Discovery of 1-[2-(2,4-Dimethylphenylsulfanyl)phenyl]piperazine (Lu AA21004): A Novel Multimodal Compound for the Treatment of Major Depressive Disorder

ARTICLE

pubs.acs.org/jmc

Discovery of 1-[2-(2,4-Dimethylphenylsulfanyl)phenyl]piperazine

(Lu AA21004): A Novel Multimodal Compound for the Treatment

of Major Depressive Disorder

Benny Bang-Andersen,*^,†Thomas Ruhland,^†Morten Jørgensen,^†Garrick Smith,^†Kristen Frederiksen,^†

Klaus Gjervig Jensen,^†Huailing Zhong,^‡Søren Møller Nielsen,^†Sandra Hogg,^†Arne Mørk,^†and

Tine Bryan Stensbøl^†

^†Neuroscience Drug Discovery Denmark, H. Lundbeck A/S, 9 Ottiliavej, DK-2500 Copenhagen-Valby, Denmark

^‡Lundbeck Research USA, 215 College Road, Paramus, New Jersey 07652-1431, United States

S Supporting Information

b

ABSTRACT: The synthesis and structureꢀactivity relation-

ship of a novel series of compounds with combined eﬀects on

5-HT_3Aand 5-HT_1Areceptors and on the serotonin (5-HT)

transporter (SERT) are described. Compound 5m (Lu

AA21004) was the lead compound, displaying high aﬃnity for

recombinant human 5-HT_1A(K_i= 15 nM), 5-HT_1B(K_i= 33 nM), 5-HT_3A(K_i= 3.7 nM), 5-HT₇(K_i= 19 nM), and noradrenergic

β₁(K_i= 46 nM) receptors, and SERT (K_i= 1.6 nM). Compound 5m displayed antagonistic properties at 5-HT_3Aand 5-HT₇

receptors, partial agonist properties at 5-HT_1Breceptors, agonistic properties at 5-HT_1Areceptors, and potent inhibition of SERT.

In conscious rats, 5m signiﬁcantly increased extracellular 5-HT levels in the brain after acute and 3 days of treatment. Following the

3-day treatment (5 or 10 (mg/kg)/day) SERT occupancies were only 43% and 57%, respectively. These characteristics indicate that

5m is a novel multimodal serotonergic compound, and 5m is currently in clinical development for major depressive disorder.

’ INTRODUCTION

shows modest eﬃcacy, and the majority of patients with MDD

fail to achieve full remission with current therapies.⁶In addition, a

relatively long time to onset of symptom relief and issues with

relapse characterize marketed antidepressants. Future drugs will

also need to address cognitive impairment, pain, and drug-related

side eﬀects such as sexual dysfunction, nausea/emesis, weight

gain, and potential cardiovascular eﬀects.⁴

For many years, the medical community and pharmaceutical

companies have been focusing on developing superior pharma-

cotherapeutic strategies for the treatment of depression and

anxiety.^7,8Clinical research has demonstrated that the antide-

pressant activity of SSRIs may be augmented by coadministration

of pindolol, an eﬀect that has been attributed to pindolol’s partial

agonistic eﬀect at the 5-HT_1Areceptor.^9,10This observation has

inspired many pharmaceutical companies to launch research

programs with the aim of discovering designed multiple ligands

(DMLs)¹¹with dual activity at the serotonin transporter (SERT)

and the 5-HT_1Areceptor. Other proﬁles such as triple reuptake

inhibitors [SERT, NE transporter (NET), and dopamine (DA)

transporter (DAT) inhibition] and compounds inhibiting SERT

plus various combinations of 5-HT receptor subtypes have also

been pursued.^7,12

The approximate lifetime prevalence of major depressive

disorder (MDD) and anxiety disorders in the United States is

17% and 29%, respectively, and onset is typically in childhood

or adolescence.¹According to the World Health Organization’s

Global Burden of Disease project, MDD will become the second

leading cause of disability worldwide within the next 10 years.²

Pharmacotherapy for MDD and anxiety disorders has been

available since the late 1950s, when the tricyclic antidepressants

(TCAs) and the monoamine oxidase inhibitors (MAOIs) be-

came clinically available. During the 1980s, the ﬁrst selective

serotonin (5-HT) reuptake inhibitors (selective SRIs or SSRIs)

were introduced and since then have become the most widely

prescribed medication for the treatment of depression and

related disorders.³SSRIs such as ﬂuoxetine, sertraline, parox-

etine, ﬂuvoxamine, and citalopram (1) show similar eﬃcacy but

clear advantages in terms of improved tolerability and a superior

safety proﬁle compared to TCAs and MAOIs.⁴The serotonin

and norepinephrine (NE) reuptake inhibitors (SNRIs) venlafax-

ine, duloxetine, desvenlafaxine, and milnacipran, as well as the

allosteric serotonin reuptake inhibitor escitalopram, have been

marketed in the 1990s and during the new millennium. In

contrast to SSRIs, SNRIs display modest improved eﬃcacy and

a slightly faster onset of antidepressant action but lower

tolerability.^4,5Despite more than 5 decades of research, there

are still high unmet needs in the treatment of depression. For

example, the treatment with currently available antidepressants

At H. Lundbeck A/S, DML programs were pursued with the

goal of ﬁnding a new generation of antidepressants. These were

Received: November 11, 2010

Published: April 12, 2011

r

2011 American Chemical Society

3206

dx.doi.org/10.1021/jm101459g J. Med. Chem. 2011, 54, 3206–3221

|

Journal of Medicinal Chemistry

ARTICLE

deﬁned prior to initiation of the lead generation program (5-

HT_2Cantagonism and SRI),^13,14or they emerged from the

knowledge that was generated during such programs (5-HT₃

receptor antagonism, 5-HT_1Areceptor agonism, and SRI). The

objective of the current program was to ﬁnd a compound

that combines 5-HT₃receptor antagonism, 5-HT_1Areceptor

agonism, and SERT inhibition in a single molecule and that will

exert superior antidepressant eﬀects in humans compared to

marketed drugs. Recently, the term multimodal was coined for

compounds that combine at least two separate pharmacological

modes of action that complement each other in terms of eﬃcacy

or tolerability.^15,16Here we describe the biological rationale for a

novel class of putative multimodal antidepressants, which is

proposed to act via diﬀerent modes of action, namely, blocking

ligand-gated ion channels (5-HT₃receptor) and monoamine

transporters (SERT), as well as activating G-protein-coupled

receptors (5-HT_1Areceptor). SERT inhibition is important for

antidepressant and anxiolytic activity.⁴The simultaneous use of

an eﬃcacious 5-HT_1Areceptor agonist, together with a SERT

inhibitor, is expected to rapidly desensitize the inhibitory soma-

todendritic 5-HT_1Aautoreceptors and, at the same time, to

mediate part of the antidepressant eﬀect through activation of

postsynaptic 5-HT_1Areceptors.^17,18The rationale for adding

5-HT₃receptor antagonism was based on preclinical studies

demonstrating that 5-HT₃receptor antagonists eliminate the

inhibitory tonus of 5-HT₃receptors on the release of NE and

acetylcholine (AcCh) in the forebrain, via 5-HT₃hetero-

receptors located on inhibitory γ-aminobutyric acid (GABA)-

ergic interneurons.^19ꢀ22These studies indicate that 5-HT₃

receptor antagonists may exert positive eﬀects on mood and

cognitive impairment in patients with depression. Adding to this

point, 5-HT_1Areceptor agonists such as buspirone (2) are used

as anxiolytics in the clinic, whereas 5-HT₃receptor antagonists

such as ondansetron (3) are used to treat the nausea associated

with chemotherapy and have shown some success in alleviating

the gastrointestinal distress associated with SSRIs.^23,24Interest-

ingly, litoxetine, initially described as an SSRI, was later shown to

possess antiemetic properties in ferrets that were ascribed to its

moderate 5-HT₃receptor antagonistic properties.²³Hence, a

combined 5-HT₃receptor antagonist, 5-HT_1Areceptor agonist,

and SERT inhibitor was hypothesized to have the potential to

treat signiﬁcant clinical unmet needs in MDD and generalized

anxiety disorder (GAD).

(data not shown). Structurally closely related members of the

series displayed substantially higher aﬃnity for h5-HT_3Aand/or

h5-HT_1Areceptors (Table 1). In contrast to compound 4b, some

of the analogues showed good human microsomal stability and/

or minimal cytochrome P450 (CYP450) enzyme inhibition (data

not shown). Thus, the series was a useful starting point for a

number of DML programs, including the structureꢀactivity

relationship (SAR) studies reported herein that led to the

discovery of 5m (Chart 1).

’ RESULTS

Chemistry. Several different strategies were employed for the

synthesis of the target compounds 4ꢀ6. The compound class

was initially prepared by solid-phase synthesis in order to

generate compound libraries for lead discovery programs. The

key step of this synthetic strategy was the generation of aromatic

carbonꢀheteroatom bonds through iron-assisted nucleophilic

aromatic substitution.^26,27This was a suitable method for the

preparation of focused libraries in a short space of time. However,

a different approach was necessary in order to synthesize such

molecules in larger quantities and in order to prepare molecules

with a more complex substitution pattern on the central benzene

ring. For this purpose, the target compounds were synthesized

either by two sequential palladium-catalyzed carbonꢀheteroa-

tom bond formations^28,29(methods A and B, Schemes 1 and 2)

or through cyclization of anilines with bis(2-bromoethyl)amine

(method C). In method A, the synthesis started from 2-bromo-

substituted N-phenylpiperazines 7 that were readily available in

multigram quantities through a selective monoamination of

dihalogenated benzenes.³⁰The thioether linkage was subse-

quently formed using the method reported by Schopfer and

Schlapbach.³¹In the analogous method B, the order of steps was

reversed. For both strategies, the required starting materials,

halogenated benzenes, were either commercially available or

could be readily obtained from the corresponding anilines.

Finally, the resulting tert-butyloxycarbonyl (Boc)-protected

intermediates were deprotected to give the target molecules

4ꢀ6 as their hydrochloride salts. Overall, method B turned out

to be more practical than methods A and C, since the Boc-

protected target molecules were easily separated by chromato-

graphy from unreacted diaryl sulfides and hydrodebrominated

byproducts.

In a focused screening program we identiﬁed the arylpiper-

azine 4b and a number of structural analogues as a lead series.

Arylpiperazines are well-known motifs that are typically found in

compounds possessing activities toward a broad range of targets,

including various 5-HT receptor subtypes. For example, some

1-(2-phenoxyaryl)piperazines, although structurally similar to

our lead series, were recently found to display a slightly diﬀerent

pharmacological proﬁle, such as high 5-HT_1Areceptor aﬃnity

but only modest SERT inhibition.²⁵This and other examples

indicate that the design of molecules, interacting in a predeﬁned

manner with three diﬀerent target classes (i.e., ligand-gated ion

channels, G-protein-coupled receptors, and monoamine trans-

porters), will be challenging, and it was also anticipated that

compounds fulﬁlling our design hypothesis might well interact

with additional targets in a subsequent broader target proﬁling.

Indeed, we found that compound 4b displays moderate human

(h) 5-HT_3Areceptor aﬃnity, very weak h5-HT_1Areceptor

aﬃnity, and potent rat (r) SERT inhibition (Table 1), as well

as high aﬃnity (K_i< 100 nM) for h5-HT_2Cand rR₁receptors

StructureꢀActivity Relationship Studies. The affinities of

target compounds 4ꢀ6 for the cloned h5-HT_3Aand h5-HT_1A

receptors were determined by use of in vitro binding competition

assays with membranes prepared from cells expressing the

relevant receptors, and data are reported as K_ivalues (see

Supporting Information for pK_i( SEM values). The inhibition

of uptake of [³H]5-HT, [³H]NE, and [³H]DA into rat brain

synaptosomes through the monoamine transporters was also

measured, and data are reported as mean IC₅₀values (see

Supporting Information for pIC₅₀( SD/SEM values) of at least

two independent determinations. These data are shown in

Tables 1ꢀ3 and 5, along with the structure of the compounds.

While affinity measures were used at the cloned h5-HT_3Aand h5-

HT_1Areceptors, for historical reasons a functional readout

was used as the SAR driving assay for assessing potency at

the SERT. Internal data and literature data, however, have

demonstrated a close correlation between affinity at hSERT

and potency in [³H]5-HT uptake assays, suggesting that values

are comparable.³²The compounds were prioritized for further

3207

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Table 1. StructureꢀActivity Relationship of Compounds 4aꢀ1 Compared with Reference Compounds Citalopram (1),

Buspirone (2), and Ondansetron (3) at Human (h) 5-HT_3Aand h5-HT_1AReceptors and the Rat (r) Serotonin Transporter

(rSERT)

K_i(nM)^a

IC₅₀(nM)^a

b

c

compd

synthesis method

R₁

R₅

h5-HT_3A

h5-HT_1A

rSERT^d

4a

B

A

H

CH₃

H

310

190

500

440

260

64

1100

4000

2500

1100

2800

30

150

7.9

4b

OCH₃

Cl

4c

9.5

4d

F

68

4e

CH₃

H

20

4f

390

8.0

4g

OCH₃

Cl

H

36

130

69

4h

H

36

13

4i

F

H

53

39

67

4j

CH₃

CF₃

C(CH₃)₃

H

56

44

18

4k

H

140

450

NT

>10000

2.8

240

2500

NT

21

6.9

4l

H

2100

3.9^e

>1000

>2000

1, citalopram

2, buspirone

3, ondansetron

NT

^aData are the mean of a minimum of two determinations covering 6 decades for each compound. K_ivalues were calculated from the ChengꢀPrusoﬀ

equation (K_i= IC₅₀/(1 þ [L]/K_d)). NT, not tested. ^b[³H]Granisetron binding to cloned 5-HT_3Areceptors. ^c[³H]5-Carboxamidotryptamine binding

to cloned 5-HT_1Areceptors. ^d[³H]5-HT uptake measured in rat synaptosomes. ^eReference 59.

Chart 1. Structures of citalopram (1), buspirone (2), ondansetron (3), target compounds (4-6) and 5m

SAR studies based on the ratios between the in vitro affinities/

potencies at h5-HT_3A, and h5-HT_1Areceptors and the rSERT.

It has been demonstrated in both preclinical and clinical studies

that a relatively high hSERT occupancy of around 80ꢀ90% is

required for SSRIs and SNRIs to exert an antidepressant

activity.^33ꢀ35In contrast, the relationship between the in vivo

3208

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Scheme 1. Synthesis of Target Compounds by Method A

occupancy in experimental animals and humans and the efficacy

at 5-HT₃and 5-HT_1Areceptors is less well described. However,

receptor occupancies of >50% have been reported for the

antiemetic activity of 5-HT₃receptor antagonists including

ondansetron (3),³⁶whereas much lower receptor occupancies

have been estimated to be efficacious for buspirone (2) at

5-HT_1Areceptors.^37ꢀ39Hence, compounds were selected that

displayed relative affinities/potencies (K_i/IC₅₀) at h5-HT_3Aand

h5-HT_1Areceptors and rSERT corresponding to rSERT/h5-

HT_3A/h5-HT_1A≈ 1:(1ꢀ5):(5ꢀ15). From a hypothesis that the

in vitro profile alone would reflect the relative in vivo occupan-

cies, these ratios would correspond to in vivo occupancy levels of

45ꢀ80% at 5-HT₃receptors and 20ꢀ45% at 5-HT_1Areceptors

for a compound with an in vivo occupancy of 80% at SERT

(Figure 1). Furthermore, the compounds were only considered

if the 5-HT_1Areceptor affinity was <100 nM in order to limit

the amount of compounds for broader in vitro and in vivo

profiling.

We ﬁrst focused on the R₁substituent in the phenylsulfanyl

ring and the R₅substituent in the central benzene ring of lead

compound 4b (Table 1). A number of analogues of 4b were

available at the onset of the project. In contrast to the methoxy in

4b, these analogues were furnished with either electron-with-

drawing substituents (R₁= F or CF₃) or substituents with limited

electronic eﬀects (R₁= Cl, CH₃, and tert-butyl). The series of

compounds was extended by synthesizing analogues carrying a

hydrogen atom at R₅instead of the methyl substituent of 4b.

These latter compounds demonstrated higher aﬃnities at the

5-HT_3Aand 5-HT_1Areceptors in comparison with compounds

furnished with a methyl substituent at this position, i.e., 4aꢀe

compared with 4fꢀj. All compounds with relatively small

substituents at R₁and either a hydrogen atom or a methyl

substituent at R₅(4b,c,e,g,h,j,k) inhibited rSERT in the same

range, whereas compounds with a ﬂuoro atom (4d and 4i) or a

hydrogen atom (4a and 4f) exhibited weaker rSERT inhibition.

The compound 4l with the large tert-butyl substituent was a

signiﬁcant weaker inhibitor of rSERT, demonstrating that small

R₁substituents are preferred for potent rSERT inhibition. It was

also shown that rSERT inhibition seemed to be less aﬀected by

the electronic eﬀects of the R₁substituent as long as the

substituent was above a certain size. The compounds with an

electron-donating methoxy substituent (4b and 4g), a more or

less neutral methyl or chloro substituent (4e, 4j, 4c, and 4h), or

an electron-withdrawing triﬂuoromethyl substituent (4k) were

more or less equipotent rSERT inhibitors. From this initial series,

compounds 4h and 4j were selected for further in vitro evaluation

(Tables 3 and 4), whereas compounds 4aꢀe,g,k,l were discarded

because of their relative low h5-HT_1Areceptor aﬃnities and

compounds 4f and 4i were discarded because of their low rSERT

inhibition and suboptimal ratio between rSERT inhibition and

h5-HT_1Areceptor aﬃnity.

We extended the series of compounds through the synthesis

of analogues carrying a hydrogen atom at R₅and methoxy,

chloro, and/or methyl groups at other positions of the phenyl-

sulfanyl ring. Thus, the eﬀects of moving the substituent from

R₁to either R₂or R₃and of having two substituents (two

identical or two diﬀerent substituents) in the phenylsulfanyl

ring were investigated (Table 2). In general, all of these

3209

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Scheme 2. Synthesis of Target Compounds by Method B

compounds (5aꢀp) turned out to have relative high aﬃnities

for h5-HT_3Areceptors (K_ifrom 7.3 to 110 nM). These

compounds also showed moderate to high aﬃnities at the h5-

HT_1Areceptor (K_ifrom 14 to 330 nM) and moderate to high

activity at rSERT (IC₅₀from 4.3 to 330 nM). Compounds with

a methoxy or a methyl substituent at R₂or R₃(5aꢀd) were

weaker rSERT inhibitors than the corresponding R₁analogues

(4g and 4j). Compound 5d, with a monomethyl substituent at

R₃, showed high aﬃnity for h5-HT_1Areceptors, and the

corresponding R₁and R₂analogues 4j and 5b showed weaker

aﬃnities for h5-HT_1Areceptors compared with 5d by factors of 3

and 7, respectively.

Six compounds were synthesized with two identical adjacent

substituents, i.e., compounds with the same substituent at R₁and

R₂(5e,f,g) or at R₂and R₃(5h,i,j). These compounds were all

relatively potent rSERT inhibitors with favorable ratios between

the three targets. However, only compounds 5g, 5i, and 5j were

selected for further in vitro investigations (Tables 3 and 4). Three

compounds with identical substituents at R₁and R₃were also

synthesized (5k,l,m), and all these compounds had the desired h5-

HT_3A/rSERT ratio. Whereas compound 5k showed weak h5-

HT_1Areceptor aﬃnity, compounds 5l and 5m were selected for

further in vitro proﬁling (Tables 3 and 4). Finally, three analogues

with diﬀerent substituents at R₁and R₃were synthesized (5n,o,p).

Only compound 5p was selected for further in vitro proﬁling

(Tables 3 and 4), since compounds 5n and 5o showed low 5-HT_1A

receptor aﬃnities, i.e., compound 5n in relation to inhibition of

SERT and 5o with K_i> 100 nM.

Compounds 4h, 4j, 5g, 5i, 5j, 5l, 5m, and 5p were tested for

rNET and rDAT inhibition. The ratios between rDAT and

rSERT and between rNET and rSERT inhibition are shown in

Table 3. The rNET/rSERT ratios ranged from 7 to 150, whereas

the rDAT/rSERT ratios ranged from 1 to 170. Compound 5m

exhibited the highest ratio (170) between rDAT and rSERT

inhibition. Compound 5g showed the highest ratio (150)

between rNET and rSERT inhibition, and a group of compounds

including 5m showed rSERT selectivity with ratios ranging from

20 to 35. Furthermore, drug metabolism and pharmacokinetic

(DMPK) properties of the compounds in the series were

optimized by monitoring their in vitro metabolic stability in rat

and human liver microsomes and by assessing the inhibitory

potential of the compounds toward recombinant (cDNA-ex-

pressed) human CYP450 enzymes. As shown in Table 4, the

compounds generally showed good stability in human liver

microsomes; i.e., the intrinsic clearances (CL_int) of all but one

(5i) were below that for the human liver blood ﬂow (LBF).

Conversely, in rat liver microsomes the CL_intwere all above the

rat LBF of 20 mL/min, indicating high clearance in vivo. This was

conﬁrmed in vivo for all compounds tested in the project, e.g., for

compounds 4h and 5m the hepatic clearances and oral bioavail-

abilities in rats were found to be 5.8 and 7.1 (L/h)/kg and 26%

and 16%, respectively. Furthermore, inhibition of CYP1A2,

3210

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Table 2. StructureꢀActivity Relationship of Compounds 5aꢀp at Human (h) 5-HT_3Aand h5-HT_1AReceptors and the Rat (r)

Serotonin Transporter (rSERT)

K_i(nM)^a

IC₅₀(nM)^a

rSERT^d

synthesis

method

b

c

compd

R₁

R₂

R₃

h5-HT_3A

h5-HT_1A

5a

5b

5c

5d

5e

5f

A

B

A

B

A

B

A

B

H

OCH₃

CH₃

H

62

140

100

59

53

64

54

OCH₃

CH₃

H

52

330

51

H

32

14

OCH₃

Cl

OCH₃

Cl

86

160

68

15

H

110

33

4.7

5.8

37

5g

5h

5i

CH₃

H

CH₃

OCH₃

Cl

H

53

OCH₃

Cl

80

68

H

67

75

17

5j

H

CH₃

H

CH₃

OCH₃

Cl

29

31

4.4

7.6

5.2

5.3/5.4^g

4.3

11

5k

5l

OCH₃

Cl

32

330

H

26

23/3.7^e

78

5m

5n

5o

5p

CH₃

OCH₃

CH₃

H

CH₃

Cl

39/15^f

94

H

12

H

OCH₃

Cl

17

210

78

H

7.3

9.7

^aData are the mean of a minimum of two determinations covering 6 decades for each compound. K_ivalues were calculated from the ChengꢀPrusoﬀ

c

equation (K_i= IC₅₀/(1 þ [L]/K_d)). ^b[³H]Granisetron binding to cloned h5-HT_3Areceptors. [³H]5-Carboxamidotryptamine binding to cloned

5-HT_1Areceptors. ^d[³H]5-HT uptake from rat synaptosomes. ^eExternal data obtained from Cerep, catalogue number 808-3h. ^fExternal data obtained

from MDS Pharmaservices, catalogue number 271110. ^g[³H]5-HT uptake from CHO cells expressing hSERT. For compound 5m, in-house and

external data are shown for h5-HT_3Aand h5-HT_1Areceptors, and for SERT both rat and human data are shown.

CYP2C9, and CYP3A4 was found to be in the high micromolar

range for all compounds, while inhibition toward CYP2D6

ranged from 0.4 μM for compound 4h to 9.8 μM for compound

5m. On the basis of the favorable in vitro and DMPK proﬁle,

compound 5m was selected for further SAR exploration, as well

as for broader in vitro and in vivo proﬁling.

The SAR of 5m was further explored by a “methyl walk”

around the entire molecular framework (Table 5). In addition,

the piperazine ring was replaced with a homopiperazine (6f) and

the sulfur atom with an oxygen atom (6g). In comparison to 5m,

all of these compounds showed weaker rSERT inhibition, and

only compound 6g had acceptable h5-HT_1Areceptor aﬃnity.

They also had relatively large variations in their aﬃnity for h5-

HT_3Areceptors, with K_iranging from 45 to 670 nM. Compound

6c (R₆= CH₃) and homopiperazine 6f showed some selectivity

for h5-HT_3Areceptors, whereas the oxygen analogue 6g, in

agreement with literature data, had some selectivity for h5-HT_1A

receptors.²⁵

using Xenopus oocytes expressing cloned homomeric h5-HT_3A

receptors (Figure 2). Conversely, at higher concentrations, 5m

showed agonistic properties (EC₅₀= 2100 nM), reaching a

maximal response of 64% relative to 5-HT (Figure 3). The

agonistic response could be blocked by the 5-HT₃receptor

antagonist ondansetron (data not shown). At r5-HT_3Areceptors,

no agonistic response was found over a large concentration range

(0.1ꢀ100000 nM), but in agreement with the findings at

h5-HT_3Areceptors, 5m was found to be a potent antagonist at

r5-HT_3Areceptors (IC₅₀= 0.18 nM) (Figure 2). Compound 5m

displayed high binding affinity for cloned h5-HT_1Areceptors in

two different assays (K_i= 39 nM (in-house) and K_i= 15 nM

(MDS Pharmaservice), whereas relative low affinity binding was

found at r5-HT_1Areceptors (K_i= 230 nM). In a [³⁵S]guanosine

5⁰-O-(3-thio)triphosphate ([³⁵S]GTPγS) binding assay with

h5-HT_1Areceptors, 5m was found to be an agonist (EC₅₀

=

200 nM, efficacy = 96%, Figure 4). In vitro characterization

showed that 5m was a potent rSERT inhibitor (IC₅₀= 5.3 nM)

when assessed in rat brain synaptosomes, and a similar potency

was found on hSERT when expressed in Chinese hamster ovary

(CHO) cells (IC₅₀= 5.4 nM). Comparably, binding affinity at

hSERT was found to be 1.6 nM. Weak potencies (IC₅₀> 100

nM) of 5m were found at rNET and rDAT (Table 3). At 1 μM,

5m was tested against a panel of more than 75 ion channels,

Further Pharmacological and Pharmacokinetic Charac-

terization of 5m. Compound 5m displayed high binding affinity

for cloned homomeric h5-HT_3Areceptors [K_i= 23 nM (in-house)

and K_i= 3.7 nM (Cerep)]. In agreement with this, 5m was shown

to be a potent functional antagonist (IC₅₀= 12 nM, tested against

10 μM 5-HT) when tested in an electrophysiological setup

3211

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Table 4. In Vitro Metabolic Stability in Rat and Human Liver

Microsomes and Inhibition of Recombinant (cDNA-ex-

pressed) Human CYP Enzymes

Cl_int

(L/min) (mL/min)

IC₅₀(μM)

compd human

rat

CYP1A2 CYP2C9 CYP2D6 CYP3A4

4h

4j

0.6

1.1

0.5

1.9

0.7

0.5

0.6

1.4

70

260

290

260

120

38

53

40

26

32

24

27

40

34

31

40

34

40

34

39

40

0.4

2.7

2.8

0.5

3.3

0.9

9.8

1.5

6.8

16

11

6.1

5g

5i

5j

8.7

4.4

5l

5m

5p

LBF^a

55

10

5.5

120

20

^aLBF, liver blood ﬂow.

Figure 1. Calculated relationships between the occupancy levels at

5-HT_1Aor 5-HT₃receptors relative to serotonin (5-HT) transporter

(SERT) occupancy. The line with circles (ꢀbꢀ) shows the relationship

between the occupancies if the ratio between the SERT and 5-HT_1Aor

5-HT₃is 1. The line with triangles (ꢀ2ꢀ) shows the relationship when the

ratio is 5, and the line with squares (ꢀ9ꢀ) shows the relationship when the

ratio is 15; i.e., a compound with occupancy of 80% at SERT would give

occupancy of 45% at 5-HT_1Areceptors if the aﬃnity at 5-HT_1Areceptors is

5-fold lower than at SERT, whereas only 20% occupancy would be obtained

if the 5-HT_1Areceptors aﬃnity is 15-fold lower than at SERT. The two

hatched areas depict estimated clinically relevant occupancy levels at

5-HT_1Aand 5-HT₃receptors based on these considerations.

Table 3. StructureꢀActivity Relationship of Key Compounds

Selected from Tables 1 and 2 at Rat Serotonin Transporter

(rSERT), Rat Dopamine Transporter (rDAT), and Rat Nor-

epinephrine Transporter (rNET), as well as the Calculated

Ratios rDAT/rSERT and rNET/rSERT

IC₅₀(nM)^a

compd rSERT^brDAT^b

ratio

rNET^b

37/45^c160/190^c

rDAT/rSERT rNET/rSERT

4h

4j

13

2.8/3.5^d

6.1

21

12/15^d

25

18

110

120^c

21^c

170^c

250^c

890

320^c

460

880^c

120^c

90^c

5g

5i

5.8

17

150

7

1

5j

4.4

5.2

5.3

9.7

38

20

5l

120^c

140

340^c

48

23

5m

170

33

26

5p

35

^aData are the mean of a minimum of two values. IC₅₀values were

determined using drug concentrations covering at least 3 decades.

^b[³H]5-HT, [³H]DA, and [³H]NE uptake from rat synaptosomes.

^cCerep catalogue numbers 711 and 712 for rNET and rDAT, respec-

tively. ^dFor compound 4h, ratios from in-house and external data

are shown.

Figure 2. Compound 5m and ondansetron (3) were tested for antag-

onistic eﬀects on human (h) and rat (r) 5-HT_3Areceptors expressed in

oocytes. Compound 5m inhibited the h5-HT_3Aand r5-HT_3Areceptors

with IC₅₀equal to 12 and 0.18 nM, respectively. Ondansetron (3)

showed similar inhibition of both h5-HT_3Aand r5-HT_3Areceptors with

IC₅₀≈ 0.1 nM.

G-protein-coupled receptors, enzymes, and transporters

(Table 6). Modest binding affinities were found at histamine

hH₂, melanocortin hMC₄, noradrenergic hβ₂, serotonergic

h5-HT_2C, h5-HT_5A, and h5-HT₆receptors, whereas more potent

3212

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Figure 4. Dose relationship depicting potency and intrinsic activity

(IA) of 5m, buspirone, and 5-OH DPAT in a [³⁵S]guanosine 5⁰-O-(3-

thio)triphosphate ([³⁵S]GTPγS) binding assay in CHO membranes

expressing h5-HT_1Areceptors. IA_maxwas deﬁned by 5-HT binding. Raw

data were obtained by MDS Pharmaservices and were then analyzed by

sigmoidal dose response curve-ﬁttings using GraphPad Prism 4 to

generate EC₅₀and intrinsic activity (IA) results. Data shown in the

graph were averaged from n = 3.

enhanced hippocampal 5-HT levels by 700% at the dose

yielding maximal response (data not shown). Three days of

treatment with 5m (5 or 10 (mg/kg)/day) also resulted in

significantly higher basal levels of 5-HT in the medial

prefrontal cortex (mPFC) compared to animals treated with

vehicle. The occupancy of SERT by 5m after 3 days of

treatment with 5 or 10 (mg/kg)/day was 43% and 57%,

respectively (Figure 6). Furthermore, 5m showed a robust

antidepressant- and anxiolytic-like profile in multiple preclini-

cal assays, such as the mouse forced-swim and tail-suspension

tests and the rat social interaction and conditioned-fear

paradigms.⁴⁰

The prediction of the hepatic clearance of 5m in humans

was done by in vitro/in vivo extrapolation from the predicted

hepatic clearance in rats and humans and the observed

systemic in vivo clearance in rats. According to the well-stirred

model, the hepatic clearance in each species (CL_pred) was

predicted from hepatic microsomal intrinsic clearance (CL_int)

according to

Figure 3. Compound 5m displayed an agonistic response on human

(h) 5-HT_3Areceptors (arrowhead) but not on rat (r) 5-HT_3Areceptors

(arrow). At h5-HT_3Areceptors, 5m had an EC₅₀of 2100 nM, reaching a

maximal response of 64% compared with the maximal response with

5-HT. Note that after only one application of 5m for 30 s, the

subsequent application of either 5m or 5-HT (10 μM) did not induce

any response at all, indicating that 5m induces functional antagonism.

That means that it is only possible to test one concentration of 5m on

each oocyte, so the agonism dose-reponse curves for 5m has been

obtained from a test of each single concentration of 5m using three to six

separate oocytes.

ðLBFÞðCL_intÞðf_uÞ

CL_pred

¼

LBF þ ðCL_intÞðf_uÞ

binding (K_i< 180 nM) was found at noradrenergic hβ₁receptors

(K_i= 46 nM) and serotonergic h5-HT_1B(K_i= 33 nM) and h5-

HT₇(K_i= 19 nM) receptors (Table 6). At h5-HT_1B

receptors, 5m proved to be a partial agonist (EC₅₀= 120 nM,

efficacy = 55%) when tested in a [³⁵S]GTPγS binding assay,

whereas 5m was shown to be an antagonist (K_i= 450 nM) at h5-

HT₇receptors when tested in a cAMP dependent fluore-

scence based assay using cloned h5-HT₇receptors expressed in

HEK-CNG cells. The unique in vitro pharmacological properties

of 5m (Table 7) translate into an in vivo profile that differs

from those of currently used antidepressants. Compound 5m

(2.5, 5, or 10 mg/kg sc) increased the extracellular levels of

5-HT in the ventral hippocampus in conscious rats (Figure 5).

In a similar experiment administration of the SSRI escitalopram

where LBF is liver blood ﬂow and f_uis the unbound fraction in

plasma. Human clearance was then predicted from the observed

in vivo rat clearance and the clearances that are predicted based

on in vitro clearances from each species (CL_pred,humanand

CL_pred,rat) according to

CL_pred;human

CL_human¼ CL_rat

CL_pred;rat

The plasma protein binding of 5m was determined to 98.3%

and 99.2% for human and rat, respectively, resulting in a

predicted low human hepatic clearance of 35 L/h, which

turned out to be oﬀ by only a few liters when measured in

3213

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Table 5. StructureꢀActivity Relationship of Compounds 6aꢀg Compared with 5m at Human (h) 5-HT_3Aand h5-HT_1AReceptors

and the Rat Serotonin Transporter (rSERT)

K_i(nM)^a

IC₅₀(nM)^a

rSERT^d

b

c

compd

synthesis method

R₄

R₅

R₆

R₇

R₈

n

X

h5-HT_3A

h5-HT_1A

6a

6b

6c

6d

6e

6f

B

A

B

C

A

CH₃

H

1

2

1

S

O

S

160

150

45

420

44

CH₃

H

2700

1400

540

25

H

CH₃

H

200

540

680

130

250

5.3/5.4^g

H

CH₃

H

670

130

27

H

CH₃

H

120

H

310

6g

5m

H

270

23/3.7^e

65

40/15^f

H

^aData are the mean of a minimum of two determinations covering 6 decades for each compound. K_ivalues were calculated from the ChengꢀPrusoﬀ

equation (K_i= IC₅₀/(1 þ [L]/K_d]. ^b[³H]Granisetron binding to cloned 5-HT_3Areceptors. ^c[³H]5-Carboxamidotryptamine binding to cloned 5-HT_1A

receptors. ^d[³H]5-HT uptake from rat synaptosomes. ^eExternal data obtained from Cerep, catalogue number 808-3h. ^fExternal data obtained from

MDS Pharmaservices, catalogue number 271110. ^g[³H]5-HT uptake from CHO cells expressing hSERT. For compound 5m, in-house and external data

are shown for h5-HT_3Aand h5-HT_1Areceptors, and for SERT both rat and human data are shown.

Table 6. Broad Screening of 5m at Cerep, MDS Pharmaservices, or H. Lundbeck A/S^a

Targets

adenosine: hA₁, hA_2A, hA₃

angiotensin II: hAT₁, hAT₂

benzodiazepine: BZD (central); BZD (peripheral)

bradykinin: hB₂

bombesin: BB (nonselective)

calcitonin: hCGRP

cannabinoid: hCB₁

cholecystokinin: hCCK₁, hCCK₂

endothelin: hET_A,hET_B,

dopamine: hD₁, hD₂, hD₃, hD_4.4, hD₅

γ-aminobutyric acid (nonselective)

growth factors: platelet-derived growth factor,

h interleukin-8B, tumor necrosis factor-R, chemokine

receptor 1

galanin: hGAL1, hGAL2

glutamate (nonselective): phenylcyclidine binding site

histamine: hH₁, hH₂(K_i= 180 nM)

melatonin: hML₁

melanocortin: hMC₄(% inh = 51)

muscarinic: hM₁, hM₂, hM₃, hM₄, hM₅

neurokinin: hNK₁, hNK₂, hNK₃

noradrenergic (nonselective): R₁(nonselective), R₂(nonselective), β₁(K_i= 46 nM),

β₂(K_i= 560 nM)

neuropeptide Y: hY₁, hY₂

neurotensin: hNT₁

opiate: hδ, κ, hμ

orphanin: hORL₁

pituitary adenylate cyclase activating polypeptide: PAC1

purinic: P2X, P2Y

prostaglandin: human thromboxane A/PGH₂, hPGI₂

serotonergic: h5-HT_1A(% inh = 98),^bh5-HT_1B(% inh = 100),^bh5-HT_2A, h5-HT_2C

(K_i= 180 nM), h5-HT_3A(% inh 100),^bh5-HT_5A(K_i= 220 nM), h5-HT₆

(K_i= 330 nM), h5-HT₇(K_i= 19 nM)

σ (nonselective)

somatostatin (nonselective): sst

vasoactive intestinal peptide: hVIP₁(VPAC₁)

ion channels: Ca^2þ(L, verapamil site), K^þv channel,

SK^þ_Cachannel, Na^þchannel (site 2), Cl^ꢀchannel

vasopression: hV_1a

transporters: hNE transporter (% inh = 88),^bhDA transporter

^aFor targets where % inhibition (inh) of binding at 1000 nM was greater than 50%, K_ivalues or % inhibition are speciﬁed. Data were generated in a high-

throughput proﬁle at Cerep 2002. Follow-up was made at Cerep, MDS Pharmaservices or at H. Lundbeck A/S. Data are the mean of a minimum of two

determinations covering 6 decades for each compound. K_ivalues were calculated from the ChengꢀPrusoﬀ equation (K_i= IC₅₀/(1 þ [L]/K_d)). ^bIC₅₀/

K_ivalues are reported elsewhere in manuscript.

human healthy volunteers (data not shown). Thus, compound

5m was indeed predicted to display promising pharmacokinetic

characteristics and low CYP450 enzyme inhibitory potential in

humans.

3214

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Figure 5. Eﬀects of administration of 5m (2.5, 5, and 10 mg/kg) or vehicle subcutaneous on extracellular 5-HT levels in the rat ventral hippocampus.

Data are expressed as the mean ( SEM. There was a dose and time-dependent increase in extracellular 5-HT following drug administration [F(27,159) =

7.081; P < 0.001]. Extracellular concentrations of 5-HT were signiﬁcantly increased relative to baseline from t = 60 until t = 120 min after administration of

2.5 mg/kg 5m and from t = 40 until t = 180 min after 5 or 10 mg/kg 5m (p < 0.05).

Table 7. Primary Human (h) in Vitro Binding Proﬁle of 5m

target

K_i(nM)

h5-HT_1A

h5-HT_1B

h5-HT_3A

h5-HT₇

hSERT

15

33

3.7

19

1.6

Figure 6. Eﬀects of 3-day treatment with vehicle or 5m (5 or 10 mg/kg

sc) on extracellular 5-HT levels in the rat medial prefrontal cortex

(mPFC) and SERT occupancy. Data are expressed as the mean ( SEM.

Three days of treatment with 5m (5 or 10 (mg/kg)/day) resulted in

signiﬁcantly higher basal levels of 5-HT in the mPFC compared to

animals treated with vehicle (p < 0.05).

caution when predicting the clinical potential of 5m. Additional

proﬁling revealed that compound 5m exerted high aﬃnity (K_i<

100 nM) at noradrenergic hβ₁(K_i= 46 nM), serotonergic

h5-HT_1B(K_i= 33 nM), and h5-HT₇(K_i= 19 nM) receptors.

In vitro functional characterization at these receptors showed

that 5m was a partial agonist at h5-HT_1Breceptors and an

antagonist at h5-HT₇receptors. Taken together, compound 5m

exerts a multimodal serotonergic action, by inhibiting SERT and

interacting with h5-HT_1A, h5-HT_1B, h5-HT₃, and h5-HT₇

receptors. Indeed, the action of an eﬃcacious 5-HT_1Areceptor

agonist is expected to rapidly desensitize the somatodendritic

5-HT_1Aautoreceptor⁴¹and simultaneously activate postsyn-

aptic 5-HT_1Areceptors.^42,17,18Moreover, the 5-HT₃receptor

antagonistic eﬀect of 5m may counteract the inhibitory tonus of

the 5-HT₃receptor on the noradrenergic and cholinergic sys-

tems, leading to a facilitating eﬀect on these systems,^19,20both of

which are known to be important in the pathology of

mood disorders.^43,44Additionally, we have shown that the

5-HT₃receptor antagonist ondansetron potentiates citalo-

pram-induced increases in 5-HT levels in the rat brain,⁴⁵

’ DISCUSSION AND CONCLUSION

A novel series of compounds with combined eﬀects on h5-

HT_3Aand h5-HT_1Areceptors and on the rSERT was discovered.

Compound 5m was identiﬁed as the lead compound from this

series, displaying the desired relative aﬃnities/potencies (K_i/

IC₅₀) between 5-HT_3A, 5-HT_1A, and SERT, as well as appro-

priate pharmacokinetic properties for further development.

Compound 5m met the criteria set for eﬃcacy at these targets,

displaying antagonistic properties at h5-HT_3Areceptors, agonis-

tic properties at h5-HT_1Areceptors, and inhibition of hSERT.

Whereas 5m showed comparable potencies at both rat and

human SERT, species diﬀerences were found at 5-HT_1Aand

5-HT_3Areceptors. Hence, rat models should be used with

3215

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

demonstrating that 5-HT₃receptor blockade also facilitates

5-HT release. While the aﬃnity of 5m for the noradrenergic

hβ₁receptor did not translate to an obvious pharmacology

in vivo, the h5-HT_1Band h5-HT₇receptor aﬃnity may con-

tribute to the in vivo pharmacological proﬁle. The 5-HT_1B

receptors are present on 5-HT nerve terminals, and a number

of studies have shown that stimulating 5-HT_1Breceptors leads to

a decrease in 5-HT output. Conversely, the combination of

SERT blockade and 5-HT_1Breceptor antagonism has been

shown to augment 5-HT levels in microdialysis studies.⁴⁶We

have demonstrated that 5m exerts a partial agonistic proﬁle at the

h5-HT_1Breceptors, which could translate to an inhibitory

control of 5-HT_1Breceptors in vivo, thus further substantiating

the potential of 5m in a clinical setting. Similarly, 5-HT₇

receptor blockade has been shown to be eﬃcacious in animal

models of depression, e.g., the mouse forced-swim test and tail-

suspension test.⁴⁷The combination of an SSRI with 5-HT₇

receptor blockade synergistically augments the levels of 5-HT in

the prefrontal cortex,⁴⁸suggesting that 5-HT₇receptor antagon-

ism alone or in combination with SERT blockade has antide-

pressant potential.

In this study we demonstrate that 5m aﬀects 5-HT levels in

both the ventral hippocampus and mPFC, two important brain

regions involved in regulation of mood.^49,50After acute admin-

istration in rat microdialysis experiments, compound 5m induced

a robust increase in extracellular 5-HT levels in the ventral

hippocampus. This increase is higher than those obtained after

acute administration of SSRIs⁵¹and SNRIs^52,53at doses that

maximally block SERT,⁵²suggesting that the combined pharma-

cological eﬀects of 5m in rats on SERT and 5-HT_1B, 5-HT₃, and

5-HT₇receptors lead to an augmentation of the extracellular

5-HT levels when compared to that of SSRIs or SNRIs. As

mentioned above, activation by 5-HT_1Areceptor agonists leads

to rapid desensitization of somatodendritic 5-HT_1Areceptors,⁴¹

suggesting an additional beneﬁcial eﬀect in humans, which might

be diﬃcult to model in rats because the r5-HT_1Areceptor aﬃnity

is low. Thus, the eﬀects of 5m on the 5-HT receptors may cancel

out inhibitory feedback eﬀects exerted on acute SERT

inhibition.⁵⁴In the rat microdialysis experiments, 3 days of

treatment with 5m resulted in signiﬁcant increases in extracel-

lular 5-HT levels at low SERT occupancy, again suggesting a

multimodal mode of action of 5m. This is in contrast to

treatment with SSRIs, where chronic treatment is generally

required to overcome the inhibitory feedback mechanisms and

to produce a positive 5-HT response in preclinical models.⁵⁵In

conclusion, these characteristics indicate that 5m is a novel

multimodal serotonergic compound that has the potential to

beneﬁt symptomatologies associated with mood disorders. Com-

pound 5m (Lu AA21004) is currently in clinical development for

MDD.

recorded on a Bruker Daltonics MicrOTOF operating with external

calibration with lithium formate (for ESI) and PEG200/PEG400/

PEG600 (1:1:1) [for atmospheric pressure photo ionization (APPI)].

Liquid chromatography/MS (LC/MS), solvent system A was water/

trifluoroacetic acid (TFA) (100:0.05) and B was acetonitrile/water/

TFA (95:5:0.035). Instruments were API 150EX and LC system

Shimadzu LC10ADvp/SLC-10Avp with Sedere Sedex75 ELS and

column operating at 60 °C. LC/MS method 1 involved the following:

column 30 mm ꢁ 4.6 mm Waters Symmetry C18 with 3.5 μm particles

operating at room temperature; linear gradient elution with 10% B to

100% B in 2.4 min at a flow rate of 3.3 mL/min. LC/MS method 2

involved the following: as above but instead with a 30 mm ꢁ 4.6 mm

Waters Atlantis column with 3 μm particles operating at 40 °C; linear

gradient elution with 2% B to 100% B in 2.4 min at a flow rate of 3.3 mL/

min. LC/MS method 3 involved the following: solvent system A

consisting of water/ammonia (99.9:0.1) and B consisting of acetoni-

trile/water/ammonia (94.9:5:0.1); instruments were Sciex API300 and

Waters Acquity UPLC with diode array detector (DAD) and evaporative

light scattering (ELS) (Waters); column 50 mm ꢁ 2.1 mm Waters BEH

with 1.7 μm particles operatomg at 60 °C. Linear gradient elution with

10% B to 100% B in 1 min at a flow rate of 1.2 mL/min. All

pharmacologically characterized compounds were >95% pure as deter-

mined by LC/MS, with the exception of 6c and 6d, which had ultraviolet

(UV) purities of 94% and 93%, respectively. These compounds were

also characterized by ¹H NMR and ¹³C NMR as well as either elemental

1

analysis or HRMS. Synthesis intermediates were characterized by H

NMR, ¹³C NMR, and HRMS data for new compounds, except for

synthesis intermediates 13b, 13c, and 13d for which no ¹³C NMR data

are reported due to rotamers.

General Procedure 1 for the Formation of Diaryl

Sulfides³¹. Potassium tert-butoxide (1.1 equiv) and a solution of an

aryl bromide/iodide (1.0 equiv, if this was a solid) were added to

Pd₂dba₃(2.5 mol %) and DPEphos (10 mol %). Subsequent toluene

(the amount required to make a ∼10% w/w solution of the aryl

bromide/iodide), the aryl bromide/iodide (1.0 equiv, if this was a

liquid), and the thiophenol (1.0 equiv) were added. The flask was

immersed in an oil bath preheated to 100 °C. The mixture was stirred at

this temperature until the reaction was complete (typically <2 h for aryl

iodides and overnight for aryl bromides). The crude mixture was loaded

onto a silica gel column and eluted with heptane/ethyl acetate to give the

products in >95% purity as determined by ¹H NMR.

General Procedure 2 for the Formation of Boc-Protected

Arylpiperazines from Aryl Bromides⁵⁶. Aryl bromide (1.0 equiv,

if this was a solid), Boc-piperazine (1.2 equiv), and sodium tert-butoxide

(1.4 equiv) were added to Pd₂dba₃(2.5 mol %) and racemic BINAP (7.5

mol %). Subsequently toluene (the amount required to make a ∼10%

w/w solution of the aryl bromide) and the aryl bromide (1.0 equiv, if this

was a liquid) were added. The flask was immersed in an oil bath

preheated to 100 °C. The mixture was stirred at this temperature

overnight. The crude mixture was loaded onto a silica gel column and

eluted with heptane/ethyl acetate to give the products in >95% purity as

determined by ¹H NMR.

General Procedure 3 for the Cleavage of Boc Groups. The

substrate was dissolved in methanol (the amount required to make a

solution of approximately 20% w/w of the substrate). An HCl solution

(2 M) in Et₂O (approximately 3 equiv) was added to the mixture, and

the resulting mixture was stirred at room temperature overnight. If

required, the mixture was cooled in an ice/water bath and diluted with

Et₂O to precipitate the product as the hydrochloride salt. No attempts

were made to isolate additional product from the filtrate.

The synthesis procedure and compound characterization data for the

ﬁnal compounds in Table 3 are summarized below. The synthesis and

characterization data for the remaining ﬁnal compounds and all inter-

mediates are in the Supporting Information.

’ EXPERIMENTAL SECTION

Chemistry. Reagents and solvents were obtained from commercial

sources and used as received. Thin layer chromatography was carried out

on Merck silica gel 60 F254. Flash chromatography was performed using

Merck silica gel 60 (40ꢀ63 μm). ¹H and ¹³C nuclear magnetic

resonance (NMR) spectra were recorded at ambient temperature on a

Bruker Avance AV-500 at 500.13 and 125.77 MHz, respectively.

Chemical shifts (δ) are reported in parts per million relative to

tetramethylsilane (TMS) or residual solvent, and coupling constants

(J) are given in hertz. High-resolution mass spectrometry (HRMS) was

3216

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

Compounds Prepared According to Method

A

1-[2-(2,4-Dimethylphenylsulfanyl)phenyl]piperazine Hydrochlor-

ide (5m). 5m was prepared according to general procedure 3 starting

from intermediate 12m in a yield of 78%. ¹H NMR (500 MHz, DMSO-

d₆) δ 9.39 (s, 2H), 7.33 (d, J = 7.7, 1H), 7.24 (s, 1H), 7.17ꢀ7.07 (m,

3H), 6.96 (dd, J = 7.6, 6.0, 1H), 6.41 (d, J = 7.8, 1H), 3.21 (broad s, 8H),

2.31 (s, 3H), 2.24 (s, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 148.22,

142.04, 139.68, 136.11, 133.74, 132.14, 128.46, 127.19, 126.40, 126.13,

125.46, 120.64, 48.47 (2C), 43.67 (2C), 21.10, 20.47. HRMS calcd for

C₁₈H₂₂N₂S þ H, 299.1576; found, 299.1584. LC/MS (method 1): t_R=

(Scheme 1). These compounds were prepared by initial formation

of 2-bromophenylpiperazines accordoing to general method 2 followed

by formation of the thioether linkage according to general procedure 1.

Deprotection and salt formation were done according to general method 3.

1-[2-(4-Chlorophenylsulfanyl)phenyl]piperazine

Hydrochloride

(4h). 4h was prepared according to general procedure 3 starting from

intermediate 11h in a yield of 31%. ¹H NMR (500 MHz, DMSO-d₆) δ

9.20 (s, 2H), 7.48 (d, J = 6.4, 2H), 7.42ꢀ7.34 (m, 2H), 7.27 (t, J = 7.2,

1H), 7.20 (d, J = 7.7, 1H), 7.06 (d, J = 7.0, 1H), 6.92 (d, J = 7.7, 1H),

3.22ꢀ3.08 (m, 8H). ¹³C NMR (126 MHz, DMSO-d₆) δ 149.73, 134.19

(2C), 133.10, 132.69, 131.70, 130.14, 130.01 (2C), 128.27, 125.58,

121.10, 48.54 (2C), 43.62 (2C). HRMS calcd for C₁₆H₁₇ClN₂S þ H,

305.0874; found, 305.0876. LC/MS (method 1): t_R= 0.98 min, UV

1.02 min, UV purity 97%, ELS purity 100%. Anal. (C₁₈H₂₂N₂S HCl)

3

C, H, N.

Compounds Prepared According to Method

B

(Scheme 2). These compounds were prepared by initial formation

of brominated diarylsulfides according to general procedure 1 followed

by aryl amination under the conditions described for general procedure

2. Deprotection and salt formation were done according to general

method 3.

purity 97%, ELS purity 100%. Anal. (C₁₆H₁₇ClN₂S HCl) C, H, N.

3

1-(2-p-Tolylsulfanylphenyl)piperazine Hydrochloride (4j). 4j was

prepared according to general procedure 3 from impure 4-[2-(4-

methylphenylsulfanyl)phenyl]piperazine-1-carboxylic acid tert-butyl es-

ter (11j) in a yield of 60%. 11j was prepared starting from 7a and

4-methylthiophenol according general procedure 1. ¹H NMR (500

MHz, DMSO-d₆) δ 9.21 (s, 2H), 7.35 (d, J = 8.0, 2H), 7.28 (d, J =

8.0, 2H), 7.22ꢀ7.13 (m, 2H), 7.05ꢀ6.96 (m, 1H), 6.69 (d, J = 7.6, 1H),

3.18 (broad s, 8H), 2.34 (s, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ

148.51, 138.82, 134.33 (2C), 134.13, 130.91 (2C), 128.69, 127.86,

126.94, 125.44, 120.69, 48.53 (2C), 43.67 (2C), 21.14. HRMS calcd for

C₁₇H₂₀N₂S þ H, 285.1420; found, 285.1422. LC/MS (method 1): t_R=

1-[2-(2,3-Dimethylphenylsulfanyl)phenyl]piperazine Hydrochlor-

ide (5j). 5j was prepared according to general procedure 3 starting from

intermediate 12j in a yield of 48%. ¹H NMR (500 MHz, DMSO-d₆) δ

9.50 (s, 2H), 7.29ꢀ7.25 (m, 2H), 7.15 (t, J = 6.4, 3H), 6.97 (dt, J = 8.3,

4.3, 1H), 6.48 (d, J = 7.8, 1H), 3.25ꢀ3.17 (m, 8H), 2.31 (s, 3H), 2.25 (s,

3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 148.42, 140.37, 138.43,

133.54, 133.31, 131.23, 131.10, 127.03, 126.97, 126.63, 125.46,

120.72, 48.48 (2C), 43.61 (2C), 20.92, 17.16. HRMS calcd for

C₁₈H₂₂N₂S þ H, 299.1576; found, 299.1582. LC/MS (method 1): t_R=

1.05 min, UV purity 98%, ELS purity 100%.

1-[2-(2-Chloro-4-methylphenylsulfanyl)phenyl]piperazine Hydro-

chloride (5p). 5p was prepared according to general procedure 3 starting

from intermediate 12p in a yield of 44%. ¹H NMR (500 MHz, DMSO-d₆)

δ 9.47 (s, 2H), 7.48 (s, 1H), 7.35ꢀ7.25 (m, 4H), 7.04 (t, J = 7.5, 1H),

6.68 (d, J = 7.8, 1H), 3.24ꢀ3.07 (m, 8H), 2.34 (s, 3H). ¹³C NMR (126

MHz, DMSO-d₆) δ 149.33, 140.95, 136.48, 135.19, 131.47, 131.04,

129.49, 128.72, 128.27, 127.85, 125.60, 121.07, 48.44 (2C), 43.58 (2C),

20.74. HRMS calcd for C₁₇H₁₉ClN₂S þ H, 319.1030; found, 319.1032.

LC/MS (method 1): t_R= 1.04 min, UV purity 98%, ELS purity 100%.

Compound Prepared According to Method C. Final com-

pound 6g was prepared by aromatic nucleophilic substitution of

2-fluoronitrobenzene with 2,4-dimethylphenol leading to 2,4-dimeth-

yl-1-(2-nitrophenoxy)benzene (14) that was reduced to 2-(2,4-di-

methylphenoxy)phenylamine (15), which was reacted further as

described below.

0.96 min, UV purity 98%, ELS purity 100%. Anal. (C₁₇H₂₀N₂S HBr) C,

3

H, N for a batch of the hydrobromide salt.

1-[2-(3,4-Dimethylphenylsulfanyl)phenyl]piperazine Hydrochlor-

ide (5g). 5g was prepared according to general procedure 3 from impure

4-[2-(3,4-dimethylphenylsulfanyl)phenyl]piperazine-1-carboxylic acid

tert-butyl ester (12g) in a yield of 70% over two steps. 12g was prepared

starting from 7a and 3,4-dimethylthiophenol according general proce-

dure 1. ¹H NMR (500 MHz, DMSO-d₆) δ 9.38 (s, 2H), 7.30ꢀ7.10 (m,

5H), 7.05ꢀ6.94 (m, 1H), 6.66 (d, J = 7.8, 1H), 3.18 (broad s, 8H), 2.25

(s, 3H), 2.23 (s, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 153.57, 143.74,

143.01, 140.66, 139.74, 137.36, 136.54, 133.80, 132.80, 131.96, 130.65,

125.84, 53.75 (2C), 48.87 (2C), 24.85, 24.75. HRMS calcd for

C₁₈H₂₂N₂S þ H, 299.1576; found, 299.1586. LC/MS (method 1): t_R=

1.03 min, UV purity 97%, ELS purity 100%.

1-[2-(2,3-Dichlorophenylsulfanyl)phenyl]piperazine Hydrochloride

(5i). 5i was prepared according to general procedure 3 starting from

impure 4-[2-(2,3-dichlorophenylsulfanyl)phenyl]piperazine-1-carboxylic

acid tert-butyl ester (12i) in a yield of 6%. 12i was prepared from 7a and

2,3-dichlorothiophenol according general procedure 1. ¹H NMR (500

MHz, DMSO-d₆) δ 8.98 (s, 2H), 7.59 (dd, J = 8.0, 1.1, 1H), 7.40 (dd,

J = 10.9, 4.3, 1H), 7.34ꢀ7.24 (m, 2H), 7.19ꢀ7.07 (m, 2H), 7.03ꢀ6.97

(m, 1H), 3.22ꢀ3.13 (m, 4H), 3.04 (s, 4H). ¹³C NMR (126 MHz,

DMSO-d₆) δ 150.34, 135.99, 131.80, 130.41, 129.20, 129.01, 128.38,

128.12, 127.25, 124.81, 120.70, 112.10, 47.62 (2C), 42.72 (2C). HRMS

calcd for C₁₆H₁₆Cl₂N₂S þ H, 339.0484; found, 339.0486. LC/MS

(method 1): t_R= 1.03 min, UV purity 99%, ELS purity 100%.

1-[2-(2,4-Dichlorophenylsulfanyl)phenyl]piperazine Hydrochloride

(5l). 5l was prepared according to general procedure 3 starting from

impure 4-[2-(2,4-dichlorophenylsulfanyl)phenyl]piperazine-1-carboxylic

acid tert-butyl ester (12l) in a yield of 6%. 12l was prepared from 7a and

2,4-dichlorothiophenol according general procedure 1. ¹H NMR (500

MHz, DMSO-d₆) δ 9.66 (s, 2H), 8.33 (s, 1H), 8.00ꢀ7.85 (m, 2H), 7.79

(d, J = 7.9, 1H), 7.74ꢀ7.63 (m, 2H), 7.53 (d, J = 7.8, 1H), 3.72 (s, 4H),

3.62 (s, 4H). ¹³C NMR (126 MHz, DMSO-d₆) δ 150.69, 135.51, 133.93,

133.22, 132.68, 131.55, 129.90, 129.39, 128.95, 128.69, 125.79, 121.56,

48.52 (2C), 43.66 (2C). HRMS calcd for C₁₆H₁₆Cl₂N₂S þ H, 339.0484;

found, 339.0492. LC/MS (method 1): t_R= 1.07 min, UV purity 96%, ELS

purity 100%.

1-[2-(2,4-Dimethylphenoxy)phenyl]piperazine Hydrochloride (6g).

Compound 15 (2.13 g) and bis(2-bromoethyl)amine hydrobromide⁵⁷

(3.89 g) were suspended in chlorobenzene (50 mL). The mixture was

boiled under reflux for 4 h. The solvent was evaporated off, and the

residual oil was partitioned between water (100 mL) and ethyl acetate

(50 mL). The aqueous layer was basified and extracted with ethyl acetate

(2 ꢁ 50 mL). The combined organic layers were washed with brine

(100 mL), dried over MgSO₄, filtered, and concentrated in vacuo to

afford an oil. This material was purified by column flash chromatography

(eluent EtOAc/MeOH 9:1 f EtOAc/MeOH/Et₃N 9:1:1) to afford the

crude title compound. This material was partitioned between water/

brine and methylene chloride. The organic layer was dried over MgSO₄,

filtered, and concentrated in vacuo to afford the free base. This material

was dissolved in ethyl acetate and treated with 2 M HCl in Et₂O to

1

precipitate the title compound (0.83 g, 26%). H NMR (500 MHz,

DMSO-d₆) δ 9.54 (s, 2H), 7.13ꢀ7.08 (m, 2H), 7.06 (t, J = 7.4, 1H),

7.00ꢀ6.92 (m, 2H), 6.69ꢀ6.60 (m, 2H), 3.31 (s, 4H), 3.08 (s, 4H), 2.24

(s, 3H), 2.16 (s, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 150.17,

147.60, 139.38, 130.82, 130.23, 126.23, 126.01, 122.14, 122.01, 118.10,

116.47, 116.07, 45.41 (2C), 41.24 (2C), 18.66, 14.25. HRMS calcd for

C₁₈H₂₂N₂O þ H, 283.1805; found, 283.1818. LC/MS (method 1): t_R=

0.98 min, UV purity 100%, ELS purity 98%.

3217

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

In Vitro Pharmacology. The assays used for the SAR studies are

described below. Details and all other pharmacology are described in

Supporting Information. In general for binding affinity assays, data

points were expressed as percent of the specific binding, and the IC₅₀

values were determined by nonlinear regression analysis using a

sigmoidal variable slope curve fitting. The dissociation constant (K_i)

was calculated from the ChengꢀPrusoff equation (K_i= IC₅₀/(1 þ ([L]/

K_d)), where the concentration of free radioligand L is approximated to

the concentration of added radioligand in the assay, and K_dequals the

affinity of the radioligand to the receptor. In general IC₅₀values are

based on 6ꢀ8 different concentrations covering 4ꢀ6 decades.

h5-HT_3ABinding Affinity Assay. [³H]Granisetron (1 nM)

(Perkin-Elmer) binding to h5-HT_3Areceptors was evaluated in mem-

branes from a HEK-293 cell line stably expressing the hHT_3Areceptor.

Nonspecific binding was measured in the presence of 10 μM bemese-

tron. After incubation, the membrane preparations were rapidly filtered

under a vacuum using glass-fiber filters. Compounds were tested at least

three times over a 6 log concentration range. IC₅₀values were

determined by nonlinear regression analysis using Hill equation curve

fitting, and K_ivalues were calculated based on the ChengꢀPrusoff

equation.

compound over the incubation time. The test compounds were in-

cubated at 1 μM with human and rat microsomes (BD Biosciences) for

60 min, using NADPH as cofactor. An NADPH regenerating system

containing NADP^þ, glucose 6-phosphate dehydrogenase, glucose

6-phosphate, sodium citrate, and MgCl₂(cofactor mix) was used as a

source of NADPH. Human and rat microsomes were thawed at room

temperature, and cofactor mix was added. The mixture was stirred on a

vortex and put in a water bath at 37 °C for 10 min. The reaction was

initiated by adding test compound (final concentration 1 μM, 0.5 mg of

protein/mL). Separate incubations were made for each test compound

and time point. The samples were incubated for 0, 5, 15, 30, and 60 min,

and the reactions were stopped by adding 100 μL of acetonitrile and

transferred to a 96-well stop plate with seals. The stop plates were

centrifuged for 10 min at 3300 rpm and 4 °C before being analyzed by

liquid chromatography coupled to a tandem mass spectrometer

(LCꢀMS/MS, Waters QuattroMicro, Manchester, U.K.).

Plasma Binding by Equilibrium Dialysis Measurements.

The method was a modification of that reported by Kalvass et al.⁵⁸

Briefly, a 96-well equilibrium dialysis apparatus was used to determine

the free fraction in the plasma (HT Dialysis LLC, Gales Ferry, CT, U.S.).

Membranes (3 kDa cutoff) were conditioned in deionized water for 40

min, followed by conditioning in 80:20 deionized water/ethanol for 20

min, and then rinsed in deionized water before use. Plasma was diluted

1:1 with PBS. Diluted plasma was spiked with the 1 μM test compound

and loaded into the 96-well equilibrium dialysis plate. Dialysis versus

PBS (100 uL) was carried out for 5 h in a temperature controlled

incubator at 37 °C, before being analyzed by liquid chromatography

coupled to a tandem mass spectrometer (LCꢀMS/MS, Waters Quat-

troMicro, Manchester, U.K.).

In Vivo Pharmacokinetics in Rats. In vivo pharmacokinetics was

studied in rats following single intravenous (1 mg/kg) or oral (2 mg/kg)

administration of compounds. Serial blood samples were collected at

various time points up to 6 h after dosing. In vivo pharmacokinetic

parameters were obtained following compartmental modeling. Bioana-

lysis of samples were analyzed by liquid chromatography coupled to a

tandem mass spectrometer (LCꢀMS/MS, Waters QuattroMicro, Man-

chester, U.K.).

Microdialysis in Rats. Rats (300ꢀ350 g; Harlan, Horst, The

Netherlands) were used for the experiments. For acute administration

5m was dissolved in 10% hydroxypropyl-β-cyclodextrin and adminis-

tered subcutaneously (sc) in a volume of 1 mL/kg. For 3-day treatment

with 5m minipumps were used to deliver the solutions. Under isoflurane

anesthesia osmotic minipumps (2ML2, Alzet, U.S.) were implanted in a

subcutaneous pocket created on the right side parallel to the spine of the

animal. By stereotaxic surgery I-shaped microdialysis probes (Hospal

AN 69 membrane, 4 mm exposed surface; BrainLink) were inserted into

the ventral hippocampus or medial prefrontal cortex under isoflurane

anesthesia. Microdialysis experiments were performed 24ꢀ48 h after

surgery by perfusing the probes with artificial CSF (147 mM NaCl,

3.0 mM KCl, 1.2 mM CaCl₂, and 1.2 mM MgCl₂) at a flow rate of 1.5

μL/min. Microdialysis samples were collected at 20 or 30 min intervals

into minivials. After the experiment the rats were sacrificed and the

position of each probe was histologically verified by making coronal

sections of the brain.

Serotonin Analysis. Aliquots were injected onto a HPLC column.

Chromatographic separation was performed using an isocratic mobile

phase. Mobile phase was run through the system at a flow rate of 0.4 mL/

min, and 5-HT was detected electrochemically at þ500 mV. For acute

experiments four consecutive pretreatment samples were taken as

baseline and their mean concentration was set to 100%. Drug effects

were expressed as percentages of basal level (mean ( SEM) within the

same animal. Moreover, differences between basal outputs of 5-HT

(fmol/30 min sample) after treatment with 5m or vehicle for 3 days were

compared.

h5-HT_1ABinding Affinity Assay. [³H]5-Carboxamidotrypta-

mine (0.15 nM) (Perkin-Elmer) binding to cloned human 5-HT_1A

receptors in CHO cells or HeLa was assayed in a membrane-based

scintillation proximity assay. The level of nonspecific binding was

obtained in the presence of 10 μM 5-HT or metergoline. Compounds

were tested at least three times over a 6 log concentration range. IC₅₀

values were determined by nonlinear regression analysis using Hill

equation curve fitting, and K_ivalues were calculated based on the

ChengꢀPrusoff equation.

Inhibition of [³H]NE, [³H]DA, and [³H]5-HT Uptake by Rat

Synaptosomes. NET, DAT, and SERT transporter inhibition (IC₅₀)

in rat synaptosomes was measured using [³H]NE, [³H]DA, and [³H]5-

HT, respectively. After incubation, the membrane preparations were

rapidly filtered under a vacuum using glass-fiber filters. The filters were

washed with ice-cold buffer using a harvester. Bound radioactivity was

measured by scintillation counting using a scintillation cocktail. Com-

pounds were tested at least twice over a 6 log concentration range. IC₅₀

values were determined by nonlinear regression analysis using Hill

equation curve fitting.

Cytochrome P450 Enzyme Inhibition. CYP enzyme inhibition

was performed using recombinant enzymes (Supersome, Gentest),

expressing individual CYP enzymes (CYP1A2, CYP2C9, CYP2D6

and CYP3A4). The inhibition of CYP enzymes by the test compounds

was assessed by incubating each enzyme with its substrate (7-ethoxy-3-

cyanocoumarin was used for CYP1A2 (10 μM); dibenzylfuorescein (2

μM) was used for CYP2C9; resorufin benzyl ether (100 μM) was used

for CYP3A4; and 3-[2-(N,N-diethyl-N-methylamino)ethyl]-7-methoxy-

4-methylcoumarin (3 μM) was used for CYP2D6) alone or in the

presence of the test compounds at eight different concentrations

(40ꢀ0.02 μM). All incubations (28 min for CYP1A2, 30 min for

CYP3A4, and 45 min for CYP2C9 and CYP2D6) were performed in

phosphate buffer with an NADPH regenerating system containing

NADP^þ, glucose 6-phosphate dehydrogenase, glucose 6-phosphate,

sodium citrate and MgCl₂. At the end of the incubation period, the

amount of metabolite formed was measured using a fluorescence

spectrophotometer. The excitation and emission wavelengths in

nanometers are 405 (ex) and 465 (em) for CYP1A2 and CYP2D6,

485 (ex) and 535 (em) for CYP2C9, and 530 (ex) and 590 (em) for

CYP3A4. Inhibitions were expressed relative to controls (without

inhibitors), and IC₅₀values were calculated using Microsoft Excel

software.

Microsomal Stability Determination. The microsomal intrin-

sic clearance was determined by assessing the elimination of test

3218

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

SERT Occupancy in Rats. After the microdialysis experiments rats

were sacrificed and the brains were removed and stored at ꢀ80 °C until

cutting. Sections of 20 μm containing the striatum were cut and air-dried

before the binding experiment. For SERT occupancy 0.5 nM [³H]DASB

was used as radioligand. The control for nonspecific binding was 1 μM

escitalopram. The binding buffer consisted of 50 mM Tris, 150 mM

NaCl, and 5 mM KCl, pH 7.4, and the incubation time was 90 min. Slides

were washed in the respective buffers at 4 °C three times for 5 min, then

briefly dipped in distilled water and air-dried. The slides were analyzed in

a Biospace β-imager 2000 for at least 6 h.

C.; Tylee, A. Prescribing patterns of antidepressants in Europe: results

from the Factors Inﬂuencing Depression Endpoints Research

(FINDER) study. Eur. Psychiatry 2008, 23, 66–73.

(4) Rosenzweig-Lipson, S.; Beyer, C. E.; Hughes, Z. A.; Khawaja, X.;

Rajarao, S. J.; Malberg, J. E.; Rahman, Z.; Ring, R. H.; Schechter, L. E.

Diﬀerentiating antidepressants of the future: eﬃcacy and safety. Phar-

macol. Ther. 2007, 113, 134–153.

(5) Papakostas, G. I.; Thase, M. E.; Fava, M.; Nelson, J. C.; Shelton,

R. C. Are antidepressant drugs that combine serotonergic and nora-

drenergic mechanisms of action more eﬀective than the selective

serotonin reuptake inhibitors in treating major depressive disorder? A

meta-analysis of studies of newer agents. Biol. Psychiatry 2007,

62, 1217–1227.

’ ASSOCIATED CONTENT

(6) Trivedi, M. H.; Rush, A. J.; Wisniewski, S. R.; Nierenberg, A. A.;

Warden, D.; Ritz, L.; Norquist, G.; Howland, R. H.; Lebowitz, B.;

McGrath, P. J.; Shores-Wilson, K.; Biggs, M. M.; Balasubramani, G. K.;

Fava, M. Evaluation of outcomes with citalopram for depression using

measurement-based care in STAR*D: implications for clinical practice.

Am. J. Psychiatry 2006, 163, 28–40.

(7) Moltzen, E. K.; Bang-Andersen, B. Serotonin reuptake inhibi-

tors: the corner stone in treatment of depression for half a century—a

medicinal chemistry survey. Curr. Top. Med. Chem. 2006, 6, 1801–1823.

(8) Mathew, S. J.; Manji, H. K.; Charney, D. S. Novel drugs and

therapeutic targets for severe mood disorders. Neuropsychopharmacology

2008, 33, 2080–2092.

S

Supporting Information. Experimental procedures for

b

synthesis of intermediates and target compounds; combustion

analysis and HRMS data; descriptions for in vitro and in vivo

assays; pK_i( SEM values for h5-HT_3Aand h5-HT_1Areceptors;

and pIC₅₀( SD/SEM values for rSERT, rNET, and rDAT. This

material is available free of charge via the Internet at http://pubs.

acs.org.

’ AUTHOR INFORMATION

Corresponding Author

*Phone: þ45 3643 3247. Fax: þ45 3643 8237. E-mail:

(9) Artigas, F.; Perez, V.; Alvarez, E. Pindolol induces a rapid

improvement of depressed patients treated with serotonin reuptake

inhibitors. Arch. Gen. Psychiatry 1994, 51, 248–251.

ban@lundbeck.com.

(10) Artigas, F.; Adell, A.; Celada, P. Pindolol augmentation of

antidepressant response. Curr. Drug Targets 2006, 7, 139–147.

(11) Morphy, R.; Rankovic, Z. Designed multiple ligands. An

emerging drug discovery paradigm. J. Med. Chem. 2005, 48, 6523–6543.

(12) Millan, M. J. Dual- and triple-acting agents for treating core and

co-morbid symptoms of major depression: novel concepts, new drugs.

Neurotherapeutics 2009, 6, 53–77.

(13) Cremers, T. I. F. H.; Wikstr€om, H. V.; Den Boer, J. A.; Bosker,

F. J.; Westerink, B. H.; Bøgesø, K. P.; Hogg, S. The Combination of a

Serotonin Reuptake Inhibitor and a 5-HT_2CAntagonist, Inverse Agonist

or Partial Agonist. WO 200141701 (A2), 2001.

’ ACKNOWLEDGMENT

We thank Drs. Kim Andersen and Ejner K. Moltzen for

valuable input regarding design of compounds and Dr. Klaus P.

Bøgesø for helpful input in the discovery phase and insightful

discussions to improve the manuscript.

’ ABBREVIATIONS USED

AcCh, acetylcholine; APPI, atmospheric pressure photo-

ionization; Boc, tert-butyloxycarbonyl; CHO, Chinese ham-

ster ovary; CYP450, cytochrome P450; DA, dopamine; DAD,

diode array detector; DAT, dopamine transporter; DML, de-

signed multiple ligand; ELS, evaporative light scattering;

GABA, γ-aminobutyric acid; GAD, generalized anxiety disor-

der; GTPγS, guanosine 5⁰-O-(3-thio)triphosphate; HEK, hu-

man embryonic kidney; HRMS, high-resolution mass spectro-

metry; MAOI, monoamine oxidase inhibitor; MDD, major de-

pressive disorder; mPFC, medial prefrontal cortex; NE, norepi-

nephrine; NET, norepinephrine transporter; SAR, structureꢀ

activity relationship; SD, standard deviation; SEM, standard er-

ror of the mean; SERT, serotonin transporter; SNRI, serotonin

and norepinephrine reuptake inhibitor; SSRI, selective serotonin

reuptake inhibitor; TCA, tricyclic antidepressant; TFA, triﬂuor-

oacetic acid

(14) Cremers, T. I.; Giorgetti, M.; Bosker, F. J.; Hogg, S.; Arnt, J.;

Mørk, A.; Honig, G.; Bøgesø, K. P.; Westerink, B. H.; den Boer, H.;

Wikstr€om, H. V.; Tecott, L. H. Inactivation of 5-HT_2Creceptors

potentiates consequences of serotonin reuptake blockade. Neuropsycho-

pharmacology 2004, 29, 1782–1789.

(15) Nutt, D. J. Beyond psychoanaleptics—can we improve anti-

depressant drug nomenclature?. J. Psychopharmacol. 2009, 23, 343–345.

(16) Chang, T.; Fava, M. The future of psychopharmacology of

depression. J. Clin. Psychiatry 2010, 71, 971–975.

(17) Blier, P.; Ward, N. M. Is there a role for 5-HT_1Aagonists in the

treatment of depression?. Biol. Psychiatry 2003, 53, 193–203.

(18) Blier, P.; Bergeron, R.; de Montigny, C. Selective activation of

postsynaptic 5-HT_1Areceptors induces rapid antidepressant response.

Neuropsychopharmacology 1997, 16, 333–338.

(19) Giovannini, M. G.; Ceccarelli, I.; Molinari, B.; Cecchi, M.;

Goldfarb, J.; Blandina, P. Serotonergic modulation of acetylcholine

release from cortex of freely moving rats. J. Pharmacol. Exp. Ther.

1998, 285, 1219–1225.

(20) Matsumoto, M.; Yoshioka, M.; Togashi, H.; Tochihara, M.;

Ikeda, T.; Saito, H. Modulation of norepinephrine release by seroto-

nergic receptors in the rat hippocampus as measured by in vivo

microdialysis. J. Pharmacol. Exp. Ther. 1995, 272, 1044–1051.

(21) Morales, M.; Battenberg, E.; de Lecea, L.; Bloom, F. E. The type

3 serotonin receptor is expressed in a subpopulation of GABAergic

neurons in the rat neocortex and hippocampus. Brain Res. 1996,

731, 199–202.

’ REFERENCES

(1) Kessler, R. C.; Berglund, P.; Demler, O.; Jin, R.; Merikangas,

K. R.; Walters, E. E. Lifetime prevalence and age-of-onset distributions of

DSM-IV disorders in the National Comorbidity Survey Replication.

Arch. Gen. Psychiatry 2005, 62, 593–602.

(2) Murray, C. J. L.; Lopez, A. D. Evidence-based health policy:

lessons from the Global Burden of Disease Study. Science 1996,

274, 740–743.

(22) Yan, Z. Regulation of GABAergic inhibition by serotonin

signaling in prefrontal cortex: molecular mechanisms and functional

implications. Mol. Neurobiol. 2002, 26, 203–216.

(3) Bauer, M.; Monz, B. U.; Montejo, A. L.; Quail, D.; Dantchev, N.;

Demyttenaere, K.; Garcia-Cebrian, A.; Grassi, L.; Perahia, D. G. S.; Reed,

3219

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

(23) Angel, I.; Schoemaker, H.; Prouteau, M.; Garreau, M.; Langer,

S. Z. Litoxetine: a selective 5-HT uptake inhibitor with concomitant

5-HT₃receptor antagonist and antiemetic properties. Eur. J. Pharmacol.

1993, 232, 139–145.

(24) Bergeron, R.; Blier, P. Cisapride for the treatment of nausea

produced by selective serotonin reuptake inhibitors. Am. J. Psychiatry

1994, 151, 1084–1086.

(25) Gray, D. L.; Xu, W.; Campbell, B. M.; Dounay, A. B.; Barta, N.;

Boroski, S.; Denny, L.; Evans, L.; Stratman, N.; Probert, A. Discovery

and pharmacological characterization of aryl piperazine and piperidine

ethers as dual acting norepinephrine reuptake inhibitors and 5-HT1A

partial agonists. Bioorg. Med. Chem. Lett. 2009, 19, 6604–6607.

(26) Ruhland, T.; Bang, K. S.; Andersen, K. Iron-assisted nucleo-

philic aromatic substitution on solid phase. J. Org. Chem. 2002,

67, 5257–5268.

(27) Ruhland, T.; Smith, G. P.; Bang-Andersen, B.; Pueschl, A.;

Moltzen, E. K.; Andersen, K. Phenyl-piperazine Derivatives as Serotonin

Reuptake Inhibitors. WO 2003029232 (A1), 2003.

(28) Louie, J.; Hartwig, J. F. Palladium-catalyzed synthesis of aryla-

mines from aryl halides: mechanistic studies lead to coupling in the

absence of tin reagents. Tetrahedron Lett. 1995, 36, 3609–3612.

(29) Guram, A. S.; Rennels, R. A.; Buchwald, S. L. A simple catalytic

method for the conversion of aryl bromides to arylamines. Angew. Chem.,

Int. Ed. Engl. 1995, 34, 1348–1350.

(40) Moore, N. A.; Bang-Andersen, B.; Brennum, L. T.; Frederiksen,

K.; Hogg, S.; Mørk, A.; Sanchez, C.; Smith, D. G.; Stensbøl, T. B.; Zhong,

H. Lu AA21004: A Novel Potential Treatment for Mood Disorders.

Presented at the 21st European College of Neuropsychopharmacology

(ECNP), Barcelona, Spain, August 30ꢀSeptember 3, 2008; Poster

P.2.b.015. Moore, N. A.; Bang-Andersen, B.; Brennum, L. T.; Freder-

iksen, K.; Hogg, S.; Mørk, A.; Sanchez, C.; Smith, D. G.; Stensbøl,

T. B.; Zhong, H. Neuropsychopharmacology 2008, 18 (suppl. 4),

S321.

(41) Assiꢀe, M.-B.; Lomenech, H.; Ravailhe, V.; Faucillon, V.; New-

man-Tancredi, A. Rapid desensitization of somatodendritic 5-HT_1A

receptors by chronic administration of the high-eﬃcacy 5-HT_1Aagonist,

F13714: a microdialysis study in the rat. Br. J. Pharmacol. 2006,

149, 170–178.

(42) Haddjeri, N.; Blier, P.; de Montigny, C. Long-term antidepres-

sant treatments result in a tonic activation of forebrain 5-HT_1A

receptors. J. Neurosci. 1998, 18, 10150–10156.

(43) Brunello, N.; Blier, P.; Judd, L. L.; Mendlewicz, J.; Nelson, C. J.;

Souery, D.; Zohar, J.; Racagni, G. Noradrenaline in mood and anxiety

disorders: basic and clinical studies. Int. Clin. Psychopharmacol. 2003,

18, 191–202.

(44) Laje, R. P.; Berman, J. A.; Glassman, A. H. Depression and

nicotine: preclinical and clinical evidence for common mechanisms.

Curr. Psychiatry Rep. 2001, 3, 470–474.

(30) Larsen, S. B.; Bang-Andersen, B.; Johansen, T. N.; Jørgensen,

M. Palladium-catalyzed monoamination of dihalogenated benzenes.

Tetrahedron 2008, 64, 2938–2950.

(31) Schopfer, U.; Schlapbach, A. A general palladium-catalysed

synthesis of aromatic and heteroaromatic thioethers. Tetrahedron

2001, 57, 3069–3073.

(45) Mørk, A.; Brennum, L. T.; Fallon, S.; Bisulco, S.; Frederiksen,

K.; Bang-Andersen, B.; Lassen, A. B.; Zhong, H.; Patel, J. G.; Poon, P.;

Smith, D. G.; Sanchez, C.; Stensbøl, T. B.; Hogg, S. Pharmacological

Proﬁle of Lu AA21004, a Novel Multi-Target Drug for the Treatment of

Mood Disorders. Presented at the Society for Neuroscience Meeting,

Chicago, IL, October 17ꢀ21, 2009; Poster.

(46) de Groote, L.; Klompmakers, A. A.; Olivier, B.; Westenberg,

H. G. M. Role of extracellular serotonin levels in the eﬀect of 5-HT_1B

receptor blockade. Psychopharmacology 2003, 167, 153–158.

(47) Sarkisyan, G.; Roberts, A. J.; Hedlund, P. B. The 5-HT₇

receptor as a mediator and modulator of antidepressant-like behavior.

Behav. Brain Res. 2010, 209, 99–108.

(32) Lenox, R. H.; Frazer, A. Mechanism of Action of Antidepres-

sants and Mood Stabilizers. In Neuropsychopharmacology: The Fifth

Generation of Progress: An Oﬃcial Publication of the American College of

Neuropsychopharmacology ; Davis, K. L., Charney, D., Coyle, J. T.,

Nemeroﬀ, C., Eds.; Lippincott Williams & Wilkins: Philadelphia, PA,

2002; pp 1139ꢀ1163.

(48) Bonaventure, P.; Kelly, L.; Aluisio, L.; Shelton, J.; Lord, B.;

Galici, R.; Miller, K.; Atack, J.; Lovenberg, T. W.; Dugovic, C. Selective

blockade of 5-hydroxytryptamine (5-HT)₇receptors enhances 5-HT

transmission, antidepressant-like behavior, and rapid eye movement

sleep suppression induced by citalopram in rodents. J. Pharmacol. Exp.

Ther. 2007, 321, 690–698.

(49) Drevets, W. C.; Price, J. L.; Furey, M. L. Brain structural and

functional abnormalities in mood disorders: implications for neurocir-

cuitry models of depression. Brain Struct. Funct. 2008, 213, 93–118.

(50) Fanselow, M. S.; Dong, H. W. Are the dorsal and ventral

hippocampus functionally distinct structures?. Neuron 2010, 65, 7–19.

(51) Mørk, A.; Kreilgaard, M.; Sanchez, C. The R-enantiomer of

citalopram counteracts escitalopram-induced increase in extracellular

5-HT in the frontal cortex of freely moving rats. Neuropharmacology

2003, 45, 167–173.

(33) Meyer, J. H.; Wilson, A. A.; Sagrati, S.; Hussey, D.; Carella, A.;

Potter, W. Z.; Ginovart, N.; Spencer, E. P.; Cheok, A.; Houle, S.

Serotonin transporter occupancy of ﬁve selective serotonin reuptake

inhibitors at diﬀerent doses: an [¹¹C]DASB positron emission tomo-

graphy study. Am. J. Psychiatry 2004, 161, 826–835.

(34) Larsen, A. K.; Brennum, L. T.; Egebjerg, J.; Sꢀanchez, C.;

Halldin, C.; Andersen, P. H. Selectivity of [³H]MADAM binding

to 5-hydroxytryptamine transporters in vitro and in vivo in mice;

correlation with behavioural eﬀects. Br. J. Pharmacol. 2004, 141

1015–1023.

(35) Owens, M. J.; Krulewicz, S.; Simon, J. S.; Sheehan, D. V.; Thase,

M. E.; Carpenter, D. J.; Plott, S. J.; Nemeroﬀ, C. B. Estimates of

serotonin and norepinephrine transporter inhibition in depressed

patients treated with paroxetine or venlafaxine. Neuropsychopharmacol-

ogy 2008, 33, 3201–3212.

(52) Beyer, C. E.; Boikess, S.; Luo, B.; Dawson, L. A. Comparison of

the eﬀects of antidepressants on norepinephrine and serotonin con-

centrations in the rat frontal cortex: an in-vivo microdialysis study. J.

Psychopharmacol. 2002, 16, 297–304.

(53) Sanchez, C.; Brennum, L. T.; Storustovu, S.; Kreilgard, M.;

Mork, A. Depression and poor sleep: the eﬀect of monoaminergic

antidepressants in a pre-clinical model in rats. Pharmacol., Biochem.

Behav. 2007, 86, 468–476.

(54) Blier, P. The pharmacology of putative early-onset antidepres-

sant strategies. Eur. Neuropsychopharmacol. 2003, 13, 57–66.

(55) Gardier, A. M.; Malagie, I.; Trillat, A. C.; Jacquot, C.; Artigas, F.

Role of 5-HT1A autoreceptors in the mechanism of action of serotoni-

nergic antidepressant drugs: recent ﬁndings from in vivo microdialysis

studies. Fundam. Clin. Pharmacol. 1996, 10, 16–27.

(56) Wolfe, J. P.; Buchwald, S. L. Scope and limitations of the Pd/

BINAP-catalyzed amination of aryl bromides. J. Org. Chem. 2000,

65, 1144–1157.

(36) Yamada, Y.; Sugiura, M.; Higo, K.; Ozeki, T.; Takayanagi, R.;

Okuyama, K.; Yamamoto, K.; Satoh, H.; Sawada, Y.; Iga, T. Receptor

occupancy theory-based analysis of antiemetic eﬀects and standard

doses of 5-HT₃receptor antagonists in cancer patients. Cancer Che-

mother. Pharmacol. 2004, 54, 185–190.

(37) Farde, L.; Ginovart, N.; Ito, H.; Lundkvist, C.; Pike, V. W.;

McCarron, J. A.; Halldin, C. PET-characterization of [carbonyl-¹¹C]WAY-

100635 binding to 5-HT_1Areceptors in the primate brain. Psychopharma-

cology 1997, 133, 196–202.

(38) Bantick, R. A.; Rabiner, E. A.; Hirani, E.; de Vries, M. H.; Hume,

S. P.; Grasby, P. M. Occupancy of agonist drugs at the 5-HT_1Areceptor.

Neuropsychopharmacology 2004, 29, 847–859.

(39) Wong, H.; Dockens, R. C.; Pajor, L.; Yeola, S.; Grace, J. E., Jr.;

Stark, A. D.; Taub, R. A.; Yocca, F. D.; Zaczek, R. C.; Li, Y.-W. 6-

Hydroxybuspirone is a major active metabolite of buspirone: assessment

of pharmacokinetics and 5-hydroxytryptamine_1Areceptor occupancy in

rats. Drug Metab. Dispos. 2007, 35, 1387–1392.

3220

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Journal of Medicinal Chemistry

ARTICLE

(57) Pettit, G. R.; Chamberland, M. R.; Blonda, D. S.; Vickers, M. A.

Antineoplastic agents: XI. N-Bis(2-bromoethyl)amines. Can. J. Chem.

1964, 42, 1699–1706.

(58) Kalvass, J. C.; Maurer, T. S. Inﬂuence of nonspeciﬁc brain and

plasma binding on CNS exposure: implications for rational drug

discovery. Biopharm. Drug Dispos. 2002, 23, 327–338.

(59) Sꢀanchez, C.; Bergqvist, P. B. F.; Brennum, L. T.; Gupta, S.;

Hogg, S.; Larsen, A.; Wiborg, O. Escitalopram, the S-(þ)-enantiomer of

citalopram, is a selective serotonin reuptake inhibitor with potent eﬀects

in animal models predictive of antidepressant and anxiolytic activities.

Psychopharmacology 2003, 167, 353–362.

3221

dx.doi.org/10.1021/jm101459g |J. Med. Chem. 2011, 54, 3206–3221

Article Doi

Discovery of 1-[2-(2,4-Dimethylphenylsulfanyl)phenyl]piperazine (Lu AA21004): A Novel Multimodal Compound for the Treatment of Major Depressive Disorder

DOI: 10.1021/jm101459g

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm101459g

Full text of DOI:10.1021/jm101459g

Products guided by the article

R&D Labs maybe for 1293343-33-5

Relevant to this article

Hot Product