Synthesis and SAR of highly potent dual 5-HT1A and 5-HT 1B antagonists as potential antidepressant drugs

Article abstract of DOI:10.1016/j.bmcl.2005.04.077

Novel 5-HT₁ autoreceptor ligands based on the N-4-aryl-piperazinyl-N′-ethyl-5,6,7,8-tetrahydropyrido[4′, 3′:4,5]thieno[2,3-d]pyrimidin-4(3H)-one core are described. Aiming at antidepressants with a novel mode of action our objective was to identify potent antagonists showing balanced affinities and high selectivity for the 5-HT _1A and 5-HT_1B receptors. Strategies for the development of dual 5-HT_1A and 5-HT_1B antagonists based on 1 and 2 as leads and the corresponding results are discussed. Isoquinoline analogue 33 displayed high affinity and an antagonistic mode of action for the 5-HT _1A and the 5-HT_1B receptors and was characterized further with respect to selectivity, electrically stimulated [³H]5-HT release and in vivo efficacy.

Full text of DOI:10.1016/j.bmcl.2005.04.077

Bioorganic & Medicinal Chemistry Letters 15 (2005) 5567–5573
Synthesis and SAR of highly potent dual 5-HT1A
and 5-HT1B antagonists as potential antidepressant drugs
Andreas Kling,a,* Udo E. W. Lange,a Helmut Mack,a Margot H. M. Bakker,a
Karla U. Drescher,a Wilfried Hornberger,a Charles W. Hutchins,b Achim Mo¨ller,a
Reinhold Muller,a Martin Schmidt,a Liliane Unger,a Karsten Wicke,a
¨
Kurt Schellhaasc and Gerd Steinerc
aNeuroscience Discovery, Abbott GmbH & Co. KG, D-67008 Ludwigshafen, Germany
bAdvanced Technologies, Abbott Laboratories, Abbott Park, IL 60064-6101, USA
cBASF AG, D-67056 Ludwigshafen, Germany
Received 28 February 2005; revised 25 April 2005; accepted 28 April 2005
Available online 10 October 2005
Abstract—Novel 5-HT1 autoreceptor ligands based on the N-4-aryl-piperazinyl-N0-ethyl-5,6,7,8-tetrahydropyrido[40, 30:4,5]thi-
eno[2,3-d]pyrimidin-4(3H)-one core are described. Aiming at antidepressants with a novel mode of action our objective was to iden-
tify potent antagonists showing balanced affinities and high selectivity for the 5-HT1A and 5-HT1B receptors. Strategies for the
development of dual 5-HT1A and 5-HT1B antagonists based on 1 and 2 as leads and the corresponding results are discussed.
Isoquinoline analogue 33 displayed high affinity and an antagonistic mode of action for the 5-HT1A and the 5-HT1B receptors
and was characterized further with respect to selectivity, electrically stimulated [3H]5-HT release and in vivo efficacy.
Ó 2005 Published by Elsevier Ltd.
Major depression is a mental disorder characterized by
persistent periods of a set of symptoms such as de-
pressed mood, irritability, low self-esteem, feelings of
hopelessness and worthlessness, loss of interest in activ-
ities or pleasurable stimuli, decreased energy, and recur-
rent thoughts of death and suicide.1 The disease is
associated with considerable impairment in a patientÕs
quality of life, functional disability, and loss of produc-
tivity, and poses a major public and economic health
problem.2–4
serotonin reuptake inhibitors (SSRIs) have emerged as
first-line antidepressant therapy due to their improved
side-effect profile.8 However, a major limitation of the
current SSRIs is their slow onset of action of up to six
weeks, and their lack of efficacy in more than 30% of
the depressed patients.9 Hence, there is a large medical
need for the development of novel antidepressants with
improved efficacy, faster onset of action, and better
tolerability.
Amongst the various 5-HT receptors, the 5-HT autore-
ceptors, in particular, have received considerable atten-
tion as potential targets for antidepressant action.10
5-HT efflux in the dorsal raphe nucleus is regulated by
the 5-HT1A, 5-HT1B, and 5-HT1D receptor subtypes,
each differentially located and with slightly different,
possibly even parallel, roles.11–13 Comparing the phar-
macological properties across species, only the rodent
5-HT1B subtype stands out. Rat and human 5-HT1B
receptor (the latter previously termed as 5-HT1Db) show
a single amino acid modification and were found to have
different pharmacology, whereas the pharmacological
profile for both the 5-HT1A and the 5-HT1D subtypes
was shown to be well conserved among different species.
Although a number of new hypotheses have emerged
during the last years, the pathophysiological mecha-
nisms underlying depression remain poorly understood.5
Numerous clinically effective antidepressants act via
increasing extracellular serotonin (5-HT) levels either
by blocking 5-HT reuptake or degradation, thus sup-
porting the hypothesis that enhancement of 5-HT neu-
rotransmission might underlie the therapeutic response
to these different antidepressants.6,7 Today selective
*
Corresponding author. Tel.: +49 621 589 34 49; fax: +49 621
58963449; e-mail: andreas.kling@abbott.com
0960-894X/$ - see front matter Ó 2005 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2005.04.077

5568
A. Kling et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5567–5573
In contrast, and despite their relatively low homology,
the human 5-HT1D and 5-HT1B subtypes are pharmaco-
logically remarkably similar and thus are virtually indis-
tinguishable in native tissue.14
R
N
O
N
N
R =
H3C
N
N
S
OMe
N
For several years 5-HT1A agonists have been evaluated
for antidepressant efficacy, but despite the fact that 5-
HT1A agonists like buspirone have been shown to exert
anxiolytic and antidepressant effects in humans, results
from clinical trials have been disappointing so far.15 In
contrast, 5-HT1A antagonists were recently shown to ex-
ert anxiolytic-like effects.16 Additional data suggest that
co-administration of a 5-HT1A antagonist accelerates
and potentiates the antidepressant efficacy of SSRIs.
However, first trials were proving to be inconsistent.17,18
Blockade of 5-HT1B/D autoreceptors has been proposed
as an alternative approach towards novel antidepres-
sants.19 Recently, the efficacy of AR-A2, a selective 5-
HT1B/D antagonist, has been demonstrated in behavioral
models of depression, and the compound is reported to
be in phase II clinical trials.20
1
2
Figure 2. Lead structures 1 and 2.
Herein, we describe the synthesis and characterization of
new 5-HT1A and 5-HT1B ligands based on the N-
aryl-piperazinyl-N0-ethyl-5,6,7,8-tetrahydropyrido[40,30:
4,5]thieno[2,3-d]pyrimidin-4(3H)-one core. We also
present SAR derived from modification in the aryl part
with respect to affinity, selectivity, and functional prop-
erties. Selected compounds were examined for selectivity
against a set of related cloned human receptors, in vitro
5-HT release, PK properties, and in vivo efficacy in a
behavioral model of depression.
Sequence alignment and molecular modeling23 indicated
only minor differences in the ligand binding site of 5-
HT1B and 5-HT1A receptors. Thus, our goal of achieving
combined antagonists seemed to be feasible. The tetra-
hydropyridothienopyrimidinones 1 and 2 were chosen
as attractive leads since they already offered high affinity
and the required balanced binding profile for the 5-HT1
autoreceptors, whereas functional antagonism remained
the primary goal for further optimization (Fig. 2).
It has been suggested that 5-HT autoreceptor activation
at least contributes to the limitations of SSRI efficacy in
two ways: reduction in 5-HT neuronal firing due to
an activation of somatodendritic 5-HT1A autoreceptors
by locally released 5-HT, and second, a reduction of
5-HT release in terminal areas resulting from increased
5-HT in the synapse activating 5-HT1B/D receptors.
Chronic antidepressant treatment will then lead to auto-
receptor desensitization, and the time required for
desensitization relates to the delayed onset of action ob-
served in antidepressant therapy.
Careful analysis of available SAR data led to the
hypothesis that the functional properties within this ser-
ies are determined by the nature of the N-aryl-piperazine
moiety. This hypothesis was supported by modeling sug-
gesting that the orientation of Asp95 on TM224 deter-
mines the functional properties of both the 5-HT1A
and the 5-HT1B receptors.25,26 Modeling revealed an
interaction between the isoquinolinyl-nitrogen in 1 and
Asp95, aligning the carboxyl group in the Ôagonist orien-
tationÕ (Fig. 3A); a similar model could be assembled for
the corresponding 2-methoxyphenyl analogue 2. We
hypothesized that disruption of this interaction via mod-
ification in the aryl part of the molecule would alter
Asp95 to an ÔantagonistÕ orientation. Hence, we focused
our efforts on modification of the Ôright-hand sideÕ aro-
matic moiety since the tetrahydropyridothienopyrimidi-
none core and the C2 linker had been found to be
optimal with respect to 5-HT1 affinity in previous
studies.
Hence, we envisaged that a strategy for circumventing
these adverse autoinhibitory effects could be to block
the 5-HT autoreceptors simultaneously, which should
lead to a rapid and massive increase in 5-HT levels,
and is expected to result in a faster onset of action and
improved antidepressant efficacy. This hypothesis was
supported by in vivo microdialysis studies showing that
concurrent blockade of the 5-HT1A and 5-HT1B/D recep-
tor subtypes resulted in a significant and synergistic in-
crease of extracellular 5-HT in guinea pig frontal
cortex.21 However, to date, SB-272183 is the only 5-
HT1A/B/D antagonist described in the literature, but nei-
ther in vivo data nor any further development has been
reported (Fig. 1).22
N
Early on in this project a1 cross-activity had been iden-
tified as potential issue in our lead series, since inhibition
of a adrenergic receptors is known to interfere with the
cardiovascular system and thus might lead to side ef-
fects.27 Previous results in this series indicated that a1
affinity could be affected by modification in the tetrahy-
dropyridothienopyrimidinone core, either by exchange
of the N-methyl group or by variation adjacent to the
pyrimidinone carbonyl, thus impeding the interaction
between carbonyl group and the receptor. We decided,
however, to focus first on achieving the desired dual
antagonism for both the 5-HT1A and 5-HT1B receptors
CH3
N
O
N
N
H
N
H3C
O
NH
N
O
N
N
CH3
Cl
AR-A2
SB-272183
Figure 1.

A. Kling et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5567–5573
5569
Figure 3. (a) Modeling structure of compound 1 in the 5-HT1B receptor. Shown are the amino acid residues assembling the ligand binding site. In
green: Asp95 and Asp129, the latter binding to the piperazine N (numbering of amino acids cf. Ref. 24). (b) Structure of compound 21 in the 5-HT1A
receptor. Leu 356 (see text) is indicated in purple.
as the more challenging, and in terms of proof of con-
cept, the more important task, and to address a1 selec-
tivity subsequently. Nevertheless, affinity data for a1
have been determined routinely for all analogues and
are included in Tables 1–4.
Table 2. Effect of additional substituents in 2-methoxy-arylpiperazine
series
X
O
N
N
N
OMe
Scheme 1 depicts the general synthesis of N-aryl-piper-
azinyl-N0-ethyl-5,6,7,8-tetrahydropyrido[40, 30:4,5]thie-
N
N
derivatives.28
S
H3C
no[2,3-d]pyrimidin-4(3H)-one
based
Condensation of 4-methylpiperidine-4-one with ethyl
cyanoacetate and sulfur according to a procedure origi-
nally described by Gewald,29 assembly of the annelated
3-substituted pyrimidinone and conversion into the
chloride afforded key intermediate 5, which was further
reacted with the respective piperazine derivatives to give
final products 1, 2, and 13–41. The various (het)aryl pip-
Compounds
X
Ki (nM)a/functionb
a1 Ki (nM)a
5-HT1A
5-HT1B
2
21
22
23
24
25
26
27
H
5-Cl
0.7/AN
3.0/AN
24.2/AN
5.6/pA
1.4/pA
0.4
0.1
70.3/AN
50.2/AN
11.9/AN
42.3/nd
116.4/nd
3.3/AN
5-OMe
5-CH3
5-CF3
5-Ph
4-Cl
8.4
12.1
18.0
0.1
103.9/nd
429.0/nd
18.8/pA
2.5/pA
Table 1. Variation of aryl substituent in position 2
0.5
7.0
3-Cl
0.5/pA
a Values are geometric means of two or three experiments.
b AN = antagonist, pA = partial agonist, A = agonist as described.34
O
N
N
N
X
N
N
S
H3C
Compounds
X
Ki (nM)a/functionb
a1 Ki (nM)a
erazines applied in the synthesis were prepared following
one of the routes as exemplified in Scheme 2 for interme-
diates 8 and 12. Nucleophilic amination according to
Buchwald-Hartwig30 using tert-butyl piperazine-1-car-
boxylate (quinoline-, isoquinoline-piperazines, and
benzothien-7-ylpiperazine, cf. Scheme 1) with subse-
quent deprotection, or assembly of the piperazine start-
ing from the corresponding aniline precursor using
bis(chloroethyl)amine (various substituted aryl pipera-
zines, 2,3-dihydro-1-benzofuran-7-ylpiperazine31), gave
the desired piperazine derivatives. The required starting
materials were either commercially available or prepared
according to generally known procedures.32
5-HT1A
5-HT1B
2
13
14
15
16
17
18
19
20
OMe
CH3
Ph
0.7/AN
2.9/AN
4.1/pA
2.8/pA
2.5/pA
0.7/pA
2.5/pA
0.9/pA
1.9/pA
1.4/pA
3.5 / pA
1.1/AN
1.7/pA
1.4/A
0.4
1.7
1.1
1.0
1.8
0.1
0.2
0.5
4.0
Cl
CN
OiPr
2.5/AN
5.4/AN
1.1/pA
1.2/pA
O
OPh
OBzl
a Values are geometric means of two or three experiments.
b AN = antagonist, pA = partial agonist, A = agonist as described.34

5570
A. Kling et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5567–5573
Table 3. SAR of poly-substitued or bicyclic (het)aryl substituents
Table 4. SAR within the quinoline and isoquinoline series
R
R
O
N
O
N
N
N
N
N
N
N
N
N
S
H3C
S
H3C
Ki (nM)a/
functionb
a1 Ki (nM)a
Compounds
R
Ki (nM)a/functionb
a1 Ki (nM)a
Compounds
R
5-HT1A
5-HT1B
5-HT1A 5-HT1B
2.5/pA 4.9/AN
Cl
1
33
34
35
36
37
38
39
N
N
0.7/pA
0.5/pA
41.1
13.7
118.1
26.7
129.1
1.2
Cl
Cl
28
29
30
31
32
0.2
14.8
4.5
OMe
Cl
4.7/AN
6.5/AN
1.5/AN
0.9/pA
0.3/pA
0.2/pA
2.5/AN
9.7/AN
1.7/pA
0.9/pA
0.1/pA
0.3/pA
25.8/A
3.8/pA
N
OMe
N
11.8/pA 5.4/pA
0.6/pA 0.5/AN
0.4/pA 0.4/AN
OMe
N
N
0.2
O
0.5
S
a Values are geometric means of two or three experiments.
b AN = antagonist, pA = partial agonist, A = agonist as described.34
1.2
N
N
N
93.4/AN 319.6/A
6780
N
Final compounds 1, 2, and 13–41 were evaluated for
receptor affinity via radioligand binding assays using hu-
man recombinant 5-HT1A, 5-HT1B, and a1 receptors.33
Furthermore, affinity for the human recombinant 5-
HT1D receptor was determined routinely. As expected,
affinity for 5-HT1D was found to parallel 5-HT1B affini-
ty, hence for reasons of clarity these data have not been
included in the SAR tables. Functional profiles were
determined for compounds displaying Ki values for 5-
HT1A and 5-HT1B <100 nM using either [35S]GTPcS
binding (5-HT1A) or FLIPR (5-HT1B co-expressed with
the G-protein chimera Gqo5 in HEK293 cells). Assign-
ment of functional properties in vitro was calculated as
percentage of the maximal response to 5-HT (Emax),
and an intrinsic activity qualifier was assigned.34 The re-
sults are depicted in Tables 1–4.
40
41
9.1/pA
0.7/pA
36.6/pA
0.5/pA
121.1
13.2
a Values are geometric means of two or three experiments.
b AN = antagonist, pA = partial agonist, A = agonist as described.34
for 5-HT1A and 5-HT1B in the case of the alkyl ana-
logues (17, 18), whereas the phenoxy and benzyloxy
derivatives 19 and 20 were shown to be partial agonists
in vitro.
Next, we examined the effect of additional substitution
in the aryl moiety of 2 (Table 2). Starting with chlorine
as substituent, 2-methoxy-5-chloro analogue 21 dis-
played the desired antagonist profile. Modeling of 21
suggested that the 5-chlorine atom interferes with
Leu356 residue on TM6, shifting the benzene ring of
21 toward TM2 so that there is no room for the interac-
tion of the 2-methoxy group with Asp82 via a water
molecule (Fig. 3b). As a consequence, the Asp82 side-
chain conformation is changed, and compound 21 is
an antagonist at both the 5-HT1A and 5-HT1B receptors.
However, the achievement of dual antagonism was asso-
Following our hypothesis of Asp95 interaction as main
determinant for intrinsic activity, we started with simple
modification of the 2-methoxy group in 2 to explore ba-
sic SAR for this substituent (Table 1, compounds 13–
20). Whereas methyl analogue 13 retained the affinity
and efficacy of 2, phenyl analogue 14 unexpectedly
showed reversed intrinsic activity for the 5-HT1A and
5-HT1B receptors, respectively, an effect which is not
understood. Exchange of methoxy with electron-with-
drawing substituents (15, 16) resulted in retained affinity
but shifted the intrinsic activity for both receptor sub-
types to (partial) agonism. Variation of the alkoxy sub-
stituent (17–20) again led to reversed intrinsic activity
ciated with a 50-fold reduction in affinity versus 5-HT1B.
A similar trend, although more pronounced in the case

A. Kling et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5567–5573
5571
R
Cl
O
N
O
N
N
CO2Et
NH2
O
a
b, c, d
e
N
N
N
N
N
H3C
S
H3C
N
N
H3C
S
S
H3C
3
4
5
1, 2, 13-41
Scheme 1. Reagents and conditions: (a) S8, ethyl cyanoacetate, morpholine, EtOH/60 °C (89%); (b) triethyl orthoformiate, acetic acid anhydride/
reflux; (c) aminoethanol, EtOH/reflux; (d) SOCl2, dichloroethane/rt (61% from 4); (e) N-arylpiperazine, NaBr, DIEA, NMP/110 °C.
N
a
b, c
d
e, f
N
N
g
O
x TFA
N
N
O
O
O
Br
CO2H
NH2
N
H
N
H
6
7
8
9
10
11
12
Scheme 2. Reagents and conditions: (a) NBS, H2SO4/ꢀ20 °C (52%); (b) tert-butyl piperazine-1-carboxylate, NaOtBu, BINAP, Pd2(dba)3, toluene/
80 °C (72%); (c) TFA, CH2Cl2 (91%); (d) BuLi, TMEDA, CO2, n-heptane/ꢀ70 °C to rt (70%); (e) NaN3, TFA, TFAA/rt; (f) NaOH, CH3OH/reflux
(58% from 10); (g) bis(chloroethyl)amine hydrochloride, chlorobenzene/reflux (91%).
of 5-HT1A, was observed with the corresponding 2,5-di-
methoxy derivative 22. Ki values >50 nM for 5-HT1B
prevented us from further advancing compounds 21
and 22. Additional variation in position 5 either led to
partial agonism toward the 5-HT1A receptor (23) or
resulted in a significant loss in affinity for all 5-HT1
receptor subtypes in parallel (24, 25). Migration of the
additional chloro substituent to the 4- or 3-position
(26, 27) again enhanced especially 5-HT1B affinity, with
2-methoxy-3-chloro analogue 26 showing the most bal-
anced affinity in this series. To exploit the beneficial ef-
fect observed for 5-chloro substitution in 21, analogues
featuring increased bulk in the aryl moiety were pre-
pared (Table 3). The corresponding bis-chloro ana-
logues 28, and 29 and naphthyl derivative 30 showed
improved 5-HT1B affinity, but none of these compounds
retained the initial promising functional profile of 21 or
22. Constraining the methoxy group of 2 by incorpora-
tion into a dihydrofuran ring (31) led to high affinities
for all three 5-HT1 receptor subtypes, but also for a1.
The corresponding naphthyl analogue 41 again showed
high affinity, but did not display balanced antagonism.
Besides the effects already discussed here, we found that
most of the compounds from the quinoline/isoquinoline
series (Table 4) showed a1 selectivity higher than those
of their corresponding ÔarylÕ analogues (Tables 1–3).
Summarizing the data presented in Tables 1–4, most
analogues derived from aryl variation in 1 and 2 re-
tained the high affinity for the 5-HT1A/B receptors. With
respect to intrinsic activity, the SAR we obtained is
inconclusive for both the 5-HT1A and 5-HT1B receptors.
Whereas the dual antagonist profile of compounds like
22, 23, or 33 corroborates our initial working hypothe-
sis, the intrinsic activity observed for compounds such
as 30 or 41 does not comply with our functional model.
One explanation for this discrepancy could be a different
binding mode of these analogues. Nevertheless, these re-
sults either reflect the difficulties in predicting the mode
of action and/or reveal our incomplete understanding of
the correlation between 5-HT1A/B structure and func-
tion, respectively.
Dihydrofuran 31 displayed antagonism for 5-HT1B
,
but only partial agonism for 5-HT1A. A similar result
was obtained with the corresponding thienyl analogue
32. All analogues from the Ôaryl seriesÕ (Tables 1–3) were
found to show low a1 selectivity, which was only mar-
ginally affected by variation in the aryl moiety.
Based on the results depicted in Table 5 (Ki values 5-
HT1A/B <10 nM and an antagonist mode of action for
5-HT1A and 5-HT1B), compounds 33 and 34 were
selected for further characterization. Affinity for the
corresponding rat receptors was found to be in a com-
parable range for 5-HT1A, and about a factor of 10
lower for 5-HT1B with Ki values for r5-HT1A of
16.3 nM (33) and 23.0 nM (34), and for r5-HT1B of
33.1 nM (33) and 116.8 nM (34). Although displaying
limited selectivity versus a1 adrenoceptors, selectivity
of at least 50-fold against related human serotonergic
and dopaminergic receptors was observed (Table 6).
Moreover, at concentrations of up to 10 lM no inhi-
bition of monoamine transporter function was found
(5-HT, dopamine or norepinephrine transporters; re-
sults not shown). To further evaluate the effect on
5-HT autoreceptor function, compounds were investi-
gated for their ability to influence electrically stimu-
We then directed our efforts to the related quinoline ser-
ies as alternative approach (Table 4). Based on 1 as
starting point migration of nitrogen around the ring sys-
tem (entries 33–38) revealed that N in the 3- or 4-posi-
tion resulted in the desired antagonist profile for 5-
HT1A and 5-HT1B. As already observed for analogue
21, affinity was diminished, in this case by a factor of
5–20. Binding affinity in this series was highest for deriv-
atives 1, 37, and 38, but these analogues did not exhibit
the desired functional profile. Combination of the favor-
able features from 1 and 34 by incorporating N in the
2- and 4-position in quinazoline 39 led to dramatic
reduction in affinity. Exchange of the N-location and
the piperazine linkage (40) resulted in diminished affini-
ty, again especially pronounced in the case of 5-HT1B
.

5572
A. Kling et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5567–5573
Table 5. Functional profiles for 33 and 34
Compounds
[
35S]GTPc S Emax (%)
5-HT1AIC50 (nM)a
FLIPR Emax (%)
5-HT1B IC50 (nM)a
33
34
4
6
316
1304
7
9
35
85
a Values are geometric means of two or three experiments.
Acknowledgments
We thank S. Heitz, P. Karl, V. Ott, S. Pister, K. Britze,
H. Schulke, C. Thiem, and S. Triebel for chemical syn-
¨
thesis, S. Bopp, P. Goeck-Sturm, E. Ka¨fer, M. Loucka,
M. Mayrer, S. Muller, S. Petersen-Werner, S. Manzano,
¨
M. Nebel, G. Plotzky, M. Schanzenba¨cher, and T.
Schowalter for assay development and screening, L.
King for PK studies, S. Butty, B. Degner, and I. Groth
for in vivo experiments, and C. Krack, M. Diehl, T.
Kro¨nung, K. Lang, and M. Schienemann for analytical
support. Many thanks to Hans Schoemaker for his
contributions and inspiring discussions.
Table 6. Counter screening results for 33 and 34
Receptor Ki (nM)
33
34
5-HT2A
5-HT2C
5-HT5A
5-HT7
D3
180
1.200
5.500
>10.000
>10.000
520
>1.000
4.480
1.300
1.460
118
1.180
1.200
13.7
D2L
a1
Values are geometric means of two or three experiments.
lated [3H]5-HT-release from rat brain cortical slices.
At concentrations of 0.1 and 1 lM, compound 33
evoked a dose-dependent increase of [3H]5-HT-release
confirming the antagonist mode of action. Pharmaco-
kinetic studies of 33 in rat showed metabolic stability
and bioavailability after iv- and po-dosing (2 and
10 mg/kg, respectively), and good CNS penetration
(10 mg/kg po; F: 11 %; brain plasma ratio of 10.2/
AUC0–8h). Upon in vivo examination in the mouse
forced swim test, a behavioral model to assess antide-
pressant efficacy,35 compound 33 did not show any
significant activity up to 30 mg/kg ip.
References and notes
1. Doris, A.; Ebmeier, K.; Shajahan, P. Lancet 1999, 354,
1369.
2. Hirschfeld, R. M.; Montgomery, S. A.; Keller, M. B.;
Kasper, S.; Schatzberg, A. F.; Moller, H. J.; Healy, D.;
Baldwin, D.; Humble, M.; Versiani, M.; Montenegro, R.;
Bourgeois, M. J. Clin. Psychiatry 2000, 61, 268.
3. Holden, C. Science 2000, 288, 39.
4. Pelissolo, A.; Lepine, J.-P. In Perspectives in Psychiatry
Vol. 8, Epidemiology of Depression and Anxiety Disorder;
Montgomery, S. A., den Boer, J. A., Eds., 2nd ed.; Wiley
& Sons: Chichester, UK, 2001.
Nevertheless, tool compound 33 will be examined in
additional behavioral assays and different species to
confirm our hypothesis and to further validate the ratio-
nale for the synthesis of dual 5-HT1A/B antagonists.
5. Slattery, D. A.; Hudson, A. L.; Nutt, D. J. Fundam. Clin.
Pharm. 2004, 18, 1.
6. Nestler, E. J.; Barrot, M.; Dileone, R. J.; Eisch, A. J.;
Gold, S. J.; Monteggia, L. M. Neuron 2002, 34, 13.
7. (a) Sun, M. K.; Alkon, D. L. Curr. Drug Targets 2002, 1,
575; (b) Vetulani, J.; Nalepa, I. Eur. J. Pharmacol. 2000,
405, 351.
8. (a) Snow, V.; Lascher, S.; Mottur-Pilson, C. Ann. Intern.
Med. 2000, 132, 738; (b) Williams, J. W.; Mulrow, C. D.;
Chiquette, E.; Noel, P.; Aguilar, C.; Cornell, J. Ann.
Intern. Med. 2000, 132, 743; (c) Nutt, D. Eur. Neuropsy-
chopharmacol. 2000, 10(Suppl. 4), S433.
9. (a) Spinks, D.; Spinks, G. Curr. Med. Chem. 2002, 9, 799;
(b) Peretti, S.; Judge, R.; Hindmarch, I. Acta. Psychiatr.
Scand. 2000, 403, 17, Supplementum.
10. (a) Lanfumey, L.; Hamon, M. Curr. Drug Targets—
CNS Neurol. Disord. 2004, 3, 1; (b) Middlemiss, D. N.;
Price, G. W.; Watson, J. M. Curr. Opin. Pharmacol.
2002, 2, 18.
11. Stamford, J. A.; Davidson, C.; McLaughlin, D. P.;
Hopwood, S. E. Trends Neurosci. 2000, 23, 459.
12. (a) Martin-Ruiz, R.; Ugedo, L. Neuropharmacology 2001,
41, 72; (b) Pullar, I. A.; Boot, J. R.; Broadmore, R. J.;
Eyre, T. A.; Cooper, J.; Dsanger, G. J.; Wedley, S.;
Mitchell, S. N. Eur. J. Pharm. 2004, 493, 85; (c) Moret, C.;
Briley, M. Neuropharmacology 1997, 36, 1713.
13. Bidmon, H. J.; Schleicher, A.; Wicke, K.; Gross, G.; Zilles,
K. Naunyn-SchmiedebergÕs Arch. Pharmacol. 2001, 363,
364.
In summary, we have identified novel and potent com-
bined 5-HT1 autoreceptor ligands containing the N-4-
aryl-piperazinyl-N0-ethyl-5,6,7,8-tetrahydropyrido[40,
30:4,5]thieno[2,3-d]pyrimidin-4(3H)-one
core.
The
compounds prepared displayed affinities in the low nano-
molar range for human 5-HT1A, 5-HT1B, and 5-HT1D
receptor subtypes. Furthermore, several analogues
exhibited an antagonist mode of action for both the 5-
HT1A and 5-HT1B receptors. In general, the structural
modifications we applied had only limited effect on in vi-
tro binding, but major impact on the intrinsic activity for
the 5-HT1A and 5-HT1B receptor subtypes. Although we
achieved our initial goal, the rational design of balanced
5-HT1A/B antagonists still presents a major challenge
since the SAR with respect to the mode of action is not
fully understood. Compounds 33 and 34 were chosen
as most promising representatives and submitted to fur-
ther characterization in in vitro and in vivo testing. In
counter screening, both compounds showed high selec-
tivity against a set of related receptors with the exception
of a1 adrenoceptors. Although 33 was shown to increase
[3H]5-HT release from rat brain cortical slices, an antide-
pressant-like effect in vivo in the mouse forced swim test
could not be demonstrated. Further studies using 33 as
tool compound will be conducted to evaluate the poten-
tial of dual 5-HT1A/B antagonists.
14. Barnes, N. B.; Sharp, T. Neuropharmacology 1999, 38,
1083.
15. Heiser, J. F.; Wilcox, C. S. CNS Drugs 1998, 10, 343.

A. Kling et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5567–5573
5573
16. (a) Barros, M.; Mello, E. L., Jr.; Maior, R. S.; Muller, c.
¨
Hancock, A. A. Neuropharmacology 2003, 44, 773; (b)
von Geldern, T. W.; Hutchins, C. W.; Kester, J. A.; Wu-
Wong, J. R.; Chiou, W.; Dixon, D. B.; Opgenorth, T. J.
J. Med. Chem. 1996, 39, 957.
P.; de Souza Silva, M. A.; Carey, R. J.; Huston, J. P.;
Tomaz, C. Eur. J. Pharm. 2003, 482, 197; (b) Cao, B. J.;
Rodgers, R. J. Neurosci. Biobehav. Rev. 1998, 23, 247.
17. (a) Artigas, F.; Romero, L.; de Montigny, C.; Blier, C.
Trends Neurosci. 1996, 19, 378; (b) Rabiner, E. A.; Gunn,
R. N.; Castro, M. E.; Sargent, P. A.; Cowen, P. J.; Koepp,
M. J.; Meyer, J. H.; Bench, C. J.; Harrison, P. J.; Pazos,
A.; Sharp, T.; Grasby, P. M. Neuropsychopharmcology
2000, 23, 285.
26. The model discussed here is in contradiction to a
functional model suggested by other authors Gaster, L.
M.; Blaney, F. E.; Davies, S.; Duckworth, D. M.; Ham,
P.; Jenkins, S.; Jennings, A. J.; Joiner, G. F.; King, F. D.;
Mulholland, K. R. J. Med. Chem. 1998, 41, 1218.
27. Hoffmann, B. B.; Lefkowitz, R. J.. In Gilmann, A. G.,
Rall, T. W., Nies, A. S., Taylor, P., Eds., 8th ed.; The
Pharmacological Basis of Therapeutics; Pergamon Press:
New York, USA, 1990.
18. Mewshaw, R. E.; Zhou, D.; Zhou, P.; Shi, X.; Hornby, G.;
Spangler, T.; Scerni, R.; Smith, D.; Schechter, L. E.;
Andree, T. H. J. Med. Chem. 2004, 47, 3823.
19. Moret, C.; Briley, M. Eur. J. Pharmacol. 2000, 404, 1.
20. (a) Ahlgren, C.; Eriksson, A.; Tellefors, P.; Ross, S. B.;
Stenfors, C.; Malmberg, A. Eur. J. Pharmacol. 2004, 499,
67; (b) Stenfors, C.; Hallerba¨ck, T.; Larsson, G.; Wallsten,
C.; Ross, S. B. Naunyn-SchmiedebergÕs Arch. Pharmacol.
2004, 369, 330;(c) AstraZeneca development pipeline:
NCEs and line extensions; AstraZeneca plc Company
World Wide Web site October 06, 2004.
21. (a) Roberts, C.; Price, G. W.; Jones, B. J. Eur. J.
Pharmacol. 1997, 326, 23; (b) Stenfors, C.; Magnusson,
T.; Larsson, L. G.; Yu, H.; Hallbus, M.; Magnusson, O.;
Ross, S. B. Naunyn-SchmiedebergÕs Arch. Pharmacol.
1999, 359, 110.
22. (a) Watson, J.; Roberts, C.; Scott, C.; Kendall, I.; Collin,
L.; Day, N. C.; Harries, M. H.; Soffin, E.; Davies, C. H.;
Randall, A. D.; Heightman, T.; Gaster, L.; Wyman, P.;
Parker, C.; Price, G. W.; Middlemiss, D. N. Br. J.
Pharmacol. 2001, 133, 797; (b) Roberts, C.; Price, G. W.;
Middlemiss, D. N. Brain Res. Bull. 2001, 56, 463.
23. Methods and modeling software according to: Hutchins,
C. W. Endocrin. J. 1994, 2, 7.
28. Detailed synthetic protocols in: Steiner, G.; Schellhass, K.;
Lubisch, W.; Holzenkamp, U.; Starck, D.; Knopp, M.;
Szabo, L.; Emling, F.; Garcia-Ladona, F. X.; Hofmann,
H. P.; Unger, L. DE 19900545, 2000; Chem. Abstr. 2000,
133, 89541.
29. Sabnis, R. W.; Rangnekar, D. W.; Sonawane, N. D.
J. Heterocycl. Chem. 1999, 36, 333.
30. Wolfe, J. P.; Wagaw, S.; Buchwald, S. L. J. Am. Chem.
Soc. 1996, 118, 7215.
31. Joiner, G. F.; Gaster, L. M. WO 9511243, 1995; Chem.
Abstr. 1995, 123, 198831.
32. Martinez-Esparza, J.; Oficialdegui, A. M.; Perez-Silanes,
S.; Begona, H.; Orus, L.; Palop, J. A.; Lasheras, B.; Roca,
J.; Mourelle, M.; Bosch, A.; Del-Castillo, J. C.; Tordela,
R.; Del Rio, J.; Monge, A. J. Med. Chem. 2001, 44, 418.
33. Assay setup and screening procedures: Bakker, H. M.;
Hornberger, W.; Kling, A.; Lange, U. E. W.; Mack, H.;
Moeller, A., Schellhaas, L.; Schmidt, M.; Steiner, G. ;
Wicke, K. WO 2004055024, 2004; Chem. Abstr. 2004, 141,
89095.
34. The intrinsic activity qualifier was assigned according to:
A = agonist (GTPcS: Emax > 70%; FLIPR: Emax > 80%);
pA = partial agonist (FLIPR: 20% < Emax < 80%; GTPc
S: 10% < Emax < 70%); AN = antagonist (GTPc S:
Emax < 10%; FLIPR: Emax < 20%).
24. The amino acid numbers mentioned in the text refer to the
residues in the h5-HT1B sequence and correspond to the
5-HT1A sequence as follows: Asp95/5-HT1B-Asp82/5-
HT1A; Asp129/5-HT1B to Asp116/5-HT1A
.
25. (a) Yao, B. B.; Hutchins, C. W.; Carr, T. L.; Cassar, S.;
Masters, J. N.; Bennani, Y. L.; Esbenshade, T. A.;
35. Cryan, J. F.; Markou, A.; Lucki, I. Trends Pharmacol. Sci.
2002, 23, 238.