New fused benzazepine as selective D3 receptor antagonists. Synthesis and biological evaluation. Part 2: [g]-Fused and hetero-fused systems

Article abstract of DOI:10.1016/j.bmcl.2007.12.042

The synthesis and the SAR of a new series of potent and selective dopamine D₃ receptor antagonists is reported. The new scaffolds of the [g]-fused and the hetero-fused tricyclic benzazepine are here reported together with their pharmacokinetic profile.

Full text of DOI:10.1016/j.bmcl.2007.12.042

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 908–912
New fused benzazepine as selective D3 receptor
antagonists. Synthesis and biological evaluation.
Part 2: [g]-Fused and hetero-fused systems
Fabrizio Micheli,a,* Giorgio Bonanomi,a Simone Braggio,a Anna Maria Capelli,b
Federica Damiani,d Romano Di Fabio,a Daniele Donati,a Gabriella Gentile,a
Dieter Hamprecht,a Ornella Perini,b Marcella Petrone,a Giovanna Tedesco,b
Silvia Terreni,a Angela Worbyc and Christian Heidbredera,
aGlaxoSmithKline, Psychiatry Centre of Excellence for Drug Discovery, Via Fleming, 4, 37135 Verona, Italy
bMolecular Discovery Research, Via Fleming, 4, 37135 Verona, Italy
cHarlow NFSP, Essex, UK
dEuropean Patent Office, Munich, DE
Received 22 November 2007; revised 18 December 2007; accepted 19 December 2007
Available online 4 January 2008
Abstract—The synthesis and the SAR of a new series of potent and selective dopamine D3 receptor antagonists is reported. The new
scaffolds of the [g]-fused and the hetero-fused tricyclic benzazepine are here reported together with their pharmacokinetic profile.
Ó 2007 Elsevier Ltd. All rights reserved.
A growing body of evidence indicates that dopamine D3
receptors are implicated in the control of drug-seeking
behaviour, and may play an important role in both the
physiology and pathology of impulse control disorders,
drug addiction and schizophrenia. This hypothesis has
been difficult to prove due to the lack of compounds
with the appropriately high selectivity for the D3 recep-
tors. Recently, however, the synthesis and characteriza-
tion of new potent and selective DA D3 receptor
antagonists1 has allowed the characterization of the D3
receptor in a wide range of preclinical animal models.
GSK showed a long-standing interest in this field and
contributed to the discovery of selective D3 receptor
antagonists.2–5 As described in the first part of this man-
uscript,5 the possibility to increase the complexity in the
benzazepine (BAZ) portion of derivative 2 (Fig. 1) was
investigated. The introduction of [h]-fused BAZ led to
new chemical entities (NCE) endowed with promising
properties in terms of potency, selectivity and pharma-
cokinetic (PK) properties. To further expand the explo-
ration of the new class, [g]-fused tricyclics and anellated
systems were prepared and are described in the follow-
ing paragraphs (Scheme 1).
Considering that in the [h]-fused system series, the oxaz-
olyl moiety gave rise to a new template that was en-
dowed with good potency and promising rat PK
properties, it was attempted to introduce this fragment
in the new [g]-fused template to probe potential analo-
gies between the series. In contrast with the [h] bond
of the BAZ, the [g] one is no more intersected by a C2
symmetry axis and, therefore, the introduction of the
new fragment generated two different templates on
unsubstituted systems. In agreement with previous re-
ports,5 each NCE prepared was assayed for agonistic
versus antagonistic properties using a functional GTPcS
assay expressing the human dopamine D3 receptor. All
the compounds here reported proved to be antagonists
at the D3 receptor.5 A 100-fold selectivity versus dopa-
mine D2 and histamine H1 receptors (functional assays)
was set as a primary criterion for further progression
along the screening cascade of this specific series. In
addition, a 100-fold selectivity versus the hERG ion
channel (Dofetilide binding assay) was also required.
Keywords: Rat; Dopamine; D3 antagonists; Selective; Drug
dependence.
*
Corresponding author. Tel.: +39 045 8218515; fax: +39 045
8218196; e-mail: fm20244@gsk.com
Present address: Philip Morris USA, Richmond, VA, USA.
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.12.042

F. Micheli et al. / Bioorg. Med. Chem. Lett. 18 (2008) 908–912
909
N
N
N
N
O
N
N
N
S
N
N
N
N
1
2
Figure 1. GSK selective D3 antagonist.
R1
O
N
R1
N
O
R4
N
NH
S
N
N
R2
N
4
N
O
N
S
Ar
31-34
N
N
S
Ar
R2
N
R3
9-25
O
R3
4
7
N
N
O
N
NH
5
N
N
5
N
N
N N
6
Cl
N
N
n
S
Ar
S
Ar
S
Ar
N
N
N
30
3
N
O
26-29
R5
NH
R6
6
N
N
N
N
R4
N
S
Ar
N
N
S
36-42
NH
7
O
N
S
R5
N
N
N
N
N
N
c
R6
7R
N
Ar =
a
b
d
e
f
g
NH
N
8
Scheme 1. Reagents and condition: K2CO3, DMF 60 °C, cat NaI and appropriate tricyclic 4–8, overnight, n = 0, 1, 2. R as defined in each Table.
Generic developability screens such as CYPEX bacto-
some P450 inhibition and rat and human in vitro clear-
ance in liver microsomes (Cli) were included early in the
screening cascade.
to a slight decrease in D3 affinity, which was paralleled,
as for the previous [h]-fused system, by a decrease in the
hERG affinity. The introduction of the ethyl group in
derivatives 11 and 12 led to the increase in potency as
suggested by receptor modelling; as both H1 and hERG
values were significantly below the screening cascade’s
selectivity threshold it was decided to further progress
derivative 12 to in vivo assays which provided favour-
able in vitro PK properties. The PK profile, probably
The first derivative prepared (9, Table 1) belonged to the
7,8,9,10-tetrahydro-6H-[1,3]oxazolo[4,5-g][3]benzaze-
pine series and showed a comparable potency to its [h]-
fused derivative. Also the PK properties (both in vitro
and in vivo6) were comparable showing rat blood clear-
ance (Clb) of 48 ml/min/kg, a relatively high distribution
volume (Vd) of 6.3 l/kg, with moderate half-life (T1/2) of
1.7 h and bioavailability (F) of 15%. More importantly
for a potential central nervous system (CNS) drug, this
compound had a good brain to blood ratio (B/B)
(2.5), measured through sampling of the blood and
brain6 1 h after iv administration. The replacement of
the quinolinyl fragment with an oxazolyl one (10) led
controlled by both the polar surface area (PSA)7 (86
2
vs 73 A for 12 vs 9) and the lower clogD8 (3.5 vs
˚
5.1), showed a much reduced Clb (5 ml/min/kg) with
low Vd (0.6 l/kg), and a similar T1/2 (1.9 h), but a much
higher F (73%).
Not too surprisingly, given the above-reported PK data,
the B/B ratio was only 0.1. The introduction of a lipo-
philic trifluoromethyl moiety in the tricyclic system (13

910
F. Micheli et al. / Bioorg. Med. Chem. Lett. 18 (2008) 908–912
Table 1. Affinity results
R1
N
O
N N
N
S
Ar
N
R2
D3 fpKi
2
PSA (A )
R1
R2
Ar
D2 fpKi
H1 fpKi
hERG pIC50
ACD logD
˚
Entry
1
2
Not applicable
8.4
8.8
7.2
7.0
8.1
7.9
8.4
8.8
9.1
9.0
9.1
8.2
8.4
7.8
8.5
8.3
6.5
8.4
7.6
6.4
6.5
6.2
6.1
5.7
5.7
5.6
4.9
5.2
5.3
5.7
5.9
5.6
6.6
6.5
5.9
5.8
5.4
6.6
6.1
5.8
6.1
6.1
69
65
73
86
73
86
73
86
86
73
86
86
73
73
60
73
73
91
104
3.8
5.3
5.1
3.0
5.6
3.5
6.3
4.1
3.9
8.3
6.2
4.7
5.3
5.3
5.9
5.3
6.0
6.0
3.8
Not applicable
Me
Me
9
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
a
b
a
b
a
b
b
a
b
c
<6.2
<6.3
<6.3
<6.4
<6.1
<6.0
7.8
<6.1
<5.5
6.3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Et
Et
<5.5
6.3
CF3
CF3
<5.6
5.8
CF2CH3
CF2CF3
CF2CF3
CF3
<6.5
<6.5
<6.4
<6.4
<6.4
<6.3
<6.4
<6.2
<6.4
<6.4
5.7
<5.7
5.8
CF3
CF3
d
e
5.7
<5.5
5.9
CF3
CF3
f
g
a
a
b
6.0
Me
pir*
pir*
<5.8
6.0
<5.5
SEM for D3 GTPcS, H1 FLIPR and HERG data sets is 0.1 and for the D2 GTPcS data is 0.2. pir* = 1,3-dimethyl-1H-pyrazol-5-yl.
and 14) led to increased D3 affinity which was paralleled
by the hERG value for compound 13 only, while deriv-
ative 14 showed an almost 1000-fold selectivity. Deriva-
tive 13 (clogD = 6.3) showed relatively high Clb (60 ml/
min/kg), with Vd equal to 5.8 l/kg, T1/2 = 1.3 h and
F = 36%. The B/B was back to high values (7.3). The
reduction in clogD shown by 14 (4.1) once again led
to a lower Vd (1.6 l/kg), accompanied by lower Clb
(35 ml/min/kg); F was high (63%) and T1/2 was of
0.6 h. The B/B ratio was 0.8 with a relatively high con-
centration of compound in the brain, making this com-
pound a promising molecule to be further progressed.
The introduction of the CF2CH3 portion (15), with an
almost unchanged clogD value with respect to deriva-
tive 14, led to nanomolar affinity at the D3 receptor,
but raised unexpected activity at H1 and D2 receptors.
The unexpected profile was also seen in vivo, where,
despite a comparable PK profile to 14 (Vd = 0.8 l/kg;
T1/2 = 1.2 h; Clb = 9 ml/min/kg; F = 54%), the B/B ratio
was reduced to 0.2. A further increase in lipophilicity
with the introduction of the CF2CF3 group (16 and
17) maintained nanomolar affinity at the D3 receptor,
but this time with more than 1000-fold selectivity over
the D2 receptor; selectivity over hERG and H1 was also
good. The exploration of this series continued with the
decoration of the thiotriazole moiety showing that both
basic (18–20) and neutral (21 and 22) derivatives were
well tolerated in terms of D3 affinity and selectivity,
while the presence of basic moieties was beneficial to
achieve lower hERG values. The introduction of a chlo-
rine atom on position 5 of the tricyclic moiety (23) had a
greater impact on the D3 receptor than on the hERG
channel, the introduction of a further heterocycle into
the system was well tolerated by the quinolinyl deriva-
tive (24), but showed a higher decrease in D3 affinity
in the oxazolyl one (25), probably due to the much lower
clogD achieved by this latter derivative (clogD = 3.8 vs
6.0 for 24).
The 7,8,9,10-tetrahydro-6H-[1,3]oxazolo[5,4-g][3]benza-
zepine template (Table 2) gave similar results to its regio-
isomer previously described (Table 1) in terms of
in vitro affinity, even if slightly lower D3 affinity was
achieved when the same substituents (26–29) were com-
pared. From the PK point of view, however, it is impor-
tant to note that derivative 28 showed a clearly different
profile with higher Clb (43 ml/min/kg) than correspond-
ing compound 15, leading to lower F (8%); however, its
higher Vd (3.3 l/kg) led to a much better brain penetration
(B/B = 4.0).
To further investigate the role of the relative positions of
the oxygen and nitrogen in the tricyclic template, the
synthesis of one of the isomeric 3-methyl-7,8,9,10-tetra-
hydro-6H-isoxazolo[5,4-g][3]benzazepines was com-
pleted. Specifically, when the compound was coupled
with the pyrazine fragment (c) on the thiotriazolyl moi-
ety (30), minimal residual activity at the D3 receptor was
observed (Table 3).
The synthesis of the 7,8,9,10-tetrahydro-6H-[1,3]thiazol-
o[5,4-g][3]benzazepine was completed to analyze the role
of the replacement of the oxygen with a sulfur atom.
Similar, but not completely superimposable, results

F. Micheli et al. / Bioorg. Med. Chem. Lett. 18 (2008) 908–912
911
Table 2. Affinity results
R3
O
N
N N
N
S
Ar
N
2
PSA (A )
R3
Ar
D3 fpKi
D2 fpKi
H1 fpKi
hERG pIC50
ACD logD
˚
Entry
26
27
28
29
CF3
CF3
a
b
a
b
8.3
7.7
8.7
8.6
<5.8
<5.8
<6.1
<6.7
5.4
<5.5
5.8
6.0
5.3
5.6
5.3
73
86
73
86
6.2
4.0
6.0
3.8
CF2CH3
CF2CH3
<5.7
SEM for D3 GTPcS, H1 FLIPR and HERG data sets is 0.1 and for the D2 GTPcS data is 0.2.
Table 3. Affinity results
N
O
N N
N
N
N
S
N
2
PSA (A )
˚
Entry
D3 fpKi
6.1
D2 fpKi
<6.1
H1 fpKi
5.6
hERG pIC50
nt
ACD logD
30
86
4.1
SEM for D3 GTPcS, H1 FLIPR and HERG data sets is 0.1 and for the D2 GTPcS data is 0.2. nt, not tested.
(Table 4) to the ones achieved on the [h]-fused system5
were observed. Compounds with similar shape, bulki-
ness and clogD (31–32 vs 26–27), but with a slightly dif-
ferent electronic component showed slightly lower
affinity values in this series with respect to their oxygen-
ated counterparts.
was only marginally modified by this new substitution
pattern, the affinity at the D3 receptor was significantly
affected.
In summary, substitution of an isolated BAZ ring with
a more complex tricyclic structure can be tolerated in
terms of D3 affinity, and remarkable examples of selec-
tivity versus D2, H1 receptors and hERG channel were
identified. Promising rat PK profiles were also
achieved, showing balanced properties and generally
high brain penetration values. Overall, the exploration
work of [h]-fused, [g]-fused and anellated tricyclic scaf-
folds contributed to shed further light into the nature
of the D3 receptor and has enriched the portfolio of
selective D3 receptor antagonists. Some of these have
been used in in vivo animal models belonging to differ-
ent therapeutic areas and will be the subject of future
communications.
Further exploration on this template was performed
with the introduction of more hydrophilic derivatives
in the tricyclic moiety. The 2,3,4,5-tetrahydro-1H-pyr-
ido[10,20:1,2]imidazo[4,5-d]azepine was chosen as a
probe for two main reasons. The first reason was ‘geo-
metric’ in nature, providing this structure a shape which
is positioned between a [h]-fused and a [g]-fused system;
the second reason was physicochemical to have PSA val-
2
˚
ues ranging from 60 to 80 A , while maintaining clogD
values between 4 and 6. The results are clearly reported
in Table 5, and while the affinity at the hERG channel
Table 4. Affinity results
R4
S
N
N N
N
S
Ar
N
2
PSA (A )
R4
Ar
D3 fpKi
D2 fpKi
H1 fpKi
hERG pIC50
ACD logD
˚
Entry
31
32
33
34
Me
Me
Et
a
b
a
b
7.5
<6.9
8.7
<6.0
<6.0
<6.5
<6.4
6.6
<5.6
6.4
5.5
4.9
6.8
5.5
60
73
60
73
5.6
3.4
6.1
3.9
Et
8.2
<5.6
SEM for D3 GTPcS, H1 FLIPR and HERG data sets is 0.1 and for the D2 GTPcS data is 0.2.

912
F. Micheli et al. / Bioorg. Med. Chem. Lett. 18 (2008) 908–912
Table 5. Affinity results
R5
R6
N
N N
N
S
Ar
N
N
2
PSA (A )
R5
R6
Ar
D3 fpKi
D2 fpKi
H1 fpKi
hERG pIC50
ACD logD
˚
Entry
35
36
37
38
39
40
41
42
H
Me
H
a
a
a
a
a
b
a
b
6.8
<6.6
<6.6
<6.6
6.6
<5.9
<5.9
<5.9
<5.9
<6.1
<6.1
<6.1
<6.1
<5.5
<5.5
6.0
6.6
5.6
5.8
5.6
6.6
6.1
5.3
5.7
64
64
64
64
64
77
64
77
4.4
4.7
5.3
4.3
5.6
3.7
5.2
3.3
Me
Cl
H
H
H
5.7
6.1
H
CF3
CF3
H
H
6.7
6.8
<5.5
<5.5
<5.5
CF3
CF3
H
6.4
SEM for D3 GTPcS, H1 FLIPR and HERG data sets is 0.1 and for the D2 GTPcS data is 0.2.
J.; Routledge, C.; Stemp, G.; Thewlis, K. M.; Trail, B.;
Vong, A. K.; Hagan, J. J. J. Pharmacol. Exp. Ther. 2000,
294, 1154.
Acknowledgments
We thank Dr. Carla Marchioro and her analytical group
for the high quality support received.
3. Macdonald, G. J.; Branch, C. L.; Hadley, M. S.; Johnson,
C. N.; Nash, D. J.; Smith, A. B.; Stemp, G.; Thewlis, K. M.;
Vong, A. K.; Austin, N. E.; Jeffrey, P.; Winborn, K. Y.;
Boyfield, I.; Hagan, J. J.; Middlemiss, D. N.; Reavill, C.;
Riley, G. J.; Watson, J. M.; Wood, M.; Parker, S. G.;
Ashby, C. R., Jr. J. Med. Chem. 2003, 46, 4952.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2007.12.042.
4. Micheli, F.; Bonanomi, G.; Blaney, F. E.; Braggio, S.;
Capelli, A. M.; Checchia, A.; Curcuruto, O.; Damiani, F.;
Di-Fabio, R.; Donati, D.; Gentile, G.; Gribble, A.;
Hamprecht, D.; Tedesco, G.; Terreni, S.; Tarsi, L.; Light-
foot, A.; Pecoraro, M.; Petrone, M.; Perini, O.; Piner, J.;
Rossi, T.; Worby, A.; Pilla, M.; Valerio, E.; Griffante, C.;
Mugnaini, M.; Wood, M.; Scott, C.; Andreoli, M.; Lacroix,
L.; Schwarz, A.; Gozzi, A.; Bifone, A.; Ashby, C. R., Jr.;
Hagan, J. J.; Heidbreder, C. J. Med. Chem. 2007, 50, 5076.
5. BMCL previous article.
References and notes
1. (a) Micheli, F.; Heidbreder, C. Recent Patents CNS Drug
Discovery 2006, 1, 271; (b) Zhang, A.; Kan, Y.; Li, F.
Expert Opin. Ther. Patent 2006, 16, 587.
2. Reavill, C.; Taylor, S. G.; Wood, M. D.; Ashmeade, T.;
Austin, N. E.; Avenell, K. Y.; Boyfield, I.; Branch, C. L.;
Cilia, J.; Coldwell, M. C.; Hadley, M. S.; Hunter, A. J.;
Jeffrey, P.; Jewitt, F.; Johnson, C. N.; Jones, D. N.;
Medhurst, A. D.; Middlemiss, D. N.; Nash, D. J.; Riley, G.
6. The research complied with national legislation and with
company policy on the Care and Use of Animals and with
related codes of practice.
7. Ertl, P.; Rohde, B.; Selzer, P. J. Med. Chem. 2000, 43, 3714.
8. ACD Labs suite is available from http://www.acdlabs.com.