Modeling of the D2 dopamine receptor arylpiperazine binding site for 1-[2-[5-(1H-benzimidazole-2-thione)]ethyl]-4-arylpiperazines.

Article abstract of DOI:10.1002/ardp.200400901

Docking of several 1-[2-[5-(1H-benzimidazole-2-thione)]ethyl]-4- and 1-benzyl-arylpiperazines to the D(2) dopamine receptor (DAR) was examined. The binding pocket of the D(2) DAR defined according to Teeter and DuRand [1] was extended using the Insight II software. It was found that (i) the interaction of the protonated N1 of the piperazine ring with Asp86, (ii) the hydrogen bond formation between the benzimidazole part of the ligand and Ser141, as well as Ser122, and (iii) the edge-to-face interactions of the aromatic ring or arylpiperazine part of the ligand with Phe178, Tyr216 and Trp182 of the receptor represent the mayor stabilizing forces. Besides, the hydrogen bond acceptor group in position 2 of the phenylpiperazine aromatic ring could form one more hydrogen bond with Trp182. Bulky substituents in position 4 are not tolerated, due to the unfavorable sterical interaction with Phe178. Substituents in positions 2 and 3 are sterically well tolerated. Electron-attractive groups (F, Cl, CF(3), and NO(2)) decreased, while electron donors (-OMe) and the second aromatic ring (naphthyl) increased the binding affinity, as compared to that of the parent compound 1. This can be explained by strong edge-to-face interactions of negative electrostatic surface potential (ESP) in the center of aromatic residues of the ligand with positive-ESP protons in the aromatic residues of the receptor. Thus, besides the salt bridges and hydrogen bonds, edge-to-face interactions significantly contribute to arylpiperazine ligands forming complexes with the D(2) DAR.

Full text of DOI:10.1002/ardp.200400901

ˇ
ˇ
´
502 Soskic et al.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
a
ˇ
´
Vladimir Sukalovic ,
Modeling of the D2 Dopamine Receptor Aryl-
piperazine Binding Site for 1-{2-[5-(1H-benz-
imidazole-2-thione)]ethyl}-4-arylpiperazines
b
´
Mario Zlatovic ,
Deana Andric ,
Goran Roglic ,
b
´
´
b
´
ˇ ´a
Sladjana Kostic-Rajacic ,
a,c
ˇ
´
ˇ
´
Vukic Soskic
Docking of several 1-{2-[5-(1H-benzimidazole-2-thione)]ethyl}-4- and 1-benzyl-
arylpiperazines to the D2 dopamine receptor (DAR) was examined. The binding
pocket of the D2 DAR defined according to Teeter and DuRand [1] was extended
using the Insight II software. It was found that (i) the interaction of the protonated
N1 of the piperazine ring with Asp86, (ii) the hydrogen bond formation between
the benzimidazole part of the ligand and Ser141, as well as Ser122, and (iii) the
edge-to-face interactions of the aromatic ring or arylpiperazine part of the ligand
with Phe178, Tyr216 and Trp182 of the receptor represent the mayor stabilizing
forces. Besides, the hydrogen bond acceptor group in position 2 of the phenyl-
piperazine aromatic ring could form one more hydrogen bond with Trp182. Bulky
substituents in position 4 are not tolerated, due to the unfavorable sterical inter-
action with Phe178. Substituents in positions 2 and 3 are sterically well toler-
ated. Electron-attractive groups (F, Cl, CF3, and NO2) decreased, while electron
donors (-OMe) and the second aromatic ring (naphthyl) increased the binding
affinity, as compared to that of the parent compound 1. This can be explained
by strong edge-to-face interactions of negative electrostatic surface potential
(ESP) in the center of aromatic residues of the ligand with positive-ESP protons
in the aromatic residues of the receptor. Thus, besides the salt bridges and
hydrogen bonds, edge-to-face interactions significantly contribute to arylpiperaz-
ine ligands forming complexes with the D2 DAR.
a Centre for Chemistry,
Institute for Chemistry,
Technology and Metallurgy,
11000 Belgrade, Serbia
Montenegro
b Faculty of Chemistry,
University of Belgrade,
11000 Belgrade, Serbia
Montenegro
c ProteoSys AG,
55129 Mainz, Germany
Keywords: Arylpiperazines; D2 receptor; Modeling; Interaction; Binding pocket
Received: April 30, 2004; Accepted: July 1, 2004 [FP901]
DOI 10.1002/ardp.200400901
Introduction
pared ligands for the D2 DAR depends on both the
benzimidazole structure and the structure of the arylpi-
During the recent years, the identification of multiple
dopamine (DA) receptor subtypes has been ac-
companied by the development of agents that alter DA
neurotransmission [2]. For many years, the D2 DAR
was a major target for neurobiological research and
drug development, since dopamine antagonists have
been proven to be effective antipsychotics [3]. In the
course of a program aimed at the discovery of new
dopaminergic ligands, we have synthesized a series
of benzimidazoles that could be considered as non-
catechol bioisosteres of catecholamines [4]. The most
active compounds of this type were obtained by con-
necting the benzimidazolethione ring through the flex-
ible ethylene linker with N-arylpiperazines, which af-
forded compounds of the general structure I (Figure
1). It was noticed that the binding affinity of the pre-
perazine part of the molecule, but the effect of the lat-
ter was more pronounced. However, the physico-
chemical basis of the above interactions is still far from
being fully understood. This prompted us to study the
effect of the electron density distribution (electrostatic
surface potential; ESP) in the arylpiperazine part of
this class of ligands on their binding affinity for the D2
DAR. The binding pocket of the D2 receptor was de-
ˇ
´
ˇ ´
Correspondence: Vukic Soskic, ProteoSys AG, Carl
Zeiss Str. 51, D-55129 Mainz, Germany. Phone: +49
6131 5019246, Fax: +49 6131 5019211, e-mail:
vukic.soskic@proteosys.com
Figure 1. Structure of 1-{2-[5-(1H-benz-imidazole-2-
thione)]ethyl}-4-arylpiperazines.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
Modelling the D2 Dopamine Receptor Arylpiperazine Binding Site 503
Table 1. Chemical structures of 1-{2-[5-(1H-benzimidazole-2-thione)]ethyl}-4-arylpiperazine and 1-benzyl-4-aryl-
piperazine ligands tested for docking in the D2 DAR binding pocket. Compounds 9Ϫ12 were newly synthesized,
while all others were already described. For references, see text..
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

ˇ
ˇ
´
504 Soskic et al.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
a) EtOH, SnCl2, reflux; b) dioxane, 1N NaOH, di-tert-butyl-dicarbonate; c) DMF, K2CO3,
KI, substituted piperazines, 80 °C; d) EtOH, 4N HCl; e) EtOH, KOH, CS2, reflux; f) EtOH,
N2H4, Ra-Ni
Scheme 1. Synthesis of 1-{2-[5-(1H-benzimidazole-2-thione)]ethyl}-4-arylpiperazines (9؊12).
fined according to Teeter and DuRand [1]. Special at-
tention has been paid to hydrophobic-type interactions
(e.g. stacking or edge-to-face interactions), which play
a significant role in the formation of the receptor-ligand
complexes [1, 5]. These attractive interactions occur
between aromatic moieties devoid of polar substitu-
ents. “Edge-to-face” interactions, though modest in
energy terms, can play an important role in diverse
areas such as protein folding, base pair stacking in
DNA, host-guest binding in supramolecular assembl-
ies, crystal engineering, drug-receptor interactions,
and other molecular recognition processes [6]. Ener-
getically, they can stabilize the system by up to Ϫ2.5
kcal/mol [7] Edge-to-face interactions between recep-
tors and their ligands should be exclusively dependent
on the shape of the ligand molecule and its ability to
interact with the aromatic residues in the binding
pocket of the receptor [6, 7]. Complementarities of
negative ESP in the center of aromatic residues of the
ligands and positive ESP of the protons in aromatic
residues of the receptor, as well as a proper orien-
tation of molecular entities forming the complex, are
prerequisites for this type of interactions. The data ob-
tained throughout the present study could serve as a
useful basis for further rational design of D2 receptor li-
gands.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
Modelling the D2 Dopamine Receptor Arylpiperazine Binding Site 505
Results and discussion
Table 2. List of amino acids considered to be part of
the 1-{2-[5-(1H-benzimidazole-2-thione)]ethyl}-4-aryl-
piperazine binding site in the D2 DAR.
Several new 1-{2-[5-(1H-benzimidazole-2-thione)]-
ethyl}-4-arylpiperazines (compounds 9Ϫ12; Table 1)
were synthesized as shown in Scheme 1, and their
affinity for binding the D2 DAR was determined.
Shortly, 4-(2-chloroethyl)-2-nitroaniline (21) was re-
duced with stannous chloride in absolute ethanol, and
the resulting diamine 22 was converted into di-tBOC
derivative 23, using di-tertbutyldicarbonate. Com-
pound 23 readily alkylated substituted piperazines in
the presence of sodium carbonate and potassium iod-
ide in DMF. Compound 30 was prepared in the same
manner, directly from nitroaniline 21. Diamines 27Ϫ29
were obtained by hydrolyzing di-tBOC derivatives
24؊26 with 4 N HCl in ethanol. Diamine 31 was pro-
duced by reducing nitroaniline 30 with Ra-Ni/hydra-
zine. Benzimidazole-2-thiones 9Ϫ12 were synthesized
from the corresponding diamines (27؊29 and 30) with
CS2/KOH in EtOH.
Residue Position Residue Position Residue Position
Asp
Trp
Phe
Val
Asp
Met
Cys
Ser
Trp
46
56
82
83
86
89
90
93
115
Ser
Ser
Ser
Ser
Phe
Phe
Cys
Phe
Phe
118
122
141
144
145
178
181
185
186
His
Tyr
Phe
Thr
Gly
Tyr
189
208
211
212
215
216
219
222
Ser
Asn
The main features of the D2 DAR model shown in Fig-
ure 2a, using compound 1 as a ligand, were (i) close
interaction of protonated N1 of the piperazine ring with
In binding experiments, synaptosomal membranes of
the bovine caudate nuclei as a source of the D2 DAR
and [3H]spiperone as a specific radioligand were used.
The new compounds, along with a number of pre-
viously described ligands (Table 1), were tested for
their docking in the D2 DAR binding site.
˚
Asp86 (calculated distance 1.68 A), (ii) hydrogen bond
formation between the benzimidazole part of the li-
gand and Ser141 and Ser122, and (iii) edge-to-face
interactions of the aromatic ring or the arylpiperazine
part of the ligand with Phe178, Tyr216 and Trp182 of
the receptor. Similar results were obtained with 2,3-
dimethylphenyl and naphthyl substituents in the pip-
erazine ring (compounds 2 and 3, respectively). Gen-
erally, introduction of the substituent in position 2 of
the phenyl ring in the piperazine part of a ligand led to
the same docking to the receptor as with ligand 1. This
holds true for all ligands tested in the present study
(compounds 4Ϫ6, as well as ligand 2). In addition, 2-
methoxy derivative 4 could form one more hydrogen
bond with Trp182.
The binding pocket of the D2 DAR was defined accord-
ing to the model of D2 DAR proposed by Teeter and
DuRand [1]. Initially, the model of the D2 DAR trans-
membrane helices was constructed directly from the
bacteriorhodopsin (bR) coordinates derived from two-
dimensional electron diffraction experiments, but the
orientations of all TM domains were subsequently ad-
justed in order to mimic the topology of the TM do-
mains of rhodopsin [8]. This model was tested for its
ability to accommodate rigid agonist and semi-rigid an-
tagonist molecules which were docked into the puta-
tive binding pocket with stabilizing interactions. The
model is consistent with structure-activity relationships
of agonists and antagonists that interact with the re-
ceptor [9] and with site-directed mutagenesis data
[9Ϫ11].
Ligands with substituents in position 4 of the arylpiper-
azine ring (5, 8 and 11) could not dock to the receptor
as previously described for compound 1. This is the
result of an unfavorable steric interaction of bulky sub-
stituents with Phe178 in the receptor binding pocket
(Figure 2c). As a consequence, formation of a salt
bridge between Asp86 and the protonated N1 of the
piperazine ring is hindered. The calculated distance
Docking of 1-{2-[5-(1H-benzimidazole-2-thione)]ethyl}-
4-arylpiperazine to the thus defined binding site could
not explain the experimentally obtained values for the
corresponding ligands. Therefore, the binding pocket
was enlarged using the Insight II software, by taking
into account all receptor amino acid side groups (Table
2) that could interact after initial positioning of the li-
gands against amino acid residues Asp86 and
Ser141. The binding pocket designed in this way pro-
vided results matching the obtained experimental re-
sults.
˚
between these two entities was increased to 3.51 A.
4-Fluorophenylpiperazine derivative 12 did not fit into
this scheme, since fluorine is similar in size to a hydro-
gen atom. The decrease in affinity of ligand 12 in com-
parison with that of compound 1 can be explained in
terms of a strong negative inductive effect of the fluor-
ine atom, reducing the energy of edge-to-face interac-
tions.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

ˇ
ˇ
´
506 Soskic et al.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
Figure 2. Schematic representation of the interaction of ligands 1Ϫ4 with the D2 dopamine receptor. The 3D
model describes a possible interaction of compounds 1 (A), 2 (B), 3 (C) and 4 (D) and the theoretical dopamine
D2 receptor model.
Docking analyses of the ligands with substituents in
position 3 of the piperazine phenyl ring (7, 10, 13 and
17) revealed that the substituents in this position are
tolerated, since no large reduction of affinity was ob-
served. In contrast, substituents with electron with-
drawal effect in this position, like trifluoromethyl (13),
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
Modelling the D2 Dopamine Receptor Arylpiperazine Binding Site 507
Figure 3. Electrostatic surface potentials of several 1-{2-[5-(1H-benzimidazole-2-thione)]ethyl}-4-arylpiperazines.
For simpler comparisons, the ESP values were mapped on the electron density surface. Values in blue indicate
a strong negative ESP, whereas those in red correspond to a strong positive ESP. Compounds 1 (A), 3 (B), 6
(C), and 4 (D).
chloro (7) and nitro groups (10), affect the affinity by
decreasing the electron density in the benzene ring of
these ligands.
lower compared to that of compound 4 (Table 2). On
the other hand, ligand 6 shows the same activity as
compounds 1 and 2, pointing out that some more fac-
tors, apart from edge-to-face interactions, are playing
a role in the explanation of structure-activity relation-
ships in this part of molecule.
From data presented in the literature, it is obvious that
the receptor-ligand complexes presented here are in
agreement with the published site-directed mu-
tagenesis data, as far as the benzimidazole D2 DAR
binding domain and Asp86 are concerned [9Ϫ11]. To
our knowledge, such data are not available for the ar-
ylpiperazine binding part of D2 DAR.
For a further evaluation of the effects of electron with-
drawing groups on dopaminergic activity, several new
benzimidazole arylpiperazines (compounds 9Ϫ13,
Scheme 1) were synthesized. Groups that differ in size
and electron withdrawal properties (fluoro, nitro, and
trifluoromethyl) were chosen. These substituents were
introduced at positions 2, 3 and 4 of the phenyl group
attached to the piperazine ring of parent compound 1.
Regardless of the position of substitution, reduction of
the binding affinity was expected. All new compounds
behaved as predicted, with the exception of the 2-nitro
derivative 9. This might have been expected for the 2-
ESP calculations on compounds 1Ϫ4 demonstrated
that they were involved in edge-to-face interactions
with the receptor molecule (Figure 3a, b, d). Exchange
of the 2-methoxy group of ligand 4 with the isosteric
chlorine atom (compound 6) partially reduced the elec-
tron density in the aromatic ring, thereby reducing the
energy of edge-to-face interactions (Figure 3c). As a
consequence, the affinity of ligand 6 was 12 times
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

ˇ
ˇ
´
508 Soskic et al.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
nitro substitution, since this group forms one additional
hydrogen bond with Trp182 (similar to the one pro-
posed for the 2-OMe group of compound 4). This ad-
ditional hydrogen bond can compensate for the nega-
tive effect of the nitro group on edge-to-face interac-
tions.
a salt bridge with Asp86, (ii) edge-to-face interactions
with aromatic residues (Tyr216, Phe178, and Trp182),
and (iii) hydrogen bonds formed between Trp182 and
2-OMe (i.e. the 2-nitro substituent of the ligands).
The results of the docking analyses, along with the
biological activities, are listed in Table 3. Generally, all
compounds of this class had a lower activity than their
benzimidazole counterparts. Benzyl ligands (14Ϫ20)
docked according to the proposed model. A salt bridge
between the protonated N1 of the piperazine ring and
Asp86, edge-to-face interactions of the aromatic ring
or the arylpiperazine part of the ligand and Phe178,
Tyr216 and Trp182 of the receptor, and a hydrogen
bond between Trp182 and 2-OMe or 2-nitro substitu-
ents (compounds 16 and 19, respectively) represent
the interacting points with the receptor (Figure 4). In-
troduction of bulky substituents at position 4 (4-OMe-
The results of the docking analyses, along with the
biological activities of all tested compounds, are sum-
marized in Table 3.
To determine the contribution of the arylpiperazine part
of the ligands to D2 DAR binding, 1N-benzyl com-
pounds 14Ϫ20 were prepared, according to previously
published procedures (for references see Experimen-
tal). These ligands lack the benzimidazole part, which
eliminates the contribution of hydrogen bonds through
serines in TM IV and V. Driving forces for the formation
of the complex with the D2 receptor molecule were (i)
Table 3. Correlation of the arylpiperazine-D2 DAR binding affinity and the type of ligand-receptor interactions.
Ligand
No.
Asp86
Ser141
Ser122
Phe178
Tyr216
Trp182
Ki (nM)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
ETF
ETF
ETF
ETF
bulk
ETF
n/a bulk
bulk
n/a
ETF
ETF
ETF
ETF
ETF
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
ETF
ETF
ETF
ETF
ETF
n/a
ETF
ETF
ETF
15.7 2.0
15.8 2.3
3.4 0.4
1.7 0.4
110 12
20.7 2.2
250 35
500 37
hydrogen bond
ETF
n/a
n/a
n/a
salt bridge hydrogen bond
n/a
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
hydrogen bond 11.79 0.9
n/a
bulk
n/a
n/a
n/a
n/a
n/a
ETF
159.3 15
>1000
salt bridge hydrogen bond
n/a
salt bridge hydrogen bond hydrogen bond
salt bridge hydrogen bond hydrogen bond
73.8 2.1
134 15
639 43
580 36
28 4.2
577 37
>1000
n/a
salt bridge
salt bridge
salt bridge
salt bridge
salt bridge
salt bridge
salt bridge
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
ETF
ETF
ETF
ETF/bulk
bulk
n/a
ETF
hydrogen bond
ETF
ETF
hydrogen bond
n/a
198 26
>1000
bulk
n/a
Ki values are the means of three independent experiments done in triplicate performed at eight competing ligand
concentrations (0.1 nMϪ0.1 mM) and 0.2 nM [3H]spiperone. The observed salt bridge lengths ranged from 1.65
˚
˚
to 1.95 A for active compounds and exceeded 1.95 A for inactive compounds. For effective hydrogen bonds, the
˚
˚
observed lengths were under 2.5 A, while the criterion for ETF interactions was a length under 3 A measured
from the phenyl centered part of the ligand to the nearest hydrogen atom of the listed amino acid residues. *ETF
is the abbreviation for edge-to-face interactions.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
Modelling the D2 Dopamine Receptor Arylpiperazine Binding Site 509
Figure 4. Schematic representation of the interaction of 1-benzyl-4-arylpiperazine ligands and the D2 DAR. Sche-
matic model of the proposed interaction of the studied compounds 16 (A) and 19 (B) with the D2 receptor.
phenyl in 18 and 4-nitro-phenyl in 20) completely
blocked the interaction of the ligands with the receptor.
The ligands that can form a hydrogen bond with
Trp182 (2-OMe-phenyl, 16, and 2-nitro-phenyl, 19)
were the most active. They were followed by those
that can take part in edge-to-face interactions (phenyl,
naphthyl and 3-MeO-phenyl; i.e. 14, 15 and 17,
respectively), whereas the ligands with bulky substitu-
ents introduced at position 4 (18 and 20) were inactive.
Conclusions
The results of our docking studies on 1-{2-[5-(1H-
benzimidazole-2-thione)]ethyl}-4-aryl-piperazine-D2
DAR complexes revealed that (i) close interaction of
the protonated N1 of the piperazine ring with Asp86,
(ii) hydrogen bond formation between the benzimida-
zole part of the ligand and Ser141, as well as Ser 122,
and (iii) edge-to-face interactions of the aromatic ring
or the arylpiperazine part of the ligand and Phe178,
Tyr216 and Trp182 of the receptor represent the main
stabilizing forces. In addition, the 2-methoxy derivative
4 could form one additional hydrogen bond with
Trp182.
Löber et al. [12] used a similar strategy for rationally
based efficacy tuning of 2-[4-(4-chlorophenyl)pipera-
zin-1-ylmethyl]pyrazolo[1,5-a]pyridines of D4 DAR ac-
tivity, which resulted in a different docking model from
the one presented in our paper. This is probably due
to the different subtype of DAR and the different arylpi-
perazine ligands considered. Beyond that, a vast
amount of literature about structure-activity relation-
ship of the arylpiperazine class of dopaminergic li-
gands exists (as an example, see the paper of Cha et
al. [13] and references cited therein). The aim of this
paper is not to give a general explanation for all arylpi-
perazine-D2 DAR interactions, but is limited to the fam-
ily of arylpiperazines presented, which are currently
under study in our laboratory.
Bulky substituents in position 4 of the aromatic part of
the phenylpiperazine ring are not tolerated because of
unfavorable steric interactions with Phe178. This re-
sult agrees well with the data of Simpson et al. [14]
and Löber et al. [15].
Substituents in position 2 and 3 of phenylpiperazine
are sterically well tolerated. Electron-attractive groups
such as F, Cl, CF3 and NO2 decreased the binding
affinity, while electron donors like -OMe and the se-
cond aromatic ring (naphthyl) increased the affinity in
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

ˇ
ˇ
´
510 Soskic et al.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
Modelling of the D2 DAR
comparison with the parent compound 1. These ef-
fects can be explained by strong edge-to-face interac-
tions of negative ESP in the center of aromatic resi-
dues of the ligands and positive ESP of the protons of
the receptor aromatic residues.
The D2 DAR model proposed by Teeter and DuRand [1, 8]
was used to explain structure-activity relationships of our li-
gands. Initially, the model of the D2 DAR transmembrane heli-
ces was constructed directly from the bR coordinates derived
from two-dimensional electron diffraction experiments, but
the orientations of all TM domains were subsequently ad-
justed in order to mimic the topology of the TM domains of
rhodopsin [10]. Primary sequences of bR and D2 DAR were
aligned, superimposing conservative amino acid residues.
The seven TM helices were arranged using constructed heli-
cal wheels, matching hydrophobic and hydrophilic inter-heli-
ces interactions. Energy minimization was performed on the
obtained model, to relieve all close van-der-Waals contacts.
No global energy minimization was performed, nor were mo-
lecular dynamics run.
The presented data give us a good explanation of the
behavior of this type of ligands, which can be used as
the basis for further rational design of active dopami-
nergic compounds.
Acknowledgments
This work was supported by the Ministry for Science,
D2 DAR complexes were done using the Docking module
within the Insight II software (Accelerys Inc., Cambridge, UK)
on an SGI Octane 2 workstation (Silicon Graphics Inc., Moun-
tain View, USA). Docking of the ligands described here was
performed as follows: Initially, using SA docking algorithm
100, structures were generated applying a Monte Carlo
method. Each structure was further minimized for 4000
Technology and Development of Serbia, Grant #1698
ˇ
´
(V. Sukalovic, M. Z., G. R., S. H., S. K.-R. and D. A.)
ˇ
ˇ
´
and ProteoSys AG, Mainz, Germany (V. Soskic).
˚
cycles or until 0.01 kcal/mol/A was reached. Minimization was
Experimental
performed by fixing all the protein backbone atoms and by
keeping ligand and amino acid residues in the binding site
flexible. In this way, the relaxation of van-der-Waals interac-
tions was permitted. Subsequently, all structures were filtered
using the general rule that the best structure is the one with
the shortest salt bridge between the ligand and Asp86 and
with a maximum number of hydrogen bonds with the D2 DAR.
The obtained results were visualized using DS View software
(Accelerys Inc., Cambridge, UK).
General
A Boetius PHMK apparatus (VEB Analytic, Dresden, Ger-
many) was used to determine melting points, presented here
1
as uncorrected. H-NMR and 13C-NMR spectra, recorded on
a Gemini 2000 spectrometer (Varian, Palo Alto, CA, USA)
with CDCl3 as solvent unless otherwise stated, are reported
in ppm downfield from the internal standard tetramethyl-
silane.
Chemistry
IR spectra were run on a Perkin Elmer 457 Grating Infrared
Spectrophotometer (Perkin Elmer, Beaconsfield, UK). Mass
spectra were determined by a Finnigan Mat 8230 mass spec-
trometer (Finnigan, Bremen, Germany). High-resolution mass
spectra were acquired on a Bruker Biflex MALDI TOF
(Bruker, Bremen, Germany). For analytical thin-layer chroma-
tography, F-256 plastic-backed thin-layer silica gel plates
from E. Merck (Darmstadt, Germany) were used. Chromato-
graphic purifications were performed on Merck-60 silica gel
columns, 230Ϫ400 mesh ASTM, under medium pressure
(MPLC). Solutions were routinely dried over anhydrous
Na2SO4 prior to evaporation.
1-{2-[5-(1H-benzimidazole-2-thione)]ethyl}-4-arylpiperazines
1Ϫ8 were prepared as previously described [21]. 1-Benzyl-
4-aryl-piperazines 14Ϫ20 were synthesized according to the
procedures of other authors as follows: compound 14 [22],
15 [23], 16Ϫ18 [24], 19 [25] and 20 [26].
Synthesis of 1-chloro-2-(3,4-di-tBOC-phenyl)ethane (23)
Stannous chloride (47.5 g, 0.23 mol) was added to a solution
of 4-(2-chloroethyl)-2-nitroaniline (21) (9.5 g, 40 mmol) in
absolute ethanol (85 mL). After refluxing for 4 h, the solution
was poured into ice, made alkaline with 5 M NaOH and ex-
tracted with EtOAc. Extracts were dried, the solvent was re-
moved in vacuo, and the resulting diamine 22 was immedi-
ately used without further purification. Obtained diamine 22
(3.5 g, 21 mmol) was dissolved at 0°C in a mixture of dioxane
(65 mL) and 1 N NaOH (65 mL). To this solution, di-tert-butyl
dicarbonate (6.9 g, 31.5 mmol) was added at 0°C. Cooling
was stopped after 2 h, and the reaction mixture was stirred
overnight at room temperature. Excess of solvent was evapo-
rated in vacuo, and the obtained residue was extracted with
CH2Cl2. The obtained product was purified by MPLC using
CH2Cl2 as the eluent. Yield: 4.5 g (60%), colorless oil; IR
(KBr): (cmϪ1) 773, 1065, 1160, 1276, 1480, 1523, 1684,
2988, 3330. 1H-NMR: δ 1.51 (s, 18H, CH3); 3.01 (t, J = 7.6
Hz, 2H, CH2Cl); 3.67 (t, J = 7.6 Hz, 2H, ArCH2); 6.61 (s, 1H,
NH); 6.80 (s, 1H, NH); 6.96 (dd, J = 6.2 Hz, J = 2 Hz, 1H,
ArH); 7.35Ϫ7.43 (m, 2H, ArH). MS m/z 370 (M+). Anal. calc.
for (C18H27ClN2O4): C, 58.29; H, 7.34; N, 7.55; found: C,
58.24; H, 7.38; N, 7.51.
Molecular modeling
Modeling of the ligands
Ligand models were constructed using Hyperchem version
7.0 software (Hypercube, Inc. Gainesville, USA) and an in-
built PM3 routine for molecule geometry optimization. It was
postulated that the ligands are bound to the receptors in pro-
tonated form [16, 17]; therefore, a formal charge of +1 was
added to the piperazine nitrogen (1N). The results obtained
were further optimized in a Gaussian 98, Rev. A.9 (Gaussian,
Inc., Pittsburgh, USA). ESP were calculated in a Gaussian G
98W using the DFT B3LYP method and a 6-31g basis set
[18, 19]. The ESP cube output from Gaussian G 98W was
visualized in a gOpenMol software [20] following the recom-
mended Gaussian procedure to display the calculated
properties.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
Modelling the D2 Dopamine Receptor Arylpiperazine Binding Site 511
General procedure for the synthesis of 1-{2-[3,4-di(tBOC-
amino)phenyl]ethyl}-4-aryl-piperazines (24Ϫ26) and 1-[2-(3-
nitro-4-aminophenyl)ethyl]-4-(4-fluorophenyl)piperazine (30)
General procedure for the synthesis of 1-{2-[5-(1H-benzimid-
azole-2-thione)]ethyl}-4-arylpiperazines (9Ϫ12)
Carbon disulfide (0.36 mL, 6.0 mmol) and KOH solution (0.37
g in 0.90 mL water) were added to a solution of 3.0 mmol of
either one of diamines 27؊29 and 31 in 5.0 mL EtOH. After
refluxing for 3 h, the solvent was removed in vacuo. The reac-
tion mixture was diluted with 100 mL ice-cold water; the ob-
tained suspension was neutralized with 2 M HCl, and the pre-
cipitate was collected by filtration. The resulting crude benzi-
midazolethiones 9؊12 were purified by recrystallization from
hot EtOH.
To a solution of 10.0 mmol of either arylpiperazine in 50.0 mL
DMF, 12.0 mmol of 1-chloro-2-arylethane (21 or 23), 6.0 g
K2CO3 and 0.1 g KI were added. The mixture was stirred at
80°C for 12 h. After cooling, the precipitate was removed and
the filtrate was evaporated in vacuo. The residue was dis-
solved in CH2Cl2, and the obtained products were purified by
MPLC using CH2Cl2 as the eluent.
(24): Yield: 3.2 g (58%), yellow oil; IR (KBr): (cmϪ1) 766,
1048, 1156, 1245, 1332, 1518, 1710, 2977, 3352. 1H-NMR:
δ 1.52 (s, 18H, CH3); 2.65Ϫ2.70 (m, 6H); 2.78Ϫ2.84 (m, 2H);
3.08Ϫ3.13 (m, 4H); 6.60 (s, 1H, NH); 6.76 (s, 1H, NH);
6.94Ϫ7.18 (m, 3H, ArH); 7.32Ϫ7.52 (m, 3H, ArH); 7.62 (dd,
J = 6.4 Hz, J = 1.6 Hz, 1H, ArH). MS m/z 541 (M+). Anal.
calc. for (C28H39N5O6): C, 62.09; H, 7.26; N, 12.93; found: C,
62.07; H, 7.30; N, 12.95.
(9): Yield: 0.56 g (50%); m.p.: 238Ϫ240°C. IR (KBr): (cmϪ1
)
742, 1110, 1188, 1343, 1461, 1490, 1518, 1603, 2834, 2948,
3069. 1H-NMR (DMSO): δ 2.49Ϫ2.56 (m, 2H); 2.76Ϫ2.83 (m,
4H); 2.97Ϫ3.01 (m, 6H); 6.98Ϫ7.16 (m, 4H, ArH); 7.32 (dd,
J = 7.2 Hz, J = 1 Hz, 1H, ArH); 7.58 (t, J = 2.6 Hz, 1H, ArH);
7.78 (d, J = 6.8 Hz, 1H, ArH); 12.47 (s, 2H, NH). MS m/z 384,
142 (MH+). Anal. calc. for (C19H21N5O2S): C, 59.51; H, 5.52;
N, 18.26; found: C, 59.54; H, 5.49; N, 18.30.
(25): Yield: 3.6 g (66%), yellow oil; IR (KBr): (cmϪ1) 760,
1050, 1159, 1244, 1347, 1528, 1725, 2977, 3402. 1H-NMR:
δ 1.52 (s, 18H, CH3); 2.63Ϫ2.73 (m, 6H); 2.79Ϫ2.86 (m, 2H);
3.30Ϫ3.35 (m, 4H); 6.60 (s, 1H, NH); 6.80 (s, 1H, NH); 6.97
(dd, J = 6.6 Hz, J = 1.8 Hz, 1H, ArH); 7.19 (dd, J = 6 Hz, J =
1.6 Hz, 1H, ArH); 7.31Ϫ7.44 (m, 3H, ArH); 7.63Ϫ7.72 (m,
2H, ArH). MS m/z 541 (M+). Anal. calc. for (C28H39N5O6): C,
62.09; H, 7.26; N, 12.93; found: C, 62.12; H, 7.30; N, 12.96.
(10): Yield: 0.45 g (40%); m.p.: 235Ϫ237°C. IR (KBr): (cmϪ1
741, 1129, 1188, 1245, 1343, 1461, 1490, 1522, 1617, 2817,
1952, 3129. 1H-NMR (DMSO):
2.50Ϫ2.60 (m, 6H);
2.78Ϫ2.85 (m, 2H); 3.25Ϫ3.29 (m, 4H); 6.98Ϫ7.08 (m, 3H,
ArH); 7.40Ϫ7.65 (m, 4H, ArH); 12.46 (s, 2H, NH). MS m/z
384, 139 (MH+). Anal. calc. for (C19H21N5O2S): C, 59.51; H,
5.52; N, 18.26; found: C, 59.49; H, 5.50; N, 18.30.
)
δ
(11): Yield: 0.42 g (39%); m.p.: 232Ϫ234 °C. IR (KBr): (cmϪ1
)
(26): Yield: 4.9 g (90%), yellow oil; IR (KBr): (cmϪ1) 754,
1049, 1162, 1246, 1326, 1598, 1730, 2978, 3314. 1H-NMR:
δ 1.51 (s, 18H, CH3); 2.60Ϫ2.67 (m, 6H); 2.76Ϫ2.83 (m, 2H);
3.41Ϫ3.47 (m, 4H); 6.80Ϫ6.85 (m, 3H, NH, ArH); 6.96 (dd,
J = 6.4 Hz, J = 1.8 Hz, 1H, ArH); 7.32 (d, J = 8.2 Hz, 1H,
ArH); 7.45 (s, 1H, ArH); 8.12 (d, J = 9.4 Hz, 2H, ArH). MS
m/z 541 (M+). Anal. calc. for (C28H39N5O6): C, 62.09; H, 7.26;
N, 12.93; found: C, 62.05; H, 7.22; N, 12.96.
740, 1118, 1240, 1325, 1490, 1520, 1609, 2824, 2950, 3102.
1H-NMR (DMSO): δ 2.41Ϫ2.96 (m, 8H); 3.21Ϫ3.38 (m, 4H);
6.96Ϫ7.43 (m, 5H, ArH); 8.11 (d, J = 9.2 Hz, 2H, ArH); 12.46
(s, 2H, NH). MS m/z 384, 144 (MH+). Anal. calc. for
(C19H21N5O2S): C, 59.51; H, 5.52; N, 18.26; found: C, 59.50;
H, 5.56; N, 18.28.
(12): Yield: 0.51 g (42%); m.p.: 197Ϫ199°C. IR (KBr): (cmϪ1
)
816, 1128, 1193, 1223, 1348, 1461, 1510, 1618, 2837, 2958,
3072. 1H-NMR (DMSO): δ 2.52Ϫ2.84 (m, 8H); 3.09Ϫ3.20 (m,
4 H); 6.05 (s, 2H, NH2); 6.92Ϫ7.10 (m, 7H, ArH); 12.46
(s, 2H, NH). MS m/z 357, 149 (MH+). Anal. calc. for
(C19H21FN4S): C, 64.02; H, 5.94; N, 15.72; found: C, 64.06;
H, 5.90; N, 15.75.
(30): Yield: 1.8 g (53%), colorless oil; IR (KBr): (cmϪ1) 820,
1189, 1342, 1515, 2945. 1H-NMR: δ 2.64Ϫ2.68 (m, 6H);
2.73Ϫ2.81 (m, 2H); 3.41Ϫ3.45 (m, 4H); 6.05 (s, 2H, NH2);
6.52Ϫ6.81 (m, 5H, ArH); 7.24 (d, J = 8 Hz, 1H, ArH); 7.94 (s,
1H, ArH). MS m/z 344 (M+). Anal. calc. for (C18H21FN4O2): C,
62.78; H, 6.15; N, 16.27; found: C, 62.77; H, 6.19; N, 16.30.
Synaptosomal membrane preparation, binding assays and
data analysis
General procedure for the synthesis of 1-[2-(3,4-diamino-
phenyl)ethyl]-4-arylpiperazines (27Ϫ29)
Synaptosomal membranes of the bovine caudate nuclei used
as a source of the dopamine D2 receptors were prepared ex-
actly as described previously [27]. [3H]spiperone (spec. act.
70 Ci mmolϪ1) used to label the D2 receptor was purchased
from Amersham Buchler GmbH (Braunschweig, Germany).
Briefly, [3H]spiperone binding was assayed in a binding buffer
at 37°C for 20 min in a total volume of 0.5 mL. Binding of the
radioligand to 5-HT2 receptors was prevented by addition of
50 nM ketanserin. Ki values were determined by competition
binding at 0.2 nM radioligand, and eight to ten concentrations
of each ligand were tested (0.1 nMϪ0.1 mM). Nonspecific
binding was measured in the presence of 1.0 mM (+)-butacla-
mol. The reaction was terminated by rapid filtration through
Whatman GF/C filters, before being washed three times with
5.0 mL ice-cold incubation buffer. Radioligand binding for
each concentration of the tested compounds was determined
in triplicate. Retained radioactivity was measured by introduc-
ing dry filters into 10 mL of toluene-based scintillation liquid
and counting in a 1219 Rackbeta Wallac scintillation counter.
Competition binding data were analyzed by the non-linear
least-squares curve-fitting program LIGAND [28].
Hydrochloric acid (37%, 10 mL) was added to a stirring solu-
tion of 1-{2-[3,4-di(tBOC-amino)phenyl]ethyl}-4-arylpiperaz-
ines 24؊26 (5.0 mmol) in 20 mL EtOH at room temperature.
After 60 min, solvent and HCl were evaporated in vacuo. The
residue was extracted with a mixture of 20 mL 10% NaHCO3
and 20 mL chloroform; the organic phase was separated,
dried over Na2SO4 and evaporated in vacuo. The products
obtained were used without further purification for the syn-
thesis of compounds 9Ϫ11.
Synthesis of 1-[2-(3,4-diaminophenyl)ethyl]-4-(4-fluorophe-
nyl)piperazine (31)
Ra-Ni (0.3 g) was added in small portions to a stirring solution
of 30 (6 mmol) in 6 mL EtOH, 12 mL 1,2-dichloroethane, and
2.0 mL (40 mmol) hydrazine hydrate, at 30°C. After addition
of Ra-Ni was completed, the mixture was heated in a water
bath (50°C, 60 min) and filtered through celite. The filtrate
was evaporated in vacuo, and the product was used without
further purification for the synthesis of compound 12.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

ˇ
ˇ
´
512 Soskic et al.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 502−512
[14] M. M. Simpson, J. A. Ballesteros, V. Chiappa, J. Chen,
M. Suehiro, D. S. Hartman, T. Godel, L. A. Snyder, T. P.
Sakmar, J. A. Javitch, Mol. Pharmacol. 1999, 56,
1116Ϫ1126.
References
[1] M. M. Teeter, C. J. DuRand, Drug Design Discovery
1996, 13, 49Ϫ62.
[2] P. Sokoloff, J. C. Schwartz, Trends Pharmacol. Sci.
1995, 16, 270Ϫ275.
[15] S. Lober, H. Hübner, P. Gmeiner, Bioorg. Med. Chem.
Lett. 1999, 9, 97Ϫ102.
[3] T. Högberg, U. Norinder, in A Texbook of Drug Design
and Development (eds. P. Krogsgaaed-Larsen, H.
Bundgaard), Harwood Academic Publ., Chur, 1991; pp.
56Ϫ91.
[16] H. E. Katerinopoulos, D. I. Schuster, Drugs of the Future,
1987, 12, 233Ϫ253.
[17] R. Fausto, M. J. S. Ribeiro, J. J. Pedroso de Lima, J.
Mol. Structure 1999, 484, 181Ϫ196.
ˇ
ˇ
´
´
[4] V. Soskic, J. Joksimovic, Curr. Med. Chem. 1998, 5,
493Ϫ512.
[18] J. J. P. Stewart, J. Comput. Chem. 1989, 10, 209Ϫ220.
[19] A. D. Becke, J. Chem. Phys. 1993, 98, 5648Ϫ5652.
[20] L. Laaksonen, J. Mol. Graph. 1992, 10, 33Ϫ34.
[5] R. Wilcox, T. Tseng, M. Kim, B. Ginsburg, R. Pearlman,
M. Teeter, C. DuRand, S. Starr, K. Neve, J. Med. Chem.
1998, 41, 4385Ϫ4391.
ˇ
´
ˇ ´
ˇ
´
´
[21] S. Kostic-Rajacic, V. Soskic, J. Joksimovic, Arch. Pharm.
1998, 331, 22Ϫ26.
[6] W. B. Jennings, B. M. Farell, J. F. Malone, Acc. Chem.
Res. 2001, 34, 885Ϫ894.
[22] G. B. McGaughey, R. E. Mewshaw, Bioorg. Med. Chem.
1999, 7, 2453Ϫ2456.
[7] S. Tsuzuki, K. Honda, T. Uchimaru, M. Mikami, K. Tan-
abe, J. Am. Chem. Soc. 2002, 124, 104Ϫ112.
[23] A. Leonardi, G. Motta, C. Riva, R. Testa, J. W. Corbett,
WO 2003031436, 2003; Chem. Abstr. 2003, 138,
304309.
[8] M. M. Teeter, M. Froimowitz, J. B. Stec, C. J. DuRand,
J. Med. Chem. 1994, 37, 2874Ϫ2888.
[9] R. E. Wilcox, W. H. Huang, M. Y. Brusniak, D. M. Wilcox,
R. S. Pearlman, M. M. Teeter, C. J. DuRand, B. L.
Wiens, K. A. Neve, J. Med. Chem. 2000, 43,
3005Ϫ3019.
[24] R. E. Mewshaw, M. Husbands, E. S. Gildersleeve, M. B.
Webb, X. Shi, H. Mazandarani, M. I. Cockett, R. Ochal-
ski, J. A. Brennan, M. Abou-Gharbia, K. Marquis, G. B.
McGaughey, J. Coupet, T. H. Andree, Bioorg. Med.
Chem. Lett. 1998, 8, 295Ϫ300.
[10] K. A. Neve, M. G. Cumbay, K. R. Thompson, R. Yang,
D. C. Buck, V. J. Watts, C. J. DuRand, M. M. Teeter, Mol.
Pharmacol. 2001, 60, 373Ϫ381.
[25] Y. Ohtake, K. Kawano, A. Naito, K. Naito, WO 9965874,
1999; Chem. Abstr. 1999, 132, 49970.
[11] J. A. Ballesteros, L. Shi, J. A. Javitch, Mol. Pharmacol.
2001, 60, 1Ϫ19.
[26] S. S. Greenberg, W. C. Lumma, Jr., K. Nickisch, R. A.
Wohl, Eur. Pat. Appl. 1989, 39, EP 332570, 1989; Chem.
Abstr. 1989, 112, 138763.
[12] S. Löber, H. Hübner, W. Utz, P. Gmeiner, J. Med. Chem.
2001, 44, 2691Ϫ2694.
ˇ
ˇ
´
´
´
´
[27] V. Soskic, A. Maelicke, G. Petrovic, B. Ristic, J. Petrovic,
J. Pharm. Pharmacol. 1990, 43, 27Ϫ31.
[13] M. Y. Cha, B. C. Choi, K. H. Kang, A. N. Pae, K. I. Choi,
Y. S. Cho, H. Y. Koh, H.-Y. Lee, D. Jung, J. Y. Kong,
Bioorg. Med. Chem. Lett. 2002, 12, 1327Ϫ1330.
[28] P. J. Munson, D. Rodbard, Anal. Biochem. 1980, 107,
220Ϫ239.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim