Synthesis, binding, and modeling studies of new cytisine derivatives, as ligands for neuronal nicotinic acetylcholine receptor subtypes

Article abstract of DOI:10.1021/jm900225j

The availability of drug affecting neuronal nicotinic acetylcholine receptors (nAChRs) may have important therapeutic potential for the treatment of several CNS pathologies. Pursuing our efforts on the systematic structural modification of cytisine and N-

Full text of DOI:10.1021/jm900225j

J. Med. Chem. 2009, 52, 4345–4357 4345
DOI: 10.1021/jm900225j
Synthesis, Binding, and Modeling Studies of New Cytisine Derivatives, as Ligands for Neuronal
Nicotinic Acetylcholine Receptor Subtypes
Bruno Tasso,† Caterina Canu Boido,† Emanuela Terranova,† Cecilia Gotti,‡ Loredana Riganti,‡ Francesco Clementi,‡
Roberto Artali,§ Gabriella Bombieri,§ Fiorella Meneghetti,§ and Fabio Sparatore*,†
†Dipartimento di Scienze Farmaceutiche, Universitaꢀ degli Studi di Genova, Viale Benedetto XV 3, 16139 Genova, Italy, ‡Dipartimento di
Farmacologia “E. Trabucchi”, Universitaꢀ degli Studi di Milano, via Vanvitelli 32, 20129 Milano, Italy, Istituto di Neuroscienze del CNR,
via Vanvitelli 32, 20129 Milano, Italy, and Dipartimento di Scienze Farmaceutiche “P. Pratesi”, Universitaꢀ degli Studi di Milano,
§
Via Mangiagalli 25, 20133 Milano, Italy
Received February 20, 2009
The availability of drug affecting neuronal nicotinic acetylcholine receptors (nAChRs) may have
important therapeutic potential for the treatment of several CNS pathologies. Pursuing our efforts
on the systematic structural modification of cytisine and N-arylalkyl and N-aroylalkyl cytisines were
synthesized and tested for the displacement of [3H]-epibatidine and [125I]-R-bungarotoxin from the most
widespread brain nAChRs subtypes R4β2 and R7, respectively. While the affinity for R7 subtype was
rather poor (Ki from 0.4 to >50 μM), the affinity for R4β2 subtype was very interesting, with nanomolar
Ki values for the best compounds. The N-substituted cytisines were docked into the rat and human R4β2
nAChR models based on the extracellular domain of a molluscan acetylcholine binding protein. The
docking results agreed with the binding data, allowing the detection of discrete aminoacid residues of the
R and β subunits essential for the ligand binding on rat and human nAChRs, providing a novel
structural framework for the development of new R4β2 selective ligands.
Introduction
large (∼200 amino acids) N-terminal hydrophilic domain
containing the multiple loops of the neurotransmitter binding
site, (2) the highly variable C-terminal hydrophilic domain
that is extracellular, and (3) a set of four closely spaced
transmembrane domains (termed M1-M4) immediately
following the large extracellular domain. The M2 domain is
believed to form the wall of the ion channel. The loops
that comprise the agonist binding site contain conserved
residues, many of which possess aromatic side chains
(Trp and Tyr), which are proposed to make cationic-π
interactions with agonists.1
The achievement of subtype selective agents that can bind
and modify the function of nAChRs has been attempted
mainly by structural modification and synthesis of analogues
of nicotine and epibatidine, two very potent natural agonists
(Chart 1). A large number of compounds have been prepared
and tested,6-8 but only a few of them exhibited promising
characteristics, particularly tebanicline (5-[(2R)-2-azetidinyl-
methoxy]-2-chloropyridine; ABT-594) as an analgesic9 and
ABT-418((S)-3-methyl-5-(1-methyl-2-pyrrolidinyl)isoxazole)
as a cognition enhancing agent.10 However, for both com-
pounds, the clinical trials have been discontinued due to
adverse effects. Another important nicotinic agonist, which
could disclose new perspective and opportunities for develop-
ing new agonists and/or antagonists, is represented by cytisine
(1), an alkaloid mainly obtained from seeds of Laburnum
anagyroides but present also in several other plants of the
Leguminosae family, to which it confers intoxicant properties.
Until recently, Texas Mountain Laurel, also known as mescal
beans (Sophora secundiflora), was believed to have been used
as a divinatory medium even prior to the discovery of peyote
by many American Indian tribes of the South West.11 Despite
the name, mescal beans do not contain mescaline but cytisine
Neuronal nicotinic acetylcholine receptors (nAChRsa)
form a family of pentameric ACh-gated cation channel,
made up of different subtypes, each of which has a
specific pharmacology, physiology, and anatomical distribu-
tion in brain and ganglia.1,2 They are widely distributed in
peripheral and central nervous systems, where they act as
postsynaptic receptors exciting neurons, or as presynaptic
receptors modulating the release of many neurotransmitters.
Neuronal nAChRs are involved in complex cerebral processes
as learning, memory consolidation, nociception, locomotor
activity, as well as in a growing number of degenerative
diseases (Alzheimer’s and Parkinson’s diseases) and nervous
pathologies such as autism, ADHD, anxiety, and schizo-
phrenia. Thus they are interesting targets for the treatment
of a variety of CNS disorders, particularly Alzheimer’s
and Parkinson’s diseases, opiate resistant chronic pain, and
tobacco smoking addiction, which may involve specific
nicotinic receptor subtypes, among which R4β2 is the
most abundant in the brain.3-5 Each of the nAChR subunits
displays a characteristic phenotype of structural features
extending from the N-terminus to the C-terminus: (1) a
*To whom correspondence should be addressed. Phone: þ39-010-
3538359. Fax: þ39-010-3538358. E-mail: sparator@unige.it.
aAbbreviations: AA, amino acid; ACh, acetylcholine; AChBP, acet-
ylcholine binding protein; ADHD, attention deficit hyperactivity dis-
order; AMBER, assisted model building with energy refinement; CC,
column chromatography; CNS, central nervous system; CPK, Corey-
Pauling-Koltun; CV, coefficient of variation; MM, molecular mechan-
ical; nAChRs, nicotinic acetylcholine receptors; NPT, normal pressure
and temperature; PDB, protein data bank; QM, quantum mechanical;
rmsd, root-mean-square deviation; SMP CPU, symmetric multiproces-
sing central processing unit; TRIS HCl, tris(hydroxymethyl)amino-
methane hydrochloride.
3
r
2009 American Chemical Society
Published on Web 06/23/2009
pubs.acs.org/jmc

4346 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Tasso et al.
Chart 1
through homology modeling techniques based on the
X-ray structure of the Aplysia californica AChBP (PDP
code: 2BYQ).32,33
Results and Discussion
Chemistry. The cytisine derivatives 2, 3, 6, 9, 12, 19, 23, 27,
and 29 were already described by Canu Boido and Spara-
tore.21 The novel N-arylalkyl and aroylalkyl derivatives 7, 8,
13, 14, 16-18, 20-22, and the N-(4-oxopentyl)cytisine 28
were obtained by reacting cytisine with the suitable haloder-
ivatives in a ratio 2:1. For preparation of N-[(pyridin-3/4-yl)-
methyl]- and N-[3-(pyridin-3-yl)propyl]cytisines (4, 5, 15),
an excess of cytisine was reacted with the relevant bromoalk-
ylpyridine hydrobromide. It is worth noting that it was
impossibile to obtain the N-[(pyridin-2-yl)methyl]cytisine
because the 2-bromomethylpyridine reacted with itself as
soon it was liberated from the hydrobromide.
To obtain compound 24, cytisine was reacted with iodoa-
cetamide in the presence of anhydrous K2CO3. The prepara-
tion of 2- and 4-pyridylethyl (10, 11), carbamoylethyl (25),
and methylsulfonylethyl-derivative (26) was effected
through a Michael addition of cytisine on the corresponding
unsaturated compounds (2- and 4-vinylpyridine, acryla-
mide, and methyl-vinylsulfone) (Scheme 1). The structures
of the prepared compounds were supported by elemental
analyses and spectral data.
In Vitro Receptor Binding. The results of the binding
assay of cytisine (1) and its N-substituted derivatives
2-29 on R4β2 and R7 rat receptor subtypes are collected in
Table 1. Most compounds exhibited very poor affinity to
R7 subtype, often more than 100-fold lower than cytisine,
with Ki > 50000 nM. Only two compounds (20, 21) showed a
Ki value rather close to that of cytisine (Ki = 331 nM). On
the other hand, according to their affinity toward R4β2 sub-
type, the tested compounds can be grouped in three clusters:
(a) eight compounds with high affinity (Ki = 2.6-63 nM), (b)
13 with moderate affinity (Ki = 330-965 nM), and (c) seven
with poor affinity (Ki = 1300-13000 nM).
Binding studies for cytisine and 12 of its derivatives
were also performed in rat cortex R4β2 receptor subtype
labeled with [3H]-cytisine. The observed Ki values (see
Supporting Information) are well concordant with those of
Table 1 ([3H]-epibatidine displacement).
and other related lupine alkaloids. Cytisine showed very high
affinity for nAChRs12 and is able to discriminate among
some subtypes, with a higher affinity for R4β2 than R3β4
subtype.13 Preference for R4β2 versus R3β4 subtype has been
observed also for N-substituted cytisine derivatives.14
Alhough the pharmacological profile of cytisine has
been thoroughly studied,15 this alkaloid has not found any
therapeutic application in western countries, while in the
former Soviet Union it was preferred to lobeline as a respira-
tory analeptic16 and was used to treat tobacco dependence
for the last 40 years in several East European Countries.17 Only
recently cytisine has received some attention as a lead for
structural analogues or derivatives able to interact, directly
or allosterically, with one or more nAChR subtypes. Four
patents18-20 suggested hypoglycemic, anti-inflammatory,
antiaddiction, and neuroprotective activities for cytisine, its
N-methyl (cauliphylline), and pyridone-substituted deriva-
tives. Recently Canu Boido and Sparatore14,21-24 prepared
and assayed in a number of tests a variety of cytisine deriva-
tives with substitution in the pyridone ring or N-substituted.
Afterward, several authors addressed their interest to the
study of different cytisine derivatives with substitution in
the pyridone ring or with N- or C-substitutions on the
bispidine moiety.25-30 Finally, Coe et al.,31 through a progres-
sive modification of the molecular frame of cytisine obtained
varenicline (6,7,8,9-tetrahydro-6,10-methano-6H-pyrazino-
[2,3-h][3]benzazepine), which displayed high binding affinity
and selectivity for rat R4β2 subtype and an interesting ther-
apeutic potential for smoking cessation and treatment of
alcohol dependence.
In line with these studies, we continue the investigation
on additional N-substituted derivatives of cytisine mainly
because the introduction of substituents on cytisine amino
group still represent the simplest way to enhance the
molecular lipophilicity and improve passage through cell
membranes and the blood-brain barrier. Although the
N-substitution is commonly considered detrimental for
affinity to nAChRs, the increased lipophilicity could balance
this negative effect.
To contribute to the understanding of the structural
requirements for targeting the nAChRs we have: (A) syn-
thesized several N-[(ω-aryl/heteroaryl)alkyl]cytisines and
N-(ω-aroylalkyl)cytisines (2-23) and, for comparison, a few
cytisine derivatives bearing aliphatic substituents (24-29)
(Chart 2) that were assayed for the displacement of [3H]-
epibatidineand[125I]-bungarotoxinfromR4β2 andR7 receptor
subtypes of rat cortex,1b respectively. Both these subtypes
have been proposed as therapeutic targets for neurological
pathologies and degenerative diseases. (B) In view of the
observed higher affinity for R4β2 versus R7 subtypes, we
docked the N-substituted cytisines in the structural model
of the nAChRs extracellular ligand binding domain of
the rat (R4)2(β2)3 type (PDB code: 1OLE), as well as in
our human (R4)2(β2)3 type (PDB code: 2GVT), developed
The reasons for the differences in the observed binding
affinity can depend from the interplay of the following
structural features:
(i) the length of the aliphatic chain that is interposed
between the basic nitrogen and the aromatic ring:
one C (2-5), two C (6-11, 18, 19), three C (12-15,
20, 21), four C (16, 17, 22, 23);
(ii) the presence of an electron withdrawing group (car-
bonyl, sulphonyl) in the chain and its distance from
the basic nitrogen (18-28);
(iii) the presence and the nature of substituents on the
aromatic moiety.
With an unsubstituted benzene nucleus, the elongation
of the aliphatic chain has a fluctuating effect, improv-
ing and then decreasing the affinity: N-phenylethyl- and
N-phenylpropylcytisine have an affinity more than 1 order
of magnitude higher than N-benzyl- and N-phenylbutylcy-
tisine. On the contrary, in N-(4-fluorophenyl)alkylcyti-
sines, the affinity decreases steadily with the increasing
number of methylene groups. Evidently, the para-fluoro

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4347
Chart 2
Scheme 1a
Table 1. Binding Affinity (Ki, nM) of Compounds 1-29 to R4β2 and
R7 Rat Nicotinic Receptor Subtypes, Labeled with [3H]-Epibatidine and
[
125I]-Bungarotoxin, Respectivelya
Ki, nM (%CV)
Ki, nM (%CV)
R4β2 R7
R4β2
R7
1
0.48(20)
850(20)
330(21)
6600(30)
675(27)
28(28)
331(28)
16
17
18
19
20
21
22
23
24
25
26
27
28
29
3900(29)
13000(26)
17(32)
nd
2
>50000
78700(50)
>100000
>100000
25000
>50000
2250(26)
nd
3
4
409(30)
2.6(23)
5.3(18)
816(30)
965(27)
7300(30)
35(31)
5
550(22)
433(22)
>50000
>50000
>50000
3600(38)
>50000
1240(22)
>50000
39700(20)
6
7
1300(25)
727(29)
495(24)
941(32)
7.2(26)
>50000
>50000
>50000
>100000
4200(29)
13000(17)
>50000
>50000
>100000
8
9
10
11
12
13
14
15
564(24)
8.7(24)
757(29)
332(22)
63(17)
5000(35)
7200(26)
744(37)
aReagents and conditions: (a) ratio 1: RBr/Cl = 2:1; MeCN;
90-100 °C; 20-24 h; (b) ratio 1: R0Br = 3:1; MeCN; 100 °C; 36 h;
(c) ratio 1: R00I = 1:1; K2CO3; 100 °C, 7 h; (d) ratio reactants = 1:1;
EtOH; AcOH; reflux, 24 h.
a The Ki values shown were the mean (% coefficient of variation) of
three-six independent measurements.
The introduction of a carbonyl group on the connecting
aliphatic chain (18-23) had always a positive influence on
affinity, but this effect became outstanding when the carbo-
nyl was in the β-position with respect to the basic nitrogen
(20, 21), also overcoming the negative influence of the para-
fluoro substitution. This observation still holds in the case of
compounds devoid of an aromatic ring, such as 25 and 27,
and suggests the possibility that all these compounds could
undergo to β-elimination reaction during the overnight
incubation, with release of cytisine, to which should be due
the displacement of the labeled ligand. However the release
of cytisine in the binding experimental conditions must be
substitution improved a little the affinity of the benzyl
derivative but strongly lowered that of phenylethyl- and
phenylpropyl- derivatives. A similar deleterious effect
was observed also for the 4-methyl and 4-methoxy substitu-
tion. The replacement of the benzene ring with the pyridine
affected the affinity depending on the joining position:
the 4-pyridylmethyl- and the 4-pyridylethylcytisines (5
and 11) behaved similarly to the corresponding phenyl
derivatives (2 and 6), while the 2- and 3-pyridylalkyl
derivatives (4, 10, and 15) exhibited a reduced affinity with
respect to 2, 6, and 12.

4348 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Tasso et al.
The minimum energy conformation of the examined
compound based on this model give rise to several docking
conformations, which after a visual inspection were clustered
and further analyzed through hybrid QM/MM geometry
optimization. Hybrid QM/MM methods have become a
standard tool for the characterization of complex molecular
systems, as permitted to analyze quantum mechanically the
fraction of the system that undergoes to the most significant
changes during the substrate binding, while the rest of the
system has been simulated with the traditional molecular
mechanic methods. This process was used to rerank the
structures generated by docking and to simulate the struc-
tural adaptations occurring in the ligand-receptor binding.
First, the more relevant docked conformations of the com-
plex ligand-rat (R4)2(β2)3 were equilibrated for 1.5 ns by
molecular dynamics, at constant temperature and pressure in
a periodic cubic box, using the TIP3P model for water
molecules (results not shown). The systems were subse-
quently optimized using the combined QM/MM approach,
with a flexible receptor environment allowing simulation of
the modification of the receptor upon ligand binding. This
procedure gives rise to a rearrangement of the residues
forming the binding site around the ligand (“induced fit”),
leading to the situation showed in Figure 2 for 20.
Figure 1. Crystal structure of one of the two independent molecules
of 20 (color by atom type: gray, carbon; blue, nitrogen; red, oxygen;
white, hydrogen).34
.
ruled out because the N-(benzoylethyl)cytisine, one of the
best ligands versus both R4β2 and R7 subtypes, was recovered
unchanged even after prolonged incubation in the usual
buffer solution. Thus the observed high affinity is peculiar
to compounds 20, 21, 25, and 27, which together with
compounds 6, 11, and 18 are the most interesting of the
whole set. The ratio between the affinity constants of com-
pounds 25 and 20 was 13.46, while that between the corre-
sponding lower homologues 24 and 18 was 429.4; therefore,
when the carbonyl group is placed in β position to the basic
nitrogen, the NH2 group exerts only a minor negative
influence, while in the shorter chain, it probably prevents
the correct positioning of the ligand on the receptor.
Molecular Docking. With the aim to a better understand-
ing of the influence of the N-substitution at the cytisine
system in the new synthesized compounds, we have exam-
ined the lower energy docking pose of cytisine (the reference
compound) in the binding pockets of rat and human
nAChRs. In both models, we have found the protonated
nitrogen lying in the center of the “aromatic cluster” formed
by the aromatic residues present at the ligand binding
pocket.32 In the rat receptor model, the pyridone stacks onto
Phe117 and the pose is stabilized by a H-bond with OTrp147,
while in the human model the charged nitrogen is sand-
wiched between Trp55 and Tyr195 and a hydrogen bond
involving the pyridone oxygen and NHETrp147 is present.
A detailed presentation of AA forming the rat and human
binding sites of the receptor is given as Supporting Informa-
tion, Table S2.
This binding site optimization leads to a small difference in
the residue geometry with respect to the starting conditions
˚
(rmsd all atoms of 1.08 A), while the number of hydrogen
bonds and ligand-residue contacts are almost unaffected.
The binding of 2-29 is characterized by hydrogen bonds
and/or cation-π interactions, with the protonated cytisine
nitrogen oriented toward the aromatic cage of the binding
pocket. The presence of a hydrogen bond between the
hydrogen of the protonated piperidine nitrogen N2 (arbi-
trary numbering, see Chart 2) and the macromolecule is
a stabilizing factor for the orientation of the ligands.
Although the majority share the same cation-π interaction
with the receptor, others adopt a somewhat diverse orienta-
tion within the binding pocket with different interaction
patterns. The best score binding pose of the most potent
compound 20 presents two hydrogen bonds: the protonated
amine moiety (N2H) interacts with the Trp147 backbone
˚
oxygen at the distance of 2.69 A, while the ketonic oxygen
˚
is 2.40 A apart from OHTyr195. The cytisine pyridone is
sandwiched in the aromatic cage, with the side chain pointing
toward the R4 subunit. The binding mode of 20 to rat
(R4)2(β2)3 nAChR model is shown in Figure 3 (left). The
derivative 21 follows the same behavior of compound
20. The introduction of a fluorine atom gives rise to a
comparable conformation of the side chain and of the
cytisine moiety, determining however two hydrogen bonds
between the protonated amine moiety (N2H) and the Trp
Before discussing separately the docking of the 28 cytisine
derivatives on the rat and human receptor models, it is
important to observe that in both models the charged
nitrogen of the N-substituted derivatives is inside the aro-
matic cage, but the cytisine pyridone ring points toward the
R subunit, in opposite direction with respect to the other
compounds, in which the pyridone ring is directed toward
the β subunit. The constantly higher affinity of the unsub-
stituted alkaloid might be related to this different orientation
inside the binding pocket.
˚
147 backbone oxygen at the distance of 3.00 A and between
˚
the pyridone oxygen O1 with NHTyr195 at 2.08 A, without
any interaction between O2 and HOTyr195. Compounds 18,
25, and 27, all bearing a carbonyl group in the side chain but
largely differing for the terminal moiety, share the interac-
tion pattern of 21. It is worth noting that the docking free
energy value of compound 25 ranks higher than expected on
the base of the experimental data.
Rat (r4)2(β2)3 nAChRs. The X-ray structure of 2034 that
exhibits the highest affinity has been used for molecular
docking analysis.
In its neutral form, 20 is represented by two independent
molecules with the same conformation of the cytisine moiety
with the pyridone ring almost planar, the adjacent ring in
envelope conformation, and the “bispidine” scaffold in chair
conformation (Figure 1).
The influence of the pyridine nitrogen in compounds 10
and 11 shows that the shifting from the 4-position to the
2-position in 10 leads to a significant decreasing in affinity.
These two compounds dock in a quite similar orienta-
tion as shown for 10 in Figure 3 (right), but compound

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4349
Figure 2. Particular of the QM region used in the QM/MM optimization of the rat receptor model and the 20 bioactive conformation in stick
(CPK). Residues in the starting conformation in blue sticks and after the QM/MM optimization in yellow sticks.
Figure 3. Representation of the best docked conformations rendered in capped sticks of 20 (left) and 10 (right) into the ligand-binding domain
(lines). Putative intermolecular hydrogen bonds are highlighted by green dashed lines.
11 has a higher consensus score in agreement with its
better affinity (Table 1).
of Leu119, giving rise to a different orientation of the side
chains (Figure 4). Similarly, all the most extended molecules
hardly fit in the binding pocket and those devoid of a
carbonyl group in the chain (14, 16, and 17) can have only
very poor interactions with the receptor.
In general, with the exception of compound 3, the docking
calculations evidence that the 4-substitution in the aromatic
moiety accounts for a worse interactions between the aro-
matic macromolecular moieties and the ligand, lowering
the consensus score. This effect is, however, overwhelmed
by the insertion of a carbonyl group which hydrogen bonds
residues in the binding pocket, as observed, for example,
in the case of the most extended and least potent compound
17 (ΔG = 7.02 kcal/mol) and the corresponding ketone
23 (ΔG=9.55 kcal/mol, see Table 2).
Independently from the length and the nature of the side
chain, compound 12, as well 2-5, 7-10, 14-17, 19, 22-24,
26, 28, and 29, present a productive cation-π interaction
between the protonated amine moiety (N2H) and Trp147
˚
(average distance 3.9 A). For these compounds, with the
exception of the most extended molecules, the different
orientations of the cytisine moiety still allow the formation
of one H-bond between the protonated amine moiety (N2H)
and the Trp147 backbone oxygen.
The introduction of the fluorine atom in 4-position in
compound 13 produces a 2 orders of magnitude decrease
in affinity with respect to 12, and this fact could be justified
by the fluorine atom repulsion with the backbone nitrogen

4350 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Tasso et al.
Figure 4. Representation of the best docked conformations rendered in capped sticks of 12 (left) and 13 (right) into the ligand-binding domain
of rat model (lines). The highlighted region evidence the Leu119 position with the 13 fluorine atom as orange spot. Putative intermolecular
hydrogen bonds are represented by green dashed lines. No C-F HN bond is present.
3 3 3
Table 2. Best Docking Scores of Cytisine and Derivatives 2-29 for Rat
and Human (R4)2(β2)3 nAChRs Models Compared with Experimental
Dataa
ΔGdock
rat
ΔGdock
human
|
ΔGdock ΔGdock
ΔGexp
ΔGexp
rat
human
1
2
3
4
5
6
7
8
9
-12.71 -12.66
-12.82
-10.62
-10.94
-9.63
-11.02
-12.10
-10.21
-10.94
-8.17
16 -7.38
17 -6.67
18 -10.60 -10.33 -12.16
19 -8.72 -9.21 -9.29
20 -11.71 -11.81 -12.48
21 -11.29 -11.78 -12.41
22 -8.31
23 -8.21
24 -7.01
25 -10.17 -10.99 -12.08
26 -8.53 -9.45 -11.19
27 -11.00 -10.49 -12.28
28 -8.35
29 -8.84
-7.13
-7.02
-7.02
-8.62
-8.28
-8.84
-7.07
-8.42
-9.12
-9.99
-8.12
-8.02
-10.30 -10.10
-8.03
-8.37
-8.60
-9.11
-7.91
-9.31
-9.11
-9.14 -10.62
-9.55 -10.24
-7.98
-9.14
10 -8.22
11 -11.11 -11.46
12 -9.82 -10.13
13 -7.23
14 -7.02
15 -8.36
-9.45
-12.31
-11.98
-8.96
-8.78
-10.87
Figure 5. Superimposition of the nonconserved residues of the
human (green) and rat (red) (R4)2(β2)3 nAChR models in the range
99-104.
-6.87
-8.45
-8.17
-8.87
-8.45
-8.73
-8.83
aΔG values are in kcal/mol.
˚
OHTyr195 at 2.35 A. The cytisine pyridone ring is sand-
wiched in the aromatic cage between Tyr195 (centroid dis-
˚
˚
The destabilizing influence of the halogen substitution and
the possible repulsive interactions lead to a detrimental
binding, as previously reported in our docking study on the
AChBP with 6-chloropyridazin-3-yl derivatives.35
tance 4.2 A) and Tyr188 (centroid distance 3.9 A). The side
chain points toward the R4 subunit and the benzene moiety
stacks nearly parallel to the Phe117 aromatic ring at about
˚
4 A (centroid distance). For compound 21, the introduction
Human (r4)2(β2)3 nAChRs. With the aim to have new
tools for predicting the binding affinities of the studied
compounds for the human receptor type, compounds 2-29
were docked in our human (R4)2(β2)3 nAChR model. Its
amino acid sequence is highly conserved (overall rmsd
0.9 A). Main differences are in the β subunit, in particular in
the range 99-104 (Figure 5).
of a fluorine atom generates a different conformation of the
side chain modifying the π interaction of the cytisine moiety,
which results less deeply positioned within the aromatic cage.
The decreasing of these π interactions lowers the consensus
score, nevertheless 21 forms one hydrogen bond between
˚
˚
O1 NHETrp147 at 2.26 A (Figure 6 right).
3 3 3
The shortening of the side chain of 20, to give compound
18, produces several changes in the docking. The cytisine
system is linked by the charged nitrogen to NHCys190 at
In the compounds 2, 4, 8-13, 15, 16, 18-21, 26, and 29,
the piperidine NH makes an hydrogen bond with the macro-
molecule. Although most of them share the same cation-π
interaction with the receptor, others adopt a distinct orienta-
tion within the binding pocket with different interactions.
The best binding pose of 20, shown in Figure 6 (left), is again
characterized by two hydrogen bonds between the proto-
nated amine moiety (N2H) with Trp147 backbone oxygen at
˚
2.34 A and its aromatic heterocycle lies parallel to Trp147
˚
and to Tyr188 (centroid distances 4.1 and 4.6 A, respec-
tively). These contacts do not influence the stacking of the
˚
aromatic moiety of the chain, which is 4.2 A apart from the
Phe117 aromatic ring. However, the ketone carbonyl
˚
strongly hydrogen bonds to NHETrp147 (1.99 A) and to
˚
HOThr148 (2.8 A) instead of OHTyr195 as observed for 20.
˚
the distance of 2.10 A and between the ketonic oxygen with

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4351
Figure 6. Representation of the best docked conformations rendered in capped sticks of 20 (left) and 10 (right) into the human ligand-binding
domain (lines). Putative intermolecular hydrogen bonds are highlighted by green dashed lines.
The exchange of the ketone group for a methylene in
20 and 18 gives rise to the phenylpropyl- (12) and the
phenylethyl- (6) derivative respectively, with a decrease of
affinity that is stronger in the first case (Ki from 2.6 to 63 nM)
than in the second one (Ki from 17 to 28 nM).
The phenylpropyl derivative 12 presents a tight interaction
of the cytisine ring system with the receptor, characteri-
zed by the strong hydrogen bond of piperidine N2H with
The substitution of the aromatic ring of the side chain
with smaller groups as for 25 and 27 do not change the
position of the cationic cytisine inside the human aromatic
binding pocket and allows the formation of additional
hydrogen bonds. In particular, in 25, the amide is strongly
linked to HNCys191 and to OGlu189 with the nitrogen
and to HNCys191 with the oxygen, while in 27, the carbonyl
group hydrogen bonds HOTyr195, increasing its consensus
score. The comparison of the molecular docking results
of compounds 24 and 25 evidence novel discriminating
receptor interactions, unexpected considering their small
chemical diversity. The addition of a methylene group
in 25 side chain allows the formation of two hydrogen
bonds involving the cytisine protonated nitrogen, which
can be considered, also in the light of this finding, important
to elicit the nicotinic activity. The 25 amidic moiety leads
to unfavorable electronic interactions with respect to
the phenyl group of 20, which shows about 15-fold higher
affinity in rat receptor.
Concluding, the eight compounds with the highest affinity
(20, 21, 11, 27, 18, 6, 25, and 12, in order of decreasing
affinity: Ki from 2.6 to 63 nM) behave similarly in both
rat and human receptor models and exhibit docking free
energy values that parallel the experimental ones, with the
only exception of compound 25, which ranks higher than
expected in rat model.
The parallelism between the experimental affinity data and
the docking free energy values is partially upset for compounds
of the groups with moderate (Ki = 330-965 nM) and poor
(Ki = 1300-13000 nM) affinity, in particular for that concern-
ing the human model. The most striking deviation is observed
for compound 9, which ranks as the last but one in the human
model, while it displays a moderate affinity (Ki = 495 nM) in
the binding experiments.
˚
˚
NHETrp147 at 2.08 A and of O1 with NHCys190 at 2.35 A.
This system is further stabilized by a stacking interaction
˚
of pyridone ring with Tyr188 at 3.9 A and of the terminal
phenyl ring over Phe117.
For compound 13, the repulsion between the para fluo-
rine atom and the Gln58 carbonyl oxygen could influence
the receptor affinity, as it was in rat model, with respect to 12.
The orientation of 12 is retained in the lower homologue 6,
which despite the shorter side chain maintains the productive
π interactions with Trp147 and Phe117. The phenylethyl
derivative 6 is the only one which presents, as best score pose,
the cytisine oxygen engaged in a hydrogen bond with
˚
NHCys191 at 2.35 A, while the pyridone ring is yet closely
˚
packed to Tyr195 and Tyr188 at about 4.01 A distance, as
seen for compound 20.
The passage from 20 and 21 to compound 11, despite the
major structural changes (absence of ketone group, shorter
linker, and pyridine in place of the benzene ring) leads to a
quite equipotent derivative, but the interactions at the bind-
ing site are rather different. The piperidine NH of this
˚
compound hydrogen bonds with OTrp147 at 3.05 A, and
this interaction is reinforced by the good fit of the cationic
head into the aromatic-rich binding pocket and by two
favorable π interactions of the pyridone ring with the aro-
˚
matic side chain of Tyr188 (centroid distance 4.1 A) and of
˚
the pyridine with Phe117 at 4.8 A distance.
The isomeric compound 10 has a similar orientation as 11
in the human model docking, but despite the presence of
the stabilizing hydrogen bonds between the pyridine nitro-
gen and NHETrp147 and HOThr148, presents a lower
consensus score than 11. Thus, the 2 orders of magnitude
decrease of affinity, due to the shift from 4- (11, Ki = 7.2 nM)
to 2-pyridine substitution (10, Ki = 941 nM), remains hard
to explain.
The statistical analysis of the docking scoring functions
point out the higher quality for the rat model with respect to
human (R4)2(β2)3 (Table 2 and Figure 7): an expected result,
taking into account that the pharmacological data are for rat
receptor. In particular, the correlation coefficient calculated
between ΔGexp and ΔGdock is the most relevant element for
the model quality assessment, which is 0.85 for the rat model
and 0.67 for the human model, as shown by the obtained

4352 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Tasso et al.
Figure 7. Correlation between the experimental (ΔGexp) and docking (ΔGdock) free energies calculated in ligand-binding [GB1]domain of rat
(left) and human (right) (R4)2(β2)3 nAChR model.
linear regression equations:
form hydrogen bonds. Further, the computational evaluation
has evidenced that the potency of these compounds could be
modulated by a larger number of hydrophobic interactions
inside the binding site.
ΔGexp ¼ 0:220þ0:981ΔGmodel
n ¼ 29 r2 ¼ 0:85 q2 ¼ 0:85 s ¼ 0:61 F ¼ 156:03
ΔGexp ¼ -0:25þ0:829ΔGmodel
The applied flexible docking has shown a good correlation
between the estimated free energy binding and the experi-
mental binding data. Although the values of the correlation
coefficients obtained for the docking scoring functions indi-
cate a better fit for the rat (R4)2(β2)3 model with respect to the
human, as expected because the binding data were obtained
on rat receptors, it is interesting to observe that these data are
not very different from those obtained using human receptor
model. These findings indicate that experimental binding data
on rat cortex preparations are useful also to predict the
binding affinity on human receptor subtypes.
Compounds 20, 21, 11, 27, 18, 6, and25, whichare endowed
with the highest affinity for R4β2 receptor subtype (Ki in
the range 2.6-35 nM), are characterized by a quite higher
lipophilicity than cytisine and therefore more suitable
for crossing the blood-brain barrier. Thus they deserve
further investigations to define their in vivo pharmacological
activities.
n ¼ 29 r2 ¼ 0:68 q2 ¼ 0:67 s ¼ 0:91 F ¼ 57:33
In the light of these findings, the close analogy between the
same receptor subtype R4β2 of the two species and the rather
similar docking interactions permit consideration of our
human receptor model useful to predict the activities of these
compounds in human cell lines and could be a valuable tool
for structure based drug design of new selective R4β2 nAChR
ligands.
Conclusions
We have synthesized a new series of N-(arylalkyl)- and
N-(aroylalkyl)-cytisines and found by binding studies on rat
R4β2 and R7 receptor subtypes that a large number of them has
a high affinity for R4β2 subtype, with Ki values in the low
nanomolar range for the best compounds and a rather poor
affinity for R7 subtype.
Finally, the present study may be useful to derive guidelines
for the rational search of even more potent compounds.
Experimental Section
Three of these compounds (3, 27, and 29) have been tested
previously, in heterologously expressed nicotinic subtypes,
for their functional activity by using Ca2þ dynamic and
electrophysiological recording.14 All the three compounds
showed antagonist/partial agonist activity toward the
different subtypes, thus indicating that the N substituent
greatly affect the efficacy of the cytisine derivatives.
From these previous data, we expect that the new compounds
can have antagonist/partial agonist activity, properties
that have been explored for possible positive therapeutic
effects.1-5,36 However, the possibility that all or some of these
compounds exhibit “receptor desensitizing” properties, as
observed for the bulky 10-substituted cytisine derivatives
described by Kozikowski et al.,27d deserves to be fully inves-
tigated.
Chemistry. Melting points were taken in open glass capil-
laries on a Buchi apparatus and were uncorrected. H NMR
1
spectra were recorded in CDCl3 or DMSO-d6 on a Varian
Gemini 200 spectrometer. Chemical shifts (δ) are reported
in parts per million from the peak for internal Me4Si.
Values of the coupling constants (J) are reported in hertz.
Cytisine protons are indicated as “R-pyr” or as “bisp” if
pertinent, respectively, to the R-pyridone or to the bispidine
moiety. Column chromatography (CC) was performed by using
basic alumina (Across). Elemental analyses were performed on a
Carlo Erba EA 1110 CHNS-O instruments in the Microanalysis
Laboratory of the Department of Pharmaceutical Sciences of
the University of Genoa. The analytical results are within
(0.3% of calculated values. The results of NMR spectra and
elemental analyses indicated that the purity of all compounds
was higher than 95%.
Intermediates. The required arylalkylhalides, ω-aroylalkylha-
lides, iodoacetamide, 2- and 4-vinylpyridine, acrylamide, methylvi-
nylsulfone, 2-, 3-, and 4-hydroxymethylpyridine, and 3-(3-
hydroxypropyl)pyridine were purchased from Aldrich. 3-(3-
Bromopropyl)pyridine hydrobromide was prepared according
to the method of Fabio et al.37
The docking at the rat and human R4β2 neuronal nicotinic
receptor model permitted rationalization of the observed
increasing potency produced by the presence in the ligand of
a R terminal aromatic group (able to stack with Phe144)
linked to the cytisine nitrogen by a short chain (two or three
carbon atoms) bearing a group, like a carbonyl moiety, able to

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4353
The 2-, 3-, and 4-bromomethylpyridines hydrobromides were
1-Phenyl-3-(cytisin-12-yl)-1-propanone (20). mp: 115-117 °C
(Et2O) Yield: 59%. 1H NMR (CDCl3) δ: 1.54-1.90 (m, 2H, bisp),
2.21-2.44 (m, 3H, bisp), 2.64 (t, 2H, J = 6.8, CH2), 2.76-2.99
(m, 5H, 3H, bisp þ 2H, CH2), 3.80 (dd, 1H, J = 15.4, 6.4, bisp),
3.91 (d, 1H, J = 15.4, bisp), 5.88 (dd, 1H, J = 6.8, 1.3, R-pyr), 6.34
(dd, 1H, J = 9, 1.3, R-pyr), 7.18 (dd, 1H, J = 9, 6.8, R-pyr), 7.29-
7.55 (m, 3H, aromatic protons), 7.71-7.82 (m, 2H, aromatic
protons). Anal. (C20H22N2O2) C, H, N.
1-(4-Fluorophenyl)-3-(cytisin-12-yl)-1-propanone (21). Oil (CC,
Al2O3, CH2Cl2). Yield: 75%. 1H NMR (CDCl3) δ: 1.55-1.92 (m,
2H, bisp), 2.16-2.41 (m, 3H, bisp), 2.61 (t, 2H, J=6.7, CH2C(O)),
2.72-2.98 (m, 5H, 3H, bisp þ 2H, CH2), 3.78 (dd, 1H, J = 15.1,
6.1, bisp), 3.90 (d, 1H, J = 15.1, bisp), 5.86 (dd, 1H, J = 6.8, 1.4,
R-pyr), 6.33 (dd, 1H, J = 9, 1.4, R-pyr), 6.92-7.09 (m, 2H,
aromatic protons), 7.17 (dd, 1H, J = 9, 6.8, R-pyr), 7.68-7.84
(m, 2H, aromatic protons).
prepared by the method of Bixler and Niemann.38 The melting
points corresponded to the literature.39
Preparation of Compounds 7, 8, 13, 14, 16-18, 20-22, and 28:
General Method. In a Aldrich pressure tube, to a solution of
cytisine (3-6 mmol) in MeCN (5-8 mL) the proper haloder-
ivative (1.5-3 mmol) was added. The tube was flushed with N2,
closed, and heated to 100 °C for 20-24 h. A shorter time of
heating (7 h) was sufficient for compound 18, while compounds
13 and 17 required 36 h of heating. After cooling, the precipitate
was filtered and the solution was concentrated to dryness. The
residue was taken up in acidic H2O, extracted with ether, and
after alkalinization extracted with CH2Cl2. Compounds were
purified as indicated in each case.
N-[2-(4-Fluorophenyl)ethyl]cytisine (7). mp: 105-107 °C
(Et2O). Yield: 64%. 1H NMR (CDCl3) δ: 1.57-1.88 (m, 2H, bisp),
2.17-2.55 (m, 7H, 3H, bisp þ 2H, N-CH2 þ 2H, CH2-Ar),
2.73-2.94 (m, 3H, bisp), 3.79 (dd, 1H, J = 15.3, 5.9, bisp), 3.92
(d, 1H, J = 15.3, bisp), 5.88 (dd, 1H, J = 7, 1.3, R-pyr), 6.39
(dd, 1H, J = 9, 1.3, R-pyr), 6.67-6.90 (m, 4H, aromatic protons),
7.19 (dd, 1H, J = 9, 7, R-pyr). Anal. (C19H21FN2O) C, H, N.
N-[2-(4-Methylphenyl)ethyl]cytisine (8). mp: 139-140 °C (ace-
tone). Yield: 72%. 1H NMR (CDCl3) δ: 1.58-1.91 (m, 2H, bisp),
2.12-2.54 (m, 10H, 3H, bisp þ 2H, N-CH2 þ 2H, CH2-Ar,
superimposed on s at 2.21, 3H, CH3), 2.74-2.98 (m, 3H, bisp), 3.81
(dd, 1H, J = 15.3, 5.9, bisp), 3.93 (d, 1H, J = 15.3, bisp), 5.90 (dd,
1H, J = 6.8, 1.2, R-pyr), 6.39 (dd, 1H, J = 9, 1.2, R-pyr), 6.74-7.02
(m, 4H, aromatic protons), 7.21 (dd, 1H, J = 9, 6.8, R-pyr). Anal.
(C20H24N2O) C, H, N.
The following data are for the hydrochloride. mp: 173-
175 °C. Anal. (C20H21FN2O2 HCl 0.25H2O) C, H, N.
3
3
1-Phenyl-4-(cytisin-12-yl)-1-butanone (22). mp: 114-115 °C
(acetone) Yield: 51%. 1H NMR (CDCl3) δ: 1.63-1.98 (m, 4H, 2H,
bisp þ 2H, CH2-CH2-CH2), 2.18-2.50 (m, 5H, 3H, bisp þ 2H,
CH2), 2.68 (t, 2H, J=7.6, CH2C(O)), 2.80-2.99 (m, 3H, bisp), 3.87
(dd, 1H, J = 15.8, 6.4, bisp), 4.09 (d, 1H, J = 15.8, bisp), 5.92 (dd,
1H, J = 7, 1.6, R-pyr), 6.34 (dd, 1H, J = 9.2, 1.6, R-pyr), 7.14 (dd,
1H, J = 9.2, 7, R-pyr), 7.35-7.60 (m, 3H, aromatic protons), 7.72-
7.84 (m, 2H, aromatic protons). Anal. (C21H24N2O2) C, H, N.
1-Methyl-4-(cytisin-12-yl)-1-butanone (28). mp: 94-96 °C
(Et2O) Yield: 27%. 1H NMR (CDCl3) δ: 1.37-1.58 (m, 2H,
CH2), 1.62-1.93 (m, 5H, 2H, bisp superimposed on s at 1.86
3H, CH3), 1.99-2.42 (m, 7H, 3H, bisp þ 2H, N-CH2 þ 2H,
CH2-C(O)), 2.65-2.94 (m, 3H, bisp), 3.78 (dd, 1H, J = 15.4, 6.5,
bisp), 3.98 (dd, 1H, J = 15.4, bisp), 5.89 (dd, 1H, J = 7, 1.2,
R-pyr), 6.36 (dd, 1H, J = 9, 1.2, R-pyr), 7.20 (dd, 1H, J = 9, 7,
R-pyr). Anal. (C16H22N2O2) C, H, N.
N-[3-(4-Fluorophenyl)propyl]cytisine (13). Oil °C (CC, Al2O3,
1
CH2Cl2). Yield: 89%. H NMR (CDCl3) δ: 1.46-1.72 (m, 2H,
CH2-CH2-CH2), 1.75-2.02 (m, 2H, bisp), 2.03-2.54 (m, 7H, 3H,
bisp þ 2H, N-CH2 þ 2H, CH2-Ar), 2.78-3.10 (m, 3H, bisp), 3.92
(dd, 1H, J = 15.6, 7.6, bisp), 4.13 (d, 1H, J = 15.6, bisp), 6.03 (dd,
1H, J = 7.6, 1.6, R-pyr), 6.48 (dd, 1H, J = 8.8, 1.6, R-pyr), 6.82-
6.89 (m, 4H, aromatic protons), 7.32 (dd, 1H, J = 8.8, 7.6, R-pyr).
The following are data for the hydrochloride. mp: 215-
Preparation of N-Pyridinylalkyl-cytisines 4, 5, 15: General
Method. In a Aldrich pressure tube, to a solution of cytisine
(0.57 g, 3 mmol) in MeCN (8 mL) the appropriate bromoalk-
ylpyridine hydrobromide (1 mmol) was added. The tube
was flushed with N2, closed, and heated to 100 °C for
36 h. After cooling, the precipitate was filtered and the
solution was concentrated to dryness. The residue was dissolved
in acidic water, the acidic solution was extracted with ether
and, after alkalinization, extracted with CH2Cl2. The organic
phase was dried (Na2SO4), and the solvent was removed
under vacuum; the residue was purified as indicated for each
compound.
N-[(Pyridin-3-yl)methyl]cytisine (4). mp: 128-129 °C (Et2O).
Yield: 64%. 1H NMR (CDCl3) δ: 1.64-1.96 (m, 2H, bisp), 2.19-
2.48 (m, 3H, bisp), 2.66-2.97 (m, 3H, bisp), 3.23-3.51 (AB
system, 2H, CH2), 3.83 (dd, 1H, J = 16, 7.8, bisp), 4.05 (d, 1H,
J = 16, bisp), 5.83 (dd, 1H, J = 7.4, 1.3, R-pyr), 6.42 (dd, 1H, J =
9.1, 1.3, R-pyr), 7.04 (dd, 1H, J = 9.1, 7.4, R-pyr), 7.14-7.30 (m,
2H, aromatic protons), 8.13-8.25 (m, 1H, aromatic proton),
8.31-8.42 (m, 1H, aromatic proton). Anal. (C17H19N3O) C, H, N.
N-[(Pyridin-4-yl)methyl]cytisine (5). mp: 121-122 °C (Et2O).
Yield: 57%. 1H NMR (CDCl3) δ: 1.58-1.94 (m, 2H, bisp), 2.14-
2.45 (m, 3H, bisp), 2.53-2.95 (m, 3H, bisp), 3.13-3.47 (AB system,
2H, CH2), 3.77 (dd, 1H, J = 15.4, bisp), 4.04 (d, 1H, J = 15.4,
bisp), 5.81 (dd, 1H, J = 7.6, 1.4, R-pyr), 6.38 (dd, 1H, J = 9.2, 1.4,
R-pyr), 6.64-6.88 (m, 2H, aromatic protons), 7.17 (dd, 1H, J =
9.2, 7.6, R-pyr), 8.14-8.38 (m, 2H, aromatic protons). Anal.
(C17H19N3O) C, H, N.
217 °C. Anal. (C20H23FN2O HCl) C, H, N.
3
N-[3-(4-Methylphenyl)propyl]cytisine (14). mp: 81-82 °C
(Et2O). Yield: 83%. 1H NMR (CDCl3) δ: 1.35-1.95 (m, 4H,
2H, bisp, þ 2H, CH2-CH2-CH2), 1.98-2.45 (m, 10H, 3H bisp þ
2H, N-CH2 þ 2H, CH2-Ar, superimposed on s at 2.22, 3H,
CH3), 2.66-2.98 (m, 3H, bisp), 3.83 (dd, 1H, J = 15.7, 6.6, bisp),
4.02 (d, 1H, J = 15.7, bisp), 5.93 (dd, 1H, J = 6.8, 1.3, R-pyr), 6.38
(dd, 1H, J = 9, 1.3, R-pyr), 6.70-7.04 (m, 4H, aromatic protons),
7.23 (dd, 1H, J = 9, 6.8, R-pyr). Anal. (C21H26N2O) C, H, N.
N-(4-Phenylbutyl)cytisine (16). mp: 93-94 °C (Et2O). Yield:
1
87%. H NMR (CDCl3) δ: 1.12-1.49 (m, 4H, CH2-CH2-CH2-
CH2), 1.58-1.92 (m, 3H, bisp), 2.02-2.54 (m, 6H, 2H, bisp þ 2H,
N-CH2 þ 2H, CH2-Ar), 2.68-2.97 (m, 3H, bisp), 3.81 (dd, 1H,
J = 15.5, 6.2, bisp), 3.97 (d, 1H, J = 15.5, bisp), 5.89 (dd, 1H, J =
6.7, 1.3, R-pyr), 6.35 (dd, 1H, J = 9, 1.3, R-pyr), 6.89-7.35 (m, 6H,
5H, aromatic protons þ 1H, R-pyr). Anal. (C21H26N2O) C, H, N.
N-[4-(4-Fluorophenyl)butyl]cytisine (17). mp: 95-96 °C
(Et2O). Yield: 52%. 1H NMR (CDCl3) δ: 1.12-1.43 (m, 4H, 2H,
bisp, þ 2H, CH2-CH2-CH2), 1.62-1.91 (m, 3H, bisp), 2.05-2.46
(m, 6H, 2H, bisp þ 2H, N-CH2 þ 2H, CH2-Ar), 2.67-2.94 (m,
3H, bisp), 3.81 (dd, 1H, J = 15.1, 6.7, bisp), 3.97 (d, 1H, J = 15.1,
bisp), 5.88 (dd, 1H, J = 6.8, 1.3, R-pyr), 6.34 (dd, 1H, J = 9, 1.3, R-
pyr), 6.77-7.02 (m, 4H, aromatic protons), 7.18 (dd, 1H, J = 9, 6.8,
R-pyr). Anal. (C21H25FN2O) C, H, N.
1-Phenyl-2-(cytisin-12-yl)-1-ethanone (18). mp: 138-140 °C
1
(acetone) Yield: 82%. H NMR (CDCl3) δ: 1.72-1.97 (m, 2H,
N-[3-(Pyridin-3-yl)propyl]cytisine (15). Oil (CC, Al2O3, CH2Cl2).
Yield: 45%. 1HNMR(CDCl3)δ:1.24-1.78 (m, 4H, 2H bisp þ 2H,
CH2CH2CH2), 1.84-2.37 (m, 6H, 2H, bisp þ 2H, N-CH2 þ 2H,
CH2-Ar), 2.54-2.89 (m, 3H, bisp), 3.08 (s, 1H, bisp), 3.72 (dd, 1H,
J= 14.9, 5.8, bisp), 3.93 (d, 1H, J= 14.9, bisp), 5.88 (dd, 1H, J=8,
1.2, R-pyr), 6.26 (dd, 1H, J = 8.8, 1.2, R-pyr), 6.82-7.08 (m, 2H,
aromatic protons), 7.16 (dd, 1H, J = 8.8, 8, R-pyr), 7.92-8.04
(m, 1H, aromatic proton), 8.12-8.25 (m, 1H, aromatic proton).
bisp), 2.37-2.48 (m, 1H, bisp), 2.53-2.70 (m, 2H, bisp), 2.75-3.00
(m. 3H, bisp), 3.44-3.63 (AB system, 2H, CH2), 3.82 (dd, 1H, J =
15.4, 6.2, bisp), 3.91 (d, 1H, J = 15.4, bisp), 5.83 (dd, 1H, J = 7,
1.6, R-pyr), 6.37 (dd, 1H, J = 9.2, 1.6, R-pyr), 6.22-6.36 (m, 2H,
aromatic protons), 6.41-6.54 (m, 1H, aromatic proton), 7.14 (dd,
1H, J = 9.2, 7.2, R-pyr), 7.72-7.83 (m, 2H, aromatic protons).
Anal. (C19H20N2O2) C, H, N.

4354 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Tasso et al.
The following data are for the hydrocloride. mp: 194-197 °C
homogenates were resuspended in the same buffer containing
20 μg/mL of the protease inhibitors leupeptin, bestatin, pepsta-
tin A, and aprotinin. Receptor expression ranged from 50 to
70 fmol/mg of protein.
(d). Anal. (C19H23N3O HCl 1.75 H2O) C, H, N.
3
3
(Cytisin-12-yl)acetamide (24). In an Aldrich pressure tube,
to a solution of cytisine (0.57 g, 3 mmol) in MeCN (6 mL),
iodoacetamide (0.56 g, 3 mmol) and anhydrous K2CO3 (0.41 g,
3 mmol) were added. The tube was flushed with N2, closed, and
heated at 100 °C for 7 h. After cooling, the precipitate was
filtered and the solution was concentrated to dryness. The
residue was taken up in acidic water, the acidic solution was
extracted with ether and, after alkalinization, extracted with
CH2Cl2. After drying (Na2SO4), the solution was concentrated
to dryness and the residue crystallized from acetone. It is worth
noting that this compound was previously isolated, as an oil
from Sophora exigua.40
[3H]-Epibatidine Binding. (()-[3H]-epibatidine with a specific
activity of 56-60 Ci/mmol was purchased from Perkin-Elmer
(Boston MA); the nonradioactive R-bungarotoxin, nicotine,
and epibatidine were purchased from Sigma. It has been pre-
viously reported that [3H]-epibatidine also binds to R-bungar-
otoxin binding receptors with nM affinity.41 To prevent the
binding of [3H]-epibatidine to the R-bungarotoxin binding
receptors, the membrane homogenates were preincubated with
2 μM R-bungarotoxin and then with [3H]-epibatidine. Prelimin-
ary time course experiments were performed before saturation
and competition analyses to determine the time required for
[3H]-epibatidine to reach equilibrium with the R4β2 nAChRs. In
the epibatidine saturation experiments, aliquots of cortex
homogenates were incubated overnight at 4 °C with concentra-
tions of [3H]-epibatidine ranging between 0.005 and 2.5 nM
diluted in buffer A. Nonspecific binding was determined in
parallel by means of incubation in the presence of 100 nM
unlabeled epibatidine. At the end of the incubation, the samples
were filtered on GFC filters presoaked in polyethylenimine
through an harvester apparatus, and the filters were counted
in a β counter. We determined a Kd value of [3H]-epibatidine of
68 pM (CV = 15%). To test the ability of compounds to inhibit
[3H]-epibatidine binding, drugs were dissolved in water or
DMSO and then diluted in buffer A just before use. The
inhibition of radioligand binding by epibatidine, nicotine, cyti-
sine and test compounds was measured by preincubated cortex
homogenates with increasing doses (10 pM-10 mM) of the
reference nicotinic agonists, epibatidine or nicotine, and the drug
to be tested for 30 min at r.t., followed by overnight incubation
with a final concentration of 0.005-0.1 nM [3H]-epibatidine
(concentration in the Kd range of [3H]-epibatidine) at 4 °C.
mp: 173-174 °C. Yield: 81%. 1H NMR (CDCl3) δ: 1.73-2.04
(m, 2H, bisp), 2.41-2.65 (m, 3H, bisp), 2.79-3.07 (m, 5H, 3H,
bisp þ 2H, CH2), 3.88 (dd, 1H, J = 16, 6.8, bisp), 4.18 (d, 1H,
J = 16, bisp), 5.24 (s, 1H, NH collapses with D2O), 6.06 (s, 1H,
NH collapses with D2O, superimposed on dd at 6.02 dd, 1H, J =
6.9, 1.2, R-pyr), 6.45 (dd, 1H, J = 9.4, 1.2, R-pyr), 7.31 (dd, 1H,
J = 9.4, 6.9, R-pyr). Anal. (C13H17N3O2) C, H, N.
Preparation of Compounds 10, 11, 25, and 26: General Method.
Cytisine (0.38 g, 2 mmol) was dissolved in EtOH (10 mL) and
treated with 2 mmol of the appropriate vinylcompound (2- and
4-vinylpyridine, acrylamide, methylvinylsulfone) and acetic
acid (0.25 mL). The solution was refluxed for 24 h, under a
stream of N2, and afterwards was concentrated to dryness and
taken up in acidic water. The acid solution was extracted with
ether, made alkaline and extracted with CH2Cl2. The dichlor-
omethane solution was dried (Na2SO4) and the solvent re-
moved.
N-[2-(Pyridin-2-yl)ethyl]cytisine (10). mp: 92-93 °C (Et2O).
Yield: 83%. 1H NMR (CDCl3) δ: 1.59-1.88 (m, 2H, bisp),
2.18-2.42 (m, 3H, bisp), 2.48-2.76 (m, 4H, 2H, N-CH2 þ 2H,
CH2-Ar), 2.78-2.95 (m, 3H, bisp), 3.64-3.85 (AB system, 2H,
bisp), 5.85 (dd, 1H, J = 7, 1.4, R-pyr), 6.33 (dd, 1H, J = 9, 1.4,
R-pyr), 6.66-6.78 (m, 1H, aromatic proton), 6.89-7.01 (m, 1H,
aromatic proton), 7.16 (dd, 1H, J = 9, 7, R-pyr), 7.29-7.44
(m, 1H, aromatic proton), 8.28-8.38 (m, 1H, aromatic proton).
Anal. (C18H21N3O) C, H, N.
N-[2-(Pyridin-4-yl)ethyl]cytisine (11). mp: 86-87 °C (Et2O).
Yield: 77%. 1H NMR (CDCl3) δ: 1.35-1.66 (m, 2H, bisp),
1.94-2.35 (m, 8H, 4H, bisp þ 2H, N-CH2 þ 2H, CH2-Ar),
2.49-2.72 (m, 2H, bisp), 3.52 (dd, 1H, J = 15.5, 6.3, bisp), 3.67 (d,
1H, J= 15.5, bisp), 5.67 (dd, 1H, J=7.1, 1.5, R-pyr), 6.16 (dd, 1H,
J = 9, 1.5, R-pyr), 6.46-6.65 (m, 2H, aromatic protons), 6.97 (dd,
1H, J =9, 7.1, R-pyr), 7.95-8.16 (m, 2H, aromatic protons). Anal.
(C18H21N3O) C, H, N.
3-(Cytisin-12-yl)propionamide (25). mp: 185-187 °C (acetone).
Yield: 71%. 1H NMR (CDCl3) δ: 1.74-1.99 (m, 2H, bisp), 2.09-
2.61 (m, 7H, 3H, bisp þ 2H, N-CH2 þ 2H, CH2C(O)), 2.94-3.06
(m, 3H, bisp), 3.80 (dd, 1H, J = 15.3, 6.1, bisp), 4.07 (dd, 1H, J =
15.3, bisp), 4.82 (s, 1H, NH collapses with D2O), 5.96 (dd, 1H, J =
7.1, 1.2, R-pyr), 6.36 (dd, 1H, J = 9.3, 1.2, R-pyr), 6.95 (s, 1H, NH
collapses with D2O), 7.23 (dd, 1H, J = 9.3, 7.1, R-pyr). Anal.
(C14H19N3O2) C, H, N.
N-[2-(Methylsulfonyl)ethyl]cytisine (26). mp: 83-84°C (Et2O).
Yield: 78% 1H NMR (CDCl3) δ: 1.58-1.97 (m, 2H, bisp),
2.21-2.58 (m, 8H, 3H, bisp þ 2H, CH2, superimposed on s at
2.39, 3H, CH3), 2.62-3.08 (m, 5H, 3H, bisp þ 2H, CH2), 3.80
(dd, 1H, J = 15.5, 6, bisp), 3.96 (d, 1H, J = 15.5, bisp), 5.94
(dd, 1H, J = 6.8, 1.1, R-pyr), 6.34 (dd, 1H, J = 9.2, 1.1,
R-pyr), 7.20 (dd, 1H, J = 9.2, 6.8, R-pyr). Anal. (C14H20N2O3S)
C, H, N.
[
125I]-r-Bungarotoxin Binding. [125I]-R-Bungarotoxin with a
specific activity of 200 Ci/mmol was purchased from Amer-
sham. The saturation binding experiments were performed
using aliquots of cortex membrane homogenates incubated
overnight with 0.1-10 nM concentrations of [125I]-R-bungar-
otoxin at rt. Nonspecific binding was determined in parallel by
means of incubation in the presence of 1 μM unlabeled
R-bungarotoxin. After incubation, the samples were filtered as
described above and the bound radioactivity was directly
counted in a γ counter. We determined a Kd value of [125I]-R-
bungarotoxin of 0.8 nM (CV=25%).The inhibition of radioli-
gand binding by epibatidine, nicotine, and the test compounds
was measured by preincubating cortex homogenates with in-
creasing doses (10 pM-10 mM) of the reference nicotinic
agonists, epibatidine or nicotine, and the drug to be tested for
30 min at rt, followed by overnight incubation with a final
concentration of 1 nM [125I]-R-bungarotoxin (concentration in
the Kd range of [125I]-R-bungarotoxin) at the same temperatures
as those used for the saturation experiments.
Data Analysis. For each compound, the experimental data
obtained from the three saturation and three competition bind-
ing experiments were analyzed by means of a nonlinear least-
squares procedure using the LIGAND program as described by
Munson and Rodbard.42 The binding parameters were calcu-
lated by simultaneously fitting three independent saturation
experiments, and the Ki values were determined by fitting the
data of three independent competition experiments. The errors
in the KD and Ki values of the simultaneous fits were calculated
using the LIGAND software and were expressed as percentage
coefficients of variation (% CV). When final compound con-
centrations up to 200 μM did not inhibit radioligand binding,
the Ki value was defined as being >100 μM based on the Cheng
and Prusoff’s equation.
Binding Studies. Preparation of Tissue Sample. Cortex from
21 day old rats were dissected and immediately frozen . Frozen
tissue was homogenized using a Potter homogenizer in an excess
of buffer A (50 mM Tris-HCl pH 7, 120 mM NaCl, 5 mM KCl,
1 mM MgCl2, 2.5 mM CaCl2, and 2 mM phenylmethylsulfonyl
fluoride), centrifuged (60 min at 30000g), and rinsed twice. The
Molecular Modeling. The three-dimensional structure of the
29 examined ligands in their protonated forms were built and

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4355
energy minimized within Ghemical43 starting from the X-ray
structure of 20.34 All calculations were performed on a 3.0 GHz
Quad-Xeon 64-bit workstation running under the CentOS4.4
x86_64 Linux distribution. The Fortran software sources for
“Modeler” and “Tinker” were recompiled using the 64-bit
optimizations to suit the 64-bit and SMP CPU architecture.
Molecular Docking. The rat and human R4β2 models were
used for docking studies with the program AutoDock (version
4.0).44 First, it was checked if AutoDock was able to find the
correct position of the cocrystallized ligand (Aplysia californica
AChBP cocrystallized with epibatidine, PDP code: 2BYQ). The
docking software was able to detect the conformation of epiba-
tidine molecule within the 10 best docking poses with an rmsd
˚
interactions, with a long-range cutoff of 14 A and a short-range
˚
cutoff of 8 A.
Acknowledgment. Financial support from MIUR (Minis-
ꢀ
tero dell’Istruzione, dell’Universita e della Ricerca, Rome,
Italy) is gratefully acknowledged.
Supporting Information Available: Binding affinity (Ki, nM)
of some cytisine derivatives to rat cortex R4β2 receptor subtype
labeled with [3H]-epibatidine and [3H]-cytisine respectively;
amino acid content of the rat and human binding pockets
˚
within 5 A radius; amino acids involved in the cytisine binding;
elemental analysis results for the novel N-substituted cytisine
derivatives. This material is available free of charge via the
Internet at http://pubs.acs.org.
˚
value below 1.5 A. The cytisine derivatives (1-29) were docked
at the putative binding site by using a two-step docking process.
The docking procedure was at first applied to the whole protein
target, without imposing any binding site, using the so-called
“blind docking” approach.45 A grid map was generated for the
whole protein target, centered at the middle of the receptor
References
(1) (a) Gotti, C.; Fornasari, D.; Clementi, F. Human neuronal nico-
tinic acetylcholine receptors. Prog. Neurobiol. 1997, 53, 199–237.
(b) Gotti, C.; Zoli, M.; Clementi, F. Brain nicotinic acetylcholine
receptors: native subtypes and their relevance. Trends Pharmacol.
Sci. 2006, 27, 482–491.
(2) Holladay, M. W.; Dart, M. J.; Lynch, J. K. Neuronal nicotinic
acetylcholine receptors as targets for drug discovery. J. Med. Chem.
1997, 40, 4169–4194.
˚
models and using a grid resolution of 0.55 A. The resulting
docked conformations were clustered into families of similar
binding modes, with a root-mean-square deviation (rmsd) clus-
˚
tering tolerance of 2 A. In almost all cases, the lowest docking-
energy conformations were included in the largest cluster found
(which usually contains 80-100% of total conformations),
then the lowest docking-energy conformations were considered
as the most stable orientations. Second, we docked the ligands
in the identified binding site (“refined docking”): a radius of
(3) Gotti, C.; Clementi, F. Neuronal nicotinic receptors: from struc-
ture to pathology. Prog. Neurobiol. 2004, 74, 363–396.
ꢁ
(4) Cassels, B. K.; Bermudez, I.; Dajas, F.; Abin-Carriquiry, J. A.;
Wonnacott, S. From ligand design to therapeutic efficacy: the
challenge for nicotinic receptor research. Drug Discovery Today
2005, 10, 1657–1665.
˚
7 A centered on the best-scored conformation previously ob-
˚
tained was considered, with a finer grid resolution (∼0.36 A).
(5) Jensen, A. A.; Frolund, B.; Liljefors, T.; Krogsgaard-Larsen,
P. Neuronal nicotinic acetylcholine receptors: structural revela-
tions, target identifications, and terapheutic inspirations. J. Med.
Chem. 2005, 48, 4705–4745.
The resulting orientations were again clustered, considering a
˚
root-mean-square deviation (rmsd) tolerance of 2.0 A, into
families, and the lowest docking-energy conformations were
then equilibrated for 1.2 ns by unrestrained MD. The simula-
tions were performed at constant temperature and pressure
(NPT ensemble) in a periodic cubic box of TIP3P water mole-
cules. The water bond distances and angle forced using the
SETTLE algorithm,46 while the bond lengths within the protein
were constrained with the LINCS algorithm.47 The coupling
time was set to 1.0 ps, and the isothermal compressibility was set
to 4.6 ꢀ 10-5 bar-1. The protein, ligand, and solvent were
independently coupled to a temperature of 298 K, coupling time
0.1 ps, and the pressure was held at 1 bar, coupling time 0.2 ps,
using a Berendsen thermostat to maintain temperature and
pressure unvarying. The time step used was 1.0 fs. Snapshots
of the receptor-substrate system were saved every 0.2 ps.
Finally a total of 6000 snapshots were saved. Hydrogen bonds
and contacts were automatically identified using contact mod-
ule of CCP4,48 while the other interactions were identified
visually.
(6) Lin, N.-H.; Meyer, M. D. Recent developments in neuronal
nicotinic acetylcholine receptor modulators. Expert Opin. Ther.
Patents 1998, 8, 991–1015.
(7) Bunnelle, W. H.; Dart, M. J.; Schrimpf, M. R. Design of ligands for
the nicotinic acetylcholine receptors: the quest for selectivity. Curr.
Top. Med. Chem. 2004, 4, 299–334.
(8) Glennon, R. A. Medicinal Chemistry of R4β2 nicotinic cholinergic
receptor ligands. Prog. Med. Chem. 2004, 42, 55–123.
(9) (a) Holladay, M. W.; Wasicak, J. T.; Lin, N. H.; Ne, Y.; Ryther,
K. B.; Bannon, A. W.; Buckley, M. J.; Kim, D. J.; Decker, M. W.;
Anderson, D. J.; Campbell, J. E.; Kuntzweiler, T. A.; Donnelly-
Roberts, D. L.; Piattoni-Kaplan, M.; Briggs, C. A.; Williams, M.;
Arneric, S. D. Identification and initial structure-activity relation-
ships of (R)-5-(2-azetidinylmethoxy)-2-chloropyridine (ABT-594),
a potent, orally active, non-opiate analgesic agent acting via
neuronal nicotinic acetylcholine receptors. J. Med. Chem. 1998,
41, 407–412. (b) Bannon, A. W.; Decker, M. W.; Holladay, M. W.;
Curzon, P.; Donnelly-Roberts, D.; Puttfarcken, P. S.; Bitner, R. S.; Diaz,
A.; Dickenson, A. H.; Porsolt, R. D.; Williams, M.; Arneric, S. P. Broad
spectrum, non-opioid analgesic activity by selective modulation of
neuronal nicotinic acetylcholine receptors. Science 1998, 279, 77–81.
(10) (a) Arneric, S. P.; Sullivan, J. P.; Briggs, C. A.; Donnelly-Roberts,
O.; Anderson, D. J.; Raszkiewicz, J. L.; Hughes, M. L.; Cadman, E.
D.; Adams, P.; Garvey, D. S. (S)-3-Methyl-5-(1-methyl-2-pyrroli-
dinyl)isoxazole (ABT-418): a novel cholinergic ligand with cogni-
tion-enhancing and anxiolytic activities: I. In vitro
characterization. J. Pharmacol. Exp. Ther. 1994, 270, 310–318.
(b) Decker, M. W.; Brioni, J. D.; Sullivan, J. P.; Buckley, M. J.; Radek,
R. J.; Raszkiewicz, J. L.; Kang, C. H.; Kim, D. J.; Giardina, W. J.;
Wasicak, J. T. (S)-3-Methyl-5-(1-methyl-2-pyrrolidinyl)isoxazole
(ABT-418): a novel cholinergic ligand with cognition-enhancing and
anxiolytic activities: II. In vivo characterization. J. Pharmacol. Exp.
Ther. 1994, 270, 319–328.
(11) (a) Bourn, W. M.; Keller, W. J.; Bonfiglio, J. F. Comparison
of mescal bean alkaloids with mescaline, Δ9-THC and other
psychotogens. Life Sci. 1979, 25, 1043–1054. (b) El-Seedi, H. R.;
De Smet, P. A. G. M.; Beck, O.; Possnert, G.; Bruhn, J. G. J.
Ethnopharmacol. 2005, 101, 238–242.
(12) (a) Wonnacott, S. Brain nicotine binding sites. Hum. Toxicol. 1987,
6, 343–353. (b) Pabreza, L. A.; Dhawan, S.; Kellar, K. J. [3H]-Cytisine
binding to nicotine cholinergic receptors in brain. Mol. Pharmacol.
1991, 39, 9–12. (c) Hall, M.; Zerbe, L.; Leonard, S.; Freedman, R.
Characterization of [3H]cytisine binding to human brain membrane
preparations. Brain Res. 1993, 600, 127–133.
Hybrid QM/MM Calculations. In the current study, we
used the pseudobond ab initio QM/MM approach, as implemen-
ted in Gaussian03.49 This methodology circumvents the major
deficiency of the conventional link-atom QM/MM approach,
providing a consistent and well-defined ab initio QM/MM
potential energy surface. For the QM/MM calculations, the
receptor-ligand systems resulting from the docking study were
first partitioned into a QM subsystem and then in an MM
subsystem. The reaction system used a smaller QM subsystem,
˚
formed by the ligand and amino acid side chains within 3.5 A
from the ligand, while the rest of the protein (the MM subsystem)
was treated using the AMBER force field. together with a
low memory convergence algorithm. The boundary problem
between the QM and MM subsystems was solved using the
pseudobond approach. With this receptor-substrate QM/MM
system, an iterative optimization procedure was applied to
the QM/MM system, using a B3LYP/3-21G* QM/MM calcula-
tion, leading to an optimized structure for the reactants. The
convergence criterion used was set to obtain an energy gradient
less than 10-4 using the twin-range cutoff method for nonbonded

4356 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Tasso et al.
(13) (a) Papke, R. L.; Heinemann, S. F. Partial agonist properties
of cytisine on neuronal nicotinic receptors containing the beta
2 subunit. Mol. Pharmacol. 1994, 45, 142–149. (b) Chavez-Noriega,
L. E.; Crona, J. H.; Washburn, M. S.; Urrutia, A.; Elliott, K. J.; Johnson,
E. C. Pharmacological characterization of recombinant human and
neuronal nicotinic acetylcholine receptors hR2β2, hR2β4, hR3β2,
hR3β4, hR4β2, hR4β3, andR7 expressed in Xenopus oocytes. J. Phar-
macol. Exp. Ther. 1997, 280, 346–356. (c) Xiao, Y.; Kellar, K. J. The
comparative pharmacology and up-regulation of rat neuronal nicotinic
receptor subtype binding sites stably expressed in transfected mamma-
liam cells. J. Pharmacol. Exp. Ther. 2004, 310, 98–107.
(14) Carbonnelle, E.; Sparatore, F.; Canu Boido, C.; Salvagno, C.;
Baldani Guerra, B.; Terstappen, G.; Zwart, R.; Vijverberg, H.;
Clementi, F.; Gotti, C. Nitrogen substitution modifies the activity
of cytisine on neuronal nicotinic receptor subtypes. Eur. J. Phar-
macol. 2003, 471, 85–96.
(15) (a) Barlow, R. B.; McLeod, L. J. Some studies on cytisine and
its methylated derivatives. Br. J. Pharmacol. 1969, 35, 161–174.
(b) Sloan, J. W.; Martin, W. R.; Bostwik, M.; Hook, R.; Wala, E. The
comparative binding characteristics of nicotinic ligands and their
pharmacology. Pharmacol., Biochem. Behav. 1988, 30, 255–267.
(c) Museo, E.; Wise, R. A. Cytisine induced behavioural activation:
delineation of neuroanatomical locus of action. Brain Res. 1995, 670,
257–263. (d) Rao, T. S.; Correa, L. D.; Reid, R. T.; Lloyd, G. K.
Evaluation of antinociceptive effects of neuronal nicotinic acetylcholine
receptors ligands in rat tail-flick assay. Neuropharmacology 1996, 35,
393–405. (e) Seale, T. W.; Singh, R. S.; Basmadjian, G. Inherited
selective hypoanalgesic response to cytisine in the tail-flick test in CF-1
mice. NeuroReport 1998, 9, 201–205. (f) Ferger, B.; Spratt, C.;
Teismann, P.; Seitz, K.; Kuschinsky, K. Effects of cytisine on hydroxyl
radicals in vitro and MPTP-induced dopamine depletion in vivo. Eur. J.
Pharmacol. 1998, 360, 155–163.
D. Halogenated and isosteric cytisine derivatives with increased affinity
and functional activity at nicotinic acetylcholine receptors. Bioorg.
Med. Chem. Lett. 2005, 15, 1221–1224.
(26) (a) Houlihan, L. M.; Slater, Y.; Guerra, D. L.; Peng, J.-H.; Kuo,
ꢁ
Y.-P.; Lukas, R. J.; Cassels, B. K.; Bermudez, I. Activity of cytisine
and its brominated isosteres on recombinant human R7, R4β2, and
R4β4 nicotinic acetylcholine receptors. J. Neurochem. 2001, 78,
1029–1043. (b) Slater, Y. E.; Houlihan, L. M.; Maskell, P. D.; Exley,
R.; Bermꢁudez, I.; Lukas, R. J.; Valdivia, A. C.; Cassels, B. K.
Halogenated cytisine derivatives as agonists at human neuronal nico-
tinic acetylcholine receptor subtypes. Neuropharmacology 2003, 44,
503–515. (c) Abin-Carriquiry, J. A.; Voutilainen, M. H.; Barik, J.;
ꢁ
Cassels, B. K.; Iturriaga-Vasquez, P.; Bermꢁudez, I.; Durand, C.; Dajas,
F.; Wonnacott, S. C3-halogenation of cytisine generates potent and
efficacious nicotinic receptor agonists. Eur. J. Pharmacol. 2006, 536,
1–11. (d) Abin-Carriquiry, J. A.; Costa, G.; Urbanavicius, J.; Cassels,
B. K.; Rebolledo-Fuentes, M.; Wonnacott, S.; Dojes, F. In vivo
modulation of dopaminergic nigrostriatal pathways by cytisine deriva-
tives: Implications for Parkinson's Disease. Eur. J. Pharmacol. 2008,
589, 80–84.
ꢀ
(27) (a) Marriere, E.; Rouden, J.; Tadino, V.; Lasne, M.-C. Synthesis of
analogues of (-)-cytisine for in vivo studies of nicotinic receptors
using positron emission tomography. Org. Lett. 2000, 2, 1121–
1124. (b) Roger, G.; Lagnel, B.; Rouden, J.; Besret, L.; Valette, H.;
Demphel, S.; Gopisetti, J. M.; Coulon, C.; Ottaviani, M.; Wrenn, L. A.;
Letchworth, S. R.; Bohme, G. A.; Benavides, J.; Lasne, M.-C.;
Bottlaender, M.; Dolle, F. Synthesis of a [2-pyridinyl-18F]-labelled
ꢁ
fluoro derivative of (-)-cytisine as a candidate radioligand for brain
nicotinic R4β2 receptor imaging with PET. Bioorg. Med. Chem. 2003,
11, 5333–5343. (c) Chellappan, S. K.; Xiao, Y.; Tueckmontel, W.;
Kellar, K. J.; Kozikowski, A. P. Synthesis and pharmacological evalua-
tion of novel 9- and 10-substituted cytisine derivatives. Nicotinic
ligands of enhanced subtype selectivity. J. Med. Chem. 2006,
49, 2673–2676. (d) Kozikowski, A. P.; Chellapan, S. K.; Xiao, Y.;
Bajjuri, K. M.; Yuan, H.; Kellar, K. J.; Petukhov, P. A. Chemical
medicine: novel 10-substituted cytisine derivatives with increased
selectivity for R4β2 nicotinic acetylcholine receptors. ChemMedChem
2007, 2, 1157–1161.
(16) (a) Dallemagne, M. G.: Heymans, C. H. Respiratory stimulants. In
Manske, R. H. F. Ed.; The Alkaloids, vol. 5; Academic Press: New
York, 1955; pp 109-139. (b) Zakusov, V. V. Farmacologiia, Medgiz:
Moscow, 1960, p 203.
(17) (a) Paun, D.; Franze, J. Raucherentwohnung mit cytisinhaltigen
“Tabex” Tabletten. Dtsch. Gesundheitswes. 1968, 23, 2088–2091.
(b) Benndorf, S.; Kempe, G.; Scharfenberg, G.; Wendekamm, R.;
Winkelvoss, E. Ergebnisse der medikamentosen Raucherentwohnung
mit Cytisin (Tabex). Dtsch. Gesundheitswes. 1968, 23, 2092–2096.
(c) Seeger, R. Cytisin als Raucherentwohnungmittel? Med. Mon-
atschr. Pharm. 1992, 15, 20–21. (d) Etter, J.-F. Cytisine for smoking
cessation. A literature review and a meta-analysis. Arch Intern. Med.
2006, 166, 1553–1559. (e) Etter, J.-F.; Lukas, R. J.; Benowitz, N. L.;
West, R.; Dresler, C. M. Cytisine for smoking cessation: A research
agenda. Drug Alcohol Dep. 2008, 92, 3–8.
(18) (a) Murakoshi, I.; Fuji, Y.; Takedo, S.; Arai, J. Lupin alkaloids
as antidiabetics. Jpn. Kokai Tokkyo Koho, Japanese Patent
04-295480, 1992; Chem Abstr. 1993, 118, 45733. (b) Murakoshi,
I.; Fuji, Y.; Kawamura, H.; Marayama, H. Lupine alkaloids as antiin-
flammatory agents. Jpn. Kokai Tokkyo Koho, Japanese Patent
04-295479 (1992); Chem Abstr. 1993, 118, 45734
(19) Lippiello, P. M.; Cladwell, W. S. Neurodegenerative diseases
treatment with cytisine compounds (RJ Reynolds Tobacco Co.).
U.S. Patent US 5,242,916, 1993; Chem. Abstr. 1993, 119, 217429
(20) O’Neill, B. T. Preparation of pyridone-fused azabicyclic- or cyti-
sine derivatives for use in addiction therapy. Patent PCT Int. Appl.
WO98 1998, 18, 798, 1998; Chem. Abstr. 1998, 129, 4774
(21) (a) Canu, C.; Sparatore, F. Cytisine derivatives as ligands for
nicotinic receptor subtypes. Ital. Hung. Polish Joint Meet. on
Med. Chem., Giardini Naxos, Sept 28-Oct 1, 1999. Abstracts book
p 206. (b) Canu Boido, C.; Sparatore, F. Synthesis and preliminary
pharmacological evaluation of some cytisine derivatives. Farmaco
1999, 54, 438–451.
(28) (a) Rouden, J.; Ragot, A.; Gouault, S.; Cahard, D.; Plaquevent,
J.-C.; Lasne, M.-C. Regio- and diastereoselective functionalization
of (-)-cytisine: an unusual N-C acyl migration. Tetrahedron:
Asymmetry 2002, 13, 1299–1305. (b) Houllier, N.; Gouault, S.; Lasne,
M.-C.; Rouden, J. Regio and diastereoselective functionalization of
(-)-cytisine. Tetrahedron 2006, 62, 11679–11686.
(29) (a) Kozikowski, A. P.; Musachio, J. L.; Kellar, K. J.; Xiao,Y.;
Wei, Z.-L. Ligands for nicotinic acetylcholine receptors,
and methods of making and using them. Patent PCT WO2005/
000806 A2, 2005. (b) Tilotta, M. C. Novel nicotinic acetylcholine
receptor ligands based on cytisine, ferruginine, anatoxin-a and
choline: in vitro evaluation and structure-activity relationships.
http://hss.ulb.unibonn.de/diss_online/math_nat_fak/2004/tilotta_mar-
ia/0464.pdf.December2, 2005.
(30) (a) Rakhimov, Sh.B.; Vinogradova, V. J.; Mirzaev, Y. R.; Vyrova,
N. L.; Kuzantseva, D. S. Synthesis and biological activity of N-
benzyl derivatives of cytisine. Chem. Nat. Compd. 2006, 42, 462–
469. (b) Safrykine, V. A.; Vinogradova, V. J.; Ambartsumova, R. F.;
Ibragimov, T. F.; Shakhidoyatov, K. M. Thiazole and benzothiazole
derivatives of cytisine. Chem. Nat. Compd. 2006, 42, 470–472.
(c) Frasinyak, M. S.; Turov, A. V.; Vinogradova, V. I.; Khilya, V. P.
Aminomethylation of cytisine by 3-hetaryl-7-hydroxychromones.
Chem. Nat. Compd. 2007, 43, 285–290. (d) Nizri, E.; Irony-Tur-Sinai,
M.; Lavon, I.; Meshulam, H.; Amitai, G.; Brenner, T. IBU-Octyl-
cytisine, a novel bifunctional compound eliciting anti-inflammatory
and cholinergic activity, ameliorates CNS inflammation by inhibition of
T-cell activity. Int. Immunopharm. 2007, 7, 1129–1139.
(22) Nicolotti, O.; Canu Boido, C.; Sparatore, F.; Carotti, A. Cytisine
derivatives as high affinity nAChR ligands: synthesis and com-
parative molecular field analysis. Farmaco 2002, 57, 469–478.
(23) Canu Boido, C.; Tasso, B.; Boido, V.; Sparatore, F. Cytisine
derivatives as ligands for neuronal nicotinic receptors and with
various pharmacological activities. Farmaco 2003, 58, 265–277.
(24) Riganti, L.; Matteoni, C.; Di Angelantonio, S.; Nistri, A.;
Gaimarri, A.; Sparatore, F.; Canu Boido, C.; Clementi, F.; Gotti,
C. Long-term exposure to the new nicotinic antagonist 1,2-bis-N-
cytisinylethane upregulates nicotinic receptor subtypes of SH-
SY5Y human neuroblastoma cells. Br. J. Pharmacol. 2005, 146,
1096–1109.
(25) (a) Imming, P.; Klaperski, P.; Stubbs, M. T.; Seitz, G.; Gundisch,
D. Syntheses and evaluation of halogenated cytisine derivatives
and of bioisosteric thiocytisine as potent and selective nAChR
ligands. Eur. J. Med. Chem. 2001, 36, 375–388. (b) Fitch, R. W.;
Kaneko, Y.; Klaperski, P.; Daly, J. W.; Seitz, G.; Gundisch,
(31) (a) Coe, J. W.; Brooks, P. R.; Vetelino, M. G.; Wirtz, M. C.;
Arnold, E. P.; Huang, J.; Sands, S. B.; Davis, T. I.; Lebel, L. A.;
Fox, C. B.; Shrikhande, A.; Heym, J. H.; Schaeffer, E.; Rollema,
H.; Lu, Y.; Mansbach, R. S.; Chambers, L. K.; Rovetti, C. C.;
Schulz, D. W.; Tingley, F. D.; O’Neill, B. T. Varenicline, an R4β2
nicotinic receptor partial agonist for smoking cessation. J. Med.
Chem. 2005, 48, 3474–3477. (b) Coe, J. W.; Vetelino, M. G.; Bashore,
C. G.; Wirtz, M. C.; Brooks, P. R.; Arnold, E. P.; Lebel, L. A.; Fox, C. B.;
Sands, S. B.; Davis, T. I.; Schulz, D. W.; Rollema, H.; Tingley, F. D.;
O'Neill, B. T. In pursuit of R4β2 nicotinic partial agonists for smoking
cessation: carbon analogs of (-)-cytisine. Bioorg. Med. Chem. Lett.
2005, 15, 2974–2979. (c) Coe, J. W.; Brooks, P. R.; Wirtz, M. C.;
Bashore, C. G.; Bianco, K. E.; Vetelino, M. G.; Arnold, E. P.; Lebel, L.
A.; Fox, C. B.; Tingley, F. D.; Schulz, D. W.; Davis, T. I.; Sands, S. B.;
Mansbach, R. S.; Rollema, H.; O'Neill, B. T. 3,5-Bicyclic aryl-
piperidines: a novel class of R4β2 neuronal nicotinic receptor partial
agonists for smoking cessation. Bioorg. Med. Chem. Lett. 2005, 15,

Article
4889–4897. (d) Steensland, P.; Simms, J. A.; Holgate, J.; Richards, J.
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4357
(41) Balestra, B.; Moretti, M.; Longhi, R.; Mantegazza, R.; Clementi,
F.; Gotti, C. Antibodies against neuronal nicotinic receptor sub-
types in neurological disorders. J. Neuroimmunol. 2000, 102, 89–97.
(42) Munson, P. J.; Rodbard, D. LIGAND: a versatile computerized
approach for characterization of ligand-binding systems. Anal.
Biochem. 1980, 107, 220–239.
K.; Bartlett, S. E. Varenicline, as R4β2 nicotinic acetylcholine receptor
partial agonist, selectively decreases ethanol consumption and
seeking. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 12518–12523.
ꢁ
(e) Eisenberg, M. J.; Filion, K. B.; Yavin, D.; Belisle, P.; Mottillo, S.;
Joseph, B.; Gervais, A.; O'Loughlin, J.; Paradis, G.; Rinfret, S.; Pilote,
L. Pharmacotherapies for smoking cessation: a meta-analysis of rando-
mized controlled trials. CMAY 2008, 179, 135–144.
(43) Ghemical molecular modelling package. [http://www.bioinfor-
matics.org/ghemical.
(44) (a) 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 Empirical Binding Free En-
ergy Function. J. Comput. Chem. 1998, 19, 1639–1662. (b) Huey, R.;
Morris, G. M.; Olson, A. J.; Goodsell, D. S. A Semiempirical Free
Energy Force Field with Charge-Based Desolvation. J. Comput. Chem.
2007, 28, 1145–1152.
(32) (a) Brejc, K.; van Dijk, W. J.; Klaassen, R. V.; Schuurmans, M.;
van der Oost, J.; Smit, A. B.; Sixma, T. K. Crystal structure of an
ACh-binding protein reveals the ligand-binding domain of nico-
tinic receptors. Nature 2001, 411, 269–276. (b) Unwin, N. Refined
˚
structure of the nicotinic acetylcholine receptor at 4 A resolution. J.
ꢀ
Mol. Biol. 2005, 346, 967–989. (c) Le Novere, N.; Grutter, T.;
Changeux, J.-P. Models of the extracellular domain of nicotinic recep-
tors and of agonist and Caþ2 binding sites. Proc. Natl. Acad. Sci. U.S.
A. 2002, 99, 3210–3215.
ꢁ
(45) (a) Hetenyi, C.; van der Spoel, D. Efficient docking of peptides to
proteins without prior knowledge of the binding site. Protein Sci.
ꢁ
2002, 11, 1729–1737. (b) Hetenyi, C.; van der Spoel, D. Blind docking
(33) Bourne, Y.; Talley, T. T.; Hansen, S. B.; Taylor, P.; Marchot,
P. Crystal structure of a Cbtx-AchBP complex reveals essential
interactions between snake alpha-neurotoxins and nicotinic recep-
tors. EMBO J. 2005, 24, 1512–1522.
(34) Bombieri, G.; Meneghetti, F.; Artali, R.; Tasso, B.; Canu Boido,
C.; Sparatore, F. The influence of the nitrogen substitution in three
cytisine derivatives as ligands for the neuronal nAChRs: a struc-
tural and theoretical study. Chem. Biodiversity 2008, 5, 1867–1878.
(35) Artali, R.; Bombieri, G.; Meneghetti, F. Docking of 6-chloropyr-
idazin-3-yl derivatives active on nicotinic acetylcholine receptors
into molluscan acetylcholine binding protein (AChBP). Farmaco
2005, 60, 313–320.
of drug-sized compounds to proteins with up to a thousand residues.
FEBS Lett. 2006, 580, 1447–1450.
(46) Miyamoto, S.; Kollman, P. A. SETTLE;an analytical version of
the shake and rattle algorithm for rigid water models. J. Comput.
Chem. 1992, 13, 952–962.
(47) Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J. G. E. M.
LINCS: a linear constraint solver for molecular simulations, J.
Comput. Chem., 1997, 18, 1463-1472.
(48) Collaborative Computational Project, Number 4. The CCP4 Suite:
Programs for Protein Crystallography. Acta Crystallogr., Sect. D:
Biol Crystallogr. 1994, 50, 760-763.
(36) (a) Dwoskin, L. P.; Xu, R.; Ayers, J. T.; Crooks, P. A. Recent
developments in neuronal nicotinic acetylcholine receptor antago-
nists. Expert Opin. Ther. Patents 2000, 10, 1561–1581. (b) Dwoskin,
L. P.; Crooks, P. A. Competitive neuronal nicotinic receptor antagonists:
a new direction for drug discovery. J. Pharmacol. Exp. Ther. 2001,
298, 395–402.
(37) Fabio, P. F.; Lang, S. A.; Lin, Y.-I; Tomcufcik, A. S. Antiamebic
amidines and sulfonamides of 5- and 6-amino-2,3-bis(4-alkyl-1-
piperazinyl)quinoxalines. J. Med. Chem. 1980, 23, 201–206.
(38) Bixler, L. B.; Nieman, C. Synthesis of b-(4-pyridyl)-DL-alanine and
of b-(4-pyridyl-1-oxide)-DL, D- and L-alanine. J. Org. Chem. 1958,
23, 575–584.
(39) Coombes, R. G.; Johnson, H. D.; Winterton, N. s-Bonded organo-
transition-metal ions. Part I. The formation of air-stable penta-
aquopyridiomethylchromium (III) ions. J. Chem. Soc. 1965, 7029–
7035.
(40) Takamatsu, S; Saito, K; Ohmiya, S.; Ruangrangsi, N.; Murakoshi,
I. Lupin alkaloids from Sophora exigua. Phytochemistry 1991, 30,
3793–3795.
(49) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi,
J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.;
Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H.
P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts,
R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth,
G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.;
Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz,
J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov,
B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin,
R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.;
Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.; Gaussian 03,
revision A.1; Gaussian: Pittsburgh, PA, 2003 .