Synthesis and structure-activity relationship studies in serotonin 5-HT1A receptor agonists based on fused pyrrolidone scaffolds

Article abstract of DOI:10.1016/j.ejmech.2013.01.044

A new class of serotonin 5-HT_1A receptor ligands related to NAN-190, buspirone and aripiprazole has been designed using our potent 5-HT ₃ receptor ligands as templates. The designed pyrrolidone derivatives 10a-n were prepared by means of the straightforward chemistry consisting in the reaction of the appropriate γ-haloester derivatives with the suitable arylpiperazinylalkylamines. The nanomolar 5-HT_1A receptor affinity and the agonist-like profile shown by fused pyrrolidone derivatives 10k,m stimulated the rationalization of the interaction with an homology model of the 5-HT_1A receptor and the evaluation of their selectivity profiles and the pharmacokinetic properties. Interestingly, the results of the profiling assays suggested for close congeners 10k,m a significantly divergent binding pattern with compound 10m showing an appreciable selectivity for 5-HT _1AR.

Full text of DOI:10.1016/j.ejmech.2013.01.044

European Journal of Medicinal Chemistry 63 (2013) 85e94
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis and structureeactivity relationship studies in serotonin
5-HT1A receptor agonists based on fused pyrrolidone scaffolds
,
a
a
a
Andrea Cappelli a , Monica Manini , Salvatore Valenti , Federica Castriconi ,
Germano Giuliani a, Maurizio Anzini a, Simone Brogi a, Stefania Butini a,
Sandra Gemma a, Giuseppe Campiani a, Gianluca Giorgi b, Laura Mennuni c,
Marco Lanza c, Antonio Giordani c, Gianfranco Caselli c,
*
Ornella Letari c, Francesco Makovec c
a Dipartimento Farmaco Chimico Tecnologico and European Research Centre for Drug Discovery and Development, Università degli Studi di Siena,
Via A. Moro, 53100 Siena, Italy
b Dipartimento di Chimica, Università degli Studi di Siena, Via A. Moro, 53100 Siena, Italy
c Rottapharm Madaus, Via Valosa di Sopra 9, 20052 Monza, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A new class of serotonin 5-HT1A receptor ligands related to NAN-190, buspirone and aripiprazole has
been designed using our potent 5-HT3 receptor ligands as templates. The designed pyrrolidone
derivatives 10aen were prepared by means of the straightforward chemistry consisting in the reaction of
Received 12 November 2012
Received in revised form
14 January 2013
Accepted 17 January 2013
Available online 8 February 2013
the appropriate g-haloester derivatives with the suitable arylpiperazinylalkylamines. The nanomolar 5-
HT1A receptor affinity and the agonist-like profile shown by fused pyrrolidone derivatives 10k,m
stimulated the rationalization of the interaction with an homology model of the 5-HT1A receptor and the
evaluation of their selectivity profiles and the pharmacokinetic properties. Interestingly, the results of
the profiling assays suggested for close congeners 10k,m a significantly divergent binding pattern with
compound 10m showing an appreciable selectivity for 5-HT1AR.
Keywords:
Serotonin
5-HT1A receptors
Arylpiperazine
Binding
Ó 2013 Elsevier Masson SAS. All rights reserved.
1. Introduction
8-OH-DPAT, 1 and NAN-190, 2 Fig. 1). Although 2 was earlier con-
sidered a selective 5-HT1AR antagonist, evidence was later collected
Serotonin (5-hydroxytryptamine, 5-HT) is an important neuro-
transmitter both in central nervous system and in periphery. 5-HT
has to date been shown to bind to distinct receptors subdivided into
14 separate subtypes (5-HT1-7: 5-HT1A-F, 5-HT2A-C, 5-HT3, 5-HT4, 5-
HT5A, 5-HT6, 5-HT7) [1]. The 5-HT1A receptor (5-HT1AR) subtype
pharmacology has been extensively studied by means of a huge
series of developed selective ligands as pharmacological tools (e.g.
about its alpha2-adrenergic receptor antagonist properties [2]. 5-
HT1AR agonists and partial agonists are clinically used in the
treatment of anxiety and depression [3]. Buspirone (3, Fig. 1) is an
example of a psychoactive drug used in the treatment of general-
ized anxiety disorder of very mild to moderate intensity without
panic attacks. Compound 3 was reported to behave as 5-HT1A
R
partial agonist endowed with moderate 5-HT1AR selectivity (due to
its affinity toward other monoaminergic receptors such as dop-
amine D2 and adrenergic alpha1 and alpha2 receptor subtypes) [4].
Accordingly, the selectivity issue represents a crucial feature in the
development of buspirone-related analogs. Clinically, the inter-
action with 5-HT1AR is also responsible for the lack of undesired
side-effects in some atypical antipsychotic drugs [5]. The seroto-
ninergic system plays a pivotal role in regulation of prefrontal
cortex (PFC) functions, including emotional control, cognitive
behavior and working memory. PFC pyramidal neurons and GABA
interneurons contain a high density of 5-HT1AR and 5-HT2AR. It was
demonstrated in PFC that NMDA receptors channels are the target
of 5-HT1AR and both receptors modulate the excitability of cortical
Abbreviations: 5-HT, 5-hydroxytryptamine; GPCRs, G-protein coupled recep-
tors; 5-HT1AR, 5-HT1A receptor; PFC, prefrontal cortex; GABA, gamma aminobutyric
acid; NMDA, N-methyl-D-aspartate; 5-HT2AR, 5-HT2A receptor; EPS, extrapyramidal
symptoms; 5-HT3R, 5-HT3 receptor; FPC, fused pyrrolidone component; APC,
arylpiperazine component; AC, aryl component; CNS, central nervous system; CA,
Central Activity; CR, central reactivity; NR, neurovegetative reflexes; NT, neuro-
motor tonus; AS, autonomic system; DMEM, Dulbecco’s modified eagle medium;
IBMX, 3-isobutyl-1-methylxanthine; OPLS-AA, optimized potentials for liquid
simulations-all atom; PRCG, Polak-Ribierè conjugate gradient; GB/SA, generalized-
born/surface-area.
* Corresponding author. Tel.: þ39 0577 234320; fax: þ39 0577 234333.
E-mail addresses: andrea.cappelli@unisi.it, cappelli@unisi.it (A. Cappelli).
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.01.044

86
A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
plays a significant role in the modulation of the short range con-
tacts. On the other hand, in the series of 3,4-dihydropyrazino[1,2-a]
indol-1(2H)-one-2-yl series of analogs (exemplified by 9a,b) the 5-
HT1AR affinity appeared to be modulated by the functionalization at
the arylpiperazine moiety [15c,d].
Taken together, these analyses, prompted us to further explore
5-HT1AR pharmacology. Thus, in the progress of our large program
focused on the medicinal chemistry of 5-HT receptor ligands,
pyrrolidone derivatives 5e8, NAN-190 (2), and compounds 9a,b
were considered as structural templates in the design of new 5-
HT1AR ligands 10 (Fig. 2), in which the basic heterocyclic moiety
employed was the arylpiperazine scaffold widely embodied in
most of the known serotoninergic ligands (e.g. 2 and 4). As the
arylpiperazine scaffold is known to interact with a broad range of
different receptors, our working hypothesis was the possible
modulation of the selectivity by means of the appropriate varia-
tion and combination of three structural components, namely: the
fused pyrrolidone component (FPC), the spacer between piper-
azine and pyrrolidone, and the substituents of the arylpiperazine
component (APC).
Fig. 1. Structures of some relevant serotonin 5-HT1AR ligands.
neurons thus affecting cognitive functions [6]. Indeed, a variety of
preclinical data has suggested that the 5-HT1AR may be a ther-
apeutic target for the development of improved antipsychotics.
Although the role of 5-HT1AR in antipsychotic drug efficacy profile
is still under debate, 5-HT1AR affinity contributes to the clinical
efficacy of most of the atypical antipsychotics (e.g. clozapine,
olanzapine, aripiprazole, 4 Fig. 1) and contributes to their low EPS
liability [7]. Experimental evidences show that 5-HT1AR activation
attenuates antipsychotic-induced side effects in humans [8].
Accordingly, an association was postulated between agonist activity
at 5-HT1AR and anxiolytic or antidepressant effects, improvements
in cognitive and negative symptoms [9], and decreased develop-
ment of EPS in schizophrenia [10]. Furthermore, in vivo studies
demonstrated that activation of 5-HT1AR can play a role in
aripiprazole-mediated behavior in rats [11,12]. Moreover, since
glutamatergic transmission is dysfunctional in schizophrenia and
glutamate release is decreased by 5-HT1AR activation [13], agonist
properties at postsynaptic 5-HT1AR may be relevant to the ther-
apeutic profile of atypical antipsychotic agents, improving negative
symptoms and cognitive deficits [14].
The present paper describes the synthesis and the preliminary
pharmacological characterization of 5-HT1AR ligands 10 as well as
the pharmacokinetic properties and the rationalization of the
ligandereceptor interaction of the most interesting compounds of
the series.
2. Results and discussion
2.1. Chemistry
Target pyrrolidone derivatives 10aen were prepared by means
of the simple general procedure (Scheme 1) previously developed
for the synthesis of our tropane 5-HT3R antagonists 5e8 [16]. The
reaction of appropriate g-haloester derivative (11e13) [16] with the
suitable arylpiperazinylalkylamine (14e18) [17] afforded the
expected tricyclic ligand (10aen) in acceptable to good yields.
The structure of compound 10c was confirmed by crystallo-
graphic studies (Fig. 3) and used as an input in the molecular
modeling studies.
Our earlier interest in the search of new serotoninergic agents
led to development of a series of chemically diverse compounds
endowed with different affinity profiles, selectivity and pharma-
cological properties [15]. Among the multitude of developed 5-HT
receptors ligands, derivatives 5e8 (Fig. 2) based on the pyrroli-
done structure [16], proved to behave as potent 5-HT3R ligands.
These analogs showed different functional profiles ranging from
the antagonist properties of tropane derivatives (compounds 5e8
with Het ¼ endo-tropan-3-yl) to the full range of intrinsic effi-
cacies shown by the quinuclidine derivatives (compounds 5e8
with Het ¼ quinuclidin-3-yl). On the whole, the results obtained
suggested that the basic heterocyclic (endo-tropan-3-yl or quinu-
clidin-3-yl) moiety of compounds 5-8 governs the long range
interaction with the 5-HT3R and the fused pyrrolidone scaffold
2.2. Binding assays, structureeactivity relationships (SARs), and
molecular modeling studies
The newly developed arylpiperazines 10aen were evaluated for
their potential activity in inhibiting the specific binding of [3H]8-
OH-DPAT [18] (1) to 5-HT1AR in rat hippocampus membranes in
comparison with unlabeled 1, and the results of the binding studies
are summarized in Table 1.
Fig. 2. Design of 5-HT1AR ligands 10aen. For the specific structure of compounds 10aen see Table 1.

A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
87
Table 1
5-HT1AR binding affinities of compounds 10aen.
Compd
FPC
Spacer
AC
Ki (nM) ꢀ SEMa
10a
10b
10c
10d
10e
10f
10g
10h
10i
10j
10k
10l
10m
10n
1
A
B
A
A
B
B
C
C
A
A
B
B
C
C
e
e(CH2)3e
e(CH2)3e
e(CH2)4e
e(CH2)4e
e(CH2)4e
e(CH2)4e
e(CH2)4e
e(CH2)4e
e(CH2)5e
e(CH2)5e
e(CH2)5e
e(CH2)5e
e(CH2)5e
e(CH2)5e
e
A
A
A
B
A
B
A
B
A
B
A
B
A
B
e
104 ꢀ 50
46 ꢀ 21
15 ꢀ 6
Scheme 1. Synthesis of target compounds 10aen. For the specific structure of com-
pounds 10aen see Table 1. Reagents: (i) C2H5OH.
30 ꢀ 9
13 ꢀ 1
In order to better understand the ligand-5-HT1AR interactions
for the compounds herein reported, and aiming at investigating the
binding mode of these new inhibitors, in the absence of a 3D X-ray
structure of the 5-HT1A GPCR, a docking procedure (performed with
AutoDock [19]) was applied by using a specific homology model of
the 5-HT1AR generated by means of a multiple templates-based
homology modeling approach (see experimental section for
details) [20]. We chose to make use of multiple templates in the
modeling the 5-HT1AR for increasing the quality of the obtained
model [21,22]. Indeed, this method successfully allowed us to
increase the simulation time for minimizing and optimizing the 5-
HT1AR model. The 5-HT1AR affinities shown by compounds 10aen
were found to vary from the submicromolar to the nanomolar
range with a modulation appropriate for structure affinity rela-
tionship analysis. The most outstanding results were obtained with
compounds 10k,m, which are very potent ligands showing Ki val-
ues in the low nanomolar range. In these two arylpiperazine
ligands, the combination of the angularly fused pyrrolidone com-
ponents with the pentamethylene spacer and the 2-
methoxyphenyl arylpiperazine component represents the optimal
scaffold for the interaction with the 5-HT1AR binding site, which
appears to be practically unable to distinguish between compound
10k (Ki ¼ 1.6 nM) and its closely related analog 10m (Ki ¼ 1.7 nM).
Interestingly, our computational studies revealed that 10k and 10m
share virtually identical interactions within 5-HT1AR binding site
(Fig. 4).
31 ꢀ 11
17 ꢀ 5
31 ꢀ 10
20 ꢀ 6
58 ꢀ 19
1.6 ꢀ 0.2
490 ꢀ 125
1.7 ꢀ 0.3
878 ꢀ 228
1.2 ꢀ 0.2
a
Each value is the mean ꢀ SEM of 3 independent determinations performed in
duplicate and represents the concentration giving half the maximum inhibition of
[3H]1 (final concentration 0.5 nM) specific binding to rat cerebral hippocampus
membranes.
determined lower activity of compounds 10i and 10j (Table 1),
since, when compared with 10k and 10m, they only revealed the H-
bond with the Asn217 (Fig. 5).
The alteration of the optimized combination of the above-
mentioned structural components (namely: the angularly fused
pyrrolidone components, the pentamethylene spacer, and the 2-
methoxyphenyl arylpiperazine component) produces a significant
decrease in the 5-HT1A
R binding affinity. For instance, the
replacement of the 2-methoxyphenyl arylpiperazine component of
the best ligands 10k,m, with the 2,3-dichlorophenyl counterpart
(as in compounds 10l,n) leads to a dramatic decrease in the 5-
HT1AR binding affinity (10l vs 10k and 10n vs 10m). Docking
studies performed with our arylpiperazine derivatives 10i,k,m,
highlight that the methoxy group of these compounds is accom-
modated into a small binding pocket, which appears to be less
tolerant for the two chlorine atoms of 10j (Fig. 6).
On the other hand, the same 2-methoxyphenyl to 2,3-
dichlorophenyl replacement appears to be well tolerated in the
remaining couples of arylpiperazine ligands reported in Table 1. The
very similar affinity shown by 10bej suggests that apart from the
above cited optimized combination of the structural components, a
broad range of different combinations is recognized by the receptor
with very similar binding energies producing Ki values about one
order of magnitude higher than those shown by the most potent
Particularly, both compounds can establish a T-shaped inter-
action with the hydrophobic Phe166, and H-bonds with both
Asn217 and Arg244. An additional charge-assisted H-bond can take
place between the protonated piperazine nitrogen and the car-
boxylic moiety of Asp78. This pattern of interaction is in line with
the nanomolar potency of these 5-HT1AR ligands. The proposed
binding modes are also consistent with the experimentally
compounds 10k,m. From another point of view, the 5-HT1A
R
binding site appears to be barely discriminating for compounds
bearing the tetramethylene spacer, while it becomes more sensitive
when the spacer is homologated to the pentamethylene one.
However, when the pentamethylene spacer is combined with the
linearly-fused pyrrolidone moiety as in compounds 10i,j, the 2-
methoxyphenyl to 2,3-dichlorophenyl replacement appears to be
well tolerated as already observed in the tetramethylene
Fig. 3. Structure of compound 10c found by crystallography. Ellipsoids enclose 50%
probability.

88
A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
Fig. 4. Interactions of 10k (panel A) and 10m (panel B) with the 5-HT1AR homology model: Only the relevant aminoacidic residues of the binding site are displayed. Compounds 10k
and 10m are represented as sticks. The H-bonds are represented as dashed red lines. For sake of clarity only polar hydrogens of the interacting residues and for protonation of
piperazine N are displayed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
derivatives 10ceh (compare 10i vs 10j). The superimposition of the
computational binding complexes of 10k and 10j with the receptor
depicted in Fig. 6 clearly evidences the perfect fitting of the pyr-
rolidone component and of the 2-methoxyphenyl arylpiperazine
component of 10k within the two binding cavities of the receptor.
It is noteworthy that the structureeaffinity relationship analysis
in the pentamethylene derivatives reveals a discrepancy between
the 2-methoxyphenyl and the 2,3-dichloromethyl subseries. In fact,
the angularly fused derivatives 10k,m belonging to the 2-
methoxyphenyl subseries are about one order of magnitude more
potent than the corresponding linearly fused 10i, while angularly
fused derivatives 10l,n belonging to the 2,3-dichlorophenyl subs-
eries are about one order of magnitude less potent than the cor-
responding linearly fused 10j.
The most potent compounds binding 5-HT1AR in rat hippo-
campus membranes (i. e. 10k,m) were assayed at Cerep (Poitiers,
France) for their potential interaction with 79 off-target receptors
(see Supplementary material) in order to evaluate the level of the
selectivity reached by these novel arylpiperazine derivatives. The
selected ligands were tested firstly at high concentration (e.g.
10,000 nM) and the receptors interacting at significant extents with
the ligands were considered for further studies at lower concen-
trations (e.g. 100 and 10 nM) and affinity estimation for those
compounds showing significant inhibitions in the low nanomolar
range (Table 2). The profiling assays showed a significantly diver-
gent binding pathway for 10k,m. Very interestingly, 5-HT1AR ligand
10m revealed an appreciable selectivity for the receptor of interest
since it was found to interact with only two off-target receptors
with affinity values in the low nanomolar range (the estimated
affinity values shown for D3, 5-HT2B resulted about one order of
magnitude lower than that shown for 5-HT1AR). On the other hand,
compound 10k was found to interact also with dopamine D2S
Finally, in compounds 10a,b based on a trimethylene spacer the
replacement of the angularly-fused pyrrolidone component of 10b
with the linearly-fused counterpart of 10a leads to a negligible
(two-fold) increase in the 5-HT1AR binding affinity.
Fig. 5. Interactions of compounds 10i (panel A) and 10j (panel B) with the 5-HT1AR homology model: Only the relevant aminoacidic residues of the binding site are displayed.
Compounds 10i and 10j are represented as sticks. The H-bonds are represented as dashed red lines. For sake of clarity only polar hydrogens of the interacting residues and for
protonation of piperazine N are displayed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
89
results are summarized in Table 3. Both compounds are charac-
terized by a very high volume of distribution, being approx 20 and
80 times the theoretical total body water in mice, respectively, and
by a high clearance, being their Cl values approx 2 and 4-fold the
normal hepatic blood flow in mice [24]. Nevertheless, their abso-
lute oral bioavailability was found to be moderate for 10k (20%) and
relatively high for 10m (45%), suggesting
a relatively good
absorption from the gastrointestinal tract.
2.5. In vivo safety assays
Compounds 10k,m were evaluated in the Irwin test [25] in order
to obtain preliminary data about the safety of compounds. In
summary, male CD1 mice (n ¼ 3/group) were dosed with 10k,m up
to 80 mg/kg, and kept under observation for the suitable time after
drug administration. Analysis of the results showed that neither
10k nor 10m affected the behavioral and physiological state of mice
as assessed by evaluating 43 different parameters categorized into
5 fundamental activities: Central Activity (CA), Central Reactivity
(CR), Neurovegetative Reflexes (NR), Neuromotor Tonus (NT), and
Autonomic System (AS).
Fig. 6. Superimposition of the computational complexes of ligands 10k and 10j with
the 5-HT1AR homology model binding site. Ligands are represented as licorice colored
green and brown for 10k and 10j respectively. Protein is represented as transparent
brown surface. (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
3. Conclusion
receptors (e.g. at 100 and 10 nM, 10k produced inhibition values
very similar to those observed with 5-HT1AR) and, to a lesser extent,
with D3, D4, 5-HT2B, and 5-HT7. Based on these in vitro binding data,
compound 10k can rather be considered a “dirty drug” being
endowed with an in vitro profile highly resembling that of aripi-
prazole [10]. It should be stressed that the divergent binding profile
showed by close congeners 10k,m is accompanied by about the
same affinity for 5-HT1AR. This observation suggests that in GPCR
family the selectivity of ligandereceptor interaction is affected by
very minor structural alterations.
A new series of 5-HT1AR ligands was designed using our pyr-
rolidone 5-HT3R ligands 5e8, compounds 9a,b, NAN-190, buspir-
one, and aripiprazole as templates. Target derivatives 10aen were
prepared by means of the straightforward chemistry (previously
developed for the synthesis of our tropane 5-HT3R antagonists 5e
8) based on the reaction of the appropriate
g-haloester deriva-
tives with the proper arylpiperazinylalkylamines. Among the
newly-synthesized compounds, 10k,m revealed 5-HT1AR affinity
values in the low nanomolar range that were easily rationalized in
terms of interaction with an homology model of the 5-HT1AR.
Moreover, profiling assays revealed for the close congeners 10k,m a
significantly divergent binding pattern. In particular, compound
10m showed an appreciable selectivity for 5-HT1AR, while 10k
appeared to be a “dirty drug” showing an in vitro profile, which
resembles that of aripiprazole. In cellular functional studies, fused
pyrrolidone derivatives 10k,m showed clear-cut agonist-like pro-
files. The combination of these results with both the safety profile
emerged from the Irwin test results and the reasonable pharma-
cokinetic profiles suggested that the biological features of arylpi-
perazine derivatives could be tailored by means of an appropriate
chemical design in order to obtain dirty drugs or cleaner agents
following the specific needs.
2.3. In vitro 5-HT1AR cellular functional assay
Based on their interesting 5-HT1AR binding affinities, com-
pounds 10k,m were selected for a suitable functional character-
ization in a cellular model. Thus, the agonist-like properties of
10k,m at human 5-HT1A receptors expressed in transfected cell
lines were determined in vitro by measuring a single signal trans-
duction representing inhibition of adenylyl cyclase activity (cAMP)
[23]. Both 10k,m showed agonist-like properties by inhibiting
forskolin-stimulated cAMP accumulation: 100 nM 10k,m inhibited
by 63% and 60% the adenylyl cyclase signals, respectively. Con-
versely, 10k,m were unable to reverse the cAMP response evoked
by 8-OH-DPAT showing no antagonist-like effect in this cellular
system.
4. Experimental
2.4. In vivo pharmacokinetic study
4.1. Chemistry
Because of their potent interaction with 5-HT1AR and the
promising agonist-like properties, the pharmacokinetic profile of
10k,m was evaluated in mice by using standard protocols and the
4.1.1. General procedures
All chemicals used were of reagent grade. Yields refer to purified
products and are not optimized. Melting points were determined in
open capillaries on a Gallenkamp apparatus and are uncorrected.
Merck silica gel 60 (230e400 mesh) was used for column chro-
matography. Merck TLC plates, silica gel 60 F254 were used for TLC.
1H NMR spectra were recorded by means of a Bruker AC 200, a
Varian Mercury-300, or a Bruker DRX 400 AVANCE spectrometers
in the indicated solvents (TMS as internal standard); the values of
the chemical shifts are expressed in ppm and the coupling con-
stants (J) in Hz. Mass spectra were recorded on either a Thermo-
Finnigan LCQ-Deca or an Agilent 1100 LC/MSD. The purity of
compounds 10aen was assessed by RP-HPLC and was found to be
higher than 95%. An Agilent 1100 Series system equipped with a
Table 2
Interaction of compounds 10k,m with off-target receptors.
10k
10k
10k IC50 10m
10m
10m
Inhibitiona Inhibitiona (nM)
at 100 nM at 10 nM
Inhibitiona Inhibitiona IC50 (nM)
at 100 nM at 10 nM
5-HT2B 70
5-HT7 65
28
18
74
39
13
43
62
3.5
24
65
65
42
26
74
28
18
24
NT
38
10
63
>100
>100
31
D2S
D3
90
86
60
D4
>100
a
The inhibition studies were performed at Cerep (Poitiers, France)

90
A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
Table 3
Pharmacokinetic parameters of compounds 10k,m in mice.
Compd
Route
Dose (mg/kg)
Cl (mL/h/kg)
Vz (mL/kg)
t1/2 (h)
AUC 0-t (h*ng/mL)
AUC inf (h*ng/mL)
F%
Cmax (ng/mL)
Tmax (h)
10k
10k
10m
10m
IV
PO
IV
2
10
2
9540
e
17552
e
1.3
4.7
1.5
4.5
207
212
70
210
325
76
e
e
e
20
e
56
e
0.3
e
26400
e
58708
e
PO
10
157
239
45
35
1.0
Lichrocart 125-4 (4 ꢁ 125 mm, Purospher Star RP-18C, 5
m
m) col-
J ¼ 7.3, 1H), 7.94 (d, J ¼ 8.0, 1H). 8.09 (d, J ¼ 8.5, 1H), 8.55 (s, 1H). MS
(ESI): m/z 431 (M þ Hþ).
umn was used in the HPLC analysis with acetonitrile-water-
methanol (80:10:10) as the mobile phase at a flow rate of 0.5 mL/
min. UV detection was achieved at 254 nm.
4.1.6. 2-[4-[4-(2,3-Dichlorophenyl)piperazin-1-yl]butyl]-2,3-
dihydro-1H-pyrrolo[3,4-b]quinolin-1-one (10d)
4.1.2. General procedure for the synthesis of compounds 10aen
This compound was synthesized from g-chloroester 11 [16]
A mixture of the appropriate
g
-haloester derivative (11e13, 1.0
(0.10 g, 0.40 mmol) and 4-[4-(2,3-dichlorophenyl)piperazin-1-yl]
butan-1-amine [17b] (16, 0.24 g, 0.80 mmol, reaction time ¼ 6 h)
and purified by flash-chromatography with ethyl acetate-
triethylamine (8:2) as the eluent to obtain pure 10d as a white
solid (0.13 g, yield 70%). An analytical sample was obtained by
recrystallization from n-hexane-dichloromethane (mp 154e
155 ꢂC). 1H NMR (400 MHz, CDCl3): 1.58e1.85 (m, 4H), 2.48 (t,
J ¼ 7.3, 2H), 2.62 (m, 4H), 3.04 (m, 4H), 3.76 (t, J ¼ 7.1, 2H), 4.59 (s,
2H), 6.93 (m, 1H), 7.12 (m, 2H), 7.62 (t, J ¼ 7.4, 1H), 7.83 (t, J ¼ 7.0,
1H), 8.07 (d, J ¼ 8.2, 1H), 8.14 (d, J ¼ 8.4, 1H), 8.60 (s, 1H). MS (ESI):
m/z 469 (M þ Hþ).
equivalent) in ethanol with the suitable arylpiperazinylalkylamine
(14e18, 2.0 equivalents) was heated under reflux for the appro-
priate time (the reaction was monitored by TLC). The reaction
mixture was concentrated under reduced pressure and the residue
was partitioned between dichloromethane and water. The organic
layer was dried over sodium sulfate and concentrated under
reduced pressure. Purification of the residue by flash-
chromatography with the appropriate eluent gave the corre-
sponding target arylpiperazine derivative (10aen).
4.1.3. 2-[3-[4-(2-Methoxyphenyl)piperazin-1-yl]propyl]-2,3-
dihydro-1H-pyrrolo[3,4-b]quinolin-1-one (10a)
4.1.7. 2-[4-[4-(2-Methoxyphenyl)piperazin-1-yl]butyl]-2,3-
This compound was prepared from
g-chloroester 11 [16] (0.10 g,
dihydro-1H-benzo[e]isoindol-1-one (10e)
0.40 mmol) and 3-[4-(2-methoxyphenyl)piperazin-1-yl]propan-1-
amine [17a] (14, 0.20 g, 0.80 mmol, reaction time ¼ 9 h) and
purified by flash-chromatography with ethyl acetate-triethylamine
(9:1) as the eluent to obtain pure 10a as a white solid (0.13 g, yield
78%). An analytical sample was obtained by recrystallization from
n-hexane-dichloromethane (mp 118e119 ꢂC). 1H NMR (400 MHz,
CDCl3): 1.98 (m, 2H), 2.52 (t, J ¼ 7.3, 2H), 2.65 (m, 4H), 3.05 (m, 4H),
3.78 (t, J ¼ 7.1, 2H), 3.83 (s, 3H), 4.61 (s, 2H), 6.82e6.98 (m, 4H), 7.61
(m,1H), 7.81 (m, 1H), 7.99 (d, J ¼ 8.2,1H), 8.13 (d, J ¼ 8.4,1H), 8.60 (s,
1H). MS (ESI): m/z 417 (M þ Hþ).
This compound was synthesized from g-bromoester 12 [16]
(0.18 g, 0.61 mmol) and amine 15 [17a] (0.32 g, 1.2 mmol, reac-
tion time ¼ 5 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (9:1) as the eluent to obtain pure 10e as a
pale yellow oil, which crystallized on standing (0.14 g, yield 53%).
An analytical sample was obtained by recrystallization from n-
hexane-dichloromethane (mp 128e129 ꢂC). 1H NMR (400 MHz,
CDCl3): 1.60e1.79 (m, 4H), 2.48 (t, J ¼ 7.2, 2H), 2.65 (m, 4H), 3.08 (m,
4H), 3.75 (t, J ¼ 7.0, 2H), 3.84 (s, 3H), 4.42 (s, 2H), 6.82e6.98 (m, 4H),
7.45e7.65 (m, 3H), 7.91e7.95 (m, 2H), 9.25 (d, J ¼ 8.2, 1H). MS (ESI):
m/z 430 (M þ Hþ).
4.1.4. 2-[3-[4-(2-Methoxyphenyl)piperazin-1-yl]propyl]-2,3-
dihydro-1H-benzo[e]isoindol-1-one (10b)
4.1.8. 2-[4-[4-(2,3-Dichlorophenyl)piperazin-1-yl]butyl]-2,3-
This compound was synthesized from
g-bromoester 12 [16]
dihydro-1H-benzo[e]isoindol-1-one (10f)
(0.13 g, 0.44 mmol) and amine 14 [17a] (0.22 g, 0.89 mmol, reac-
tion time ¼ 6 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (9:1) as the eluent to obtain pure 10b as a
pale yellow glassy solid (0.12 g, yield 65%). 1H NMR (200 MHz,
CDCl3): 1.97 (m, 2H), 2.52 (t, J ¼ 7.3, 2H), 2.66 (m, 4H), 3.11 (m, 4H),
3.74 (t, J ¼ 7.0, 2H), 3.83 (s, 3H), 4.46 (s, 2H), 6.81e6.99 (m 4H),
7.50e7.68 (m, 3H), 7.88e7.98 (m, 2H), 9.25 (d, J ¼ 8.1, 1H). MS (ESI):
m/z 416 (M þ Hþ).
This compound was synthesized from g-bromoester 12 [16]
(0.18 g, 0.61 mmol) and amine 16 [17b] (0.37 g, 1.2 mmol, reaction
time ¼ 6 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (8:2) as the eluent to obtain pure 10f as a pale
yellow oil, which crystallized on standing (0.075 g, yield 26%, mp
92e93 ꢂC). 1H NMR (200 MHz, CDCl3): 1.58e1.82 (m, 4H), 2.47 (t,
J ¼ 7.4, 2H), 2.62 (m, 4H), 3.04 (m, 4H), 3.72 (t, J ¼ 7.1, 2H), 4.45 (s,
2H), 6.93 (m,1H), 7.12 (m, 2H), 7.48-7.68 (m, 3H), 7.90 (d, J ¼ 8.1,1H),
7.98 (d, J ¼ 8.4, 1H). 9.23 (d, J ¼ 9.0, 1H). MS (ESI): m/z 468 (M þ Hþ).
4.1.5. 2-[4-[4-(2-Methoxyphenyl)piperazin-1-yl]butyl]-2,3-
dihydro-1H-pyrrolo[3,4-b]quinolin-1-one (10c)
4.1.9. 4-Chloro-2-[4-[4-(2-methoxyphenyl)piperazin-1-yl]butyl]-
This compound was synthesized from
g
-chloroester 11 [16]
2,3-dihydro-1H-pyrrolo[3,4-c]quinolin-1-one (10g)
(0.12 g, 0.48 mmol) and 4-[4-(2-methoxyphenyl)piperazin-1-yl]
butan-1-amine [17a] (15, 0.32 g, 1.2 mmol, reaction time ¼ 18 h)
and purified by flash-chromatography with ethyl acetate-
triethylamine (9:1) as the eluent to obtain pure 10c as a white
solid (0.16 g, yield 77%). An analytical sample was obtained by
recrystallization from n-hexane-dichloromethane (mp 124e
125 ꢂC). 1H NMR (200 MHz, CDCl3): 1.51e1.83 (m, 4H), 2.43 (t,
J ¼ 7.2, 2H), 2.59 (m, 4H), 3.03 (m, 4H), 3.71 (t, J ¼ 7.0, 2H), 3.80 (s,
3H), 4.53 (s, 2H), 6.78e6.97 (m, 4H), 7.56 (t, J ¼ 7.3, 1H), 7.77 (t,
This compound was prepared from g-bromoester 13 [16] (0.41 g,
1.25 mmol) with amine 15 [17a] (0.66 g, 2.5 mmol, reaction
time ¼ 6 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (9:1) as the eluent to give compound 10g as a
pale yellow oil, which crystallized on standing (0.27 g, yield 46%,
mp 126e127 ꢂC). 1H NMR (200 MHz, CDCl3): 1.57e1.86 (m, 4H),
2.46 (t, J ¼ 7.2, 2H), 2.63 (m, 4H), 3.06 (m, 4H), 3.72 (t, J ¼ 6.9, 2H),
3.81 (s, 3H), 4.47 (s, 2H), 6.80e6.99 (m, 4H), 7.63e7.81 (m, 2H), 8.06
(d, J ¼ 8.1, 1H), 8.98 (d, J ¼ 7.9, 1H). MS (ESI): m/z 465 (M þ Hþ).

A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
91
4.1.10. 4-Chloro-2-[4-[4-(2,3-dichlorophenyl)piperazin-1-yl]butyl]-
2,3-dihydro-1H-pyrrolo[3,4-c]quinolin-1-one (10h)
This compound was prepared from -bromoester 13 [16] (0.41 g,
4.1.15. 4-Chloro-2-[5-[4-(2-methoxyphenyl)piperazin-1-yl]pentyl]-
2,3-dihydro-1H-pyrrolo[3,4-c]quinolin-1-one (10m)
g
This compound was synthesized from g-bromoester 13 [16]
1.25 mmol) with amine 16 [17b] (0.75 g, 2.5 mmol, reaction
time ¼ 6 h) and purified by flash-chromatography with ethyl ace-
tate as the eluent to obtain compound 10h as a white solid (0.28 g,
yield 45%). An analytical sample was obtained by recrystallization
from n-hexane-dichloromethane (mp 140e141 ꢂC). 1H NMR
(200 MHz, CDCl3): 1.55e1.88 (m, 4H), 2.47 (t, J ¼ 7.2, 2H), 2.62 (m,
4H), 3.05 (m, 4H), 3.76 (t, J ¼ 7.0, 2H), 4.52 (s, 2H), 6.92 (m, 1H), 7.13
(m, 2H), 7.66e7.85 (m, 2H), 8.11 (d, J ¼ 8.4, 1H), 9.03 (d, J ¼ 8.5, 1H).
MS (ESI): m/z 503 (M þ Hþ).
(12 g, 36.5 mmol) and amine 17 [17c] (20 g, 72 mmol, reaction
time ¼ 6 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (9:1) as the eluent to obtain pure 10m as a
pale glassy solid (2.5 g, yield 14%). 1H NMR (200 MHz, CDCl3): 1.35e
1.85 (m, 6H), 2.42 (t, J ¼ 7.3, 2H), 2.64 (m, 4H), 3.07 (m, 4H), 3.70 (t,
J ¼ 7.2, 2H), 3.83 (s, 3H), 4.47 (s, 2H), 6.80e7.00 (m, 4H), 7.63-7.82
(m, 2H), 8.07 (d, J ¼ 8.5, 1H), 9.01 (d, J ¼ 8.1, 1H). MS (ESI): m/z 479
(M þ Hþ).
4.1.16. 4-Chloro-2-[5-[4-(2,3-dichlorophenyl)piperazin-1-yl]
pentyl]-2,3-dihydro-1H-pyrrolo[3,4-c]quinolin-1-one (10n)
4.1.11. 2-[5-[4-(2-Methoxyphenyl)piperazin-1-yl]pentyl]-2,3-
dihydro-1H-pyrrolo[3,4-b]quinolin-1-one (10i)
This compound was prepared from g-bromoester 13 [16] (0.10 g,
This compound was synthesized from
g-chloroester 11 [16]
0.30 mmol) and amine 18 [17b] (0.19 g, 0.61 mmol, reaction
time ¼ 5 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (9:1) as the eluent to obtain pure 10n as a
white solid (0.092 g, yield 59%). An analytical sample was obtained
by recrystallization from ethyl acetate-dichloromethane (mp 181e
182 ꢂC). 1H NMR (200 MHz, CDCl3): 1.43e1.86 (m, 6H), 2.42 (t,
J ¼ 7.3, 2H), 2.61 (m, 4H), 3.03 (m, 4H), 3.73 (t, J ¼ 7.2, 2H), 4.50 (s,
2H), 6.90 (m, 1H), 7.12 (m, 2H), 7.66-7.84 (m, 2H), 8.11 (d, J ¼ 8.6,
1H), 9.05 (d, J ¼ 8.0, 1H). MS (ESI): m/z 517 (M þ Hþ).
(0.10 g, 0.40 mmol) and 5-[4-(2-methoxyphenyl)piperazin-1-yl]
pentan-1-amine [17c] (17, 0.22 g, 0.80 mmol, reaction time ¼ 21 h)
and purified by flash-chromatography with ethyl acetate-
triethylamine (9:1) as the eluent to obtain pure 10i as a pale yel-
low oil (0.14 g, yield 79%). An analytical sample was obtained by
recrystallization from n-hexane-dichloromethane (mp 116e
118 ꢂC). 1H NMR (200 MHz, CDCl3): 1.34e1.78 (m, 6H), 2.38 (t,
J ¼ 7.4, 2H), 2.58 (m, 4H), 3.04 (m, 4H), 3.69 (t, J ¼ 7.1, 2H), 3.81 (s,
3H), 4.53 (s, 2H), 6.78-6.98 (m, 4H), 7.57 (t, J ¼ 7.3, 1H), 7.78 (t,
J ¼ 7.1, 1H), 7.95 (d, J ¼ 8.6, 1H), 8.10 (d, J ¼ 8.5, 1H), 8.56 (s, 1H). MS
(ESI): m/z 445 (M þ Hþ).
4.2. X-ray crystallography
Single crystals of compound 10c were submitted to X-ray data
4.1.12. 2-[5-[4-(2,3-Dichlorophenyl)piperazin-1-yl]pentyl]-2,3-
dihydro-1H-pyrrolo[3,4-b]quinolin-1-one (10j)
collection on a Siemens P4 diffractometer with a graphite mono-
ꢀ
chromated Mo-K
a
radiation (
l
¼ 0.71073 A) at 293 K. The structure
This compound was synthesized from
g-chloroester 11 [16]
was solved by direct methods implemented in SHELXS-97 program
[26]. The refinement was carried out by full-matrix anisotropic
least-squares on F2 for all reflections for non-H atoms by means of
the SHELXL-97 program [27]. Crystallographic data (excluding
structure factors) for the structure solved in this paper have been
deposited with the Cambridge Crystallographic Data Centre as
supplementary publication CCDC 877453. Copies of the data can be
obtained, free of charge, on application to CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK; (fax: þ 44 (0) 1223 336 033; or e-mail:
deposit@ccdc.cam.ac.uk).
(0.10 g, 0.40 mmol) and 5-[4-(2,3-dichlorophenyl)piperazin-1-yl]
pentan-1-amine [17b] (18, 0.25 g, 0.79 mmol, reaction time ¼ 24 h)
and purified by flash-chromatography with ethyl acetate-
triethylamine (8:2) as the eluent to obtain pure 10j as a pale yel-
low oil, which crystallized on standing (0.13 g, yield 67%). An ana-
lytical sample was obtained by recrystallization from n-hexane-
dichloromethane (mp 93e95 ꢂC). 1H NMR (200 MHz, CDCl3): 1.38e
1.84 (m, 6H), 2.40 (t, J ¼ 7.3, 2H), 2.59 (m, 4H), 3.02 (m, 4H), 3.72 (t,
J ¼ 7.1, 2H), 4.56 (s, 2H), 6.90 (m, 1H), 7.11 (m, 2H), 7.60 (t, J ¼ 7.6,
1H), 7.81 (t, J ¼ 7.8, 1H), 7.98 (d, J ¼ 7.8, 1H), 8.12 (d, J ¼ 8.5, 1H), 8.59
(s, 1H). MS (ESI): m/z 483 (M þ Hþ).
4.3. In vitro pharmacological studies
4.1.13. 2-[5-[4-(2-Methoxyphenyl)piperazin-1-yl]pentyl]-2,3-
dihydro-1H-benzo[e]isoindol-1-one (10k)
4.3.1. Binding assays
Male Wistar rats (Charles River Italia, Calco, CO, Italy) weighing
200e250 g were used. Animal care and handling throughout the
experimental procedures were in accordance with the European
Communities Council Directive of 24 November 1986 (86/609/EEC).
Rats were sacrificed by decapitation, brains were rapidly dis-
sected and hippocampi were used for binding assay preparation
according to Hall et al. [18] Crude membranes were diluted in the
binding buffer (Tris-Cl 50 mM, pH 7.4) in order to obtain the final
protein concentration.
This compound was synthesized from
g-bromoester 12 [16]
(4.1 g, 14 mmol) and amine 17 [17c] (7.8 g, 28 mmol, reaction
time ¼ 6 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (9:1) as the eluent to obtain pure 10k as a
pale yellow glassy solid (2.6 g, yield 42%). 1H NMR (200 MHz,
CDCl3): 1.34e1.81 (m, 6H), 2.40 (t, J ¼ 7.4, 2H), 2.62 (m, 4H), 3.05 (m,
4H), 3.67 (t, J ¼ 7.1, 2H), 3.83 (s, 3H), 4.41 (s, 2H), 6.80e7.01 (m, 4H),
7.44e7.67 (m, 3H), 7.86e7.97 (m, 2H), 9.24 (d, J ¼ 8.2, 1H). MS (ESI):
m/z 444 (M þ Hþ5.
[3H]8-OH-DPAT specific binding was measured in 0.5 mL total
4.1.14. 2-[5-[4-(2,3-Dichlorophenyl)piperazin-1-yl]pentyl]-2,3-
volume. Aliquots (50
mL) of tissue suspension (3 mg/mL protein
dihydro-1H-benzo[e]isoindol-1-one (10l)
concentration) were incubated with 50
m
L of 0.5 nM labeled ligand
This compound was prepared from g-bromoester 12 [16] (0.12 g,
(Perkin Elmer Life and Analytical Sciences), and different concen-
trations of the compound under investigation. Five to eight con-
centrations of each compound were assayed, each performed in
duplicate. The incubation was performed in polypropylene test
tubes at 37 ꢂC for 10 min. The bound radioligand was separated by
rapid filtration on glass fiber Whatman GF/C filters. Filtrates were
washed four times with 4 mL of cold binding buffer, before the filter
disks were transferred in minivials filled with 4 mL of Ultima Gold
(Perkin Elmer Life and Analytical Sciences).
0.41 mmol) and amine 18 [17b] (0.26 g, 0.82 mmol, reaction
time ¼ 12 h) and purified by flash-chromatography with ethyl
acetate-triethylamine (9:1) as the eluent to obtain pure 10l as a pale
yellow oil, which crystallized on standing (0.080 g, yield 40%, mp
90e91 ꢂC). 1H NMR (200 MHz, CDCl3): 1.42e1.82 (m, 6H), 2.41 (t,
J ¼ 7.3, 2H), 2.60 (m, 4H), 3.02 (m, 4H), 3.70 (t, J ¼ 7.2, 2H), 4.43 (s,
2H), 6.88 (m, 1H), 7.09 (m, 2H), 7.46e7.67 (m, 3H), 7.87e7.98 (m,
2H), 9.24 (d, J ¼ 8.3, 1H). MS (ESI): m/z 482 (M þ Hþ).

92
A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
The specific binding of [3H]8-OH-DPAT (final concentration
0.5 nM) was defined as the difference between the total binding
sequence database. In the second step, the previously obtained
profile was used to scan the PDB database of known protein struc-
ture for possible templates. Usage of the exhaustive profile drasti-
cally increases the probability to find a suitable template structure
[35]. Following this approach three best templates with higher
sequence identity were selected in order to build the 3D structure of
and the nonspecific binding determined in the presence of 10
HT and represented about 80e85% of the total binding.
mM 5-
Competition experiments were analyzed by the “Allfit” program
[28] to obtain the concentration of unlabeled drug that caused 50%
inhibition of [3H]8-OH-DPAT specific binding (IC50). Apparent
affinity constants (Ki) were derived from the IC50 values according
to the Cheng and Prusoff equation [Ki ¼ IC50/(1 þ L/Kd)] [29].
The binding of compounds 10k,m to human recombinant D2S
receptors expressed in HEK-293 cells was assayed following
described methods [30]. The specific binding of [3H] spiperone
(final concentration 0.3 nM; incubation 22 ꢂC for 1 h) was deter-
mined as the difference between the total binding and the non-
the 5-HT1AR. The sequences of h
b2-adrenergic receptor (Protein
Data Bank code 2RH1; sequence identity of chain A 36.9%), t
b
1-
adrenergic receptor (2VT4; sequence identity of chain B 29.7%)
and hA2A adenosine receptor (3EML; sequence identity of chain A
22.7%) were imported into Prime [36] and aligned to the query
sequence. Subsequently, these templates 3D structures aligned were
used for “Comparative Modeling” methods implemented in Prime.
Since we aligned multiple templates, which in Prime can be used in
several ways to build a model, we specified, in the “build structure
step”, which method was used for building the 5-HT1AR 3D struc-
ture. Consensus model optionwas employed to build the model; this
option allowed us to take into account all the previously selected
templates since the model was built as an average of all templates.
The obtained model was submitted to refinement protocol in Prime
environment. In particular, we have performed a side-chains opti-
mization and loops refinement by Prime and default settings were
employed. Further structure optimization was carried out by Mac-
roModel application implemented in Maestro [37] suite 2011 using
the Optimized Potentials for Liquid Simulations-all atom (OPLS-AA)
force field 2005 [38]. The solvent effects were simulated using the
analytical Generalized-Born/Surface-Area (GB/SA) model [39], and
no cutoff for non-bonded interactions was selected. Polak-Ribierè
Conjugate Gradient (PRCG) method with 100,000 maximum iter-
ations and 0.001 gradient convergence threshold was employed.
Moreover, the minimized 5-HT1AR homology model was submitted
to Protein Preparation Wizard Workflow implemented in Maestro
suite 2011 [40]. This protocol allowed us to obtain a reasonable
starting structure of the receptor for our molecular docking calcu-
lations, by a series of computational steps. In particular, we per-
formed three steps in order to: i) add missed hydrogens; ii) optimize
the orientation of hydroxyl groups, Asn, and Gln, and the proto-
nation state of His; and iii) perform a constrained refinement with
the impref utility, setting the max root-mean-square deviation
specific binding determined in the presence of 1.0
(þ)-butaclamol.
mM unlabeled
4.3.2. 5-HT1AR cellular functional assays
Cyclic AMP accumulation was determined in CHO cells trans-
fected with the human 5-HT1AR receptor as previously described
[31]. Briefly, cells were incubated (22 ꢂC for 15 min) with com-
pounds 10k or 10m in DMEM,10 mM HEPES,100 mM forskolin, and
100 mM 3-isobutyl-1-methylxanthine (IBMX). The reaction was
stopped by aspiration of the medium and addition of 0.1 N HCl. Data
were expressed as % of the response obtained with 100 nM 8-OH-
DPAT (agonist effect) and/or as % of inhibition of control 10 nM 8-
OH-DPAT stimulus (antagonist effect).
4.4. In vivo pharmacokinetic studies
Male CD1 mice (20e30 g; Charles River Labs, France) were fas-
ted overnight prior to dosing. Two groups of mice (n ¼ 3) were
dosed intravenously (iv) at 2 mg/kg with both 10k,m. Sampling
time points were 5, 15, 30, 60, 120, 240, 360, and 1440 min after iv
administration of each tested compound. Two additional groups of
mice (n ¼ 3) were dosed orally at 10 mg/kg. Sampling time points
were 15, 30, 60, 120, 240, 360, 480 and 1440 min after oral
administration of each tested compound. Blood was drawn by a
cardiac puncture at each time point and collected in tubes coated
with lithium heparin, mixed gently, then kept on ice and centri-
fuged at 2500 g for 15 min at 4 ꢂC, within 1 h of collection. Plasma
was then harvested and stored at ꢃ20 ꢂC prior to analysis.
ꢀ
(RMSD) of 0.30 A. The impref utility consists of a cycles of energy
minimizations based on the impact molecular mechanics engine
and on the OPLS_2005 force field.
The modeled protein was validated by Ramachandran plot (see
Supplementary material) generated by RAMPAGE webserver [41].
The residues of protein were for 82.4% (336 amino acids) in the
favored region of the plot, 17.2% (70 amino acids not involved in the
binding site) of the residues lie in additional allowed region and
only 0.5% (2 amino acids not involved in the binding site) of the
residues were located in the disallowed region as depicted in. The
results of RAMPAGE webserver revealed that over 99% of the resi-
dues of our 5-HT1AR refined model sit in the allowed regions of
Ramachandran Plot. This value is more than the cut-off value
(96.1%) defined for the most reliable models [42]. Consequently, the
stereo chemical quality of our 5-HT1AR homology model was found
acceptable displaying a very low percentage of residues having phi/
psi angles in outlier region.
4.5. In vivo safety assay (Irwin test)
General behavioral observations were recorded by camera. Male
CD1 mice (n ¼ 3/group; Harlan, S.Pietro al Natisone, UD, Italy) were
dosed orally with 10k,m up to 80 mg/kg. Compounds were sus-
pended in a solution containing 0.1% Tween 80 þ 0.5% methyl
cellulose (SigmaeAldrich, Milan) and administered in a volume of
10 mL/kg. Animals were kept under observation for a total time of
24 h after a single oral administration of drugs.
4.6. Molecular modeling
4.6.1. Homology modeling
The sequence of rat 5-HT1AR was taken in fasta format from the
UniProtKB (entry P19327) [32]. The template selection was per-
formed by HOMER-A (HOmology ModellER) using automatic tem-
plate search method (version 1.3) [33]. In the automatic template
selection mode the web-server application searched for a template
structure for the modeling of the target sequence using the PDB-
BLAST protocol [34]. This procedure was performed in two steps:
in the first step, the target sequence was used as a seed to construct
an exhaustive sequence profile on the non-redundant protein
4.6.2. Ligand setup
Starting from our experimental X-ray data the structures of the
ligands used in the docking simulations were generated by means
of Macmolplt [43], and were optimized by means of GAMESS [44].
Ligand Optimizations and partial atomic point charges calculations
were performed with RHF and 6-31g* as basis set. All the relevant
torsion angles were treated as rotatable during the docking process,
thus allowing a search of the conformational space.

A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
93
(c) E. Richelson, T. Souder, Binding of antipsychotic drugs to human brain recep-
tors. Focus on newer generation compounds, Life Sci. 68 (2000) 29e39.
[8] K. Yoshida, T. Sugita, H. Higuchi, Y. Hishikawa, Effect of tandospirone on tardive
dyskinesia and parkinsonian symptoms, Eur. Psychiatry 13 (1998) 421e422.
[9] M.J. Millan, Improving the treatment of schizophrenia: focus on serotonin 5-
HT1A receptors, J. Pharmacol. Exp. Ther. 295 (2000) 853e861.
[10] D.A. Shapiro, S. Renock, E. Arrington, L.A. Chiodo, L.X. Liu, D.R. Sibley,
B.L. Roth, R. Mailman, Aripiprazole, a novel atypical antipsychotic drug with
a unique and robust pharmacology, Neuropsychopharmacology 28 (2003)
1400e1411.
4.6.3. Receptor setup
The homology model of 5-HT1AR was setup for docking as fol-
lows: receptor was prepared by protein preparation wizard proto-
col implemented in Maestro as described in the homology
modeling section, and Kollman united-atom partial charges were
as-signed. The ADDSOL utility of AutoDock was used to add sol-
vation parameters to the protein structures, and the grid maps
representing the proteins in the docking process were calculated
using AutoGrid. The grids, one for each atom type in the inhibitor
plus one for the electrostatic interactions, were chosen to be large
enough to include not only the hypothetical channel binding site
sites but also a significant part of the protein around it. As a con-
sequence, the dimensions of the grid maps were 100 ꢁ 100 ꢁ 100
[11] D. Marona-Lewicka, D.E. Nichols, Aripiprazole (OPC-14597) fully substitutes
for the 5-HT1A receptor agonist LY293284 in the drug discrimination assay in
rats, Psychopharmacology 172 (2004) 415e421.
[12] S. Jordan, V. Koprivica, R. Chen, K. Tottori, T. Kikuchi, C.A. Altar, The anti-
psychotic aripiprazole is
a potent, partial agonist at the human 5-HT1A
receptor, Eur. J. Pharmacol. 441 (2002) 137e140.
[13] F. Mauler, T. Fahrig, E. Horvath, R. Jork, Inhibition of evoked glutamate release
by the neuroprotective 5-HT1A receptor agonist BAY X 3702 in vitro and
in vivo, Brain Res. 888 (2001) 150e157.
ꢀ
points, with a grid-point spacing of 0.375 A.
[14] H.Y. Meltzer, T. Sumiyoshi, Does stimulation of 5-HT1A receptors improve
cognition in schizophrenia? Behav. Brain Res. 195 (2008) 98e102.
[15] (a) A. Cappelli, M. Anzini, S. Vomero, L. Mennuni, F. Makovec, M. Hamon,
P.G. De Benedetti, M.C. Menziani, The interactions of the 5-HT3 receptor with
arylpiperazine, tropane, and quinuclidine ligands, Curr. Top. Med. Chem. 2
(2002) 599e624;
4.6.4. Docking calculations
All the docked compounds were subjected to 256 independent
runs of docking simulations by means of the AutoDock 4.2 [19]
using the Lamarckian genetic algorithm and through a protocol
with an initial population of 300 randomly placed individuals, a
maximum number of 50 million energy evaluations, a mutation
rate of 0.02, a crossover rate of 0.80, and an elitism value of 1. The
pseudo-Solis and Wets algorithm with a maximum of 300 inter-
actions was applied for the local search. 256 independent docking
runs were carried out for each ligand, and the resulting con-
(b) A. Cappelli, S. Butini, A. Brizzi, S. Gemma, S. Valenti, G. Giuliani, M. Anzini,
L. Mennuni, G. Campiani, V. Brizzi, S. Vomero, The interactions of the 5-HT3
receptor with quipazine-like arylpiperazine ligands: the journey track at the
end of the first decade of the third millennium, Curr. Top. Med. Chem. 10
(2010) 504e526;
(c) G. Campiani, S. Butini, F. Trotta, C. Fattorusso, B. Catalanotti, F. Aiello,
S. Gemma, V. Nacci, E. Novellino, J.A. Stark, A. Cagnotto, E. Fumagalli,
F. Carnovali, L. Cervo, T. Mennini, Synthesis and pharmacological evaluation of
potent and highly selective D3 receptor ligands: inhibition of cocaine-seeking
behavior and the role of dopamine D3/D2 receptors, J. Med. Chem. 46 (2003)
3822e3839;
(d) S. Butini, S. Gemma, G. Campiani, S. Franceschini, F. Trotta, M. Borriello,
N. Ceres, S. Ros, S.S. Coccone, M. Bernetti, M. De Angelis, M. Brindisi, V. Nacci,
I. Fiorini, E. Novellino, A. Cagnotto, T. Mennini, K. Sandager-Nielsen,
J.T. Andreasen, J. Scheel-Kruger, J.D. Mikkelsen, C. Fattorusso, Discovery of a
new class of potential multifunctional atypical antipsychotic agents targeting
dopamine D3 and serotonin 5-HT1A and 5-HT2A receptors: design, synthesis,
and effects on behavior, J. Med. Chem. 52 (2009) 151e169.
ꢀ
formations that differed by less than 2.0 A in positional RMSD were
clustered together. Cluster analysis was performed by selecting the
lowest energy solution of the most populated cluster. The lowest
energy solutions resulting from the docking calculations in each
case belong to the most populated cluster; the most populated
clusters were taken into account for their consistency with exper-
imental data. All estimated binding energies was in agreement with
the nanomolar range of the biological data. All protonated con-
formations were optimized and the point charges were obtained by
means of ab-initio calculations, and then used into the docking
simulations.
[16] A. Cappelli, M. Anzini, S. Vomero, L. Mennuni, F. Makovec, E. Doucet,
M. Hamon, M.C. Menziani, P.G. De Benedetti, G. Giorgi, C. Ghelardini,
S. Collina, Novel potent 5-HT3 receptor ligands based on the pyrrolidone
structure: synthesis, biological evaluation, and computational rationalization
of the ligand-receptor interaction modalities, Bioorg. Med. Chem. 10 (2002)
779e801.
[17] (a) A. Hackling, R. Ghosh, S. Perachon, A. Mann, H.D. Holtje, C.G. Wermuth,
J.C. Schwartz, W. Sippl, P. Sokoloff, H. Stark, N-(omega-(4-(2-methoxyphenyl)
piperazin-1-yl)alkyl)carboxamides as dopamine D2 and D3receptor ligands,
J. Med. Chem. 46 (2003) 3883e3899;
Acknowledgments
The authors are grateful to Dr. Francesco Berrettini (CIADS,
Università di Siena) for the X-ray data collection. This work was
partly supported by MIUR (Ministero dell’Istruzione, dell’Università
e della Ricerca) - PRIN (Programmi di ricerca di Rilevante Interesse
Nazionale).
(b) M.J. Robarge, S.M. Husbands, A. Kieltyka, R. Brodbeck, A. Thurkauf,
A.H. Newman, Design and synthesis of [(2,3-dichlorophenyl)piperazin-1-yl]
alkylfluorenylcarboxamides as novel ligands selective for the dopamine D3
receptor subtype, J. Med. Chem. 44 (2001) 3175e3186;
(c) M. Resimont, J.F. Liegeois, Synthesis and in vitro binding studies of
piperazine-alkyl-naphthamides: impact of homology and sulphonamide/car-
boxamide bioisosteric replacement on the affinity for 5-HT1A, alpha2A, D4.2,
D3 and D2L receptors, Bioorg. Med. Chem. Lett. 20 (2010) 5199e5202.
[18] M.D. Hall, S. el Mestikawy, M.B. Emerit, L. Pichat, M. Hamon, H. Gozlan, [3H]8-
hydroxy-2-(di-n-propylamino)tetralin binding to pre- and postsynaptic 5-
hydroxytryptamine sites in various regions of the rat brain, J. Neurochem.
44 (1985) 1685e1696.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2013.01.044.
[19] G.M. Morris, D.S. Goodsell, R.S. Halliday, R. Huey, W.E. Hart, R.K. Belew,
A.J. Olson, Automated docking using a Lamarckian genetic algorithm and an
empirical binding free energy function, J. Comput. Chem. 19 (1998) 1639e1662.
[20] (a) A. Rayan, New vistas in GPCR 3D structure prediction, J. Mol. Model. 16
(2010) 183e191;
References
[1] J. Hannon, D. Hoyer, Molecular biology of 5-HT receptors, Behav. Brain Res.
195 (2008) 198e213.
(b) M. Li, H. Fang, L. Du, L. Xia, B. Wang, Computational studies of the binding
site of alpha1A-adrenoceptor antagonists, J. Mol. Model. 14 (2008) 957e966.
[21] (a) J.C. Mobarec, R. Sanchez, M. Filizola, Modern homology modeling of G-
protein coupled receptors: which structural template to use? J. Med. Chem. 52
(2009) 5207e5216;
[2] J.P. Foong, J.C. Bornstein, 5-HT antagonists NAN-190 and SB 269970 block
alpha2-adrenoceptors in the guinea pig, Neuroreport 20 (2009) 325e330.
[3] R. Schreiber, J. De Vry, 5-HT1A receptor ligands in animal models of anxiety,
impulsivity and depression: multiple mechanisms of action? Prog. Neuro-
Psychopharmacol. Biol. Psychiatry 17 (1993) 87e104.
(b) P. Sokkar, S. Mohandass, M. Ramachandran, Multiple templates-based
homology modeling enhances structure quality of AT1 receptor: validation
by molecular dynamics and antagonist docking, J. Mol. Model. 17 (2011)
1565e1577.
[4] G. Tunnicliff, Molecular basis of buspirone’s anxiolytic action, Pharmacol.
Toxicol. 69 (1991) 149e156.
[5] A.C. McCreary, C.A. Jones, Antipsychotic medication: the potential role of 5-
HT(1A) receptor agonism, Curr. Pharm. Des. 16 (2010) 516e521.
[6] P. Zhong, E.Y. Yuen, Z. Yan, Modulation of neuronal excitability by serotonin-
NMDA interactions in prefrontal cortex, Mol. Cell. Neurosci. 38 (2008) 290e299.
[7] (a) S.L. Mason, G.P. Reynolds, Clozapine has sub-micromolar affinity for 5-
HT1A receptors in human brain tissue, Eur. J. Pharmacol. 221 (1992) 397e398;
(b) J. Elliott, G.P. Reynolds, Agonist-stimulatedGTPgamma[35S]binding to5-HT1A
receptors in human post-mortem brain, Eur. J. Pharmacol. 386 (1999) 313e315;
[22] N. Fernandez-Fuentes, B.K. Rai, C.J. Madrid-Aliste, J.E. Fajardo, A. Fiser, Com-
parative protein structure modeling by combining multiple templates and
optimizing sequence-to-structure alignments, Bioinformatics 23 (2007)
2558e2565.
[23] A. Newman-Tancredi, M.B. Assie, N. Leduc, A.M. Ormiere, N. Danty, C. Cosi,
Novel antipsychotics activate recombinant human and native rat serotonin 5-

94
A. Cappelli et al. / European Journal of Medicinal Chemistry 63 (2013) 85e94
HT1A receptors: affinity, efficacy and potential implications for treatment of
schizophrenia, Int. J. Neuropsychopharmacol. 8 (2005) 341e356.
[24] B. Davies, T. Morris, Physiological parameters in laboratory animals and
humans, Pharm. Res. 10 (1993) 1093e1095.
[25] S. Irwin, Comprehensive observational assessment: Ia. A systematic, quanti-
tative procedure for assessing the behavioral and physiologic state of the
mouse, Psychopharmacologia 13 (1968) 222e257.
[34] L. Rychlewski, L. Jaroszewski, W. Li, A. Godzik, Comparison of sequence pro-
files. Strategies for structural predictions using sequence information, Protein
Sci. 9 (2000) 232e241.
[35] http://protein.bio.unipd.it/homer/helpauto.html, access date: 10.04.12.
[36] Prime, Version 3.0, Schrödinger, LLC, New York, NY, 2011.
[37] MacroModel, Version 9.9, Schrödinger, LLC, New York, NY, 2011.
[38] (a) W.L. Jorgensen, D.S. Maxwell, J. TiradoRives, Development and testing of
the OPLS all-atom force field on conformational energetics and properties of
organic liquids, J. Am. Chem. Soc. 118 (1996) 11225e11236;
(b) G.A. Kaminski, R.A. Friesner, J. Tirado-Rives, W.L. Jorgensen, Evaluation and
reparametrization of the OPLS-AA force field for proteins via comparison with
accurate quantum chemical calculations on peptides, J. Phys. Chem. B 105
(2001) 6474e6487.
[39] W.C. Still, A. Tempczyk, R.C. Hawley, T. Hendrickson, Semianalytical treatment
of solvation for molecular mechanics and dynamics, J. Am. Chem. Soc. 112
(1990) 6127e6129.
[40] Protein Preparation Wizard Workflow, http://www.schrodinger.com/
supportdocs/18/16., 2011.
[41] S.C. Lovell, I.W. Davis, W.B. Arendall 3rd, P.I. de Bakker, J.M. Word,
M.G. Prisant, J.S. Richardson, D.C. Richardson, Structure validation by Calpha
geometry: phi, psi and Cbeta deviation, Proteins 50 (2003) 437e450.
[42] R. Luthy, J.U. Bowie, D. Eisenberg, Assessment of protein models with 3-
dimensional profiles, Nature 356 (1992) 83e85.
[43] B.M. Bode, M.S. Gordon, MacMolPlt: a graphical user interface for GAMESS,
J. Mol. Graph. Model. 16 (1998) 133e138.
[44] K.K. Baldridge, M.S. Gordon, J.H. Jensen, N. Matsunaga, M.W. Schmidt,
T.L. Windus, J.A. Boatz, T.R. Cundari, Applications of parallel GAMESS, in:
T.G. Mattson (Ed.), Parallel Computing in Computational Chemistry, vol. 592,
American Chemical Society, Washington, DC, 1995, pp. 29e46.
[26] G.M. Sheldrick, SHELXS-97, Rel. 97e2,
A
Program for Automatic
Solution of Crystal Structures, University of Göttingen, Göttingen (Ger-
many), 1997.
[27] G.M. Sheldrick, SHELXL-97, Rel. 97e2,
A Program for Crystal Structure
Refinement, University of Göttingen, Göttingen (Germany), 1997.
[28] A. DeLean, P.J. Munson, D. Rodbard, Simultaneous analysis of families of sig-
moidal curves: application to bioassay, radioligand assay, and physiological
dose-response curves, Am. J. Physiol. 235 (1978) E97eE102.
[29] Y. Cheng, W.H. Prusoff, Relationship between the inhibition constant (Ki) and
the concentration of inhibitor which causes 50 per cent inhibition (I50) of an
enzymatic reaction, Biochem. Pharmacol. 22 (1973) 3099e3108.
[30] D.K. Grandy, M.A. Marchionni, H. Makam, R.E. Stofko, M. Alfano,
L. Frothingham, J.B. Fischer, K.J. Burke-Howie, J.R. Bunzow, A.C. Server, et al.,
Cloning of the cDNA and gene for a human D2 dopamine receptor, Proc. Natl.
Acad. Sci. U. S. A. 86 (1989) 9762e9766.
[31] A. Newman-Tancredi, L. Verriele, M.J. Millan, Differential modulation
by GTPgammaS of agonist and inverse agonist binding to h5-HT1A receptors
revealed by [3H]-WAY100,635, Br. J. Pharmacol. 132 (2001) 518e524.
[32] M. Magrane, the UniProt consortium, UniProt Knowledgebase: A Hub of
Integrated Protein Data Database, 2011: Bar009 (2011).
[33] S.C.E. Tosatto, The VICTOR Package for 3D Protein Structure Modelling, http://
protein.cribi.unipd.it/homer 2009.