Discovery of a new class of potential multifunctional atypical antipsychotic agents targeting dopamine d3 and serotonin 5-HT 1A and 5-HT2A receptors: Design, synthesis, and effects on behavior

J. Med. Chem. 2009, 52, 151–169

151

Discovery of a New Class of Potential Multifunctional Atypical Antipsychotic Agents Targeting

Dopamine D₃and Serotonin 5-HT_1Aand 5-HT_2AReceptors: Design, Synthesis, and Effects on

Behavior

Stefania Butini,^†,‡Sandra Gemma,^†,‡Giuseppe Campiani,*^,†,‡Silvia Franceschini,^†,‡Francesco Trotta,^†,‡Marianna Borriello,^†,§

Nicoletta Ceres,^†,§Sindu Ros,^†,‡Salvatore Sanna Coccone,^†,‡Matteo Bernetti,^†,‡Meri De Angelis,^†,‡Margherita Brindisi,^†,‡

Vito Nacci,^†,‡Isabella Fiorini,^†,‡Ettore Novellino,^†,§Alfredo Cagnotto,^†,|Tiziana Mennini,^†,|Karin Sandager-Nielsen,

Jesper Tobias Andreasen, Jorgen Scheel-Kruger, Jens D. Mikkelsen,^#and Caterina Fattorusso^†,§

European Research Centre for Drug DiscoVery and DeVelopment, UniVersity of Siena, Italy, Dipartimento Farmaco Chimico Tecnologico,

UniVersita´ di Siena, Via Aldo Moro, 53100 Siena, Italy, Dipartimento di Chimica delle Sostanze Naturali e Dipartimento di Chimica

Farmaceutica e Tossicologica, UniVersita` di Napoli Federico II, Via D. Montesano 49, 80131 Napoli, Italy, Istituto di Ricerche

Farmacologiche Mario Negri, Via La Masa 19, 20156 Milano, Italy, NeuroSearch A/S, PederstrupVej 93, Ballerup DK-2750, Denmark, and

Neurobiological Research Unit, UniVersity Hospital Rigshospitalet, Copenhagen, Denmark

ReceiVed June 6, 2008

Dopamine D₃antagonism combined with serotonin 5-HT_1Aand 5-HT_2Areceptor occupancy may represent

a novel paradigm for developing innovative antipsychotics. The unique pharmacological features of 5i are

a high afﬁnity for dopamine D₃, serotonin 5-HT_1Aand 5-HT_2Areceptors, together with a low afﬁnity for

dopamine D₂receptors (to minimize extrapyramidal side effects), serotonin 5-HT_2Creceptors (to reduce the

risk of obesity under chronic treatment), and for hERG channels (to reduce incidence of torsade des pointes).

Pharmacological and biochemical data, including speciﬁc c-fos expression in mesocorticolimbic areas,

conﬁrmed an atypical antipsychotic proﬁle of 5i in vivo, characterized by the absence of catalepsy at

antipsychotic dose.

Introduction

pine still remains invaluable for psychosis, presenting high

clinical efﬁcacy and a reduced incidence of EPS and hyper-

prolactinemia. Olanzapine may precipitate or unmask diabetes

in susceptible patients,⁵and its use was associated with a 12%

increase in excessive appetite compared to haloperidol.⁶Mean-

while, ziprasidone, risperidone, and quetiapine may be respon-

sible for drug-induced long QT syndrome (risk of malignant

ventricular arrhythmia).

A new antipsychotic agent (aripiprazole 4, Chart 1) has

recently been launched. It acts as a partial agonist at D₂R.^7,8Its

unique mechanism of action might underlie its efﬁcacy and low

risk of side effects seen with other antipsychotics. However,

unmet clinical needs still remain and include (i) more effective

antipsychotic therapy for treatment of refractory patients, (ii)

improved treatment of negative symptoms, and (iii) improved

treatment of cognitive dysfunction.

Several strategies for the development of novel antipsychotics

have started to provide additional tools for relieving the

symptoms of schizophrenia. In recent years, the pharmaceutical

industry and academia have shown signiﬁcant interest in the

development of novel antipsychotics characterized by interaction

with less obvious receptors such as metabotropic glutamate

receptors and tachykinin receptors.⁹

Second-generation antipsychotics combine D₂R occupancy

with activity at serotoninergic receptors (such as 5-HT_1A

receptors (5-HT_1AR) and 5-HT_2Areceptors (5-HT_2AR), i.e.,

clozapine) to provide drug therapies for resistant schizophrenic

patients, with prompter therapeutic beneﬁts and the improvement

of cognitive symptoms.¹⁰In contrast, we decided to exploit a

novel multireceptor afﬁnity proﬁle approach, which combines

antagonism at dopamine D₃receptors (D₃R) and at 5-HT_2AR

and partial agonism at serotonin 5-HT_1AR, with a low afﬁnity

for D₂R (no liability of EPS at antipsychotic doses) and 5-HT_2C

receptors (5-HT_2CR) (reducing the risk of obesity under chronic

treatment).

Schizophrenia is among the most serious mental illnesses; it

has a considerable social and economic impact and globally it

affects approximately 1% of the world population.¹

Antipsychotic medication is the main therapeutic intervention

for schizophrenia. Although the pathophysiology of the disease

has yet to be clearly deﬁned, the development of antipsychotic

drugs in recent decades has been heavily inﬂuenced by the

dopamine hypothesis, mainly supported by the capability of

antipsychotic drugs of interfering with dopamine receptors in

vivo and in vitro and by evidence that the clinical efﬁcacy of

typical antipsychotic drugs is correlated with their occupancy

at dopamine D₂receptors (D₂R).²

The early agents for the treatment of psychosis, the “typical”

antipsychotics (haloperidol, 1, Chart 1), were breakthrough

therapies for the positive symptoms of schizophrenia, but they

failed to manage its negative symptoms and cognitive impair-

ment. Nevertheless, typical antipsychotics carry a heavy side-

effect burden (i.e., extrapyramidal symptoms (EPS) and hyper-

prolactinemia) and are ineffective in one-third of schizophrenic

patients.³

The “atypical” antipsychotics (e.g., clozapine (2) and olan-

zapine (3)), are characterized by a multireceptor afﬁnity proﬁle,

which combines a potent antagonism for serotonin 5-HT_2Awith

a dopamine D₂and D₃receptors blockade.⁴Among the atypical

antipsychotic agents approved by regulatory authorities, cloza-

* To whom correspondence should be addressed. Phone: 0039-0577-

234172. Fax: 0039-0577-234333. E-mail:campiani@unisi.it.

†

European Research Centre for Drug Discovery and Development.

Universita` degli Studi di Siena.

Universita` degli Studi di Napoli “Federico II”.

Istituto Mario Negri, Milano.

‡

§

|

NeuroSearch A/S.

University Hospital Rigshospitalet.

#

10.1021/jm800689g CCC: $40.75

2009 American Chemical Society

Published on Web 12/15/2008

152 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Chart 1. Title and Reference Compounds

Butini et al.

evidence that activation of 5-HT_1AR can play a role in

aripiprazole-mediated behavior in rats.^16,17Furthermore,

because glutamatergic transmission is dysfunctional in schizo-

phrenia and because glutamate release is decreased by

5-HT_1AR activation,¹⁸agonist properties at postsynaptic

5-HT_1AR may be relevant to the therapeutic proﬁle of atypical

antipsychotic agents, improving negative symptoms and

cognitive deﬁcits. This topic has recently been reviewed by

Meltzer and Sumiyoshi,¹⁹who underline the additional

beneﬁcial effects found in many preclinical models based

on the reciprocal interaction with the 5-HT_1AR (agonists) and

5-HT_2AR (antagonists). Serotonin, through its interaction with

5-HT_2AR, inhibits neuronal activity in the substantia nigra

and ventral tegmental area. 5-HT_2AR antagonists may increase

the ﬁring rate of midbrain dopaminergic neurons in a state-

dependent manner and potentiate the increase in the activity

of nigrostriatal dopamine-containing neurons in response to

moderate D₂R antagonism by antipsychotics. Therefore,

5-HT_2AR antagonism may prevent or alleviate EPS induced

by acute or long-term treatment with typical drugs such as

1.

Although the role of dopamine D₂R and serotonin 5-HT_2AR

has been deﬁned, the role of D₃R is still controversial. Recently,

D₃R has been claimed as a potential target for antischizophrenic

drug development.²⁰This is mainly based on the speciﬁc dis-

tribution of D₃Rs in the mesolimbic dopaminergic system, along

with the observation of elevated levels of these receptors in

patients that are off antipsychotics. D₃R functions are primarily

related to the mesolimbic, rather than the nigrostriatal, dopam-

inergic system, thus the blockade of D₃R does not elicit EPS.

Furthermore, because dopamine through D₃R modulates the

cholinergic system at the prefrontal cortex level, D₃R antagonists

(devoid of muscarinic effects) may robustly enhance acetyl-

choline release in frontal cortex.²¹Consequently, D₃R antago-

nists might improve cognitive deﬁcits poorly treated by currently

available agents, including clozapine. The persistent concern is

whether selective D₃R blockade can relieve positive symptoms

of schizophrenia. Investigation of dopamine D₃R function was

limited because of the lack of highly selective ligands.

To validate the novel approach to antipsychotics and to

achieve an optimum interaction with dopamine and serotonin

receptors, we exploited our knowledge acquired in the develop-

ment of tricyclic atypical antipsychotics.²²Thus, we selected

the arylalkylpiperazine system as a ﬂexible scaffold for achiev-

ing a ﬁne balancing of dopamine D₃R and serotonin 5-HT_1AR

and 5-HT_2AR afﬁnity, reducing D₁R, D₂R, and 5-HT_2CR

occupancy. Starting from highly selective D₃R hits²³that lack

signiﬁcant behavioral effects, the present article deals with the

design, synthesis, and behavioral investigation of a rationally

designed set of arylpiperazines as novel and potent antipsy-

chotics, characterized by high afﬁnity for D₃R, 5-HT_1AR, and

5-HT_2AR. Their structure-activity relationships (SARs) for

dopamine and serotonin receptors were associated with variation

of the aromatic system and the substituents on the arylpiperazine

moiety. Among the analogues synthesized and tested (5a-m),

compound 5i, a selective antagonist at D₃R and 5-HT_2AR, and

partial agonist at 5-HT_1AR, was selected for further biological

investigation. Its atypical antipsychotic proﬁle is herein dis-

cussed, together with the molecular modeling study. Despite

its structural similarity to aripiprazole (nanomolar afﬁnity for

D₂R, D₃R, 5-HT_1AR, and 5-HT_2AR, Table 2), 5i shows a unique

pharmacological proﬁle, being designed to validate the novel

approach to atypical antipsychotics based on a D₃, 5-HT_1A, and

5-HT_2Amultireceptor afﬁnity proﬁle.

The serotoninergic system plays an important role in the

regulation of prefrontal cortex (PFC) functions, including

emotional control, cognitive behavior, and working memory.

PFC pyramidal neurons and GABA interneurons contain several

serotonin receptor subtypes with a particularly high density of

5-HT_1AR and 5-HT_2AR. It has recently been demonstrated in

PFC that NMDA receptor channels are the target of 5-HT_1AR

and that both receptors modulate the excitability of cortical

neurons, thus affecting cognitive functions.¹¹Indeed, a variety

of preclinical data has suggested that the 5-HT_1AR may be a

therapeutic target for the development of improved antipsychotic

drugs. Although the role of 5-HT_1AR in antipsychotic drug

efﬁcacy proﬁle remains speculative, 5-HT_1AR afﬁnity contributes

to the clinical efﬁcacy of most of the atypical antipsychotic drugs

(clozapine, olanzapine, aripiprazole) and their low liability for

EPS.¹²A growing number of studies show that 5-HT_1AR

activation attenuates antipsychotic-induced side effects in hu-

mans,¹³nonhuman primates,¹⁴and antipsychotic-induced cata-

lepsy in rats. An association has indeed been postulated

between agonist activity at 5-HT_1AR and anxiolytic or

antidepressant effects, improvements in cognitive and nega-

tive symptoms, and decreased development of EPS in

schizophrenia.¹⁵Recent studies have provided the ﬁrst in vivo

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 153

Table 1. Physical and Chemical Data for Compounds 5a-m

^aYields refer to isolated and puriﬁed materials. ^bRecrystallization solvent: methanol. ^cAll the compounds were analyzed within (0.4% of the theoretical

values.

Furthermore, long QT syndrome, experienced with several

tion of hERG channel blockade is now a key step along the

drug discovery trajectory of the pharmaceutical industry.

Therefore, hERG channel interaction was introduced as a

further parameter in our design strategy of the novel

antipsychotics.

antipsychotics, has prompted considerable efforts in trying

to deﬁne the molecular basis of this phenomenon. Conse-

quently, efforts to predict long QT syndrome risk have been

focused on assays testing hERG channel activities. Investiga-

154 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

Table 2. Binding Afﬁnities for D₁, D₂, D₃, 5-HT_1A, 5-HT_2A, and 5-HT_2CReceptors (K_inM ( SD)^aand hERG Channels (K_iµM) of Compounds 5a-m

and Reference Antipsychotics

b

c

d

e

f

g

compd

D₁

D₂

D_3r

5-HT_1A

5-HT_2A

NTⁱ

41 ( 5

23.4 ( 6

164 ( 12

149 ( 39

389 ( 34

300 ( 63

83 ( 28

5-HT_2C

NTⁱ

hERG^h

5a

5b

5c

5d

5e

5f

5g

5h

5i

>10000

859 ( 121

>10000

0.54 ( 0.07

0.099 ( 0.002

0.27 ( 0.02

33.5 ( 3.5

6.0 ( 0.5

3.7 ( 0.5

3.2 ( 0.4

0.105 ( 0.002

4.5 ( 0.8

8.2 ( 0.9

32.92 ( 3

0.14 ( 0.03

8.6 ( 0.9

876 ( 22

5.0 ( 0.3

14.7 ( 0.6

28.0 ( 1.3

118 ( 16

NTⁱ

0.39 (2)

0.20 (3)

0.30 (4)

1.84 (2)

NTⁱ

4600 ( 870

5500 ( 1130

>10000

136 ( 25

173 ( 23

1086 ( 190

434 ( 47

NTⁱ

1039 ( 190

1012 ( 29

6475 ( 646

339 ( 86

229 ( 76

122 ( 37

263 ( 31

63 ( 6

215 ( 15

75 ( 7

3530 ( 918

2.6 ( 0.5

210 ( 31

20 ( 17

>10000

4315 ( 880

4370 ( 852

3290 ( 418

NA^j

NTⁱ

9.3 ( 1.9

11.9 ( 1.7

4.0 ( 0.7

24.2 ( 2

18.0 ( 4.3

33 ( 3

1800 ( 310

160 ( 30

610 ( 85

5.6 ( 0.2

190 ( 15

b

18.4 ( 3.0

206 ( 14

NTⁱ

0.20 (4)

0.93 (2)

NTⁱ

0.16 (2)

0.21 (2)

0.26 (2)

0.12 (8)

17 (3)

15.3 ( 3.2

127 ( 28

122.76 ( 15

57 ( 11

5j

5k

5l

5m

NTⁱ

7124 ( 488

>10000

318 ( 22

353 ( 67

25 ( 3.5

1960 ( 180

50 ( 2

154 ( 14

1360 ( 191

4700 ( 500

4.8 ( 0.4

4.1 ( 0.9

22 ( 1.7

32 ( 2.2

c

200 ( 40

164 ( 88

10 ( 2

1

2

3

4

0.9 ( 0.04

319 ( 55

39 ( 5.9

4.0 ( 1.0

8.7 ( 0.9

0.15 ( 0.02

36 (2)

1.02 (2)

0.92 (2)

0.8 ( 0.07

3.8 ( 0.3

3.3 ( 0.5

Risperidone

6.7 ( 0.7

^aEach value is the mean ( SD of three determinations and represents nM K_ivalue. [³H]SCH 23390, rat striatum. [³H]spiperone, rat striatum.

^d[³H]-7-OH-DPAT, Sf9 cells. [³H]-8-OH-DPAT, rat hippocampus. [³H]ketanserin, rat cortex. [³H]mesulergine, guinea pig cortex. ^hEach value is the

mean of 2-8 determinations (number of determinations are given in parentheses), SD were all within 10% of the mean. ⁱNT not tested. ^jNA not active at

100 µM.

e

f

g

Table 3. Multiple Sequence Alignment of the Trans-Membrane Ligand-Binding Domain of Considered Human DAR and 5-HTR Subtypes

TM1-7

similarity (%)

receptor

subtype

TM2 sequence^{a b}

TM3 sequence^{a b}

TM4 sequence^{a b}

TM5 sequence^{a b}

TM6 sequence^{a b}

TM7 sequence^{a b}

,

D₃

D₂

5-HT_2C

5-HT_1A

5-HT_2A

D₁

LVMPWVVYL

LVMPLSLLA

LVLPMAALY

LVMPVSMLT

LVMPWKAVA

FVTLDVMMC

WISLDVLFS

FIALDVLCC

WIYLDVLFS

WVAFDIMCS

ITAVWVLAF

ISIVWVLSF

IAIVWAISI

ISLTWLIGF

IIAVWTISV

ISVAWTLSV

FVIYSSVVSFYLP

FVVYSSIVSFYVP

FVLIGSFVAFFIP

YTIYSTFGAFYIP

FVLIGSFVSFFIP

YAISSSVISFYIP

FIVCWLPFFL

FIICWLPFFI

FLIMWCPFFI

FILCWLPFFI

FVVMWCPFFI

FVCCWLPFFI

SATTWLGYV

SAFTWLGYV

NVFVWIGYV

AIINWLGYS

NVFVWIGYL

DVFVWFGWA

81

63

61

59

57

^aBold: conserved Asp residue on TM3, serine residues on TM5, and aromatic residues lining the ligand binding pocket. ^bUnderlined: residues displayed

in Figure 1A,C.

Scheme 2. Synthesis of Intermediate 13

Chemistry

The synthesis of compounds 5a-m (Chart 1 and Table 1) is

described in the Schemes 1-4.

The secondary amine synthons 8 and 13 were synthesized

following the synthetic strategies described in Schemes 1 and

2. Compound 8 was prepared starting from previously described

ethyl 7-methoxyindole-2-carboxylate 6,²⁴which was N-alkylated

with bromoacetonitrile to afford the cyano-derivative 7.²⁵

Reduction of nitrile 7 using an excess of lithium aluminum

hydride gave rise to a cyclic lactam intermediate that was further

reduced in situ to afford the target amine 8.²⁵

The synthesis of the known ꢀ-carboline 13 was realized by

a modiﬁcation of the literature procedure²⁶as described in

Scheme 2. Formylation of the indole-derivative 9 at C3 was

performed using N-chlorosuccinimide (NCS), triphenylphos-

phine, and N,N-dimethylformamide (DMF).²⁷The aldehyde 10

thus obtained was treated with nitromethane and ammonium

hydride, and the corresponding tryptamine derivative 12 was

acetate to afford 4-methoxy-3-(2-nitrovinyl)-1H-indole 11.²⁸

submitted to a classic Pictet-Spengler reaction, affording the

This latter was subsequently reduced by lithium aluminum

ꢀ-carboline 13.²⁶

The synthetic pathway for obtaining compounds 5a-i,m is

shown in Scheme 3. The bromo-derivatives 14a,b and 15 were

Scheme 1. Synthesis of Intermediate 8

obtained according to the procedure previously reported in the

literature.²³Subsequent alkylation of suitable arylpiperazines

(commercially available, or obtained by a previously reported

methodology,²³or according to Schemes 1 and 2) gave the ﬁnal

compounds 5a-i,m.

The ether-derivatives 5j-l (Scheme 4) were synthesized

starting from the aromatic hydroxy-derivatives 16a,b, which

were O-alkylated with the appropriate commercially available

dibromoalkane in the presence of cesium carbonate as a base.

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 155

Scheme 3. Synthesis of Compounds 5a–i,m

proﬁle). We had previously acquired experience in tricyclic

antipsychotics²²and in arylpiperazines designed to selectively

interact with D₃R to modulate cocaine-seeking behavior.²³In

this work, we therefore decided to exploit speciﬁc modiﬁcations

of the versatile arylpiperazine scaffold. We were following an

innovative pharmacological hypothesis aimed at modulating

activity toward highly homologous dopamine receptors, such

as D₂R and D₃R, while still maintaining a good afﬁnity for

serotonin 5-HT_1Aand 5-HT_2Areceptors and minimizing afﬁnity

for the D₁R and 5-HT_2CR subtypes. Furthermore, to improve

the drugability of the novel antipsychotics, we directed our

design strategy at limiting hERG potassium channel inhibition

by up to 1 µM, thus minimizing the risk of long QT syndrome

and cardiac side effects.

To drive the structural modiﬁcations of this new series

of arylpiperazine derivatives, we performed a conformational

and electronic analysis of the ligands, along with a structural

and bioinformatic analysis of the above-mentioned serotonin

and dopamine receptor subtypes (see Experimental Section for

details). The amino acid (AA) sequences of the receptors were

aligned²⁹with the aim of mapping their active sites’ homology

and a model of the putative binding site for our competitive

ligands was produced (Homology, Insight2005, Accelrys, San

Diego) (Table 3, Figure 1A,C). Indeed, all the considered

receptors (i.e., D₁, D₂, D₃, 5-HT_1A, 5-HT_2A, and 5-HT_2C) belong

to the 7tm_1 family of G-protein coupled receptors (GPCRs).

In particular, the two endogenous aromatic amines, dopamine

and serotonin, share a homologous binding site.³⁰Many GPCRs

models have been proposed, derived from analogies with high-

resolution structures of rhodopsin. Despite this, the actual seven

trans-membrane (TMs) helices connectivity, arrangement, and

3D structure in the different GPCR subtypes is still unknown.

However, biophysical and mutagenesis studies on rhodopsin and

a number of ligand-activated GPCRs provided evidence that

the formation of the receptor active state involves movements

of TM5, TM6, and TM7 with respect to TM3.^31-34In particular,

different clusters of aromatic residues, critical to receptor

structure and ligand recognition, were identiﬁed.^31,35,36The very

recently solved structure of the ꢀ₂-adrenergic receptor in

complex with the inverse agonist carazolol^37-39conﬁrmed the

data obtained from mutagenesis studies³⁵and offered a new

template for the generation of receptor 3D models (see

Experimental Section).

Scheme 4. Synthesis of Compounds 5j-l

The bromo-derivatives 17a-c were reacted with 2,3-dichlo-

rophenylpiperazine to give compounds 5j-l.²³

Results and Discussion

The generated receptor models provided the rational bases

for driving structural modiﬁcation of the arylpiperazine skeleton

in order to achieve the desired dopamine/serotonin receptor

subtype selectivity. Consequently, we ﬁrst identiﬁed, in the

arylpiperazine skeleton, different pharmacophoric moieties

(Chart 2) aimed at binding to the corresponding interaction sites

identiﬁed in different secondary structures of the receptor models

(Figure 1). In this view, the polymethylene tether length

is crucial for guaranteeing the required structural ﬂexibility in

order to adapt the different pharmacophoric moieties to multiple

receptors’ binding sites. Indeed, in agreement with the reported

structural and biochemical data on dopamine and serotonin

receptors, we hypothesized that our competitive ligands bind

by placing their heteroaromatic moiety (namely the “head”) into

For in vitro and in vivo testing, compounds 5a-m were used

as mono-, di-, or trihydrochloride salts.

The binding afﬁnities for 5-HT_1A, 5-HT_2C, 5-HT_2A, D₁, D₂,

and D₃receptors of compounds 5a-m, together with clozapine,

olanzapine, and haloperidol, are given in Table 2. Furthermore,

Table 2 lists the afﬁnity of the tested compounds for hERG

channels. Figure 3 summarizes the effects of 5i in animal models

sensitive to mesolimbic mediated antipsychotic activity (meth-

amphetamine (MAMP)) and phencyclidine (PCP) induced

hyperactivity) and striatal mediated side effects (catalepsy).

Figure 4 reports the effect of 5i on Fos protein induction in the

nucleus accumbens shell subregion, and core subregion and in

the dorsolateral part of the rostral striatum. Table 4 reports a

summary of the behavioral effects of 5c, 5h, 5i, 5l, and the

reference compounds.

the “relevant aromatic pocket”^22cformed by the motif W_i, F_i+3

,

F_i+4on the TM6 helix, which is conserved throughout the

biogenic amine receptors (Table 3, Figure 1).^35,38At the same

time, the H-bonding group of our compounds is thought to

interact with the equally conserved Ser/Thr residues on TM5

helix, reported to be important for receptor interaction/activation

as well as for ligand speciﬁcity (see the discussion below for

1. Rational Design, Binding Studies, and Structure-

Activity Relationships (SARs). Design Strategy. The challenge

in designing new potential antipsychotics is to balance, in a

single molecular structure, the required activity and selectivity

toward speciﬁc multiple receptor subtypes (multireceptor afﬁnity

156 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

Figure 1. Binding sites (A: longitudinal view; C: transversal view) of 5-HTR and DAR 3D models. The inactive state (template: ꢀ₂adrenergic

receptor; PDB code: 2RH1) of D₂R (yellow), D₃R (white), 5-HT_2AR (cyan), and 5-HT_2CR (green), and the active state (template: bovine rhodopsine;

PDB code: 2I37) of 5-HT_1AR (orange) are superimposed by CR atoms. Key residues on TM3, TM5, TM6, and TM7 are displayed. AA involved

in subtype selectivity are highlighted by white circles. Serine residues on TM5 are labeled as discussed in the main text. Longitudinal (B) and

transversal (D) view of 5b (magenta), 5j (yellow), 5k (orange), and 5l (green) AM1 global minimum conformers, superimposed by ﬁtting their

pharmacophoric moieties (i.e., “head” ring centroid; protonated piperazine nitrogen hydrogen; H-bond donor group; “tail” ring centroid). The

hypothesized interactions with TM3, TM5, TM6, and TM7 receptor helices are evidenced.

all the considered receptors have in common a further Trp

residue on TM4 and TM7 helices (dopamine subtypes also on

TM2), lining the substrate binding site (Figure 1A,C and Table

3). Mutagenesis studies in the 5-HT_2AR also support an

interaction of serotonergic ligands with the conserved Trp on

TM7.⁴¹However, crucial changes in the AA composition of

dopamine and serotonin binding sites determine the molecular

bases for substrate and ligand selectivity (Figure 1 and Table

3). In particular, residues that are critical for pharmacological

speciﬁcity are often one turn of helix away from the critical

“conserved” residues.^35,38

Hence, taking into account the structural homology among

our dopamine and serotonin receptor models, as well as the

structural comparison with other known dopamine/serotonin

competitive ligands, we rationally modiﬁed the electronic and

conformational features of our ligands on the basis of our

experience in the ﬁeld of antipsychotics²²and in the develop-

ment of selective D₃R ligands.²³Accordingly, we maintained

the 4-methylene linker in the amide series of compounds,

Figure 2. Illustrated representation of the central pore cavity of hERG

because it was previously demonstrated to be the optimal spacer

channels according to the model reported by Stansfeld et al.⁴⁶The

for high D₃R afﬁnity,⁴²while we tested 4- and 5-methylene

central pore helix P and the S6 helix are indicated by orange and blue

linkers in the ether series (5j-l, Table 1). However, we varied

brackets, respectively. Key binding residues for blockers identiﬁed by

the chemical-physical parameters of: (i) the heteroaromatic

system at the “head”, (ii) the functional group (X) connecting

the “head” to the methylene linker, (iii) the phenyl ring at the

“tail”, by modifying the nature, position, and number of

substituents (Chart 2 and Table 1).

Following this rationale, we were able to obtain the optimal

afﬁnity proﬁle for entering into further in vivo experiments.

Dopamine Receptors. According to dopamine receptors’

classiﬁcation (D₁-like and D₂-like subfamilies), dopamine D₁R

showed the lowest degree of sequence homology with both D₂R

and D₃R, and with 5-HTRs, presenting relevant AA changes in

mutagenesis studies are evidenced.

details).^31,40Finally, the Asp residue on TM3, present in all

GPCRs amine receptors, is likely to interact with the protonated

piperazine nitrogen of our ligands (Table 1 and 3, Figure 1).

Moreover, because the binding sites of both dopamine and

serotonin contain additional aromatic pockets, modiﬁcation of

the chemical-physical features of the “tail” of our compounds

could drive the molecule to assume alternative binding modes

in the active site, adapting (or not) to different receptor subtypes.

Indeed, in addition to the above-mentioned Trp residue on TM6,

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 157

Figure 3. Effect of compound 5i on spontaneous exploratory locomotor activity (A), MAMP (B), and PCP-induced hyperactivity (C). (A) Compound

5i caused a dose-dependent reduction in spontaneous exploratory locomotor activity, marginally reducing activity at 3 mg/kg (p < 0.05) and

potently reducing activity at 10 mg/kg (p < 0.001). (B) Compound 5i dose-dependently reduced MAMP-induced hyperactivity in doses of 3 and

10 mg/kg when compared to VEH + MAMP (p < 0.01 and p < 0.001 respectively). (C) Likewise, PCP-induced hyperactivity was also signiﬁcantly

reduced by compound 5i, reaching statistical signiﬁcance at 3 and 10 mg/kg (p < 0.001 for both) when compared to VEH + PCP. n ) 7 pr dose

group. Results are expressed as means ( SEM of distance traveled. Statistical evaluation was performed by two-way ANOVA followed by Tukey

test for multiple comparisons. ###: p < 0.001 versus vehicle treatment; **: p < 0.01 versus MAMP/PCP treatment; ***: p < 0.001 versus MAMP/

PCP treatment.

Figure 4. Effect of a single intraperitoneal dose of 5i on Fos induction 60 min after the treatment in the nucleus accumbens shell subregion (A),

the nucleus accumbens core subregion (B), and the dorsolateral part of the rostral striatum (C). The results are expressed as means ( SEM of the

number of Fos-immunoreactive cells/mm²in all regions. *P < 0.05, **P < 0.01, ***P < 0.001 when compared to respective vehicle-treated group.

Figure 5. Outline of 5i multireceptor afﬁnity proﬁle and its signiﬁcance for optimizing the atypical antipsychotic proﬁle.

TM1, TM2, TM5, TM6, and TM7 helices involved in

ligand-receptor interaction/activation (Table 3).^35,40Accord-

ingly, we easily designed arylpiperazines devoid of D₁R afﬁnity

(Table 2). However, despite the high degree of identity between

D₂R and D₃R active sites, we were able to selectively modulate

D₂R occupancy in our series of compounds.

The D₂R and D₃R substrate binding site both present on TM2,

TM4, and TM7 helices, an aromatic residue placed one turn of

helix away (i.e., at the i+3/i+4 position) with respect to the

conserved Trp residue, thus forming additional aromatic pockets

(Table 3). Nevertheless, on D₃R TM7 helix, there is a Thr

residue, not conserved in D₂R, just before the conserved Trp

158 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

Table 4. Summary of Behavioral Effects of 5c, 5h, 5i, and 5l

Compared to the Effects of 18, 19, Aripiprazole, Clozapine, and

Haloperidol Tested Under Identical Conditions

maintaining high D₃R afﬁnity through the whole series of

compounds (Table 2).

The crucial role in D₂R/D₃R selectivity played by the

orientation of H-bond donor/acceptor group(s) (supposed to

interact with the serine stretch S_i, S_i+1, S_i+4on TM5) with respect

to the extra volume occupied by the heteroaromatic ring

(supposed to contact TM6) (Figure 1) was already observed by

us,²³and it is conﬁrmed by the afﬁnity proﬁle of the newly

designed compounds (Table 2). In fact, when an ether function

(X ) O) is replaced by an amide bond (X ) CONH) (Charts

1, 2 and Table 1), D₂R afﬁnity dramatically decreased and the

compound become extremely selective toward D₃R (5k vs 5b,

Table 2). Indeed, in addition to the above-mentioned Thr residue

on TM7, D₂R and D₃R present further differences in the AA

composition of their binding site. In particular, in the hydro-

phobic cleft between TM5 and TM6, bulky Ile residues are

present in D₂R (I195(TM5) and I391(TM6), human numbering;

Figure 1, Table 3), while at the D₃R level, these isoleucines are

replaced by the less hindered Val residues. This may account

for the potency and selectivity of the designed compounds

toward D₃R. Accordingly, the different amino acid composition

of the regions shaped between TM5/TM6 and TM6/TM7 helixes

was exploited by us to obtain the desired D₃R/D₂R selectivity

proﬁle by modifying the “head” of title compounds, (Chart 2,

Tables 1-3, Figure 1A-D). In particular, the electronic partial

charge distribution of 5j-l, in their AM1 global minimum

conformer (Figure 1B,D), evidenced the bridging oxygen (full

optimized AM1 partial charge ) -0.26) and the heteroaromatic

nitrogen (full optimized AM1 partial charge ∼-0.20) lone pairs

as putative H-bond acceptor groups. Our computational analysis

also highlighted that the high ﬂexibility due to the ether linker

(i.e., at least ﬁve sp³atoms), together with the formation of an

H-bond interaction between the electron-dense ether oxygen

(conjugated to an heteroaromatic ring) and the protonated

piperazine nitrogen (Figure 1B,D), strongly stabilized a “but-

terﬂy-shaped” conformation, ﬁtting the D₂R pharmacophoric

requirements reported for tricyclic D₂R antagonists.^22cThus,

the ether derivatives may be able to: (i) interact with the serine

stretch S_i, S_i+1, S_i+4on TM5, (ii) accommodate the heteroaro-

matic “head” in the region between TM5 and TM6 establishing

favorable interactions with the relevant aromatic pocket on TM6

(W_i, F_i+3, F_i+4motif), (iii) interact with the Asp residue on TM3,

and (iv) place the “tail” in the additional aromatic pocket on

TM7 (Figure 1). Because the ﬂexibility of the alkyl tether

determines the disposition of the pharmacophoric moieties, it

plays a key role in dopamine receptors (DAR) afﬁnity.²³In the

present series, this is conﬁrmed by the ether derivative 5l (5l

vs 5k, Figure 1B,D, Table 1, 2). However, when an amide bond

is in place (5b), our calculations revealed the occurrence of

intramolecular H-bonds between: (i) the amide hydrogen and

the isoquinoline nitrogen, and (ii) the carbonyl oxygen and the

protonated piperazine nitrogen. Accordingly, we calculated a

strong conformational preference for this family of “H-bonded”

conformers. In consequence, the amide derivatives are charac-

terized by a different orientation of the heteroaromatic “head”

with respect to the H-bonding groups (Chart 2, Figure 1B,D).

Indeed, in the case of the amide bridged analogues, the carbonyl

oxygen (full optimized AM1 partial charge ) -0.46), the

aromatic heteroatom (full optimized AM1 partial charge ∼-0.15),

and the amide hydrogen (partial charge ) 0.26) represent the

most probable H-bond donor/acceptor to interact with the above-

mentioned serine stretch on TM5 helix (Figure 1). To establish

this latter interaction and, at the same time, to accommodate

the heteroaromatic “head” in the “relevant aromatic pocket” on

expl. locom.^aMAMP locom.^bPCP locom.^c

ED₅₀

compd

(mg/kg)

catalepsy

5c

5h

5i

5l

18

19

14

>30

5.5

4.2

3.7

2.1

6.7

>3

2.6

>30

>10

>30

0.2

Aripiprazole

Clozapine

0.6

2.4

<0.3

2.5

>30

0.19

1.1

Haloperidol

0.08

0.06

0.09

^aExpl. Locom.: exploratory locomotor activity; ^bMAMP locom.:

methamphetamine-induced locomotor activity; ^cPCP locom.: PCP-induced

locomotor activity.

Chart 2. General Structure of Title Compounds

Chart 3. Structure and Binding Proﬁle of Compounds 18

and 19

residue (T369, human D₃numbering; Table 3). This could offer

a further H-bond interaction site (i.e., besides the Ser stretch

S_i, S_i+1, S_i+4on TM5) close to the relevant aromatic pocket on

TM6 and the Asp residue on TM3 (Figure 1A,C). Consequently,

the arylpiperazine skeleton is suitable for high D₃R afﬁnity,

likely placing (i) the protonated N in contact with the Asp

residue on TM3, (ii) the heteroaromatic “head” (Chart 2) on

the relevant aromatic pocket on TM6, establishing H-bond

interactions with TM5 or TM7, and (iii) the aromatic “tail” in

one of the additional aromatic pockets on TM2, TM4, and TM7

(Figure 1). It cannot be excluded that signiﬁcant structural

changes in the molecule, such as the introduction of further

H-bond donor/acceptor groups at the “tail”, could drive this latter

into the relevant aromatic pocket on TM6, leading to putative

additional binding modes.^22cTo this end, we succeeded in

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 159

TM6, the molecule should project toward the above-mentioned

hydrophobic region between TM5 and TM6, where bulky Ile

residues are present in D₂R (Figure 1, Table 3). This may

account for the very low afﬁnity of 5b vs 5k/5l. According to

these observations, the conformationally constrained analogues

5f and 5h, lacking the H-bond acceptor group (NH), proved to

be more potent than 5a and 5b on D₂R, although the potency

on D₃R was comparable (Table 2).

It could also allow in 5-HTR subtypes alternative binding modes

for the arylpiperazine-like skeleton.

The intrinsic activity showed by compound 5i (see next

paragraph) indicated an agonist like effect on 5-HT_1AR only.

Interestingly, in the activated state of GPCRs (Figure 1 protein

orange), TM5 turns clockwise, shifting one residue away.

Although in 5-HT_1AR S_i+1is missing, the clockwise shifting

during the receptor activation pushes the S_iserine residue

(conserved in 5-HT_1AR) to occupy the S_i+1serine position of

the inactivated state. The position of the S_i+1serine residue of

the inactivated state now corresponds to the position of the S_i

serine residue of the activated state (Figure 1A,C). Thus 5i binds

5-HT_1AR in the activated state. This observation may explain

the intrinsic activity of 5i at 5-HT_1AR.

The higher tolerance of 5-HT_2AR with respect to D₂R has

already been evidenced and exploited by us in the design of

new atypical antipsychotics based on a tricyclic scaffold.²²It

is noteworthy that the pharmacophoric requirements reported

to be responsible for D₂R/5-HT_2AR selectivity in the tricyclic

series, such as the distance between the aromatic ring binding

to the relevant aromatic pocket on TM6 and the protonated

nitrogen binding to TM3, are similarly related to the selectivity

proﬁle of our new arylpiperazine-like derivatives (e.g., 5b,

distance ) 7.42 Å, K_iD2> 10000 nM, K_i5-HT2A) 41 nM vs 5j,

distance ) 5.63 Å, K_iD2) 63 nM, K_i5-HT2A) 127 nM).

The analogues 5b-e, 5h,i, and 5l,m were also tested on

5-HT_2CR. Of these, only the indole-based ligand 5h, character-

ized by the presence of alkyl-substituted nitrogens at the “head”

of the structure (Table 1), showed a signiﬁcant afﬁnity (K_i)

18.4 nM, Table 2). Accordingly, in 5-HT_2CR only S_i+1is

conserved, while S_iand S_i+4are replaced by a Gly and an Ala

residue, respectively. Moreover, bulky hydrophobic residues are

present around S_i+1on TM5 and TM6 (Table 3, Figure 1).

Biochemical studies^4,34,46demonstrated that mutation of these

residues causes a substantial alteration of ligand speciﬁcity and

selectivity.

Once we had selected the amidic function at the “head” (X

group, Chart 2) to get the best afﬁnity pattern at D₂R and D₃R,

we explored the inﬂuence of speciﬁc substitutions at the “tail”

level in order to obtain the desired pharmacological proﬁle. On

the basis of our previous ﬁndings,²³different electron withdraw-

ing groups were introduced on the phenyl ring at the “tail” to

lower D₂R afﬁnity (Table 1). Among the arylpiperazines tested

in the present series, the 2,3-dichlorophenyl piperazine (5b, 5g,

5h, 5j-l) provided analogues with signiﬁcant D₂R occupancy

and the best D₃R afﬁnity. However, they differed, as predicted,

in the D₂R/D₃R afﬁnity ratio by 1-5 orders of magnitude (Table

2). By shifting the meta-chlorine atom at the para position, the

afﬁnity decreased on both subtypes, although this effect was

more pronounced on D₃R than on D₂R (5h vs 5g, Table 2).

Contrary to the trend previously observed for D₂R on chlorine-

substituted arylpiperazines,²³the introduction of electron-

withdrawing substituents at the meta position of the phenyl ring

increased D₃R afﬁnity (5a vs 5i, Table 2). On these bases,

compound 5i, bearing a methyl group at the meta position,

despite showing a decreased D₃R receptor afﬁnity with respect

to 5a, still maintained a great selectivity proﬁle, exhibiting, in

addition, low afﬁnity for hERG channel (see below).

Finally, in agreement with our design hypothesis, the D₃R

binding site can tolerate signiﬁcant structural changes. In

particular, according to the presence of T198 on TM7, which

may allow alternative binding modes, introducing further

H-bonding groups at the “tail” of the arylpiperazine skeleton

(i.e., compounds 5d, 5e, and 5m) we maintained good D₃R

afﬁnity and D₂R/D₃R selectivity ratios.

In conclusion, we based our design strategy on the differences

among receptor subtypes’ composition of TM5/6/7 helices. We

have thus been able to design compounds with weak afﬁnity

for 5-HT_2CR while maintaining the desired afﬁnity and activity

proﬁle on 5-HT_1AR and 5-HT_2AR.

Following our rationale, the optimal dopamine/serotonin

receptors’ afﬁnity balance was obtained for compounds 5b, 5c,

and 5i (Table 2) characterized by: (i) the isoquinoline (5b and

5c) or benzofurane (5i) rings as heteroaromatic systems (Table

1), (ii) the amide bond as X group, and (iii) the presence of

small electron-withdrawing (5b and 5c) and/or -donating (5i)

substituents at 2 and/or 3 phenyl ring position of the arylpip-

erazine moiety.

Serotonin Receptors. 5-HT_1AR afﬁnity is not greatly inﬂu-

enced by the position of the H-bonding group at the “head” of

the arylpiperazine skeleton (Chart 2, Table 2). Indeed, com-

pounds 5b,c, 5h-j, and 5l, characterized by different heteroaro-

matic moieties and X groups (Chart 2, Table 1), presented a

5-HT_1AR afﬁnity ranging from 5.0 to 18.0 nM (Table 2).

However, the afﬁnity of the above-cited set of compounds on

5-HT_2AR (ranging from 15.3 to 127 nM, Table 2) demonstrated

the different sensitivity of the two serotonin receptor subtypes

to structural modiﬁcations at the “head” of our arylpiperazine

scaffold. Similarly, the variation of the alkyl tether length in

the ether series differently affected serotonin receptor subtype

afﬁnity (5l vs 5k, Tables 1 and 2). In fact, these serotonin

receptor subtypes present critical AA changes in the TM5 helix

(Figure 1A,C, Table 3), which can account for the observed

Ether-a-gogo-Related Gene (hERG) KC Channels. Reduc-

ing the risk of drug-induced cardiac arrhythmia is recognized

as a major hurdle in the successful development of new drugs.

To date, investigation of hERG channel blockade has been a

signiﬁcant step along the drug discovery trajectory of the

pharmaceutical industry. The most common problem is acquired

long QT syndrome. This is caused by drugs that block human

ether-a-gogo-related-gene (hERG) KC channels, delay cardiac

repolarization, and increase the risk of torsades de pointes

arrhythmia (TdP). Indeed, efforts to predict long QT syndrome

risk have been focused on assays testing hERG channel

activities. This is because hERG channel blockade is an

important indicator of potential pro-arrhythmic liability. Drug-

induced arrhythmia by noncardiac drugs is rare. However,

commonly used medications can induce ventricular arrhythmia.

SARs.^34,43,44In particular, the DAR serine stretch (S_i, S_i+1, S_i+4

)

on TM5 is not conserved in 5-HT_1AR, where S_iis still present,

but S_i+1and S_i+4are replaced by a Thr and an Ala residue,

respectively. In contrast, in the 5-HT_2AR subtype S_iis replaced

by a Gly residue, while S_i+1and S_i+4are conserved (Table 3).

Moreover, 5-HT_1AR binding site presents an Asn residue on

TM7 (N386, human 5-HT_1AR human numbering, Table 3),

corresponding to the ꢀ-adrenergic receptor TM7 Asn residue

involved in carazolole binding.^37-39As with D₃R (reported

above), this could represent a putative additional H-bond

interaction site besides the Ser stretch on TM5 (Figure 1A,C).

160 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

These include antipsychotics, antimalarials, and antidepressants.

Although these drugs represent different therapeutic classes and

chemical structures, they share the ability to prolong the QT

interval, which is an indirect measure of ventricular-repolar-

ization time measured from the ECG and a known risk factor

for TdP. Sertindole and thioridazine were recently withdrawn

from the market for this reason.

This strategy led to the discovery of 5i, a potential atypical

antipsychotic agent characterized by low afﬁnity for hERG

(Table 2).

Following this approach, we have been able to modulate

hERG channel occupancy within the same structural class of

compounds while maintaining the desired receptor binding

proﬁle (5a vs 5b and 5i, Tables 1 and 2) and in vivo

pharmacological properties.

To date, long QT syndrome has alarmed regulatory authorities

and has been most often associated with drug-induced ventricu-

lar arrhythmia. We therefore believe that weeding out potential

hERG channel blocker activity as early as possible, preferably

in the drug-design stage, will eliminate a lot of safety concerns

that have previously accompanied the development of numerous

drugs, especially antipsychotics. To this end, we decided to

investigate hERG channel interaction at the earliest stages of

the drug design and discovery process. Starting from the

published protein-based and ligand-based pharmacophore mod-

els (hERG cartoon at Figure 2), we exploited our experience in

drug design and, consequently, developed our own strategy for

the structural class of compounds under investigation.

Miscellaneous Receptor Binding and 5i Intrinsic Activ-

ity at Selected Receptors. On the basis of its unique binding

proﬁle, 5i was therefore tested against a panel of other receptors

including histamine H₁receptors (K_i) 15 nM), central

imidazole I₂receptors (K_i) 138 nM) and adrenergic receptors

R₁and R₂(K_iof 6 and 43 nM, respectively). The afﬁnity for

alpha receptors is a critical aspect of clozapine/olanzapine

binding proﬁle. The R₂or 5-HT_1AR⁵⁴occupancy might explain

the marked increase in dopamine output in the prefrontal cortex

induced by clozapine, which is beneﬁcial to cognitive functions.

The R₂afﬁnity of 5i might enhance its clinical efﬁcacy, while

its R₁occupancy might protect against dopamine deﬁcits.

The intrinsic activity of compound 5i, at D_2L, D_2S, and D₃

receptors was determined in [³⁵S]-GTPγS binding assays, at

5-HT_2AR was determined in rat aortic ring contraction model,

and at 5-HT_1AR in guinea pig ileum relaxation model. Com-

pound 5i behaves as antagonist at D_2Lreceptors (IC₅₀) 584

Scanning alanine mutagenesis studies have identiﬁed key

binding-sensitive residues for a growing list of known hERG

blockers in the inner helix (S6) and in the loop connecting S6

to the pore helix, close to the potassium selectivity ﬁlter.^47-50

In particular, the binding site of drugs blocking hERG func-

tionality is contained within the central pore cavity of the pore

domain, located below the selectivity ﬁlter and ﬂanked by the

four S6 helices of the tetrameric channel. Ligand-binding

sensitivity to Ala mutations has been demonstrated for (i) T623,

S624, and V625 residues, at the base of the selectivity ﬁlter

and for (ii) G648, Y652, F656, and V659 residues, all of which

lie on the same face of S6 helix (i, i + 4 positions) (Figure 2).

nM), at D_2Sreceptors (IC₅₀) 1240 nM), and at D₃R (IC₅₀

)

143 nM). On 5-HT_2AR, 5i displayed antagonist-like effects (IC₅₀

) 230 nM), while on 5-HT_1AR displayed agonist like effects

(IC₅₀) 180 nM).

Taking into account the great potency of 5i on serotonin,

dopamine, and adrenergic receptors, this compound, together

with 5c,h,k, was selected as a promising atypical antipsychotic

agent and subjected to in vivo pharmacological characterization.

All potent hERG blockers are highly sensitive to Y652A and

F656A mutations. The presence of these two aromatic amino

acids (Y652 and F656) in S6 of hERG is unique to the ether-

a-go-go family of K⁺channels. Indeed, despite the lack of

negative charged residue on the hERG S6 inner helix, the

binding of the cationic species is guaranteed by the overall

negative electrostatic ﬁeld inside the hERG cavity. This is due

to Tyr and Phe side chains lining the inner pore in a 4-fold

symmetry. However, the tetrameric assembling of the S6 and

pore helices suggests the possibility of alternative binding modes

for blockers. This is also supported by data obtained from

mutagenesis studies.⁵¹Therefore, hERG pore structure can

account for the observed unintentional binding of biogene amine

competitive agonists/antagonists, such as antipsychotics. Ac-

cordingly, published hERG blockers pharmacophores^52,53share

similar features with the dopamine/serotonin pharmacophores,

consisting of two or more hydrophobic/aromatic groups sur-

rounding a basic center. Thus, it is not surprising that our

arylpiperazine derivatives showed blocking properties on hERG,

although at high nanomolar concentrations (Table 2). Our

strategy, to lower hERG afﬁnity to the micromolar level, was

based on the principle that, although hERG channels and

dopamine/serotonin receptors share a similar and highly polar-

ized binding site for cationic species (reﬂected in the similarity

of the predicted hERG blockers and antipsychotics pharma-

cophores), the polarization of the endogenous ligand and,

consequently, the polarization of the active site itself, must differ.

Accordingly, once we had established the key pharmacophoric

features necessary for obtaining the desired receptor binding

proﬁle, we tried to modulate hERG afﬁnity by changing the

polarization of the phenyl group at the “tail” of the compounds.

2. Behavioral Studies. 2.1. Can Selective D₃R Blockade

Relieve Positive Symptoms? To validate the approach herein

proposed for the treatment of schizophrenia, the ﬁrst question

to be addressed was whether selective D₃R occupancy alone

relieves positive symptoms of schizophrenia. Previously, we

reported the synthesis of the arylpiperazines 18 and 19 (Chart

3) designed to investigate the role of D₂R/D₃R on cocaine-

seeking behavior.²³Arylpiperazines 18 and 19 proved to be two

of the most potent and selective D₃R ligands showing partial

agonist properties (18) or full antagonism (19) toward D₃R.²³

Both compounds were used in the present study as pharmaco-

logical tools to investigate the role played by D₃R occupancy

on schizophrenia models. They did not show signiﬁcant

behavioral effect in the animal models herein described.

Together with the arylpiperazine aripiprazole (4), an analo-

gous arylpiperazine characterized by high potency against D₂R,

D₃R, 5-HT_1AR, and 5-HT_2AR, compounds 18 and 19 were tested

in the PCP-induced hyperactivity animal model for psychosis.

While aripiprazole showed a dose-dependent reduction of

hyperactivity (Table 4), compounds 18 and 19 did not signiﬁ-

cantly reduce hyperactivity (ED₅₀> 30 mg/kg, Table 4),

indicating that D₃R occupancy alone lacks efﬁcacy in animal

models predictive for antipsychotic properties. We concluded

that D₃R selective ligands may not relieve positive symptoms

of schizophrenia. Accordingly, we designed arylpiperazines

capable of interacting with D₃R, 5-HT_1AR, and 5-HT_2AR.

2.2. Behavioral Effects of a Selected Set of Novel Arylpip-

erazines. An initial behavioral screening was performed on

compounds 5c, 5h, 5i, and 5l, selected on the basis of their

multireceptor afﬁnity proﬁle. Clinically, the term “atypical

antipsychotic” has been used for drugs, like clozapine, that

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 161

relieve positive symptoms at doses that do not cause side effects

such as prolactin increase and extrapyramidal effects. Preclini-

cally, this translates into drugs that show effect in models of

psychosis at doses that do not cause dopamine D₂R related side

effects. In this study, side-effect liability was evaluated by the

horizontal bar test, which is very sensitive for catalepsy induced

by dopamine D₂R blockade. Antipsychotic potential of the

compounds was assessed in mice rendered hyperactive by

MAMP and PCP. Both tests are sensitive to mesolimbic

mediated antipsychotic activity. Meth- (or D-) amphetamine-

induced hyperactivity are potently reduced by D₂R antagonists,

such as haloperidol, whereas PCP-induced hyperactivity has

been reported to be more sensitive to atypical antipsychotic

compounds when compared to typical antipsychotics.⁵⁵

dependently reduced hyperactivity induced by PCP with ED₅₀

values at 6.7 and 2.6 mg/kg, respectively. Sedative effects were

observed at approximately 2-fold higher ED₅₀values in the

exploratory locomotor activity test (13.8 mg/kg for compound

5c and 5.5 mg/kg for compound 5i). The atypical antipsychotic

compounds aripiprazole and clozapine both dose-dependently

reduced PCP-induced hyperactivity. Aripiprazole was admin-

istered at doses of 0.1, 0.3, and 1 mg/kg and inhibited PCP-

induced hyperactivity with an ED₅₀value of 0.2 mg/kg.

Clozapine, dosed in 0.3, 1, and 3 mg/kg, reduced PCP-induced

hyperactivity with an ED₅₀value of 1.1 mg/kg. Both compounds

showed sedative effects in the exploratory locomotor activity

test with ED₅₀values approximately twice the ED₅₀values

necessary to reduce PCP-induced hyperactivity. Furthermore,

the typical antipsychotic compound haloperidol, tested at doses

of 0.01, 0.025, and 0.05 mg/kg, reduced PCP-induced hyper-

activity, reaching ED₅₀at 0.09 mg/kg. This is in the same dose

range as the ED₅₀for causing sedation in the exploratory

locomotor activity test (ED₅₀) 0.08 mg/kg). In summary,

compounds 5c and 5i both showed favorable proﬁles in MAMP-

and PCP-induced hyperactivity when compared to typical as

well as atypical antipsychotic reference compounds (Table 4,

and Figure 3 for 5i). On the basis of these results, 5i was selected

as the lead compound of the new series and was subjected to

side-effect evaluation to further characterize its atypical antip-

sychotic proﬁle.

2.3. Effects of 5c, 5h, 5i, 5l, and Reference Compounds

on Spontaneous Exploratory Locomotor Activity. In the

exploratory locomotor activity test, the mice are transferred to

new cages immediately before the test began and consequently

used the ﬁrst 30 min or so to explore their new environment.

Compounds that cause sedation or induce motor disturbances

will result in reduction in spontaneous exploratory locomotor

activity. Pretreatment with 3, 10, and 30 mg/kg of compound

5h and 5l only mildly reduced exploratory locomotor activity

with ED₅₀values >30 mg/kg (Table 4). Compounds 5c and 5i

dose-dependently reduced spontaneous locomotor activity with

ED₅₀values at 13.8 and 5.5 mg/kg, respectively (Table 4, and

Figure 3 for 5i). For comparison, aripiprazole, clozapine, and

haloperidol were tested. All three compounds reduced explor-

atory locomotor activity with ED₅₀values of 0.6, 2.4, and 0.08

mg/kg, respectively.

2.6. Comparison of Cataleptogenic Effects of Aripipra-

zole, Clozapine, Haloperidol and Compound 5i. Compound

5i did not cause catalepsy in the horizontal bar test at doses up

to 30 mg/kg (Table 4). As expected, the typical antipsychotic

compound haloperidol dose-dependently resulted in catalepsy

with an ED₅₀value of 0.19 mg/kg, while the atypical antipsy-

chotic compounds aripiprazole and clozapine did not induce

catalepsy at doses up to 30 mg/kg.

2.4. Effects of 5c, 5h, 5i, and 5l on Methamphetamine-

Induced Locomotor Activity. As shown in Table 4 and Figure

3 for 5i, pretreatment with 1, 3, and 10 mg/kg of compound 5l

did not reduce MAMP-induced hyperactivity when compared

to MAMP alone (ED₅₀value >10 mg/kg). In contrast, pretreat-

ment with 1, 3, and 10 mg/kg of compounds 5c, 5h, and 5i all

caused a dose-related reduction in MAMP-induced hyperactivity,

with ED₅₀values of 4.2, 3.7, and 2.1 mg/kg, respectively. These

effects cannot be accounted for by drug-induced sedation

because sedative effects were only observed in the exploratory

locomotor activity test at ED₅₀values ranging from 2.6- to 8.1-

fold higher doses (see above). For comparison, aripiprazole,

clozapine, and haloperidol were also tested. All three compounds

dose-dependently reduced MAMP-induced hyperactivity. Arip-

iprazole reduced MAMP-induced hyperactivity with ED₅₀< 0.3

mg/kg, i.e., at least 2-fold lower than the ED₅₀value for sedative

effects in the exploratory locomotor activity test (ED₅₀) 0.6

mg/kg). Clozapine reduced MAMP-induced hyperactivity with

ED₅₀) 2.5 mg/kg, which was in the same dose range as needed

2.7. Mesocorticolimbic Selectivity and Atypical Antipsy-

chotic Proﬁle of 5i. The immunochemistry of the Fos protein

has been shown to be useful in mapping functional pathways

in the central nervous system and especially in identifying brain

areas that are targets for antipsychotics. Moreover, the ability

of these drugs to increase Fos protein expression in the striatal

complex has been considered useful in discriminating between

typical and atypical antipsychotic compounds.⁵⁶In contrast to

haloperidol, but like most atypical antipsychotics, 5i did not

affect Fos immunoreactivity in the regions implicated in the

control of extrapyramidal motor function, such as the dorso-

lateral striatum. However, like clozapine, it markedly enhanced

the expression of Fos protein in the mesocorticolimbic regions,

which are involved in the control of affective and motivational

behaviors, i.e., the shell subregion of the nucleus accumbens

(Figure 4). In this respect, 5i clearly behaves as an atypical

antipsychotic. However, it is worth noting that not all of the

atypical antipsychotics, which similarly increase Fos expression

in limbic areas, can enhance this expression in the medial

prefrontal cortex too.^56a

to cause sedation in the exploratory locomotor test (ED₅₀

)

2.4 mg/kg). Haloperidol reduced MAMP-induced hyperactivity

with ED₅₀) 0.06 mg/kg, also in the same dose range as needed

to cause sedation (ED₅₀) 0.08 in exploratory locomotor

activity). Accordingly, compounds 5c, 5h, and 5i showed a

favorable proﬁle when compared to both atypical and typical

reference compounds in this test, with a 2-8 fold separation

between doses causing sedation and doses reducing MAMP-

induced hyperactivity.

A single systemic injection of 5i produced a dose-dependent

and site-selective induction of the transcription factor Fos in

the limbic striatum (Figure 4). There were clear topographical

differences in the induction of Fos in different striatal areas.

While 5i markedly increased the number of Fos-immunoreactive

cells in the nucleus accumbens (ACC) ACCshell at a dose of 3

mg/kg, higher doses were necessary to produce an induction in

the ACCcore and an even higher in the dorsolateral (DL)

DLstriatum (Figure 4). In the ACCcore, 5i had no effect at a

low dose, but at 10 mg/kg and higher, the compound induced

2.5. Effects of Compounds 5c, 5h, 5i, and 5l on PCP-

Induced Locomotor Activity. Pretreatment with 0.3, 1, and 3

mg/kg of compound 5h and 1, 3, and 10 mg/kg of compound

5l was not able to reduce PCP-induced increase in locomotor

activity (ED₅₀> 3 and > 10 mg/kg, respectively). In contrast,

compounds 5c and 5i, both dosed in 1, 3, and 10 mg/kg, dose-

162 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

Fos in this region. Similarly, in the DLstriatum, 5i increased

the Fos induction, but only at the extremely high dose of 30

mg/kg.

clozapine. Accordingly, 5i shows an “atypical” antipsychotic

activity without liability for side effects as measured by standard

behavioral testing paradigms. This is supported by the preferred

increase in expression of Fos immunoreactive cells in the

ACCshell when compared to ACCcore and DLstriatum. More-

over, the data would suggest a therapeutic index of at least

10-15-fold between efﬁcacy (methamphetamine or PCP mod-

els) and side effects (catalepsy). These ﬁndings indirectly

illustrate the high and selective activity of 5i toward the

mesolimbic and mesocortical dopaminergic system. 5i may pave

the way for the development of a novel class of drugs for the

treatment of neuropsychiatric disorders.

Interestingly, it has been recently shown⁵⁷that the 5-HT_1AR

agonist 8-OH-DPAT is able to convert the “typical” pattern of

haloperidol on c-fos expression into a pattern resembling that

of clozapine, suggesting that the proﬁle of Fos protein expression

of 5i might be linked to partial agonist properties at 5-HT_1AR.

3. Summary of the in Vivo Characterization of the

Atypical Antipsychotic Agent 5i and Its Therapeutic Poten-

tial. As already discussed, 5i (1, 3, 10 mg/kg, sc) caused a dose-

dependent reduction in MAMP- and PCP- induced hyperactivity

in mice, reaching ED₅₀values of 2.1 and 2.6 mg/kg, respectively.

In the exploratory locomotor activity test, sedation was only

observed with ED₅₀values approximately twice the value needed

to reduce psychostimulant-induced hyperactivity. This is in the

same range observed in this study by marketed atypical

antipsychotics such as aripiprazole and clozapine. The ratio

between ED₅₀for reducing MAMP-/PCP-induced hyperactivity

and ED₅₀for sedation ranges between 1 and 3. D₂R-related side-

effects liability was evaluated by the horizontal bar test. Neither

compound 5i nor aripiprazole nor clozapine caused catalepsy

up to the highest dose tested (30 mg/kg). In contrast, haloperidol

potently induced catalepsy with an ED₅₀value of 0.19 mg/kg.

In conclusion, 5i has an approximate 200-fold selectivity toward

the D₃R (K_iD2) 764 nM; K_iD3) 4.5 nM) and high afﬁnity for

5-HT_1AR and 5-HT_2AR with (K_i) 12 and 15 nM, respectively).

Taking this into account, 5i was found effective in psychosis

models without showing effect in the striatal-mediated side-

effect models, suggesting antipsychotic activity with no pro-

pensity for causing D₂R-related side effects.

Experimental Section

Reagents were purchased from Aldrich and were used as

received. Reaction progress was monitored by TLC using Merck

silica gel 60 F₂₅₄(0.040-0.063 mm) with detection by UV. Merck

silicagel60(0.040-0.063mm)wasusedforcolumnchromatography.

Melting points were determined in Pyrex capillary tubes using

1

an Electrothermal 8103 apparatus and are uncorrected. H NMR

and ¹³C NMR spectra were recorded on Bruker 200 MHz or Varian

300 MHz spectrometer with TMS as internal standard. Splitting

patterns are described as singlet (s), doublet (d), triplet (t), quartet

(q), and broad (br); the value of chemical shifts (δ) are given in

ppm and coupling constants (J) in Hertz (Hz).

GC-MS were performed on a Saturn 3 (Varian) or Saturn 2000

(Varian) GC-MS System using a Chrompack DB5 capillary column

(30 mm × 0.25 mm i.d.; 0.25 µm ﬁlm thickness). FAB-MS spectra

were performed using a VG 70-250S spectrometer. ESI-MS, APCI-

MS spectra were performed by an Agilent 1100 series LC/MSD

spectrometer and by LCQDeca-THERMOFINNIGAN spectrometer.

Elemental analyses were performed in a Perkin-Elmer 240C

elemental analyzer, and the results were within (0.4% of the

theoretical values, unless otherwise noted.

Yields refer to puriﬁed products and are not optimized. All

moisture-sensitive reactions were performed under argon atmo-

sphere using oven-dried glassware and anhydrous solvents. All the

organic layers were dried using anhydrous sodium sulfate.

1-(2,3-Dichlorophenyl)piperazine, 1-(3-methylphenyl)piperazine,

and 1-(3-chlorophenyl)piperazine were commercially available,

1-(2,4-dichlorophenyl)piperazine was synthesized as previously

described,²³and 1-(3-cyanophenyl)piperazine was obtained starting

from 3-cyanobromobenzene and piperazine following the same

procedure described for the synthesis of 1-(2,4-dichlorophenyl)pip-

erazine (spectroscopic data are consistent with those reported in

the literature).⁶⁰

For testing, compounds 5a-m were transformed into the

corresponding hydrochloride salts by a standard procedure.

Ethyl 1-(Cyanomethyl)-7-methoxy-1H-indol-2-carboxylate

(7). A mixture of sodium hydride (60% dispersion in mineral oil,

(849.2 mg, 21.23 mmol) and ethyl 7-methoxyindole-2-carboxylate

6 (3.1 g, 14.15 mmol) in dry DMF (15.0 mL) was stirred at room

temperature for 30 min before the addition of bromoacetonitrile

(2.0 mL, 28.3 mmol) in dry DMF (2.0 mL). The reaction mixture

was then heated to 65 °C for 30 min and stirred for further 6 h at

room temperature, and treated with ice. The separated solid was

ﬁltered and puriﬁed by means of ﬂash chromatography (33%

n-hexane in dichloromethane) to give 7 (30.1% yield) as a white

solid: mp (ethanol) 99-101 °C. ¹H NMR (CDCl₃) δ 1.41 (t, 3H, J

) 7.2 Hz), 3.99 (s, 3H), 4.40 (q, 2H, J ) 7.1 Hz), 5.96 (s, 2H),

6.80 (d, 1H, J ) 7.7 Hz), 7.10 (t, 1H, J ) 7.9 Hz), 7.25 (m, 1H),

7.33 (s, 1H). GC-MS m/z 258 (100) [M]⁺, 232, 213, 201, 187, 172,

144, 130, 114, 89.

The in vivo studies, after acute administration of 5i, clearly

showed that this compound has an antipsychotic potential. As

with all other antipsychotics tested so far, 5i induces Fos in the

nucleus accumbens shell.^58,59Furthermore, 5i displays an

atypical proﬁle, activating neurons exclusively in the limbic/

ventral striatal subdivisions rather than the sensory-motor

portions in the dorsolateral striatum.

Conclusions

In summary, starting from highly potent and selective

dopamine D₃R ligands previously described, with no overt

behavioral effects in animal model for schizophrenia, we

discovered novel arylpiperazine atypical antipsychotic agents

characterized by speciﬁc occupancy of D₃R, 5-HT_1AR, and

5-HT_2AR. Interaction with D₂R and 5-HT_2CR was minimized

in order to reduce any liability of EPS and to reduce the risk of

obesity under chronic treatment, respectively. The analogue 5i,

which has unique receptor afﬁnity properties on dopamine,

serotonin, and adrenergic receptors, was identiﬁed as a potential

atypical drug candidate. The binding proﬁle of 5i suggests a

complex interaction on the cortical receptors involved in the

regulation of the activity of prefrontal cortical cells innervated

by ventral tegmental area neurones (5-HT_2A, dopaminergic, and

R adrenergic receptor subtypes). Moreover, 5i showed a low

afﬁnity for hERG channels. 5i validated the hypothesis based

on speciﬁc interaction with D₃R, 5-HT_1AR, and 5-HT_2AR for

the discovery of innovative antipsychotic drugs (Figure 5). 5i

reduced methamphetamine as well as PCP-induced locomotor

activity without causing catalepsy (up to 30 mg/kg, sc).

Moreover, ED₅₀for reducing MAMP- and PCP-induced hyper-

activity was approximately a factor 2-fold lower than the ED₅₀

for causing sedation. This was found to be in the same range

for the two marketed atypical antipsychotics, aripiprazole and

1,2,3,4-Tetrahydro-6-methoxypyrazino[1,2-a]indole (8). A

suspension of 7 (500.0 mg, 1.93 mmol) in dry diethyl ether (Et₂O)

(20.0 mL) was added slowly to a well-stirred slurry of lithium

aluminum hydride (LiAlH₄) (293.4 mg, 7.72 mmol) in dry Et₂O

(10.0 mL). The mixture was reﬂuxed for 4 h and then poured into

ice-water. 1 N NaOH (10.0 mL) was added and the aqueous phase

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 163

was extracted with EtOAc (3 × 30 mL). The collected organic

layers were dried and evaporated. The crude product was chro-

matographed (10% methanol in chloroform) to afford 8 as a yellow

solid (40.6% yield): mp (methanol) 120-122 °C. ¹H NMR (CDCl₃)

δ 1.89 (br s, 1H), 3.26 (t, 2H, J ) 5.7 Hz), 3.90 (s, 3H), 4.17 (s,

2H), 4.47 (t, 2H, J ) 5.8 Hz), 6.14 (s, 1H), 6.58 (d, 1H, J ) 7.7

Hz), 6.97 (t, 1H, J ) 7.7 Hz), 7.14 (d, 1H, J ) 7.8 Hz). ESI-MS

m/z 405 [2 M + H]⁺, 203 (100) [M + H]⁺.

1,2,3,4-Tetrahydro-5-methoxy-ꢀ-carboline (13). 4-Methox-

ytryptamine 12 (375.0 mg, 1.97 mmol) was transformed into the

corresponding hydrochloride salt by a standard procedure. To a

solution of 4-methoxytryptamine hydrochloride (445.0 mg, 1.97

mmol) in water (50.0 mL) glyoxylic acid monohydrate (181.2 mg,

1.97 mmol) was added and the mixture was stirred under reﬂux

for 1 h. After cooling to room temperature, a 20% solution of NaOH

was added and the mixture was extracted with EtOAc (3 × 30

mL). The organic layers were dried and evaporated. The crude

product was chromatographed (CHCl₃/MeOH/NH₄OH 20:5:0.5 v/v)

to give 13 as an amorphous solid (63.2% yield). ¹H NMR (CDCl₃)

δ 1.67 (br s, 1H), 2.96 (m, 2H), 3.13 (m, 2H), 3.88 (s, 3H), 3.98

(s, 2H), 6.47 (d, 1H, J ) 7.6 Hz), 6.89 (d, 1H, J ) 8.1 Hz), 7.01

(t, 1H, J ) 7.9 Hz), 7.75 (br s, 1H).

7-(4-Bromobutyloxy)quinoline (17a). To a solution of 7-hy-

droxyquinoline 16a (500.0 mg, 3.45 mmol) in dry DMF (15.0 mL),

1,4-dibromobutane (2.22 mL, 10.34 mmol) was added and the

mixture was stirred at room temperature for 10 min. Then cesium

carbonate (1.12 g, 3.45 mmol) was added and the mixture was

heated to 65 °C for 12 h. After cooling to room temperature, methyl-

tert-butylether (MTBE) (40.0 mL) and water (30.0 mL) were added

and the mixture was extracted with MTBE (3 × 35 mL). The

collected organic layers were dried, ﬁltered, and evaporated. The

residue was chromatographed (dichloromethane) to afford 645.0

mg of pure 17a as a yellow oil (67.0% yield). ¹H NMR (CDCl₃) δ

2.00 (m, 4H), 3.44 (t, 2H, J ) 5.9 Hz), 4.09 (t, 2H, J ) 5.1 Hz),

7.16 (m, 2H), 7.35 (d, 1H, J ) 2.5 Hz), 7.63 (d, 1H, J ) 8.9 Hz),

7.99 (m, 1H), 8.77 (m, 1H). ESI-MS m/z 280, (100) [M + H]⁺,

198.

3-(4-Bromobutoxy)isoquinoline (17b). Compound 17b was

obtained starting from isoquinolin-3-ol 16b (250.0 mg, 1.72 mmol)

and following the above-described procedure for 17a. Compound

17b was obtained as a yellow oil (52.0% yield). ¹H NMR (CDCl₃)

δ 2.07 (m, 4H), 3.50 (t, 2H, J ) 6.3 Hz), 4.38 (t, 2H, J ) 5.8 Hz),

6.36 (s, 1H), 7.35 (m, 1H), 7.55 (m, 1H), 7.67 (d, 1H, J ) 8.2 Hz),

7.86 (d, 1H, J ) 8.2 Hz), 8.92 (s, 1H). ESI-MS m/z 302 [M +

Na]⁺, 280 [M + H]⁺, 199 (100).

dry acetonitrile (10.0 mL), 1-(2,3-dichlorophenyl)piperazine hy-

drochloride (87.7 mg, 0.33 mmol) and triethylamine (74.0 µL, 0.53

mmol) were added and the solution was reﬂuxed overnight under

stirring. The solvent was removed under reduced pressure, water

was added, and the mixture was extracted with dichlorometane (3

× 10 mL). The organic layers were dried and concentrated and the

crude product was chromatographed (10% MeOH in CHCl₃) to give

1

5b (60.2% yield) as a colorless oil. H NMR (CDCl₃) δ 1.67 (m,

4H), 2.47 (t, 2H, J ) 6.9 Hz), 2.64 (m, 4H), 3.06 (m, 4H), 3.57 (q,

2H, J ) 6.1 Hz), 6.92 (m, 1H), 7.11 (m, 2H), 7.70 (m, 2H), 7.98

(t, 2H, J ) 8.3 Hz), 8.33 (br s, 1H), 8.60 (s, 1H), 9.12 (s, 1H). ¹³

C

NMR (CDCl₃) δ 24.3, 27.6, 39.3, 51.3, 53.3, 58.0, 118.5, 120.1,

124.4, 127.3, 127.4, 127.5, 128.0, 128.7, 129.6, 130.9, 133.9, 136.0,

143.7, 150.9, 151.3, 164.7. ESI-MS m/z 457 [M + H]⁺, ESI-MS/

MS of [M + H]⁺285, 227 (100). Anal. (C₂₄H₂₆Cl₂N₄O) C, H, N.

N-[4-[4-(3-Chlorophenyl)piperazin-1-yl]butyl]isoquinoline-3-

carboxamide (5c). The title compound was prepared starting from

14b (190.0 mg, 0.62 mmol) and 1-(3-chlorophenyl)piperazine

hydrochloride (144.0 mg, 0.62 mmol) following the above-described

procedure for 5b. Compound 5c was obtained as a white solid

1

(50.4% yield): mp (methanol) 156-157 °C. H NMR (CDCl₃) δ

1.65 (m, 4H), 2.46 (t, 2H, J ) 6.7 Hz), 2.60 (t, 4H, J ) 4.9 Hz),

3.21 (t, 4H, J ) 5.0 Hz), 3.57 (q, 2H, J ) 6.5 Hz), 6.78 (m, 2H),

6.86 (d, 1H, J ) 1.6 Hz), 7.14 (t, 1H, J ) 8.0 Hz), 7.72 (m, 2H),

8.00 (t, 2H, J ) 8.2 Hz), 8.33 (br s, 1H), 8.61 (s, 1H), 9.14 (s,

1H). ¹³C NMR (CDCl₃) δ 24.6, 27.9, 39.5, 48.7, 53.1, 58.2, 114.0,

115.9, 119.4, 120.4, 127.8, 128.3, 128.9, 129.8, 130.2, 131.2, 135.1,

136.2, 144.0, 151.2, 152.5, 165.0. ESI-MS m/z 445 (100) [M +

Na]⁺, 423 [M + H]⁺, ESI-MS/MS of [M + H]⁺251, 227 (100).

Anal. (C₂₄H₂₇ClN₄O) C, H, N.

N-[4-(3,4-Dihydro-6-methoxypyrazino[1,2-a]1ndol-2(1H)-yl)-

butyl]isoquinoline-3-carboxamide (5d). To a suspension of 1,2,3,4-

tetrahydro-6-methoxypyrazino[1,2-a]indole (8) (120.0 mg, 0.32

mmol) and K₂CO₃(160.0 mg, 1.22 mmol) in dry acetonitrile (5.0

mL), bromo-derivative 14b (98.0 mg, 0.32 mmol) and a catalytic

amount of NaI were added and the resulting mixture was heated

under reﬂux for 18 h. Thereafter the mixture was ﬁltered and the

ﬁltrate was evaporated to dryness under reduced pressure. The

residue was suspended in water (10.0 mL) and extracted with Et₂O

(2 × 25 mL) and dichloromethane (1 × 25 mL). The combined

organic layers were evaporated under reduced pressure, and the

crude product was puriﬁed by means of ﬂash chromatography (0.5%

MeOH in CHCl₃) affording to 5d as a yellow oil (55.0% yield).

¹H NMR (CDCl₃) δ 1.75 (m, 4H), 2.60 (m, 2H), 2.89 (t, 2H, J )

5.5 Hz), 3.57 (q, 2H, J ) 6.1 Hz), 3.79 (s, 2H), 3.86 (s, 3H), 4.49

(t, 2H, J ) 5.6 Hz), 6.12 (s, 1H), 6.53 (d, 1H, J ) 7.6 Hz), 6.93

(t, 1H, J ) 7.7 Hz), 7.10 (d, 1H, J ) 7.9 Hz), 7.70 (m, 2H), 7.97

(m, 2H), 8.39 (br s, 1H), 8.59 (s, 1H), 9.05 (s, 1H). ESI-MS m/z

879 [2 M + Na]⁺, 451 [M + Na]⁺, 429 (100) [M + H]⁺. Anal.

(C₂₆H₂₈N₄O₂·¹/₂H₂O) C, H, N.

N-[4-(1,2,3,4-Tetrahydro-5-methoxy-ꢀ-carbolin-2-yl)butyl]iso-

quinoline-3-carboxamide (5e). The title compound was obtained

following the procedure described for 5d, starting from 1,2,3,4-

tetrahydro-5-methoxy-ꢀ-carboline (13) (94.0 mg, 0.55 mmol) and

14b (137.0 mg, 0.45 mmol). Compound 5e was obtained as a yellow

oil (30.0% yield). ¹H NMR (CDCl₃) δ 1.74 (m, 4H), 2.64 (m, 2H),

2.82 (m, 2H), 3.03 (m, 2H), 3.57 (m, 4H), 3.86 (s, 3H), 6.44 (d,

1H, J ) 7.5 Hz), 6.92 (m, 2H), 7.72 (m, 2H), 7.97 (m, 3H), 8.39

(br s, 1H), 8.59 (s, 1H), 9.07 (s, 1H). ESI-MS m/z 429 (100) [M +

H]⁺, 256, 227. Anal. (C₂₆H₂₈N₄O₂) C, H, N.

3-(5-Bromopentyloxy)isoquinoline (17c). Compound 17c was

obtained starting from isoquinolin-3-ol 16b (200.0 mg, 1.37 mmol)

and 1,5-dibromopentane following the above-described procedure

for 17a. Compound 17c was obtained as a yellow oil (49.8% yield).

¹H NMR (CDCl₃) δ 1.64 (m, 2H), 1.89 (m, 4H), 3.43 (t, 2H, J )

6.5 Hz), 4.34 (t, 2H, J ) 6.3 Hz), 6.97 (s, 1H), 7.33 (m, 1H), 7.54

(t, 1H, J ) 7.2 Hz), 7.66 (d, 1H, J ) 8.3 Hz), 7.85 (d, 1H, J ) 8.2

Hz), 8.92 (s, 1H). ESI-MS m/z 295 (100) [M + H]⁺, 146.

N-[4-[4-(3-Cyanophenyl)piperazin-1-yl]butyl]benzo[b]furan-

2-carboxamide (5a). To a stirred solution of N-[4-(1-bromo)bu-

tyl]benzo[b]furan-2-carboxamide (14a) (50.0 mg, 0.17 mmol) in

dry acetonitrile (3.0 mL), 1-(3-cyanophenyl)piperazine (31.7 mg,

0.17 mmol) and triethylamine (38.2 µL, 0.27 mmol) were added;

the solution was reﬂuxed for 12 h under stirring. The solvent was

removed under reduced pressure, water was added, and the mixture

was extracted with dichloromethane (3 × 10 mL). The organic

layers were dried and concentrated and the crude product was

chromatographed (10% MeOH in CHCl₃) to give 60.0 mg of 5a

N-[4-[4-(3-Cyanophenyl)piperazin-1-yl]butyl]3,4-dihydropy-

razino[1,2-a]indol-1(2H)-one (5f). The title compound was pre-

pared starting from 15 (190.0 mg, 0.59 mmol) and 1-(3-

cyanophenyl)piperazine (110.3 mg, 0.59 mmol) following the same

procedure described for compound 5a. Compound 5f was obtained

1

(90.0% yield) as colorless oil. H NMR (CDCl₃) δ 1.72 (m, 4H),

2.46 (t, 2H, J ) 6.7 Hz), 2.61 (t, 4H, J ) 4.9 Hz), 3.24 (t, 4H, J

) 5.0 Hz), 3.52 (q, 2H, J ) 6.1 Hz), 6.89 (br s, 1H), 7.09 (m, 3H),

7.36 (m, 5H), 7.66 (d, 1H, J ) 7.5 Hz). FAB-MS m/z 403 [M +

H]⁺, 147. Anal. (C₂₄H₂₆N₄O₂) C, H, N.

1

as a colorless oil (60.0% yield). H NMR (CD₃OD) δ 1.83 (m,

4H), 3.24 (m, 6H), 3.70 (m, 4H), 3.92 (m, 4H), 4.37 (m, 2H), 7.13

(m, 2H), 7.23 (m, 1H), 7.33 (m, 3H), 7.44 (m, 2H), 7.66 (m, 1H).

N-[4-[4-(2,3-Dichlorophenyl)piperazin-1-yl]butyl]isoquinoline-

3-carboxamide (5b). To a stirred solution of N-[1-(4-bromo)bu-

tyl]isoquinoline-3-carboxamide (14b) (100.0 mg, 0.33 mmol) in

ESI-MS m/z 428, (100) [M

(C₂₆H₂₉N₅O) C, H, N.

+

H]⁺, 241, 199. Anal.

164 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

N-[4-[4-(2,4-Dichlorophenyl)piperazin-1-yl]butyl]3,4-dihydro-

pyrazino[1,2-a]indol-1(2H)-one (5g). The title compound was

prepared starting from bromoderivative 15 (200.0 mg, 0.63

mmol) and 1-(2,4-dichlorophenyl)piperazine (145.0 mg, 0.63

mmol) following the same procedure described for 5a. Com-

) 7.4 Hz), 7.52 (t, 1H, J ) 7.5 Hz), 7.64 (d, 1H, J ) 8.4 Hz), 7.83

(d, 1H, J ) 8.2 Hz), 8.91 (s, 1H). ESI-MS m/z 466 [M + Na]⁺,

444 (100) [M + H]⁺, 299. Anal. (C₂₄H₂₇Cl₂N₃O) C, H, N.

3,4-Dihydro-2-[4-(3,4-dihydro-6-methoxypyrazino[1,2-a]indol-

2(1H)-yl)butyl]pyrazino[1,2-a]indol-1(2H)-one (5m). The title

compound was prepared starting from bromo-derivative 15 (47.7

mg, 0.15 mmol) and 1,2,3,4-tetrahydro-6-methoxypyrazino[1,2-

a]indole 8 (30.0 mg, 0.15 mmol) following the procedure described

for the synthesis of compound 5d. Compound 5m was obtained as

a yellow oil (62.0% yield). ¹H NMR (CDCl₃) δ 1.69 (m, 4H), 2.58

(t, 2H, J ) 6.6 Hz), 2.88 (t, 2H, J ) 5.5 Hz), 3.66 (t, 2H, J ) 6.7

Hz), 3.77 (m, 4H), 3.88 (s, 3H), 4.24 (t, 2H, J ) 5.8 Hz), 4.45 (t,

2H, J ) 5.6 Hz), 6.10 (s, 1H), 6.54 (d, 1H, J ) 7.6 Hz), 6.93 (t,

1H, J ) 7.8 Hz), 7.13 (m, 3H), 7.29 (m, 2H), 7.70 (d, 1H, J ) 7.9

Hz). ¹³C NMR (CDCl₃) δ 24.3, 25.3, 29.6, 40.2, 45.5, 45.9, 46.1,

51.2, 51.7, 55.3, 57.2, 97.1, 101.8, 106.0, 109.5, 112.9, 120.6, 122.6,

124.3, 125.9, 127.5, 129.4, 130.3, 134.5, 136.3, 147.7, 159.9. ESI-

MS m/z 907 [2 M + Na]⁺, 884 [2 M + H]⁺, 443 (100) [M + H]⁺.

Anal. (C₂₇H₃₀N₄O₂·¹/₃H₂O) C, H, N.

Molecular Modeling. Molecular modeling calculations were

performed on SGI Origin 200 8XR12000, while molecular modeling

graphics were carried out on SGI Octane 2 and Octane workstations.

Homology Modeling. Originally we choose the bovine rhodop-

sin’s crystal structures in the inactivated and partially activated states

downloaded from the PDB data bank (http://www.rcsb.org/pdb/;

PDB IDs: 1U19 and 2I37, respectively) as templates for the

modeling of the 7TMs region of human D₂R and D₃R and human

5HT_1AR, 5HT_2AR, and 5HT_2CR. Subsequently, after the publication

of the X-ray structures of the human ꢀ₂-adrenergic receptor, we

also used this receptor as template, choosing the structure with the

higher resolution³⁸(PDB ID: 2RH1, R ) 2.4 Å) for the modeling

of dopaminergic and serotoninergic receptors TMs in the partially

inactivated state. Hereafter, we will refer to a common computa-

tional procedure applied for the generation of all receptor models

in each activation state.

To be sure of correctly predicting the “7TMs” region of our

receptor models, we ﬁrst performed a secondary structure prediction

test on the template sequences (i.e., bovine rhodopsin and ꢀ₂-

adrenergic receptors) by using the Structure Prediction and Sequence

Analysis (PredictProtein) server (http://www.predictprotein.org/).

The correctness of the prediction was then veriﬁed against the X-ray

structures. On the basis of the results obtained, D₂R, D₃R, 5-HT_1AR,

5-HT_2AR, and 5-HT_2CR sequences were subjected to the same

PredictProtein server (http://www.predictprotein.org/) calculation.

Accordingly, we modeled the 7TMs regions of the dopamine and

serotonin receptors by aligning the following sequence portions of

the templates (PDB numbering). Bovine rhodopsin: TM1 ) 33-66,

TM2 ) 70-100, TM3 ) 106-141, TM4 ) 151-173, TM5 )

199-230, TM6 ) 244-278, TM7 ) 285-310; human ꢀ₂-

adrenergic receptor: TM1 ) 29-62, TM2 ) 66-96, TM3 )

102-137, TM4 ) 148-170, TM5 ) 195-225, TM6 ) 265-299,

TM7 ) 305-330, with the dopamine/serotonin receptors segments

predicted to assume an R-helix secondary structure and to have a

transmembranarian location.

1

pound 5g was obtained as a yellow oil (87.4% yield). H NMR

(CDCl₃) δ 1.68 (m, 4H), 2.51 (m, 2H), 2.66 (m, 4H), 3.05 (m,

4H), 3.65 (t, 2H, J ) 7.1 Hz), 3.78 (t, 2H, J ) 5.7 Hz), 4.28

(m, 2H), 6.94 (d, 1H, J ) 8.4 Hz), 7.16 (m, 2H), 7.31 (m, 4H),

7.71 (d, 1H, J ) 7.8 Hz). ESI-MS m/z 471, (100) [M + H]⁺,

285, 241, 199. Anal. (C₂₅H₂₈Cl₂N₄O·¹/₂H₂O) C, H, N.

N-[4-[4-(2,3-Dichlorophenyl)piperazin-1-yl]butyl]3,4-dihydro-

pyrazino[1,2-a]indol-1(2H)-one (5h). The title compound was

prepared starting from 15 (270.0 mg, 0.84 mL) and 1-(2,3-

dichlorophenyl)piperazine (193.0 mg, 0.84 mmol) following the

same procedure described for 5a. Compound 5h was obtained as a

1

white solid (78.2% yield): mp (methanol) 160-161 °C. H NMR

(CDCl₃) δ 1.65 (m, 4H), 2.46 (t, 2H, J ) 7.2 Hz), 2.62 (m, 4H),

3.06 (m, 4H), 3.64 (t, 2H, J ) 6.5 Hz), 3.78 (t, 2H, J ) 6.0 Hz),

4.26 (t, 2H, J ) 5.3 Hz), 6.93 (m, 2H), 7.14 (m, 3H), 7.31 (m,

2H), 7.70 (d, 1H, J ) 8.0 Hz). ESI-MS m/z 472 (100) [M + H]⁺,

285, 241, 199, 172, 144. Anal. (C₂₅H₂₈Cl₂N₄O) C, H, N.

N-(4-(4-(m-Tolyl)piperazin-1-yl)butyl)benzo[b]furan-2-car-

boxamide (5i). To a stirred solution of 14a (620.0 mg, 2.09 mmol)

in dry acetonitrile (30.0 mL) under argon, 1-(m-tolyl)piperazine

dihydrochloride (443.0 mg, 2.09 mmol) and triethylamine (620 µL,

4.60 mmol) were added and the solution was reﬂuxed overnight

under stirring. The solvent was removed under reduced pressure,

water was added, and the mixture was extracted with dichlo-

romethane (3 × 30 mL). The organic layers were dried and

concentrated, and the crude product was chromatographed (6%

MeOH in CHCl₃) to give 0.42 g of 5i (52.0% yield) as white solid:

1

mp (methanol) 119-120 °C. H NMR (CDCl₃) δ 1.70 (m, 4H),

2.31 (s, 3H), 2.46 (t, 2H, J ) 6.6 Hz), 2.62 (t, 4H, J ) 4.9 Hz),

3.23 (t, 4H, J ) 4.9 Hz), 3.52 (q, 2H, J ) 6.1 Hz), 6.70 (m, 3H),

7.00 (br s, 1H), 7.13 (t, 1H, J ) 4.4 Hz), 7.33 (m, 4H), 7.66 (d,

1H, J ) 7.7 Hz). ¹³C NMR (CDCl₃) δ 22.0, 24.5, 27.7, 39.4, 49.3,

53.5, 58.1, 110.5, 111.9, 113.4, 117.1, 120.9, 122.9, 123.9, 127.0,

127.9, 129.2, 139.0, 149.2, 151.5, 154.9, 159.1. ESI-MS m/z 805

(100) [2 M + Na]⁺, 414 [M + Na]⁺, 392 [M + H]⁺. Anal.

(C₂₄H₂₉N₃O₂) C, H, N.

7-[4-[4-(2,3-Dichlorophenyl)piperazin-1-yl]butyloxy]quino-

line (5j). The title compound was prepared starting from 17a (160.0

mg, 0.57 mmol) and 1-(2,3-dichlorophenyl)piperazine hydrochloride

(153.4 mg, 0.57 mmol) following the procedure described for the

synthesis of 5b. Compound 5j was obtained as a colorless oil

(62.4% yield). ¹H NMR (CDCl₃) δ 1.85 (m, 4H), 2.50 (t, 2H, J )

7.4 Hz), 2.66 (m, 4H), 3.06 (t, 4H, J ) 4.3 Hz), 4.16 (t, 2H, J )

6.0 Hz), 6.93 (m, 2H), 7.14 (m, 1H), 7.24 (m, 2H), 7.40 (d, 1H, J

) 2.3 Hz), 7.68 (d, 1H, J ) 8.9 Hz), 8.05 (d, 1H, J ) 8.0 Hz),

8.81 (m, 1H). ESI-MS m/z 430 [M + H]⁺. Anal. (C₂₃H₂₅Cl₂N₃O)

C, H, N.

3-[4-[4-(2,3-Dichlorophenyl)piperazin-1-yl]butoxy]isoquino-

line (5k). The title compound was prepared starting from 17b (250.0

mg, 0.89 mmol) and 1-(2,3-dichlorophenyl)piperazine hydrochloride

(237.0 mg, 0.89 mmol) following the procedure described for the

synthesis of 5b. Compound 5k was obtained as a white amorphous

solid (61.0% yield). ¹H NMR (CDCl₃) δ 1.81 (m, 4H), 2.52 (t, 2H,

J ) 7.3 Hz), 2.66 (m, 4H), 3.06 (m, 4H), 4.38 (t, 2H, J ) 6.3 Hz),

6.94 (m, 2H), 7.13 (m, 2H), 7.36 (m, 1H), 7.56 (t, 1H, J ) 7.3

Hz), 7.67 (d, 1H, J ) 8.2 Hz), 7.87 (d, 1H, J ) 8.1 Hz), 8.94 (s,

1H). ESI-MS m/z 452 [M + Na]⁺, 430 (100) [M + H]⁺. Anal.

(C₂₃H₂₅Cl₂N₃O) C, H, N.

Second, to be sure of properly aligning the sequences of the

templates and of the receptors to be modeled, we performed an

alignment test on the template sequences by using either Bioinfo

(http://cheminfo.u-strasbg.fr) or ClustalW (http://www.ebi.ac.uk/

Tools/clustalw2/).

The ligands, 11-cis-retinal (1U19) and carazolol (2RH1), were

removed using the Unmerge command in the Biopolymer module

of Insight2005 (Accelrys, San Diego). A.car and an.mdf ﬁle were

generated for the coordinates of the 7TMs region of each template.

Template sequences were extracted from the 3D structures by using

the Extract command in the Sequence pulldown of the Homology

module (Insight2005) and manually aligned to that of each receptor

subtype, taking into account the results obtained from: (a) Predict-

Protein (http://www.predictprotein.org/) analysis, (b) ConSeq (http://

conseq.bioinfo.tau.ac.il/) analysis, (c) ClustalW (http://www.ebi.ac.uk/

Tools/clustalw2/) multiple alignment, (d) Bioinfo (http://cheminfo.u-

strasbg.fr) multiple alignment, (e) mutagenesis studies,³⁵and (f)

3-[5-[4-(2,3-Dichlorophenyl)piperazin-1-yl]pentyloxy]isoquin-

oline (5l). The title compound was prepared starting from 17c (325.0

mg, 1.11 mmol) and 1-(2,3-dichlorophenyl)piperazine hydrochloride

(295.0 mg, 1.11 mmol) following the procedure described for

synthesis of 5b. Compound 5l was obtained as a white amorphous

solid (54.3% yield), mp (methanol) 133-134 °C. ¹H NMR (CDCl₃)

δ 1.53 (m, 4H), 1.86 (m, 2H), 2.42 (m, 2H), 2.61 (m, 4H), 3.03

(m, 4H), 4.32 (m, 2H), 6.92 (m, 2H), 7.10 (m, 2H), 7.32 (t, 1H, J

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 165

Eukaryotic Linear Motif server sequence analysis (http://elm.eu.

org/) and protein structural domains analysis,⁶¹aimed at identifying

linear functional motifs and structural domains possibly involved

in protein-protein interactions. The resulting multiple sequence

alignment is reported in the Supporting Information.

as: (a) the distance between the centroid of the phenyl ring at the

tail and the centroid of the heteroaromatic ring at the head, (b) the

orientation of the piperazine nitrogen hydrogen with respect to

the position of the aromatic rings, (c) the conformation of the

piperazine ring (i.e., chair, half-chair, and boat), (d) the dihedral

angle value between the heteroatom of the heteroaryl moiety and

the heteroatom X (amid or ether oxygen) of the linker, (e) the

rotation of the phenyl ring with respect to the piperazine ring (i.e.,

dihedral angle value), and (f) the conﬁguration of the amide group

(i.e., cis or trans).

To properly analyze the electronic properties of the compounds,

the most stable conformer of each family, within the range of 5

kcal/mol from the global minimum conformer (∆E_GM< 5 kcal/

mol), was subjected to a full geometry optimization through

semiempirical calculations, using the quantum mechanical method

AM1 in the Mopac2007 package.⁶⁵The EF (eigenvector Following

routine) algorithm of geometry optimization was used, with a

GNORM value set to 0.01. The max step size parameter was set to

0.05 to tackle the convergence problems and to properly reach a

full geometry optimization. With the same aim, the criteria for

terminating all optimizations was increased by a factor of 100 using

the keyword PRECISE. All the resulting conformers were subse-

quently ranked in different families using the same energy and

geometry criteria reported above.

The dipole moments were calculated using partial charges

obtained by the quantum mechanical method AM1 (Mopac2007)

and visualized as vector by using the Decipher module of

Insight2005.

Pharmacology. Procedures involving animals and their care were

conducted in conformity with the institutional guidelines that are

in compliance with national (D.L. n. 116, G.U. suppl. 40, 18

Febbraio 1992, Circolare no. 8, G.U. Fourteen Luglio 1994) and

international laws and policies (EEC Council Directives 86/609,

OJL 358, 1, Dec. 12, 1987; Guide for the Care and Use of laboratory

Animals, U.S. National Research Council, 1996).

The coordinates of the 7TMs for each model were assigned as

follows. Boxes were created and frozen around each sequence

segment of the model aligned with the corresponding TM of the

template protein (Homology module; Insight2005). These sequence

segments represented the putative 7TMs of the receptor models

and were deﬁned as structural conserved regions (SCR). Coordinates

were assigned using the SCR-AssignCoords command of the

Sequence pulldown. In particular, for template/model identical side

chains, the coordinates for the whole amino acid were transferred.

However, when the side chain of the model differed from that of

the template, backbone coordinates were transferred while side chain

atoms coordinates were assigned by aligning the side chain dihedral

angles with the corresponding template residue. This was to preserve

the conformation of the reference side chain as much as possible.

Eventual bumps were removed by manually changing the torsion

angles or by means of the Manual_Rotamer command of the

Residue pulldown of the Homology module. Hydrogens were added

to all amino acids considering a pH value of 7.2. The C- and N-

terminus of each helix were capped with an aldheydic and an N_sp2H₂

group, respectively, to avoid unrealistic ionic contacts.

The receptor model optimization procedure was carried out by

using the Discover3 module of Insight2005. Every structure was

progressively relaxed (Cell Multipole summation method for

nonbond interactions,⁶²ε ) 4r, CVFF force ﬁeld).⁶³First, an initial

optimization of side chain interactions was performed, keeping the

backbone atoms ﬁxed (algorithms: Steepest Descent until the

maximum rms derivative was less than 0.5 kcal/Å, Conjugate

Gradient from a derivative of 0.5 to less than 0.1 kcal/Å).

Subsequently, the backbone was relaxed through six minimization

cycles by using the above-reported combination of Steepest Descent

and Conjugate Gradient algorithms and applying different tether

force values for each minimization cycle: from an initial value of

5 kcal/Å²to a ﬁnal value of 0.1 kcal/Å². A ﬁnal energy minimization

round was performed with a residual tether force of 0.1 Kcal/ Å²

applied only to backbone CR atoms. Each energy minimization

cycle was followed by a structural check by using the Struct_Check

command of the ProStat pulldown in the Homology module to

verify the correctness of the geometry optimization procedure before

moving to the next minimization cycle. Checks included ꢁ, Ψ, ꢂ₁,

ꢂ₂, ꢂ₃, and ω dihedral angles, CR virtual torsions, and Kabsch and

Sander main chain H-bond energy evaluation. A full unconstrained

geometry optimization has been attempted but produced unrealistic

results considering that the system upon study is lacking in structural

elements (i.e., loops, lipid bilayer, water molecules, etc.).

Conformational Analysis for Compounds 5a-m. Compounds

5a-m were built using the Insight2005 Builder module. The

apparent pK_avalues were estimated using the ACD/pK_aDB version

11.00 software (Advanced Chemistry Development Inc., Toronto,

Canada). According to the resulting values (Table 2 of the

Supporting Information), in all subsequent calculations, all com-

pounds were considered protonated at pH ) 7.2 (cytoplasmic value)

on the aliphatic nitrogen of the piperazine ring. Partial charges were

assigned by using the CFF91 force ﬁeld.⁶⁴

The conformational space of compounds 5a-m was sampled

through 200 cycles of simulated annealing. An initial temperature

of 1000 K was applied to the system for 1000 fs with the aim of

surmounting torsional barriers. Successively, temperature was

linearly reduced to 200 K with a decrement of 0.5 K/fs. The

resulting structures were subjected to two protocols of energy

minimization within Insight2005 Discover module (CFF91 force

ﬁeld, Conjugate Gradient algorithm; maximum rms derivative <

0.001 kcal/Å), which differed in the dielectric constant value, i.e.

ε ) 80r and ε ) 1, in order to maximize inter- or intramolecular

interactions, respectively. All the resulting conformers were sub-

sequently ranked in different families taking into account their

conformational energy as well as some geometric parameters, such

1. In Vitro Binding Assays on GPCRs.^22a,b,23Serotonin

and Dopamine Receptors. Male CRL:CD(SD)BR-COBS rats and

male CRL:(HA) BR albino guinea pig (Charles River, Italy) were

killed by decapitation; their brains were rapidly dissected into the

various areas (rat striatum for D₁R and D₂R, rat hippocampus for

5-HT_1AR, rat cortex for 5-HT_2AR and guinea pig cortex for

5-HT_2CR) and stored at - 80 °C until assay. Binding was calculated

on rat tissue homogenates (D₁, D₂, 5-HT_2A, and 5-HT_2Creceptors)

and on Sf9 cell membranes (D_3r). Tissues were homogenized in

about 50 volumes of ice-cold Tris HCl, 50 mM, pH 7.4 (for D₁R,

D₂R, and 5-HT_2AR) using an Ultra-Turrax TP-1810 homogenizer

(2 × 20 s) and centrifuged at 50000g for 10 min at 4 °C (Beckman

J-25 centrifuge). Each pellet was resuspended in the same volume

of fresh buffer, incubated at 37 °C for 10 min, and centrifuged

again at 50000g for 10 min at 4 °C. The pellet was then washed

once by resuspension in fresh buffer and centrifuged as before. The

resulting pellets were resuspended just before the binding assay in

the appropriate incubation buffer 50 mM Tris HCl, pH 7.4,

containing 10 µM pargyline, 0.1% ascorbic acid, 120 mM NaCl, 5

mM KCl, 2 mM CaCl₂, 1 mM MgCl₂for D₁R and D₂R; 50 mM

Tris HCl, pH 7.4, containing 10 µM pargyline and 4 mM CaCl₂

for 5-HT_1AR; 50 mM Tris HCl, pH 7.7 for 5-HT_2AR; 50 mM Tris

HCl, pH 7.4, containing 10 µM pargyline, 4 mM CaCl₂; and 0.1%

ascorbic acid for 5-HT_2CR). For D_3rreceptors, Sf9 cells membranes

expressing D_3rdopamine receptors (Signal Screen) were resus-

pended just before the binding assay in 50 mM Tris HCl, pH 7.4,

containing 5 mM EDTA, 5 mM MgCl₂, 5 mM KCl, 1.5 mM CaCl₂,

120 mM NaCl.

[³H]-SCH 23390 (speciﬁc activity, 75.5 Ci/mmol; NEN) binding

to D₁R was assayed in a ﬁnal incubation volume of 0.5 mL,

consisting of 0.25 mL of membrane suspension (striatum 2.5 mg

of tissue/sample), 0.25 mL of [³H]ligand (0.4 nM) and 10 µL of

displacing agent or solvent. Nonspeciﬁc binding was obtained in

the presence of 10 µM (-)cis-ﬂupentixol.

[³H]-Spiperone (speciﬁc activity, 25 Ci/mmol; NEN) binding to

D₂R was assayed in a ﬁnal incubation volume of 1 mL, consisting

166 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

of 0.5 mL of membrane suspension (striatum 2.5 mg of tissue/

sample), 0.5 mL of [³H]ligand (0.2 nM), and 20 µL of displacing

agent or solvent. Nonspeciﬁc binding was obtained in the presence

of 100 µM (-)sulpiride.

resuspended in Tris-HCl buffer (50 mM, pH 7.4) and stored at -80

°C until use. Rabbit kidney membranes (200 µg of protein) were

incubated with 5 nM [³H]idazoxan (Amersham, 43 Ci/mmol) in

the absence or presence of a range of 10-12 concentrations of

competing ligand drug in a total volume of 400 µL of assay buffer.

To mask adrenoreceptors, 10 µM (-)-norepinephrine (in the

presence of 0.005% ascorbic acid) was added to all tubes.

Nonspeciﬁc binding was determined with 10 µM of cirazoline.

Speciﬁc binding represented about 90% of the total binding at 5

nM [³H]-idazoxan. Following equilibrium (45 min at 25 C), bound

radioactivity was separated from free by ﬁltration as described

above. Each point was performed in triplicate.

[³H]-7-OH-DPAT (speciﬁc activity, 159 Ci/mmol; Amersham)

binding to D_3rreceptors was assayed in a ﬁnal incubation volume

of 1 mL, consisting of 0.5 mL of membrane suspension (12 µg

protein/sample), 0.5 mL of [³H]ligand (0.7 nM), and 20 µL of

displacing agent or solvent. Nonspeciﬁc binding was obtained in

the presence of 1 µM dopamine.

[³H]-8-OH-DPAT (speciﬁc activity, 217 Ci/mmol; Amersham)

binding to 5-HT_1AR was assayed in a ﬁnal incubation volume of

0.5 mL, consisting of 0.25 mL of membrane suspension (5 mg

tissue/sample), 0.25 mL of [³H]ligand (1 nM), and 10 µL of

displacing agent or solvent. Nonspeciﬁc binding was obtained in

the presence of 10 µM serotonin.

Dose-inhibition curves were analyzed by the Allﬁt⁶⁷program

to obtain the concentration of unlabeled drug that caused 50%

inhibition of ligand binding. The K_ivalues were derived from the

IC₅₀values according to the method of Cheng and Prusoff.⁶⁸

2. In Vitro Assays for hERGAfﬁnity Evaluation. Plasmids.

The human ERG and KCNE1 (potassium voltage-gated channel,

Isk-related family, member 1) were subcloned into the mammalian

expression vectors pNS1n and pNS1z, respectively, to produce the

plasmid constructs pNS1n_hERG and pNS1Z-minK.

HEK 293 cells stably expressing hERG + KCNQ1 (potassium

voltage-gated channel, KQT-like subfamily, member 1). HEK 293

tissue culture cells were transfected with equal amounts of the

plasmids pNS1n_hERG and pNS1Z-minK using lipofection (Li-

pofectamin, Life Technologies). Expression of functional hERG

channels was veriﬁed by patch-clamp measurements.

[³H]-Ketanserin (speciﬁc activity, 88 Ci/mmol; Amersham)

binding to 5-HT₂R was assayed in a ﬁnal incubation volume of 1

mL, consisting of 0.5 mL of membrane suspension (5 mg of tissue/

sample), 0.5 mL of [³H]ligand (0.7 nM), and 20 µL of displacing

agent or solvent. Nonspeciﬁc binding was obtained in the presence

of 1 µM methisergide.

[³H]-Mesulergine (speciﬁc activity, 90 Ci/mmol; Amersham)

binding to 5-HT_2cR was assayed in a ﬁnal incubation volume of 1

mL, consisting of 0.5 mL of membrane suspension (30 mg of tissue/

sample), 0.5 mL of [³H]ligand (1 nM), and 20 µL of displacing

agent or solvent. Nonspeciﬁc binding was obtained in the presence

of 10 µM mesulergine.

Whole-Cell Recordings. Cells plated on coverslips were placed

in a 15 µL perfusion chamber (ﬂowrate ∼1 mL/min ) full exchange

every 1 s). All experiments were performed at room temperature

(20-22 °C) using EPC-9 patch-clamp ampliﬁers (HEKA-electron-

ics, Lambrect, Germany). Series resistances as well as capacitance

compensation were updated before each stimulus. Usually the cell

capacitance ranged from 5 to 20 pF and the series resistances were

in the range 3-6 MΩ.

Incubations (15 min at 25 °C for D₁R, 15 min at 37 °C for D₂R

and 5-HT_2AR, 30 min at 25 °C for 5-HT_1AR, 60 min at 25 °C for

D

_3rreceptors, 30 min at 37 °C for 5-HT_2CR) were stopped by rapid

ﬁltration under vacuum through GF/B (for D₁R, D₂R, 5-HT_1AR,

5-HT_2AR, and 5-HT_2CR) or GF/B presoaked with 0.3% polyethyl-

eneimine for D₃R ﬁlters which were then washed with 12 mL (4

mL × 3 times) of ice-cold buffer (50 mM Tris HCl, pH 7.4) using

a Brandel M-48R cell harvester. The radioactivity trapped on the

ﬁlters was counted in 4 mL of Ultima Gold MV (Packard) in a ꢀ

counter (Wallac) with a counting efﬁciency of 50%. For all binding

assays the radioactivity trapped on the ﬁlters was counted in 4 mL

of Ultima Gold MV (Packard) in a LKB 1214 rack beta liquid

scintillation spectrometer with a counting efﬁciency of 50%.

Histamine H₁Receptor. Histamine H₁receptor. Binding afﬁnity

was tested according to the procedure of Dini et al.⁶⁶Binding was

determined using membranes prepared from guinea pig cerebellum

with [³H]pyrilamine (0.5 nM) as radioligand.

Fitting Procedure. We assumed that the drugs (D) interacted

with the receptors (R) in the following way:

k_on

R + DSRD

k_off

This is a simple bimolecular reaction, which integrated under

nonequilibrium conditions are described by equation 1:

_on+k_off

I_t) I₀(1 - (C ⁄ C + K_i* (1 - exp^-t(C*k

⁾)))

(1)

Where I_t) current at time 1 in nA, I₀) unblocked current in

nA, C ) drug concentration in M, k_off) off-rate in s^-1, k_on) on-

rate in M^-1s^-1, K_i) k_off/k_on. At t ) ∞ eq 1 simpliﬁes to

Michaelis-Menten equation with K_i) IC₅₀.

The peak tail-current was plotted versus time. The blocker-

induced decrease in current versus time was ﬁtted to eq (1) to give

k_on*k_offand thereby k_i. Each value is the mean ( SD of 2-8

determinations and represents µM values.

Adrenergic r₁Receptors. Binding was determined using

membranes prepared from rat cerebral cortex homogenized in

phosphate buffer. The homogenate was centrifuged at 44000g for

10 min, and the pellet was suspended in phosphate buffer and

washed two times more; 500 µL of this homogenate was incubated

with 500 µL of 0.2 nM [³H]prazosin and 20 µL of test compound

for 30 min at 25 °C and then ﬁltered through a Whatman GF/B

ﬁlter (Whatman International Ltd.). Radioactivity on the ﬁlter was

measured with a liquid scintillation counter. Complete (100%)

inhibition of [³H]prazosin binding was determined in the presence

of 10 µM prazosin. Nonspeciﬁc binding was determined in the

presence of 10 µM prazosin.

Adrenergic r₂Receptors. Binding was determined using

membranes prepared from rat cerebral cortex homogenized in

phosphate buffer. The homogenate was centrifuged at 44000g for

10 min, and the pellet was suspended in phosphate buffer and

washed two times more; 500 µL of this homogenate was incubated

with 500 µL of 1 nM [³H]clonidine and 20 µL of test compound

for 30 min at 25 °C and then ﬁltered through a Whatman GF/B

ﬁlter (Whatman International Ltd.). Nonspeciﬁc binding was

determined in the presence of 10 µM clonidine.

3. Experimental in Vitro Pharmacology for Intrinsic Activ-

ity Assessment. 3.1. Cell-Based Assays. Dopamine D_2LRecep-

tors. For evaluation of [³⁵S]-GTPγS binding to D_2Lreceptors,

human CHO cells Chinese hamster ovary in DMSO (0.4%) vehicle

were resuspended in Hepes 20 mM, pH 7.4, containing 100 mM

NaCl, 1 mM EDTA, 10 mM MgCl₂, 1 mM DTT, in a glass/Teﬂon

potter. The binding assay was carried out in a ﬁnal incubation

volume of 120 µL, consisting of 100 µL of membrane suspension

(about 15 µg of protein/sample), 10 µL of [³⁵S]-GTPγS (ﬁnal

concentration 1 nM), and 10 µL of displacing agent or solvent.

Nonspeciﬁc binding was obtained in the presence of 10 µM GTPγS;

samples were preincubated for 15 min at 30 °C without [³⁵S]-

GTPγS and then for 15 min at 30 °C with [³⁵S]-GTPγS.

Imidazole I₂Receptors. Rabbit kidney was homogenized in 10

volumes of Tris-HCl buffer (50 mM, pH 7.4) and 250 mM sucrose

and centrifuged at 500g for 10 min. The supernatant was centrifuged

at 28000g for 30 min, and the resulting pellet was washed twice

with the same buffer without sucrose. The ﬁnal pellet was

Dopamine D_2SReceptors. For evaluation of [³⁵S]-GTPγS

binding to D_2Sreceptors, human CHO cells Chinese hamster ovary

in DMSO (1%) vehicle were resuspended in Hepes 20 mM, pH

7.4, containing 100 mM NaCl, 1 mM EDTA, 10 mM MgCl₂, 1

mM DTT, in a glass/Teﬂon potter. The binding assay was carried

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 167

out in a ﬁnal incubation volume of 120 µL, consisting of 100 µL

of membrane suspension (about 15 µg of protein/sample), 10 µL

of [³⁵S]- GTPγS (ﬁnal concentration 1 nM), and 10 µL of displacing

agent or solvent.

reference compounds were administered 30 min before test start at

the following doses: 5c, 5h, and 5l: 3, 10, 30 mg/kg; 5i: 1, 3, 10

mg/kg; aripiprazole: 0.1, 0.3, 1.0, 3.0 mg/kg; clozapine: 0.3, 1, 3

mg/kg; haloperidol: 0.025, 0.05, 0.1, 0.2 mg/kg.

Nonspeciﬁc binding was obtained in the presence of 10 µM

GTPγS; samples were preincubated for 15 min at 30 °C without

[³⁵S]-GTPγS and then for 30 min at 30 °C with [³⁵S]-GTPγS.

Dopamine D₃Receptors. For evaluation of [³⁵S]-GTPγS binding

to D₃receptors, human CHO-K1 cells Chinese hamster ovary in

DMSO (1%) vehicle were resuspended in Hepes 20 mM, pH 7.4,

containing 100 mM NaCl, 1 mM EDTA, 10 mM MgCl₂, 1 mM

DTT, in a glass/Teﬂon potter. The binding assay was carried out

in a ﬁnal incubation volume of 120 µL, consisting of 100 µL of

membrane suspension (about 15 µg of protein/sample), 10 µL of

[³⁵S]- GTPγS (ﬁnal concentration 1 nM), and 10 µL of displacing

agent or solvent.

Nonspeciﬁc binding was obtained in the presence of 10 µM

GTPγS; samples were preincubated for 15 min at 30 °C without

[³⁵S]-GTPγS and then for 30 min at 30 °C with [³⁵S]-GTPγS.

3.2. Tissue Assays. Serotonin 5-HT_1AReceptors. For 5-HT_1AR

intrinsic activity assessment, Duncan Hartley guinea pig ileum 325

( 25 g were taken in DMSO (0.1%) vehicle and were put in the

incubation buffer (Krebs, pH 7.4) and were incubated 5 min at 32

°C. The ﬁnal volume of the bath was brought to 10 mL. After 5

min, the quantitation method was isometric (gram changes).

Signiﬁcant criteria: agonism g50% reduction of neurogenic twitch

relative to 0.12 µM 8-OHDPAT response. Signiﬁcant criteria:

antagonism g50% inhibition of 0.12 µM 8-OHDPAT-induced

relaxation.

Serotonin 5-HT_2AReceptors. For 5-HT_2AR intrinsic activity

assessment, Wistar rat 275 ( 25 g aortic ring was taken in DMSO

(0.1%) vehicle and was put in the incubation buffer (Krebs at pH

7.4) and was incubated 5 min at 37 °C. The ﬁnal volume of the

bath was brought to 10 mL. After 5 min, the quantitation method

was isometric (gram changes). Signiﬁcant criteria: agonism g50%

contraction response relative to 3 µM 5-HT response. Signiﬁcant

criteria: antagonism g50% inhibition of 3 µM 5-HT induced

contraction.

4. Behavioral Tests. All experimental procedures carried out

in this study were in compliance with the European Communities

Council Directive of 24 November 1986 (86/609/EEC) and ap-

proved by the Danish Animal Welfare Committee, appointed by

the Danish Ministry of Justice.

Male Wistar rats (200-225 gr) or female NMRI mice (Taconic

M&B, P.O. Box 1079, DK-8680 Ry, Denmark) kept in a ventilated

closed rack (Scantainer, Scanbur Ltd., Denmark) at constant

temperature (21 °C) and humidity (60-70%) with a 7:00 a.m. light/

7:00 p.m dark cycle were used in these studies. Food (Altromin

rat pellets) and water were available ad libitum. Experiments were

conducted during daytime in the light phase. The animals were

habituated to the experimental room approximately 24 h before start

of the experiments.

5c, 5h, 5i, 5l. Aripiprazole (Sequoia Research Products Ltd.),

and Clozapine (RBI) were dissolved in 10% Tween 80 and diluted

to ﬁnal concentration by adding appropriate volume of 0.9% NaCl.

Haloperidol (Seranase, Janssen-Cilag, ampules of 5 mg/mL), S-(+)-

Methamphetamine-HCl (RBI) and Phencyclidine (PCP, synthesized

at NeuroSearch A/S) were dissolved in 0.9% NaCl. All test- and

reference compounds were administered subcutaneously (sc) 30 min

before test start in a dose volume of 10 mL/kg. MAMP and PCP

were dosed intraperitoneally (ip) and sc, respectively, and admin-

istered in a dose volume of 10 mL/kg immediately before test start.

Appropriate doses of 5c, 5h, 5i, and 5l were chosen based on initial

behavioral observations.

MAMP- and PCP-Induced Hyperactivity. MAMP- and PCP-

induced hyperactivity was measured in automated activity frames

as described above. MAMP and PCP were administered in a dose

of 2 and 4 mg/kg, respectively. Both compounds were dosed

immediately before test start. In the MAMP study, test or reference

compounds were pretreated 30 min before test start at the following

doses: 5h, 5c, 5i, and 5l: 1, 3, and 10 mg/kg; aripiprazole: 0.1, 0.3,

and 1.0 mg/kg; clozapine: 0.3, 1, and 3 mg/kg; haloperidol: 0.025,

0.05, and 0.1 mg/kg. Locomotor activity was measured for a total

of 120 min. In the PCP study, test or reference compounds were

administered 30 min before test start at the following doses: 5h:

0.3, 1, 3 mg/kg; 5c, 5i, and 5l: 1, 3, and 10 mg/kg; aripiprazole:

0.1, 0.3, and 1.0 mg/kg; clozapine: 0.3, 1.0, and 3.0 mg/kg;

haloperidol: 0.01, 0.025, and 0.05 mg/kg. Locomotor activity was

measured for 60 min.

Catalepsy. Cataleptogenic potential of the compounds was

assessed in the horizontal bar test consisting of a horizontal bar

with a diameter of 2 mm and elevated 4 cm above ground. Test or

reference compound was administered sc 30 min before the ﬁrst

catalepsy assessment. Hereafter, the degree of catalepsy was scored

every 15 min for a total of 45 min.

Statistics. Drug effects on exploratory and stimulant-induced

locomotor activity were analyzed by applying a two-way ANOVA

with drug effect and time interval as factors. Drug-induced catalepsy

was evaluated by one-way ANOVA with treatment as factor. P

values < 0.05 were considered statistically signiﬁcant.

5. Fos Induction After Acute Administration of 5i. Rats (n

) 5 pr group) were injected i.p. with a single dose of 5i (3, 10, 15,

or 30 mg/kg) or vehicle (5% Chremophor) and returned to their

home-cage. One hour after drug administration, the rat was deeply

anaesthetized with mebumal (50 mg/mL, 3 mL/kg) and perfused

transcardially with 0.1 M phosphate buffered saline (PBS; pH )

7.4) followed by ﬁxation in 4% paraformaldehyde-PBS for 10

min.⁶⁹Then 40 µm serial coronal sections were cut through the

forebrain in series of four and processed for c-Fos immunoreactivity

according to earlier report.⁶⁹Prior to the immunocytochemical steps,

the sections were rinsed for 3 × 10 min in 0.01 M PBS, incubated

for 10 min in 1% H₂O₂-PBS to block endogenous peroxidase

activity, and for a minimum of 20 min in 0.01 M PBS with 0.3%

Triton X-100 (TX), 5% swine serum, and 1% bovine serum albumin

(BSA) to block nonspeciﬁc binding sites. The sections were then

incubated at 4 °C for 24 h in the primary antiserum diluted 1:4000

in 0.01 M PBS with 0.3% Triton X-100 and 1% BSA. The primary

polyclonal antiserum (1:4000) was generated in a rabbit against

the N-terminal peptide similar to amino acids 2-17 of the rat c-Fos

protein in our laboratory and characterized previously.⁷⁰After

incubation in primary antiserum, immunoreactivity was detected

by means of the avidin-biotin method using diaminobenzidine as

chromagen as previously described.⁶⁹The number of c-Fos positive

cells were counted by means of light microscopy (20× magniﬁca-

tion) using a counting grid (500 µm × 500 µm) placed over either

the shell of nucleus accumbens (ACCshell), core of nucleus

accumbens (ACCcore), or dorsolateral part of the rostral striatum

(DLstriatum) (Figure 4) by an observer blind to treatment regimens.

The number of c-Fos positive cells was averaged from two adjacent

sections of each animal, and statistical analysis was performed on

groupmeans(SEMusingaone-wayANOVAwithNewman-Keul’s

posthoc test.

Acknowledgment. We thank MIUR-Rome, Italy (PRIN),

and NeuroSearch A/S Ballerup, Denmark, for ﬁnancial support.

Exploratory Locomotor Activity. Exploratory locomotor activ-

ity was assessed in 14 automated activity frames (TSE Home Cage

Activity Monitoring System, MoTil, TSE Technical and Scientiﬁc

Equipment GmbH, Germany) equipped with infrared photobeam

emitters and sensors. To assess drug effect on exploratory locomotor

activity, the mice were transferred to new home cages immediately

before test start and activity was measured for 30 min. Test or

Supporting Information Available: Multiple sequence align-

ment of the 7TMs region of bovine rhodopsin, human ꢀ₂-adrenergic

receptors, between them and with the 7TMs region modelled of

human D₂R, D₃R, 5-HT_1AR, 5-HT_2AR, and 5HT_2CR, percentage of

neutral and protonated form on the aliphatic nitrogen of the

168 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1

Butini et al.

Cagnotto, A.; Mennini, T. New Antipsychotic Agents with Serotonin

and Dopamine Antagonist Properties Based on a Pyrrolo[2,1-

b][1,3]benzothiazepine Structure. J. Med. Chem. 1998, 41, 3763–3772.

(b) Campiani, G.; Butini, S.; Gemma, S.; Nacci, V.; Fattorusso, C.;

Catalanotti, B.; Giorgi, G.; Cagnotto, A.; Goegan, M.; Mennini, T.;

Minetti, P.; Di Cesare, M. A.; Mastroianni, D.; Scafetta, N.; Galletti,

B.; Stasi, M. A.; Castorina, M.; Paciﬁci, L.; Ghirardi, O.; Tinti, O.;

Carminati, P. Pyrrolo[1,3]benzothiazepine atypical antipsychotic agents.

Synthesis, structure-activity relationships, molecular-modelling, and

biological studies. J. Med. Chem. 2002, 45, 344–359. (c) Campiani,

G.; Butini, S.; Fattorusso, C.; Catalanotti, B.; Gemma, S.; Nacci, V.;

Morelli, E.; Cagnotto, A.; Mereghetti, I.; Mennini, T.; Carli, M.;

Minetti, P.; Di Cesare, M. A.; Mastroianni, D.; Scafetta, N.; Galletti,

B.; Stasi, M. A.; Castorina, M.; Paciﬁci, L.; Vertechy, M.; Di Serio,

S.; Ghirardi, O.; Tinti, O.; Carminati, P. Pyrrolo[1,3]benzothiazepine-

based serotonin and dopamine receptor antagonists. Molecular model-

ing, further structure-activity relationship studies, and identiﬁcation

of novel atypical antipsychotic agents. J. Med. Chem. 2004, 47, 143–

157. (d) Campiani, G.; Butini, S.; Fattorusso, C.; Trotta, F.; Gemma,

S.; Catalanotti, B.; Nacci, V.; Fiorini, I.; Cagnotto, A.; Mereghetti, I.;

Mennini, T.; Minetti, P.; Di Cesare, M. A.; Stasi, M. A.; Di Serio, S.;

Ghirardi, O.; Tinti, O.; Carminati, P. Novel atypical antipsychotic

agents: rational design, an efﬁcient palladium-catalyzed route, and

pharmacological studies. J. Med. Chem. 2005, 48, 1705–1708.

(23) Campiani, G.; Butini, S.; Trotta, F.; Fattorusso, C.; Catalanotti, B.;

Aiello, F.; Gemma, S.; Nacci, V.; Novellino, E.; Stark, J. A.; Cagnotto,

A.; Fumagalli, E.; Carnovali, F.; Cervo, L.; Mennini, T. Synthesis

and pharmacological evaluation of potent and highly selective D₃

receptor ligands: inhibition of cocaine-seeking behavior and the role

of dopamine D₃/D₂receptors. J. Med. Chem. 2003, 46, 3822–3839.

(24) Suzuki, H.; Gyoutoku, H.; Yokoo, H.; Shinba, M.; Sato, Y.; Yamada,

H.; Murakami, Y. Unexpected formation of quinolone derivatives in

Reissert indole synthesis. Synlett 2000, 8, 1196–1199.

(25) Rajur, S. B.; Merwade, A. Y.; Flendi, S. B.; Basanagoudar, I. D.

Synthesis of 1,2,3,4-tetrahydropyrazino[1,2-a]indoles and ethyl 1-(2-

amino-ethyl)indole-2-carboxylates. Indian J. Chem. 1989, 28B, 1065–

1068.

(26) Repke, D. B., III. Synthesis of 5-methoxy-and 5-hydroxy-1,2,3,4-

tetrahydro-9H-pyrido[3,4-b]indoles. J. Heterocycl. Chem. 1982, 19,

845–848.

(27) Sugimoto, O.; Mori, M.; Moriya, K.; Tanji, K. Application of

phosphonium salts to the reactions of various kinds of amides. HelV.

Chim. Acta 2001, 84, 1112–1118.

(28) Robertson, A. A. B.; Botting, N. P. Synthesis of deuterium labelled

desulfoglucosinolates as internal standards for LC-MS analysis.

Tetrahedron 1999, 55, 13269–13284.

(29) GPCR alignments; www.bioinfo-pharma.u-strasbg.fr/gpcrdb/jss.html.

(30) Terakita, A. The opsins. Genome Biol. 2005, 6, 213.

(31) Gether, U.; Kobilka, B. K. G protein-coupled receptors. II. Mechanism

of agonist activation. J. Biol. Chem. 1998, 273, 17979–17982.

(32) Ballesteros, J. A.; Jensen, A. D.; Liapakis, G.; Rasmussen, S. G.; Shi,

L.; Gether, U.; Javitch, J. A. Activation of the beta 2-adrenergic

receptor involves disruption of an ionic lock between the cytoplasmic

ends of transmembrane segments 3 and 6. J. Biol. Chem. 2001, 276,

29171–29177.

(33) Shapiro, D. A.; Kristiansen, K.; Weiner, D. M.; Kroeze, W. K.; Roth,

B. L. Evidence for a model of agonist-induced activation of 5-hy-

droxytryptamine 2A serotonin receptors that involves the disruption

of a strong ionic interaction between helices 3 and 6. J. Biol. Chem.

2002, 277, 11441–11449.

piperazine ring at pH ) 7.2 of compounds 5a-m accordingly to

apparent pK_avalue estimated with ACD/pK_aDB, experimental

procedure for the synthesis of compounds 10-12, and elemental

analysis data. This material is available free of charge via the

Internet at http://pubs.acs.org.

References

(1) Practice guideline for the treatment of patients with schizophrenia.

American Psychiatric Association. Am. J. Psychiatry 1997, 154, 1-

63.

(2) Carlsson, A. The current status of the dopamine hypothesis of

schizophrenia. Neuropsychopharmacology 1988, 1, 179–186.

(3) Meltzer, H. Y. Clinical studies on the mechanism of action of

clozapine: the dopamine-serotonin hypothesis of schizophrenia. Psy-

chopharmacology (Berlin) 1989, 99, S18–S27.

(4) Campbell, M.; Young, P. I.; Bateman, D. N.; Smith, J. M.; Thomas,

S. H. The use of atypical antipsychotics in the management of

schizophrenia. Br. J. Clin. Pharmacol. 1999, 47, 13–22.

(5) Koller, E. A.; Doraiswamy, P. M. Olanzapine-associated diabetes

mellitus. Pharmacotherapy 2002, 22, 841–852.

(6) Geddes, J.; Freemantle, N.; Harrison, P.; Bebbington, P. Atypical

antipsychotics in the treatment of schizophrenia: systematic overview

and meta-regression analysis. Br. Med. J. 2000, 321, 1371–1376.

(7) McGawin, J. K.; Goa, K. L. Aripiprazole. CNS Drugs 2002, 16, 779–

786.

(8) Buckley, P. F. Aripiprazole: Efﬁcacy and Tolerability Proﬁle of a

Novel-Acting Atypical Antipsychotic. Drugs Today 2003, 2, 145–

151.

(9) Butini, S.; Campiani, G.; De Angelis, M.; Fattorusso, C.; Nacci, V.;

Fiorini, I. Novel antipsychotic agents: recent advances in the drug

treatment of schizophrenia. Expert Opin. Ther. Patents 2003, 13, 425–

448.

(10) Morphy, R.; Rankovic, Z. Designed multiple ligands. An emerging

drug discovery paradigm. J. Med. Chem. 2005, 48, 6523–6543.

(11) Zhong, P.; Yuen, E. Y.; Yan, Z. Modulation of Neuronal Excitability

by Serotonin-NMDA Interactions in Prefrontal Cortex. Mol. Cell.

Neurosci. 2008, 38, 290–299.

(12) (a) Mason, S. L.; Reynolds, G. P. Clozapine has submicromolar afﬁnity

for 5-HT_1Areceptors in human brain tissue. Eur. J. Pharmacol. 1992,

221, 397–398. (b) Elliott, J.; Reynolds, G. P. Agonist-stimulated

GTPgamma[³⁵S] binding to 5-HT_1Areceptors in human post-mortem

brain. Eur. J. Pharmacol. 1999, 386, 313–315. (c) Richelson, E.;

Souder, T. Binding of antipsychotic drugs to human brain receptors.

Focus on newer generation compounds. Life Sci. 2000, 68, 29–39.

(13) Yoshida, K.; Sugita, T.; Higuchi, H.; Hishikawa, Y. Effect of

tandospirone on tardive dyskinesia and parkinsonian symptoms. Eur.

Psychiatry 1998, 13, 421–422.

(14) Christoffersen, C. L.; Meltzer, L. T. Reversal of haloperidol-induced

extrapiramidal side effects in cebus monkeys by 8-hydroxy-2-(di-N-

propylamino)tetralin and its enantiomers. Neuropsychopharmacology

1998, 18, 399–402.

(15) Millan, M. J. Improving the treatment of schizophrenia: focus on

serotonin 5-HT_1Areceptors. J. Pharmacol. Exp. Ther. 2000, 295, 853–

861.

(16) Marona-Lewicka, D.; Nichols, D. E. Aripiprazole (OPC-14597) fully

substitutes for the 5-HT_1Areceptor agonist LY293284 in the drug

discrimination assay in rats. Psychopharmacology (Berlin) 2004, 172,

415–421.

(34) Schwartz, T. W.; Frimurer, T. M.; Holst, B.; Rosenkilde, M. M.; Elling,

C. E. Molecular mechanism of 7TM receptor activationsa global

toggle switch model. Annu. ReV. Pharmacol. Toxicol. 2006, 46, 481–

519.

(35) Shi, L.; Javitch, J. A. The binding site of aminergic G protein-coupled

receptors: the transmembrane segments and second extracellular loop.

Annu. ReV. Pharmacol. Toxicol. 2002, 42, 437–467.

(36) Javitch, J. A.; Ballesteros, J. A.; Weinstein, H.; Chen, J. A cluster of

aromatic residues in the sixth membrane-spanning segment of the

dopamine D2 receptor is accessible in the binding-site crevice.

Biochemistry 1998, 37, 998–1006.

(37) Rasmussen, S. G.; Choi, H. J.; Rosenbaum, D. M.; Kobilka, T. S.;

Thian, F. S.; Edwards, P. C.; Burghammer, M.; Ratnala, V. R.;

Sanishvili, R.; Fischetti, R. F.; Schertler, G. F.; Weis, W. I.; Kobilka,

B. K. Crystal structure of the human beta2 adrenergic G-protein-

coupled receptor. Nature 2007, 450, 383–387.

(38) Cherezov, V.; Rosenbaum, D. M.; Hanson, M. A.; Rasmussen, S. G.;

Thian, F. S.; Kobilka, T. S.; Choi, H. J.; Kuhn, P.; Weis, W. I.;

Kobilka, B. K.; Stevens, R. C. High-resolution crystal structure of an

engineered human beta2-adrenergic G protein-coupled receptor. Sci-

ence 2007, 318, 1258–1265.

(39) Rosenbaum, D. M.; Cherezov, V.; Hanson, M. A.; Rasmussen, S. G.;

Thian, F. S.; Kobilka, T. S.; Choi, H. J.; Yao, X. J.; Weis, W. I.;

(17) Jordan, S.; Koprivica, V.; Chen, R.; Tottori, K.; Kikuchi, T.; Altar,

C. A. The antipsychotic aripiprazole is a potent, partial agonist at the

human 5-HT_1Areceptor. Eur. J. Pharmacol. 2002, 441, 137–140.

(18) Mauler, F.; Fahrig, T.; Horvath, E.; Jork, R. Inhibition of evoked

glutamate release by the neuroprotective 5-HT_1Areceptor agonist BAY

X 3702 in vitro and in vivo. Brain Res. 2001, 888, 150–157.

(19) Meltzer, H. Y.; Sumiyoshi, T. Does stimulation of 5-HT_1Areceptors

improve cognition in schizophrenia. BehaV. Brain Res. 2008, 195, 98–

102.

(20) (a) Schwartz, J. C.; Diaz, J.; Pilon, C.; Sokoloff, P. Possible

implications of the dopamine D₃receptor in schizophrenia and in

antipsychotic drug actions. Brain Res. Brain Res. ReV. 2000, 31, 277–

287. (b) Joyce, J. N.; Millan, M. J. Dopamine D₃receptor antagonists

as therapeutic agents. Drug DiscoVery Today 2005, 10, 917–925.

(21) Millan, M. J.; Loiseau, F.; Dekeyne, A.; Gobert, A.; Flik, G.; Cremers,

T. I.; Rivet, J. M.; Sicard, D.; Billiras, R.; Brocco, M. S33138 (N-

[4-[2-[(3aS,9bR)-8-cyano-1,3a,4,9b-tetrahydro[1] benzopyrano[3,4-

c]pyrrol-2(3H)-yl)-ethyl]phenyl-acetamide), a preferential dopamine

D₃versus D₂receptor antagonist and potential antipsychotic agent:

III. Actions in models of therapeutic activity and induction of side

effects. J. Pharmacol. Exp. Ther. 2008, 324, 1212–1226.

(22) (a) Campiani, G.; Nacci, V.; Bechelli, S.; Ciani, S. M.; Garofalo, A.;

Fiorini, I.; Wikstrom, H.; de Boer, P.; Liao, Y.; Tepper, P. G.;

Antipsychotics Targeting D₃5-HT_1A, and 5-HT_2AReceptors

Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 169

Stevens, R. C.; Kobilka, B. K. GPCR engineering yields high-

resolution structural insights into beta2-adrenergic receptor function.

Science 2007, 318, 1266–1273.

100907 in rats reﬂects antagonism of 5-HT_2Areceptors. Eur. J. Phar-

macol. 1995, 280, R9–R11. (b) Gleason, S. D.; Shannon, H. E.

Blockade of phencyclidine-induced hyperlocomotion by olanzapine,

clozapine and serotonin receptor subtype selective antagonists in mice.

Psychopharmacology (Berlin) 1997, 129, 79–84.

(40) Kalani, M. Y.; Vaidehi, N.; Hall, S. E.; Trabanino, R. J.; Freddolino,

P. L.; Kalani, M. A.; Floriano, W. B.; Kam, V. W.; Goddard, W. A.,

3rd. The predicted 3D structure of the human D2 dopamine receptor

and the binding site and binding afﬁnities for agonists and antagonists.

Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 3815–3820.

(41) Roth, B. L.; Shoham, M.; Choudhary, M. S.; Khan, N. Identiﬁcation

of conserved aromatic residues essential for agonist binding and second

messenger production at 5-hydroxytryptamine2A receptors. Mol.

Pharmacol. 1997, 52, 259–266.

(56) (a) Deutch, A. Y.; Lee, M. C.; Iadarola, M. J. Regionally speciﬁc

effects of atypical antipsychotic drugs on striatal fos expression: the

nucleus accumbens shell as a locus of antipsychotic action. Mol. Cell.

Neurosci. 1992, 3, 332–341. (b) Robertson, G. S.; Fibiger, H. C.

Neuroleptics increase c-fos expression in the forebrain: contrasting

effects of haloperidol and clozapine. Neuroscience 1992, 46, 315–

328.

(42) Boeckler, F.; Gmeiner, P. The structural evolution of dopamine D₃

receptor ligands: structure-activity relationships and selected neu-

ropharmacological aspects. Pharmacol. Ther. 2006, 112, 281–333.

(43) Boess, F. G., Jr.; Meyer, V.; Zwingelstein, C.; Sleight, A. J. Interaction

of tryptamine and ergoline compounds with threonine 196 in the ligand

binding site of the 5-hydroxytryptamine₆receptor. Mol. Pharmacol.

1997, 52, 515–523.

(44) Shapiro, D. A.; Kristiansen, K.; Kroeze, W. K.; Roth, B. L. Differential

modes of agonist binding to 5-hydroxytryptamine(2A) serotonin

receptors revealed by mutation and molecular modeling of conserved

residues in transmembrane region 5. Mol. Pharmacol. 2000, 58, 877–

886.

(45) Johnson, M. P.; Loncharich, R. J.; Baez, M.; Nelson, D. L. Species

variations in transmembrane region V of the 5-hydroxytryptamine type

2A receptor alter the structure-activity relationship of certain ergolines

and tryptamines. Mol. Pharmacol. 1994, 45, 277–286.

(46) Almaula, N.; Ebersole, B. J.; Ballesteros, J. A.; Weinstein, H.; Sealfon,

S. C. Contribution of a helix 5 locus to selectivity of hallucinogenic

and nonhallucinogenic ligands for the human 5-hydroxytryptamine_2A

and 5-hydroxytryptamine_2Creceptors: direct and indirect effects on

ligand afﬁnity mediated by the same locus. Mol. Pharmacol. 1996,

50, 34–42.

(47) Recanatini, M.; Cavalli, A.; Masetti, M. Modeling hERG and its

Interactions with Drugs: Recent Advances in Light of Current

Potassium Channel Simulations. ChemMedChem. 2008, 3, 523–535.

(48) Stansfeld, P. J.; Gedeck, P.; Gosling, M.; Cox, B.; Mitcheson, J. S.;

Sutcliffe, M. J. Drug block of the hERG potassium channel: insight

from modelling. Proteins 2007, 68, 568–580.

(49) Farid, R.; Day, T.; Friesner, R. A.; Pearlstein, R. A. New insights

about hERG blockade obtained from protein modeling, potential energy

mapping, and docking studies. Bioorg. Med. Chem. 2006, 14, 3160–

3173.

(50) Kamiya, K.; Niwa, R.; Mitcheson, J. S.; Sanguinetti, M. C. Molecular

determinants of hERG channel block. Mol. Pharmacol. 2006, 69,

1709–1716.

(51) Myokai, T.; Ryu, S.; Shimizu, H.; Oiki, S. Topological mapping of

the asymmetric drug binding to the human ether-a`-go-go-related gene

product (hERG) potassium channel by use of tandem dimers. Mol.

Pharmacol. 2008, 73, 1643–1651.

(57) Tremblay, P. O.; Gervais, J.; Rouillard, C. Modiﬁcation of haloperidol-

induced pattern of c-fos expression by serotonin agonists. Eur.

J. Neurosci. 1998, 10, 3546–3555.

(58) Robertson, G. S.; Matsumura, H.; Fibiger, H. C. Induction Patterns

of Fos-Like Immunoreactivity in the Forebrain as Predictors of

Atypical Antipsychotic Activity. J. Pharmacol. Exp. Ther. 1994, 271,

1058–1066.

(59) Sumner, B. E.; Cruise, L. A.; Slattery, D. A.; Hill, D. R.; Shahid, M.;

Henry, B. Testing the validity of c-fos expression proﬁling to aid the

therapeutic classiﬁcation of psychoactive drugs. Psychopharmacology

(Berlin) 2004, 171, 306–321.

(60) Belﬁeld, A. J.; Brown, G. R.; Foubister, A. J.; Ratcliffe, P. D. Synthesis

of meta-substituted aniline derivatives by nucleophilic substitution.

Tetrahedron 1999, 55, 13285–13300.

(61) Butini, S.; Campiani, G.; Borriello, M.; Gemma, S.; Panico, A.; Persico,

M.; Catalanotti, B.; Ros, S.; Brindisi, M.; Agnusdei, M.; Fiorini, I.;

Nacci, V.; Novellino, E.; Belinskaya, T.; Saxena, A.; Fattorusso, C.

Exploiting Protein Fluctuations at the Active-Site Gorge of Human

Cholinesterases: Further Optimization of the Design Strategy to

Develop Extremely Potent Inhibitors. J. Med. Chem. 2008, 51, 3154–

3170.

(62) Ding, H. Q.; Karasawa, N.; Goddard, W. A III atomic level simulations

on a million particles: the cell multipole method for Coulomb and

London nonbond interactions. J. Chem. Phys. 1992, 97, 4309–4315.

(63) Dauber-Osguthorpe, P.; Roberts, V. A.; Osguthorpe, D. J.; Wolff, J.;

Genest, M.; Hagler, A. T. Structure and energetics of ligand binding

to proteins: Escherichia coli dihydrofolate reductase-trimethoprim, a

drug-receptor system. Proteins 1988, 4, 31–47.

(64) Maple, J. R.; Hwang, M. J.; Stockﬁsch, T. P.; Dinur, U.; Waldman,

M.; Ewig, C. S.; Hagler, A. T. Derivation of class II force ﬁelds. I.

Methodology and quantum force ﬁeld for the alkyl functional group

and alkane molecules. J. Comput. Chem. 1994, 15, 162–182.

(65) Stewart, J. P. MOPAC; Stewart Computational Chemistry: Colorado

Springs, CO, 2007; http://OpenMOPAC.net.

(66) Dini, S.; Caselli, G. F.; Ferrari, M. P.; Giani, R.; Clavenna, G.

Heterogenity of [³H]-mepyramine binding sites in guinea pig cerebel-

lum and lung. Agents Actions 1991, 33, 181–184.

(67) DeLean, K. W.; Munson, P. J.; Rodbard, D. Simultaneous analysis of

families of sigmoidal curves: application to bioassay, radioligand assay

and physiological dose-response curves. Am. J. Physiol. 1978, 235,

E97–E102.

(68) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition constant

(K1) and the concentration of inhibitor which causes 50% inhibition

(IC₅₀) of an enzymatic reaction. Biochem. Pharmacol. 1973, 22, 3099–

3108.

(69) Hansen, H. H.; Timmermann, D. B.; Peters, D.; Walters, C.; Damaj,

M. I.; Mikkelsen, J. D. Alpha-7 nicotinic acetylcholine receptor

agonists selectively activate limbic regions of the rat forebrain: an

effect similar to antipsychotics. J. Neurosci. Res. 2007, 85, 1810–

1818.

(70) Mikkelsen, J. D.; Vrang, N.; Mrosovsky, N. Expression of Fos in the

circadian system following nonphotic stimulation. Brain Res. Bull.

1998, 47, 367–376.

(52) Pearlstein, R. A.; Vaz, R. J.; Kang, J.; Chen, X. L.; Preobrazhenskaya,

M.; Shchekotikhin, A. E.; Korolev, A. M.; Lysenkova, L. N.;

Miroshnikova, O. V.; Hendrix, J.; Rampe, D. Characterization of hERG

potassium channel inhibition using CoMSiA 3D QSAR and homology

modeling approaches. Bioorg. Med. Chem. Lett. 2003, 13, 1829–1835.

(53) Cavalli, A.; Poluzzi, E.; De Ponti, F.; Recanatini, M. Toward a

pharmacophore for drugs inducing the long QT syndrome: insights

from a CoMFA study of hERG K(+) channel blockers. J. Med. Chem.

2002, 45, 3844–3853.

(54) Ichikawa, J.; Ishii, H.; Bonaccorso, S.; Fowler, W. L.; O’Laughlin,

I. A.; Meltzer, H. Y. 5-HT_2aand D₂receptor blockade increases cortical

DA release via 5-HT_1areceptor activation: a possible mechanism of

atypical antipsychotic-induced cortical dopamine release. J. Neuro-

chem. 2001, 76, 1521–1531.

(55) (a) Maurel-Remy, S.; Bervoets, K.; Millan, M. J. Blockade of

phencyclidine-induced hyperlocomotion by clozapine and MDL

JM800689G

Article Doi

Discovery of a new class of potential multifunctional atypical antipsychotic agents targeting dopamine d3 and serotonin 5-HT 1A and 5-HT2A receptors: Design, synthesis, and effects on behavior

DOI: 10.1021/jm800689g

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm800689g

Full text of DOI:10.1021/jm800689g

Products guided by the article

R&D Labs maybe for 898532-77-9

Relevant to this article

Hot Product