Dopamine Receptor Ligands. Part VII [1]: Novel 3-Substituted 5-Phenyl-1,2,3,4,5,6-hexahydro-azepino-[4,5-b]indoles as Ligands for the Dopamine Receptors

Article abstract of DOI:10.1002/ardp.200300777

A number of 5-phenyl-1,2,3,4,5,6-hexahydro-azepino-[4,5-b]indoles 3 were synthesized with different substituents at the azepine-N position (methyl-, allyl-, 2-phenyl-ethyl-, cyclopropylmethyl- and unsubstituted). Furthermore, the indole-N-methylated compound was generated and by using norephedrines and norpseudoephedrines as a chiral pool, 4-methyl-5-phenyl-1,2,3,4,5,6-hexahydro-azepino-[4,5-b]indoles were prepared which contained racemisation at the reacting C-atom. These compounds, as well as the ring-open amino-alcohols, were screened for their affinity to the hD ₁-, hD₅-, hD_2L-, and hD₄-receptors (s please check sentence). They had micromolar affinities for the receptors and showed the highest affinity to the D₁-subtype family. The cyclic compounds possessed the highest affinity, with the cyclo-propylmethyl-(3 c) and methyl-substituents (3 e) being the most active of the tested compounds. Based on an intracellular cAMP-assay, the unsubstituted compound (at the azepine-N position) turned out to be an agonist for the D ₁- and D₅-subtype family, whereas the substituted compounds showed (partial) agonistic, or even inverse agonistic activity.

Full text of DOI:10.1002/ardp.200300777

466 Decker and Lehmann
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Michael Decker,
Jochen Lehmann
Dopamine Receptor Ligands.
Part VII [1]: Novel 3-Substituted
Institut für Pharmazie,
5-Phenyl-1,2,3,4,5,6-hexahydro-azepino-[4,5-b]indoles
as Ligands for the Dopamine Receptors
Pharmazeutische/Medizinische
Chemie, Friedrich-Schiller
Universität Jena, Germany
A number of 5-phenyl-1,2,3,4,5,6-hexahydro-azepino-[4,5-b]indoles 3 were synthe-
sized with different substituents at the azepine-N position (methyl-, allyl-, 2-phenyl-
ethyl-, cyclopropylmethyl- and unsubstituted). Furthermore, the indole-N-methylat-
ed compound was generated and by using norephedrines and norpseudoephe-
drines as a chiral pool, 4-methyl-5-phenyl-1,2,3,4,5,6-hexahydro-azepino-[4,5-
b]indoles were prepared which contained racemisation at the reacting C-atom.
These compounds, as well as the ring-open amino-alcohols, were screened for their
affinity to the hD1-, hD5-, hD2L-, and hD4-receptors (çs please check sentence).They
had micromolar affinities for the receptors and showed the highest affinity to the D1-
subtype family.The cyclic compounds possessed the highest affinity, with the cyclo-
propylmethyl-(3 c) and methyl-substituents (3 e) being the most active of the tested
compounds. Based on an intracellular cAMP-assay, the unsubstituted compound
(at the azepine-N position) turned out to be an agonist for the D1- and D5-subtype
family, whereas the substituted compounds showed (partial) agonistic, or even in-
verse agonistic activity.
Keywords: Indolo-azepines; Dopamine receptors; Chiral pool; Inverse agonists
Received: January 23, 2003; Accepted: February 21, 2003 [FP777]
DOI 10.1002/ardp.200300777
Introduction
A number of compounds are known to interact selective-
ly with the hD1-receptor including the well known hD1-se-
lective antagonist SCH 23390 2 a [2], the agonist SKF
38393 2 b [3] and the indoloazezine LE 300 1, a new
lead structure containing both a tryptamine unit and a
β-phenylethylamine structure [4, 5] (sç please check sen-
tence). We were interested in synthesizing correspond-
ing indoloazepines 3 a–i with the goal of generating a
new class of antipsychotics selective for the hD1-recep-
tor family but which lack the extrapyramidal side effects
of classical antipsychotic drugs [6] (Figure 1).
The stereochemistry of a compound determines wheth-
er it will have a high affinity interaction with members of
the dopamine receptor family. We wanted to develop a
synthesis for dopamine receptor ligands, in which we
could make use of the chiral pool of norephedrines 4 and
ephedrines 5 (Scheme 1; the [1R, 2S]-enantiomers are
shown) as synthones (sç please check sentence).
Figure 1. Lead structures (LE 300 1, SCH 23390 2 a,
SKF 38393 2 b) and novel 5-phenyl-1,2,3,4,5,6-hexahy-
dro-[4,5-b]indoles 3 a–i.
Results
Correspondence: Jochen Lehmann, Institut für Pharmazie,
Pharmazeutische/Medizinische Chemie, Friedrich-Schiller-
Universität Jena, Philosophenweg 14, D-07743 Jena,
Germany. Phone: +49 3641 949803, Fax: +49 3641 949 802,
e-mail: j.lehmann@uni-jena.de
Chemistry
In order to use norephedrines as synthones, we made
use of a modified synthesis procedure described by
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Novel dopamine receptor ligands 467
Scheme 1. Synthesis of compounds 8 a–c, 9 a–c and 3 a.
Elliot et al.[7].Amides (8 a–c) were prepared by first acti-
vating 1H-indol-3-ylacetic acid with N, NЈ-carbonyldi-
imidazole and subsequently reacting it with 2-amino-1-
phenyl-ethanol 7 or norephedrines 4 and ephedrines 5,
respectively.The amides were then reduced to the corre-
sponding amines (9 a–c) using lithium aluminium hy-
dride. Subsequent cyclisation of the compounds to
azepines was performed with polyphosphoric acid (PPA)
(Scheme 1).The indoloazepines were directly alkylated
to tertiary amines with the appropriate alkyl bromides
using potassium carbonate in dimethylformamide (DMF)
as described for the benzoazepines [8] (Scheme 2).
Methylation at the azepine-N atom was achieved by re-
action with ethyl chloroformate followed by reduction
with lithium aluminium hydride [6].Methylation at both ni-
trogen atoms was achieved by the use of methyl iodide
with sodium hydride in tetrahydrofuran (THF) (Scheme
2). The use of ephedrine and norephedrine allowed the
syntheses of a variety of ring-open compounds.By using
As previously described [10], racemisation occurs dur-
ing cyclisation to the seven-membered ring at the chiral
centre which contains a hydroxy-group (Scheme 3). Op-
tical purity during the course of the reaction, and final ra-
cemisation, which resulted in the formation of diaster-
eomers (3 g–i), was determined using capillary electro-
phoresis (NMR spectra at 300 MHz were identical) with
heptakis(2,3-O-diacetyl-6-sulfato)b-cyclodextrin as
a
chiral selector [10, 11].The chemical and analytical data
for the respective compounds has been previously de-
scribed as were the detailed conditions for analytical
separation using CE [10, 11].
Pharmacology
The compounds were tested in radioligand binding stud-
ies using Chinese hamster ovary (CHO)-cells stably ex-
pressing hD1-, hD2L-, hD4- and hD5-receptors.In an initial
screen, the decrease in receptor-bound radioactivity, by
a 10 µM solution of the test compound, was measured.
When the decrease was greater then 70 %, the KI-value
(1-methyl-1H-indol-3-yl)acetic acid, prepared by
a
Fischer-indole reaction [9], the corresponding indole-N-
methylated compounds could be prepared (seeTable 2).
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

468 Decker and Lehmann
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Table 1. Specifications of indoloazepines 3 a–i and their radioligand binding results (see Figure 1 for the basic struc-
ture). A100 % decrease in radioactive binding was reached by using fluphenazine (10 µM) for the D1 and D5 receptors
and haloperidol (1 µM) for the D2 and D4 receptors.
Compound
R1
R2
R3
Decrease in receptor-
bound radioactivity
by a 10 µM solution
KI-value
± SEM
3 a
H
H
H
D1: –82
D2L: –33
D4: –31
D5: –59
2112 ± 61 nM
3901 ± 27 nM
3 b
3 c
3 d
3 e
3 f
2-Phenylethyl-
H
H
H
H
H
H
H
D1: –73
D2L: –23
D4: –24
D5: –19
Cyclopropylmethyl-
D1: –99
D2L: –76
D4: –56
D5: –85
214 ± 19 nM
1572 ± 26 nM
1424 ± 52 nM
2781 ± 43 nM
Allyl-
CH3
CH3
H
H
D1: –75
D2L: –36
D4: –51
D5: –41
H
D1: –95
D2L: –59
D4: –56
D5: –72
750 ± 34 nM
2413 ± 31 nM
CH3
D1: –66
D2L: –35
D4: –19
D5: –48
3 g
3 h
3 i
CH3
(bound to
R-configurated C-atom)
H
D1: –54
D2L: –39
D4: –22
D5: –46
H
CH3
(bound to
S-configurated C-atom)
H
D1: –45
D2L: –4
D4: –13
D5: –22
CH3
CH3
(bound to
R-configurated C-atom)
H
D1: –66
D2L: –5
D4: –7
D5: –69
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Novel dopamine receptor ligands 469
Scheme 2. Synthesis of compounds 3 b–d and 3 f.
was also determined.The results are summarized in Ta-
bles 1 and 2. Curves for the most potent compounds at
the hD1 receptor are shown in Figure 2.
Three of the cyclic compounds were also tested in a cel-
lular assay, in which the change in cAMP formation was
measured after incubation with the test compounds.
Specifically, Human embryonic kidney (HEK)-cells ex-
pressing the hD1, hD2L and hD5 receptors were prein-
cubated with forskolin, which unspecifically stimulates
adenylate cyclase. Preincubation was necessary to pro-
duce adequate levels of cAMP for detection. Cells that
expressed the hD1 and hD5 receptors, could then be
Scheme 3. Synthesis and partial racemisation of nore-
phedrinoid indoloazepines [10].
Table 2. Radioligand binding results for ring–open compounds 9 a–e (preparation and properties of 9 d and e have
been previously described [10]). A 100 % decrease in radioactive binding was reached by using fluphenazine (10 µM)
for the hD1 and hD5 receptors and haloperidol (1 µM) for the hD2 and hD4 receptors (çs please check sentence).
Starting materials
Compound
Decrease in receptor-bound
radioactivity by a 10 µM solution
2-Amino-1-phenyl-ethanol
and 1H-indol-3-ylacetic acid
D1: –45
D2L: –26
D4: –27
D5: –22
(1R, 2S)-Ephedrine and
(1-methyl-1H-indol-3-yl)acetic acid
D1: –55
D2L: –51
D4: –27
D5: –42
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

470 Decker and Lehmann
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Table 2. (continued).
Starting materials
Compound
Decrease in receptor-bound
radioactivity by a 10 µM solution
(1R, 2S)-Norephedrine
and (1-methyl-1H-indol-3-yl)acetic
acid
D1: –69
D2L: –48
D4: –71°
D5: –46
(1R, 2S)-Norephedrine
and 1H-indol-3-ylacetic acid
D1: –42
D2L: –23
D4: –31
D5: –23
(1S, 2R)-Norephedrine
and 1H-indol-3-ylacetic acid
D1: –22
D2L: –54
D4: –23
D5: –9
° (KI = 492 ± 21 nM)
Figure 2a.
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Novel dopamine receptor ligands 471
Figure 2b.
Figure 2. Heterologous competition experiments at the hD1 receptor using [3H]-SCH23390 and compounds 3 c and
3 e. Shown is a representative example of one of the three experiments performed.
Figure 3. Influence of test compounds 3 a, c, e and the hD1 receptor agonist SKF 38393 (2 b) on intracellular cAMP for-
mation in HEK-D1 cells (Gs coupled).
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

472 Decker and Lehmann
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Figure 4. Influence of the test compounds 3 a, c, e and the hD1 receptor agonist SKF 38393 (2 b) on intracellular cAMP
formation in HEK-D5 cells (Gs coupled).
Figure 5.Influence of the test compounds 3 a, c, e and the hD2 receptor agonist quinpirole on intracellular cAMP forma-
tion in HEK-D2L cells (Gi coupled).
stimulated with agonists (Figures 3 and 4), and dose-re-
sponse curves generated. An increase in cAMP forma-
tion indicates that the test compound possesses intrinsic
affinity and acts therefore as an agonist (Gs coupling).
With the D2L receptor, which is Gi coupled, an agonist
lowers cAMP formation. The results are summarized in
Figures 3, 4 and 5.
Discussion
While structure-activity relationships cannot be deduced
from the initial screen, general trends may become
apparent from the data. Both ring-open as well as cyclic
compounds showed affinity in the micromolar range
to both dopamine receptor subtype families. The
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Novel dopamine receptor ligands 473
cyclic compounds showed the highest affinity for the
receptors and also appeared to have greater selectivity
for the D1-receptor family (3 a–i; 9 a–e). A methyl-group
at the azepine-N atom (3 e, Figure 2) resulted in a
compound with higher affinity to the hD1 receptor com-
pared to the N-H compound (3 a). With the exception of
the cyclopropylmethyl-group (3 c, Figure 2), sterically
larger substituents gave lower affinities (3 b, d). 3 c was
the most potent of the tested compounds. A second
methyl-group at the indole-N seemed to decrease the
corresponding affinities, especially for the cyclic com-
pounds (3 f). This also seemed to be the case for a
methyl-group (originating from an ephedrine or nore-
phedrine) bound to the aliphatic part of the molecules
(3 g–i).
Experimental
Chemistry
Melting points were determined on a “Melting Point Apparatus”
by Gallenkamp in open capillary tubes and are uncorrected.
Elemental analyses were carried out on “Vario EL” by Elemen-
tar. 1H-NMR spectral data were obtained on a “Varian XL 300”
(300 MHz). Unless otherwise stated, the solvent was d6-
DMSO. IR-data were measured with a “1420” by Perkin Elmer.
General procedure for synthesis of 3-alkyl-5-phenyl-
1,2,3,4,5,6-hexahydroazepino-[4,5-b]indoles 3 b–d
5-Phenyl-1,2,3,4,5,6-hexahydroazepino-[4,5-b]indole
(3 a,
2.3 mmol, 0.6 g) and anhydrous potassium carbonate
(2.5 mmol, 0.35 g) were dissolved or suspended, respectively,
in a mixture of 6 mL of DMF and 0.25 mL of water. A solution of
2.5 mmol of the appropriate alkylbromide in 3 mL of dichlo-
romethane was added over a 30 min period and the mixture
was stirring under nitrogen over night.The solution was poured
into 50 mL of water, the organic phase separated, the aqueous
phase extracted twice with 10 mL of dichloromethane, and the
combined organic phases washed with water and brine. It was
dried over anhydrous calcium chloride and the solvent re-
moved under reduced pressure. The crude product was puri-
fied by column chromatography (200 g SiO620, CH2Cl2/MeOH/
NH3conc. = 85/14/1).
From the results of the cAMP testing at the hD1 and
hD5 receptors (Figures 3 and 4), it could be shown that
only the N-H compound (3 a) is an agonist, that stimu-
lates adenylate cyclase via the corresponding receptor.
This has already been shown for the benzo-azepines
(2 a and 2 b, the latter showed a stronger signal in our
experiments due to its much higher affinity for the re-
ceptors) [2, 3]. The N-methyl compound (3 e) also
showed some intrinsic activity at the hD5 receptor.
However, considering the fact that it generally showed
a higher affinity at the dopamine receptor and there
was only a moderate increase in cAMP formation, the
compound seems to be a partial agonist with fairly
small agonistic activity.
3-Phenylethyl-5-phenyl-1,2,3,4,5,6-hexahydro-[4,5-b]indole 3 b
Yield: 0.67 g (76 %), yellow crystals, mp 121 °C; 1H-NMR: 2.6–
3.19 (m, 10 H, aliphat. -CH2-), 4.3 (s, 1 H, -CH-), 6.9–7.4 (m,
14 H, aromatic H), 10.3 (s, 1 H, indole-NH); IR (KBr, cm–1):
3446, 2926, 1454, 1336, 1123, 750, 702; C26H26N2 · H2O
(384.2) calc. (CHN): 81.2, 7.3, 7.3 found: 81.1, 6.9, 7.0.
3-Cyclopropylmethyl-5-phenyl-1,2,3,4,5,6-hexahydro-[4,5-
b]indole 3 c
Another remarkable feature is the moderate decreas-
es of cAMP formation (apart from compound 3 e at the
hD1 receptor, all of the values determined are statisti-
cally significantly different from the blank) during incu-
bation with the cyclopropylmethyl-compound 3 c,
which was also observed with many “antagonistic”
compounds not described in this paper, including LE
300 [1]. Furthermore, this observation is consistent
with recent findings by M. W. Martin et al., which
showed that typical antipsychotics (including fluphen-
azine and thioridazine) showed inverse agonist activity
at the rat D1-like receptor [12]. For the hD2L receptor,
the standard agonist quinpirole lowered cAMP forma-
tion as expected, but none of the test compounds al-
tered its formation at the concentrations tested (Figure
5).
Yield:0.61 g (81 %), yellow-brown crystals, mp 45 °C;1H-NMR:
0.02 (m, 2 H, cyclopropane -CH2-), 0.38 (m, 2 H, cyclopropane
-CH2-), 0.7 (m, 1 H, cyclopropane -CH-), 2.4 (dq, 2J = 11 Hz, 3J
= 6.5 Hz, 2 H, cyclopropyl-CH2-); 2.8–3.1 (m, 5 H, aliphat. H);
3.43 (dd, 2J = 13 Hz, 3J = 6 Hz, 1 H, Ph-CHCH-), 4.3 (dd, 3J =
6 Hz, 3J = 5.5 Hz, 1 H, Ph-CH-), 6.9–7.48 (m, 9 H, aromatic H),
10.3 (s, 1 H, indole-NH); IR (KBr, cm–1): 3409, 2361, 1654,
1458, 742, 700; C22H20N2 · ½ H2O (325.2) calc. (CHN): 81.2,
7.7, 8.6 found: 80.7, 7.5, 8.7.
3-Allyl-5-phenyl-1,2,3,4,5,6-hexahydro-[4,5-b]indole 3 d
Yield: 0.52 g (73 %), yellow-brownish crystals, mp 77 °C; 1H-
NMR: 2.7–3.22 (m, 8 H, aliphatic -CH2-), 4.31 (dd, 3J = 3.5 Hz,
3J = 3 Hz, 1 H, Ph-CH), 5.0–5.15 (m, 2 H, C=CH2), 5.6–5.75 (m,
1 H, -CH=CH2), 6.9–7.45 (m, 9 H, aromatic H), 10.3 (s, 1 H, in-
dole-NH); IR (KBr, cm–1): 3422, 2361, 1654, 1458, 668;
C21H22N2 · ¼ H2O (306.9) calc. (CHN): 82.2, 7.4, 9.1 found:
82.1, 7.3, 9.1.
3,6-Dimethyl-5-phenyl-1,2,3,4,5,6-hexahydro-azepino-[4,5-
In summary, cyclic compounds with micromolar affinities
were synthesized (3 a–e), that showed some selectivity
for the hD1 receptor (3 c had ~eight times higher affinity
to hD1 than to hD2L or hD5).Interestingly, in contrast to the
azepine N-H compound (3 a) which is an agonist, the da-
ta suggests that compounds 3 c and 3 e are a partial and
an inverse agonist, respectively.
b]indole 3 f
5-Phenyl-1,2,3,4,5,6-hexahydroazepino-[4,5-b]indole
(3 a,
7.6 mmol, 1.98 g) was dissolved in 10 mL of dried THF.The so-
lution was poured, under nitrogen and ice-cooling, into a sus-
pension of 0.125 mol (3 g) sodium hydride in 150 mL of THF.
Under cooling, a solution of 30 mmol (4.26 g) methyliodide in
15 mL of anhydrous THF was added dropwise. Stirring was
continued overnight at room temperature. Sodium hydride was
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

474 Decker and Lehmann
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
destroyed with a 1:1 mixture of water/methanol, the precipita-
tion filtered and then resuspended in THF. It was refiltered, the
combined filtrates dried over anhydrous sodium sulphate and
the solvent removed in vacuo. The crude product was purified
by column chromatography (200 g SiO620, CH2Cl2/MeOH/
NH3conc. = 85/14/1).
1054, 742, 668; C18H18N2O2 · ¼ H2O (299.9) calc. (CHN): 72.1,
6.5, 9.3 found: 72.4, 6.1, 9.5.
N-[(1S, 2R)-2-Hydroxy-1-methyl-2-phenylethyl]-N-methyl-2-
(1-methyl-1H-indol-3-yl)acetamide 8 b
Starting materials: (1-methyl-1H-indol-3-yl)acetic acid, (1R,
2S)-ephedrine 5; yield: 14.2 g (82 %), yellowish powder, mp
51 °C; IR (KBr, cm–1): 3368, 2934, 2361, 1616, 1474, 1330,
1117, 743, 701; C21H24N2O2 · ½ H2O (345.4) calc. (CHN): 73.0,
7.3, 8.1 found: 72.5, 7.3, 7.6 (satisfying elemental analysis was
difficult to obtain due to the hygroscopy of the compound).
Yield:0.24 g (10.3 %), yellow crystals, mp 135 °C; 1H-NMR: 2.3
(s, 3 H, azepine-N-methyl), 2.75 (m, 1 H, Ph-CH-CH2-), 3.05–
3.1 (m, 4 H, aliphat.H), 3.4 (d, 2J = 10 Hz, 3J = 5 Hz, 1 H, Ph-CH-
CH2-), 3.45 (s, 3 H, indole-methyl);4.43 (dd, 3J = 5 Hz, 3J = 5 Hz,
1 H, Ph-CH-), 7.1–7.6 (m, 9 H, aromatic H); IR (KBr, cm–1):
3458, 2927, 2361, 1472, 1372, 1123, 735, 701; C20H22N2 · H2O
(308.4) calc. (CHN): 77.9, 7.8, 9.1 found: 78.5, 7.7, 8.2. The
NMR-data confirmed the structure.Satisfying elemental analy-
sis of the hygroscopic compound could not be obtained.
N-[(1S, 2R)-2-Hydroxy-1-methyl-2-phenylethyl]-2-(1-methyl-1H-
indol-3-yl)acetamide 8 c
Starting materials: (1-methyl-1H-indol-3-yl)acetic acid, (1R,
2S)-norephedrine 4; yield: 13.4 g (82 %), light yellow powder,
mp 128.5 °C; IR (KBr, cm–1): 3285, 2361, 1654, 1543, 1253,
736; C21H24N2O2 · ¼ H2O (326.9) calc. (CHN): 73.5, 6.9, 8.6
found: 73.8, 6.9, 8.6.
5-Phenyl-1,2,3,4,5,6-hexahydroazepino-[4,5-b]indole 3 a
2-{[2-(1H-Indol-3-yl)ethyl]amino}-1-phenylethanol (9 a, 0.025
mol, 6.98 g) was added to a stirred mixture of 500 mL of chloro-
form and 200 mL of PPA.The mixture was stirred and boiled un-
der reflux (of chloroform) for 90 min. After cooling, the organic
phase was decanted, and under ice-cooling and heavy stirring,
1 L of water was added until complete homogeneity was
reached. The solution was then made basic (pH = 8), under
continued ice-cooling, with 6N NaOH.The product was extract-
ed twice with 500 mL of ethylacetate. The combined organic
phases were washed twice with water, dried with anhydrous so-
dium sulphate and the solvent was removed under reduced
pressure. Chromatography on silica (500 g SiO620, CH2Cl2/Me-
OH/NH3conc. = 85/14/1) gave 1.03 g of 3 a (15 %).
General procedure for the synthesis of 9 a–c
The corresponding amide (8 a–c, 0.05 mol) was dissolved in
80 mL of dried THF. The solution was slowly added to an ice-
cold suspension of 2 mol (75.9 g) lithium aluminium hydride in
300 mL of dried diethylether.The suspension was boiled under
reflux and stirred for 18 hours. After cooling, excess LiAlH4 was
destroyed with 0.5 N NaOH, which was carefully added drop-
wise under ice-cooling until the formation of hydrogen was fin-
ished.
1
yellowish crystals, mp 141 °C; H-NMR (deuteriochloroform):
The inorganics were filtered, washed intensively twice with
ether and the filtrate was dried with MgSO4 and evaporated in
vacuo. The products were purified by column chromatography
on silica (500 g SiO260, CH2Cl2/MeOH/NH3conc. = 85/14/1).
3.0–3.38 (m, 6 H, aliphat. H), 4.23 (dd, J = 4 Hz, 3J = 3.5 Hz,
1 H, Ph-CH-), 7.04–7.68 (m, 9 H, aromatic H), 10.23 (s, 1 H, in-
dole-NH);IR (KBr, cm–1):3399, 2361, 1456, 743, 701;C18H18N2
· ¾ H2O (275.5) calc.(CHN):78.4, 7.4, 9.7 found:78.7, 6.9, 9.8.
3
Synthesis of the analogous compounds 9 d, e (see Table 2) out
of 1H-indol-3-ylacetic acid and norephedrines was previously
described [10].
General procedure for the synthesis of 8 a–c
1H-indol-3-ylacetic acid (0.05 mol, 8.76 g) or (1-methyl-1H-in-
dol-3-yl)acetic acid (0.05 mol, 9.46 g), were dissolved in 50 mL
of driedTHF and 0.05 mol (8.11 g) of N, NЈ-carbonyldiimidazole
(CDI) was added. Following formation of carbon dioxide
(~15 min), the solution of activated acid was allowed to stand
for three hours. It was then slowly added dropwise, during vig-
orous stirring, to a solution of 0.05 mol of the corresponding
amino-alcohol (7, ephedrine 5, or norephedrine 4, respective-
ly).The resulting mixture was stirred at room temperature over
night. The solution was washed once with 25 mL of 1N H2SO4
and twice with water.During the course of the second wash with
water the separation of phases does not take place and addi-
tional water has to be added for separation. This time, the or-
ganic phase is the lower one. The organic phase is dried over
MgSO4 and the solvent is removed under reduced pressure.
2-[2-(1H-Indol-3-yl)ethylamino]-1-phenylethanol 9 a
Yield: 8.4 g (58 %), yellow powder, mp 120 °C; 1H-NMR: 2.75–
2.9 (m, 6 H, -CH2-groups), 4.65 (t, 3J = 5 Hz, 1 H, -CH-), 6.9–7.5
(m, 10 H, aromatic H), 10.8 (s, 1 H, indole-NH); IR (KBr, cm–1):
3295, 2894, 2361, 1451, 1421, 1066, 819, 734, 702;C18H20N2O
· ½ H2O (289.4) calc.(CHN):75.0, 7.0, 9.7 found:75.0, 7.2, 9.8.
(1R, 2S)-2-{Methyl[2-(1-methyl-1H-indol-3-yl)ethyl]amino}-1-
phenylpropan-1-ol 9 b
Yield: 5.3 g (32 %), yellow powder, mp 51 °C; 1H-NMR: 0.96 (d,
3J = 7 Hz, 3 H, aliphatic -CH3), 2.36 (s, 3 H, N-CH3), 2.48–2.9
(m, 5 H, aliphatic H), 3.65 (s, 3 H, indole-CH3), 4.6 (s[b], 1 H,
Ph-CH-), 5.02 (s[b], 1 H, -OH), 6.9–7.45 (m, 10 H, aromatic H);
IR (KBr, cm–1): 3422, 2361, 1473, 740, 668; C21H26N2O · ½ H2O
(331.5) calc. (CHN): 76.1, 8.2, 8.5 found: 76.5, 8.2, 8.3.
The synthesis of analogous compounds out of 1H-indol-3-
ylacetic acid and norephedrines was previously described [10].
(1R, 2S)-2-{[2-(1-Methyl-1H-indol-3-yl)ethyl]amino}-1-phenyl-
propan-1-ol 9 c
N-(2-Hydroxy-2-phenylethyl)-2-(1H-indol-3-yl)acetamide 8 a
Yield: 5.6 g (36 %), yellow oil; 1H-NMR: 0.91 (d, 3J = 7 Hz, 3 H,
N-CH3), 2.74–2.9 (m, 5 H, aliphat. H), 3.78 (s, 3 H, indole-CH3),
4.62 (d, 3J = 5 Hz, 1 H, -CH-Ph), 7.0–7.5 (m, 10 H, aromatic H);
IR (KBr, cm–1): 3446, 2361, 1654, 1474, 736, 698, 668;
C20H24N2O (308.4) calc. (CHN): 77.9, 7.8, 9.1 found: 77.5, 7.7,
8.8
Starting materials:1H-indol-3-ylacetic acid, 2-amino-1-phenyl-
ethanol 7; yield: 12.8 g (86 %), white powder, mp 132 °C; 1H-
NMR: 3.1–3.6 (m, 4 H, -CH2-groups), 4.57 (s, 1 H, Ph-CH-),
5.45 (s, 1 H, amide-NH), 6.95–7.6 (m, 10 H, aromatic H), 10.9
(s, 1 H, indole-NH); IR (KBr, cm–1): 3348, 2369, 1643, 1558,
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
Novel dopamine receptor ligands 475
Pharmacology
Protein® (Beckman, USA) using a Beckman LS 6000 SC scintil-
lation counter.The competition binding data was analyzed with
GraphPad PrismTM software using nonlinear least squares fit.
Experiments for determining the KI values were carried out
three times using two different dilutions of drugs.
The methods used have been previously described in further
detail with special regards to the pharmacological behaviour of
compound 1 (LE 300) [5].
Materials
cAMP assay
[3H]-Spiperone and [3H]-SCH 23390 were purchased from
Amersham, UK, and had a specific activity of 97,0 Ci/mmol and
83,0 Ci/mmol, respectively.
A commercially available cAMP Assay Kit from Amersham, UK,
was used to measure the formation of cAMP. Human D1, D2L
and D5 receptors were stably expressed in HEK 293 cells as
previously described by Dal Toso et al.[15].The densities of D1
and D5 receptors measured with [3H]-SCH 23390 were
8183.72 fmol/mg protein and 7021.1 fmol/mg protein, respec-
tively. The density of D2 receptors measured with [3H]-Spiper-
one was 2022.45 fmol/mg protein. The cells were grown at
37 °C under a humified atmosphere of 5 % CO2:95 % air in Dul-
becco’s modified Eagles minimal essential medium (MEM,
Sigma-Aldrich) with 15 mM HEPES, pyridoxine and NaHCO3,
supplemented with 10 % fetal bovine serum, 1mM L-glutamine,
20 U/ml penicillin G, 20 µg/ml streptomycin and 0,2 µg/ml G 418
(all from Sigma-Aldrich). HEK cells were harvested as de-
scribed above. The cells were washed with 10 mL of Krebs-
HEPES buffer, pH = 7.4, and the resulting pellet resuspended
in 3 mL of the same buffer.
Methods
Cell culture
Human D1, D2L, D4, and D5 receptors were stably expressed in
Chinese hamster ovary (CHO) cells as previously described by
Sunahara et al. [13]. The cDNAs for the transfection were pro-
vided by Dr. D.Grandy (Portland, OR, USA) (D1, D5) and Dr.
Shine (Darlinghurst, AUS) (D2L). The stably transfected CHO-
D4.4 cell line was kindly donated to us by Dr. H.H.M. Van Tol
(Toronto, CA).The donations are gratefully acknowledged.The
densities of receptors measured with [3H]-SCH 23390 were
307.15 fmol/mg protein for the D1 receptor and 679.44 fmol/mg
protein for the D5 receptor.The densities of receptors measured
with [3H]-Spiperone were 2020.92 fmol/mg protein for the D2L
receptor and 137.21 fmol/mg protein for the D4 receptor. Cells
were grown at 37 °C under a humidified atmosphere of 5 %
CO2: 95 % air in HAM/F12-medium (Sigma-Aldrich) supple-
mented with 10 % fetal bovine serum, 1mM L-glutamine,
20 U/ml penicillin G, 20 µg/ml streptomycin and 0,2 µg/ml G
418 (all by Sigma-Aldrich).
Assays were performed in duplicate two independent times
(means are given in Figures 3, 4 and 5). The incubation was
carried out in a volume of 100 µL which contained the whole-
cell-suspension (80 µL), forskolin solution (10 µL, final concen-
tration 10 µM) and 10 µL distilled water (for determining the
blank) or test substance solution (10 µM final concentration).
Reaction tubes were incubated for 15 min at 37 °C and the re-
action stopped by the addition of 100 µL of ethanolic HCl (1 mL
1N HCl/100 mL EtOH). The mixture was allowed to stand for
5 min and then centrifuged (14 000 rot/min, 4 °C, 5 min).
Supernatants were stored at –24 °C without any change in
cAMP content.
Preparation of whole-cell-suspensions
Human D2L, D4, D5 and D1 receptor cell lines were grown to
85 % confluency on T 175 culture dishes (Nunc). The medium
was removed and the cells were incubated with 6 mL trypsine-
EDTA-solution (Sigma-Aldrich) to remove the cells from the
culture dish. The resulting suspension was centrifuged
(1000 rot/min, 4 °C, 4 min), the pellet resuspended in 10 mL of
PBS (ice-cooled, calcium- and magnesium-free), pelleted and
the procedure repeated.The resulting pellet was resuspended
in 12 mL ofTris-Mg2+-buffer (5 mM magnesium chloride, 50 mM
Tris-HCl, pH = 7,4) and used directly for the radioligand binding
assay.
To measure the amount of cAMP produced, 80 µL of the super-
natant was dried with a refrigerated vapour trap (SpeedVac SC
100®, RefrigeratedVaporTrap RVT 100®, Savant, UK) and then
dissolved in 25 µL of Tris-EDTA-buffer (0.05 M tris, 4 mM
EDTA), pH = 7.5.
Radio-immuno-assay
The radio-immuno-assay was performed as described by
Amersham. Aliquots of the obtained radioactive solution
(100 µL) were diluted in five mL of scintillation cocktail Ready-
Protein® (Beckman, USA) and counted in a Beckman LS 6000
SC scintillation counter.
Radioligand binding assay
Binding assays, with whole-cell-suspensions, were carried out
in triplicate according to the method described by Mierau et al.
[14].Specifically, assays were carried out in a volume of 1.1 mL
containing Tris-Mg2+-buffer (690 mL), [3H]-ligand (100 µL),
whole-cell-suspension (200 µL) and appropriate drugs
(110 µL). Non-specific binding in the assays containing the D2
or D4 receptors was determined using fluphenazine (10 µM, fi-
nal concentration). Haloperidol (1 µM, final concentration) was
used to determine non-specific binding in the assays contain-
ing the D1 or D5 receptors. Drugs were used at a concentration
of 100 µM in the initial screen and the percentage of removed
radioligand was determined.They were first dissolved in 0.5 mL
of DMSO, and then diluted with distilled water to the desired
concentration. Concentrations of 100, 10, 1 and 0,1 µM were
used to determine the KI-value.The incubation was initiated by
addition of the radioligand and carried out at 27 °C for 2 h.It was
stopped by rapid filtration through a glass fiber filter (GF 6,
Schleicher and Schüll, Germany) previously treated with
0.25 % polyethyleneimine solution (Sigma-Aldrich), and
washed twice with ice-cold water.The radioactivity retained on
the filters was counted in five mL of scintillation cocktail Ready-
Data analysis (cAMP assay)
Measured radioactivity (in decays per minute, dpm) was con-
verted to quantity of cAMP (pmol/tube) by comparison with a
standard curve.The amount of cAMP produced was then relat-
ed to the number of cells, as determined by a cell counter
(0.1 mm depth, 0.0025 mm2, Brand, D), so that the production
of cAMP per cell was obtained.This value was related to cAMP
produced in the absence of a test compound (blank = 1).
References
[1] Th. W. Wittig, Ch. Enzensperger, J. Lehmann, Hetero-
cycles 2003, 60(4), 887–898.
[2] J. Hyttel, Eur. J. Pharmacol. 1983, 91, 153–155.
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

476 Decker and Lehmann
Arch. Pharm. Pharm. Med. Chem. 2003, 336, 466–476
[3] M.P.Seiler in Perspectives in Medicinal Chemistry (Ed.:B.
Testa, W. Fuhrer, E. Kyburz, R. Giger), VCH, Basel, Wein-
heim, 1993, 221–237.
[10] M. Decker, R. Faust, M.Wedig, M. Nieger, U. Holzgrabe, J.
Lehmann, Heterocycles 2001, 55(8), 1455–1466.
[11] M.Wedig, M.Thunhorst, S. Laug, M. Decker, J. Lehmann,
U. Holzgrabe Fresenius J. Anal. Chem. 2001, 371, 212–
217.
[4] Th. Witt, S. Hock, J. Lehmann, J. Med. Chem. 2000, 43,
2079–2081.
[5] M. A. Kassack, B. Höfgen, M. Decker, N. Eckstein, J. Leh-
mann, Naunyn-Schmiedebergs Arch. Pharmacol. 2002,
366, 543–550.
[12] M.W.Martin, A.W.Scott, D.E.Johnston Jr., S.Griffin, R.R.
Luedtke, Eur. J. Pharmacol. 2001, 420, 73–82.
[13] R. K. Sunahara, H.-C. Guan, B. F. O’Dowd, P. Seeman, L.
G.Laurier, G.Ng, S.R.George, J.Torchia, H.H.M.VanTol,
H. B. Niznik, Nature 1991, 350, 614–619.
[6] H. Andersen, B. Nielsen, DN&P 1991, 4(3), 150–157.
[7] A. J. Elliot, E. H. Gold, H. Guzik, J. Med. Chem. 1980, 23,
1268–1274.
[14] J. Mierau, F. J. Schneider, H. A. Ensinger, C. L. Chio, M. E.
Lajiness, R. M. Huff Mol. Pharmacol. 1995, 290, 29–36.
[8] J.L.Neumeyer, N.S.Kula, R.J.Baldessarini, N.Baindur, J.
Med. Chem. 1992, 35, 1466–1471.
[15] R. Dal Toso, B. Sommer, M. Ewert, A. Herb, D. B. Pritchett,
A. Bach, B. D. Shivers, P. H. Seeburg, EMBO J. 1989, 8,
4025–4034.
[9] P.Rosenmund, G.Meyer, I.Hansal, Chem.Ber.1975, 108,
3538–3542.
International Union of Pure and
Applied Chemistry (IUPAC)
P. Heinrich Stahl / Camille G.Wermuth (Eds.)
Handbook of Pharmaceutical Salts
An essential step
in the preclinical phase
of drug development
Properties, Selection, and Use
2002. 388 pages. Hardcover. E 149.00* / sFr 220.00 / £ 85.00.
ISBN 3-906390-26-8.
*The e-Price is valid only for Germany.
O–
Cl
Na+
O
NH
Cl
Contents:
The majority of medicinal chemists in pharmaceutical industry whose primary focus is the
design and synthesis of novel compounds as future drug entities are organic chemists for whom
salt formation is often a marginal activity restricted to the short-term objective of obtaining
crystalline material. Because a comprehensive resource that addresses the preparation, selection,
and use of pharmaceutically active salts has not been available, researchers may forego the
opportunities for increased efficacy and improved drug delivery provided by selection of an
optimal salt.To fill this gap in the pharmaceutical bibliography, we have gathered an international
team of seventeen authors from academia and pharmaceutical industry who, in the contributions
to this volume, present the necessary theoretical foundations as well as a wealth of detailed
practical experience in the choice of pharmaceutically active salts.
The Physicochemical Background: Fundamentals of Ionic
Equilibria • Solubility and Dissolution of Weak Acids, Bases,
and Salts • Evaluation of Solid-State Properties of Salts •
Pharmaceutical Aspects of the Drug Salt Form • Biological
Effects of the Drug Salt Form • Salt-Selection Strategies • A
Procedure For Salt Selection and Optimization • Large-Scale
Aspects of Salt Formation: Processing of Intermediates and
Final Products • Patent Aspects of Drug-Salt Formation •
Regulatory Requirements for Drug Salts in the European
Union, Japan, and the United States • Selected Procedures
for the Preparation of Pharmaceutically Acceptable Salts of
Acids and Bases • Monographs on Acids and Bases
Wiley-VCH, Customer Service Department, P.O. Box 10 11 61 ,
D-69451 Weinheim, Germany, Fax +49 (0) 6201 606-184,
e-mail: service@wiley-vch.de, www.wiley-vch.de
John Wiley & Sons, Ltd., Customer Services Department,
1 Oldlands Way, Bognor Regis, West Sussex, PO22 9SA England,
Fax: +44 (0) 1243-843-296, www.wileyeurope.com
© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim