Fused azaindole derivatives: Molecular design, synthesis and in vitro pharmacology leading to the preferential dopamine D3 receptor agonist FAUC 725

Article abstract of DOI:10.1016/S0960-894X(02)00390-6

Computational studies based on the similarity of molecular electrostatic potential maps initiated the synthesis of the tricyclic target compounds 1 (FAUC 725) and 2. Receptor binding studies at the dopamine receptor subtypes D1, D2_long, D2_short, D3 and D4 showed that the azaindole 1 revealed D3 affinity (K_i=0.54 nM) comparable to the lead pramipexole and enhanced selectivity over D2 and D4. Mitogenesis experiments indicated substantial intrinsic activity for the D3 selective dipropylamine 1. Based on the structure of (S)-3-PPP, bioisosteric replacement and conformational restriction leading to the test compound 2 was not fruitful.

Full text of DOI:10.1016/S0960-894X(02)00390-6

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2377–2380
Fused Azaindole Derivatives: Molecular Design, Synthesis and
In Vitro Pharmacology Leading to the Preferential
Dopamine D3Receptor Agonist FAUC 725
Stefan Lober, Harald Hubner and Peter Gmeiner*
Department of Medicinal Chemistry, Emil Fischer Center, Friedrich-Alexander University, Schuhstraße 19,
D-91052 Erlangen, Germany
Received 28 March 2002; revised 21May 2002; accepted 24 May 2002
Abstract—Computational studies based on the similarity of molecular electrostatic potential maps initiated the synthesis of the
tricyclic target compounds 1 (FAUC 725) and 2. Receptor binding studies at the dopamine receptor subtypes D1, D2long, D2short, D3
and D4 showed that the azaindole 1 revealed D3 affinity (Ki=0.54 nM) comparable to the lead pramipexole and enhanced selectivity
over D2 and D4. Mitogenesis experiments indicated substantial intrinsic activity for the D3 selective dipropylamine 1. Based on the
structure of (S)-3-PPP, bioisosteric replacement and conformational restriction leading to the test compound 2 was not fruitful.
# 2002 Elsevier Science Ltd. All rights reserved.
Dopamine neurotransmission is intimately involved in
the pathophysiology of Parkinson’s disease and schizo-
phrenia.1 There is strong evidence that D2 and D3
receptors exist both postsynaptically and as auto-
receptors controlling dopamine synthesis, release and
neuronal firing. The D3 receptor subtype, which is pre-
ferentially located in the mesolimbic region of the brain,
can be considered as particularly related to affective
disorders.2 Thus, the discovery of selective dopamine
D3 receptor agonists is of current interest for the treat-
ment of unipolar major depression and for affecting the
negative symptoms of schizophrenia.
catechol bioisostere leading to the potent D3 receptor
agonist FAUC 73.5 As an extension of these structure–
activity relationship studies we were intrigued by the
question whether the pyrazolo[1,5-a]pyridine nucleus,
which proved to be a valuable pharmacophoric element
in the field of dopamine D4 receptor ligands,6 could also
serve as a potent structural framework for the con-
struction of novel D3 receptor agonists.
For the molecular design of our target compounds, we
exploited molecular electrostatic potentials (MEPs),
being derived from MOPAC ESP charges of geo-
metrically optimized structures, as markers of similarity.
Calculations were conducted by the SYBYL program
package using the AM1-algorithm as implemented in
MOPAC 6.0. Based on the comparison of molecular
electrostatic potential maps of a series of dopamine D2/
D3 agonists with special focus on pramipexole and
(S)-3-PPP, we approached the tetrahydropyrido[1,2-b]-
indazole 1 and the tetrahydrotriazafluorene 2, respec-
tively. The 10p-system of 1 and 2 requires more steric
demand and the six-membered rings are forced to adopt
an approximately coplanar orientation for 2 which is
different to energetically favorable conformations of
(S)-3-PPP. However, both target compounds show high
conformity with their lead structures with respect to the
size, shape and direction of the negative isopotentials
(À2 kcal/mol), indicated in red (Fig. 1). We supposed
this feature to be essential for high D3 affinity.
Conformationally restricted dopamine analogues
including
(R)-7-OH-dipropylaminotetralin
(7-OH-
DPAT) and (S)-3-PPP are known for their high affinity
to the D2 and D3 receptor subtypes.3 Although sub-
stantial in vitro activity is reported, these compounds
are crucial for an application in vivo which is due to
their instability towards metabolic processes. Bioiso-
steric replacement of the hydroxyphenyl substructure of
7-OH-DPAT by an aminothiazole moiety led to the DA
receptor agonist pramipexole.4 We recently reported that
a conjugated enyne moiety can serve as nonaromatic
*Corresponding author. Tel.: +49-9131-852-9383; fax: +49-9131-
852-2585; e-mail: gmeiner@pharmazie.uni-erlangen.de
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00390-6

2378
S. Lober et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2377–2380
¨
Figure 1. Molecular design of the potential dopamine D3 receptor
agonists 1 and 2 based on the similarity of negative MEPs depicted as
isopotential surfaces (À2 kcal/mol).
Scheme 1. (a) K2CO3, HOCH2CH2OH, 130 ꢀC, 16
h (38%);
(b) LiHMDS, DBAD, THF, À78 ꢀC, 0.5 h (54%); (c) LiBEt3H, THF,
À78 ꢀC, 2 h (not isolated); (d) THF, rt, 1h [(c)+(d) 82%]; (e) (1) H
,
2
Ra–Ni, Pd/C, MeOH, rt, 18 h; (2) CH3CH2CHO, NaCNBH3, rt, 6 h
(34%).
As the key reaction step for an effective synthesis of the
conformationally restricted b-arylethylamine 1 in race-
mic form, we planned to exploit our previously reported
methodology involving electrophilic amination and
subsequent reductive degradation.7 Since our efforts to
synthesize the pyridoindazolone framework according
to ref 8 failed, the tricyclic ring system was constructed
by cycloaddition reaction of N-aminopyridinium iodide
and 3-ethoxycyclohexenone, resulting in formation of
the ketone 3 in 38% yield.9
ification resulted in formation of the ethyl carboxylate
8, which was reduced to the corresponding hydroxy-
ethyl derivative 10 in almost quantitative yield. The
alcohol function was activated with MesCl and trans-
formed into the secondary amine 9 by treatment with n-
PrNH2. Finally, ring closure could be accomplished by
intramolecular Mannich reaction, leading to the desired
tetrahydrotriazafluorene 2 in 71% yield.14
Deprotonation with LiHMDS and trapping of the
resulting enolate by addition of dibenzyl azodicarb-
oxylate (DBAD) afforded the Cbz-protected hydrazine
4 (Scheme 1).10 Diastereoselective cis-reduction with
LiBEt3H at À78 ꢀC gave the protected hydrazino alco-
hol 6. Warming up of the reaction mixture to room
temperature furnished the oxazolidinone 5 in 82%
yield.11 Finally, reductive degradation involving
removal of the protecting group as well as N,N-cleavage
and hydrogenolysis of the benzylic C–O bond followed
by reductive alkylation with propanal afforded the final
product 1 in analytically pure form.12
Receptor binding profiles of the test compounds 1 and 2
were established in vitro by measuring their ability to
compete with [3H]spiperone for the cloned human
15
dopamine receptor subtypes D2long, D2short
,
D316 and
D4.417 stably expressed in Chinese hamster ovary cells
(CHO).5 D1receptor affinities were measured employ-
ing bovine striatal membranes and the D1selective
radioligand [3H]SCH 23390.5 Bovine striatal mem-
branes labeled by [3H]pramipexole selectively recogniz-
ing the high affinity ternary complex were utilized for
investigating D2 agonist properties. The resulting Ki
values are listed in Table 1compared to the dopami-
nergic lead compounds pramipexole and (S)-3-PPP.
For the preparation of the 3-PPP analogue 2, we started
from the pyrazolo[1,5-a]pyridine derivative 7, which was
synthesized by cyclization of N-aminopyridinium iodide
with dimethyl acetonedicarboxylate following the pro-
tocol of Awano and Suzue,13 which was slightly mod-
ified to give 7 in 75% yield (Scheme 2). Hydrolysis,
regioselective decarboxylation and subsequent ester-
Our initial investigations involved the observation of
[3H]pramipexole competition, clearly indicating poor
receptor recognition for the test compound 2 (Ki=3100
nM) but substantial high affinity binding of the dipro-
pylamine 1 (Ki=8.6 nM). Both azaindoles showed weak
Table 1. Binding data of the test compounds 1 (FAUC 725) and 2 for the human D2long, D2short, D3 and D4 receptors and the bovine D1and D2
receptors in correlation to the reference compounds pramipexole and (S)-3-PPP [Ki values (nM) based on the means of 2–5 experiments each per-
formed in triplicate]
Test compd
Ki values in (nM)a
[3H]SCH23390
bD1hD2
[3H]Spiperone
[3H]Pramipexole
bD2
hD2short
hD3
hD4.4
long
1
2
18,000
21,000
nd
85/6400
230/53,000
27/5400
35/3700
130/50,000
40/3600
0.54/59
16/16,000
0.87/44
25/1600
50/2200
87/23,000
8.5/130
8.6
3100
1.6 (Kd)
52
Pramipexole
(S)-3-PPP
nd
9.0/1800
27/1800
16/1300
nd, values not determined.
aHigh/low affinity binding sites as indicated by nonlinear regression analysis and further fitting of the calculated curves using PRISM (Graph Pad).

S. Lober et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2377–2380
¨
2379
concentration (EC50) of a test compound and by com-
paring the maximal effect to that of a full agonist.4,19
The data listed in Table 2 clearly show substantial
ligand efficacy of FAUC 725 (1). According to the ratio
of EC50 values, the selectivity profile being determined
by the binding experiments could be confirmed. By
contrast, the conformationally restricted (S)-3-PPP
analogue 2 did not exert significant intrinsic activity
within the investigated dose range.
In conclusion, molecular design based on the similarity
of characteristic MEP maps proved successful for the
development of the preferential D3 agonist FAUC 725
(1). Obviously, the increased steric demand compared to
the lead structure can be tolerated by the receptor
excluded volume. On the other hand, the conformational
restriction involved in the design of the (S)-3-PPP ana-
logue 2 led to a strong reduction of biological activity.
Scheme 2. (a) K2CO3, DMF, 50 ꢀC, 2 h (75%); (b) (1) H2SO4 (40%),
100 ꢀC, 3 h; (2) EtOH/H2SO4, reflux, 6 h (83%); (c) LiAlH4, THF,
0 ꢀC, 1h (99%); (d) (1) MesCl, TEA, THF, 0 ꢀC, 0.5 h; (2) PrNH2,
50 ꢀC, 12 h (86%); (e) CH2O, HOAc, rt, 0.5 h (71%).
Table 2. Intrinsic activities of the fused azaindoles 1 (FAUC 725)
and 2 and the reference compounds pramipexole and (S)-3-PPP in
Acknowledgements
relation to the full agonist quinpirole at the rat D2long, rat D2short
,
human D3 and human D4.2 receptors established by measuring the
stimulation of mitogenesis
The authors wish to thank Dr. H. H. M. Van Tol
(Clarke Institute of Psychiatry, Toronto), Dr. J.-C.
Schwartz and Dr. P. Sokoloff (INSERM, Paris) as well
as Dr. J. Shine (The Garvan Institute of Medical
Research, Sydney) for providing dopamine D4, D3 and
D2 receptor expressing cell lines, respectively. Dr. R.
Huff (Pharmacia & Upjohn, Inc., Kalamazoo, MI,
USA) is acknowledged for providing D2 and D4
expressing cell lines employed for mitogenesis. Thanks
are also due to Mrs. H. Szczepanek, Mrs. P. Schmitt
and Mrs. P. Hubner for skillful technical assistance and
J. Elsner for helpful discussions. This work was sup-
ported by the BMBF, and the Fonds der Chemischen
Industrie.
Test compd
D2long
D2short
D3
D4.2
EC50
(nM)a
EC50
(nM)a
EC50
(nM)a
EC50
(nM)a
1
2
16 (101%)
>1000
47 (92%)
>1000
2.8 (82%)
>1000
430 (73%)
>1000
Pramipexole
(S)-3-PPP
9.2 (85%)
19 (54%)
12 (103%)
19 (82%)
1.5 (93%)
9.8 (68%)
15 (84%)
60 (65%)
aRate of incorporation of [3H]thymidine as evidence for mitogenetic
activity relative to the maximal effect of the full agonist quinpirole
(100%) as the means of quadruplicates from 4 to 10 experiments. EC50
values in nM are derived from the mean curves of the experiments.
D1affinity. Employing the cloned human dopamine
receptor subtypes, careful analysis of the competition
experiments employing a large number of test con-
centrations of 1 (FAUC 725) clearly displayed biphasic
curves for the members of the D2 family when the
respective Hill coefficients and a better fit of equations
indicated a two-site competition. We determined Ki
values of 85, 35, 0.54 and 50 nM for the high affinity
binding sites of D2long, D2short, D3 and D4, respectively,
representing the G-protein coupled ternary complexes.
Furthermore, D3 selectivity proved to be superior when
compared to pramipexole, especially over D2long and
D4. In contrast, the triazafluorene derivative 2 dis-
played disappointing dopamine receptor binding. Due
to common structural features of 1 with 8-OH-DPAT
and derivatives thereof, 5HT1A recognition was eval-
uated with the aid of [3H]8-OH-DPAT-labeled porcine
brain homogenates when the resulting Ki values indi-
cated only moderate 5-HT1A affinity (Ki=2500 nM).
References and Notes
1. Neve, K. A., Neve, R. L., Eds. The Dopamine Receptors.
Humana Press: Totowa, NJ, 1997.
2. Dikeos, D. G.; Papadimitriou, G. N.; Avramopoulos, D.;
Karadima, G.; Daskalopoulou, E. G.; Souery, D.; Mendle-
wicz, J.; Vassilopoulos, D.; Stefanis, C. N. Psychiatr-Genet.
1999, 9, 189.
3. Kebabian, J. W.; Tarazi, F. I.; Kula, N. S.; Baldessarini,
R. J. Drug Discov. Today 1997, 2, 333.
4. Mierau, J.; Schneider, F. J.; Ensinger, H. A.; Chio, C. L.;
Lajiness, M. E.; Huff, R. M. Eur. J. Pharmacol. 1995, 290, 29.
5. Hubner, H.; Haubmann, C.; Utz, W.; Gmeiner, P. J. Med.
Chem. 2000, 43, 756.
6. (a) Lober, S.; Hubner, H.; Gmeiner, P. Bioorg. Med. Chem.
Lett. 1999, 9, 97. (b) Prante, O.; Lober, S.; Hubner, H.;
Gmeiner, P.; Kuwert, T. J. Labelled Cpd. Radiopharm. 2001,
44, 849. (c) Lober, S.; Hubner, H.; Utz, W.; Gmeiner, P.
J. Med. Chem. 2001, 44, 2691. (d) Lober, S.; Aboul-Fadl, T.;
Hubner, H.; Gmeiner, P. Bioorg. Med. Chem. Lett. 2002, 12,
633.
7. (a) Gmeiner, P.; Bollinger, B. Tetrahedron Lett. 1991, 32,
5927. (b) Gmeiner, P.; Bollinger, B. Liebigs Ann. Chem. 1992,
273. (c) Gmeiner, P.; Sommer, J.; Mierau, J.; Hofner, G.
Bioorg. Med. Chem. Lett. 1993, 3, 1477. (d) Gmeiner, P.; Bol-
linger, B. Tetrahedron 1994, 50, 10909.
Ligand efficacy of FAUC 725 (1) and 2 was confirmed
by mitogenesis assays measuring the rate of [3H]thymi-
dine incorporation into growing CHO10001 cells stably
expressing D2long, D2short and D4.218 and D3 receptor
expressing CHOdhfrÀ cells, respectively.16 Intrinsic
activity was quantified by determination of the effective

2380
S. Lober et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2377–2380
¨
8. Tamura, Y.; Tsujimoto, N.; Sumida, Y.; Ikeda, M. Tetra-
hedron 1972, 28, 21 .
mg). After stirring for 18 h under H2-atmosphere at ambient
temperature, the mixture was filtered and the solvent was
evaporated (Caution: The crude primary amine is extremely
sensitive to air!). The residue was resolved in CH2Cl2 (10 mL),
treated with propionaldehyde (35 mL; 0.6 mmol) and
NaCNBH3 (50 mg; 0.8 mmol) and stirred for 6 h at rt. After
addition of 2N NaOH and CH2Cl2, the organic layer was
dried (MgSO4) and evaporated and the residue was purified by
flash chromatography (EtOAc/CH2Cl2/MeOH 10:10:1) to
yield pure 1 (0.36 g, 34%) as a colorless oil. 1H NMR (CDCl3,
360 MHz): d (ppm)=0.91(t, J=7.0 Hz, 6H), 1.50 (tt, J=7.0,
7.0 Hz, 4H), 1.72–1.86 (m, 1H), 2.10–2.19 (m, 1H), 2.52 (t,
J=7.0 Hz, 4H), 2.58–2.66 (m, 1H), 2.81–2.93 (m, 2H), 3.01–
3.12 (m, 2H), 6.62 (ddd, J=7.0, 7.0/1.5 Hz, 1H), 7.01 (dd,
9. 3: A mixture of N-aminopyridinium iodide (2.0 g; 9 mmol),
K2CO3 (2.5 g; 18 mmol) and ethane-1,2-diol (30 mL) was
treated with 3-ethoxycyclohexenone (3.0 g; 21mmol) and stir-
red for 16 h at 130 ꢀC. After cooling to rt, satd NaHCO3 and
Et2O were added. The organic layer was dried (Na2SO4) and
evaporated and the residue was purified by flash chromato-
graphy (petroleum ether/EtOAc 1:1) to yield pure 38 (0.92 g,
38%) as a colorless solid.
10. 4: LiHMDS (1.0 M in hexane; 3.3 mL; 3.3 mmol) was
added to a solution of 3 (0.60 g; 3.2 mmol) in THF (30 mL) at
À78 ꢀC. After 30 min, the mixture was transferred to a pre-
cooled solution of dibenzyl azodicarboxylate (1.32 g; 4.4
mmol) in THF (15 mL) and stirred at À78 ꢀC for additional 5
min. After addition of satd NaHCO3 and EtOAc, the organic
layer was dried (Na2SO4) and evaporated and the residue was
purified by flash chromatography (EtOAc/CH2Cl2/MeOH
J=8.5/7.0 Hz, 1H), 7.32 (d, J=8.5 Hz, 1H), 8.32 (d, J=7.0
+
.
Hz, 1H). EI–MS: m/z 271(M ). Anal. calcd for C17H25N3
1/3H2O: C 73.60; H 9.33; N 15.15. Found: C 73.45; H 9.24;
N 15.17.
1
25:25:1) to yield pure 4 (0.84 g, 54%) as a colorless solid. H
13. Awano, K.; Suzue, S. Chem. Pharm. Bull. 1992, 40, 631.
14. A mixture of 9 (40 mg; 0.2 mmol), formaldehyde (40%
m/m in H2O; 20 mL; 0.27 mmol) and HOAc (2 mL) was
stirred at rt for 30 min. The solution was neutralized with
satd NaHCO3 and extracted with CH2Cl2. The organic layer
was dried (MgSO4) and evaporated and the residue was pur-
ified by flash chromatography (CH2Cl2/MeOH 10:1) to yield
pure 2 (30 mg, 71%) as a colorless oil. 1H NMR (CDCl3,
360 MHz): d (ppm) 0.97 (t, J=7.0 Hz, 3H), 1.66 (tt, J=7.0,
7.0 Hz, 2H), 2.58 (t, J=7.0 Hz, 2H), 2.89 (t, J=6.0, 2H),
3.02 (t, J=6.0, 2H), 3.71(s, 2H), 6.64 (ddd, J=7.0, 7.0, 1.5
Hz, 1H), 7.03 (dd, J=8.5, 7.0 Hz, 1H), 7.27 (d, J=8.5 Hz,
1H), 8.36 (d, J=7.0 Hz, 1H). EI–MS: m/z 215 (M+). Anal.
NMR (CDCl3, 360 MHz): d (ppm) 2.30–2.47 (m, 1H), 2.56–
2.69 (m, 1H), 3.05–3.28 (m, 2H), 4.94–5.30 (m, 5H), 6.92 (br s,
1H), 7.03 (ddd, J=7.0, 7.0, 1.5 Hz, 1H), 7.21–7.37 (m, 10H),
7.49 (dd, J=8.5, 7.0 Hz, 1H), 8.11 (br d, J=8.5 Hz, 1H), 8.49
(br d, 7.0 Hz, 1H); EI–MS: m/z 484 (M+). Anal. calcd for
.
C27H24N4O5 1/4H2O: C 66.30; H 5.05; N 11.46. Found: C
66.20; H 5.16 N 11.30.
11. 5: To a solution of 4 (0.60 g; 1.24 mmol) in THF (12 mL)
was added Super-hydride# (1 M in THF; 1.30 mL; 1.30 mmol)
at À78 ꢀC. The mixture was stirred for 2 h and warmed to rt.
After 1h, satd NaHCO 3 and CH2Cl2 were added. The organic
layer was dried (Na2SO4) and evaporated and the residue was
purified by flash chromatography (EtOAc/CH2Cl2/MeOH
10:10:1) to yield 0.39 g (82%) as a colorless solid. IR (cmÀ1):
.
calcd for C13H17N3 1/2H2O: C 69.61; H 8.09; N 18.73.
Found: C 69.98; H 7.77; N 18.77.
1
3185, 2986, 1781, 1637, 1537, 1509, 1238; H NMR (CDCl3,
15. Hayes, G.; Biden, T. J.; Selbie, L. A.; Shine, J. Mol.
Endocrinol. 1992, 6, 920.
16. Sokoloff, P.; Andrieux, M.; Besanc¸ on, R.; Pilon, C.;
Martres, M.-P.; Giros, B.; Schwartz, J.-C. Eur. J. Pharmacol.
1992, 225, 331.
17. Asghari, V.; Sanyal, S.; Buchwaldt, S.; Paterson, A.;
Jovanovic, V.; Van Tol, H. H. M. J. Neurochem. 1995, 65, 1157.
18. Hubner, H.; Kraxner, J.; Gmeiner, P. J. Med. Chem. 2000,
43, 4563.
360 MHz): d (ppm) 1.93–2.06 (m, 1H), 2.25–2.36 (m, 1H),
2.75–3.03 (m, 2H), 4.46–4.64 (m, 1H), 5.10–5.32 (m, 2H), 5.92
(d, J=7.5 Hz, 1H), 6.78 (ddd, J=7.0, 7.0, 1.5 Hz, 1H), 7.04
(br s, 0.5H rotamers), 7.06 (br s, 0.5H rotamers), 7.21(dd,
J=8.5, 7.0 Hz, 1H), 7.30–7.42 (m, 5H), 7.55 (d, J=8.5 Hz,
1H), 8.41 (d, 7.0 Hz, 1H); EI–MS: m/z 378 (M+). Anal. calcd
.
for C20H18N4O4 1/3H2O: C 62.49; H 4.89; N 14.58. Found: C
62.45; H 4.87 N 14.65.
12. 1: To a solution of 5 (150 mg; 0.4 mmol) in MeOH (10
mL) was added moist Raney–Ni (200 mg) and Pd/C (10%; 50
19. Chio, C.; Lajiness, M. E.; Huff, R. M. Mol. Pharmacol.
1994, 45, 51 .