Identification of the binding sites and selectivity of sarpogrelate, a novel 5HT 2 antagonist, to...
Transcript of Identification of the binding sites and selectivity of sarpogrelate, a novel 5HT 2 antagonist, to...
www.elsevier.com/locate/lifescie
Life Sciences 73 (2003) 193–207
Identification of the binding sites and selectivity of sarpogrelate,
a novel 5-HT2 antagonist, to human 5-HT2A, 5-HT2B and
5-HT receptor subtypes by molecular modeling
2CMamunur Rashida, Philippe Manivetb,c, Hiroaki Nishiod, Jaturong Pratuangdejkulb,c,Mazen Rajabb,c, Masaji Ishiguroe, Jean-Marie Launayb,c, Takafumi Nagatomoa,*
aDepartment of Pharmacology, Niigata University of Pharmacy and Applied Life Sciences, 5-13-2 Kamishinei-cho,
Niigata 950-2081, JapanbCR C. Bernard ‘‘Pathologie Experimentale et Communications Cellulaires’’, IFR Jules Marrey,
Service de Biochimie et de Biologie Moleculaire, Hopital Lariboisiere, 75475 Paris Cedex 10, FrancecE.A. ‘‘Biologie des maladies a prions et regulations cellulaires’’, Laboratoire de Biologie Cellulaire,
UFR des Sciences Pharmaceutiques et Biologiques, 4 Avenue de l’Observatoire, 75006 Paris, FrancedDepartment of Pharmacology, Faculty of Pharmacy and Pharmaceutical Sciences, Fukuyama University,
Hiroshima 729-0292, JapaneSuntory Institute for Bioorganic Research, 1-1-1 Wakayamadai, Shimahon-cho, Mishima-gun, Osaka 618-8503, Japan
Received 14 November 2002; accepted 17 December 2002
Abstract
The aim of the present study was to investigate the binding sites interactions and the selectivity of sarpogrelate
to human 5-HT2 receptor family (5-HT2A, 5-HT2B and 5-HT2C receptor subtypes) using molecular modeling.
Rhodopsin (RH) crystal structures were used as template to build structural models of the human serotonin-2A and
-2C receptors (5-HT2AR, 5-HT2CR), whereas for 5-HT2BR, we used our previously published three-dimensional
(3D) models based on bacteriorhodopsin (BR). Sarpogrelate, a novel 5-HT2R antagonist, was docked to the
receptors. Molecular dynamics (MD) simulations produced the strongest interaction for 5-HT2AR/sarpogrelate
complex. Upon binding, sarpogrelate constraints aromatic residues network (Trp3.28, Phe5.47, Trp6.48, Phe6.51,
Phe6.52 in 5-HT2AR; Phe3.35, Phe6.51, Trp7.40 in 5-HT2BR; Trp
3.28, Phe3.35, Phe5.47, Trp6.48, Phe6.51, Phe6.52 in 5-
HT2CR) in a stacked configuration, preventing activation of the receptor. The models suggest that the structural
origin of the selectivity of sarpogrelate to 5-HT2AR vs both 5-HT2BR and 5-HT2CR comes from the following
results: (1) The tight interaction between the antagonist and the transmembrane domain (TMD) 3. Asp3.32
0024-3205/03/$ - see front matter D 2003 Elsevier Science Inc. All rights reserved.
doi:10.1016/S0024-3205(03)00227-3
* Corresponding author. Tel.: +81-25-268-3170; fax: +81-25-268-1280.
E-mail address: [email protected] (T. Nagatomo).
M. Rashid et al. / Life Sciences 73 (2003) 193–207194
neutralizes the cationic head and interacts simultaneously with carboxylic group hydrogen of the antagonist
molecule. (2) Due to steric hindrance, Ser5.46 (vs Ala5.46 in 5HT2B and 5HT2C) prevents sarpogrelate to enter
deeply inside the hydrophobic core of the helix bundle and to interact with Pro5.50. (3) The side chain of Ile4.56 (vs
Ile4.56 in 5HT2BR and Val4.56 in 5HT2CR) constraints sarpogrelate to adjust its position by translating toward the
strongly attractive Asp3.32. These results are in good agreement with binding affinities (pKi) of sarpogrelate for
5-HT2 receptor family expressed in transfected cell.
D 2003 Elsevier Science Inc. All rights reserved.
Keywords: Sarpogrelate; Molecular modeling; 5-HT2 receptor family; Antagonist; Selectivity
Introduction
There are many physiological, behavioral and cognitive functions involved through the interaction
between the biogenic amine serotonin (5-Hydroxytryptamine, 5-HT) and a large family of serotonin
receptor on the membrane of both neurones and non-neuronal cells that belong to the super-family of
G-protein coupled receptor [16]. These receptors are divided in seven classes (5-HT1 to 5-HT7, except
5-HT3 that is a ionic channel) and more than 14 molecularly identified 5-HT receptor subtypes, some
with splice variants and others with isoforms created by mRNA editing [8].
The 5-HT2 serotonin receptors (5-HT2R) are of significant clinical interest because of their potential
involvement in cardiovascular function and in certain mental disorders. 5-HT2R family contains 5-HT2A,
5-HT2B and 5-HT2C subtypes sharing similar intron-exon distribution [3], high primary sequence
homology (68–79% in the transmembrane segments), pharmacology, and all stimulate phospholipase C
activity, leading to increased phosphoinositide hydrolysis in the cell. It is of now-a-days interest to give
more emphasis for identification of more selective ligands to fully elucidate the functional role of
different subtypes. Molecular modeling studies are useful in as much as they may allow us to understand
the activity and selectivity of currently existing agents, and furthermore, may aid in the design of
completely novel therapeutic agents [37].
Sarpogrelate, a 5-HT2 antagonist, was introduced as a therapeutic agent for the treatment of ischemic
diseases associated with thrombosis [17]. Sarpogrelate inhibits thrombus formation [13,39], lowers
platelet aggregation, [14,23], inhibits both serotonin-induced coronary artery spasm [21] and contraction
[12] in porcine model as well as vascular smooth muscle cell proliferation [30]. Presently these
pathophysiological effects are claimed to be mediated by 5-HT2A receptors. Based on radio-ligand
binding and functional studies, it was shown that sarpogrelate exibit specificity toward 5-HT2 receptors,
since it has lack of significant 5-HT1, 5-HT3, 5-HT4, a1-, a2- and h-adrenoreceptor, histamine H1, H2
and muscarinic M3 antagonistic activity [20,24,26]. However, the structural origin of the selectivity of
sarpogrelate to 5-HT2AR vs 5-HT2B and 5-HT2C receptor subtypes has not yet been reported. In the
present study, we thus intend to precise the selectivity and binding sites of sarpogrelate to 5-HT2A,
5-HT2B and 5-HT2C receptor subtypes by molecular modeling. Using bovine rhodopsin X-ray structure
(1HZX) [31] as template, 3D structural models of the human serotonin 5-HT2AR and 5-HT2CR were
constructed. For 5-HT2BR we used our previously published 3D model [19]. The receptors were then
complexed with sarpogrelate by molecular modeling techniques and submitted to MD simulations. The
antagonist-receptor complexes were then used to examine sarpogrelate binding sites and to determine key
residues that may be important for specifying sarpogrelate 5-HT2AR selectivity.
M. Rashid et al. / Life Sciences 73 (2003) 193–207 195
Materials and methods
Radioligand binding experiments
5-HT2A receptor binding assay in CHO-K1 cells expressing a cloned human 5-HT2A receptor was
performed using [3H]ketanserin as a radioligand. CHO-K1 cells stably transfected with a plasmid
encoding the human serotonin 5-HT2A receptor are used to prepare membranes in modified Tris–HCl
buffer (50 mM Tris–HCl, pH 7.7) using standard techniques [4]. In competition binding experiments, a
30 mg aliquot of membrane protein is incubated with 0.5 nM [3H]ketanserin for 60 minutes at 25 jC.For saturation binding experiments, concentrations of radioligand ranged from 0.01 nM to 2.0 nM. Non-
specific binding is estimated in the presence of 1.0 AM mianserin. Membranes are filtered and washed 3
times and the filters are counted to determine [3H]ketanserin specifically bound. Ki was calculated using
the Cheng–Prusoff equation [9]. Ki values are expressed as pKi (� log Ki).
5-HT2B receptor binding assay in CHO-K1 cells expressing a cloned human 5-HT2B receptor was
performed using [3H]LSD as a radioligand. CHO-K1 cells stably transfected with a plasmid encoding
the human serotonin 5-HT2B receptor are used to prepare membranes in modified Tris–HCl buffer (50
mM Tris–HCl, pH 7.4, 4 mM CaCl2, 0.1% ascorbic acid) using standard techniques [4]. In com-
petition binding experiments, a 30 mg aliquot of membrane protein is incubated with 1.2 nM [3H]LSD
for 60 minutes at 37 jC. For saturation binding experiments, concentrations of radioligand ranged
from 0.05 to 8.0 nM. Non-specific binding is estimated in the presence of 10 mM serotonin. Mem-
branes are filtered and washed 3 times and the filters are counted to determine [3H]LSD specifically
bound. Ki was calculated using the Cheng–Prusoff equation [9]. Ki values are expressed as pKi
(� log Ki).
5-HT2C receptor binding assay in CHO-K1 cells expressing a cloned human 5-HT2C receptor was
performed using [3H]mesulergine as a radioligand. CHO-K1 cells stably transfected with a plasmid
encoding the human serotonin 5-HT2C receptor are used to prepare membranes in modified Tris–HCl
buffer (50 mM Tris–HCl, pH 7.7, 0.1% ascorbic acid, 10 AM pargyline) using standard techniques [4].
In competition binding experiments, a 3.2 mg aliquot of membrane protein is incubated with 1.0 nM
[3H]mesulergine for 60 minutes at 25 jC. For saturation binding experiments concentrations of
radioligand ranged from 0.05 to 8.0 nM. Non-specific binding is estimated in the presence of 1 AMmianserin. Membranes are filtered and washed 3 times and the filters are counted to determine
[3H]mesulergine specifically bound. Ki was calculated using the Cheng–Prusoff equation [9]. Ki values
are expressed as pKi (� log Ki).
Protein assay
Protein contents of the membrane concentration were measured by the method of Bradford [5] using
bovine serum albumin as the standard.
Data analysis
Kd (dissociation constants) and Bmax (the density of 5-HT2 receptors binding sites) were determined
using nonlinear regression analysis. Nonlinear regression analysis of saturation and competition binding
assay were performed using Prism (GraphPad Software, San Diego, CA, USA). Data were presented as
M. Rashid et al. / Life Sciences 73 (2003) 193–207196
the mean F standard error of mean (SEM). Differences between groups were analyzed using a one-way
ANOVA, to obtain a p value.
Drugs
The drugs used and their sources were as follows: sarpogrelate from Mitsubishi Chemical Corporation
(Tokyo, Japan); ketanserin, miancerin and serotonin from Research Biochemicals International (Mary-
land, USA); [3H]ketanserin (88 Ci mmol� 1) and [3H]LSD (76.2 Ci mmol� 1) from NEN Life Science
Products (USA); [3H]mesulergine (74 Ci mmol� 1) from Amersham Pharmacia Biotech (England).
Molecular modeling
3D models of the TMD bundle of the three 5-HT2R family member were carried out by using
homology modeling technique using criteria and procedures which took into account sequence
conservation [1,25], degree of residue polarity [40], and incorporated a large number of experimental
constraints as reviewed elsewhere [2,34]. All bioamine-GPCR sequences presently available were
Fig. 1. Transmembrane domains multiple sequence alignment of human 5-HT2AR, 5-HT2BR, 5-HT2CR and bovine rhodopsin.
Most conserved residue of each TMD among GPCR super-family is shaded in gray. Residues are numbered according to a
consensus numbering scheme [2], and each helix N-terminal starting residue is numbered according to SWISSPROT genome
databank numbering scheme (5-HT2AR P28223, 5-HT2BR P41595, 5-HT2CR P28335, bovine rhodopsin P02269). Residues
involved in sarpogrelate binding are depicted in bold. Residues that are making electrostatic interactions with sarpogrelate are
indicated by circle. For clarity, only the TMD sequences of the template and of the three studied 5-HT2Rs are presented in
Fig. 1.
M. Rashid et al. / Life Sciences 73 (2003) 193–207 197
aligned with RH and BR using the CLUSTALX software [32]. For clarity, only the TMD sequences of
the two templates, BR and RH, and the three studied 5-HT2R family are represented in Fig. 1. The
residues are numbered according to both the SwissProt database and to Ballesteros and Weinstein [2], in
order to facilitate comparison between the three subtypes of 5-HT2R.
Model refinements, energy minimization, and molecular dynamics (MD) simulations were performed
with the CHARMM program [7]. A 20-A cut-off was applied to non-bond interactions. All graphical
manipulations were done with the Insight II modeling package (Accelrys Inc, San Diego, CA). Both
graphical interface and calculations were performed on Silicon Graphics O2 stations.
To minimize errors inherent to the generation of receptor 3D-models, for each receptor, five different
protocols of simulation were performed by assigning different initial velocities, leading in fine to fifteen
receptor models. For 5-HT2BR, we used our previously published 3D model [19]. The 5-HT2AR and
5-HT2CR models were built on the basis of the backbone coordinates of 1HZX [31] file from the
Protein Data Bank. Side-chains of the receptors were positioned using the Biopolymer module of Insight II
software. Interhelix contacts were adjusted by taking into account data from site-directed mutagenesis
experiments. After 700 ps of constrained MD simulations, average structures resulting from 200 ps
of free MD simulations were energy-minimized. Conformational stability of the generated 5-HT2AR,
5-HT2BR and 5-HT2CR models were systematically assessed by following the root mean square
deviations (RMSDs) of the cartesian coordinates of the backbone atoms at all steps of the simulations.
The obtained relaxed and equilibrated receptors were further used for the docking of sarpogrelate.
Sarpogrelate parameters for CHARMM force field were optimized by performing quantum chemistry
conformational analysis of sarpogrelate. Calculations were done at the DFT/6-31 + G** level. The 3D
models of 5-HT2 receptors were finally complexed to sarpogrelate by using the AutoDock 3.0 program
[22]. The most stable receptor/antagonist drug complexes were then submitted to 300 ps MD at 300 K
and finally analyzed.
Results
Radioligand binding experiments for 5-HT2 receptor family
In saturation binding experiments, specific binding of [3H]ketanserin, [3H]LSD and [3H]mesulergine
were determined using the cloned membranes of the 5-HT2A, 5-HT2B and 5-HT2C receptor subtypes,
respectively. The affinity constant (Kd) and the receptor density (Bmax) for all the membranes are
shown in the Table 1. Both sarpogrelate and ketanserin monophasically inhibited this bindings. Table 2
shows the binding affinity (pKi) for these two 5-HT2 antagonists at the three subtypes. Both the
Table 1
Saturation binding parameters for human 5-HT2 receptor family
Receptors Radioligand Kd (nM) Bmax (fmol mg� 1 protein� 1)
5-HT2A (3) [3H]ketanserin 0.2 F 0.005 510.0 F 62.0
5-HT2B (3) [3H]LSD 2.1 F 0.2 1100.0 F 62.0
5-HT2C (3) [3H]mesulergine 1.1 F 0.1 4900.0 F 843.0
The data shown indicate mean F SEM and numbers in parentheses indicate number of experiments.
Table 2
Binding affinities (pKi) of sarpogrelate and ketanserin for human 5-HT2 receptor family in CHO-K1 cells
Receptors Sarpogrelate (pKi) Ketanserin (pKi)
5-HT2A 8.52 F 0.12 (3) 9.67 F 0.12 (4)
5-HT2B 6.57 F 0.12 (3) 6.55 F 0.09 (4)
5-HT2C 7.43 F 0.03 (3) 7.39 F 0.11 (4)
The data represented the mean F SEM. Number in the parentheses indicated the number of experiments. pKi = � log (Ki),
where Ki is in nM.
M. Rashid et al. / Life Sciences 73 (2003) 193–207198
antagonists showed high binding affinity to 5-HT2A subtype vs 5-HT2B and 5-HT2C subtypes. However,
ketanserin showed 10 fold higher binding affinity to 5-HT2A subtype in comparison with sarpogrelare.
Sarpogrelate displayed higher binding affinity to 5-HT2A subtype than 5-HT2B and 5-HT2C subtypes.
The affinities of sarpogrelate and ketanserin to both 5-HT2B and 5-HT2C receptor subtypes were almost
similar.
Molecular modeling
The observation of the modeled binding sites of sarpogrelate in the three receptors subtypes shows a
strong interaction between antagonist and receptors (Figs. 3 and 4). Those interactions involve a large
number of residues (listed in Table 3) belonging to the TMDs 3, 4, 5, 6, and 7. In the three complexes
Table 3
List of the interacting residues with 5-HT2AR, 5-HT2BR and 5-HT2CR
TMDs 5HT2AR 5HT2BR 5HT2CR
VdW Elc VdW Elc VdW Elc
3 Trp3.28 Trp3.28 Val3.33 Asp3.32 Trp3.28 Asp3.32
Ile3.29 Asp3.32 Phe3.35 Leu3.31
Asp3.32 Asp3.32
Val3.33 Phe3.35
Ser3.36
Ser3.39
Met3.41
4 Ile4.56 Leu4.51 Ile4.52
Ser4.53
Val4.56
5 Ser5.46 Gly5.42 Ala5.46
Phe5.47 Ala5.46 Phe5.47
Thr5.49 Pro5.50
6 Trp6.48 Trp6.48 Phe6.51 Trp6.48
Phe6.51 Thr6.54 Phe6.51
Phe6.52 Leu6.58 Phe6.52
7 Val7.39 Glu7.36 Gly7.42 Ser7.46
Tyr7.43 Ile7.37
Trp7.40
Residues are numbered according to a consensus numbering scheme [2]. Receptor transmembrane domains (TMDs). Van der
Waals contacts (VdW). Electrostatic interactions (Elc).
M. Rashid et al. / Life Sciences 73 (2003) 193–207 199
sarpogrelate takes a similar orientation inside the helix bundle, parallel to the membrane plane. The
hydrophobic part of the antagonist molecule that is defined by two 6-membered rings (rings A and B,
Fig. 2), is tightly bound to residue side chains of the hydrophobic pocket delimited by TMDs 4, 5 and 6
through a large number of Van der Waals interactions. The charged part of sarpogrelate interacts mostly
with polar and charged residues belonging to TMD 3 and 6 for 5-HT2AR, TMD 3 for 5-HT2BR and TMD
3 and 7 for 5-HT2CR, through electrostatic interactions.
Electrostatic interactions
In the three receptor subtypes/antagonist complexes, a strong electrostatic interaction is formed in the
course of all simulations between the cationic head of sarpogrelate and the carboxylate group of Asp3.32
(Fig. 4) that is the major counter-ion for 5-HT binding to 5-HT2Rs [11,18,29,33,38].
In the 5-HT2AR, the number of electrostatic interactions established with sarpogrelate is higher than in
the two other subtypes. Indeed, in addition to the interaction with the cationic head (hydrogen atom of
the trimethyl ammonium) of sarpogrelate, Asp3.32 (acidic function oxygen) interacts with hydrogen H1
atom of carboxylate group of the antagonist. The peptide bond oxygen of Trp3.28 interacts with the
hydrogen atom of the trimethyl ammonium group of sarpogrelate and finally, indole NH group hydrogen
of Trp6.48 interacts with the carboxylate group oxygen O2 of sarpogrelate (Figs. 3a and 4a).
In the 5-HT2BR and the 5-HT2CR, sarpogrelate interacts with acidic function of Asp3.32 (Fig. 4b and
4c). Moreover, in the 5-HT2CR, carboxylate group oxygen O1 of sarpogrelate is H-bonded by Ser7.46 OH
group. In 5-HT2AR and 5-HT2BR, interactions with Ser7.45 and Ser7.46 are not possible due to the axial
position of sarpogrelate.
Fig. 2. Sarpogrelate chemical structure. Ox (x = 1–6), oxygen atom numbering scheme.
Fig. 3. Stereo view of the binding site of sarpogrelate in 5-HT2AR (a), 5-HT2BR (b), 5-HT2CR (c) receptor, view orthogonal to
the membrane plane from the extracellular side of the cell.
M. Rashid et al. / Life Sciences 73 (2003) 193–207200
Hydrophobic interactions
Whatever the receptor subtype considered, most of the Van der Waals contacts with hydrophobic
residues are done by atoms of the two 6-membered rings A and B of sarpogrelate. A major difference of
Fig. 4. Two dimensional plots of sarpogrelate/5-HT2AR (a), -/5-HT2BR (b), -/5-HT2CR (c) complexes. Hydrogen atoms were not
represented for clarity of the picture. Spiky semicircle around interacting atoms represent Van der Waals interactions. Small
lines between interacting heavy atoms symbolize hydrogen bonds. Residues are numbered according to a consensus numbering
scheme [2]. Schematic diagrams were obtained with the help of LIGPLOT software [35].
M. Rashid et al. / Life Sciences 73 (2003) 193–207 201
sarpogrelate binding among the three receptor subtypes appears in the ligand bundle axial position
(orthogonal to the membrane plane). In the 5-HT2AR, sarpogrelate is localized near the extracellular
edge of the receptor, whereas it is deeply embedded inside the hydrophobic binding pocket delimited by
TMD 4, 5 and 6, in the 5HT2CR and in a lesser extend in the 5HT2BR. This difference seems to be due to
mostly for the nature of key residues present in TMD 4 and 5. In 5HT2AR, the side chain of Ser5.46 (vs
Ala5.46 in 5HT2BR and 5HT2CR,) prevents the methoxybenzyl rings (ring B, Fig. 2) of sarpogrelate to
enter deeply inside the hydrophobic binding pocket. Due to steric hindrance, the methyl position of the
methoxy group of ring B prevents the formation of a hydrogen bond between OH group of Ser5.46 and
the methoxy oxygen O6 of sarpogrelate. During the course of molecular dynamics trajectory, driving
forces guided the ligand toward TMD 4. In the 5-HT2AR, to avoid steric clash and in order to adjust Van
der Waals contacts with Ile4.56 side chain (vs Ile4.56 in 5HT2BR and Val4.56 in 5HT2CR), sarpogrelate is
constrained to move away from Ile4.56 and Ser5.46 toward the extra-cellular side of the receptor and to
translate toward the strong attractive and negatively charged Asp3.32. The presence of a shorter length
side chain Val residue at locus 4.56 in the 5-HT2CR, makes possible the interaction of sarpogrelate with
Ile4.52 and Ser4.53.
In 5HT2CR, sarpogrelate is deeply embedded in the helix bundle and the methoxy group of aromatic
ring B can interact with Pro5.50. Additional contacts are done with Ala5.46 and Phe5.47. In 5HT2BR,
M. Rashid et al. / Life Sciences 73 (2003) 193–207202
sarpogrelate is less embedded than in 5-HT2CR. The interaction with Pro5.50 is not possible since the
translation of sarpogrelate is blocked by the presence of the Thr5.49 side-chain (vs Ile5.49 in 5HT2AR and
5HT2CR).
Interactions with aromatic residues
Sarpogrelate promotes the rearrangement by direct contacts of residues belonging to the aromatic
network upon binding, including in 5-HT2A/2CR, Phe5.47, Trp6.48, Phe6.51 and Phe6.52 and in 5-HT2BR,
only Phe6.51. The conformational changes undergone by those residue side-chains are not compatible
with activation of the receptor. Indeed, we previously described the structural modification promoted by
5-HT docking in 5-HT2BR [19]. Upon 5-HT binding, rotation of the Phe6.52 aromatic ring affected the
aromatic network within the hydrophobic pocket delimited by TMDs 4, 5, 6, and 7. The two contacted
aromatic residues, Phe6.52 and Trp6.48, adopted a ‘‘T-shape’’ oriented structure. In the case of
sarpogrelate, aromatic residues are constrained in a stacked conformation. The two biphenyl rings A
and B of sarpogrelate constrain the side-chains of those amino acids by hydrophobic interactions and
prevent the propagation of sequential side-chain rotation by maintaining stacked the ring plane of
aromatic residues.
We note also in the three complexes that some aromatic residues (Trp3.28, Trp6.48, Phe6.51 in 5-HT2AR;
Phe3.35, Phe6.51, Trp7.40 in 5-HT2BR; Phe3.35, Trp3.28, Trp6.48, Phe6.51 in 5-HT2CR) define an ‘‘ aromatic
box ’’ that surrounds and stabilizes the cationic head of sarpogrelate and maintains the antagonist
molecule in a favorable orientation to interact with Asp3.32.
Discussion
In this report, computer modeling of the 5-HT2R family allowed us to define putative binding sites of
sarpogrelate in the three subtypes and to speculate on the possible 5-HT2AR binding selectivity. The
relevance of this theoretical approach was estimated by comparing the structural predictions made on the
basis of the sole computations models to radioligand binding assays data measuring affinity of
compounds with pKi.
Sarpogrelate affinity for the three subtypes
The binding of sarpogrelate in our three 5-HT2R subtypes 3D models shows the participation
of a higher number of interacting residues, mainly through hydrophobic interactions, in 5-HT2A/
2CR than in 5-HT2BR (Table 3). In 5-HT2AR, sarpogrelate makes electrostatic interactions with
three residues (Asp3.32, Trp3.28 and Trp6.48). At first approximation, this suggests a greater affinity
of sarpogrelate for 5-HT2AR followed by 5-HT2CR and finally 5-HT2BR. Those data are in good
agreement with experimentally determined binding affinities of sarpogrelate for 5-HT2R family
expressed in transfected cell. pKi values of sarpogrelate for 5-HT2AR, 5-HT2BR and 5-HT2CR,
are respectively 8.52, 6.57, 7.43 (Table 2). Sarpogrelate is a highly active on 5-HT2AR (pKi =
8.52).
In the three receptor models, Asp3.32 is involved in the binding site of sarpogrelate. Previous
published molecular modeling studies of 5-HT2R family member/antagonist complexes have shown
M. Rashid et al. / Life Sciences 73 (2003) 193–207 203
the existence of such a pattern of interactions between TMD 3, through highly conserved Asp3.32
and ligand molecule [11,18,29,33,38]. Asp3.32 neutralizes the cationic head of sarpogrelate and
allows the antagonist molecule to prevent the fixation and the action of agonist molecules.
Sarpogrelate and aromatic residues network
Hibert et al. [15] reported that the two highly conserved aromatic residues Trp6.48 and Phe6.52
are directly involved in the binding of several neurotransmitters (5-HT, dopamine and adrenaline)
to their corresponding receptors. They also speculated that the side chain conformations of these
hydrophobic residues probably changed during the binding process. Such changes could directly
affect the conformation of the adjacent helices, in particular in the vicinity of proline residues, and
of other helices by propagation along the backbone of interacting conserved aromatic residues.
Site-directed mutagenesis experiments also supported the crucial role of the two above aromatic
residues [10,28,36]. Six conserved aromatic residues (Trp3.28, Phe3.35, Trp4.50, Phe6.44, Trp6.48, and
Trp7.40) are in hydrophobic interaction and belong to this aromatic network. We have previously
demonstrated that in 5-HT2BR/5-HT complex, upon 5-HT binding, a shift of Phe6.52 induces a
displacement of the Trp6.48 side chain. Consequently, the other aromatic amino-acids in the
hydrophobic core of 5-HT2BRs are submitted to conformational changes. In our present 3D
models, whatever the subtype considered, the two phenyl rings of sarpogrelate are embedded in the
hydrophobic core of the receptors delimited by TMD 4, 5, 6 and interact tightly with aromatic
residues. Phe6.52 and Trp6.48 side chains are constraints and can undergo a conformational change
upon sarpogrelate binding. This structural configuration explains the low RMSD of the backbone
atoms found between the starting and final MD conformations of 5-HT2AR and in a lesser extent
for 5-HT2CR.
5-HT2AR selectivity
Radioligand binding assay showed that sarpogrelate had higher binding affinity to human 5-
HT2AR than 5-HT2BR and 5-HT2CR. The sarpogrelate pKi values at 5-HT2AR, 5-HT2BR and 5-
HT2CR show a moderate selectivity for 5-HT2AR vs 5-HT2CR (10 fold) and a high selectivity for
5-HT2AR vs 5-HT2BR (100 fold) (Table 2). Our modeling results are in good agreement with
experimental data. Indeed, bearing in mind that standard MD simulation algorithms do not allow
simulations of the protein folding processes and calculations of theoretical pKi, it can be used to
appreciate the role of key residues in the compound selectivity and interpret experimental data. At
first approximation, receptor affinity can be appreciated by the number of interacting residues and
the strength of the interactions. The number and strength of interactions are easily given by listing
effective Van der Waals and electrostatic interaction between the ligand and the receptor. The
selectivity can be evaluated by comparing and analyzing the binding modes of the different
complexes.
We have shown above that axial position of sarpogrelate in the helix bundle is different for the
three 5-HT2R subtypes. This orientation could explain the difference of affinity between the three
receptors subtypes and also the selectivity of sarpogrelate for 5-HT2AR. The position of sarpogrelate
in the 5HT2A/2CR is mainly influenced by residues at the loci 4.56 and 5.46. In 5-HT2AR, the side
chain of Ser5.46 prevents sarpogrelate to deeply enter inside the helix bundle and to make a contact
M. Rashid et al. / Life Sciences 73 (2003) 193–207204
with Pro5.50 due to steric hindrance between OH group of Ser5.46 and the methoxy oxygen O6 of
sarpogrelate. The side chain of Ile4.56 constraints sarpogrelate to adjust its position by translating
toward the strongly attractive and negatively charged Asp3.32. In 5-HT2CR, Ala5.46 and Val4.56, two
shorter side-chain residues allow 5-HT2CR to interact down the helix bundle axis to Pro5.50. Since
5-HT2BR 3D model was built from a different 3D crystallized template, a small but significant
difference appears in TMD 5. The methyl group of Thr5.49 side-chain interacts with phenyl ring B
of sarpogrelate and not with Pro5.50.
Ser5.46 is not conserved among all the serotonin receptor families. Indeed, among the 90 aligned
receptors, a Ser residue is present at this locus only in five different serotonin receptors (human, pig,
Rhesus macaque 5-HT2AR ; drosophila 5-HT7R and silk moth 5-HTR) and a Thr residue carrying a
hydroxyl group is present in seven different serotonin receptors (drosophila 5-HT1AR and 5-HT1BR ;
noctuid moth 5-HTR ; human, rat and mouse 5-HT6R ; coenorhabditis 5-HTCE2LR). In the other receptor
sequence, an Ala residue is present. A long side chain hydrophobic residue may change the binding
profile of sarpogrelate. Those observations are reinforcing the hypothesis that sarpogrelate 5-HT2AR
selectivity may depend mostly on the presence of a polar residue or a long hydrophobic side-chain
residue at locus 5.46. It is interesting to note that the OH group hydrogen of Ser5.46 can not make a
hydrogen bond with the methoxy oxygen of sarpogrelate as we could expect. Indeed, the long side chain
of Ile4.56 prevents the methoxy group of sarpogrelate and the OH group of Ser5.46 to adjust their position
to optimally interact.
Comparison of sarpogrelate with other antagonist/5-HT2R family complexes
Locus 7.45 has been previously described as determinant for selective binding of ketanserin to the
5HT2AR vs 5-HT2CR [18]. The presence of a polar residue at locus 7.46 seems also to influence
interaction with locus 7.45. In our 3D models, locus 7.46 participates to antagonist binding and increases
5-HT2CR selectivity vs 5-HT2AR. Neither locus 7.45 nor locus 7.46 is participating to sarpogrelate
binding in the 5-HT2A/2BR 3D model due to the receptor axial position of the molecule due to their
positions in the helix bundle axis.
Comparison of experimental binding affinities of sarpogrelate and ketanserin shows that both ligand
are strongly active on 5-HT2AR. pKi values for sarpogrelate and ketanserin were 8.52 and 9.67 for
5-HT2AR, 6.57 and 6.55 for 5-HT2BR, 7.43 and 7.39 for 5-HT2CR, respectively (Table 2). We can
note more pronounced 5-HT2AR affinity (10 fold) and selectivity for ketanserin.
Sarpogrelate binding mode in our 5-HT2R family 3D models are different from recently published
ketanserin related compound like aminoalkylbenzo and heterocycloal [6]. In this study, antagonist
compounds were predicted to bind with TMD 2, 3 and 7 and have no contact with TMD 4, 5 and 6.
Affinity of ketanserin for Ser5.46 to Ala 5-HT2A mutant receptor has been also tested in order to
control that helix 5 is not participating to ketanserin binding [27]. However, only the lowest energy
conformation of the complexes was retained for study, selected from ligand docking software protocol.
No MD simulation was applied to those ligand/receptor configurations to test the availability of the
antagonist binding mode. Sarpogrelate is not related to the same family of compounds than ketanserin
and present a particular chemical structure, i.e., a very flexible aromatic part (ring A and B) that
makes feasible the docking of the molecule perpendicular to the bundle axis. This can explain the
difference obtained with ketanserin binding, parallel to the bundle axis since the latter molecule is
somewhat rigid.
M. Rashid et al. / Life Sciences 73 (2003) 193–207 205
Conclusion
Our molecular modeling investigations are useful to predict that the sarpogrelate binding site of 5-
HT2R family involves residues belonging to TMDs 3, 4, 5, 6 and 7. We also provide first structural
hypothesis on the 5-HT2AR selectivity of sarpogrelate that may be further examined by site-directed
mutagenesis experiments and radioligand binding assays. Important determinants for selective binding
appear to be assumed mostly by Trp3.28, Asp3.32, Ile4.56, Ser5.46. Those results are useful to axe research
in the way to improve 5-HT2AR vs 5-HT2BR/5-HT2CR selectivity by modifying the chemical structure of
sarpogrelate, and the present findings of the selectivity of sarpogrelate towards 5-HT2AR vs 5-HT2BR
and 5-HT2CR emphasises its clinical implications.
Acknowledgements
This research was in part supported by a grant from the Promotion and Mutual Aid Corporation for
Private Schools of Japan.
References
[1] Baldwin JM. Structure and function of receptors coupled to G proteins. Current Opinion in Cell Biology 1994;6:180–90.
[2] Ballesteros JA, Weinstein H. Integrated methods for the construction of three dimensional models and computational
probing of structure–function relations in G-protein coupled receptors. Methods in Neurosciences 1995;25:366–428.
[3] Baxter G, Kennett GA, Blaney F, Blackburn T. 5-HT2 receptor subtypes: a family re-united? Trends in PharmacologicalSciences 1995;16:105–10.
[4] Bonhaus DW, Bach C, DeSouza A, Salazar FHR, Matsuoka BD, Zuppan P, Chan HW, Eglen RM. The pharmacology and
distribution of human 5-hydroxytryptamine2B (5-HT2B) receptor gene products: comparison with 5-HT2A and 5-HT2C
receptors. British Journal of Pharmacology 1995;115:622–8.
[5] Bradford M. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle
of protein-dye binding. Analytical Biochemistry 1976;72:23422–6.
[6] Brea J, Rodrigo J, Carrieri A, Sanz F, Cadavid MI, Enguix MJ, Villazon M, Mengod G, Caro Y, Masaguer CF, Ravina E,
Centeno NB, Carotti A, Loza MI. New serotonin 5-HT(2A), 5-HT(2B), and 5-HT(2C) receptor antagonists: synthesis,
pharmacology, 3D-QSAR, and molecular modeling of (aminoalkyl)benzo and heterocycloalkanones. Journal of Medicinal
Chemistry 2002;45:54–61.
[7] Brooks BR, Bruccoleri RE, Olafson BD, States DJ, Swaminathan S, Karplus M. CHARMM a program for macro-
molecular energy, minimization and dynamics calculations. Journal of Computational Chemistry 1983;4:187–212.
[8] Burns CM, Chu H, Reuter SM, Hutchinson LK, Canton H, Sanders-Bush E, Emeson RB. Regulation of serotonin-2C
receptor G-protein coupling by RNA editing. Nature 1997;387:303–8.
[9] Cheng Y-C, Prusoff WH. Relationship between the inhibition constant (Ki) and the concentration of inhibitor which
causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochemical Pharmacology 1973;22:3099–108.
[10] Choudhary MS, Sachs N, Uluer A, Glennon A, Westkaemper RB, Roth BL. Differential ergoline and ergopeptine binding
to 5-hydroxytryptamine2A receptors: ergolines require an aromatic residue at position 340 for high affinity binding.
Molecular Pharmacology 1995;47:450–7.
[11] Donnelly D, Findlay JBC, Blundell TL. The evolution and structure of aminergic G-protein-coupled receptors. Receptors
Channels 1994;2:61–8.
[12] Gong H, Nakamura T, Hattori K, Ohnuki T, Rashid M, Nakazawa M, Watanabe K, Nagatomo T. A novel 5-HT2
antagonist, sarpogrelate hydrochloride, shows inbibitory effects on both contraction and relaxation mediated by 5-HT
receptor subtypes in porcine coronary arteries. Pharmacology 2000;61:263–8.
M. Rashid et al. / Life Sciences 73 (2003) 193–207206
[13] Hara H, Kitajima A, Shimada H, Tamao Y. Antithrombic effect of MCI-9042, a new antiplatelet agent on experimental
thrombosis models. Thrombosis and Haemostasis 1991;66:484–8.
[14] Hara H, Osakabe M, Kitajima A, Tamao Y, Kikumoto R. MCI-9042, a new antiplatelet agent is a selective S2-serotonergic
receptor antagonist. Thrombosis and Haemostosis 1991;65:415–20.
[15] Hibert M, Trumpp-Kallmeyer S, Bruinvels A, Hoflack J. Three-dimensional models of neurotransmitter G-binding
protein-coupled receptors. Molecular Pharmacology 1991;40:8–15.
[16] Hoyer D, Hannon JP, Martin GR. Molecular, pharmacological and functional diversity of 5-HT receptors. Pharmacology
Biochemistry and Behavior 2000;71(4):533–54.
[17] Ito K, Notsu T. Effect of sarpogrelate hydrochloride (MCI-9042) on peripheral circulation of chronic arterial occlusive
diseases. Journal of Clinical and Therapeutical Medicine 1991;7:1243–51.
[18] Kristiansen K, Dahl SG. Molecular modeling of serotonin, ketanserin and ritanserin and their 5-HT2C receptor interac-
tions. European Journal of Pharmacology 1996;306:195–210.
[19] Manivet P, Schneider B, Smith JC, Choi DS, Maroteaux L, Kellermann O, Launay JM. The serotonin binding site of
human and murine 5-HT2B receptors: molecular modeling and site-directed mutagenesis. Journal of Biological Chemistry
2002;277(19):17170–8.
[20] Maruyama K, Kinami J, Sugita Y, Takada Y, Sugiyama E, Tsuchihashi H, Nagatomo T. MCI-9042: high affinity for
serotonergic receptors as assessed by radioligand binding assay. Journal of Pharmacobio-Dynamics 1991;14:177–81.
[21] Miyata K, Shimokawa H, Higo T, Yamawaki T, Katsumata N, Kandabashi T, Tanaka E, Takamura Y, Yogo K, Egashira K,
Takeshita A. Sarpogrelate, a selective 5-HT2A serotonergic receptor antagonist, inhibits serotonin-induced coronary artery
spasm in a porcine model. Journal of Cardiovascular Pharmacology 2000;35:294–301.
[22] Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, Olson AJ. Automated docking using a lamarck-
ian genetic algorithm and and empirical binding free energy function. Journal of Computational Chemistry 1998;19:
1639–62.
[23] Nakamura K, Kariyazono H, Moriyama Y, Toyohira H, Kubo H, Yotsumoto G, Taira A, Yamada K. Effects of sarpogrelate
hydrochloride on platelet aggregation, and its relation to the release of serotonin and P-selectin. Blood Coagulation and
Fibrinolysis 1999;10:513–9.
[24] Nishio H, Inoue A, Nakata Y. Binding affinity of sarpogrelate, a new antiplatelet agent, and its metabolite for serotonin
receptor subtypes. Archives Internationales de Pharmacodynamie et de therapie 1996;331:189–202.
[25] Oliveira L, Paiva ACM, Vriend GJ. A low resolution model for the interaction of G proteins with G protein-coupled
receptors. Protein Engineering 1999;12(12):1087–95.
[26] Pertz H, Elz S. In-vitro pharmacology of sarpogrelate and the enantiomers of its major metabolite: 5-HT2A receptor
specificity, stereoselectivity and modulation of ritanserin-induced depression of 5-HT contractions in rat tail artery. Journal
of Pharmacy and Pharmacology 1995;47:310–6.
[27] Ravina E, Negreira J, Cid J, Masaguer CF, Rosa E, Rivas ME, Fontenla JA, Loza MI, Tristan H, Cadavid MI, Sanz F,
Lozoya E, Carotti A, Carrieri A. Conformationally constrained butyrophenones with mixed dopaminergic (D(2)) and
serotoninergic (5-HT(2A), 5-HT(2C)) affinities: synthesis, pharmacology, 3D-QSAR, and molecular modeling of (amino-
alkyl)benzo- and -thienocycloalkanones as putative atypical antipsychotics. Journal of Medicinal Chemistry 1999;
42(15):2774–97.
[28] Roth BL, Shoham M, Choudhary MS, Khan N. Identification of conserved aromatic residues essential for agonist binding
and second messenger production at 5-hydroxytryptamine2A receptors. Molecular Pharmacology 1997;52:259–67.
[29] Roth BL, Willins DL, Kristiansen K, Kroeze WK. 5-Hydroxytryptamine2-family receptors (5-hydroxytryptamine2A,
5-hydroxytryptamine2B, 5-hydroxytryptamine2C): where structure meets function. Pharmacology and Therapeutics
1998;79:231–57.
[30] Sharma SK, Zahradka P, Chapman D, Kumamoto H, Takeda N, Dhalla NS. Inhibition of serotonin-induced vascular
smooth muscle cell proliferation by sarpogrelate. Journal of Pharmacology and Experimental Therapeutics 1999;290:
1475–81.
[31] Teller DC, Okada T, Behnke CA, Palczewski K, Stenkamp RE. Advances in determination of a high-resolution three-
dimensional structure of rhodopsin, a model of G-protein-coupled receptors (GPCRs). Biochemistry 2001;40(26):
7761–72.
[32] Thompson JD, Plewniak F, Poch O. A comprehensive comparison of multiple sequence alignment programs. Nucleic
Acids Research 1999;27:2682–90.
[33] Trumpp-Kallmeyer S, Hoflack J, Bruinvels A, Hibert M. Modeling of G-protein-coupled receptors: application to dop-
M. Rashid et al. / Life Sciences 73 (2003) 193–207 207
amine, adrenaline, serotonin, acetylcholine, and mammalian opsin receptors. Journal of Medicinal Chemistry 1992;
35:3448–62.
[34] Visiers I, Ballesteros JA, Weinstein H. Three-dimensional representations of G protein-coupled receptor structures and
mechanisms. Methods in Enzymology 2002;343:329–71.
[35] Wallace AC, Laskowski RA, Thornton JM. A program to generate schematic diagrams of protein-ligand interactions.
Protein Engineering 1995;8:127–34.
[36] Wang CD, Gallaher TK, Shih JC. Site-directed mutagenesis of the serotonin 5-hydroxytrypamine2 receptor: identification
of amino acids necessary for ligand binding and receptor activation. Molecular Pharmacology 1993;43:931–40.
[37] Westkaemper RB, Glennon RA. Approaches to molecular modeling studies and specific application to serotonin ligands
and receptors. Pharmacology Biochemistry and Behavior 1991;40(4):1019–31.
[38] Westkaemper RB, Glennon RA. Molecular graphics models of members of the 5-HT2 subfamily. 5-HT2A, 5-HT2B,
and 5-HT2C receptors. Medicinal Chemistry Research 1993;3:317–34.
[39] Yamashita T, Kitamori K, Hashimoto M, Watanabe S, Giddings JC, Yamamoto J. Conjunctive effects of the 5HT2 receptor
antagonist, sarpogrelate, on thrombolysis with modified tissue plasminogen activator in different laser-induced thrombosis
models. Haemostasis 2000;30:321–32.
[40] Zhang D, Weinstein H. Polarity conserved positions in transmembrane domains of G-protein coupled receptors and
bacteriorhodopsin. FEBS Letters 1994;337:207–12.