Design, synthesis, biological evaluation, and comparative docking study of 1,2,4-triazolones as CB1 receptor selective antagonists

Article abstract of DOI:10.1016/j.ejmech.2013.12.018

Cannabinoids are potentially useful for the treatment of several diseases. In the present work, we report the syntheses and biological evaluations of 1,2,4-triazolone derivatives designed using a combined approach of scaffold hopping and pharmacophore-oriented method. These compounds exhibited interesting antagonistic activity to the cannabinoid CB1 receptor. The preliminary structure-activity relationships were further discussed. In addition, docking simulations were performed on the good bioactive compound 5c and the low potent compound 5d, respectively, on the basis of homology models of the CB1 and CB2 receptors, which were constructed based on human β2-adrenoreceptor and optimized in a membrane environment by MD simulations. Calculation of the binding modes gave us insights into the structural requirements for improving the cannabinoid receptor bioactivity and selectivity.

Full text of DOI:10.1016/j.ejmech.2013.12.018

European Journal of Medicinal Chemistry 74 (2014) 73e84
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Design, synthesis, biological evaluation, and comparative docking
study of 1,2,4-triazolones as CB1 receptor selective antagonists
,
Shuang Han a, Fei-Fei Zhang b, Xin Xie b, Jian-Zhong Chen a
*
a College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, Zhejiang 310058, China
b CAS Key Laboratory of Receptor Research, National Center for Drug Screening, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 July 2013
Received in revised form
21 November 2013
Accepted 19 December 2013
Available online 31 December 2013
Cannabinoids are potentially useful for the treatment of several diseases. In the present work, we report
the syntheses and biological evaluations of 1,2,4-triazolone derivatives designed using a combined
approach of scaffold hopping and pharmacophore-oriented method. These compounds exhibited inter-
esting antagonistic activity to the cannabinoid CB1 receptor. The preliminary structureeactivity re-
lationships were further discussed. In addition, docking simulations were performed on the good
bioactive compound 5c and the low potent compound 5d, respectively, on the basis of homology models
of the CB1 and CB2 receptors, which were constructed based on human b2-adrenoreceptor and opti-
Keywords:
Cannabinoid receptors
Antagonist
1,2,4-Triazolone
Scaffold hopping
Homology model
mized in a membrane environment by MD simulations. Calculation of the binding modes gave us insights
into the structural requirements for improving the cannabinoid receptor bioactivity and selectivity.
Ó 2013 Elsevier Masson SAS. All rights reserved.
1. Introduction
trisubstituted sulfonamides and bisamide derivatives as the CB2
receptor inverse agonists showing potent inhibitory activity on
Cannabinoid (CB) receptors belong to the rhodopsin-like G
protein-coupled receptors (GPCRs) family. Up to now, it has been
identified of two subtypes of cannabinoid receptors, namely, CB1
receptor and CB2 receptor [1]. The CB1 receptor is predominantly
located within the central nervous system (CNS) and also identified
in peripheral nerve terminals and other cell types, while the CB2
receptor is expressed mainly in the peripheral immune tissues and
cells. Activation of either CB1 or CB2 receptor mediates inhibition of
adenylate cyclase and activation of mitogen-activated protein
(MAP) kinase [2,3]. In addition, the CB1 receptor’s activation also
results in inhibition of N- and P/Q-type calcium channels and
stimulation of potassium channels. The endocannabinoid system
(ECS) comprises the cannabinoid receptors, endocannabinoids, and
the corresponding enzymes involved in synthesis and degradation
of endocannabinoids. The pharmacological studies have shown
that the ECS regulates many physiological functions and processes
including pain, energy balance, emotional regulation, etc [4e7]. For
instance, S-777469 [8] (structure A in Fig. 1), as a CB2 receptor
agonist, has completed phase II clinical trials for the treatment of
atopic dermatitis. Yang et al. [9e11] reported the novel
RANKL-induced osteoclast formation, which offered a wonderful
beginning to develop a novel agent for the treatment of osteopo-
rosis. The CB1 receptor selective antagonists were effective as
therapeutic agents for obesity and related chronic diseases, such as
type II diabetes, cardiovascular disease [12,13]. For example,
Rimonabant (SR141716A, structure B in Fig. 1) was developed as an
anti-obesity drug by reducing food intake in Europe. However,
Rimonabant was withdrawn from the market owing to its signifi-
cant psychiatric side effects. By now, the treatment choices of
obesity are limited while the number of obesity people is increasing
in every year, so there remains a significantly unsatisfied need for
the development of anti-obesity agents. Rimonabant displays
comprehensive pharmacological effects for obesity and associated
metabolic disorders, indicating the CB1 receptor is an attractive
target for drug discovery [14]. In order to avoid the adverse effects
in CNS, the new strategy is designing novel ligands of the CB1 re-
ceptor with lower bloodebrain barrier (BBB) penetration [15].
In recent years, the identification and optimization of novel
cannabinoid ligand has been a hot topic in drug discovery [16]. The
first reported CB1 receptor selective antagonist, SR141716A ([N-
(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-
methyl-1H-pyrazole-3-carboxamide]), was discovered by Sanofi-
Aventis in 1994 [17]. Later, SR144528 ([N-[(1S)-endo-1,3,3-
trimethylbicycle[2.2.1]-heptan-2-yl]-5-(4-chloro-3-methylphenyl
* Corresponding author. Tel.: þ86 571 88208659.
E-mail address: chjz@zju.edu.cn (J.-Z. Chen).
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.12.018

74
S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
Fig. 1. Chemical structures of some known cannabinoid ligands.
)-1-(4-methylbenzyl)-pyrazole-3-carboxamide], structure
C
in
docking simulations were performed to gain insight to the
receptor-ligand interactions and to explain the possibly selective
characteristics of the compounds with different substituents. The
docking simulations provided more information on the differences
between the two binding sites of CB1 and CB2 receptors, which
would afford guidelines for the design of new cannabinoid ligand
with improved bioactive and selective profiles.
Fig. 1), an analog of SR141716A, was developed to be the first CB2
receptor selective inverse agonist/antagonist [18,19]. With further
research of the potent diarylpyrazole derivatives, several structural
modifications were achieved on the basis of these parent com-
pounds. Some compounds with different substituents on the pyr-
azole ring or the carboxamide part were synthesized [20e24], and
others were prepared with isosteric replacement of the pyrazole
core structure [25e27]. In the meantime, a number of cannabinoid
ligands with different chemical scaffolds have emerged. As shown
in Fig. 1, bioactive cannabinoid compounds A, D [28], and E [29]
share similarly structural characteristics containing a carbonyl
group at the b-position of the carboxamide part, inducing a privi-
leged structure for designing original cannabinoid ligands. In
addition, the scaffold hopping strategy is a useful approach for drug
discovery by modifying the core structure of promising com-
pounds. Combination of scaffold hopping and bioisostere method
has been previously applied in identification of novel cannabinoid
ligands [30,31].
In the present paper, we describe the design, syntheses, bio-
logical evaluations, preliminary structureeactivity relationship
(SAR) analyses, and docking simulations of a new serious of
cannabinoid ligands with 1,2,4-triazolone scaffold. At first, the
structural skeleton was designed using a combined method of
scaffold hopping and privileged structure-oriented approaches.
Structural modifications were then carried out with guideline of a
pharmacophore model derived from the previous reported 3D-
QSAR model [32]. The cell-based calcium current assays were
applied to study the functional activity of synthesized 1,2,4-
triazolone derivatives. The bioactive tests indicated that our
designed compounds displayed interesting antagonistic activity
towards either CB1 or CB2 receptor.
In order to calculate the interaction modes of our synthesized
1,2,4-triazolones binding to the CB1 and CB2 receptors, respec-
tively, 3D homology models of both CB1 and CB2 receptors were
constructed based on the crystal structure of
b2-adrenergic re-
Fig. 2. Pharmacophore model based on the reported favorable conformation of pyr-
azole compounds for the CB1 receptor (A) and molecular alignment of both SR141716A,
which carbon atoms were drawn in white, and typical compound of 1,2,4-triazolone
with carbon atoms colored in orange (B). (For interpretation of the references to co-
lor in this figure legend, the reader is referred to the web version of this article.)
ceptor [33], and the conformations of both CB receptors were then
optimized by molecular dynamics (MD) simulations with the pro-
teins embedded in a hydrate bilayer system. Based on the gener-
ated homology models of both CB1 and CB2 receptors, flexible

S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
75
Table 1
Chemical structures and in vitro cannabinoid receptor activity of 1,2,4-triazolones (5aem).
No.
R1
Cl
R2
R3
CB1 inhibition/%a
98.6
CB1 IC50/nMb
388
CB2 inhibition/%a
94.0
CB2 IC50/nMb
5a
1510
5b
Cl
74.6
8440
85.6
20000
5c
5d
5e
5f
Cl
102.2
42.6
36.8
98.0
90.5
222
NT
46.4
87.7
NT
Cl
1930
NT
Cl
NT
ꢀ55.3
130.8
43.6
OCH3
OCH3
1840
4730
4090
NT
5g
5h
5i
CH3
CH3
60.9
89.0
17800
3750
132.3
29.9
5030
NT
5j
CH3
98.2
91.9
96.6
96.5
2100
947
58.5
ꢀ25.7
74.0
NT
5k
5l
CH3
NT
OCH3
OCH3
2280
878
NT
5m
123.5
6080
JTE-907
SR141716A
37.6
100.0
NT
13
129.5
100.0
644
9800
a
CB1 and CB2 receptors inhibition (% basal) at 10
NT ¼ not tested.
m
M ligand concentration.
b
2. Results and discussion
development of novel bioactive cannabinoid ligands. As illustrated
in Fig. 2A, we first extracted the pharmacophore model of ligand for
the CB1 receptor according to the preferred conformation of AM263
embedded in 3D-QSAR model [32]. 1,2,4-triazolone ring was
applied to replace the pyrazole ring according to the discipline of
scaffold hopping strategy. The substitutions were then designed on
the basis of 3D-QSAR models generated with the pyrazole de-
rivatives [32]. As illustrated in the reported CoMFA contour maps, a
2.1. Compound design
Chen et al. [32] carried out a combined study by computational
3D-QSAR CoMFA and 2D NMR method on pyrazole compounds to
explore the structural requirements for both CB1 and CB2 re-
ceptors, and the results provided us helpful criteria for the

76
S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
para-substituted phenyl group was good for bioactivity at 5-
position of the pyrazole ring (Fig. 1). In addition, the red area of
the CB1 3D-QSAR model in this region suggested a negatively
partial-charged unit would increase ligand’s bioactivity, while a
yellow contour in the corresponding part of the CB2 3D-QSAR
model indicated the steric bulky groups would not be favored for
the CB2 receptor. Therefore, as the general structure given in
Table 1, it was designed to introduce the groups of chlorine, methyl,
and methoxyl, respectively, to study the influence of subunits with
different electric charge at R1 part of our designed 1,2,4-triazolone
compounds. The CB1 3D-QSAR model had a red region around the
4-position on pyrazole, while by contrast, blue and yellow contours
were displayed in the reported CB2 3D-QSAR model, which implied
an increasing negatively partial-charged group was superior for
CB1 affinity. Herein, a third nitrogen atom was introduced at the
corresponding 2-position of 1,2,4-triazolone ring with the purpose
of further improving CB1 selectivity. Both CoMFA models showed
red contours near the carbonyl oxygen and blue contours near the
amide nitrogen, indicating this part was favored for both CB1 and
CB2 receptors, so the amide moiety was maintained to be H-bond
donor and acceptor, respectively, for the newly designed com-
pounds. Moreover, large yellow areas were displayed around the N-
substituted amide in both 3D-QSAR models, implying limited
binding pockets were presented in both CB receptors. Therefore, we
selected cyclohexyl and substituted phenyl subunits based on the
favorable groups of lead pyrazoles. CoMFA models gave contrasting
steric maps around the 1-position group of pyrazole, which sug-
gested this group could play a key role in affecting the selectivity for
the subtypes of CB receptors, and an aromatic ring was favored for
CB1 receptor affinity. Consequently, we introduced substituted
phenyl and benzyl groups to study their influence on bioactivity for
both receptors. As shown in Fig. 2B, a typical compound was
selected from the designed 1,2,4-triazolones to be aligned with
SR141716A based on the pharmacophore model, and their confor-
mations were consistent with each other.
Fig. 3. Whole-cell calcium assays using CHO cells expressing human CB1 receptor and
15/16. All experiments were performed in triplicate.
Ga
[35,36]. Reactions of the intermediates 4 with different isocyanates
produced the 1,2,4-triazolones 5aem.
2.3. Biological evaluations and SAR analyses
Fluorescent imaging plate reader technology [37] has played a
significant role in high-throughput screening (HTS) for GPCR and
ion channel targets [38e40]. The above synthesized 1,2,4-
triazolone derivatives, as well as the positive controls SR141716A
and JTE-907 [41] were evaluated with their functional activity for
the CB1 and CB2 receptors, respectively, on the cell-based calcium
current assays. Table 1 summarizes their structures and the cor-
responding CB1 and CB2 receptors inhibitory activity. In addition,
Fig. 3 illustrates all ligands’ IC50 curves obtained from the assays for
the CB1 receptor. The results indicated that most of the newly
synthesized compounds presented the CB1 receptor antagonistic
2.2. Chemistry
activity with IC50 values ranging from 222 nM to 17.8
compounds had the IC50 values in the range of 1.51
the CB2 receptor.
m
M, and some
mMe20 M to
m
The general procedure for the syntheses of 1,2,4-triazolone de-
rivatives was summarized in Scheme 1. Our successful syntheses of
1,2,4-triazolones were started with para-position substituted ben-
zoic acids 1. The esters 2 were formed by refluxing the acids in
MeOH with sulfuric acid as a catalyst, and then the esters were
converted to the acid hydrazides 3 by refluxing with an excess of
hydrazine hydrate [34]. Reactions of hydrazides 3 and isocyanates
gave aroylsemicarbazides, and cyclization of these semicarbazides
in refluxing aqueous sodium hydroxide afforded compounds 4
The differences at R2 and R3 regions of the compounds influ-
enced their bioactivity for the subtypes of cannabinoid receptors.
The substituted phenyl group at R2 part was demonstrated a su-
perior CB1 receptor inhibitory activity over the CB2 receptor (5aec,
5feg, 5je5m). Remarkably, a substituted benzyl group introduced
on R2 position resulted in an obvious increase in ligand’s bioactivity
and selectivity for the CB2 receptor (5d, 5h). For R3 group, either
substituted phenyl or cyclohexyl group was tolerated for the CB1
Scheme 1. Reagents and conditions: (a) conc. H2SO4, MeOH, reflux, 24 h; (b) excess NH2NH2, EtOH, reflux, 5 h; (c) R2NCO, THF, reflux, 24 h; (d) 1 N NaOH(aq), reflux, 24 h; (e)
R3NCO, DBU, CH3CN, reflux, overnight.

S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
77
Fig. 4. Sequence alignment of CB1, CB2, and
b2-adrenergic receptors. Conserved residues are displayed with blue background, and the T4L part of 2RH1 is not shown in this figure
for short. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
receptor bioactivity, and the cyclohexyl group was generally helpful
(5a > 5b, 5f > 5g, 5m > 5l). On the other hand, comparing com-
pounds 5d vs. 5e, and 5h vs. 5i, the substituted phenyl group at R3
led to abolish the CB2 receptor bioactivity. Various substituents for
the R1 part showed similar bioactivity. Usually, a chlorine atom was
more favored than methyl or methoxyl (5c > 5f, 5d > 5h) group.
The in vitro active data are congruent with the published CoMFA
results [32], as discussed in the section of Compound design, of the
parent pyrazole compounds.
regions, seven TM helixes and one juxtamembrane helix of the CB1
or CB2 receptor were produced directly based on the coordinates of
corresponding domains of 2-AR. The structures of loop regions
b
connected the TM domains were obtained using loop search from
the whole PDB bank. It was selected manually of the conformations
of loops according to the most reasonable alignment and the
smallest root mean square deviation value (RMSD) at the terminal
residues of the structurally conserved regions (SCR). In addition,
based on the references reported before [42,52], both CB1 and CB2
receptors contain a disulfide bridge in second extracellular loop. In
the present models, the SeS bonds which formed between Cys257
and Cys264 in the CB1 receptor as well as Cys174 and Cys179 in the
CB2 receptor were also taken into account.
2.4. Generations of homology models
The 1,2,4-triazolone derivatives displayed interesting antago-
nistic activity to the CB1 and CB2 receptors, which attracted our
attention for further exploring the binding modes of these com-
pounds with the cannabinoid receptors. Since both CB1 and CB2
receptors are membrane proteins, their crystal structures still
remain experimentally unresolved. In the past years, many 3D
models of the cannabinoid receptors have been generated based on
the Rhodopsin’s crystal structure [42e45]. Recent breakthroughs in
crystallographic studies of GPCRs, as well as the homology
modeling method, would help us to build more accurate homology
models of cannabinoid receptors. Recently, some successful CB1
and CB2 receptor models [46e51] were also reported based on the
2.5. MD simulations
The conformations from homology modeling were refined by
MD simulations using AMBER 11 software package [53]. Consid-
ering that both CB receptors are membrane proteins, the initially
generated homology structures were respectively embedded into a
pre-equilibrated aqueous membrane system [54]. The MD simula-
tions were then performed for each of the whole protein/DPPC/
TIP3P systems. In order to keep the active conformations of our
ligand-free receptors, in the equilibration process the force
X-ray structure of
b2-adrenergic receptor (b2-AR) [33], which
shares higher identity with the CB receptors than Rhodopsin.
However, these generated homology models were not put into a
membrane environment for MD optimizations. In order to simulate
the interaction modes of our synthesized compounds binding to
the CB1 and CB2 receptors, we built up our own coordinates for the
structural models of both CB1 and CB2 receptors based on the
crystal structure of
b2-AR [33] with a general procedure of ho-
mology modeling and MD refinements in
a
membrane
environment.
At first, we constructed the homology models of both CB1 and
CB2 receptors to represent their active state based on the X-ray
structure of b2-adrenergic receptor [33] with high resolution. Both
CB1 and CB2 receptors enclose most of the highly conserved resi-
dues and motifs in GPCRs superfamily, such as D(E)RY in the third
transmembrane domain (TM3), CWXP in TM6, and NPXXY in TM7,
which is helpful for sequence alignment and homology model
generation. Fig. 4 shows the sequence alignment results of CB1,
Fig. 5. Backbone RMSD for both apo CB1 receptor (black) and ligand-free CB2 receptor
(red) trajectories of production simulation. (For interpretation of the references to
color in this figure legend, the reader is referred to the web version of this article.)
CB2, and
b2 (PDB code: 2RH1) receptors. On the grounds that all the
GPCRs have conserved secondary structure in the transmembrane

78
S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
Fig. 6. MD refined conformations of CB1/DPPC/TIP3P complex (A) and CB2/DPPC/TIP3P system (B). The proteins display in cartoon mode, and carbon atoms of DPPC are colored in
green. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
restraint weight on the atoms of backbone was slowly decreased
from 50 to 1 kcal/(mol A) step by step. During the 30 ns production
2.6. Analyses of interaction modes
ꢀ
simulations, the backbone atoms of the helical domains were still
restrained with a small force constant of 1 kcal/(mol A), while the
Since the 1,2,4-triazolone derivatives exhibited weaker bioac-
tivity than the lead compound SR141716A under the conditions of
our bioassays, we first investigated the differences between the
binding modes of SR141716A and compound 5c, respectively, with
the generated homology model of the CB1 receptor by flexible
docking studies using Sybyl-X1.3/FlexiDock [57]. Fig. 8A shows the
docking simulated interaction mode between SR141716A and the
CB1 receptor, which correlates with its binding conformation
calculated by Salo et al. [43]. The critical lysine K192 [43] and the
amide oxygen atom of SR141716A forms a hydrogen bond with the
length and angle of 2.54 A and 170.2 , respectively. Furthermore,
McAllister et al. [58] reported the binding of SR141716A was
affected by the F200A, W279A and W356A mutations, indicating
these residues are part of the binding site. In our simulated model,
the p-Cl phenyl moiety contacts with a hydrophobic cavity
composed by the residues V196, F200, W356, and L360 of the CB1
receptor. The 2,4-dichlorophenyl group situates in another hydro-
phobic cleft formed by W279 and M363, and it is observed a par-
ꢀ
loop regions and the side chains of all the residues were free for
minimization. As shown in Fig. 5, the RMSD of the backbone atoms
were calculated for each system. The conformations of both pro-
teins kept stable during the last 15 ns of production. After 30 ns MD
simulations, the refined conformations of both receptors were
extracted for further docking study. Fig. 6 shows the final confor-
mations of both CB receptors in the DPPC/TIP3P system. As indi-
cated in Fig. 7 for the Ramachandran jeF distribution plots of the
ꢁ
ꢀ
CB receptors, 96.1% of the residues located in the favored regions
and other 3.9% resided in allowed areas for the model of the CB1
receptor, while for the CB2 receptor model, the number was 94.6%
and 5.4%, respectively. Subsequently, the accuracy and validity of
each model was tested with the profiles-3D [55] program of
Accelrys Discovery Studio 2.5.5 [56]. The MD-optimized homology
models of the CB1 and CB2 receptors were scored 145.3 and 148.7,
respectively, comparing with the corresponding expected high
value of 140.6 and 133.8 and lowest possible score of 63.3 and 60.2.
The results of profiles-3D analysis indicated both generated CB1
and CB2 models would be rationalities with high probabilities.
allel
pep stacking consisting of three aromatic rings from 2,4-
dichlorophenyl, W279, and Y275. The importance of aromatic
stacking at Y275 for ligand recognition has been demonstrated by
Fig. 7. Ramachandran plot for the MD refined CB1 receptor (A) and CB2 receptor (B) homology model. Glycine residues are displayed as triangles, prolines are shown as squares, and
other residues are illustrated as circles. The regions surrounded by blue lines are the favorable areas, and sections enclosed by purple lines are the allowed areas. (For interpretation
of the references to color in this figure legend, the reader is referred to the web version of this article.)

S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
79
McAllister et al. [59] through site mutation and modeling studies. In
addition, the piperidyl group inserts into the third pocket sur-
rounded by F177, F189, and W255. Later on, the docking-simulated
CB1-5c complex was superimposed on the CB1-SR141716A model
based on the backbone atoms of the CB1 receptor as shown in
Fig. 8A. It is indicated compound 5c shares quite similar binding
mode with SR141716A in the same active site of the CB1 receptor.
Moreover, it is accepted that the aromatic and hydrophobic in-
teractions are crucial for affinity of cannabinoid ligands [60,61].
On the other hand, the C4 methyl group on the pyrazole ring
forms hydrophobic interaction with the side chain of residue V196
in the docking-simulated CB1-SR141716A complex (Fig. 8A). How-
ever, such a hydrophobic interaction is missing in the binding mode
of compound 5c with the CB1 receptor since introducing the third
nitrogen atom at the corresponding position of its core ring. In
addition, SR141716A contains a much more hydrophobic 2,4-
dichlorophenyl moiety in the corresponding region, which can
produce stronger interaction to boost CB1 receptor activity, by
pep
comparing with the 4-methylphenyl group at R2 position of com-
pound 5c. These results probably explain that compound 5c has
lower CB1 receptor bioactivity than SR141716A. Combined with the
bioactive data obtained from calcium current assays, the docking
simulations further imply that it could be beneficial for improving
the CB1 receptor bioactivity by increasing the lipophilicity of sub-
stituents on the core scaffold. The results would provide a helpful
guideline for future structural optimization on our synthesized
compounds.
Consequently, among the 1,2,4-triazolone derivatives we syn-
thesized, compounds 5c and 5d, which only had a slight difference
at R2 part, displayed the contrast selectivity for the subtype of
cannabinoid receptors. Compound 5c with a p-tolyl group exhibi-
ted the best inhibitory activity and selectivity for the CB1 receptor
Fig. 8. (A) Superimposition of the docking results of CB1 receptor with SR141716A and 5c based on backbone atoms, and docking results of CB1-5c (B), CB1-5d (C), CB2-5c (D), and
CB2-5d (E) complexes. The carbon atoms of CB1 receptor, CB2 receptor, SR141716A, 5c, and 5d are colored by cyan, green, light blue, magentas, and yellow, respectively. The H-bonds
are shown as red dashed lines, parallel
pep interactions are illustrated by blue dashed lines, cation-p interactions are displayed by green dashed lines, and Sep interactions are
indicated by orange dashed lines. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

80
S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
but only 46.4% CB2 receptor inhibition at 10
m
M. While changing
The flexible docking studies indicated that the pocket formed by
W172, R177, Y190, and W194 would be crucial to the ligand
recognition of the CB2 receptor.
the building block to 4-methylbenzyl group, compound 5d was a
CB2 receptor selective antagonist with 3-fold lower bioactivity
comparing with the reference compound JTE-907 under the same
assay conditions, and it displayed only 42.6% CB1 receptor inhibi-
On the basis of the docking simulation results, the amide part is
the significant H-bond forming site for both CB1 and CB2 receptors.
For the 1,2,4-triazolone derivatives, the van der Waals interaction is
the dominated binding mode with the conserved aromatic resi-
dues, including F3.36, W4.64, Y5.39, W5.43, and W6.48, of canna-
binoid receptors. Comparing the docking-simulated interaction
modes of compounds 5c and 5d binding to the CB1 or CB2 receptor,
the selective property could be mainly caused by the interaction
between R2 group and the ambient residues, such as the residues
W279 and M363 of the CB1 receptor and the residues W172, R177,
and W194 of the CB2 receptor.
tory effect at the concentration of 10
mM. Thus, comparative
docking studies were carried out to better understand the binding
modes between 1,2,4-triazolone derivatives and the CB receptors
and to further predict possibly important residues affecting the
selectivity.
As shown in Fig. 8B and C, K192 forms a hydrogen bond with the
oxygen atom of amide group in both docking-simulated CB1-5c and
CB1-5d complexes. For compound 5c, the length and angle of the
ꢁ
ꢀ
H-bond are 1.91 A and 152.1 , respectively, and for 5d, the values are
ꢁ
ꢀ
2.04 A and 151.9 , correspondingly. The H-bond interaction implies
the amide linker is necessary to maintain the activity for the CB1
receptor. The identical R1 and R3 subunits of both ligands display
same interaction with CB1 receptor. Furthermore, it is observed
distinctions between hydrophobic interactions producing by
W279/M363 motif and the R2 group with phenyl or benzyl group.
In the CB1-5c complex, the p-tolyl group produces a face-to-face
3. Conclusion
In the present work, a series of 1,2,4-triazolone derivatives was
synthesized and evaluated for their in vitro cannabinoid receptor
bioactivity. It was obtained of interesting antagonistic activity
depending on various building blocks of the new compounds.
Hence, we discovered that the 1,2,4-triazolone skeleton was a novel
potential chemical scaffold for cannabinoid receptor antagonists. In
order to gain a better understanding of the receptor-ligand in-
teractions, homology models of both CB1 and CB2 receptors in
pe
p
interaction with W279, and in the meantime, W279 also
forms effect with the side chain of Y275. Whilst for ligand 5d, the
4-methylbenzyl substituent interacts with W279 through a T-sha-
ped interaction, which is obviously weaker than parallel
p
pe
p
pep
interaction in the CB1-5c mode. In addition, a Met-aromatic
active form were constructed based on the crystallized b2-
interaction is formed between p-tolyl of 5c and M363 with an Se
adrenergic receptor, and the models were embedded in a DPPC/
TIP3P bilayer system for the conformational refinements by MD
simulations. The comparison of docking-calculated CB1-SR141716A
and CB1-5c complexes implied the hydrophobic and aromatic in-
teractions would be important for maintaining high CB1 receptor
activity. Later, docking method was also applied to compute the
binding modes between typical 1,2,4-triazolone ligands and each
subtype of the CB receptors. The results indicated that K192 of the
CB1 receptor and S285 of the CB2 receptor respectively formed H-
bond with the oxygen atom of ligands’ carboxyl group. Further-
more, W279 and M363 of the CB1 receptor as well as W172, R177,
and W194 of the CB2 receptor displayed different interaction
modes with the changed R2 group, implying they might be
important residues affecting the selectivity of ligand binding to
cannabinoid receptors. By now, with the aim of identifying new
drug candidate with better safety, the major task of developing
cannabinoid ligands is improving selective property. The simulated
binding modes of our synthesized compounds with the homology
models of the CB1 and CB2 receptors in this work would be useful
for understanding the interaction mechanism and designing new
ligands with improved potency and selectivity.
ꢀ
p
distance of 3.24 A. However, in the CB1-5d complex, the Se
p
ꢀ
distance is 3.95 A. Recently, Cordomí [62] reviewed the sulfur-
containing amino acids which formed MeteMet and Met/Cys-
aromatic interactions played a key role in pharmacology and
function of seven-transmembrane receptors. In summary, changing
R2 unit may result in different interactions with residues W279 and
M363, indicating these two residues could be important residues
for the selectivity of ligand to the CB1 receptor.
Fig. 8D and E illustrate the docking simulation results of CB2-
ligand complexes. The oxygen atom of amide forms H-bond with
the hydroxyl group of S285, which is an important H-bond site for
kinds of cannabinoid ligands [49,60]. For compounds 5c and 5d, the
ꢁ
ꢁ
ꢀ
ꢀ
H-bond lengths and angles are 1.85 A, 164.2 and 1.99 A, 162.9 ,
respectively. In addition, the carboxyl group on the core ring can
form another two H-bonds with R177 in the second extraceꢁllular
ꢀ
loop (ECL2) with the bond lengths and angles of 2.41 A, 137.0 and
ꢁ
ꢁ
ꢀ
ꢀ
ꢀ
1.95 A,153.3 in CB2-5c complex as well as 2.57 A,132.7 and 1.85 A,
157.9ꢁ for CB2-5d interaction. Comparing to the binding mode of
compound 5d with the CB1 receptor, its R3 cyclohexyl ring locates
in a smaller hydrophobic pocket enclosed by F91, F94, and K278 of
the CB2 receptor, which explains the relatively flexible cyclohexyl
to be superior to the rigid phenyl group of other compound, like 5e
or 5i, for the CB2 receptor bioactivity of ligand. In addition, the p-Cl
phenyl group situates in the hydrophobic region encircled by V113,
F117, W258, V261, and C288, which is similar with the CB1-ligand
interaction mode. For the R2 site, more apparent differentiations
are obtained based on the docking simulations. For compound 5c,
the rigid R2 structure leads to a handicap for proper contact with
the pocket surrounded by the conserved aromatic residues W172,
4. Experimental section
4.1. Compound synthesis
4.1.1. Materials and instrumentation
Melting points were measured with a Büchi Melting Point B-540
apparatus (Büchi Labortechnik, Flawil, Switzerland) and are un-
corrected. High-resolution LC-MS (HRMS) data were obtained using
ESI ionization on Agilent 1290-6200 TOF LC-MS (Agilent, Palo Alto,
CA, USA). Mass spectra (MS) were taken in ESI mode on Agilent
1100 LC-MS (Agilent, Palo Alto, CA, USA). 1H NMR spectra was
recorded on Bruker 500 MHz spectrometers (Bruker Bioscience,
Billerica, MA, USA) using TMS as an internal standard. The values of
Y190, and W194, and no
pep interaction is observed from our
modeling simulation results, which could be the major factor to
induce the loss of its CB2 receptor bioactivity. By contrast, in
docking simulated CB2-5d complex, the 4-methylbenzyl group
deeply inserts into the corresponding pocket. A parallel
pep
interaction forms between 5d’s benzyl group and W172’s indolyl
ring with a distance of 4.42 A. 5d’s benzyl group also interacts with
chemical shifts (d) were given in ppm and coupling constants (J) in
Hz. The purity was determined by high-performance liquid chro-
matography (HPLC) performed on COSMOSIL C18-MS-II column
ꢀ
W194 via an edge-to-face
p stacking. Meanwhile, R177 can form
ꢀ
cation- interaction to 5d’s benzyl unit with a distance of 3.88 A.
p
(4.6 mm ꢂ 250 mm, 5
mm) using solvent of methanol/water at the

S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
81
flow rate of 1 mL/min and peak detection at 254 nm under UV. All
materials were used as received. Tetrahydrofuran (THF) and ether
(Et2O) were dried by distillation from sodium-benzophenone.
Acetonitrile was dried by distillation from P2O5. All other solvents
were used as received and were reagent grade where available.
4.1.4.4. 3-(4-Methoxyphenyl)-4-(p-tolyl)-1H-1,2,4-triazol-5(4H)-one
(4d). Light yellow solid. Yield: 54.5%. Melting Point: > 250 ꢁC. LCMS
(ESI): [M þ H]þ 282.
4.1.4.5. 4-(4-Chlorophenyl)-3-(4-methoxyphenyl)-1H-1,2,4-triazol-
5(4H)-one (4e). Light yellow solid. Yield: 57.2%. Melting Point:
203.3e204.8 ꢁC. LCMS (ESI): [M þ H]þ 302.
4.1.2. General procedure for synthesis of 4-substituted methyl
benzoate derivatives 2ae2c
4.1.4.6. 4-(4-Methylbenzyl)-3-(p-tolyl)-1H-1,2,4-triazol-5(4H)-one
(4f). Offwhite solid. Yield: 60.8%. Melting Point: > 250 ꢁC. LCMS
(ESI): [M þ H]þ 280.
4-substituted benzoic acid 1 (10 mmol) was refluxed in 10 mL
MeOH catalyzed by sulfuric acid for 24 h. The solvent was
concentrated in vacuo, and the crude product was purified by flash
silica gel chromatography (petroleum ether/ethyl acetate ¼ 10:1, v/
v) to give the intermediates 2ae2c.
4.1.4.7. 4-(4-Chlorophenyl)-3-(p-tolyl)-1H-1,2,4-triazol-5(4H)-one
(4g). Light yellow solid. Yield: 82.4%. Melting Point: 207.9e
208.9 ꢁC. LCMS (ESI): [M þ H]þ 286.
4.1.2.1. Methyl 4-chlorobenzoate (2a). White solid. Yield: 87.3%.
Melting Point: 38.6e39.9 ꢁC (Ref. melting point 39 ꢁC [63]).
4.1.5. General procedure for synthesis of 1,2,4-triazolone derivatives
5aem
4.1.2.2. Methyl 4-methoxybenzoate (2b). White solid. Yield: 93.6%.
Melting Point: 46.4e48.5 ꢁC (Ref. melting point 47e48 ꢁC [64]).
To a stirred solution of 4 (0.30 mmol) and 1,8-diazabicyclo[5.4.0]
undec-7-ene (DBU) (0.60 mmol) in 3 mL dry acetonitrile under
nitrogen was added dropwise isocyanate (0.36 mmol). After being
refluxed overnight, the reaction was diluted with ethyl acetate. The
reaction was transferred to a separatory funnel where it was
washed three times with water and once with saturated aqueous
sodium chloride solution. The organic layer was dried over anhy-
drous MgSO4 before it was evaporated at reduced pressure. The
resulting solid was purified by flash chromatography (petroleum
ether/ethyl acetate ¼ 3:1, v/v) to give the target compound 5aem.
4.1.2.3. Methyl 4-methylbenzoate (2c). White solid. Yield: 93.0%.
Melting Point: 30.0e32.0 ꢁC (Ref. melting point 32e35 ꢁC [65]).
4.1.3. General procedure for synthesis of 4-substituted
benzohydrazide 3ae3c
A solution of hydrazine hydrate (20.00 mmol) in 2 mL EtOH was
added dropwise to the ester 2 (5.00 mmol). The mixture was
refluxed for 5 h and filtered, and the corresponding acid hydrazide
3 was obtained by washing the residue with ice water.
4.1.5.1. 3-(4-Chlorophenyl)-N-cyclohexyl-4-(4-methoxyphenyl)-5-
oxo-4,5-dihydro-1H-1,2,4-triazole-1-carboxamide (5a). Light yellow
solid. Yield: 18.3%. Melting Point: 71.6e72.9 ꢁC. HPLC (90% meth-
anol in water): tR ¼ 5.986 min, 97.0%. 1H NMR (CDCl3, 500 MHz):
4.1.3.1. 4-Chlorobenzohydrazide (3a). White solid. Yield: 93.4%.
Melting Point: 162.0e163.7 ꢁC (Ref. melting point 162e163 ꢁC [66]).
d
8.16 (d, J ¼ 8.0 Hz, 1H), 7.38 (d, J ¼ 8.0 Hz, 2H), 7.28 (d, J ¼ 8.0 Hz,
2H), 7.15 (d, J ¼ 8.5 Hz, 2H), 6.97 (d, J ¼ 8.5 Hz, 2H), 3.95e3.89 (m,
1H), 3.85 (s, 3H), 2.00e1.98 (m, 2H), 1.73e1.70 (m, 2H), 1.60e1.58
(m,1H),1.44e1.31 (m, 4H),1.27e1.26 (m, 1H). HRMS (ESI): m/z calcd
for C22H23ClN4O3 [M]þ 426.1459; found: m/z 426.1463.
4.1.3.2. 4-Methoxybenzohydrazide (3b). White solid. Yield: 70.0%.
Melting Point: 138.0e139.7 ꢁC (Ref. melting point 136e140 ꢁC [67]).
4.1.3.3. 4-Methylbenzohydrazide (3c). White solid. Yield: 75.5%.
Melting Point: 113.1e114.4 ꢁC (Ref. melting point 112e114 ꢁC [68]).
4.1.5.2. 3-(4-Chlorophenyl)-4-(4-methoxyphenyl)-5-oxo-N-(p-tolyl)-
4,5-dihydro-1H-1,2,4-triazole-1-carboxamide (5b). Light yellow
solid. Yield: 15.6%. Melting point: 162.4e164.8 ꢁC. HPLC (90%
methanol in water): tR ¼ 5.379 min, 100.0%. 1H NMR (CDCl3,
4.1.4. General procedure for synthesis of 3,4-disubstituted 1h-1,2,4-
triazol-5(4H)-one 4ae4g
A stirred mixture of 3 (3.00 mmol) and 10 mL dry THF was
warmed until a homogenous solution was obtained, and then iso-
cyanate (3.60 mmol) was added via syringe. After being refluxed for
24 h, the mixture was diluted with Et2O and the residue was
collected by filtration. The precipitate was dissolved in 20 mL 1 N
NaOH aqueous solution and refluxed for 24 h. The mixture was
cooled in an ice bath and slowly acidified with 2 N HCl aqueous
solution. The resulting precipitate was collected by filtration to give
compound 4.
500 MHz):
d
10.17 (s, 1H), 7.49 (d, J ¼ 8.5 Hz, 2H), 7.38 (d, J ¼ 8.5 Hz,
2H), 7.28 (d, J ¼ 8.5 Hz, 4H), 7.24 (d, J ¼ 8.5 Hz, 2H), 7.17 (d,
J ¼ 9.0 Hz, 4H), 6.99 (d, J ¼ 9.0 Hz, 2H), 3.85 (s, 3H), 2.33 (s, 3H).
HRMS (ESI): m/z calcd for C23H19ClN4O3 [M]þ 434.1146; found: m/z
434.1151.
4.1.5.3. 3-(4-Chlorophenyl)-N-cyclohexyl-5-oxo-4-(p-tolyl)-4,5-
dihydro-1H-1,2,4-triazole-1-carboxamide (5c). White solid. Yield:
19.1%. Melting point: 163.7e164.9 ꢁC. HPLC (90% methanol in wa-
ter): tR ¼ 5.892 min, 97.1%. 1H NMR (CDCl3, 500 MHz):
d 8.17 (d,
4.1.4.1. 3-(4-Chlorophenyl)-4-(4-methoxyphenyl)-1H-1,2,4-triazol-
5(4H)-one (4a). Brown solid. Yield: 53.5%. Melting Point: 204.2e
205.8 ꢁC. HRMS (ESI): m/z calcd for C15H12ClN3O2 [M]þ 301.0618;
found: m/z 301.0614. LCMS (ESI): [M þ H]þ 302.
J ¼ 7.5 Hz,1H), 7.35 (d, J ¼ 8.5 Hz, 2H), 7.27 (d, J ¼ 7.0 Hz, 4H), 7.10 (d,
J ¼ 7.5 Hz, 2H), 3.92e3.91 (m, 1H), 2.41 (s, 3H), 2.00e1.98 (m, 2H),
1.73e1.70 (m, 2H), 1.60e1.58 (m, 1H), 1.43e1.33 (m, 4H), 1.26e1.24
(m, 1H). HRMS (ESI): m/z calcd for C22H23ClN4O2 [M]þ 410.1510;
found: m/z 410.1512.
4.1.4.2. 3-(4-Chlorophenyl)-4-(p-tolyl)-1H-1,2,4-triazol-5(4H)-one
(4b). Light brown solid. Yield: 55.0%. Melting Point: 209.0e
210.3 ꢁC. LCMS (ESI): [M þ H]þ 286.
4.1.5.4. 3-(4-Chlorophenyl)-N-cyclohexyl-4-(4-methylbenzyl)-5-oxo-
4,5-dihydro-1H-1,2,4-triazole-1-carboxamide (5d). White solid.
Yield: 19.7%. Melting Point: 156.2e157.5 ꢁC. HPLC (90% methanol in
4.1.4.3. 3-(4-Chlorophenyl)-4-(4-methylbenzyl)-1H-1,2,4-triazol-
5(4H)-one (4c). Offwhite solid. Yield: 64.1%. Melting point:
>250 ꢁC. LCMS (ESI): [M þ H]þ 300.
water): tR ¼ 7.922 min, 99.3%. 1H NMR (CDCl3, 500 MHz):
d 8.08 (d,
J ¼ 7.5 Hz,1H), 7.44 (d, J ¼ 8.5 Hz, 2H), 7.39 (d, J ¼ 8.5 Hz, 2H), 7.12 (d,
J ¼ 8.0 Hz, 2H), 6.99 (d, J ¼ 7.5 Hz, 2H), 4.88 (s, 2H), 3.89e3.87 (m,

82
S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
1H), 2.32 (s, 3H), 2.03e2.00 (m, 2H), 1.74e1.71 (m, 2H), 1.61e1.59
(m,1H),1.44e1.30 (m, 4H),1.26e1.22 (m,1H). HRMS (ESI): m/z calcd
for C23H25ClN4O2 [M]þ 424.1666; found: m/z 424.1663.
tR ¼ 6.812 min, 99.3%. 1H NMR (CDCl3, 500 MHz):
d 10.06 (s, 1H),
7.41 (dd, J ¼ 8.5 Hz and 5.0 Hz, 4H), 7.33 (d, J ¼ 9.5 Hz, 2H), 7.07 (dd,
J ¼ 8.5 Hz and 5.5 Hz, 4H), 6.95 (d, J ¼ 8.5 Hz, 2H), 2.33 (s, 3H), 2.30
(s, 3H). HRMS (ESI): m/z calcd for C23H19ClN4O2 [M]þ 418.1197;
found: m/z 418.1195.
4.1.5.5. 3-(4-Chlorophenyl)-4-(4-methylbenzyl)-5-oxo-N-(p-tolyl)-
4,5-dihydro-1H-1,2,4-triazole-1-carboxamide (5e). Light yellow
solid. Yield: 19.3%. Melting Point: 146.1e148.4 ꢁC. HPLC (90%
methanol in water): tR ¼ 7.235 min, 98.8%. 1H NMR (CDCl3,
4.1.5.12. 4-(4-Chlorophenyl)-3-(4-methoxyphenyl)-5-oxo-N-(p-
tolyl)-4,5-dihydro-1H-1,2,4-triazole-1-carboxamide
(5l). Yellow
500 MHz):
d
10.14 (s, 1H), 7.49 (dd, J ¼ 7.5 Hz and 6.0 Hz, 4H), 7.41
solid. Yield: 38.6%. Melting Point: 144.8e145.6 ꢁC. HPLC (90%
methanol in water): tR ¼ 5.406 min, 97.9%. 1H NMR (CDCl3,
(d, J ¼ 8.0 Hz, 2H), 7.16 (t, J ¼ 9.0 Hz, 4H), 7.03 (d, J ¼ 7.5 Hz, 2H), 4.93
(s, 2H), 2.34 (s, 6H). HRMS (ESI): m/z calcd for C24H21ClN4O2 [M]þ
432.1353; found: m/z 432.1358.
500 MHz):
d
10.09 (s, 1H), 7.49 (d, J ¼ 8.5 Hz, 2H), 7.46 (d, J ¼ 9.0 Hz,
2H), 7.35 (d, J ¼ 8.5 Hz, 2H), 7.21 (d, J ¼ 8.5 Hz, 2H), 7.17 (d,
J ¼ 8.0 Hz, 2H), 6.83 (d, J ¼ 9.0 Hz, 2H), 3.81 (s, 3H), 2.33 (s, 3H).
HRMS (ESI): m/z calcd for C23H19ClN4O3 [M]þ 434.1146; found: m/z
434.1149.
4.1.5.6. N-Cyclohexyl-3-(4-methoxyphenyl)-5-oxo-4-(p-tolyl)-4,5-
dihydro-1H-1,2,4-triazole-1-carboxamide (5f). White solid. Yield:
38.0%. Melting Point: 193.6e194.7 ꢁC. HPLC (90% methanol in wa-
ter): tR ¼ 5.194 min, 95.8%. 1H NMR (CDCl3, 500 MHz):
d
8.18 (d,
4.1.5.13. 4-(4-Chlorophenyl)-N-cyclohexyl-3-(4-methoxyphenyl)-5-
J ¼ 7.5 Hz, 1H), 7.34 (d, J ¼ 9.0 Hz, 2H), 7.24 (d, J ¼ 8.0 Hz, 2H), 7.09
(d, J ¼ 8.0 Hz, 2H), 6.77 (d, J ¼ 9.0 Hz, 2H), 3.93e3.88 (m,1H), 3.77 (s,
3H), 2.39 (s, 3H), 1.98e1.94 (m, 2H), 1.71e1.68 (m, 2H), 1.58e1.56
(m,1H),1.42e1.29 (m, 4H),1.27e1.22 (m,1H). HRMS (ESI): m/z calcd
for C23H26N4O3 [M]þ 406.2005; found: m/z 406.2011.
oxo-4,5-dihydro-1H-1,2,4-triazole-1-carboxamide
(5m). Yellow
solid. Yield: 13.9%. Melting Point: 98.4e99.5 ꢁC. HPLC (90% meth-
anol in water): tR ¼ 5.384 min, 100.0%. 1H NMR (CDCl3, 500 MHz):
d
8.07 (d, J ¼ 8.0 Hz, 1H), 7.43 (d, J ¼ 9.0 Hz, 2H), 7.32 (d, J ¼ 9.0 Hz,
2H), 7.17 (d, J ¼ 9.0 Hz, 2H), 6.81 (d, J ¼ 9.0 Hz, 2H), 3.93e3.91 (m,
1H), 3.80 (s, 3H), 2.01e1.98 (m, 2H), 1.73e1.70 (m, 2H), 1.60e1.58
(m,1H),1.43e1.33 (m, 4H),1.27e1.24 (m,1H). HRMS (ESI): m/z calcd
for C22H23ClN4O3 [M]þ 426.1459; found: m/z 426.1462.
4.1.5.7. N-(4-Chlorophenyl)-3-(4-methoxyphenyl)-5-oxo-4-(p-tolyl)-
4,5-dihydro-1H-1,2,4-triazole-1-carboxamide (5g). White solid.
Yield: 13.9%. Melting Point: 91.0e92.6 ꢁC. HPLC (90% methanol in
water): tR ¼ 5.475 min, 100.0%. 1H NMR (CDCl3, 500 MHz):
d
10.34
4.2. Biological assay
(s, 1H), 7.35 (d, J ¼ 9.0 Hz, 2H), 7.26 (d, J ¼ 9.0 Hz, 4H), 7.11 (d,
J ¼ 8.0 Hz, 2H), 7.02 (d, J ¼ 8.5 Hz, 2H), 6.89 (d, J ¼ 8.0 Hz, 2H), 3.79
(s, 3H), 2.38 (s, 3H). HRMS (ESI): m/z calcd for C23H19ClN4O3 [M]þ
434.1146; found: m/z 434.1150.
Chinese hamster ovarian (CHO) cells stably co-expressing G
16 with either CB1 or CB2 receptor were plated onto 96-well plates
and incubated for 24 h. Cells were loaded with 2 M fluo-4 AM in
a15/
m
Hanks balanced salt solution (HBSS, containing 5.4 mM KCl, 0.3 mM
Na2HPO4, 0.4 mM KH2PO4, 4.2 mM NaHCO3, 1.3 mM CaCl2, 0.5 mM
4.1.5.8. N-Cyclohexyl-4-(4-methylbenzyl)-5-oxo-3-(p-tolyl)-4,5-
dihydro-1H-1,2,4-triazole-1-carboxamide (5h). Yellow solid. Yield:
27.5%. Melting Point: 99.3e100.8 ꢁC. HPLC (90% methanol in water):
MgCl2, 0.6 mM MgSO4, 137 mM NaCl, 5.6 mM
sulfinpyrazone, pH 7.4) at 37 ꢁC for 45 min. After moving the excess
dye, cells were washed with HBSS and 50 L HBSS containing vari-
able concentrations of test compounds (10 pMe100 M), Rimona-
bant, JTE-907 (positive control) or DMSO (negative control) was
added. After 10 min incubation at room temperature, 25 L CP55940
D-glucose and 250 mM
tR ¼ 7.723 min, 97.4%. 1H NMR (CDCl3, 500 MHz):
d
8.10 (d, J ¼ 8.0
m
Hz, 1H), 7.39 (d, J ¼ 8.0 Hz, 2H), 7.21 (d, J ¼ 8.0 Hz, 2H), 7.11 (d,
J ¼ 8.0 Hz, 2H), 7.00 (d, J ¼ 8.0 Hz, 2H), 4.87 (s, 2H), 3.91e3.84 (m,
1H), 2.38 (s, 3H), 2.31 (s, 3H), 2.01e1.98 (m, 2H), 1.74e1.71 (m, 2H),
1.61e1.57 (m, 1H), 1.42e1.31 (m, 4H), 1.26e1.23 (m, 1H). HRMS
m
m
was dispensed into the wells using a FlexStation microplate reader,
and intracellular calcium change was recorded at an excitation
wavelength of 485 nm and an emission wavelength of 525 nm. All
experiments were performed in triplicate. IC50 values were analyzed
with sigmoidal doseeresponse curve fitting using GraphPad Prism.
(ESI): m/z calcd for
C
24H28N4O2 [M]þ 404.2212; found: m/z
404.2215.
4.1.5.9. N-(4-Chlorophenyl)-4-(4-methylbenzyl)-5-oxo-3-(p-tolyl)-
4,5-dihydro-1H-1,2,4-triazole-1-carboxamide (5i). Light yellow
solid. Yield: 13.5%. Melting Point: 77.9e79.2 ꢁC. HPLC (90% meth-
anol in water): tR ¼ 7.416 min, 97.9%. 1H NMR (CDCl3, 500 MHz):
4.3. Computational methods
d
10.25 (s,1H), 7.38 (d, J ¼ 7.5 Hz, 2H), 7.28 (d, J ¼ 9.0 Hz, 2H), 7.20 (d,
4.3.1. Homology models of both CB1 and CB2 receptors
J ¼ 8.0 Hz, 2H), 7.12 (d, J ¼ 8.0 Hz, 2H), 7.10 (d, J ¼ 8.5 Hz, 2H), 6.99
(d, J ¼ 8.0 Hz, 2H), 4.86 (s, 2H), 2.36 (s, 3H), 2.29 (s, 3H). HRMS (ESI):
m/z calcd for C24H21ClN4O2 [M]þ 432.1353; found: m/z 432.1356.
The sequences of CB1 and CB2 receptors were got from the NCBI
protein database. Owing to the flexibility of the N- and C-terminal
domains as well as no direct receptor-ligand interaction reported
for these regions, residues 109e417 of CB1 receptor and residues
26e319 of CB2 receptor were used for the setup of the homology
4.1.5.10. 4-(4-Chlorophenyl)-N-cyclohexyl-5-oxo-3-(p-tolyl)-4,5-
dihydro-1H-1,2,4-triazole-1-carboxamide (5j). White solid. Yield:
37.1%. Melting Point: 161.0e162.3 ꢁC. HPLC (90% methanol in wa-
models. The X-ray structure of human b2-adrenergic receptor (PDB
ter): tR ¼ 5.835 min, 100.0%. 1H NMR (CDCl3, 500 MHz):
d
8.09 (d,
code: 2RH1, resolution ¼ 2.40 A) was downloaded from the RCSB
Protein Data Bank website (http://www.pdb.org) as the basic
template. Considering both CB1 and CB2 receptors contain many of
the structurally conserved motifs associated with GPCRs, as shown
in Fig. 4, the sequential alignment was carried out based on the
structural characterization of 2RH1. Since the seven trans-
membrane helical regions of GPCRs are highly conserved, the co-
ꢀ
J ¼ 7.5 Hz,1H), 7.35 (d, J ¼ 8.0 Hz, 2H), 7.28 (d, J ¼ 8.0 Hz, 2H), 7.19 (d,
J ¼ 7.5 Hz, 2H), 7.02 (d, J ¼ 8.0 Hz, 2H), 3.85e3.83 (m, 1H), 2.34 (s,
3H), 1.93e1.91 (m, 2H), 1.66e1.63 (m, 2H), 1.53e1.50 (m, 1H), 1.36e
1.25 (m, 4H), 1.18e1.16 (m, 1H). HRMS (ESI): m/z calcd for
C
22H23ClN4O2 [M]þ 410.1510; found: m/z 410.1512.
4.1.5.11. 4-(4-Chlorophenyl)-5-oxo-N,3-di-p-tolyl-4,5-dihydro-1H-
1,2,4-triazole-1-carboxamide (5k). Yellow solid. Yield: 21.9%.
Melting Point: 215.7e217.1 ꢁC. HPLC (90% methanol in water):
ordinates for the residues in these regions of b2-AR were directly
assigned to the corresponding residues of CB receptors using DS
2.5.5/protein module [56]. The conformations of extracellular and

S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
83
intracellular loops were obtained through the loop search method,
and the beginning and ending regions were generated by endpair
module.
technology office in Zhejiang province (No. 2010R10048) for
funding our work.
Appendix A. Supplementary data
4.3.2. MD Simulations
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2013.12.018.
Each of the initial homology models was inserted into a pre-
equilibrated DPPC/TIP3P hydrate bilayer membrane system by
CHARMM GUI Membrane Builder (http://www.charmm-gui.org/)
[54,69]. The refinement and minimization of CB receptor homology
models was performed by molecular dynamics using Amber 11
package [53]. The force field parameters for DPPC molecule were
created using general Amber force field (GAFF) [70] and restrained
electrostatic potential (RESP) partial charges [71], and the AMBER
ff99SB force field [72] was selected for the proteins. Chloride ions
were added to the system until neutrality was achieved.
References
[1] P.H. Reggio, Curr. Med. Chem. 17 (2010) 1468e1486.
[2] R.G. Pertwee, Pharmacol. Ther. 74 (1997) 129e180.
[3] G.A. Thakur, R. Tichkule, S. Bajaj, A. Makriyannis, Expert Opin. Ther. Pat. 19
(2009) 1647e1673.
[4] M. Maccarrone, G. Bernardi, A.F. Agrò, D. Centonze, Brit. J. Pharmacol. 163
(2011) 1379e1390.
[5] R. Tanasescu, C.S. Constantinescu, Immunobiology 215 (2010) 588e597.
[6] P. Caraceni, M. Domenicali, F. Giannone, M. Bernardi, Best Pract. Res. Cl. En. 23
(2009) 65e77.
[7] P. Anand, G. Whiteside, C.J. Fowler, A.G. Hohmann, Brain Res. Rev. 60 (2009)
255e266.
[8] M. Odan, N. Ishizuka, Y. Hiramatsu, M. Inagaki, H. Hashizume, Y. Fujii,
S. Mitsumori, Y. Morioka, M. Soga, M. Deguchi, K. Yasui, A. Arimura, Bioorg.
Med. Chem. Lett. 22 (2012) 2803e2806.
[9] P. Yang, K.-Z. Myint, Q. Tong, R. Feng, H. Cao, A.A. Almehizia, M.H. Alqarni,
L. Wang, P. Bartlow, Y. Gao, J. Gertsch, J. Teramachi, N. Kurihara,
G.D. Roodman, T. Cheng, X.-Q. Xie, J. Med. Chem. 55 (2012) 9973e9987.
[10] Q. Ouyang, Q. Tong, R. Feng, K.-Z. Myint, P. Yang, X.-Q. Xie, ACS Med. Chem.
Lett. 4 (2013) 387e392.
[11] P. Yang, L. Wang, R. Feng, A.A. Almehizia, Q. Tong, K.-Z. Myint, Q. Ouyang,
M.H. Alqarni, L. Wang, X.-Q. Xie, J. Med. Chem. 56 (2013) 2045e2058.
[12] G. Kunos, J. Tam, Br. J. Pharmacol. 163 (2011) 1423e1431.
[13] M. Alvarado, J. Decara, M.J. Luque, L. Hernandez-Folgado, M. Gomez-Cañas,
M. Gomez-Ruiz, J. Fernandez-Ruiz, J. Elguero, N. Jagerovic, A. Serrano, P. Goya,
F.R. de Fonseca, Bioorg. Med. Chem. 21 (2013) 1708e1716.
[14] D. Jones, Nat. Rev. Drug. Discov. 7 (2008) 961e962.
Each system was minimized in two steps. First, the whole pro-
ꢀ
tein was kept fixed with a force constant of 300 kcal/(mol A), while
the lipids, water, and ions were allowed to minimize, and second,
only the atoms of backbone were restrained with a force of 50 kcal/
ꢀ
(mol A). In each step, the minimization was perform by the steepest
descent algorithm for 1000 steps and then conjugated gradient
method for another 5000 steps. Later, the systems were gradually
heated from 0 to 300 K over 100 ps in the NVT ensemble. Then,
equilibration of each system was performed in an NPT ensemble at
300 K under 1.0 atm pressure. During this procedure, the force
restraint weight to the atoms of backbone was decreased from 50 to
ꢀ
1 kcal/(mol A) gradually over total 3500 ps.
Finally, 30 ns production MD simulations were carried out for
the protein/DPPC/TIP3P systems in the NPT ensemble at a tem-
perature of 300 K and a pressure of 1 atm. Considering our ho-
mology models are apo proteins in the active state, in the
production process the atoms of backbone in the 7-TM regions
[15] Y.-K. Wu, C.-F. Yeh, T.W. Ly, M.-S. Hung, Curr. Top. Med. Chem. 11 (2011)
1421e1429.
[16] P. Pacher, S. Batkai, G. Kunos, Pharmacol. Rev. 58 (2006) 389e462.
[17] M. Rinaldi-Carmona, F. Barth, M. Heaulme, D. Shire, B. Calandra, C. Congy,
S. Martinez, J. Maruani, G. Neliat, FEBS Lett. 350 (1994) 240e244.
[18] M. Rinaldi-Carmona, F. Barth, J. Millan, J.-M. Derocq, P. Casellas, C. Congy,
D. Oustric, M. Sarran, M. Bouaboula, B. Calandra, M. Portier, D. Shire, J.-
C. Brelière, G. Le Fur, J. Pharmacol. Exp. Ther. 284 (1998) 644e650.
[19] M. Portier, M. Rinaldi-Carmona, F. Pecceu, T. Combes, C. Poinot-Chazel,
B. Calandra, F. Barth, G. Le Fur, P. Casellas, J. Pharmacol. Exp. Ther. 288 (1999)
582e589.
ꢀ
were still restrained with a force constant of 1 kcal/(mol A) to
maintain the active conformation. All the MD simulations were
performed with the integration time step of 2 fs, and the co-
ordinates and velocities were saved every 2 ps.
4.3.3. Flexible docking calculations
[20] H.J. Seo, M.J. Kim, S.H. Lee, S.-H. Lee, M.E. Jung, M.-S. Kim, K. Ahn, J. Kim, J. Lee,
Bioorg. Med. Chem. 18 (2010) 1149e1162.
The aforementioned MD refined CB receptor homology models
were employed for molecular docking studies. The active pockets of
the receptors were predicted based on the combination of Connolly
solvent-accessible protein channel analysis in the Sybyl-X1.3/
MOLCAD program [57] and the literature data [43,60]. The initial
structures of 5c and 5d were constructed by SKETCH module of
Sybyl-X1.3 [57], and minimized by Tripos force field with Gasteiger-
Hückel atomic charge. Then the ligands were put into the binding
cavities surrounded by the helices III, V, VI, and VII of the two re-
ceptors, respectively. All the single bonds of the ligand as well as
the side chains of residues inside the binding pocket were regarded
as rotatable bonds, and the H-bonding sites were marked for
suitable atoms of both receptor and ligand. Flexible docking of each
compound with either CB1 or CB2 receptor was performed
applying Sybyl-X1.3/FlexiDock [57] program to search the possible
binding conformations of ligands and receptors. The binding
interaction energy was calculated including the terms of van der
Waals, electrostatics, and torsion energy defined in the Tripos force
field, and top 20 scored receptor-ligand complexes were kept for
further analysis.
[21] P.K. Sasmal, R. Talwar, J. Swetha, D. Balasubrahmanyam, B. Venkatesham,
K.A. Rawoof, B. Neelima Devi, V.P. Jadhav, S.K. Khan, P. Mohan, D. Srinivasa
Reddy, V.K. Nyavanandi, S. Nanduri, K. Shiva Kumar, M. Kannan, P. Srinivas,
P. Nadipalli, H. Chaudhury, V.J. Sebastian, Bioorg. Med. Chem. Lett. 21 (2011)
4913e4918.
[22] F. Piscitelli, A. Ligresti, G. La Regina, V. Gatti, A. Brizzi, S. Pasquini, M. Allarà,
M.A. Maria Carai, E. Novellino, G. Colombo, V. Di Marzo, F. Corelli, R. Silvestri,
Eur. J. Med. Chem. 45 (2011) 5641e5653.
[23] S.R. Donohue, R.F. Dannals, C. Halldin, V.W. Pike, J. Med. Chem. 54 (2011)
2961e2970.
[24] A. Fulp, K. Bortoff, H. Seltzman, Y. Zhang, J. Mathews, R. Snyder, T. Fennell,
R. Maitra, J. Med. Chem. 55 (2012) 2820e2834.
[25] J. Bostrom, K. Berggren, T. Elebring, P.J. Greasley, M. Wilstermann, Bioorg.
Med. Chem. 15 (2007) 4077e4084.
[26] J.H.M. Lange, M.A.W. van der Neut, A.J.M. Borst, M. Yildirim, H.H. van Stui-
venberg, B.J. van Vliet, C.G. Kruse, Bioorg. Med. Chem. Lett. 20 (2010) 2770e
2775.
[27] R.A. Smith, Z. Fathi, S.-E. Brown, S. Choi, J. Fan, S. Jenkins, H.C.E. Kluender,
A. Konkar, R. Lavoie, R. Mays, J. Natoli, S.J. O’Connor, A.A. Ortiz, B. Podlogar,
C. Taing, S. Tomlinson, T. Tritto, Z. Zhang, Bioorg. Med. Chem. Lett. 17 (2007)
673e678.
[28] L. Botta, T. Semeraro, C. Mugnaini, A. Ligresti, E. Palazzo, S. Maione, V. Di
Marzo, F. Corelli, J. Med. Chem. 51 (2008) 5075e5084.
[29] H. Omura, M. Kawai, A. Shima, Y. Iwata, F. Ito, T. Masuda, A. Ohta, N. Makita,
K. Omoto, H. Sugimoto, A. Kikuchi, H. Iwata, K. Ando, Bioorg. Med. Chem. Lett.
18 (2008) 3310e3314.
[30] V. Gembus, C. Furman, R. Millet, R. Mansouri, P. Chavatte, V. Levacher, J.-
F. Brière, Eur. J. Med. Chem. 58 (2012) 396e404.
[31] A.J. Morrison, J.M. Adam, J.A. Baker, R.A. Campbell, J.K. Clark, J.E. Cottney,
M. Deehan, A.-M. Easson, R. Fields, S. Francis, F. Jeremiah, N. Keddie, T. Kiyoi,
D.R. McArthur, K. Meyer, P.D. Ratcliffe, J. Schulz, G. Wishart, K. Yoshiizumi,
Bioorg. Med. Chem. Lett. 21 (2011) 506e509.
Acknowledgments
We wish to thank the National Natural Science Foundation of
China (No. NSFC 30973637) and Qianjiang talents project of

84
S. Han et al. / European Journal of Medicinal Chemistry 74 (2014) 73e84
[32] J.-Z. Chen, X.-W. Han, Q. Liu, A. Makriyannis, J. Wang, X.-Q. Xie, J. Med. Chem.
49 (2006) 625e636.
[52] D. Shire, B.C.M. Delpech, X. Dumont, M. Kaghad, G. Le Fur, D. Caput, P. Ferrara,
J. Biol. Chem. 271 (1996) 6941e6946.
[33] V. Cherezov, D.M. Rosenbaum, M.A. Hanson, S.G.F. Rasmussen, F.S. Thian,
T.S. Kobilka, H.-J. Choi, P. Kuhn, W.I. Weis, B.K. Kobilka, R.C. Stevens, Science
318 (2007) 1258e1265.
[53] D.A. Case, T.E. Cheatham, T. Darden, H. Gohlke, R. Luo, K.M. Merz,
C. Simmerling, B. Wang, AMBER 11, University of California, San Francisco,
2010.
[34] M. Jansen, H. Rabe, A. Strehle, S. Dieler, F. Debus, G. Dannhardt, M.H. Akabas,
H. Lüddens, J. Med. Chem. 51 (2008) 4430e4448.
[35] J.M. Kane, B.M. Baron, M.W. Dudley, S.M. Sorensen, M.A. Staeger, F.P. Miller,
J. Med. Chem. 33 (1990) 2772e2777.
[54] S. Jo, T. Kim, W. Im, PLoS ONE 2 (2007) e880.
[55] R. Liithy, J.U. Bowie, D. Eisenberg, Nature 356 (1992) 83e85.
[56] 2.5.5, Accelrys Discovery Studio, Accelrys Software Inc., San Diego, CA, 2010.
[57] ver. x1.3, Sybyl Molecular Modeling Software Packages, Tripos Associates, Inc.,
St. Louis, MO63144, 2011.
[36] J.M. Kane, Synthesis 10 (1987) 912e914.
[37] K.S. Schroeder, B.D. Neagle, J. Biomol. Screen. 1 (1996) 75e80.
[38] L.C. Mattheakis, L.D. Ohler, Drug. Discov. Today 5 (Suppl. 1) (2000) 15e19.
[39] G.R. Monteith, G.S.J. Bird, Trends Pharmacol. Sci. 26 (2005) 218e223.
[40] T. Zhu, L.-Y. Fang, X. Xie, Acta Pharmacol. Sin. 29 (2008) 507e516.
[41] H. Iwamura, H. Suzuki, Y. Ueda, T. Kaya, T. Inaba, J. Pharm. Pharm. Sci. 296
(2001) 420e425.
[42] X.-Q. Xie, J.-Z. Chen, E.M. Billings, Proteins 53 (2003) 307e319.
[43] O.M.H. Salo, M. Lahtela-Kakkonen, J. Gynther, T. Jaervinen, A. Poso, J. Med.
Chem. 47 (2004) 3048e3057.
[44] T. Tuccinardi, P.L. Ferrarini, C. Manera, G. Ortore, G. Saccomanni, A. Martinelli,
J. Med. Chem. 49 (2006) 984e994.
[45] C. Montero, N.E. Campillo, P. Goya, J.A. Páez, Eur. J. Med. Chem. 40 (2005)
75e83.
[58] S.D. McAllister, G. Rizvi, S. Anavi-Goffer, D.P. Hurst, J. Barnett-Norris,
D.L. Lynch, P.H. Reggio, M.E. Abood, J. Med. Chem. 46 (2003) 5139e5152.
[59] S.D. McAllister, Q. Tao, J. Barnett-Norris, K. Buehner, D.P. Hurst, F. Guarnieri,
P.H. Reggio, K.W. Nowell Harmon, G.A. Cabral, M.E. Abood, Biochem. Phar-
macol. 63 (2002) 2121e2136.
[60] O.M.H. Salo, K.H. Raitio, J.R. Savinainen, T. Nevalainen, M. Lahtela-Kakkonen,
J.T. Laitinen, T. Jaervinen, A. Poso, J. Med. Chem. 48 (2005) 7166e7171.
[61] J.-J. Chen, S. Han, Y. Cao, J.-Z. Chen, Acta Pharm. Sin. 48 (2013) 1436e1449.
[62] A. Cordomí, J.C. Gómez-Tamayo, V. Gigoux, D. Fourmy, Trends Pharmacol. Sci.
34 (2013) 320e331.
[63] G.A. Olah, S.C. Narang, A. Garcia-Luna, Synthesis 1981 (1981) 790e791.
ꢁ
ꢁ
[64] N. Maras, S. Polanc, M. Kocevar, Tetrahedron 64 (2008) 11618e11624.
[65] W.J. Bailey, R. Barclay, J. Am. Chem. Soc. 81 (1959) 5393e5396.
[66] J.P. Horwitz, V.A. Grakauskas, J. Org. Chem. 19 (1954) 194e201.
[67] F. Macaev, G. Rusu, S. Pogrebnoi, A. Gudima, E. Stingaci, L. Vlad, N. Shvets,
F. Kandemirli, A. Dimoglo, R. Reynolds, Bioorg. Med. Chem. 13 (2005) 4842e
4850.
[46] P. Diaz, S.S. Phatak, J. Xu, F. Astruc-Diaz, C.N. Cavasotto, M. Naguib, J. Med.
Chem. 52 (2009) 433e444.
[47] S. Durdagi, M. Papadopoulos, P. Zoumpoulakis, C. Koukoulitsa,
T. Mavromoustakos, Mol. Divers. 14 (2010) 257e276.
[48] N.A. Osman, A.H. Mahmoud, M. Allar, R. Niess, K.A. Abouzid, V. Di Marzo,
A.H. Abadi, Bioorg. Med. Chem. 18 (2010) 8463e8477.
[49] E. Cichero, A. Ligresti, M. Allar, V. di Marzo, Z. Lazzati, P. D’Ursi, A. Marabotti,
L. Milanesi, A. Spallarossa, A. Ranise, P. Fossa, Eur. J. Med. Chem. 46 (2011)
4489e4505.
[50] G. Kuang, G. Hu, X. Sun, W. Li, G. Liu, Y. Tang, J. Mol. Model. 18 (2012) 3831e
3845.
[51] K.-i. Kusakabe, Y. Tada, Y. Iso, M. Sakagami, Y. Morioka, N. Chomei,
S. Shinonome, K. Kawamoto, H. Takenaka, K. Yasui, H. Hamana, K. Hanasaki,
Bioorg. Med. Chem. 21 (2013) 2045e2055.
[68] A. Omodei-Sale, P. Consonni, G. Galliani, J. Med. Chem. 26 (1983) 1187e1192.
[69] T.B. Woolf, B. Roux, Proteins 24 (1996) 92e114.
[70] J. Wang, R.M. Wolf, J.W. Caldwell, P.A. Kollman, D.A. Case, J. Comput. Chem. 25
(2004) 1157e1174.
[71] C.I. Bayly, P. Cieplak, W. Cornell, P.A. Kollman, J. Phys. Chem. 97 (1993)
10269e10280.
[72] V. Hornak, R. Abel, A. Okur, B. Strockbine, A. Roitberg, C. Simmerling, Proteins
65 (2006) 712e725.