Synthesis and SAR of highly potent and selective dopamine D 3-receptor antagonists: 1H-Pyrimidin-2-one derivatives

Article abstract of DOI:10.1016/j.bmcl.2005.10.068

The synthesis and SAR of novel highly potent and selective dopamine D ₃-receptor antagonists based on a 1H-pyrimidin-2-one scaffold are described. A-690344 antagonized PD 128907-induced huddling deficits in rat (ED₅₀ 6.1 mg/kg po), a social interaction paradigm.

Full text of DOI:10.1016/j.bmcl.2005.10.068

Bioorganic & Medicinal Chemistry Letters 16 (2006) 490–494
Synthesis and SAR of highly potent and selective dopamine
D3-receptor antagonists: 1H-Pyrimidin-2-one derivatives
a,
b
a
b
a
*
´
Herve Geneste, Gisela Backfisch, Wilfried Braje, Jurgen Delzer, Andreas Haupt,
¨
Charles W. Hutchins,c Linda L. King,b Andreas Kling,a Hans-Jurgen Teschendorf,a
¨
Liliane Ungera and Wolfgang Werneta
aAbbott GmbH & Co. KG, Discovery Research, D-67008 Ludwigshafen, Germany
bAbbott GmbH & Co. KG, DMPK, D-67008 Ludwigshafen, Germany
cAdvanced Technology, Abbott Laboratories, IL 60064-6101, USA
Received 18 October 2005; accepted 19 October 2005
Available online 11 November 2005
Abstract—The synthesis and SAR of novel highly potent and selective dopamine D3-receptor antagonists based on a 1H-pyrimidin-
2-one scaffold are described. A-690344 antagonized PD 128907-induced huddling deficits in rat (ED50 6.1 mg/kg po), a social
interaction paradigm.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Drug therapy of schizophrenia has traditionally in-
volved the use of neuroleptics which pharmacologically
act as antagonists at the dopamine D2 receptor. This has
been an important basis for the Ôdopamine hypothesisÕ
of schizophrenia. However, medical treatment of dopa-
mine related disorders is often limited by side effects as a
consequence of binding to various dopamine or other
related monoamine receptors. Due to its localization,
the dopamine D3 receptor subtype might be a target
for the development of antipsychotics, with a reduced
risk of side effects.1 This hypothesis is supported by data
reported for two selective D3 antagonists: A-437203
(also known as BSF-2016402) and SB-2770113 (Chart
1) which exhibit efficacy in a variety of preclinical mod-
els predictive of antipsychotic activity without inducing
catalepsy or raising plasma prolactin levels.
3H-pyrimidin-4-one pharmacophore of A-437203 has
prompted us to evaluate the possibility of using 1H-
pyrimidin-2-ones for the design of D3-antagonists
(Chart 1).
As numerous substituted uracils are commercially
available, a convenient two-step route4 was devised
(Scheme 1): alkylation5 of the pyrimidin-2-one on N1
to introduce the spacer followed by nucleophilic substi-
tution and introduction of the piperazine-pyrimidine
moiety (Q).6
The non commercially available substituted uracils were
readily accessible using published procedures7–9 or mod-
ifications4 of them (Scheme 2). 4-Alkyl (compound 2a)
and 4-cycloalkyl (compounds 2b–d) substituted 1H-pyr-
imidin-2-ones were prepared from the corresponding
b-keto acetals4,10 1a–d by reaction4,7 with urea. 4-Ph
derivative 3 was prepared from 2,4-dichloro-pyrimidine
by selective Suzuki coupling8,11 and subsequent hydroly-
sis9,11 of the remaining chloride. Introduction of a cyclic
amine (azetidine, piperidine) in position 4 (6a and b)
proceeded via thio-derivative4,12 5 in good yields
(40–55%): the original procedure13 (described in a sealed
tube) was adapted to microwave conditions.4 The
N3-substituted derivatives (8a and b) were obtained by
methylation4 of the halogenated precursors 7a and b
using iodomethane in the presence of KOH.
A program in our laboratories aimed at exploring
further the potential of the pyrimidyl-piperazine core
(Q, Chart 1) of A-437203. Here, we present the synthesis
and SAR of novel D3 antagonists based on a 1H-pyrim-
idin-2-one moiety. Its structural similarity with the
Keywords: Dopamine D3 receptor antagonists; Atypical antipsychot-
ics; 1H-Pyrimidin-2-ones.
*
Corresponding author. Tel.: +49 621 589 3423; fax: +49 621 589
63423; e-mail: herve.geneste@abbott.com
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.10.068

H. Geneste et al. / Bioorg. Med. Chem. Lett. 16 (2006) 490–494
491
N
N
N
N
N
N
N
CF3
N
CF3
O
O
N
HO
N
N
N
S
N
N
Q
Q
NC
N
HO
A-690344
SB-277011
A-437203
Chart 1. Reference compounds—from A-437203 to A-690344.
O
O
Receptor affinities for the compounds described have
been determined in binding assays4 using human
cloned dopamine D3 and D2 receptors. Compounds
with a Ki 6 10 nmol/l and a D3 versus D2 selectivity
of >50-fold were analyzed in vitro for their functional
properties with a hD3-GTPcS binding assay:14 all
compounds tested displayed antagonistic activities
(Emax < 10%).
X
a
N
NH
N
N
R1
R1
R2
R2
X = Cl
b
N
N
N
X =
N
3
CF
In the course of our studies15 to explore further the po-
tential of the pyrimidyl-piperazine core (Q, Chart 1) of
A-437203, a novel series of potent and selective D3
antagonists based on a 1H-pyrimidin-2-one scaffold
was identified. Unexpectedly, although the C-linkage
of the side chain was shifted to a N-linkage when the
3H-pyrimidin-4-one moiety of A-437203 was replaced
by a 1H-pyrimidin-2-one moiety (Chart 1), comparable
D3-affinity and selectivity versus D2 could be achieved
(A-437203 and 4, Table 1).
Scheme 1. Reagents and conditions: (a) K2CO3, 1-bromo-4-chloro-
butane, DMF, 12 h, rt; (b) 2-tert-butyl-4-piperazin-1-yl-6-(trifluoro-
methyl)pyrimidine,6 Et3N, DMF, 24 h, 110 ꢁC.
O
O
OMe
OMe
O
O
a, b, c
d
N
NH
+
R
H
OMe
R
R
1 a-d
2 a-d
R = (a) Me, (b) cyclopr., (c) cyclobut., (d) cyclohex.
Molecular modelling: A model for the D3 dopamine
receptor has been developed internally, based on the
three-dimensional models of the D1 and D2 dopamine
receptors,16 by changing the residues that are different
between the human sequences of the receptors. There-
fore, comparative modeling of a ligand in two differ-
ent receptor models (in our case D3 and D2) might
give a clue for the observed selectivity as a result
from the different interactions of the ligand with the
amino acids in the binding site of each receptor
model.
Cl
Cl
O
e
f
N
N
N
N
N
NH
Cl
Ph
Ph
3
O
O
Q
h
Q
N
N
N
N
R
N
6 a-b
R = OH
R = SH
(4)
(5)
As exemplified with A-690344 (Fig. 1), modeling sug-
gested a strong interaction of the carbonyl group in po-
sition 2 of the 1H-pyrimidin-2-one moiety with a
threonine residue (Thr 368 on TM7) present in the D3
model. In the corresponding D2 model, this threonine
is replaced by a phenylalanine residue (Phe 411) which
cannot exert any H-bond donor interaction.17 The D2
and D3 receptor models thus support the high affinity
for the D3 receptor and the high selectivity versus the
D2 receptor of A-690344, both resulting from the inter-
action of the carbonyl group in position 2: favorable
with Thr 368 in the D3 model, bad with Phe 411 in the
D2 model.
g
N
,
N
N
=
a
b
O
R
O
Cl
Me
Cl
N
N
i
N
N
HO
O
R
8 a-b
R = (a) H, (b) Me
7 a-b
Scheme 2. Reagents and conditions: (a) NaOMe, DEE, 3 h, rt; (b)
H2SO4, MeOH, 12 h, rt; (c) aq KOH (pH 8); (d) Urea, HCl, EtOH,
2 h, reflux; (e) PhB(OH)2, Pd(PPh3)4, K2CO3, PhCH3, MeOH, 3 h, rt,
70%; (f) concd HCl, 1 h, 100 ꢁC, quant.; (g) 1/2 P2S5, dioxane, 1 h,
4-Substitution: In analogy with former data,1,15 a five
single bond spacer length was recognized as optimal
for both high D3 affinity and high selectivity versus D2
(data not shown). However, based on the comparison
reflux, quant.; (h)
, EtOH, 1 h, 60 ꢁC, lW Milestone Ethos 1600,
NH
55% (azetidine), 41% (piperidine); (i) MeI, 5 equiv KOH, DMSO, 5 h,
rt, 58% (R=H); quant. (R=Me).

492
H. Geneste et al. / Bioorg. Med. Chem. Lett. 16 (2006) 490–494
Table 1. Variation of 4-substituent
demanding t-Bu group (compound 12) exhibited both
high D3 affinity and selectivity versus D2. The 4-cyclo-
alkyl analogs 13–15 and the 4-Ph analog 16 confirmed
that sterically more demanding groups were advanta-
geous in terms of selectivity versus D2, whereas D3-
affinity was hardly affected. Cyclic amino groups in
position 4 (6a and b) are not tolerated with respect to
selectivity versus D2: a 3-fold improvement of D2-bind-
ing was observed for comparable D3-affinity (14 and 6a,
15 and 6b).
N
N
N
CF3
O
N
N
N
R
b
Compound
R
H
Ki (nM)
Sel. versus D2
35
a
D3
a
D2
5-Substitution: Although a 5-substituent in the pyrimi-
din-2,4-dione series is not a prerequisite for high D3
affinity and selectivity versus D2 (compound 4), highly
diverse residues are well tolerated in position 5 (Table
2: D3-affinity of 5 nM or lower, selectivity vs D2 90-fold
or more). The highest selectivity was obtained with
R = Me (A-690344, sel. vs D2 246-fold). Increase of ste-
ric bulk in this position (R = Et, 17) reduced the selectiv-
ity versus D2 (89-fold). Interestingly, variation of the 5-
substituent influences dramatically the physico-chemical
properties of the molecule as reflected in the calculated
9
12.9
1.9
7.4
4.4
2.6
3.7
7.4
4.8
1.4
4.4
3.4
2.9
447
196
290
307
216
155
273
302
168
4
OH
OEt
Me
104
39
10
11
71
2
2
polar surface areas18 (21, 110 A compared to 87 A
˚
˚
for 18). Position 5 is thus the place of choice for fine-
tuning—for example, to adjust the pharmacokinetic
properties.
12
t-Bu
82
13
41
N3-Methylation: N3-Methylation decreased D3-affinity
(Ki > 10 nM, 6- to 8-fold higher than for the nonmethy-
lated compounds) as well as D2-affinity (Table 3, 22 and
23 compared to 4 (Table 1) and A-690344 (Table 2)).
The selectivity versus D2 remains thus high to moderate.
An unsubstituted NH is therefore required for good
affinity and selectivity. However, it is not clear whether
this effect is due to additional steric hinderance intro-
duced by the N-methyl group, or to the removal of a
hydrogen bond donor, either the N3H itself or eventual-
ly the tautomeric OH-forms in position 2 and 4, their
formation being blocked by the N-methylation.
14
37
15
63
16
Ph
122
19
N
6a
82.1
A-690344 was further characterized and displayed a
selectivity higher than a 100-fold versus the whole panel
of targets in a broad CEREP profile.19 Moreover, A-
690344 showed in vitro metabolic stability in liver
microsomes of rat and human (>97% recovery),20 oral
bioavailability, and brain penetration, and was charac-
terized by moderate permeability in the Caco-2 model
(Papp 1.94 · 10À6 cm/s).21
N
6b
89.4
26
A-437203
351
120
a Values are means of 2–3 experiments. Standard deviation of max 30%
of the mean.
b Ki D2/Ki D3.
In vivo efficacy was evaluated in a rat social interaction
paradigm. Naturally occurring social interaction in rats
(huddling) can be disrupted by a D3 preferring agonist,
such as PD 128907.22 Antipsychotics have been shown
to reverse the effect of PD 128907 (at least partially).23
A-690344 antagonized PD 128907-induced huddling
deficits in rat (ED50 6.1 mg/kg po). Deficits in huddling
might represent one component of deficient social
behavior and thus constitute an animal model for nega-
tive symptoms of schizophrenia.23
with the nonsubstituted derivative 9, a 4-substituent was
required for low nanomolar D3-affinity and high selec-
tivity versus D2 (Table 1). Pyrimidin-2,4-dione 4 was
among the best representatives in terms of D3-affinity
and selectivity versus D2 (above 100-fold). H-bond do-
nor group like OH seemed to be beneficial regarding
selectivity versus D2 when one compares 4 with OEt-
derivative 10.
In conclusion, the combination of diversely substituted
1H-pyrimidin-2-ones with the disubstituted pyrimi-
dine–piperazine moiety of A-437203 has yielded a family
Surprisingly, lipophilic alkyl groups (compounds 11, 12)
were shown to be also well tolerated: the more sterically

H. Geneste et al. / Bioorg. Med. Chem. Lett. 16 (2006) 490–494
493
Figure 1. (a) A-690344 (in stick presentation) in the D3 model. Interactions with the 1H-pyrimidin-2-one moiety.17 The semitransparent white surface
represents the molecular surface of the protein and the dashed line the interaction with Thr 368. (b) A-690344 (in stick presentation) in the D2 model.
Interactions with the 1H-pyrimidin-2-one moiety.17 The semitransparent white surface represents the molecular surface of the protein. The ligand
cannot bind in this orientation due to bad steric contacts with Phe 411.
Table 2. Variation of 5-substituent
Table 3. Effect of N3-methylation
N
N
N
N
N
CF3
N
CF3
O
O
N
N
N
N
N
N
HO
O
R
R
b
b
Compound
R
H
Ki (nM)
Sel. versus D2
Compound
R
H
Ki (nM)
Sel. versus D2
a
a
a
a
D3
D2
D3
D2
4
A-690344
1.9
1.3
1.7
3.0
2.3
5.4
4.8
196
320
148
602
220
646
473
104
246
89
22
23
11.8
10.4
610
356
52
34
CH3
CH3
17
18
19
20
21
C2H5
F
Cl
a Values are means of 2–3 experiments. Standard deviation of max 30%
of the mean.
201
97
b Ki D2/Ki D3.
CF3
CN
119
99
a Values are means of 2–3 experiments. Standard deviation of max 30%
of the mean.
b Ki D2/Ki D3.
Acknowledgments
We thank Stefan Maurus, Karlpeter Orth, Claudia
Partos, Hermann Schulke, Carsten Thiem, and Sonja
¨
of highly potent and very selective D3-antagonists.
Potency and selectivity versus D2 are supported by our
molecular model. Two critical positions of the 1H-pyr-
imidin-2-one scaffold were identified: position 4 was
shown to play an important role for the discrimination
of the related D3 and D2 receptors, whereas position 5
was recognized as the place of choice to modify the
physico-chemical properties of the molecule. In vivo effi-
cacy was shown for A-690344 after po administration.
Triebel for supporting chemical synthesis, Liliana Parra
(BASF AG) for helpful discussions on the synthesis of
some of the substituted uracils, our analytical depart-
ment, Manfred Nebel, Heidrun Ga¨rtner, and Beate
Rauprich for assay development and screening, Helga
Buschbacher for in vivo testing, and Wilfried Lubisch
and Hans Schoemaker for valuable comments and help
in the preparation of the manuscript.

494
H. Geneste et al. / Bioorg. Med. Chem. Lett. 16 (2006) 490–494
10. Preparation analog to: Andriamialisoa, Z.; Valla, A.;
References and notes
Cartier, D.; Labia, R. Helv. Chim. Acta 2002, 85, 2926.
11. Lamon, R. W. J. Heterocycl. Chem. 1968, 5, 837.
12. Prepared as described above8 or by treatment of the urea
derivative with phosphorus pentasulfide according to Lal,
B.; Dohadwalla, A. N.; Dadkar, N. K.; DÕSa, A.; de
Souza, N. J. J. Med. Chem. 1984, 27, 1470.
13. Fenick, D. J.; Carr, H. S.; Falvey, D. E. J. Org. Chem.
1995, 60, 624.
14. Wicke, K.; Garcia-Ladona, J. Eur. J. Pharmacol. 2001,
424, 85.
15. Geneste, H.; Backfisch, G.; Braje, W.; Delzer, J.; Haupt,
A.; Hutchins, C. W.; King, L. L.; Lubisch, W.; Steiner, G.;
Teschendorf, H.-J.; Unger, L.; Wernet, W. Bioorg. Med.
Chem. Lett., in press.
16. Hutchins, C. Endocr. J. 1994, 2, 7; Hutchins, C. In
Structure and Function of 7TM Receptors; Schwartz, T.
W., Hjorth, S. A., Sandholm Kastrup, J., Ed.; Alfred
Benzon Symposium 39, Munksgaard: Copenhagen, 1996;
pp 213–226.
17. The interactions with the other parts of the molecule are
not shown and not discussed in detail, as they have already
been published for other D3 ligands.1
18. The calculation of polar surface area is based on fragment
contributions: Ertl, P.; Rohde, B.; Selzer, P. J. Med.
Chem. 2000, 43, 3714.
19. Seventy-five binding assays comprising receptors, ion
channels, and enzymes: www.cerep.com.
20. Measured in % recovery of parent after 1 h incubation at
37 ꢁC with liver microsomes (rat and human) in the
presence of NADPH.
21. Artursson, P. Crit. Rev. Ther. Drug Carrier Syst. 1991, 8,
105; Hilgers, A. R.; Conradi, R. A.; Burton, P. S. Pharm.
Res. 1990, 7, 902.
22. Adapted from Kagaya, T.; Yonaga, M.; Furuya, Y.;
Hashimoto, T.; Kuroki, J.; Nishizawa, Y. Brain Res. 1996,
721, 229.
1. Hackling, A.; Ghosh, R.; Perachon, S.; Mann, A.; Ho¨ltje,
H.-D.; Wermuth, C. G.; Schwartz, J.-C.; Sippl, W.;
Sokoloff, P.; Stark, H. J. Med. Chem. 2003, 46, 3883,
and references cited therein.
2. Unger, L.; Ladona, F. J. G.; Wernet, W.; Sokoloff, P.;
Wicke, K. M.; Gross, G. Poster, 32nd Annual Meeting of
the Society for Neuroscience, Orlando, FL, Nov 2–7,
2002; Society for Neuroscience: Washington, DC, 2002;
Abstr. 894.5; Drescher, K. U.; Ladona, F. J. G.,
Teschendorf, H. J.; Traut, M.; Unger, L.; Wicke, K. M.;
Weddige, F. K.; Freeman, A. S.; Gross, G. Poster, 32nd
Annual Meeting of the Society for Neuroscience, Orlando,
FL, Nov 2–7, 2002; Society for Neuroscience: Washing-
ton, DC, 2002; Abstr. 894.6.
3. Ashby, C. R.; Minabe, Y.; Stemp, G. J. Pharmacol. Exp.
Ther. 2000, 294, 1166; Reavill, C.; Taylor, S. G.; Wood,
M. D. J. Pharmacol. Exp. Ther. 2000, 294, 1154.
4. The preparation of the building blocks mentioned, typical
procedures and assay setup are described in DE 10311065,
2004; Chem. Abstr. 2004, 141, 296038.
5. When the 4-substituent was a hydroxy group (exemplified
by 5-methyl-1H-pyrimidine-2,4-dione: for R1=OH,
R2=Me in Scheme 1), the yield of the first step was
limited by rapid N1,N3-dialkylation: after treatment with
3 equiv of K2CO3 and quenching with 1 equiv of 1-bromo-
4-chloro-butane, 13% of product were isolated accompa-
nied by N1,N3-dialkylated compound (78% yield, based on
electrophile). When the 4-substituent was an alkyl group
(exemplified by 4-methyl-1H-pyrimidin-2-one: for R1=Me,
R2=H in Scheme 1), the yield remained moderate (46%)
due to simultaneous O-alkylation (12%).
6. The preparation of the building block QH (Scheme 1) is
described in DE 19735410, 1999; Chem. Abstr. 1999, 130,
182482.
7. Brown, D. J.; Lee, T.-C. Aust. J. Chem. 1968, 21, 243;
Jones, W. D.; Huber, E. W.; Grisan, J. M.; Schnettler, R.
A. J. Heterocycl. Chem. 1987, 24, 1221.
8. Gong, Y.; Pauls, H. W. Synlett 2002, 829.
9. Benneche, T.; Undheim, K. Acta Chem. Scand. B 1984,
38, 505.
ˆ
23. Jongen-Relo, A. L.; Drescher, K. U.; Teschendorf, H. J.;
Rueter, L. E.; Unger, L.; Gross, G; Schoemaker, H.
Poster, 34th Annual Meeting of the Society for Neurosci-
ence, San Diego, CA, October 23–27, 2004; Society for
Neuroscience: Washington, DC, 2004; Abstr. 350.2.