Cyclic amidines as benzamide bioisosteres: EPC synthesis and SAR studies leading to the selective dopamine D4 receptor agonist FAUC 312

Article abstract of DOI:10.1016/S0960-894X(03)00004-0

Investigation of conformationally restricted benzamide bioisosteres led to the chiral phenyltetrahydropyrimidine derivative ent2a (FAUC 312) displaying strong and highly selective dopamine D4 receptor binding (K_ihigh=1.5 nM). Mitogenesis experiments indicated 83% ligand efficacy when compared to the unselective agonist quinpirole. The target compounds of type 2 and 3 were synthesized in enantiopure form starting from asparagine.

Full text of DOI:10.1016/S0960-894X(03)00004-0

Bioorganic & Medicinal Chemistry Letters 13 (2003) 851–854
Cyclic Amidines as Benzamide Bioisosteres:
EPC Synthesis and SAR Studies Leading to the
Selective Dopamine D4 Receptor Agonist FAUC 312
Jurgen Einsiedel, Harald Hubner and Peter Gmeiner*
Department of Medicinal Chemistry, Emil Fischer Center, Friedrich-Alexander University, Schuhstraße 19, D-91052 Erlangen, Germany
Received 28 October 2002; revised 16December 2002; accepted 17 December 2002
Abstract—Investigation of conformationally restricted benzamide bioisosteres led to the chiral phenyltetrahydropyrimidine deri-
vative ent2a (FAUC 312) displaying strong and highly selective dopamine D4 receptor binding (Ki =1.5 nM). Mitogenesis
high
experiments indicated 83% ligand efficacy when compared to the unselective agonist quinpirole. The target compounds of type 2
and 3 were synthesized in enantiopure form starting from asparagine.
# 2003 Elsevier Science Ltd. All rights reserved.
Very recently, we have reported on the dihydroimid-
azole derivative FAUC 179 (1) showing highly
selective dopamine D4 receptor binding in combi-
nation with partial agonist properties (42% ligand
efficacy).1 Drugs exhibiting that kind of activity
profile are supposed to be beneficial for an effective
pharmacotherapy of ADHD (attention deficit hyper-
activity disorders).2
be investigated for its ability to serve as a suitable bio-
isostere.
Since the receptor binding of FAUC 179 (1) proved to
be strongly stereoselective,1 we planned to establish
effective EPC syntheses giving access to both enantio-
mers. Thus, the target compounds of type 2 and 3
should be prepared optically pure starting from (S)-
asparagine and its unnatural antipode.
In conjunction with our current efforts focused on the
efficacy tuning of dopaminergics,3 we were intrigued by
the question, whether modifications of the dihydro-
imidazole substructure led to a significant change of
intrinsic activity. Conserving the basicity of the amidine
function, we decided to expand the dihydroimidazole
moiety by one carbon atom resulting in tetra-
hydropyrimidine derivatives of type 2. The modified
geometry of the six-membered ring was expected to
influence the spatial orientation of the phenylpiper-
azinylmethyl side chain. This could affect the interaction
with both the binding sites of the active and inactive
receptor state thus enhancing or reducing the intrinsic
activity. To evaluate the effect of an electron with-
drawing carbonyl function, the neutral 5,6-dihydro-3H-
pyrimidin-2-one substructure of the analogue 3 should
Whereas the synthesis of the tetrahydropyrimidine deri-
vatives 2 was envisioned by reaction of a benzimidate
with a suitable 1,3-diamine building block, the pyri-
midinone derivative 3 should be prepared by condensation
with the appropriate b-alanine equivalent. The elaboration
of smooth cyclization conditions were regarded to be
crucial, because we expected a high temperature sensitivity
of the chiral building blocks. In detail, the preparation
of the chiral 1,3-diamine derivative 10a was possible by
two alternative pathways. Starting from (S)-asparagine
(4), a previously developed protocol gave access to the
aldehyde derivative 5.4 Reductive amination to give the
*Corresponding author. Tel.: +49-9131-852-9383; fax: +49-9131-
852-2585; e-mail: gmeiner@pharmazie.uni-erlangen.de
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00004-0

852
J. Einsiedel et al. / Bioorg. Med. Chem. Lett. 13 (2003) 851–854
piperazine derivative 6 and subsequent reduction with
LiAlH4 afforded the diamine 10a with high enantio-
meric purity.5 Overall yields could be increased by
exploiting the 1,3-diaminoalcohol derivative 7 as a cen-
tral intermediate, readily available from (S)-aspar-
agine.6 The protection of the primary amine to give the
phthalimide derivative 8 was achieved by the reaction
with phthalic acid anhydride applying Dean-Stark con-
ditions.7 Our first attempts to perform a Dess-Martin
oxidation8 of the alcohol and subsequent reductive
amination using phenylpiperazine hydrochloride resul-
ted in disappointing yields. Thus, we preferred an acti-
vation of the hydroxy group by methanesulfonic
chloride followed by a nucleophilic displacement with
phenylpiperazine. This reaction sequence furnished 59%
yield of the desired product 9a and 7% of the regio-
isomer 9b, obviously formed via an aziridinium inter-
mediate.9 Hydazinolysis of both isomers afforded the
primary amine derivatives 10a and 11a. Derivatization
of 10a with (R)-phenylethyl isocyanate and careful
HPLC investigation proved the isomeric integrity of the
reaction sequence.10 Debenzylation under reductive
conditions gave access to the fully deprotected diamine
building blocks 10b and 11b (Scheme 1).
Scheme 2. (a) 12a, MeOH, rt, 72 h (2a, 13: 44%); (b) 12b, (CH2Cl)2,
reflux, 2 h, (72%).
As a model reaction for the synthesis of the 5,6-dihydro-
3H-pyrimidin-4-one scaffold, we decided to react suit-
able b-alanine equivalents with methyl benzimidate
(12a). While the condensation with b-alanine and
b-amino propionitrile required extremely high tempera-
tures, the ring closure with b-alanine amide (14)16 pro-
ceeded in refluxing ethanol to give 2-phenyl-5,6-
dihydro-3H-pyrimidin-4-one (15) (Scheme 3). Since the
position of the double bond in the amidine substructure
was not actually clear,17 we performed an ab initio
molecular orbital calculation favoring unambiguously
the tautomer 15.18
Scheme 3. (a) 12a, EtOH, reflux, 12 h (34%); (b) H2, Pd(OH)2/C, 12 h,
rt (100% crude); (c) 12a, EtOH, reflux, 1 h (28%).
Scheme 1. (a) Phenylpiperazine HCl, NaBH3CN, MeOH, rt, 24 h
(30%); (b) LiAlH4, Et2O, 0 ꢀC–rt (crude: 99%); (c) phthalic acid
anhydride, NEt3, toluene, Dean-Stark, 5 h (91%); (d) (1) MesCl,
NEt3, THF, ꢁ23 ꢀC, 1 h; (2) phenylpiperazine, DMF/THF, 0 ꢀC–rt, 24
h (9a: 59%, 9b: 7%); (e) N2H4xH2O, EtOH, reflux, 3 h (10a: 96%, 11a:
64%); (f) H2, Pd(OH)2/C, MeOH, rt, 24 h (10b: 92%, 11b: 79%).
The initial strategy for the synthesis of the target com-
pounds 3 was to build up the pyrimidinone scaffold,
followed by the introduction of phenylpiperazine. In
detail, the dibenzylamino derivative 16, being accessible
from (S)-asparagine in two steps,19 gave rise to the
b-alanine amide derivative 17 by reductive deprotection.
Unfortunately, the condensation with methyl benzimi-
date 12a resulted in a complete formation of the oxazo-
line derivative 18, obviously due to kinetic reaction
control. In order to induce the desired regiocontrol of
the ring closure, O-protection of ent16 to give silyl ether
ent19 was performed. Indeed, subsequent debenzylation
and ring closure of the precursor ent20 furnished the
pyrimidone derivative ent21 (Scheme 4). Finally, desily-
lation to give the primary alcohol ent22, subsequent
conversion into the respective methanesulfonate and
SN2 reaction with phenylpiperazine afforded 7% of the
test compound ent3 over three steps.
Employing cyclization conditions that proved suitable
for the synthesis of dihydroimidazole derivatives,11 we
were able to proceed a smooth formation of the tetra-
hydropyrimidine 2a12 by condensation of the 1,3-dia-
mine 10b with methyl benzimidate (12a)13 (Scheme 2).
In an analogous way, the isomer 13 was available. For
the 5-bromo-2-methoxy derivative 2b, it turned out
advantageous to react the 1,3-diamine 10b and the
corresponding imidate 5b (available by the reaction of
5-bromo-2-methoxy benzamide14 with Meerwein’s
salt15) in refluxing 1,2-dichloroethane.

J. Einsiedel et al. / Bioorg. Med. Chem. Lett. 13 (2003) 851–854
853
ent2a (FAUC 312) revealed high D4 affinity, when
careful analysis of the competition experiments
employing a large number of test concentrations
showed a biphasic curve (Ki =1.5 nM and Ki =27
high
low
nM). These data are comparable to the D4 binding
properties of the unselective dopamine receptor agonist
quinpirole. However, selectivity over D2long, D2short and
D3 was approximately 500-fold higher. On the other
hand, receptor binding of the enantiomer 2a was sig-
nificantly reduced. Surprisingly, only weak affinity was
observed when investigating the 5-bromo-2-methoxy
derivatives 2b and ent2b. Introducing the electron with-
drawing carbonyl group to give the pyrimidone deriva-
tives 3 and ent3 resulted in reduced D4 affinity, when
ent3, showing the same spatial orientation of the phe-
nylpiperazinylmethyl side chain as ent2a, turned out to
be the more active isomer. The diazepine derivative 13
showed moderate D4 binding (Ki=120 nM), while the
affinity of ent13 was low.
Scheme 4. (a) TBDMS-Cl, imidazole, DMF, 0 ꢀC–rt, 2 h (100%); (b)
H2, Pd(OH)2/C, rt, 6h, (97%, crude); (c) 12a, EtOH, reflux, 72 h
(49%); (d) NBu4F, THF, 0 ꢀC, 30 min (95%); (e) (1) MesCl, NEt3,
THF, ꢁ23 ꢀC, 30 min (2) phenylpiperazine, DMF, 45 ꢀC, 12 h, 75 ꢀC, 2
h (7%).
To improve the yield, we changed the reaction sequence
towards a completely functionalized b-alanine building
block. Starting from alcohol 16, periodinane promoted
oxidation and subsequent reductive amination with
phenylpiperazine hydrochloride furnished the dibenzy-
lamine 23b. During the process, the cyclic N,N-acetal
23a was formed as a side product. Careful hydro-
genolysis of 23b to give the primary amine 24 and
subsequent condensation with methyl benzimidate (12a)
afforded the target compound 3 in high optical purity
(Scheme 5).20 The enantiomers ent2a, ent2b, ent3 and
ent13 were prepared analogously starting from (R)-
asparagine.
Table 1. Receptor binding data [Ki values (nM) based on the means
of 2–4 experiments each performed in triplicate]
Compd
-X-
R1 R2 D1 D2long D2short D3
D4
2a
H
H
H 39,000 29,000 30,000 13,000 310
H 29,000 25,000 32,000 13,000 1.5/27
ent2a
2b
ent2b
OMe Br 9100 16,000 14,000 1900 3500
OMe Br 15,000 44,000 43,000 7400 1600
3
ent3
H
H
H 8900 18,000 6400 20,000 3700
H 19,000 7300 2800 12,000 460
21
ent21
H
H
H 23,000 10,000 4200 12,000 120
H 23,000 19,000 14,000 3300 4400
Quinpirole
nd 64/3100 52/4000 24/420 1.8/53
To investigate the intrinsic effect of FAUC 312 (ent2a),
an in vitro functional assay measuring the [3H]thymi-
dine uptake in growing CHO cells stably expressing the
dopamine D4.2 receptor was performed.25 In fact, a
83% stimulation of mitogenesis (compared to the full
agonist quinpirole and the D4 antagonist clozapine) was
determined (EC50=50 nM) (Table 2).
Scheme 5. (a) (1) DMP, CH2Cl2, 0 ꢀC, 1 h; (2) NaHCO3/Na2S2O3
(1:1), H2O, Et2O (27%); (3) phenylpiperazine–HCl, NaBH3CN,
MeOH, rt, 24 h (26%); (b) H2, Pd(OH)2/C, rt, 6h, (97%, crude); (c)
2a, EtOH, reflux, 5 h (52%).
Table 2. Agonist effects of the tetrahydropyrimidine ent2a, quin-
pirole and clozapine at the D4.2 receptor investigated by measuring
the stimulation of mitogenesis
The test compounds 2a,b, 3 and 13 and the optical
antipodes ent2a,b ent3 and ent13 as well as the dopa-
mine receptor agonist quinpirole were evaluated in vitro
for their abilities to displace [3H]spiperone from the
Test compounds
21
cloned human dopamine receptors D2long, D2short
,
D322 and D4.423 being stably expressed in CHO cells
(see Table 1).24 D1 affinity was determined by employ-
ing bovine striatal membrane preparations and the D1
selective antagonist [3H]SCH 23390.24
ent2a
Quinpirole
Clozapine
Agonist effect (%)a
EC50 (nM)b
83
50
100
9.9
0
nd
aRate of incorporation of [3H]thymidine (in %) relative to the max-
imal effect of the full agonist quinpirole (100%); the results are the
means of quadruplicates from 10 experiments.
bEC50 values derived from the mean curves of 10 experiments; nd, not
determined.
All test compounds showed only weak D1, D2 and D3
receptor binding. Investigation of the D4 binding prop-
erties clearly indicated that the tetrahydropyrimidine

854
J. Einsiedel et al. / Bioorg. Med. Chem. Lett. 13 (2003) 851–854
In conclusion, SAR studies on cyclic amidine benza-
mide bioisosteres led to the highly selective dopamine
D4 receptor partial agonist ent2a (FAUC 312) incor-
porating a chiral tetrahydropyrimidine substructure.
Thus, expanding the imidazoline moiety of FAUC 179
(1) by one carbon atom did not significantly change the
D4 binding, if the spatial orientation of the pheny-
piperazinylmethyl side chain was maintained. On the
other hand, the intrinsic activity was increased from 42
to 83%. Further modifications including the introduc-
tion of a carbonyl group or a substitution of the phenyl
group resulted in a significantly reduced affinity.
(according to ref 5). HPLC analysis showed no (SR)-diaster-
eomer.
11. Djerassi, C.; Scholz, C. R. J. Am. Chem. Soc. 1947, 49,
1688.
12. 2a: a2D0=+60.0ꢀ (c 0.72, CHCl3); ent2a: a2D0=ꢁ61.0ꢀ (c
1.22, CHCl3); 1H NMR (CDCl3, 360 MHz): d=1.80 (dddd,
J=13.5, 8.8, 8.0, 4.8 Hz, 1H, H-5a), 2.08 (dddd, J=13.5, 4.9,
4.8, 4.6Hz, 1H, H-5b), 2.56–2.70 (m, 4H, C H2N(CH2)2/
CH2N(CH2)2), 2.77–2.83 (m, 2H, CH2N(CH2)2), 3.20–3.23 (m,
4H, PhN(CH2)2), 3.56(ddd, J=14.4, 8.8, 4.8 Hz, 1H, H-6a),
3.76(ddd, J=14.4, 4.9, 4.8 Hz, 1H, H-6b), 3.92 (dddd, J=8.0,
7.5, 7.5, 4.6, 1H, H-4), 6.86–6.94 (m, 3H, o/p-Ph), 7.24–7.30
(m, 2H, m-Ph), 7.45–7.50 (m, 2H, m-Ph0), 7.56–7.61 (m, 1H,
p-Ph0), 7.97–8.00 (m, 2H, o-Ph0). 13C NMR (CDCl3, 90 MHz):
d=22.3 (C-5), 39.0 (C-6), 46.7 (C-4), 49.1 (PhN(CH2)2), 53.3
(CH2N(CH2)2), 60.9 (CH2N(CH2)2), 116.1 (o-Ph), 120.0 (p-
Ph), 127.5 (i-Ph), 127.5 (o-Ph0), 129.2 (m-Ph), 129.4 (m-Ph0),
133.7 (p-Ph0), 150.9 (i-Ph0), 167.1 (C-2). HRMS (EI) calcd for
C21H26N4 (M+): 334.21576; found: 334.21690.
13. Wheeler Am. Chem. J. 1895, 17, 358. Release of the free
imidate base according to: Glickman, S. A.; Cope, A. C. J.
Am. Chem. Soc. 1945, 67, 1017.
Acknowledgements
The BMBF and the Fonds der Chemischen Industrie is
acknowledged for financial support. We thank Dr. H.
H. M. Van Tol (Clarke Institute of Psychiatry, Tor-
onto), Dr. J.-C. Schwartz and Dr. P. Sokoloff
(INSERM, Paris, France) and Dr. J. Shine (The Garvan
Institute of Medical Research, Sydney, Australia) for
providing D4.4, D3 and D2 receptor expressing cell
lines. Dr. R. Huff (Pharmacia & Upjohn, Inc., Kala-
mazoo, MI, USA) is acknowledged for providing a D4-
expressing cell line employed for mitogenesis.
14. Hirwe, N. W.; Patil, B. V. J. Indian Chem. Sect. A 1937, 321.
15. Meyers, A. I.; Flisak, J. R.; Aitken, R. A. J. Am. Chem.
Soc. 1987, 109, 5446.
16. Franchimont, A. P. N.; Friedmann, H. Recl. Trav. Chim.
Pays-Bas 1906, 25, 75. In contrast to the given protocol, we
used 25% aqeuous NH3, resulting in a complete and clean
formation of the amide.
17. Boksiner, E. I.; Golubyatnikova, A. A.; Fel’dman, I. Kh.
Zh. Obshch. Khim. 1968, 38, 999. Kato, T.; Takada, A.; Ueda,
T. Chem. Pharm. Bull. 1972, 20, 901. Weis, A. L.; Frolow, F.;
Zamir, D.; Bernstein, M. Heterocycles 1984, 22, 657.
18. The structure of 15 and its tautomer were preminimized at
the MM-level with MAXIMIN2, using the Tripos force field
implemented in Sybyl 6.6. Employing the 6-31G* basis set and
the ab initio program Gamess (version of 25.03.2000), a geo-
metry optimization of both tautomers was performed, result-
ing in a significant energy difference of 5.3 kcal/mol in favour
of tautomer 15.
References and Notes
1. Einsiedel, J.; Hubner, H.; Gmeiner, P. Bioorg. Med. Chem.
Lett. 2001, 11, 2533, and references cited therein.
2. McCracken, J. T.; Smalley, S. L.; McGough, J. J.; Craw-
ford, L.; Del’Homme, M.; Cantor, R. M.; Liu, A.; Nelson,
S. F. Mol. Psychiatry 2000, 5, 531. Lichter, J. B.; Barr, C. L.;
Kennedy, J. L.; Van Tol, H. H. M.; Kidd, K. K.; Livak, K. J.
Hum. Mol. Genet. 1993, 2, 767.
19. Michel, D.; Waibel, R.; Gmeiner, P. Heterocycles 1999,
51, 365.
3. Lober, S.; Hubner, H.; Utz, W.; Gmeiner, P. J. Med. Chem.
2001, 44, 2691. Bettinetti, L.; Schlotter, K.; Hubner, H.;
Gmeiner, P. J. Med. Chem. 2002, 45, 4594.
4. Gmeiner, P.; Hummel, E.; Haubmann, C.; Hofner, G.
Arch. Pharm (Weinheim Ger.) 1995, 328, 265.
5. 10a was derivatized by the reaction with (R)-phenylethyl
isocyanate. HPLC (Nucleodex beta-PM; MeOH/HNE-
t3OAc1%,pH4 98:2; 1 mL/min; 250 nm): SR (rt=25.01 min):
RR (rt=22.35 min)=98:2.
20. 3 and ent3 were investigated by performing chiral HPLC
(Chiralcel OD; petroleum ether/iPrOH 1:1; 1 mL/min; 250
nm): 3: S (rt=16.1 min): R (rt=13.0 min):=96:4. ent3:
R:S=99.4:0.6.
21. Hayes, G.; Biden, T. J.; Selbie, L. A.; Shine, J. Mol.
Endocrinol. 1992, 6, 920.
22. Sokoloff, P.; Andrieux, M.; Besancon, R.; Pilon, C.;
Martres, M.-P.; Giros, B.; Schwartz, J.-C. Eur. J. Pharmacol.
1992, 225, 331.
6. Thomas, C.; Orecher, F.; Gmeiner, P. Synthesis 1998, 1491.
7. Wang, C.-L. J.; Taylor, T. L.; Mical, A. J.; Spitz, S.; Reilly,
T. M. Tetrahedron Lett. 1992, 33, 7667.
23. Agashari, V.; Sanyal, S.; Buchwaldt, S.; Paterson, A.;
Jovanovic, V.; Van Tol, H. H. M. J. Neurochem. 1995, 65,
1157.
8. Speicher, A.; Bomm, V.; Eicher, T. J. Prakt. Chem. 1996,
338, 588.
24. Hubner, H.; Haubmann, C.; Utz, W.; Gmeiner, P. J. Med.
Chem. 2000, 43, 756.
9. Weber, K.; Kuklinski, S.; Gmeiner, P. Org. Lett. 2000, 2, 6 47.
10. ent10a synthesized from 9a was reacted with (R)-PEI
25. Hubner, H.; Kraxner, J.; Gmeiner, P. J. Med. Chem. 2000,
43, 4563.