Dibenzazecine scaffold rebuilding-Is the flexibility always essential for high dopamine receptor affinities?

Article abstract of DOI:10.1016/j.bmc.2009.08.028

The moderately flexible 7-methyl-5,6,7,8,9,14-hexahydrodibenz[d,g]azecines are known to be potent dopamine receptor antagonists, whereas the corresponding rigid dibenzo[d,g]quinolizines are inactive. We built the scaffolds of dibenzo[c,g], [c,f] and [d,f]

Full text of DOI:10.1016/j.bmc.2009.08.028

Bioorganic & Medicinal Chemistry 17 (2009) 6898–6907
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Dibenzazecine scaffold rebuilding—Is the flexibility always essential
for high dopamine receptor affinities?
Maria Schulze a, Franziska K. U. Müller a, Jennifer M. Mason b, Helmar Görls c, Jochen Lehmann a,
Christoph Enzensperger a,
*
a Institut für Pharmazie, Lehtuhl für Pharmazeutische/Medizinische Chemie, Friedrich-Schiller-Universität Jena, Philosophenweg 14, 07743 Jena, Germany
b Industrial Research Ltd, PO Box 31-310, Lower Hutt, New Zealand
c Institut für Anorganische und analytische Chemie, Friedrich-Schiller-Universität Jena, August-Bebel-Str. 2, 07743 Jena, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The moderately flexible 7-methyl-5,6,7,8,9,14-hexahydrodibenz[d,g]azecines are known to be potent
dopamine receptor antagonists, whereas the corresponding rigid dibenzo[d,g]quinolizines are inactive.
We built the scaffolds of dibenzo[c,g], [c,f]- and -[d,f]azecines and together with their ring closed, more
rigid precursors, evaluated the affinities for the human D1–D5 receptors (radioligand binding) as well
as the functionalities (calcium assay) and thus investigated the influence of annelation and conformative
flexibility of these compounds on their affinity for human cloned dopamine receptors.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 13 May 2009
Revised 8 August 2009
Accepted 13 August 2009
Available online 20 August 2009
Keywords:
Dopamine receptor
Azecines
SAR
Heterocyclic compounds
1. Introduction
mobility of the 10-membered dibenzo[d,g]azecines facilitates the
binding to dopamine receptors. On the other hand, some dibenzo-
The dopaminergic system plays an important role in regulating
neuronal motor control, cognition, emotion, and vascular function.
Dopamine antagonists are commonly used as antipsychotic drugs.
It is now widely accepted for GPCRs of the amine cluster (i.e., dopa-
mine, serotonin, norepinephrine, or histamine receptors) that
antagonists mostly interact with two different hydrophobic areas
and an anionic aspartate in the receptor’s binding cavity.1 This
might explain why most of the antagonists have two aromatic moi-
eties to interact with the abovementioned two hydrophobic sites,
and a chargeable basic nitrogen which is crucial for the formation
of an ionogenic salt-bridge to the aspartate anion in the binding
pocket. It is not clarified yet, in what relative positions these three
pharmacophores should be, to gain maximum affinity for the
respective GPCRs. Changing the relative distances and angles of
these structural elements may give more insights into receptor li-
gand interactions.
quinolizines such as stepholidine and chloroscoulerine, exhibit
nanomolar affinities for the D1 and D2L receptors, and also the
rather constrained dihydrexidine, a dibenzoquinoline derivative,
is a highly potent D1 agonist.4–6 The scaffolds of the homologous
dibenzazecines only differ in the position of the second annulated
benzene moiety (Fig. 1). We therefore planned to synthesize dib-
enzazecines 7, 8, 10, 11 and their tetracyclic precursors 5, 6, and
9. To obtain meaningful information for SAR-studies on the chan-
ged scaffolds, dibenzo[d,g]-, -[c,f]- and -[c,g]azecines were synthe-
sized with the same substitution pattern, which is based on the
previously investigated potent dopamine antagonists 3-methoxy-
and 3-hydroxy-7-methyl-5,6,7,8,9,14-hexahydrodibenzo[d,g]aze-
cine (3, 4).2,3 The affinities of compounds 1–11 and the two
enantiomers of dibenzo[d,f]azecine 127 (Fig. 2) were measured
for all dopamine receptor subtypes by radioligand binding experi-
ments. The functionality of all compounds was determined by a
functional calcium assay.3
The rather flexible and symmetric dibenzo[d,g]azecines (e.g., 3,
4) are highly potent antagonists at the D1-receptor subtype family
(D1 and D5). For these azecines 3-hydroxy-derivatives show an
advantage over 3-methoxy-substituted compounds. In contrast,
the analogous, more rigid dibenzo[d,g]quinolizines (e.g., 1, 2) do
not show any affinity.2,3 So, obviously, the higher conformational
2. Results
2.1. Chemistry
The synthesis of compounds 7–11 is described in Schemes 1
and 2. In order to obtain the dibenzo[c,g]azecines 7, 8 and the
dibenzo[c,f]azecines 10, 11, we first synthesized the appropriate
* Corresponding author. Tel.: +49 3641 949820; fax: +49 3641 949802.
E-mail address: ch.enzensperger@uni-jena.de (C. Enzensperger).
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.08.028

M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
6899
O
HO
O
HO
HO
b
N
N
N
HO
R1
NH
23 %
5
6
known active compounds
R3
R2
a
96 %
chlorscoulerineR
=OH, R =MeO R =Cl
1
1
2
3
dihydrexidine
stepholidine
R
=MeO, R =OH R =H
2 3
HO
O
R
R
N+
N+
R
I-
Br-
b
N
N
N
43 %
13
14
R=MeO
2R=OH
9 R=MeO
1
R=MeO
R=OH
5
6
c
10 %
N
tetrahydro-
d,g
R
[
c,g ]quinolizine
[
]quinolizine
HO
O
isoindolobenzazepine
N
b
R
R
7
8
14 %
N
N
N
3
R=MeO
10R=MeO
R=OH
7
R=MeO
4 R=OH
8 R=OH
11
O
O
f
N
H2N
O
[
[
]azecine
O
]azecine
c,g
d,g
]azecine
[
c,f
O
O
39 %
15
Figure 1. The underlying tetracyclic compounds and resulting azecines.
d
O
O
O
O
O
O
a
N+
N
I-
N
N
78%
16
9
12b
12a
c
55 %
Figure 2. Dibenzo[d,f]azecines.
O
HO
e
tetracyclic precursor molecules 5 and 9 and converted them to
their quaternary salts 13 and 16 with methyl iodide in acetonitrile.
Birch conditions (Na0/NH3liq) were used to obtain 7 and 10. Cleavage
of the methyl-ether moieties of the quaternary salt 13, the
dibenzo[c,g]quinolizine 5 and the [c,g]azecine 7 was accomplished
by using hydrobromic acid in glacial acetic acid. Borontribromide
in chloroform was used to prepare the phenol 11 in excellent
yields, but for the ether cleavage of compounds 5 and 7 with
bortribromide, no product could be isolated. Compound 5 was
prepared as described previously starting from 3-methoxy-phenyl-
ethylamine and 3-isochromanone.8 Compound 9 was obtained by
lithium aluminum hydride reduction of the tetracyclic lactam
(15), which was synthesized in three steps starting from methyl-
2-formylbenzoate and 3-(3-methoxyphenyl)-propylamine via TiCl4
mediated acyliminium cyclization as described by Heaney and
Shuhaibar.9
Synthesis of the dibenzo[c,g]azecines 7 and 8 under Birch con-
ditions turned out to be more complicated than observed for the
preparation of dibenzo[d,g]azecines.2 The cleavage yielded a side
product in high yields due to the second benzylamine structure,
as outlined in Scheme 2. Thus, ring opening of the quaternary salt
14 yielded almost exclusively (>90%) 2-methyl-1-(2-methylben-
zyl)-1,2,3,4-tetrahydroisoquinolin-6-ol (17) instead of the desired
azecine 8 and cleavage of the methoxy derivative 13 gave a
mixture of 86% of 6-methoxy-2-methyl-1-(2-methylbenzyl)-
N
N
56 %
10
11
Scheme 1. Synthesis of the dibenzazecines 5–8. Reagents and conditions: (a)
methyl iodide, MeCN; (b) HBr, glacial acetic acid; (c) Na/NHl3iq; (d) LiAlH4, THF; (e)
BBr3, CHCl3; (f) (i) toluene reflux; (ii) Na, methanol; (iii) TiCl4, CH2Cl2.
1,2,3,4-tetra-hydroisoquinoline 19 and 14% of the desired azecine
7. Both compounds were separated by column chromatography.
In order to obtain better yields especially for the hydroxy deriva-
tive, we used a reaction sequence of Hofmann degradation fol-
lowed by catalytic hydrogenation, which is described for known
dibenzo[c,g]azecines.10 Unfortunately, starting with the quaternary
salt 13 we only obtained 3-(2-ethyl-4-methoxyphenyl)-2-methyl-
1,2,3,4-tetrahydro-isoquinoline (18). All by-products were charac-
terized by NMR and GC–MS and screened for their dopamine
receptor affinities.
2.2. Pharmacological investigations
Using radioligand displacement experiments, we screened all
dibenzazecines and ring-closed analogues, as well as the men-
tioned synthetic by-products, for their affinities for all human

6900
M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
HO
HO
The dibenzo[c,f]azecines are of special interest since their
N+
N+
N
N
N
N
scaffold can be seen as ring opened dihydrexidine derivative.
Dihydrexidine is described as high-affinity dopamine agonist.4
While the tetracyclic precursor 9 shows very low affinities for
all receptor subtypes, the 11-methoxy- and 11-hydroxy[c,f]aze-
cines (10, 11) show slight affinities for the D1 receptor with
micromolar Ki values. In contrast to dibenzo[d,g]azecines, but
similar to dibenzo[c,g]azecines, the hydroxy compound 11 shows
significantly lower affinies for D1 receptors than the methoxy
compound 10. Surprisingly, at D5 receptors, only the hydroxy
derivative has a weak affinity. Though dihydrexidine shows ago-
nistic activities at D1 receptors,4 the investigated dibenzo[c,g]-
azecines, which structurally resemble dihydrexidine, exhibit
weak antagonistic affinities.
Br-
b
53 %
14
43 %
17
18
19
a
b
O
O
c
I-
15 %
13
The dibenzo[d,f]azecines 12a, b exhibit low affinities for dopa-
mine receptors with micromolar Ki values, comparable to the bind-
O
O
ing affinities of the dibenzo[c,g]-homologs 7,
8 and the
dibenzo[c,f]azecines 10, 11. Interestingly, the (+)-enantiomer 12a
has slight affinities for D1 and D5 receptors, while the correspond-
ing (ꢀ)-enantiomer 12b exhibits no measureableable affinities.
This correlates to observations made for 12-chloroscoulerine,
where the two enantiomers show significant differences in their
affinities for dopamine receptors. However, the racemic 12-chlo-
roscoulerine displays almost the same activity as the active
enantiomer.12
10 %
60 %
+
7
Scheme 2. Synthetic by-products during cleavage of quaternary salt 11. Reagents
and conditions: (a) HBr, glacial acetic acid; (b) Na/NHl3iq; (c) (i) Ag2O, Methanol; (ii)
DMSO; (iii) H2/Pt2O.
In the functional calcium assay all active compounds proved to
be antagonists at D1 and D2L receptors. None of the tested com-
pounds show agonistic activities. Hence, a change in the scaffold
from dibenzo[d,g]- to other dibenzazecines seems to have no effect
on functionality at dopamine receptors, while the substitution pat-
tern, for example of dibenzo[c,g]quinolizines seems to have great
impact on functionalities at dopamine receptors.
cloned dopamine receptor subtypes D1–D5 and determined the
functionality for the D1 and D2L receptors in an intracellular cal-
cium assay. Ki values for the radioligand binding assay are given
in nM units (Table 1). A detailed protocol is described in Section 5.
3. Discussion
For the dibenzazecines 7 and 10 we could also obtain X-ray data
(Fig. 3). Both compounds turned out to be angled in the crystal.
Calculations of E-minimized conformations (DS-visualizer, MOE)
gave for the dibenzo[c,g]azecine 7 a rather straight shape, which
was also observed for E-minimized conformations of dibenzo[d,g]-
and -[c,g]quinolizines (1, 2, 5, 6). Measurement of distances be-
tween the aromatic centroids and the basic nitrogen in the crystal
structure, as well as in the simulated E-minimized conformations,
did not indicate a correlation between these distances and the ob-
served affinities for dopamine receptors. Further molecular model-
ing studies are in progress.
Both the constrained tetracyclic dibenzo[d,g]quinolizine deriva-
tives 1 and 2 and their ring opened tricyclic analogues 3 and 4 con-
tain a dopamine-like phenylethylamine moiety, but only the
tricyclic and more flexible azecines 3 and 4 display high affinities
for dopamine receptors. The 3-methoxy quinolizine derivative 1
shows slight affinity for the D1 receptor, but almost 100 times
weaker than the analogous azecine 3.
It is known from literature that the tetrahydroprotoberberines
12-chloroscoulerine and stepholidine, which are dibenzo[c,g]quin-
olizine derivatives, show high affinities for dopamine receptors.5,6
We now synthesized 3-methoxy and 3-hydroxy dibenzo[c,g]quin-
olizines (5, 6), which show good affinities as well. As already
recognized for the dibenzo[d,g]azecines,3 the hydroxy dibenzo-
quinolizine 6 exhibits a higher selectivity toward the D1/D5-recep-
tors and higher affinities at these subtypes compared to the
methoxy compound 5. The modification in substitution pattern
of stepholidine to compounds 5 and 6 lead to a change of function-
ality at D1 receptors. While stepholidine is an agonist at this recep-
tor subtype,5 our tested tetrahydroprotoberberines (5, 6) show
antagonistic activities. It was now expected that ring opening of
these potent tetracyclic compounds would increase affinities at
all receptor subtypes. But surprisingly, both the hydroxy- and
methoxy-substituted derivatives lost affinities nearly completely,
despite the increased flexibility of the ring scaffold. Only at D1
and D5 receptors micromolar affinities can be detected for the
methoxy compound (7). The hydroxy derivative does not show
any advantage over the methoxy compound. The synthetic by-
products 17–19 also show only weak micromolar or no affinities.
Here, 1-benzyl-tetrahydroisoquinolines (17, 19) turn out to be
advantageous over the 3-phenyl-tetrahydroisoquinoline derivative
(18), which shows very low affinities at all receptor subtypes. The
selectivity of compounds 17 and 19 toward D1 and D2 receptors is
an interesting observation and topic of our future investigations.
4. Conclusion
It can be summarized, that the annelation pattern in the tricy-
clic dibenzazecines and in the corresponding tetracyclic quinoli-
zine and quinolizine-like derivatives highly influences the
affinities, but not the functionalities at dopamine receptors. While
dibenzo[d,g]azecines show nanomolar affinities for all dopamine
receptors, dibenzo[c,f]-, -[c,g]- or -[d,f]azecines display micromolar
Ki values for D1 and D5 receptor subtypes. All active test com-
pounds turned out to act as antagonists.
Obviously, ring opening of dibenzo[d,g]quinolizines and the
consequent increase in conformational flexibility, increases recep-
tor affinities dramatically. The same effect occurs after ring open-
ing of isoindoloazepine 9 to the dibenzo[c,f]azezines 10 and 11,
even though less pronounced. In spite of the close resemblance
to the known dopamine receptor agonist dihydrexidine, these
more flexible compounds, which represent a new tricyclic ring sys-
tem, are antagonists with only moderate affinities.
The constrained dibenzo[c,g]quinolizines 5, 6 exhibit the high-
est receptor affinities among all investigated ring-closed com-
pounds. In contrast to the positive influence of ring opening on

M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
6901
Table 1
Affinities (Ki) for dopamine D1–D5 receptor subtypes were determined by radioligand binding experiments
Compounds
Affinity for D1, nM
Affinity for D2L, nM
Affinity for D3, nM
Affinity for D4, nM
Affinity for D5, nM
(Ki SEM)
(Ki SEM)
(Ki SEM)
(Ki SEM)
(Ki SEM)
O
N
N
N
N
2116 484b
>10,000
>10,000
>10,000
>10,000
43.4a
>10,000
>10,000
54 20a
1.5 0.5a
1
HO
>10,000
>10,000
>10,000
2
O
28.5 9.7a
13.0 9.0a
75.7 7.3a
3
HO
0.39 0.22a
17.5 1.5a
47.5 24a
11.3 1a
4
Stepholidinec
5.9
974
30 (Rat)
3748
4.4
O
N
183 27.5b
320 159a
251 56b
2565 1673a
185 97b
5
HO
N
136 42a
486 106b
306.5 61.5b
5092 1589b
41.1 24.3a
6
O
N
1727 238a
>10,000
>10,000
>10,000
1508 136b
7
HO
N
>10,000
>10,000
>10,000
>10,000
13 (Rat)
>10,000
16 (Rat)
8
Dihydrexidinec
35.7
2607
170 (Rat)
(continued on next page)

6902
M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
Table 1 (continued)
Compounds
Affinity for D1, nM
Affinity for D2L, nM
Affinity for D3, nM
Affinity for D4, nM
Affinity for D5, nM
(Ki SEM)
(Ki SEM)
(Ki SEM)
(Ki SEM)
(Ki SEM)
O
N
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
9
O
1264 165.5b
2070 76a
4748 500a
>10,000
>10,000
N
10
HO
N
1810 478a
1812 164b
>10,000
11
O
O
N
12a
O
O
N
12b
HO
N
3019 817b
2724 411b
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
17
O
N
>10,000
>10,000
18
O
N
2349 1414b
7371 2018b
>10,000
6644 2547b
>10,000
19
a
b
c
Ki values are means of at least three experiments, performed in triplicate SEM.
Ki values are means of two experiments, performed in triplicate SD.
Values from the PDSP database 11.

M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
6903
Figure 3. X-ray structures of dibenzo[c,g]azecine 7 and dibenzo[c,f]azecine 10.
receptor affinities recognized for the other dibenzazecines, ring
5.1.2. Synthesis of the quaternary salts (e.g., 13 and 16)—
general procedure
opening to the corresponding dibenzo[c,g]azecines 7, 8 abolishes
affinities. In this case, the dibenzo[c,g]quinolizines are more prone
to interact with the receptor than the more flexible dibenzo-
[c,g]azecines. Thus, higher flexibility and moderate increase in
residual mobility are not always beneficial for receptor interaction.
Contrary to dibenzo[d,g]azecines, whose hydroxy derivative 4
has distinctly higher affinities than the methoxy congener 3, the
methoxy derivatives of dibenzo[c,g]- and -[c,f]azecines (7 and 10)
unexpectedly display higher affinities. All investigated dibenzaze-
cines show an explicit selectivity toward the D1-receptor family.
Our findings might contribute to essential insights for further
molecular modeling studies on this topic.
A 10-fold molar excess of methyl iodide was added to a stirred
solution of the respective quinolizine (e.g., 5, 9) in acetonitrile. Un-
der nitrogen, the mixture was stirred for 48 h at room temperature.
The precipitated solids were isolated by filtration, washed with lit-
tle acetonitrile and dried in vacuo.
5.1.3. Ether cleavage of methoxylated compounds (5, 10, and
13)—general procedure
The methoxy compound was dissolved in a mixture of 20 mL of
glacial acetic acid and 10 mL aqueous HBr (48%) and refluxed un-
der nitrogen for 5 h. The solvents were removed in vacuo and the
residue was crystallized from methanol/diethyl ether. Hydroxy
compound 11 was obtained from the methoxy congener 10 by
cleavage with a fivefold molar excess of BBr3 in CHCl3.
5. Experimental
5.1. Synthesis and characterization of test compounds
5.1.4. 3-Methoxy-5,8,13,13a-tetrahydro-6H-isoquino[3,2-
a]isoquinoline (5)
Syntheses were performed under nitrogen with solvents and re-
agents of commercial availability with no further purification.
Melting points are uncorrected and were measured in open capil-
A solution of 3-methoxyphenethylamine (0.02 mol) and 3-iso-
chromanone (0.022 mol) in toluene (100 mL) was refluxed under
nitrogen for 6 h. After cooling down to room temperature, the
amide can be filtered off as white-yellow solid. The amide
(0.005 mol) was dissolved in a mixture of 40 mL acetonitrile and
5 mL POCl3. The solution was refluxed for 18 h. The solvents were
removed in vacuo and the resulting brown oil washed twice with
20 mL of petroleum ether. The residue was dissolved in 40 mL
methanol and cooled to 0 °C in an ice bath. NaBH4 (6 g) was added
in small portions to the cooled reaction mixture. The cooling was
removed and the suspension refluxed for 0.5 h. After evaporation
of the solvent, the residue was dissolved in water and extracted
with ethyl acetate. The organic layer was dried over MgSO4 and
evaporated in vacuo. The resulting yellow oil was crystallized from
methanol to yield 62% of a yellow powder. Mp: 224 °C (hydrochlo-
ride), 1H NMR: 250 MHz (CDCl3): d 2.59–3.42 (m, 6H); 3.59–3.86
(m, 2H); 3.82 (s, 3H, OMe); 4.03–4.08 (d, J = 14.95 Hz, 1H) 6.69–
6.70 (d, J = 2.5 Hz, 1H, 4); 6.80–6.84 (dd, J = 2.4, 8.6 Hz, 1H, 2);
7.09–7.21 (m, 5H, 1, 9, 10, 11, 12); 13C NMR: 250 MHz (DMSO-
d6): d 29.77; 36.81; 51.31; 55.23; 58.60; 59.49; 76.56; 77.07;
77.58; 112.50; 113.18; 125.82; 126.14; 126.26; 126.56; 128.74;
130.27; 134.45; 134.49; 135.83; 157.82 GC–MS: m/z 264 (100%);
250 (11%); 234 (2%); 218 (6%); 204 (2%); 191 (2%); 178 (2%); 167
(1%); 160 (82%); 147 (10%); 139 (1%); 130 (4%); 117 (13%); 104
(54%); 91 (9%); 78 (18%); 65 (4%). Anal. (C18H19NO): C, H, N.
lary tubes, using a Gallenkamp melting point apparatus. 1H and 13
C
NMR spectral data were obtained from a Bruker Advance 250 spec-
trometer (250 MHz) and Advance 400 spectrometer (400 MHz),
respectively. Elemental analyses were performed on a Hereaus
Vario EL apparatus for all test compounds. MS data were deter-
mined by GC–MS, using a Hewlett–Packard GCD-Plus (G1800C)
apparatus (HP-5MS column; J&W Scientific). EI-MS-Spectra were
recorded using LCQ Advantage by ThermoElectron and HRMS were
recorded by a TSQ Quantum AM spectrometer (Therma Electron
Corporation (Waltham, USA)).
5.1.1. Ring opening—general procedure
A 100-mL three-neck flask equipped with a balloon as an over-
flow tank was cooled in a liquid nitrogen bath. Ammonia was con-
densed into this flask until it was 3/4 filled. The cooling bath was
removed and the ammonia was allowed to liquefy. After suspend-
ing 1 mmol of the quaternary salts (13, 16) in the liquid ammonia,
rice-grain-sized pieces of sodium were added to the stirred mix-
ture until the developing blue color remained for 10–15 min. The
mixture was quenched with 1–2 drops of saturated aqueous NH4Cl.
The ammonia was evaporated under nitrogen. The residue was
portioned between ether and water and was stirred until two
phases formed. The aqueous phase was extracted with ether
(3 ꢁ 15 mL) and the pooled organic phases were dried over MgSO4
and evaporated to yield the ring open compound (e.g., 7, 10). In
case of dibenzo[c,g]azecines also tetrahydroisoquinoline deriva-
tives (17, 19) were obtained as by-products.
5.1.5. 5,8,13,13a-Tetrahydro-6H-isoquino[3,2-a]isoquinolin-3-
ol hydrobromide (6)
The synthesis was performed according to the general proce-
dures in the manuscript from 0.4 g of the methyl-ether 5. Recrystal-

6904
M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
lization from methanol/diethyl ether yielded 23% brown powder.
Mp: 245 °C, 1H NMR: 250 MHz (DMSO-d6): d 2.97–3.91 (m, 6H);
4.52–4.74 (m, 2H); 4.76–4.92 (m, 1H, 13a); 6.63–6.64 (d, J = 2,1 Hz,
1H, 4); 6.74–6.78 (dd, J = 1.9, 8.2 Hz, 1H, 2); 7.25–7.32 (m, 5H, 1, 9,
10, 11, 12); 10.52 (s, 1H, OH); 13C NMR, dept: 250 MHz (DMSO-d6):
d 23.97; 33.21; 50.45; 54.88; 59.56 (13a); 115.05; 115.19; 122.86;
126.51; 127.26; 127.35; 128.27; 128.84; 129.24; 132.15; 133.23;
157.22. Anal. (C17H18BrNO ꢁ 1/8 HBr): C, H, N.
up from the lithium aluminum hydride reduction was accom-
plished as free base. Either strongly alkaline or acidic conditions
should be avoided. This isoindole-derivative is very instable and at-
tempts to convert it into an HCl-salt failed due the immediate for-
mation of a dark green solution and no solids. The free base forms
yellow oil which solidifies to a brick red solid upon storage in the
refrigerator. Mp: 68–70 °C. Due to the instability, the pure com-
pound shows three decomposition compounds in the GC–MS.
The fragmentation of the amine is characteristic for quinolizine-
like compounds and has a mꢀ1 peak as the most stable fragment:
m/z: 265 (34%); 264 (100%); 248 (32%); 236 (9%); 220 (8%); 204
(5%); 191 (6%); 156 (6%); 144 (13%); 131 (14%). 1H NMR
250 MHz (CDCl3): d1.75–1.98 (m, 2H, 6); 2.74–3.40 (m, split in four
separate signals, each matching 1H, 4H, 5, and 6); 3.83 (s, 3H, O–
Me); 3.79–3.88 (d, J = 13 Hz, 1H, 8); 4.39–4.44 (d, J = 13 Hz, 1H,
8); 5.70 (s,1H, 13b); 6.62–6.67 (dd, J = 2.6, 8.5 Hz, 1H, 2); 6.77–
6.78 (d, J = 2.6, 1H, 4); 7.17–7.20 (d, J = 8.5 Hz, 1H, 1); 7.25–7.30
(m, 4H, 9, 10, 11, 12). Anal. (C18H19NO ꢁ H2O): C, H, N.
5.1.6. 10-Methoxy-6-methyl-5,6,7,8,13,14-hexahydrodibenzo[c,g]
azecine hydrochloride (7)
The synthesis was performed according to the general proce-
dures in the manuscript using 0.7 g of 13. Flash chromatography
on silica gel with methanol/chloroform (1:4) as eluent, gave a yel-
low oil. Addition of two drops of ethereal HCl and recrystallization
from 2-propanol/diethyl ether yielded 9.7% of white crystals. Mp:
214 °C, 1H NMR: 250 MHz (methanol-d4): d 2.80–2.94 (m, 1H, 4);
3.08–3.51 (m, 8H, 7, 8, 13, 14); 3.85–3.97 (m, 1H, 5); 4.42–4.48
(dd, J = 4.4, 10.7 Hz, 1H, 5); 5.95–5.99 (d, J = 8.4 Hz, 1H, 9); 6.35–
6.39 (dd, J = 2.5, 8.4 Hz, 1H, 11); 6.68–6.69 (d, J = 2.3 Hz, 1H, 12);
7.10–7.22 (m, 4H, 1, 2, 3, 4). HRMS m/z calcd for C19H24NO
282.1852 found 282.1850; EI-MS m/z: 281 (100%); 266 (12%);
250 (23%); 235 (13%); 223 (51%); 208 (17%); 192 (7%); 176
(56%); 165 (9%); 159 (7%); 146 (35%); 134 (23%); 115 (11%); 104
(18%); 91 (19%); 78 (10%); 65 (5%); 57 (9%); 44 (29%); 36 (17%);
GC–MS: (base) m/z: 281 (3%); 266 (3%); 250 (5%); 235 (3%); 219
(2%); 176 (10%); 162 (8%); 144 (30%); 130 (11%); 118 (20%); 104
(100%); 91 (40%); 78 (80%); 65 (20%). Anal. (C19H24ClNO ꢁ 1/9
HCl) C, N; for H calcd 7.55, found 6.92.
5.1.9. 11-Methoxy-6-methyl-5,6,7,8,9,14-hexahydrodibenzo[c,f]
azecine (10)
According the general procedure in the manuscript, 1.7 mmol
(700 mg) of the mossy green crude quaternary salt 16 were ring
opened with sodium in liq. ammonia. The amber colored oily res-
idue was dissolved in 1 mL of 2-propanol and converted into an
HCl-salt by the addition of three drops concd HCl acid. The green-
ish precipitation was recrystallized from acetone/2-propanol to
give 300 mg of white needles with mp 208–210 °C. Yield: 55% after
recrystallization. 1H NMR 250 MHz (methanol-d4): (HCl-salt) d
1.9–2.2 (m, 1H, 5, 6 or 7); 2.4–3.0 (m, 4H, 5, 6 or 7); 3.09 (s, 3H,
N-Me); 3.09–3.28 (m, 1H, 5, 6 or 7); 3.74 (s, 3H, O-Me); 3.91–
3.97 (d, J = 15 Hz, 1H, 14); 4.21–4.27 (d, J = 15 Hz, 1H, 14); 4.41–
4.46 (d, J = 13 Hz, 1H, 9); 4.76–4.80 (d, J = 13 Hz, 1H, 9); 6.72 (d,
J = 2.5 Hz, 1H, 4); 6.79–6.81 (dd, J = 2.5, 8.5 Hz 1H, 2); 7.29–7.51
(m, 4H, 10, 11, 12, 1); 7.51–7.77 (dd, J = 8.0/1.0 Hz, 1H, 13). GC–
MS: (base) m/z: 281 (37%); 266 (5%); 249 (10%); 235 (7%); 223
(19%); 209 (100%); 190 (84%); 177 (35%); 165 (32%); 152 (9%);
145 (21%); 115 (14%). Anal. (C19H23NO ꢁ HCl): C, H, N.
5.1.7. 6-Methyl-5,6,7,8,13,14-hexahydrodibenzo[c,g]azecin-10-
ol hydrochloride (8)
The synthesis was performed according to the general proce-
dures in the manuscript from 0.05 g of the methyl-ether 7. The
crude product was dissolved in a small amount of acetone. After
adding two drops of ethereal HCl and some drops of ether, brown
oil separates, which could be dried in vacuo to a brown powder.
Mp: 198 °C. Yield 14.1%. 1H NMR: 250 MHz (methanol-d4): d
2.78–2.91 (m, 1H,7); 2.99–3.09 (m, 7H, 13, 14, N–Me); 3.30–3.34
(m, 2H, 8); 3.50–3.60 (m, 1H, 7); 4.43–4.58 (q, J = 14.0 Hz, 2H, 5);
6.53–6.60 (m, 2H, 9, 11); 7.01–7.04 (d, J = 8.2 Hz, 1H, 12); 7.20–
7.26 (m, 3H, 1,2,3,); 7.42–7.49 (d, J = 8.1 Hz, 4) 13C NMR, dept:
250 MHz (methanol-d4): d 25.14; 31.74; 32.13; 42.82 (N–Me);
54.42; 55.66; 114.31; 115.46; 126.65; 127.74; 128.47; 128.66;
129.79; 130.55; 131.52; 132.29; 136.19; 142.02; 155.89. GC–MS:
(base) m/z: 267 (2%); 252 (2%); 236 (2%); 223 (1%); 210 (2%); 146
(14%); 132 (2%); 120 (20%); 104 (30%); 91 (28%); 78 (33%); 65
(15%); 52 (18%); 44 (100%). Anal. (C18H22ClNO ꢁ 4/5H2O ꢁ 1/
5HCl) C, H, N.
5.1.10. 6-Methyl-5,6,7,8,9,14-hexahydrodibenzo[c,f]azecin-11-
ol (11)
An aqueous solution of 317 mg 10 HCl (1 mmol) was alkalized
with some NaOH and extracted with CHCl3. After drying over
MgSO4, the solvent was evaporated to approximately 50 mL. With
an external cooling bath and under nitrogen atmosphere, 450 lL of
BBr3 (ꢂ5 mmol) were added through a septum. The cooling bath
was removed and the reaction mixture was brought to reflux for
1 h. When the mixture returned to room temperature, it was
poured onto chipped ice and the pH was brought to 9 with 2 N
NaOH. Extraction with CH2Cl2, drying over MgSO4 and evaporation
yielded 151 mg of white foam. Yield 56%. Mp: 139–140 °C, 1H
NMR: 250 MHz (CDCl3): (base): d 1.48–1.51 (mc, 4H, 6, 10);
2.03–2.07 (mc, 4H, 7, 9); 2.17 (s, 3H, N–Me); 2.65–2.75 (mc, 4H,
5, 11); 4.06, (s, 2H, 16); 6.47–6.52 (dd, J = 2.6, 8.2, 1H, 2); 6.61 (d,
J = 2.6, 1H, 4); 6.86–6.90 (d, J = 8.2, 1H, 1); 7.01–7.14 (m, 4H, 12,
13, 14, 15). GC–MS: (base) m/z: 267 (3%); 252 (2%); 236 (5%);
221 (3%); 209 (12%); 195 (39%); 176 (18%); 165 (45%); 152
(30%); 146 (30%); 132 (35%); 120 (46%); 115 (52%); 107 (48%);
104 (72%); 102 (71%); 91 (100%); 70 (100%); 57 (92%). Anal.
(C18H21NO ꢁ 1/5 H2O): C, H, N.
5.1.8. 3-Methoxy-6,7,9,13b-tetrahydro-5H-isoindolo[1,2-
a][2]benzazepine (9)
To a suspension of 5.0 g (0.13 mol) lithium aluminum hydride
in 150 mL of dry THF there was added slowly under nitrogen and
with cooling a solution of 5.0 g (0.18 mol) of the lactam 13 in
50 mL of dry THF. The mixture was refluxed for 1 h and stirred
overnight. With cooling, the residual lithium aluminum hydride
was decomposed by the careful addition of a saturated aqueous
solution of potassium-sodium tartrate. After filtration of inorganic
solids and washing of the filter cake with 50 mL of THF, the layers
were separated and the aqueous layer was extracted with dichloro-
methane. After evaporation of the pooled organic extracts, the res-
idue was dissolved in 50 mL of dichloromethane and dried over
Na2SO4. Evaporation of the solvent yielded 2.2 g of 9 as yellow
oil that slowly solidified after storage in the refrigerator. Work
5.1.11. 3-Methoxy-7-methyl-5,8,13,13a-tetrahydro-6H-
isoquino[3,2-a]isoquinolinium iodide (13)
The synthesis was performed according to the general proce-
dures in the manuscript using 2 g of 5. The precipitate was filtered
and washed with little ether to yield 96% of white crystals. Mp:

M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
6905
199 °C; 1H NMR: 250 MHz (DMSO-d6): d 2.82 (s, 3H, N–Me); 2.98–
3.45 (m, 3H); 3.78 (s, 3H, MeO) 3.82–4.16 (m, 3H); 4.76–4.98 (dd,
J = 15.5, 39.75 Hz, 2H, 10); 5.15–5.22 (dd, J = 4.4, 12.5 Hz, 1H, 13a);
6.87–6.99 (m, 2H); 7.28–7.45 (m, 5H); 13C NMR: 250 MHz (DMSO-
d6): d 23.78; 28.96; 33.81; 55.76; 60.57; 64.70; 65.47; 113.77;
114.35; 122.58; 126.88; 127.25; 127.76; 128.51; 128.77; 130.16;
131.02; 132.39, 159.66. Anal. (C19H22INO): C, H, N.
vents, crude 16 was obtained as dark green oil. Attempts to induce
crystallization from methanol/diethyl ether failed but the ether
insoluble oil popped up under reduced pressure to mossy green
foam which was successfully used for the ring opening procedure
without further purification. Yield 72%.
5.1.15. 2-Methyl-1-(2-methylbenzyl)-1,2,3,4-tetrahydroisoquino
lin-6-ol hydrochloride (17)
5.1.12. 3-Hydroxy-7-methyl-5,8,13,13a-tetrahydro-6H-isoquino
[3,2-a]isoquinolinium bromide (14)
The synthesis was performed according to the general proce-
dures in the manuscript using 0.5 g of 14. Removing of ether in va-
cuo yielded white foam. Addition of two drops of etheric HCl and
recrystallization from 2-propanol/diethyl ether yielded 53% of
white crystals. Mp: 224 °C; 1H NMR: 250 MHz (methanol-d4): d
1.95 (s, 3H, 20-Me); 2.98 (s, 3H, N–Me); 3.08–3.51 (m, 5H); 3.85–
3.97 (m, 1H); 4.42–4.48 (dd, J = 4.4, 10.7 Hz, 1H); 5.95–5.99 (d,
J = 8.4 Hz, 1H, 8); 6.35–6.39 (dd, J = 2.5, 8.4 Hz, 1H, 7); 6.68–6.69
(d, J = 2.3 Hz, 1H, 5); 7.13–7.20 (m, 4H, 30,40,50, 60); 13C NMR, dept:
250 MHz (methanol-d4): d 17.94 (20-Me); 22.29; 36.13; 39.40;
44.90 (N–Me); 64.21; 113.36; 114.61; 120.04; 125.89; 127.15;
129.40; 130.17; 130.51; 130.88; 133.43; 137.21; 157.56. EI-MS
m/z: 162 (100%); 147 (10%); 117 (2%); 105 (9%); 77 (5%); 65
(2%); 42 (11%); 36 (8%). Anal. (C18H22ClNO ꢁ 1/11 HCl) C, H, N.
The synthesis was performed according to the general proce-
dures in the manuscript from 1 g of 13. Recrystallization from meth-
anol yielded 43% of a brown powder. Mp: 274 °C; 1H NMR: 250 MHz
(DMSO-d6): d 2.80 (s, 3H, N–Me); 2.96–3.41 (m, 2H); 3.84–4.07 (m,
4H); 4.70–4.49 (dd, J = 15.5, 39.75 Hz, 2H, 10); 5.08–5.14 (dd,
J = 4.4, 12.5 Hz, 1H, 13a); 6.69 (d, J = 1.8 Hz, 1H, 4); 6.76–6.80 (dd,
J = 1.9, 10.7 Hz, 1H, 2); 7.26–7.38 (m, 5H, 1.9, 10, 11, 12); 9.71 (s,
1H, OH); 13C NMR, dept: 250 MHz (DMSO-d6): d 23.61; 28.99;
38.79 (N-Me); 60.62; 64.64; 65.50 (13a); 115.33; 115.44; 120.88;
127.18; 127.24; 127.74; 127.75; 128.66; 129.48; 130.90; 132.15;
157.72. Anal. (C18H20BrNO ꢁ 1/11 HBr): C, H, N.
5.1.13. 3-Methoxy-5,6,7,13b-tetrahydro-9H-isoindolo[1,2-a][2]
benzazepin-9-one (15)
5.1.16. 3-(2-Ethyl-4-methoxyphenyl)-2-methyl-1,2,3,4-
According to the general procedure, described by Heaney and
tetrahydro-isoquinoline hydrochloride (18)
Shuhaibar9,
a
mixture of 12.5 g (75.5 mmol) methyl-2-form-
To a solution of 0.48 g (1.18 mmol) quaternary salt 11 in 20 mL
methanol, 1 g of freshly prepared silver oxide was added and the
mixture refluxed for 20 min. The suspension was cooled to room
temperature, filtered and the solution treated with charcoal for
10 min at room temperature. After filtration the solvent was re-
moved in vacuo. The residue was dissolved in 5 mL of DMSO and
allowed to stand at room temperature for 15 min. Ice water
(2.5 mL) was added, and the solution extracted with benzene.
The organic layer was dried over potassium carbonate, passed
through a column of alumina, and eluted with 100 mL of benzene
to yield 208 mg of yellow oil, which was crystallized from acetone
to obtain 125 mg of white crystals. The solid was dissolved in
20 mL of dioxane and stirred with platinum oxide in hydrogen
atmosphere (balloon) at room temperature for 5 h. After filtration,
the solvent was removed in vacuo. The resulting yellow oil crystal-
lized after addition of two drops etheric HCl from 2-propanol/
diethyl ether. Yield 15%. Mp: 229 °C; 1H NMR: 250 MHz (metha-
nol-d4): d 1.21–1.27 (t, J = 7.5 Hz, 3H, Ph–CH2–CH3); 2.70 (s, 3H,
NMe); 2.61–2.93 (m, 2H, Ph–CH2–CH3); 3.24–3.65 (m, 2H); 3.84
(s, 3H, MeO); 4.65–4.92 (m, 3H); 6.74–6.99 (m, 2H); 7.28–7.33
(m, 4H); 7.51–7.59 (d; J = 8.5 Hz, 1H); 13C NMR, dept: 250 MHz
(methanol-d4): d 14.99 (Ph–CH2–CH3), 25.61; 35.03; 39.23 (N–
Me); 54.47 (MeO); 56.52; 61.29 (3); 112.76; 114.78; 124.10;
125.91; 127.00; 127.86; 127.90; 127.96; 128.20; 131.69; 145.29;
160.65. GC–MS: (base) m/z: 281 (5%); 266 (3%); 250 (8%); 235
(5%); 219 (4%); 204 (2%); 191 (2%); 176 (20%); 162 (16%); 146
(58%); 132 (18%); 118 (23%); 104 (100%); 91 (44%); 78 (68%); 65
(19%). Anal. (C19H24ClNO) C, H, N.
ylbenzoate and 12.35 g (75.5 mmol) of 3-(3-methoxyphenyl)-pro-
pylamine in 80 mL of toluene was refluxed for 4 h under
Dean–Stark conditions. After evaporation of the solvent and recrys-
tallization from ethyl acetate, 14 g of the imine were obtained as
off-white powder with >99% GC–MS purity. Yield: 60%. GC–MS:
311 (1%); 296 (6%); 177 (100%); 145 (62%); 132 (25%); 121
(20%); 91 (16%). For the rearrangement, a pea-sized piece of metal-
lic sodium was added to a solution of 1.5 g (4.8 mmol) of the imine
in 50 mL of methanol and the mixture was stirred under an inert
atmosphere for 2 days. After removal of the solvent, the residue
was dissolved in dichloromethane and washed with water to yield
after drying over Na2SO4 and evaporation of the solvent 1.2 g of a
methoxy-isoindolone as oil. (Yield: 80%.) 1H NMR: (CDCl3) d
1.90–1.99 (m, 2H, 9); 2.57–2.63 (t, J = 8.6 Hz, 2H, 10); 2.79 (s, 3H,
1-OMe); 3.18–3.29 (m, 1H, 8); 3.70 (s, 3H, 12-OMe); 3.73–3.82
(m, 1H, 8); 5.78 (s, 1H, 1); 6.62–6.73 (m, 3H, 11, 13, 15); 7.07–
7.11 (t, J = 7.8 Hz, 1H, 14); 7.44–7.48 (m, 3H, 3, 4, 5); 7.73–7.77
(m, 1H, 2). GC–MS: 311 (38%); 297 (18%); 296 (88%); 188 (10%);
147 (100%); 133 (21%); 121 (32%); 91 (32%). For the acyliminium
cyclization, 3.0 g (9.6 mmol) of the methoxy-isoindolone were dis-
solved in 50 mL of dichloromethane. The solution was cooled to
ꢀ78 °C in a dry-ice/methanol bath under an inert atmosphere
and 1.2 mL (9.6 mmol) of TiCl4 were added through a septum.
The solution was stirred in the ice bath and allowed to return to
room temperature overnight. The reaction mixture was washed
with satd solution of sodium hydrogen carbonate, the phases were
separated, the organic layer dried over Na2SO4 and evaporated to
leave the lactam as an oil which was purified by column chroma-
tography (CHCl3/MeOH) 9:1. Yield: 83%. 1H NMR: 400 MHz (CDCl3):
d 1.95–2.25 (mc, 2H, 6); 2.70–2.80 (mc, 2H, 5); 3.32–3.40 (mc, 2H,
7); 3.80 (s, 3H, OMe); 4.34–4.39 (mc, 2H, 7); 5.71 (s, 1H, 13b);
8.60 (s, 1H, 4); 6.73–6.76 (d, 1H, J = 7.5, 2) 7.17–7.20 (d, J = 7.5, 1H,
1); 7.46–7.58 (m, 3H, 10, 11, 12); 7.90–7.92 (d, J = 7.5, 13). GC–MS
m/z: 279 (100%); 279 (87%); 264 (4%); 250 (53%); 220 (30%); 178
(11%); 165 (10%); 147 (14%) 130 (11%); 102 (6%).
5.1.17. 6-Methoxy-2-methyl-1-(2-methylbenzyl)-1,2,3,4-
tetrahydroisoquinoline hydrochloride (19)
The synthesis was performed according to the general proce-
dures in the manuscript using 0.7 g of 13. Silica gel column with
methanol/chloroform (1:4) as eluent yielded a yellow oil. Addition
of two drops of etheric HCl and crystallization from 2-propanol/
diethyl ether yielded 60% of white crystals. Mp: 192 °C; 1H NMR:
250 MHz (methanol-d4): d 1.95 (s, 3H, 20-Me); 3.00 (s, 3H, N–Me);
3.10–3.51 (m, 5H); 3.76 (s, 3H, MeO); 3.89–4.00 (m, 1H); 4.48–
4.54 (dd, J = 4.3, 10.7 Hz, 1H, 1); 6.06–6.10 (d, J = 8.6 Hz, 1H, 8);
6.49–6.54 (dd, J = 2.4, 8.6 Hz, 1H, 7); 6.84–6.85 (d, J = 2.5 Hz, 1H,
5); 7.10–7.19 (m, 4H, 30,40,50, 60); 13C NMR, dept: 250 MHz (metha-
5.1.14. 3-Methoxy-8-methyl-6,7,9,13b-tetrahydro-5H-
isoindolo[1,2-a][2]benzazepinium iodide (16)
The solution of 2 g of compound 9 in acetonitrile and an excess
of methyl iodide turns deeply green. After evaporation of the sol-

6906
M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
nol-d4): d 19.38 (20-Me); 22.01; 38.42; 40.10; 44.32 (N–Me); 55.28
(MeO); 64.32; 112.66; 113.54; 120.68; 126.15; 127.45; 129.97;
130.11; 130.48; 131.14; 133.43; 137.14; 159.70. GC–MS m/z: 176
(100%); 161 (23%); 144 (6%); 132 (54%); 117 (9%); 105 (91%); 91
(11%); 77 (70%); 65 (17%). Anal. (C19H24ClNO) C, H, N.
trypsine–EDTA-solution (Sigma–Aldrich) to remove the cells from
the culture dish. After incubation, cells were suspended in 6 mL
medium in order to stop the effect of trypsine–EDTA-solution.
The resulting suspension was centrifuged (483 RCF, 4 °C, 4 min).
The pellet was re-suspended in 20 mL of ice-cooled PBS (phos-
phate-buffered saline: 137 mM NaCl, 2.7 mM, 8.1 mM Na2HPO4
and 1.4 mM KH2PO4) and again centrifuged to obtain a pellet. This
procedure was repeated once before the resulting pellet was sus-
pended in 12 mL of buffer (5 mM magnesium chloride, 50 mM
TRIS–HCl, pH 7.4). This resulting suspension was directly used for
the radioligand binding assay.
5.2. X-ray analyses of compounds 7 and 10
The intensity data for the compounds were collected on a Non-
ius KappaCCD diffractometer using graphite-monochromated Mo
Ka radiation. Data were corrected for Lorentz and polarization ef-
fects but not for absorption effects.13,14
The structures were solved by direct methods (SHELXS15) and re-
5.3.1.3. Radioligand binding assay. The binding studies were per-
formed following the protocol previously described but in 96-well
format.3 The assays with the whole-cell-suspension were carried
fined by full-matrix least squares techniques against Fo2
(SHELXL-
9716). For the whole compound 7 and for the amine-groups of 10
the hydrogen atoms were located by difference Fourier synthesis
and refined isotropically. The other hydrogen atom positions were
included at calculated positions with fixed thermal parameters.
Compound 10 crystallized as a racemic twin. All non-hydrogen
atoms were refined anisotropically.16 XP (SIEMENS Analytical X-
ray Instruments, Inc.) was used for structure representations.
out in triplicate in a volume of 550
TRIS–Mg2+-buffer (345 L), [3H]-ligand (50
sion (100 L) and appropriate drugs (55
l
L (final concentration):
L), whole-cell-suspen-
lL). As radioligands, for
l
l
l
the D1 family (D1 and D5) [3H] SCH23390, and for the D2 family
(D2, D3, D4) [3H] spiperone were used. Non-specific binding was
determined using fluphenazine (100
lM) for D1 and D5 tests and
haloperidol (10 M) for D2, D3, and D4 tests. The incubation was
l
5.2.1. Crystal data for 7
initiated by addition of the radioligand and was carried out in 96
deep well plates (Greiner bio-one, Frickenhausen) using a thermo-
cycler (Thermocycler comfortÒ, Eppendorf, Wessling) at 27 °C. The
incubation was terminated after 90 min by rapid filtration with a
C19H24ClNO, Mr = 317.84 g molꢀ1
,
colorless prism, size
0.04 ꢁ 0.04 ꢁ 0.04 mm3, monoclinic, space group P21/c, a =
14.4129(5), b = 11.1844(6), c = 11.4434(6) Å, b = 108.873(2)°,
V = 1745.50(14) Å3, T = ꢀ90 °C, Z = 4,
q
calcd = 1.209 g cmꢀ3
,
l
(Mo
Perkin–Elmer Mach III Harvester using a Perkin–Elmer Filtermat
TM
K
a
) = 2.21 cmꢀ1, F(0 0 0) = 680, 12,180 reflections in h(ꢀ18/18),
A, previously treated with a 0.25% polyethyleneimine-solution
(Sigma–Aldrich) and washed with water. The filtermat was dried
for 3 min at 400 W using a microwave oven (MW 21, Clatronic,
Kempen). The dry filtermat was placed in a filter plate (Omni filter
k(ꢀ14/12), l(ꢀ14/14), measured in the range 2.62° 6
H 6 27.46°,
completeness Hmax = 99.7%, 3987 independent reflections,
Rint = 0.0569, 2797 reflections with Fo > 4 (Fo), 295 parameters, 0
restraints, R1obs = 0.0433, wR2obs = 0.0886, R1all = 0.0775, wR2all
r
TM
=
plates , Perkin–Elmer Life Sciences) and each field of the filtermat
TM
0.1024, GOOF = 1.010, largest difference peak and hole: 0.186/
moistened with 50 lL Microscint 20 scintillation cocktail. The
ꢀ0.229 e Åꢀ3
.
radioactivity retained on the filters was counted using a Top Count
TM
NXT microplate scintillation counter (Packard, CT, USA). For deter-
5.2.2. Crystal data for 10
mining the Ki values at least two independent experiments each in
triplicate were performed.
C19H25ClNO1.50
,
Mr = 326.85 g molꢀ1
,
colorless prism, size
0.04 ꢁ 0.04 ꢁ 0.04 mm3, triclinic, space group P1, a = 7.1171(3),
The competition binding data were analyzed with GraphPad
TM
b = 14.7356(6), c = 17.1663(8) Å,
a
= 103.813(2)°, b = 95.077(3)°,
Prism software using nonlinear regression with sigmoidal dose re-
c
= 90.005(3)°, V = 1740.97(13) Å3, T = ꢀ90 °C, Z = 4, qcalcd
=
sponse equation. For calculating of mean, standard deviation and
1.247 g cmꢀ3
,
l
(Mo K
a
) = 2.25 cmꢀ1, F(0 0 0) = 700, 12,481 reflec-
standard error of the mean the software Microsoft Excel was used.
TM
tions in h(ꢀ9/8), k(ꢀ19/19), l(ꢀ21/22), measured in the range
Ki values were calculated from IC50 values applying the equation of
2.09° 6 6 27.50°, completeness Hmax = 98%, 12,481 independent
H
Cheng and Prusoff.18
reflections, Rint = 0.0000, 9649 reflections with Fo > 4r(Fo), 826
parameters, three restraints, R1obs = 0.0596, wR2 = 0.1270, R1all
=
5.3.2. Functional assay: measuring intracellular Ca2+ with a
fluorescence microplate reader3,19
5.3.2.1. Cell culture. Human D1 and D2L receptors were stably
expressed in human embryonic kidney cells (HEK293) and
cultured as mentioned above.
obs
0.0905, wR2all = 0.1453, GOOF = 1.039, Flack-parameter 0.47(6) (race-
mic twin), largest difference peak and hole: 0.395/ꢀ0.438 e Åꢀ3
.
5.3. Pharmacological assays
5.3.1. Dopamine receptor affinity
5.3.2.2. Preparation of whole-cell-suspension. Human D1 and
D2L receptor cell lines were grown on T 175 culture dishes (Greiner
bio-one, Frickenhausen) to 85–90% confluence. The medium was
removed via a suction apparatus and cells rinsed twice with 6 mL
Krebs-HEPES buffer (118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4,
5.3.1.1. Cell culture. Human D1, D2L, D3, D4.4, and D5 receptors
were stably expressed in Chinese hamster ovary (CHO) cells or hu-
man embryonic kidney cells (HEK293). D1, D2L, D3, and D5 were ex-
pressed in HEK cells and D4.4 receptors were expressed in CHO
cells, respectively. Cells were grown at 37 °C under an atmosphere
of 5% CO2: 95% air in HAM/F12-medium (Sigma–Aldrich) for CHO
cells and Dulbecco’s modified Eagles Medium Nutrient mixture F-
12 Ham for HEK293 cells, each supplemented with 10% fetal bo-
1.2 mM KH2PO4, 4.2 mM NaHCO3, 11.7 mM
D-glucose, 1.3 mM
CaCl2, 10 mM HEPES, pH 7.4) each time. After these two washes,
TM
cells were loaded with 3 lL of a 0.5 M Oregon Green 488 BAP-
TA-1/AM-solution (Molecular Probes, Eugene, OR) (in DMSO) in
vine serum, 1 mM
Sigma–Aldrich).
L
-glutamine and 0.2
l
g/mL of G 418 (all from
6 mL of Krebs-HEPES buffer containing 3 L of a 20% Pluronic F-
l
127-solution (Sigma–Aldrich) (in DMSO) for 45 min at 37 °C. After
35 min incubation, the culture dish was rapped slightly in order to
remove all cells from the dish for further incubation. To this cell
suspension 5 mL of Krebs-HEPES buffer were added again to rinse
all cells from the plate. Then the resulting suspension portioned in
10 1.5 mL Eppendorf caps and centrifuged at 10,640 RCF for 10 s.
5.3.1.2. Preparation of whole-cell-suspension17. Human D1, D2L
D3, D4, and D5 receptor cell lines were grown on T 175 culture
dishes (Greiner bio-one, Frickenhausen) to 85% confluence. The
medium was removed and the cells were incubated with 3 mL
,

M. Schulze et al. / Bioorg. Med. Chem. 17 (2009) 6898–6907
6907
The supernatant buffer was removed and the resulting 10 pellets
are divided in two portions of five pellets, which are suspended,
in 1 mL of Krebs-HEPES buffer, each. The two suspensions are cen-
trifuged again for 10 s. After removing the buffer, the two pellets
were combined and re-suspended in 1 mL buffer, diluted with
17 mL of Krebs-HEPES buffer and plated into 96-well plates (Opti-
plementary publication CCDC 730823 for 7, and CCDC 730824 for
10. Copies of the data can be obtained free of charge on application
to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [E-mail:
deposit@ccdc.cam.ac.uk].
Acknowledgments
TM
Plate HTRF-96 , Packard, Meriden, CT; Cellstar, Tissue Culture
Plate, 96 W, Greiner bio-one, Frickenhausen). Microplates were
kept at 37 °C under an atmosphere including 5% CO2 for 30 min be-
fore they were used for the assay.
The authors would like to thank Achim Meyer, who supported
us with the synthesis of the dibenzo[c,g]quinolizines, as well as
Bärbel Schmalwasser, Petra Wiecha and Heidi Traber for the reli-
able technical assistance in performing the pharmacological as-
says. Furthermore we would like to thank the DFG for financial
support of the project (EN 875/1-1).
5.3.2.3. Calcium assay3. Screening for agonistic and antagonistic
TM
activity was performed using a NOVOstar microplate reader (BMG
LabTechnologies) with a pipettor system. Agonistic activities were
tested by injecting 20 lL buffer alone as negative control, standard
References and notes
agonist in buffer as positive control, and test compounds in buffer
in rising concentrations, respectively, each into separate wells. Fluo-
rescence measurement started simultaneously to the automatic
injection. SKF 38393 was used as standard agonist for D1 receptors
1. Surgand, J. S.; Rodrigo, J.; Kellenberger, E.; Rognan, D. Proteins 2006, 62, 509.
2. Mohr, P.; Decker, M.; Enzensperger, C.; Lehmann, J. J. Med. Chem. 2006, 49,
2110.
3. Hoefgen, B.; Decker, M.; Mohr, P.; Schramm, A. M.; Rostom, S. A.; El Subbagh,
H.; Schweikert, P. M.; Rudolf, D. R.; Kassack, M. U.; Lehmann, J. J. Med. Chem.
2006, 49, 760.
4. Lovenberg, T. W.; Brewster, W. K.; Mottola, D. M.; Lee, R. C.; Riggs, R. M.;
Nichols, D. E.; Lewis, M. H.; Mailman, R. B. Eur. J. Pharmacol. 1989, 166,
111.
5. Li, J. F.; Jin, G. Z.; Shen, J. S.; Shanghai, R. A. Drugs Future 2006, 31, 379.
6. Fu, W.; Shen, J.; Luo, X.; Zhu, W.; Cheng, J.; Yu, K.; Briggs, J. M.; Jin, G.; Chen, K.;
Jiang, H. Biophys. J. 2007, 93, 1431.
and quinpirole for D2 receptors (final concentration: 1
Screening for antagonistic activities was performed by pre-incu-
bating the cells with 20 L of the test compound dilutions (final con-
centrations: 100 M, 50 M, 10 M, 5 M, 1 M, 500 nM, 100 nM,
50 nM, 10 nM, 1 nM, 0,1 nM) at 37 °C 30 min prior to injection of
20 L standard agonist per well. As standard agonists we used, as de-
lM).
l
l
l
l
l
l
l
scribed for the agonist screening, SKF38393 for D1 and quinpirole for
D2 receptors, respectively. Fluorescence measurement also started
simultaneously to the automatic injection. At least two independent
experiments each in four or six replications were performed.
Fluorescence intensity was measured at 520 nm (bandwidth
25 nm) for 30 s at 0.4 s intervals. Excitation wavelength was
485 nm (bandwidth 20 nm). Agonistic or antagonistic activities
were assessed by a dose response curve obtained by determination
of the maximum fluorescence intensity of each data set and non-
linear regression with sigmoidal dose response equation using
7. Furneaux, R. H.; Gainsford, G. J.; Mason, J. M. J. Org. Chem. 2004, 69, 7665.
8. Meise, W.; Zymalkowski, F. Arch. Pharm. Ber. Dtsch. Pharm. Ges. 1971, 304,
182.
9. Heaney, H.; Shuhaibar, K. F. Synlett 1995, 47.
10. Sawa, Y.; Hirose, K.; Maeda, S.; Hamada, Y. U.S. Patent 3,932,384, 1976.
11. NIMH Psychoactive Drug Screening Program (PDSP). 2006. Internet
Communication.
12. Chen, L. J.; Zhou, Q. T.; Dong, Z. J.; Yu, L. P.; Jin, G. Z. Acta Pharmacologica Sinica
1999, 20, 884.
13. COLLECT, Data Collection Software; Nonius B.V., Netherlands, 1998.
14. Otwinowski, Z.; Minor, W. In Methods in Enzymology Macromolecular
Crystallography; Carter, C. W., Sweet, R. M., Eds.; Part A; Academic Press,
1997; Vol. 276, p 307.
15. Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, 46, 467.
16. Sheldrick, G. M. SHELXL-97 (Release 97-2), University of Göttingen, Germany,
19.
TM
GraphPadPrism 3.0.
6. Supplementary data
17. Decker, M.; Lehmann, J. Arch. Pharm. (Weinheim) 2003, 336, 466.
18. Cheng, Y.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099.
19. Kassack, M. U.; Hofgen, B.; Lehmann, J.; Eckstein, N.; Quillan, J. M.; Sadee, W. J.
Biomol. Screen. 2002, 7, 233.
Crystallographic data (excluding structure factors) has been
deposited with the Cambridge Crystallographic Data Centre as sup-