.
These diamines were attached with five different aromatic moieties via
virtual amide bond using ChemAxon's Reactor to form 1167860
complete molecules, which were later used for docking.
In the case of GDB-13, the subset containing only carbon and
nitrogen atoms up to 11 heavy atoms was extracted (3′182′540
SMILES). From this subset, structures containing exactly two nitrogen
atoms, exactly two rings, only single bonds, and only 4−8-membered
rings were retained, resulting in the diamine subset (35′666 SMILES).
All structures were then monoacylated at every primary or secondary
amine, with either 4-chlorobenzoyl or 4-bromophenyloxycarbonyl, in
analogy to the two reference structures 1 and 2, resulting in two sets of
56′788 structures each. The two sets were submitted to ROCS
calculation with 1 and 2 as references, respectively. The number of
structures in both sets was reduced further by applying more restrictive
filters as follows: Structures were only allowed if they contain only 5−
7-membered rings, at most two acyclic carbon atoms, and at least two
bonds in fused rings. Additionally, structures containing either a
primary or a secondary nonacylated amine and structures containing a
terminal carbon atom bound to another carbon atom were removed,
resulting in the final set of 344 structures.
1
acquired on a Bruker 300 MHz or a Bruker 400 MHz instrument. H
and 13C chemical shifts are quoted relative to solvent signals. Carbons
multiplicities have been assigned either by attached proton test (APT),
distorsionless enhancement polarization transfer (DEPT), hetero-
nuclear multiple bond correlation spectroscopy (HMBC), or
heteronuclear single quantum correlation (HSQC) spectroscopy
experiments. Mass spectra were obtained by electron spray ionization
(ES-MS) on a Micromass Autospec Q (Waters/Micromass) instru-
ment in the positive mode.
Amide Formation. Procedure A. The acid (1 equiv), EDC
(1.25 equiv), HOBt (1.25 equiv), N-methyl morpholine (2.5 equiv,
4.5 equiv when the amine was a salt), and the amine (1 equiv) were
dissolved in this order in DCM (1 mL per 1 mmol of acid). The
reaction was stirred for 14 h at room temperature under argon. Once
the reaction was completed, it was evaporated under reduced pressure.
Only for final compounds, the residue was purified using a preparative
RP-HPLC, and only fractions with a purity >95% were lyophilized
(unless otherwise indicated).
Procedure B. The amine (1 equiv) was mixed with the acid
(3 equiv) and PyBOP (3 equiv) in dry DCM (10 mL per 1 mmol of
amine) under an argon atmopshere. Then, diisopropylethylamine
(5 equiv) was added, and the reaction was stirred for 14 h at room
temperature under argon. Once the reaction was completed, it was
evaporated under reduced pressure, and the residue was purified using
a preparative RP-HPLC, and only fractions with a purity >95% were
lyophilized (unless otherwise indicated).
Docking with AutoDock. The protonation state of 72495 randomly
selected compounds was set at pH 7.4. The set was then converted to
507030 stereoisomers using Corina. AutoDock 3.0.5 was applied to
dock the virtual molecules into binding cavity of Ach receptor
1UW6.pdb. The receptor was trimmed to retain only chain C and
chain D and was prepared using AutoDock Tools ADT3 with all
default settings.
Docking with Glide. The same set of 72495 compounds was also
docked using Glide (Schrodinger) docking software. A total of 72495
SMILES were converted to 3D structures using LigPrep. The resulting
.mae structures were then docked into the binding site composed of
4614
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
Carbamate Formation. Procedure C. The diamine (1 equiv) was
dissolved in water (5 mL per 1 mmol), followed by the addition to the
solution of NaOH in pellets (1 equiv). The solution was cooled down
to 0 °C. 4-Bromophenyl chloroformate (1.2 equiv) was then added
dropwise, and after some minutes, a white precipitate started
appearing. After 3 h, the reaction was extracted with CH2Cl2 (3×),
and the organics were dried over Na2SO4, filtered, and evaporated
under reduced pressure. Unless otherwise indicated, the residue was
purified using a preparative RP-HPLC, and only fractions with a purity
>95% were lyophilized.
0.081 mmol, 15% yield). 1H NMR (CDCl3, 300 MHz): δ 10.81 (br s,
1H), 7.50 (br d, J = 10 Hz, 1H), 7.47−7.40 (m, 2H), 7.09−7.01 (m,
2H), 3.85 (qd, J1 = 10 Hz, J2 = 4 Hz, 1H), 3.65 (br d, J = 12 Hz, 1H),
3.48−3.32 (m, 1H), 3.28−2.85 (m, 3H), 2.69−2.52 (m, 1H), 2.42−
2.22 (m, 1H), 2.21−1.74 (m, 8H), 1.60−1.22 (m, 5H). 13C NMR
(CDCl3, 75 MHz): δ 154.6, 150.2, 132.1, 123.6, 118.2, 67.2, 52.5,
50.2, 46.4, 33.6, 24.3, 23.6, 22.8, 22.3, 22.5. MS (ESI+) m/z 381.4
([M + 1, Br79]+), 383.4 ([M + 1, Br81]+). HRMS (ESI+) calcd for
C18H26N2O2Br79, 381.1177 ([M + 1]+]) found 381.1190; mp (°C)
145−146.
Procedure D. To a solution of 4-bromophenylchloroformate
(1.2 equiv) in dry THF (10 mL per 1 mmol) cooled to 0 °C under
argon were added dropwise Et3N(2 equiv) and the diamine (1 equiv)
diluted in THF. The reaction mixture was then stirred for 2 h at room
temperature. The reaction was quenched with brine and extracted with
EtOAc (3×), and the organics layers were dried over Na2SO4, filtered,
and evaporated under reduced pressure. Unless otherwise indicated,
the residue was purified using a preparative RP-HPLC, and only
fractions with a purity >95% were lyophilized.
N-(((R)-1-(Pent-2-ynyl)pyrrolidin-2-yl)methyl)benzo[d]thiazole-6-
carboxamide, TFA Salt (8b). Using procedure A, this product was
obtained from amine 31 as a colorless oil (103 mg, 0.241 mmol, 27%
yield). 1H NMR (CDCl3, 300 MHz): δ 12.22 (br s, 1H), 9.14 (s, 1H),
8.90 (t, J = 6 Hz, 1H), 8.59 (s, 1H), 8.14 (d, J = 8 Hz, 1H), 8.07 (d, J =
8 Hz, 1H), 4.07−3.64 (m, 6H), 3.36−3.21 (m, 1H), 2.34−1.87 (m,
6H), 1.06 (t, J = 7 Hz, 3H). 13C NMR (CDCl3, 75 MHz): δ 167.7 (C),
157.3 (CH), 154.8 (C), 133.8 (C), 130.2 (C), 125.4 (CH), 123.2
(CH), 121.8 (CH), 93.2 (C), 66.6 (C), 65.5 (CH), 53.5 (CH2), 43.9
(CH2), 39.7 (CH2), 27.8 (CH2), 23.2 (CH2), 12.9 (CH3), 12.1 (CH2).
MS (ESI+) m/z 328.1 ([M + 1]+). HRMS (ESI+) calcd for
C18H22N3OS, 328.1483 ([M + 1]+); found, 328.1485.
N-((S)-1-(2-Methylcyclopentyl)pyrrolidin-3-yl)benzo[d]thiazole-6-
carboxamide, TFA Salt (9b). Using procedure A, this product was
obtained from amine (S)-43 as a yellow oil (48 mg, 0.112 mmol, 25%
yield). 1H NMR (CDCl3, 300 MHz) diastereomeric mixture (1.5:1): δ
11.94 (br s, 0.6H), 11.80 (br s, 0.4H), 9.20 (br d, J = 8 Hz, 0.6H), 9.11
(s, 1H), 9.04 (br d, J = 8 Hz, 0.4H), 8.65 (s, 0.6H), 8.58 (d, J = 1 Hz,
0.4H), 8.19−8.04 (m, 2H), 5.28−5.08 (m, 1H), 4.06−3.81 (m, 2H),
3.32−2.87 (m, 3H), 2.70−2.51 (m, 1H), 2.49−1.50 (m, 8H), 1.12 (d,
J = 7 Hz, 1.2H), 1.07 (d, J = 7 Hz, 1.8H). 13C NMR (CDCl3,
75 MHz): δ 166.7 (C), 166.4 (C), 156.8 (C), 155.1 (C), 133.9 (C),
130.5 (C), 125.7 (CH), 125.6 (CH), 123.3 (CH), 121.9 (CH), 71.5
(CH), 71.4 (CH), 60.6 (CH2), 59.6 (CH2), 54.5 (CH2), 52.9 (CH2),
47.9 (CH), 47.6 (CH), 35.0 (CH), 34.9 (CH), 31.3 (CH2), 31.2
(CH2), 30.9 (CH2), 30.7 (CH2), 25.5 (CH2), 25.4 (CH2), 19.8 (CH2),
19.5 (CH2), 14.0 (CH3), 13.6 (CH3). MS (ESI+) m/z 330.2
([M + 1]+). HRMS (ESI+) calcd for C18H24N3OS, 330.1640 ([M + 1]+);
found, 330.1631.
(4-Chlorophenyl)(4-isopropyl-1,4-diazocan-1-yl)methanone (5a).
Using procedure A, this product was obtained from amine 46. The
final compound slowly hydrolyzed in 0.1% TFA (HPLC conditions).
For that reason, purification was done by column chromatography
(neutral alumina, CH2Cl2/MeOH 0.5%) to give 5a as a yellow oil
1
(20 mg, 0.068 mmol, 53% yield). H NMR (CDCl3, 300 MHz): δ
7.33−7.26 (m, 2H), 7.25−7.18 (m, 2H), 3.86 (br s, 1H), 3.40 (br s,
2H), 2.83−2.00 (br m, 6H), 1.70 (br s, 4H), 1.10 (br s, 6H). 13C
NMR (CDCl3, 75 MHz): δ 170.5 (C), 135.8 (C), 135.0 (C), 128.6
(CH), 127.6 (CH), 54.8 (br CH2), 54.1 (CH), 50.1 (br CH2), 39.8
(br CH2), 23.4 (CH2), 20.9 (CH3). MS (ESI+) m/z 295.4 ([M + 1,
Cl35]+), 297.2 ([M + 1, Cl37]+). HRMS (ESI+) calcd for
C16H24Cl35N2O, 295.1572 ([M + 1]+); found, 295.1566.
(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)(4-isopropyl-1,4-diazocan-
1-yl)methanone (5c). Using procedure A, this product was obtained
from 1-isopropyl-1,4-diazocane 46, prepared by a literature proce-
dure.45 The final compound slowly hydrolyzed in 0.1% TFA (HPLC
conditions) For that reason, purification was done by column
chromatography (neutral alumina, CH2Cl2/MeOH 0.5%) to give 5c
as a yellow oil (15 mg, 0.047 mmol, 37% yield). 1H NMR (CDCl3, 300
MHz): δ 6.98−6.79 (m, 3H), 4.25 (s, 4H), 4.11 (br s, 1H), 3.44 (br t,
2H), 2.91−2.27 (br m, 6H), 1.75 (br s, 4H), 1.15 (br d, J = 6 Hz, 6H).
13C NMR (CDCl3, 75 MHz): δ 171.0 (C), 144.4 (C), 143.3 (C),
130.6 (C), 130.5 (C), 119.7 (CH), 117.2 (CH), 115.7 (CH), 64.4
(CH2), 64.3 (CH2), 54.9 (br CH2), 54.1 (CH), 50.0 (br CH2), 39.9
(br CH2), 23.4 (CH2), 21.0 (CH3). MS (ESI+) m/z 319.0 ([M + 1]+).
HRMS (ESI+) calcd for C18H27N2O3, 319.2016 ([M + 1]+); found,
319.2012.
4-Chloro-N-((hexahydro-1H-pyrrolizin-7a-yl)methyl)benzamide
(6a). Using procedure A, this product was obtained from amine 63.
Purification by column chromatography (neutral alumina, CH2Cl2/
MeOH 2%) gave 6a as a yellow solid (40 mg, 0.143 mmol, 40% yield).
1H NMR (CDCl3, 300 MHz): δ 8.12 (br s, 1H), 7.91 (d, J = 8 Hz,
2H), 7.40 (d, J = 8 Hz, 2H), 3.63 (d, J = 6 Hz, 2H), 3.44−3.27 (m,
2H), 2.89−2.73 (m, 2H), 2.11−1.70 (m, 8H). 13C NMR (CDCl3, 75
MHz): δ 166.6, 137.7, 132.0, 128.8, 128.7, 78.0, 55.4, 45.6, 35.7, 24.7.
MS (ESI+) m/z 279.2 ([M + 1, Cl35]+), 2281.2 ([M + 1, Cl37]+).
HRMS (ESI+) calcd for C15H20Cl35N2O, 279.1264 ([M + 1]+); found,
279.1260; mp (°C) 158−163.
N-((Hexahydro-1H-pyrrolizin-7a-yl)methyl)benzamide, TFA Salt
(6f). Using procedure A, the compound was obtained from amine 63
as a colorless oil (22.7 mg, 0.06 mmol, 30% yield). 1H NMR (MeOD,
300 MHz): δ 7.90 (dd, J1 = 3 Hz, J2 = 9 Hz, 2H), 763−7.58 (m, 1H),
7.54−7.48 (m, 2H), 3.74 (s, 2H), 3.69−3.61 (m, 2H), 3.28−3.22 (m,
2H), 2.27−2.24 (m, 8H). 13C NMR (MeOD, 75 MHz): δ 173.4,
133.9, 133.8, 130.0, 128.8, 84.5, 56.7, 47.6, 35.9, 25.5. MS (ESI+) m/z
245.1 ([M + 1]+). HRMS (ESI+) calcd for C15H21N2O, 245.1648
([M + 1]+); found, 245.1654.
((3S,9aR)-Hexahydro-3-methyl-1H-pyrido[1,2-a]pyrazin-2(6H)-
yl)(2,3-dihydrobenzo[b][1,4]dioxin-6-yl)methanone, TFA Salt (10c).
Using procedure A, this product was obtained from amine 48 as a
1
colorless oil (22 mg, 0.053 mmol, 41% yield). H NMR (D2O, 400
MHz at 353 K): δ 7.63−7.45 (m, 3H), 5.29 (br s, 1H), 4.89 (s, 2H),
4.88 (s, 2H), 4.63−4.38 (m, 2H), 4.29−3.80 (m, 4H), 3.69−3.55 (m,
1H), 2.59−1.93 (m, 9H). 13C NMR (D2O, 100 MHz at 353 K): δ
120.9 (CH), 118.3 (CH), 116.4 (CH), 65.4 (CH2), 65.2 (CH2), 57.4
(CH), 54.0 (CH2), 47.8 (CH2), 46.2 (br CH), 42.7 (br CH2), 21.7
(CH2), 21.3 (CH2), 17.4 (CH2), 15.3 (CH3). MS (ESI+) m/z 317.4
([M + 1]+). HRMS (ESI+) calcd for C18H25N2O3, 317.1865 ([M +
1]+); found, 317.1860.
(S)-4-Bromophenyl Hexahydro-1H-pyrrolo[1,2-a][1,4]diazepine-
2(3H)-carboxylate, TFA Salt [(S)-11e]. Using procedure C, this
product was obtained from amine 51 as a colorless oil (26 mg, 0.059
1
mmol, 8% yield). H NMR (D2O, 400 MHz): δ 7.66 (d, J = 9 Hz,
2H), 7.14 (d, J = 7 Hz, 2H), 4.40−3.56 (m, 7H), 3.47−3.20 (m, 2H),
2.54−1.85 (m, 6H). 13C NMR (D2O, 75 MHz): mixture of four
isomers δ 155.8, 155.6, 155.3, 155.0, 149.8, 149.7, 149.6, 132.6, 123.7,
118.8, 118.7, 118.7, 67.0, 66.9, 66.2, 65.5, 57.2, 57.1, 56.9, 56.8, 53.7,
53.6, 51.4, 51.3, 47.3, 47.0, 46.9, 46.8, 46.2, 43.9, 43.7, 28.1, 28.0, 27.7,
27.6, 26.2, 26.0, 25.2, 25.0, 23.3, 21.7, 21.6. MS (ESI+): m/z 339.2 ([M
+ 1, Br79]+), 341.2 ([M + 1, Br81]+). HRMS (ESI+) calcd for
C15H20Br79N2O2, 339.0708 ([M + 1]+); found, 339.0698.
N-(1-Aza-bicyclo[2.2.2]oct-2-ylmethyl)-4-fluorobenzamide, TFA
Salt (rac-12k). Compound 26k (79 mg, 0.30 mmol) and Pd/C
(40 mg) were diluted in dry MeOH (6 mL), and the solution was
stirred for 4 h under an hydrogen atmosphere. The reaction mixture
was then filtered over Celite and washed with MeOH. The filtrate was
concentrated under vacuum to afford a yellow oil, which was purified
by preparative HPLC to give the desired compound as a colorless oil
4-bromophenyl-2-(piperidin-1-yl)cyclohexylcarbamate, TFA salt (7e).
Using procedure C, this product was obtained from the corresponding
amine which was prepared according ref 40 as a white solid (39 mg,
4615
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
(3 mg, 3%). 1H NMR (CDCl3, 300 MHz): δ 8.69 (bs, 1H), 7.91 (dd,
J1 = 9 Hz, J2 = 6 Hz, 2H), 7.06 (t, J = 9 Hz, 2H), 4.02 (ddd, J1 = 15 Hz,
J2 =7 Hz, J3 = 3 Hz, 1H), 3.65 (q, J = 9 Hz, 1H), 3.55−3.19 (m, 5H),
2.27−2.13 (m, 2H), 2.01−1.84 (m, 4H), 1.53−1.46 (m, 1H). 13C
NMR (CDCl3, 75 MHz): δ 167.3, 165.3 (d, J = 255 Hz, 1C), 130.1
(d, J = 15 Hz, 2C), 129.5, 115.7 (d, J = 15 Hz, 2C), 59.1, 48.6, 41.2,
40.5, 28.8, 23.7, 22.8, 20.7. 19F NMR (CDCl3, 376 MHz): δ −109.5,
−107.9. MS (ESI+): m/z 263.1 ([M + 1]+). HRMS (ESI+) calcd for
C15H20N2OF, 263.1554 ([M + 1]+); found, 263.1554.
ACKNOWLEDGMENTS
■
This work was supported financially by the University of Berne,
the Swiss National Science Foundation. N.G.-D. is a
postdoctoral fellow of the Roche Research Foundation. This
work was supported by the EEC grant Neurocypres to D.B. We
thank E. Neveu and M. Maver for technical assistance in some
of the experiments and preparation of the figures.
N-((1-Aza-bicyclo[2.2.2]oct-2-en-2-yl)methyl)-4-fluorobenzamide
(26k). Using procedure A, this compound was obtained from amine 54
as yellowish solid after filtration on neutral alumina (DCM) and
ABBREVIATIONS USED
■
GDB, Chemical Universe Generated Database; nAChR,
nicotinic acetylcholine receptor; AChBP, acetylcholine binding
protein; NMDA, N-methyl D-aspartate; ROCS, rapid overlay of
chemical structures; ZINC, zinc is not commercial
1
crystallization in hexanes (50 mg, 0.19 mmol, 38% yield). H NMR
(CDCl3, 300 MHz): δ 7.85 (dd, J = 6 Hz, 2H), 7.37 (bs, 1H), 7.05 (t,
J = 9 Hz, 2H), 6.43 (d, J = 6 Hz, 1H), 4.10 (d, J = 6 Hz, 2H), 3.11−
3.02 (m, 2H), 2.71−2.57 (m, 3H), 1.72−1.62 (m, 2H), 1.54−1.42 (m,
2H). 13C NMR (CDCl3, 75 MHz): δ 166.2, 164.9 (d, J = 247 Hz, 1C),
146.1, 130.6 (d, J = 7.5 Hz, 1C), 129.9, 129.7 (d, J = 7.5 Hz, 2C),
115.7(d, J = 15 Hz, 2C), 49.5, 41.0, 27.4, 26.9. 19F NMR (CDCl3, 376
MHz): δ −108.3 MS (ESI+) m/z 261.1 ([M + 1]+). HRMS (ESI+)
calcd for C15H18FN2O, 261.1398 ([M + 1]+); found, 261.1390.
2-Aminomethyl-1-azabicyclo[2.2.2]oct-2-ene (54). 2-Chlorometh-
yl-1-azabicyclo[2.2.2]oct-2-ene 53 (798 mg, 5.1 mmol), prepared by a
literature procedure,48 was diluted in MeOH (20 mL), and an aqueous
NH3 solution (20 mL) was added. The reaction mixture was stirred at
room temperature for 4 h, and the mixture was concentrated to
dryness to afford the desired compound in a quantitative yield. This
material was used in the next step without purification. 1H NMR
(MeOD, 300 MHz): δ 6.63 (d, J = 9 Hz, 1H), 3.53 (d, J = 3 Hz, 2H),
3.08−2.99 (m, 2H), 2.73−2.69 (m, 1H), 2.66−2.56 (m, 2H), 1.78−
1.69 (m, 2H), 1.53−1.43 (m, 2H). 13C NMR (MeOD, 75 MHz): δ
145.5, 134.5, 51.5, 50.2, 41.8, 28.7, 28.4, 25.1. MS (ESI+) m/z 139.1
[M + 1]+. HRMS (ESI+) calcd for C8H15N2, 139.1230; found,
139.1233.
Electrophysiology. Recordings were performed as previously
described.49 Briefly, oocytes were prepared using standard methods by
mechanical and enzymatical dissociation. Stage 5−6 oocytes were selected
manually under the microscope and 10 nL of water containing 0.2 μg/μL
of plasmid of interest, which contained the cDNA sequence for the
nAChR. Expression of α3β2, α4β2, α3β4, and α4β4 was obtained by
injection of 10 nL of solution containing a 1:1 ratio of the respective
subtype containing plasmid at 0.2 μg/μL final concentration. Injections
were performed using an automated system (roboinject, Multichannel
Systems, Germany). Two or more days later, the functional properties of
receptors were evaluated using an automated system equipped with two
electrode voltage clamp (HiClamp, Multichannel Systems, Germany).
The membrane potential of oocytes was clamped at a steady value of −80
or −100 mV, and currents evoked by ACh or compounds were recorded
with a sampling frequency of 100 Hz. All recordings were performed at 18
°C, and cells were superfused with OR2 medium containing in mM:
NaCl, 82.5; KCl, 2.5; HEPES, 5; CaCl2·2H2O, 1.8; and MgCl2·6H2O, 1;
pH 7.4.
REFERENCES
■
(1) D'Hoedt, D.; Bertrand, D. Nicotinic acetylcholine receptors: an
overview on drug discovery. Expert Opin. Ther. Targets 2009, 13, 395−
411.
(2) Warr, W. A. ChEMBL. An interview with John Overington, team
leader, chemogenomics at the European Bioinformatics Institute
Outstation of the European Molecular Biology Laboratory (EMBL-
EBI). J. Comput.-Aided Mol. Des. 2009, 23, 195−198.
(3) Lederberg, J.; Sutherland, G. L.; Buchanan, B. G.; Feigenbaum, E.
A.; Robertson, A. V.; Duffield, A. M.; Djerassi, C. Applications of
artificial intelligence for chemical inference. I. Number of possible
organic compounds. Acyclic structures containing carbon, hydrogen,
oxygen, and nitrogen. J. Am. Chem. Soc. 1969, 91, 2973−2976.
(4) Wermuth, C. G. Similarity in drugs: reflections on analogue
design. Drug Discovery Today 2006, 11, 348−354.
(5) Bodnar, A. L.; Cortes-Burgos, L. A.; Cook, K. K.; Dinh, D. M.;
Groppi, V. E.; Hajos, M.; Higdon, N. R.; Hoffmann, W. E.; Hurst, R.
S.; Myers, J. K.; Rogers, B. N.; Wall, T. M.; Wolfe, M. L.; Wong, E.
Discovery and structure-activity relationship of quinuclidine benza-
mides as agonists of alpha7 nicotinic acetylcholine receptors. J. Med.
Chem. 2005, 48, 905−908.
(6) Biton, B.; Bergis, O. E.; Galli, F.; Nedelec, A.; Lochead, A. W.;
Jegham, S.; Godet, D.; Lanneau, C.; Santamaria, R.; Chesney, F.;
Leonardon, J.; Granger, P.; Debono, M. W.; Bohme, G. A.; Sgard, F.;
Besnard, F.; Graham, D.; Coste, A.; Oblin, A.; Curet, O.; Vige, X.;
Voltz, C.; Rouquier, L.; Souilhac, J.; Santucci, V.; Gueudet, C.;
Francon, D.; Steinberg, R.; Griebel, G.; Oury-Donat, F.; George, P.;
Avenet, P.; Scatton, B. SSR180711, a novel selective alpha7 nicotinic
receptor partial agonist: (1) binding and functional profile. Neuro-
psychopharmacology 2007, 32, 1−16.
(7) Garcia-Delgado, N.; Bertrand, S.; Nguyen, K. T.; van Deursen, R.;
Bertrand, D.; Reymond, J.-L. Exploring a7-Nicotinic Receptor Ligand
Diversity by Scaffold Enumeration from the Chemical Universe
Database GDB. ACS Med. Chem. Lett. 2010, 1, 422−426.
(8) Schneider, G.; Hartenfeller, M.; Reutlinger, M.; Tanrikulu, Y.;
Proschak, E.; Schneider, P. Voyages to the (un)known: Adaptive
design of bioactive compounds. Trends Biotechnol. 2009, 27, 18−26.
(9) Reymond, J. L.; Van Deursen, R.; Blum, L. C.; Ruddigkeit, L.
Chemical space as a source for new drugs. Med. Chem. Commun. 2010,
1, 30−38.
ASSOCIATED CONTENT
■
S
* Supporting Information
(10) Weininger, D. Smiles, a Chemical Language and Information-
System. 1. Introduction to Methodology and Encoding Rules. J. Chem.
Inf. Comput. Sci. 1988, 28, 31−36.
(11) Bartoloni, M.; Kadam, R. U.; Schwartz, J.; Furrer, J.; Darbre, T.;
Reymond, J. L. Expanding the accessible chemical space by solid phase
synthesis of bicyclic homodetic peptides. Chem. Commun. 2011, 47,
12634−12636.
(12) Fink, T.; Bruggesser, H.; Reymond, J. L. Virtual exploration of
the small-molecule chemical universe below 160 Da. Angew. Chem., Int.
Ed. Engl. 2005, 44, 1504−1508.
(13) Fink, T.; Reymond, J. L. Virtual exploration of the chemical
universe up to 11 atoms of C, N, O, F: Assembly of 26.4 million
structures (110.9 million stereoisomers) and analysis for new ring
Details of the synthesis and characterization including spectra
of all synthesized compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Fax: +41 31 631 80 57. E-mail: jean-louis.reymond@ioc.unibe.
ch.
Notes
The authors declare no competing financial interest.
4616
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
systems, stereochemistry, physicochemical properties, compound
classes, and drug discovery. J. Chem. Inf. Model. 2007, 47, 342−353.
(14) Blum, L. C.; Reymond, J. L. 970 million druglike small
molecules for virtual screening in the chemical universe database
GDB-13. J. Am. Chem. Soc. 2009, 131, 8732−8733.
(15) Nguyen, K. T.; Syed, S.; Urwyler, S.; Bertrand, S.; Bertrand, D.;
Reymond, J. L. Discovery of NMDA glycine site inhibitors from the
chemical universe database GDB. ChemMedChem 2008, 3, 1520−
1524.
(16) Nguyen, K. T.; Luethi, E.; Syed, S.; Urwyler, S.; Bertrand, S.;
Bertrand, D.; Reymond, J. L. 3-(aminomethyl)piperazine-2,5-dione as
a novel NMDA glycine site inhibitor from the chemical universe
database GDB. Bioorg. Med. Chem. Lett. 2009, 19, 3832−3835.
(17) Luethi, E.; Nguyen, K. T.; Burzle, M.; Blum, L. C.; Suzuki, Y.;
Hediger, M.; Reymond, J. L. Identification of selective norbornane-
type aspartate analogue inhibitors of the glutamate transporter 1
(GLT-1) from the chemical universe generated database (GDB). J.
Med. Chem. 2010, 53, 7236−7250.
(18) Taly, A.; Corringer, P. J.; Guedin, D.; Lestage, P.; Changeux, J.
P. Nicotinic receptors: allosteric transitions and therapeutic targets in
the nervous system. Nat. Rev. Drug Discovery 2009, 8, 733−750.
(19) Paleari, L.; Cesario, A.; Fini, M.; Russo, P. alpha7-Nicotinic
receptor antagonists at the beginning of a clinical era for NSCLC and
Mesothelioma? Drug Discovery Today 2009, 14, 822−36.
(20) Bartos, M.; Price, K. L.; Lummis, S. C.; Bouzat, C. Glutamine 57
at the complementary binding site face is a key determinant of
morantel selectivity for α7 nicotinic receptors. J. Biol. Chem. 2009, 284,
21478−21487.
(21) Seo, S.; Henry, J. T.; Lewis, A. H.; Wang, N.; Levandoski, M. M.
The positive allosteric modulator morantel binds at noncanonical
subunit interfaces of neuronal nicotinic acetylcholine receptors. J.
Neurosci. 2009, 29, 8734−8742.
(22) Wu, T. Y.; Smith, C. M.; Sine, S. M.; Levandoski, M. M.
Morantel allosterically enhances channel gating of neuronal nicotinic
acetylcholine alpha 3 beta 2 receptors. Mol. Pharmacol. 2008, 74, 466−
475.
(23) Cesa, L. C.; Higgins, C. A.; Sando, S. R.; Kuo, D. W.;
Levandoski, M. M. Specificity Determinants of Allosteric Modulation
in the Neuronal Nicotinic Acetylcholine Receptor: A Fine Line
between Inhibition and Potentiation. Mol. Pharmacol. 2012, 81, 239−
249.
(24) Walker, D. P.; Wishka, D. G.; Piotrowski, D. W.; Jia, S.; Reitz, S.
C.; Yates, K. M.; Myers, J. K.; Vetman, T. N.; Margolis, B. J.; Jacobsen,
E. J.; Acker, B. A.; Groppi, V. E.; Wolfe, M. L.; Thornburgh, B. A.;
Tinholt, P. M.; Cortes-Burgos, L. A.; Walters, R. R.; Hester, M. R.;
Seest, E. P.; Dolak, L. A.; Han, F.; Olson, B. A.; Fitzgerald, L.; Staton,
B. A.; Raub, T. J.; Hajos, M.; Hoffmann, W. E.; Li, K. S.; Higdon, N.
R.; Wall, T. M.; Hurst, R. S.; Wong, E. H.; Rogers, B. N. Design,
synthesis, structure-activity relationship, and in vivo activity of
azabicyclic aryl amides as alpha7 nicotinic acetylcholine receptor
agonists. Bioorg. Med. Chem. 2006, 14, 8219−8248.
(25) Ulens, C.; Akdemir, A.; Jongejan, A.; van Elk, R.; Bertrand, S.;
Perrakis, A.; Leurs, R.; Smit, A. B.; Sixma, T. K.; Bertrand, D.; de Esch,
I. J. Use of acetylcholine binding protein in the search for novel alpha7
nicotinic receptor ligands. In silico docking, pharmacological screen-
ing, and X-ray analysis. J. Med. Chem. 2009, 52, 2372−2383.
(26) Reymond, J. L.; van Deursen, R.; Bertrand, D. What we have
learned from crystal structures of proteins to receptor function.
Biochem. Pharmacol. 2011, 82, 1521−1527.
(27) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart,
W. E.; Belew, R. K.; Olson, A. J. Automated docking using a
Lamarckian genetic algorithm and an empirical binding free energy
function. J. Comput. Chem. 1998, 19, 1639−1662.
(28) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.;
Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.;
Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: A new
approach for rapid, accurate docking and scoring. 1. Method and
assessment of docking accuracy. J. Med. Chem. 2004, 47, 1739−1749.
(29) Sadowski, J.; Gasteiger, J. From Atoms and Bonds to 3-
Dimensional Atomic Coordinates - Automatic Model Builders. Chem.
Rev. 1993, 93, 2567−2581.
(30) Nguyen, K. T.; Blum, L. C.; van Deursen, R.; Reymond, J. L.
Classification of organic molecules by molecular quantum numbers.
ChemMedChem 2009, 4, 1803−1805.
(31) van Deursen, R.; Blum, L. C.; Reymond, J. L. A searchable map
of PubChem. J. Chem. Inf. Model. 2010, 50, 1924−1934.
(32) van Deursen, R.; Blum, L. C.; Reymond, J. L. Visualisation of
the chemical space of fragments, lead-like and drug-like molecules in
PubChem. J. Comput.-Aided Mol. Des. 2011, 25, 649−662.
(33) Blum, L. C.; van Deursen, R.; Reymond, J. L. Visualisation and
subsets of the chemical universe database GDB-13 for virtual
screening. J. Comput.-Aided Mol. Des. 2011, 25, 637−647.
(34) Blum, L. C.; van Deursen, R.; Bertrand, S.; Mayer, M.; Burgi, J.
J.; Bertrand, D.; Reymond, J. L. Discovery of alpha7-Nicotinic
Receptor Ligands by Virtual Screening of the Chemical Universe
Database GDB-13. J. Chem. Inf. Model. 2011, 51, 3105−3112.
(35) Rush, T. S., 3rd; Grant, J. A.; Mosyak, L.; Nicholls, A. A shape-
based 3-D scaffold hopping method and its application to a bacterial
protein-protein interaction. J. Med. Chem. 2005, 48, 1489−1495.
(36) Hawkins, P. C.; Skillman, A. G.; Nicholls, A. Comparison of
shape-matching and docking as virtual screening tools. J. Med. Chem.
2007, 50, 74−82.
(37) Suzuki, T.; Oka, M.; Maeda, K.; Furusawa, K.; Mitani, T.;
Kataoka, T. N-[2-(1-azabicyclo[3.3.0]octan-5-yl)ethyl]-2-nitroaniline,
a potent muscarinic agonist. Chem. Pharm. Bull. 1997, 45, 1218−1220.
(38) Oka, M.; Baba, K.; Nakamura, K.; Dong, L. L.; Hamajima, H.;
Unno, R.; Matsumoto, Y. Synthesis of 1-azabicyclic systems by double
cyclization. J. Heterocycl. Chem. 2003, 40, 177−180.
(39) Paventi, M.; Edward, J. T. Preparation of Alpha-Amino-
thioamides from Aldehydes. Can. J. Chem. 1987, 65, 282−289.
́
(40) Gonzalez-Sabín, J.; Gotor, V.; Rebolledo, F. Chemoenzymatic
Preparation of Optically Active trans-Cyclohexane-1,2-diamine De-
rivatives: An Efficient Synthesis of the Analgesic U-(−)-50,488.
Chem.Eur. J. 2004, 10, 5788−5794.
(41) Khan, M. O. F.; Levi, M. S.; Tekwani, B. L.; Wilson, N. H.;
Borne, R. F. Synthesis of isoquinuclidine analogs of chloroquine:
Antimalarial and antileishmanial activity. Bioorg. Med. Chem. 2007, 15,
3919−3925.
(42) Du Bois, D. J.; John, K. D.; Brian, S. E.; Bernard, S. D.; Beihan,
W. N-ureido-piperidines as antagonists VIII for CCR-3 receptor.
Patent WO/2003/045937, 2003.
(43) Tanimori, S.; Sunami, T.; Fukubayashi, K.; Kirihata, M. An
efficient construction of bridged chiral tetracyclic indolidines, a core
structure of asperparaline, via stereocontrolled catalytic Pauson-Khand
reaction. Tetrahedron 2005, 61, 2481−2492.
(44) Micovic, V. M.; Mihailovic, M. L. The Reduction of Acid
Amides with Lithium Aluminum Hydride. J. Org. Chem. 1953, 18,
1190−1200.
(45) Majchrzak, M.; Kotelko, A.; Guryn, R. Derivatives of
Octahydrodiazocine-1,5 and Octahydrodiazocine-1,4 with Potential
Pharmacological Action. 1. Synthesis of N-Alkyl Derivatives of
Octahydrodiazocine-1,5 and Octahydrodiazocine-1,4. Acta Pol.
Pharm. 1975, 32, 145−148.
(46) Bradsher, C. K.; Telang, S. A. 2-Azaquinolizinium Oxides. J. Org.
Chem. 1966, 31, 941−943.
(47) Guryn, R. Synthesis of 1,5-Diazabicyclo[5.3.0]Decane and 1,5-
Diazabicyclo[5.4.0]Undecane. Pol. J. Chem. 1985, 59, 1243−1246.
(48) Rosen, T.; Guarino, K. J. Kinetically-Controlled Displacement
by Azide on an Allylic ChlorideSynthesis of a Highly Potent
Serotonin-3 Receptor Ligand Prototype. Tetrahedron 1991, 47, 5391−
5400.
(49) Hogg, R. C.; Bandelier, F.; Benoit, A.; Dosch, R.; Bertrand, D.
An automated system for intracellular and intranuclear injection. J
Neurosci. Methods 2008, 169, 65−75.
(50) Bertrand, D.; Gopalakrishnan, M. Allosteric modulation of
nicotinic acetylcholine receptors. Biochem. Pharmacol. 2007, 74, 1155−
1163.
4617
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618

Journal of Medicinal Chemistry
Article
(51) Cachelin, A. B.; Rust, G. Unusual pharmacology of
(+)-tubocurarine with rat neuronal nicotinic acetylcholine receptors
containing beta 4 subunits. Mol. Pharmacol. 1994, 46, 1168−1174.
(52) Irwin, J. J.; Shoichet, B. K. ZINCA free database of
commercially available compounds for virtual screening. J. Chem. Inf.
Model. 2005, 45, 177−182.
4618
dx.doi.org/10.1021/jm300030r | J. Med. Chem. 2012, 55, 4605−4618