11-Substituted (R)-aporphines: Synthesis, pharmacology, and modeling of D(2A) and 5-HT(1A) receptor interactions

Article abstract of DOI:10.1021/jm960189i

A series of C11-substituted (R)-aporphines and C11-oxygenated (R)- noraporphines has been synthesized and evaluated for central serotonergic and dopaminergic effects in vitro and in vive. The various C11-substituents were introduced using efficient nickel

Full text of DOI:10.1021/jm960189i

J . Med. Chem. 1996, 39, 3503-3513
3503
11-Su bstitu ted (R)-Ap or p h in es: Syn th esis, P h a r m a cology, a n d Mod elin g of D2A
a n d 5-HT1A Recep tor In ter a ction s
Martin H. Hedberg,† Tero Linnanen,† J ohanna M. J ansen,† Gunnar Nordvall,‡ Stephan Hjorth,§
Lena Unelius,‡ and Anette M. J ohansson*,†
Organic Pharmaceutical Chemistry, Uppsala University, Uppsala Biomedical Centre, Box 574, S-751 23 Uppsala, Sweden,
Preclinical R&D, Astra Arcus AB, S-151 85 So¨derta¨lje, Sweden, and Department of Pharmacology, University of Go¨teborg,
Medicinaregatan 7, S-413 90 Go¨teborg, Sweden
Received March 8, 1996X
A series of C11-substituted (R)-aporphines and C11-oxygenated (R)-noraporphines has been
synthesized and evaluated for central serotonergic and dopaminergic effects in vitro and in
vivo. The various C11-substituents were introduced using efficient nickel- and palladium-
catalyzed reactions of the corresponding triflate (R)-11-[[(trifluoromethyl)sulfonyl]oxy]aporphine
(6). Several compounds display high affinity to serotonin 5-HT1A receptors in spite of major
differences in steric bulk and electronic properties of the various C11-substituents. A change
of the N-methyl group of the nonselective 3 to H [23, (R)-11-hydroxynoraporphine] or propyl
[2, (R)-11-hydroxy-N-propylnoraporphine] increases the selectivity for 5-HT1A receptors (100-
fold) and dopamine D2A receptors (3-fold), respectively. Compounds 3 and 23 have similar
affinities to 5-HT1A receptors, whereas the propyl substituent of 2 not only enhances the
selectivity for D2A receptors but also increases the D2A affinity. Modeling of ligand-receptor
binding site interactions yielded an interaction site model for the 5-HT1A receptor that describes
a gradual change in binding mode for C11-hydroxy, -methoxy-, and -phenyl-substituted
derivatives. Hydrogen bonding is hereby gradually replaced by van der Waals interactions
involving a relatively large lipophilic pocket. The derived D2A receptor model can accommodate
both the N-propyl substituent of 2 and the C11-ethyl substituent of 11 [(R)-11-ethylaporphine].
In tr od u ction
Many aporphines possess the ability to activate
dopamine (DA) receptors. Several structural modifica-
tions of the nonselective DA receptor agonist (R)-
apomorphine (1) have been performed in order to
investigate structure-activity relationships (SAR) pri-
marily at DA D1 and D2 receptors.1 These studies have
shown that the catechol function of 1 is not a prereq-
uisite for a potent interaction of aporphines with DA
receptors. (R)-11-Hydroxy-N-propylnoraporphine (2)
has been reported by Neumeyer and co-workers1c to be
equipotent to 1 as a DA receptor agonist. In more recent
studies, however, compound 2 appears to be more potent
than 1, reportedly having higher affinity to and efficacy
at D2 receptors.2 The racemic N-methyl-substituted
analogue (()-3 has been reported to be a DA receptor
agonist.1a However, the pure R-enantiomer (3) was
found to be an antagonist at D1 receptors while also
displaying moderate affinity to D2 receptors.1e In ad-
dition, 3 was shown to have affinity to serotonin 5-HT1A
receptors and to possess weak partial 5-HT1A receptor
agonist properties.3 Compound (()-4,1e,4 the racemic
C11-methoxy analogue of 3, was almost devoid of DA
receptor agonist properties.4 In addition, the C10-
methyl analogue of 3, compound 5 (HYMAP), was
recently shown to be a potent and selective 5-HT1A
receptor agonist.3,5
phines in order to investigate the SAR of these com-
pounds at 5-HT1A
, D1, and D2A receptors. The com-
pounds were evaluated pharmacologically in vitro using
receptor binding and in vivo by use of biochemical and
behavioral assays in reserpinized rats. The differences
in binding profiles were rationalized by modeling of
ligand-receptor interactions using homology-based re-
ceptor models of the 5-HT1A
sites.
and D2A receptor binding
Ch em istr y
The syntheses of the novel aporphines and norapor-
phines presented in Table 1 are shown in Schemes 1-3,
and detailed protocols are found in the Experimental
Section.
Syn th esis. Aporphines 7-13 were prepared from
triflate 6 using efficient palladium- and nickel-catalyzed
coupling reactions (Scheme 1). Triflate 6 was synthe-
sized from 3 using a modified version of the procedure
reported by Cacchi et al.6 The palladium-catalyzed
cross-coupling reactions of 6 with stannanes were
performed following the protocol for sterically hindered
triflates reported by Saa´ et al.7 (R)-Aporphine (7),5c
synthesized as the racemate in 1924 by Gadamer,
Oberlin, and Schoeler,8 was prepared from 6 by a
In the present investigation we have synthesized a
series of C11-substituted (R)-aporphines and norapor-
* Correspondence and reprints: Dr. Anette M. J ohansson. Phone:
+46-18-174336. Fax: +46-18-174024. E-mail: Anette@bmc.uu.se.
† Uppsala University.
‡ Astra Arcus AB.
§ University of Go¨teborg.
X Abstract published in Advance ACS Abstracts, August 1, 1996.
S0022-2623(96)00189-6 CCC: $12.00 © 1996 American Chemical Society

3504 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Ta ble 1. Physical Data of Novel (R)-Aporphines
Hedberg et al.
recrystn
solventsa
[R]rt (deg)
D
compd
R1
R2
mp (°C)
yield (%)
(c 1.0, MeOH)
formula
6
8
9
OTf
Me
COMe
CHdCH2
Et
CN
CONH2
Ph
2-OMe-Ph
2-OH-Ph
OMe
OH
OMe
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Bn
Bn
H
H
2-Pr
2-Pr
233-237b
231-234b
227-230b
235-236b
234-238b
232-234b
230-231b
234-237b
233-234b
211-212b
230-235b
280-283b
291-292b,d
320-328b,e
230-232
74
70
87
57
56
66
92
60
49
97
85
77
88
94
86
92
A
B
C
D
B
E
F
G
F
H
F
I
-204.4
-51.4
C18H16F3NO3‚HCl
C18H19N‚HCl
-237.7
+88.6
C19H19NO‚HCl‚1/4H2O
C19H19N‚HCl
10
11
12
13
14
15
16
20
21
22
23
24
25
-44.5
C19H21N‚HCl‚1/4H2O
C18H16N2‚HCl
C18H18N2O‚HCl
C23H21N‚HCl
-170.9
-228.4
-111.5
-143.3
-107.3
-41.4
C24H24NO‚HCl
C23H22NO‚HCl‚1/2H2O
C24H23NO‚HCl
C23H21NO‚HCl
C17H17NO‚HCl
C16H15NO‚HCl
-24.0c
-115.3
-90.0f
-103.2
-80.9
B and J
H
F
OH
OMe
OH
C
20H23NO‚HCl
267-270b,g
K
C19H21NO‚HCl
a
A: EtOH. B: MeCN. C: MeCN/t-BuOMe. D: EtOH/ether. E: MeCN/EtOH (1:1)/t-BuOMe. F: MeCN/ether. G: MeCN/EtOH
(1:1)/ether. H: 2-PrOH/ether. I: MeOH. J : MeOH/ether. K: MeCN/MeOH. Decomposition. c (c 0.10, MeOH). Green at 269-270 °C.
b
d
e Green at 280-285 °C. f (c 0.25, MeOH). Green at 240 °C.
g
Sch em e 1a
a
Reagents: (a) PhN(Tf)2, Et3N, K2CO3, CH2Cl2, reflux; (b) (PPh3)2PdCl2, Ph2P(CH2)3PPh2, Bu3N, HCO2H, DMF, 80 °C; (c) Me4Sn,
(PPh3)2PdCl2, LiCl, PPh3, 2,6-di-tert-butyl-4-methylphenol, DMF, 120 °C; (d) i. Bu3SnC(OEt)CH2, (PPh3)2PdCl2, LiCl, PPh3, 2,6-di-tert-
butyl-4-methylphenol, DMF, 120 °C, ii. HCl (aq), THF, rt; (e) Bu3SnCHdCH2, (PPh3)2PdCl2, LiCl, PPh3, 2,6-di-tert-butyl-4-methylphenol,
DMF, 120 °C; (f) H2, 10% Pd(C), THF; (g) KCN, PPh3, (PPh3)2NiCl2, Zn, DMF, N2, 120 °C; (h) H2SO4, H2O, 120 °C; (i) PhSnBu3, (2-
furyl)3P, (dba)3Pd2, LiCl, NMP, 100 °C; (j) (2-methoxyphenyl)SnBu3, (PPh3)2PdCl2, CuI, LiCl, 2,6-di-tert-butyl-4-methylphenol, DMF, 145
°C; (k) 48% HBr (aq), N2, 120 °C. Tf ) CF3SO2-.
palladium-catalyzed hydrogenolysis.9 Analogues 8-10
were synthesized by cross-coupling of 6 with tetra-
methylstannane, tributyl(1-ethoxyvinyl)stannane, and
tributylvinylstannane, respectively. The intermediate
vinyl ether formed in the coupling of the 1-ethoxyvinyl
group was hydrolyzed to give 9. Catalytic hydrogena-
tion of 10 afforded 11. The C11-cyano derivative 12 was
prepared by a modified version of the nickel-catalyzed
cyanation method reported by Chambers and Widdow-
son.10 Longer reaction time and elevated reaction
temperature, the use of dimethylformamide as solvent,
and additions of extra portions of Ni catalyst and ligand
were found to be necessary for the reaction to proceed.
Hydrolysis of 12 in acid afforded the amide analogue
13. In the syntheses of the C11-arylated compounds 14
and 15, the protocol of Saa´ et al.7 produced unsatisfac-
tory results. Instead, compound 14 was prepared from
6 using the recently reported method of Farina and co-
workers11 with tri(2-furyl)phosphine as ligand and tris-
(dibenzylideneacetone)dipalladium [(dba)3Pd2] as pre-
catalyst in N-methylpyrrolidinone (NMP). This latter
method, however, failed in the attempted synthesis of
the 2-methoxyphenyl-substituted 15 from 6 and (2-
methoxyphenyl)stannane even at elevated temperatures
and/or using triphenylarsine as ligand. Instead, the use
of conditions in which copper(I) cocatalysis was uti-
lized12 gave the desired 15 in moderate yield. De-
methylation of 15 afforded 16. Compounds 15 and 16
most likely exist as atropisomers13 as judged by their
1
complicated H and 13C NMR spectra.

11-Substituted (R)-Aporphines
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3505
Sch em e 2a
Sch em e 3a
a
Reagents: (a) Bu3N, HCO2H, PPh2(CH2)3PPh2, (PPh3)2PdCl2,
DMF, N2, 80 °C; (b) Zn dust, 1 M H2SO4, EtOH, reflux; (c) H2,
10% Pd(C), HOAc, rt; (d) (R)-2-methoxy-2-phenylacetyl chloride,
CH2Cl2, 1 M NaOH, rt.
(Schemes 2 and 3), or by previously reported routes,3,15,21
are optically pure.
a
Reagents: (a) Bu3N, HCO2H, PPh2(CH2)3PPh2, (PPh3)2PdCl2,
DMF, N2, 80 °C; (b) 48% HBr (aq), N2, 120 °C; (c) H2, 10% Pd(C),
HOAc, rt; (d) 2-iodopropane, K2CO3, diisopropylethylamine, CH3CN,
70 °C. Tf ) CF3SO2-.
Molecu la r Mod elin g. Low-energy conformations of
selected nonprotonated (R)-aporphines were identified
by molecular mechanics MM2(91)22 calculations using
the torsion angle driver procedure as described previ-
ously.3,15 When studying intermolecular interactions
between the ligands and active site residues, the pro-
tonated form of the ligands was used.
The (R)-noraporphines 2,1c,14 18,1c,14 and 20-25 were
prepared using the synthetic sequences shown in Scheme
2. Palladium-catalyzed hydrogenolysis of triflates 1715
and 1915 gave the corresponding analogues 18 and 20,
which were further O-demethylated to phenols 2 and
21, respectively. N-Debenzylation of 20 and 21 afforded
22 and 23, respectively. Alkylation of 22 with 2-io-
dopropane gave 24, which was O-demethylated to 25.
3,15
3,23,24
Models of human 5-HT1A
and D2A
receptors
based on a presumed homology in three-dimensional
structure between bacteriorhodopsin and the G-protein-
coupled receptors have been constructed previously
according to a strategy first described for modeling of
the muscarinic m1 receptor.25 A combination of this
homology-based approach with a systematic side chain
search and minireceptor optimization using the pseu-
doreceptor/minireceptor modeling program Yak26 re-
sulted in an aporphine-derived binding site model for
the 5-HT1A receptor.15 This same procedure was applied
to the homology-based D2A receptor model3,23 using the
side chains of Asp114 and Ser193 in the systematic
search and adding Phe390 as a third key residue when
fitting 1 in the resulting models. Active site optimiza-
tion was performed in the pseudoreceptor/minireceptor
modeling program Yak26 using first 1 and then including
also 2 to derive a model for examination of putative
binding modes of related aporphines in the D2A binding
site (Figure 1).
Binding modes in the 5-HT1A and D2A receptor models
were examined by fitting the N+-hydrogen, the A-ring
centroid, and the C11, or the C11-oxygen, of the differ-
ent conformations of selected aporphines onto either 5
or 14, as docked in the active site of the 5-HT1A model,15
and onto 1, docked in the D2A site. All Yak minimiza-
tions were performed on both ligand(s) and residues
simultaneously using the all-atom Yeti force field while
keeping the backbone atoms fixed.27 Translational and
A 4% yield of derivative 26 was isolated in the
synthesis of 20 from 19 (Scheme 3). This byproduct was
probably formed by a palladium-catalyzed oxidation of
20 during the workup.16 Interestingly, Murahashi et
al. have reported that when treating (S)-N,N,1-tri-
methylpropylamine with palladium(0) in a palladium-
catalyzed oxidation reaction,17a in which a palladium
hydride-iminium complex is thought to be involved,17
the recovered starting material is partially racemized.
This result, together with the fact that 6a,7-didehy-
droaporphine analogues of 26 have been reduced to the
corresponding aporphines in high yield using metal
hydrides (lithium aluminum hydride18 and sodium
cyanoborohydride16b), led to the speculation that 20
might have been partially racemized during the workup
procedure. Compound 26 was therefore converted to 28,
the racemate of 22, by reduction19 followed by hydro-
genolysis. Racemic 28 was then used to prepare the
diastereomeric O-methylmandelic amides20 29 and 30,
thus enabling a determination of the optical purity of
22. The enantiomeric purity of 22 was determined from
the corresponding O-methylmandelic amide to be 99.8%
ee (capillary GC). This result indicates that all apor-
phines synthesized using the routes presented here

3506 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Hedberg et al.
Ta ble 2. Affinities of Selected Aporphines to Rat Brain 5-HT1A
and D1 Receptor Recognition Sites Labeled by [3H]-8-OH-DPAT
and [3H]SCH23390, Respectively, and at Cloned Human D2A
Receptors Expressed in Ltk- Cells and Labeled by
[3H]Raclopride
Ki (nM)a
[3H]-8-OH-DPAT
(5-HT1A
[3H]SCH23390
(D1)b
[3H]raclopride
b
compd
)
(D2A)
1
2
3
4
6
7
8
9
10
11
12
13
14
15
16
18
20
21
22
23
24
25
296 ( 15c
39.5 ( 6.1
9.6 ( 3.1c
5.1 ( 1.2
236d
1090 ( 85
20.4 ( 1.0c
606 ( 21
386 ( 44
717 ( 99
700 ( 137
4690 ( 1100
720 ( 106
270 ( 0.5
279 ( 35
>4000
41.9 ( 4.7c
12.7 ( 1.6
58.5 ( 9.5c
551 ( 31
195 ( 26
527 ( 296
201 ( 91
311 ( 28
177 ( 54
79.2 ( 32
183 ( 14
7.7 ( 0.2
80.0 ( 2.3
14.4 ( 1.5
57.1 ( 4.1
6.0 ( 0.6
4.5 ( 0.61
41.1 ( 0.2
198 ( 32
>10 000
1.8 ( 0.4
3630 ( 260
>3000
233 ( 3
26.9 ( 5.4
28.5 ( 2.4
31.3 ( 1.9
3220 ( 1350
1330 ( 130
1570 ( 590
826 ( 319
3750 ( 300
>5000
>10 000
>3000
>5000
1097 ( 67
>10 000
>10 000
F igu r e 1. Schematic representation of the interaction be-
tween 2 and the D2A receptor binding site showing only the
helices involved in the receptor-ligand interactions. A rein-
forced ionic bond between Asp114 and the protonated nitrogen,
a hydrogen bond between Ser193 and the C11-hydroxyl group,
and an edge-to-face interaction between Phe390 and the A-ring
of 2 represent key-interactions. Also indicated is the presence
of the “propyl cleft”.
>1000
>5000
>1000
>1000
>1000
>1000
8.3 ( 1.7
5.7 ( 1.6
375 ( 120
437 ( 240
1010 ( 325
a
The Ki values are means ( standard errors of n ) 2-3
b
experiments. None of the compounds tested produced a clear
biphasic binding profile at the D1 and/or D2A receptor. Therefore,
the results have been interpreted in terms of a one-site model (see
rotational degrees of freedom were considered for the
ligand(s), whereas only torsional degrees of freedom
were considered for the amino acids.
d
ref 3). c From ref 3. From ref 2b.
R-adrenoceptor agonists, respectively. The in vivo ef-
fects of single doses of 2 (3.2 and 32 µmol/kg), 3 (3.5
µmol/kg), 11 (3.3 µmol/kg), 14 (2.9 µmol/kg), 18 (30 µmol/
kg), and 23 (3.7 µmol/kg) on 5-HTP and DOPA ac-
cumulation following decarboxylase inhibition by means
of (3-hydroxybenzyl)hydrazine dihydrochloride (NSD1015,
117 mg/kg sc) were investigated in the limbic forebrain,
striatum, and cortex of reserpine-pretreated rats, as
described previously.15,29-31 None of the compounds was
able to decrease the synthesis of 5-HT at the doses
tested (Table 3), tentatively suggesting a lack of efficacy/
potency at 5-HT1A receptors. As expected, the D2
receptor agonist 21c significantly decreased the synthesis
of DA in limbic forebrain and striatum. Interestingly,
derivative 3, previously reported to lack D2 receptor
agonist activity,1e and 18 and 23, which display micro-
molar affinity to both D1 and D2A receptors, significantly
decreased the synthesis of DA in limbic forebrain and
striatum.
Stimulation of postsynaptic 5-HT1A receptors by means
of agonists, like (R)-8-hydroxy-2-(dipropylamino)tetralin
[(R)-8-OH-DPAT],32 results in a clear-cut motor behav-
ioral syndrome, the “5-HT syndrome”, in rats (flattened
body posture, forepaw treading, and hindlimb abduc-
tion).33,34 The ability of compounds 2, 3, 11, 14, 18, and
23 to elicit this behavior was assessed as described
previously (Table 4).15,33,34 None of the compounds
tested induced a clear “5-HT syndrome”. However, the
D2 receptor agonist 21c (3.2 and 32 µmol/kg) induced an
atypical motor stimulation syndrome accompanied by
stereotyped behavior in the rat, more akin to dopam-
inergic overactivation mixed with some 5-HT1A receptor
agonist-like effects (see Table 4). Postsynaptic dopam-
inergic agonists, such as 1, induce stereotyped behavior
(sniffing, licking chewing, etc.) in rats. In addition, they
P h a r m a cology
Recep tor Bin d in g Stu d ies. The compounds were
examined in vitro for their ability to displace [3H]-8-
OH-DPAT, [3H]SCH23390, and [3H]raclopride binding
to rat hippocampal 5-HT1A, rat striatal D1, and cloned
human D2A receptors, respectively, as described previ-
ously.3 The results are presented in Table 2, and the
affinity of (R)-apomorphine (1) to these receptors is also
included. All N-methylated compounds, except for 1 and
13, displayed moderate to high affinity to 5-HT1A
receptors. For example, derivatives 3 (OH), 4 (OMe), 6
(OTf), 10 (vinyl) 11 (Et), and 14 (Ph) have affinities <
10 nM to 5-HT1A receptors. Among the noraporphines,
only 22 and 23 showed high affinity to and selectivity
for 5-HT1A receptors. All compounds tested displayed
at least 10-fold lower affinities to D1 receptors compared
to the nonselective D1 receptor antagonist 3.1e Several
derivatives showed moderate affinity to D2A receptors,
the most potent compounds being 1-3 and 11. Inter-
estingly, the C11-ethyl analogue 11 showed only a 2-fold
lower affinity to D2A receptors in comparison to 1. In
addition, the previously reported D2 receptor agonist 21c
appeared to have a modest 3-fold selectivity for D2A
receptors versus 5-HT1A receptors. In the noraporphine
series, the N-benzyl- and N-isopropyl-substituted de-
rivatives 20, 21 and 24, 25 displayed low affinities to
all receptors tested.
Bioch em istr y a n d Beh a vior . The rationale for the
in vivo biochemical experimental protocol is based on
the well-established phenomenon of (auto)receptor-
mediated feedback inhibition of presynaptic monoam-
inergic neurons.28 Thus, the synthesis of 5-HT, DA, and
norepinephrine (NA) is inhibited by 5-HT, DA, and

11-Substituted (R)-Aporphines
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3507
Ta ble 3. Effects of Selected (R)-Aporphines on Central 5-HTP and DOPA Accumulation in Reserpine-Pretreated Rats
5-HTP accumulation:a percent of
DOPA accumulation:b percent of
saline controls, mean ( SEM
saline controls, mean ( SEM
dose
compd
(µmol/kg)
n
limbic forebrain
striatum
cortex
limbic forebrain
striatum
cortex
2
3.2
32
3.5
3.3
2.9
30
2
2
4
3
3
4
3
122 ( 10
107 ( 6
98 ( 12
83 ( 5
97 ( 4
120 ( 23
83 ( 10
95 ( 6
87 ( 2
103 ( 8
93 ( 23
104 ( 10
133 ( 7
89 ( 12
122 ( 11
66 ( 5*
58 ( 4*
67 ( 7*
87 ( 10
95 ( 6
66 ( 3*
78 ( 1*
29 ( 11*
28 ( 6*
30 ( 3*
92 ( 9
106 ( 5
112 ( 1
103 ( 9
100 ( 6
107 ( 14
82 ( 7
3
11
14
18
23
95 ( 2
106 ( 7
87 ( 5
93 ( 4
96 ( 4
108 ( 10
43 ( 4*
71 ( 4*
3.7
89 ( 6
96 ( 6
a
Shown is the amount of accumulated 5-HTP in percent of saline controls (for 11, 14 and 23, 100% limbic forebrain 235 ( 33 ng/g,
striatum 187 ( 11 ng/g, cortex 95 ( 11 ng/g, n ) 4; for 2, 3 and 18, 100% limbic forebrain 164 ( 11 ng/g, striatum 158 ( 14 ng/g, cortex
123 ( 15,n ) 3), means ( SEM, n ) 2-4. Statistical differences were calculated by one-way ANOVA followed by Fischer’s protected
b
least-significant-difference test (PLSD): *p < 0.05 vs saline controls. Shown is the amount of accumulated DOPA in percent of saline
controls (for 11, 14 and 23, 100% limbic forebrain 765 ( 50 ng/g, striatum 3526 ( 323 ng/g, cortex 103 ( 11 ng/g, n ) 4; for 2, 3 and 18,
100% limbic forebrain 674 ( 40 ng/g, striatum 2511 ( 173 ng/g, cortex 100 ( 7, n ) 3), means ( SEM, n ) 2-4. Statistical differences
were calculated by one-way ANOVA followed by Fischer’s PLSD: *p < 0.05 vs saline controls.
Ta ble 4. Postsynaptic 5-HT1A Receptor-Mediated Behaviora
tion. Figure 2 shows the gradual change in binding
mode when 3, 4, and 14 were optimized simultaneously.
The C11-oxygenated compounds 3 and 4 interact through
hydrogen bonding with Ser168 and differ from the C11-
arylated 14 in their position of the protonated nitrogen
and therefore also in the oxygen atom at Asp116 with
which they interact. Interestingly, the C11-ethyl and
C11-phenyl derivatives 11 and 14 had high affinities
to 5-HT1A receptors and moderate to low affinities to
D2A, but neither produced any effects in vivo in the doses
tested (Table 2-4). Thus, both 11 and 14 seem to
behave as weak partial agonists or antagonists, al-
though further work is needed to examine this possibil-
ity.
dose
(µmol/kg)
Σ, scores
compd
median (range)
n
saline control
0 (0-1)
1 (0-2)b
3 (3-3)b
0 (0-1)
0 (0-1)
0 (0-1)
0 (0-0)
0 (0-0)
3
2
2
3
3
3
3
3
2
3.2
32
3.5
3.3
2.9
30
3
11
14
18
23
3.7
a
Behavioral scores according to a 3-item, 4-point intensity-
based rating scale (0 ) absent, 1 ) equivocal, 2 ) clearly present,
3 ) intense). Rating period 60 s bin, 30 min after drug admin-
istration, just before NSD1015 injection. Rats were pretreated
b
with reserpine (5 mg/kg, sc) 18-20 h before. Compound 2 (32
µmol/kg) induced an atypical motor stimulation syndrome: sudden
jerks and movement “explosions”, running around, aggression,
strong stereotyped sniffing, licking and chewing, clear-cut to
intense hindlimb extension, tremor to almost convulsive extent.
The lower dose of 2 resulted in milder but still mixed DA and
5-HT1A agonist-like behavioral activation.
The D2A binding site model obtained after optimiza-
tion in Yak confirms the previously described interac-
tions: a reinforced ionic bond between the protonated
nitrogen and Asp114, a hydrogen bond from the phenolic
hydrogen to Ser193, an aromatic edge-to-face interaction
between the A-ring and Phe390, and a “propyl cleft”
accommodating the N-propyl substituent of high-affinity
D2A receptor ligands (Figure 1).3 The propyl cleft is
lined by Leu113, Asp114, Met117, Cys118, Trp386,
Phe389, Tyr408, Thr412, and Tyr416 (Figures 1 and 3).
Optimization of C11-substituted aporphines in the
obtained binding site of the D2A receptor model indicated
that in accordance with the moderate affinity of 11 to
D2A receptors, there is space available for a C11-ethyl
substituent when Ser193 and His393 are rotated.
Residues in close contact with the ethyl moiety (i.e.,
those that have atoms within 3 Å of the substituent)
are Leu171, Ser193, Ser197, Phe390, and His393 (Fig-
ure 3); those residues could be considered to contribute
to the slight enhancement of D2A affinity when compar-
ing 11 to 7 and 8. The reduction in affinity when ethyl
is substituted with methoxy (11 vs 4) is probably due
to unfavorable electrostatic interactions of the methoxy
moiety with several of these residues.
induce locomotor activity and rearing.35,36 The other
compounds were inactive at the dose tested.
Discu ssion
The pharmacological evaluation of the present series
of derivatives (Tables 2-4) reveals some interesting
SAR of C11-substituted (R)-aporphines. High affinities
to 5-HT1A receptors are displayed by a number of
aporphines carrying sterically and electronically differ-
ent C11-substituents. As was shown previously,23
a
hydrogen bond-donating group in the C11-position of the
aporphines seems to be required for an optimal interac-
tion with D2A receptors. However, hydrogen bonding
to a C11-substituent does not seem to be essential for
an interaction of C11-substituted aporphines with 5-HT1A
receptors.37 This behavior can be rationalized using a
previously described binding site model for the 5-HT1A
receptor. In this model 5 interacts with Asp116 (a
reinforced ionic bond), Ser168 (a hydrogen bond), and
Phe362 (an aromatic edge-to-face interaction).15 Com-
pound 14 has to bind in a different orientation compared
to 5 in order to fit in the binding site and to maintain
an interaction with Asp116.15 The C11-phenyl sub-
stituent of 14 then fits into a pocket lined by Ser168,
Met172, Thr196, Ser199, Thr200, Phe362, Ala365, and
Leu366. Optimization of 3 and 4 in the derived binding
site indicates that 3 assumes an orientation similar to
515 while 4 appears to accept an intermediate orienta-
The influence of the N-substituent on the affinity to
5-HT1A and D2A receptors is evident by a comparison of
derivatives 23 (H), 3 (Me), 2 (Pr), 25 (2-Pr), and 21 (Bn).
An increase in affinity and selectivity for 5-HT1A recep-
tors was observed when the size of the N-substituent
was decreased to methyl (3) or hydrogen (23), whereas
an N-propyl group as in 2 seemed to be optimal for a
potent interaction with D2A receptors.23 The propyl cleft
described above for the D2A binding site accommodates
a low-energy conformation of the N-propyl substituent

3508 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Hedberg et al.
F igu r e 2. Stereoscopic representation of a 5-HT1A receptor binding site model obtained by simultaneous optimization of 3 (red),
4 (yellow), and 14 (green) showing a gradual change in binding mode which is dependent on the C11-substituent.
F igu r e 3. Stereoscopic representation of a D2A receptor binding site model interacting with 2 (red) and 11 (green). The residues
responsible for key interactions (Asp114, Ser193, and Phe390) and some of the residues lining the N-propyl cleft and the site
accommodating the C11-ethyl substituent are shown. The two orientations shown for Ser193 and His393 correspond to those
obtained after individually optimizing the site for either ligand.
of 2 (∆Es ) 0.3 kcal/mol). Docking all 11 minimum
energy conformations of 2 into the 5-HT1A binding site
revealed that none of the low-energy conformations
could bind to both Asp116 and Ser168 without undergo-
ing steric repulsion. The lowest energy conformation
that could be docked had a relative steric energy of 2.5
kcal/mol which may explain the drop in 5-HT1A receptor
affinity observed for 2 compared to 3.
A moderate to high affinity of 2, 18, and 23 to 5-HT1A
receptors, together with their inability to decrease the
5-HT synthesis rate and to induce the 5-HT1A syndrome
in vivo at the doses tested, may indicate antagonistic
properties or lack of efficacy of these analogues at
5-HT1A receptors. As expected, the D2 receptor agonist
21c decreased the DA synthesis rate, induced stereo-
typed behavior, and displayed high affinity to D2A
receptors. Surprisingly, 18 and 23 also produced D2
agonistic effects by decreasing the DA synthesis in vivo,
although they only displayed micromolar affinity to D2A
receptors. These results indicate mixed agonist/antago-
nist effects of 2, 18, and 23 at 5-HT1A and D2A receptors.
It is also possible that pharmacokinetic factors could
contribute to the complex profile observed in vivo with
these compounds,38 and further work is needed to clarify
these issues.
5-HT1A receptor agonist.3 The moderate affinity to D2A
receptors together with the ability to decrease the DOPA
accumulation indicates that 3 may also have agonistic
properties at D2 receptors.
The present series of compounds clearly shows how
small structural changes significantly influence the
receptor binding profile and the in vivo pharmacological
activity. Although the complex in vivo pharmacological
profiles are not fully understood yet, the in vitro binding
data could be rationalized using models of the 5-HT1A
and D2A receptor binding sites.
Exp er im en ta l Section
Ch em istr y. Gen er a l Com m en ts. Melting points (uncor-
rected) were determined in open glass capillaries on an
Electrothermal melting point apparatus. Optical rotation
measurements were obtained on a Perkin-Elmer 241 polarim-
eter. Elemental analyses (C, H, N) were performed at Mik-
rokemi AB, Uppsala, Sweden, and were determined within
(0.4% of the theoretical values. The 1H and 13C NMR spectra
of the amines were recorded on a J EOL J NM-EX270 spec-
trometer at 270 and 67.8 MHz, respectively, and are referenced
to internal tetramethylsilane. In the 1H and 13C NMR spectra
of the hydrochloride salts of 6-13 in CD3OD, diastereomeric
conformations at or below the slow exchange limit (SEL) were
present.39 The spectra could not be simplified by heating the
CD3OD sample to 50 °C. Therefore, the NMR data reported
for compounds 6-16 were recorded for the amines in CDCl3,
whereas for the 11-oxygenated analogues 2, 18, and 20-25,
Compound 3 was recently reported to have antago-
nistic properties at D1 receptors1e and to be a partial

11-Substituted (R)-Aporphines
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3509
the data for the hydrochlorides in CD3OD are reported.
Compounds 15 and 16 most likely exist as mixtures of
atropisomers,13 as judged by their complicated 1H and 13C
NMR spectra. Infrared (IR) spectra were recorded on a Perkin-
Elmer 298 infrared spectrophotometer. Thin-layer chroma-
tography (TLC) was performed by using aluminum sheets
precoated with either silica gel 60 F254 or aluminum oxide 60
F254 E neutral (0.2 mm; E. Merck). For preparative TLC,
plates precoated with either silica gel 60 F254 (2.0 mm) or
aluminum oxide F254 T (1.5 mm) (E. Merck) were used.
Column chromatography was performed on silica gel 60 (230-
400 mesh; E. Merck) or aluminum oxide 90 (70-230 mesh,
activity II-III; E. Merck). Capillary GC was performed on a
Carlo Erba 6000 instrument, equipped with a SE 54 fused
silica gel capillary column (25 m). Dimethylformamide (DMF)
was distilled from calcium hydride and stored over 4 Å
molecular sieves under nitrogen. N-methyl-2-pyrrolidinone
(NMP) was used as received. Physicochemical properties of
the novel compounds are also shown in Table 1.
29.2, 35.7, 44.1, 52.8, 62.3, 125.6, 125.7, 126.0, 126.8, 127.4,
130.8, 132.7, 133.4, 133.9, 134.3, 136.0, 137.1.
(-)-(R)-11-Acetyla p or p h in e (9). A mixture of 6 (154 mg,
0.40 mmol), LiCl (75 mg, 1.8 mmol), PPh3 (180 mg, 0.69 mmol),
(PPh3)2PdCl2 (35 mg, 0.050 mmol), and tributyl(1-ethoxyvinyl)-
stannane (293 mg, 0.85 mmol) in dry DMF (10 mL) was stirred
in a closed vessel at 120 °C for 18 h. The volatiles were
evaporated in vacuo. The residue was diluted with ether,
filtered through a pad of aluminum oxide, and concentrated.
THF (10 mL) and concentrated aqueous HCl (10 mL) were
added to the residue, and the mixture was stirred for 2 min.
The volatiles were evaporated, and the residue was partitioned
between 10% aqueous NaHCO3 and CH2Cl2. The combined
organic extracts were dried (K2CO3), filtered, and concentrated
to give an oil which was chromatographed [silica gel; CH2Cl2/
MeOH (gradient 39:1-9:1) and aluminum oxide; ether/hexanes
(gradient 1:1-1:0)]. The amine was converted into the hy-
drochloride, which was recrystallized from MeCN/tert-butyl
methyl ether to give 109 mg (87%) of pure 9‚HCl.
Syn th eses. (-)-(R)-11-[[(Tr iflu or om eth yl)su lfon yl]oxy]-
a p or p h in e (6). N-Phenyltrifluoromethanesulfonimide (3.41
g, 9.54 mmol) and K2CO3 (1.12 g, 8.10 mmol) were added to a
refluxed (15 min) slurry of 3 (2.01 g, 8.00 mmol) and Et3N (3.40
mL, 24.5 mmol) in CH2Cl2 (100 mL), kept under nitrogen.
Additional portions of N-phenyltrifluoromethanesulfonimide
(400 mg, 1.11 mmol and 330 mg, 0.92 mmol) were added after
24 and 43 h. After 70 h, the heating was interrupted, and
the reaction mixture was extracted with 10% aqueous NaH-
CO3. The organic layer was dried (K2CO3), filtered, and
concentrated in vacuo. The oily residue was chromatographed
[silica gel; CHCl3/MeOH (9:1) and aluminum oxide (deactivated
with 5% H2O); ether]. The pure amine was converted into the
hydrochloride salt, which was recrystallized from EtOH to give
9: IR (film) 1684, 1463, 1374 cm-1; 1H NMR (CDCl3) δ 2.27
(s, 3 H), 2.50-2.67 (m, 2 H), 2.60 (s, 3 H), 2.80 (dd, 1 H), 3.08
(ddd, 1 H), 3.12-3.28 (m, 3 H), 7.07-7.65 (m, 6 H); 13C NMR
(CDCl3) δ 29.2, 31.0, 34.6, 44.1, 52.9, 61.7, 126.47, 126.48,
126.8, 127.4, 128.7, 130.1, 131.7, 133.0 133.5, 135.0, 137.2,
139.9, 206.4.
(+)-(R)-11-Vin yla p or p h in e (10). A mixture of 6 (429 mg,
1.11 mmol), LiCl (384 mg, 8.95 mmol), PPh3 (180 mg, 0.69
mmol), (PPh3)2PdCl2 (94 mg, 0.13 mmol), tributylvinylstannane
(720 mg, 2.3 mmol), and a few crystals of 2,6-di-tert-butyl-4-
methylphenol in dry DMF (10 mL) was stirred in a closed
vessel at 120 °C for 5 h. The volatiles were evaporated in
vacuo, and the residue was partioned between CHCl3 and 10%
aqueous NaHCO3. The combined organic layers were dried
(K2CO3), filtered, and concentrated. The residue was chro-
matographed [silica gel; CHCl3/MeOH (gradient 39:1-9:1) and
aluminum oxide (deactivated with 3.5% H2O); ether/hexanes
(gradient 1:1-3:1)]. The amine was converted into the hy-
drochloride salt, which was recrystallized from EtOH/ether to
give 191 mg (57%) of pure 10‚HCl.
2.49 g (74%) of pure 6‚HCl: IR (KBr) 2369, 1212, 1142 cm-1
.
6: 1H NMR (CDCl3) δ 2.51-2.63 (m, 2 H), 2.56 (s, 3 H), 2.77
(dd, 1 H), 3.05 (ddd, 1 H), 3.13-3.25 (m, 3 H), 7.14 (app d, 1
H), 7.13-7.16 (m, 4 H), 7.79 (app d, 1 H); 13C NMR (CDCl3) δ
29.1, 35.0, 44.1, 52.8, 61.5, 118.5 (q), 121.4, 126.0, 126.6, 128.1,
128.3, 128.7, 128.3, 129.2, 133.3, 135.2, 139.7, 146.4; 19F NMR
(CDCl3) δ -73.6.
10: 1H NMR (CDCl3) δ 2.49-2.61 (m, 2 H), 2.56 (s, 3 H),
2.77 (dd, 1 H), 3.01-3.27 (m, 4 H), 5.71 (dd, 1 H), 5.77 (dd, 1
H), 7.03 (dd, 1 H), 7.09 (app d, 1 H), 7.18-7.26 (m, 3 H), 7.45-
7.50 (m, 2 H); 13C NMR (CDCl3) δ 29.1, 35.3, 44.1, 52.9, 62.0,
114.6, 125.6, 126.9, 127.1, 127.3, 127.4, 127.7, 132.5, 132.8,
132.9, 135.3, 135.7, 136.9, 138.3.
(-)-(R)-11-Eth yla p or p h in e (11). A slurry of 10 (93 mg,
0.36 mmol) and Pd(C) (10%, 90 mg) in THF was hydrogenated
at atmospheric pressure for 5 h. The catalyst was filtered off
(aluminum oxide), and the volatiles were evaporated in vacuo.
The residue was chromatographed [silica gel; CH2Cl2/MeOH
(gradient 1:0-9:1)]. The amine was converted into the hy-
drochloride salt, which was recrystallized from MeCN to give
57 mg (56%) of pure 11‚HCl.
(-)-(R)-Ap or p h in e (7).5c A mixture of 6 (223 mg, 0.58
mmol), (diphenylphosphino)propane (dppp; 37 mg, 0.089 mmol),
(PPh3)2PdCl2 (25 mg, 0.035 mmol), Bu3N (0.55 mL, 2.31 mmol),
and formic acid (66 µL, 1.8 mmol) in dry DMF (5 mL) was
stirred at 80 °C for 15 h. The volatiles were evaporated in
vacuo, and the residue was partitioned between CH2Cl2 and
10% aqueous NaHCO3. The organic layer was dried (K2CO3),
filtered, and concentrated. The residue was chromatographed
[silica gel; CHCl3/MeOH (gradient 1:0-9:1) and aluminum
oxide ether/hexanes (gradient 1:1-6:1)]. The amine was
converted into the hydrochloride salt, which was recrystallized
from MeCN to give 80 mg (51%) of pure 7‚HCl: mp 233-234
°C (lit.5c mp 229-232 °C); [R]23 -108.2° (c 1.0, MeOH) [lit.5c
D
[R]28.5 -109.4° (c 0.0019, MeOH)].
D
11: 1H NMR (CDCl3) δ 1.34 (t, 3 H), 2.47-2.61 (m, 2 H),
2.55 (s, 3 H), 2.77 (dd, 1 H), 2.83-3.14 (m, 5 H), 3.21 (ddd, 1
H), 7.09 (app d, 1 H), 7.12 (app d, 1 H), 7.19 (app t, 1 H), 7.23
(app t, 1 H), 7.27 (app d, 1 H), 7.42 (app d, 1 H); 13C NMR
(CDCl3) δ 15.3, 26.7, 29.2, 36.0, 44.1, 52.8, 62.2, 125.5, 125.7,
125.9, 127.0, 127.4, 128.8, 132.8, 133.3, 133.7, 136.1, 137.1,
140.6.
(-)-(R)-11-Cya n oa p or p h in e (12). Dry nitrogen gas was
passed through a mixture of 6 (588 mg, 1.53 mmol), KCN (123
mg, 1.89 mmol), PPh3 (208 mg, 0.79 mmol), Zn (152 mg, 2.33
mmol), and (PPh3)2NiCl2 (255 mg, 0.39 mmol) in dry DMF (5
mL). The vessel was sealed, and the mixture was stirred at
120 °C for 48 h. Additional portions of PPh3 (78.0 mg, 0.30
mmol) and (PPh3)2NiCl2 (97 mg, 0.15 mmol) were added, and
after an additional 24 h, the mixture was diluted with CH2-
Cl2 and filtered through a pad of aluminum oxide. The
volatiles were evaporated in vacuo, and the residue was
chromatographed [silica gel; CHCl3/MeOH (gradient 39:1-9:
1) and aluminum oxide; ether/hexanes/methanol (gradient 1:2:
0-97:0:3)]. The amine was converted into the hydrochloride
salt, which was recrystallized from MeCN/EtOH (1:1)/tert-butyl
methyl ether to give 299 mg (66%) of pure 12‚HCl.
7: 1H NMR (CDCl3) δ 2.55 (ddd, 1 H), 2.57 (s, 3 H), 2.63-
2.81 (m, 2 H), 3.07 (ddd, 1 H), 3.14-3.26 (m, 3 H), 7.08 (app d,
1 H), 7.20-7.35 (m, 4 H), 7.56 (app d, 1 H), 7.72 (app d, 1 H);
13C NMR (CDCl3) δ 29.1, 34.1, 43.8, 53.4, 62.0, 121.2, 123.7,
126.8, 127.3, 127.5, 128.0, 128.3, 133.4, 133.5, 133.8, 134.4,
135.3.
(-)-(R)-11-Meth yla p or p h in e (8). A mixture of 6 (233 mg,
0.61 mmol), LiCl (130 mg, 3.07 mmol), PPh3 (98 mg, 0.37
mmol), (PPh3)2PdCl2 (52 mg, 0.074 mmol), Me4Sn (0.34 mL,
2.5 mmol), and a few crystals of 2,6-di-tert-butyl-4-methylphe-
nol in dry DMF (5 mL) was stirred in a closed vessel at 120
°C for 7 h. The volatiles were evaporated in vacuo, and the
residue was partitioned between ether and 10% aqueous
NaHCO3. The ether layer was dried (K2CO3), filtered, and
concentrated. The residue was dissolved in CHCl3, filtered
through a pad of silica gel, concentrated, and chromatographed
[aluminum oxide; ether/hexanes (gradient 1:1-1:0)]. The
amine was converted into the hydrochloride salt, which was
recrystallized from MeCN to give 121 mg (70%) of pure 8‚HCl.
8: 1H NMR (CDCl3) δ 2.48-2.60 (m, 2 H), 2.55 (s, 3 H), 2.60
(s, 3 H), 2.77 (dd, 1 H), 2.92-3.06 (m, 3 H), 3.15 (ddd, 1 H),
6.97-7.17 (m, 5 H), 7.38 (app d, 1 H); 13C NMR (CDCl3) δ 22.6,

3510 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Hedberg et al.
1
12: IR (film) 2225 cm-1; H NMR (CDCl3) δ 2.47-2.65 (m,
128.1, 128.2, 128.4, 128.8, 128.9, 129.0, 129.6, 130.4, 130.5,
130.8, 131.8, 132.2, 132.4, 132.5, 132.7, 133.1, 133.5, 134.1,
134.4, 134.8, 135.8, 137.1, 138.5, 151.8, 153.4.
2 H), 2.55 (s, 3 H), 2.77 (dd, 1 H), 3.02-3.26 (m, 4 H), 7.19
(app d, 1 H), 7.29 (app t, 1 H), 7.33 (app t, 1 H), 7,49 (app d,
1 H), 7.65 (app d, 1 H), 8.11 (app d, 1 H); 13C NMR (CDCl3) δ
28.9, 34.6, 44.0, 52.9, 61.4, 108.0, 119.9, 124.7, 126.8, 127.3,
129.8, 130.3, 132.6, 133.4, 133.9, 135.1, 137.4, 137.7.
(-)-(R)-11-Met h oxy-N-p r op yln or a p or p h in e (18).1c,14
Compound 18 was synthesized from 1715 (0.34 g, 0.78 mmol)
using the procedure described for the preparation of 4.3,21 The
reaction time was 18 h. Crude 18 was chromatographed [silica
gel; CHCl3/MeOH (39:1) and ether/hexanes (1:1) followed by
aluminum oxide; ether/hexanes (1:1)]. The amine was con-
verted into the hydrochloride salt, which was recrystallized
from MeCN to yield 0.21 g (82%) of pure 18‚HCl: mp 231-
234 °C dec (lit.14 mp 227-229 °C); [R]23D -81.1° (c 1.0, MeOH)
(-)-(R)-11-Ca r b a m oyla p or p h in e (13). A mixture of
12‚HCl (74 mg, 0.25 mmol), H2O (7 mL), and concentrated H2-
SO4 (3 mL) was stirred at 120 °C for 4 h. The mixture was
partitioned between 2 M aqueous NaOH and CHCl3. The
combined organic layers were dried (MgSO4), filtered, and
concentrated. The amine was converted into the hydrochloride
salt, which was recrystallized from MeCN/ether to give 72 mg
[lit.14 [R]26
-69.9°, [R]26
-89.32° (c 0.0515, MeOH)]; 1H
578
546
(92%) of pure 13‚HCl: IR (KBr) 1664 cm-1
.
NMR (CD3OD) δ 1.10 (t, 3 H), 1.75-2.04 (m, 2 H), 2.95 (dd, 1
H), 3.04-3.32 (m, 2 H), 3.32-3.69 (m, 4 H), 3.89 (s, 3 H), 3.81-
3.92 (m, 1 H), 4.37 (app br d, 1 H), 7.01 (app d, 1 H), 7.06 (app
d, 1 H), 7.17 (app d, 1 H), 7.27 (dd, 1 H), 7.34 (dd, 1 H), 8.22
(app d, 1 H); 13C NMR (CD3OD) δ 12.1, 19.2, 27.7, 33.8, 51.0,
57.0, 57.5, 62.9, 113.7, 122.9, 123.9, 129.3, 129.7, 129.9, 130.4,
131.1, 131.5, 134.1, 136.5, 159.2.
13: 1H NMR (CDCl3) δ 2.43-2.60 (m, 2 H), 2.54 (s, 3 H),
2.73 (dd, 1 H), 3.03 (dd, 1 H), 3.07-3.24 (m, 3 H), 5.97 (s, 1
H), 6.35 (s, 1 H), 7.07 (app d, 1 H), 7.16 (app t, 1 H), 7.21 (app
t, 1 H), 7.31 (app d, 1 H), 7.41 (app d, 1 H), 7.51 (app d, 1 H);
13C NMR (CDCl3) δ 28.9, 34.8, 44.0, 52.8, 61.6, 125.8, 126.3,
127.2, 127.6, 128.5, 129.9, 131.2, 132.0, 133.0, 133.7, 134.9,
137.2, 173.6.
(-)-(R)-11-Hyd r oxy-N-p r op yln or a p or p h in e (2).1c,14
A
(-)-(R)-11-P h en yla p or p h in e (14). A solution of 6 (100
mg, 0.26 mmol) in NMP (2.5 mL) was added to a mixture of
(2-furyl)3P (12.2 mg, 0.053 mmol), LiCl (33.2 mg, 0.78 mmol),
PhSnBu3 (196 mg, 0.53 mmol), and tris(dibenzylideneacetone-
)dipalladium(0) [(dba)3Pd2; 12 mg, 0.014 mmol]. The mixture
was stirred at 100 °C for 5 h. Ether was added, the mixture
was filtered through a pad of alumina, and the volatiles were
evaporated in vacuo. The residue was chromatographed [silica
gel; EtOAc/MeOH (9:1)], and the amine was converted into the
hydrochloride, which was recrystallized from acetonitrile/
EtOH/ether to give 54 mg (60%) of pure 14‚HCl.
slurry of 18‚HCl (167 mg, 0.506 mmol) in 48% HBr (8 mL)
was refluxed for 4 h at 120 °C under nitrogen. Absolute
ethanol was added, and the volatiles were removed in vacuo.
The residue was partitioned between ether/10% aqueous
NaHCO3, and the combined organic layers were dried (Na2-
SO4), filtered, and concentrated. Crude 2 was purified by
chromatography [silica gel; CHCl3/MeOH (39:1 and 9:1)]. The
amine was converted into the hydrochloride salt, which was
recrystallized from MeCN/ether to give 123 mg (77%) of pure
2‚HCl: mp 260-263 °C dec (lit.1c mp 257-258 °C); [R]23
D
-67.4° (c 1.0, MeOH) [lit.1c [R]25 -64.0° (c 0.289, MeOH)];
578
1H NMR (CD3OD) δ 1.07 (t, 3 H), 1.71-2.07 (m, 2 H), 2.89
(dd, 1 H), 3.00-3.70 (m, 6 H), 3.81-3.97 (m, 1 H), 4.37 (app
br d, 1 H), 6.86 (app d, 1 H), 6.87 (app d, 1 H), 7.10 (dd, 1 H),
7.15 (app d, 1 H), 7.34 (dd, 1 H), 8.38 (app d, 1 H); 13C NMR
(CD3OD) δ 11.3, 18.4, 27.1, 33.0, 50.3, 56.8, 62.2, 117.2, 120.8,
121.2, 128.2, 128.8, 129.1, 129.3, 130.0, 130.5, 134.0, 135.6,
156.2.
14: 1H NMR (CDCl3) δ 2.51-2.66 (m, 2 H), 2.58 (s, 3 H),
2.75 (dd, 1 H), 3.06 (dd, 1 H), 3.10-3.33 (m, 3 H), 6.65 (d, 1
H), 6.75 (dd, 1 H), 6.91 (d, 1 H), 7.20-7.37 (m, 8 H); 13C NMR
(CDCl3) δ 29.2, 36.0, 44.1, 52.8, 62.2, 125.2, 126.7, 127.1, 127.4,
127.8, 128.4, 128.4, 129.5, 129.5, 130.6, 132.56, 132.56, 132.61,
132.7, 135.7, 137.6, 139.7, 143.0.
(-)-(R)-11-(2-Meth oxyp h en yl)a p or p h in e (15). A mix-
ture of 6 (630 mg, 1.64 mmol), LiCl (560 mg, 13 mmol), CuI
(127 mg, 0.67 mmol), (PPh3)2PdCl2 (183 mg 0.26 mmol), (2-
methoxyphenyl)SnBu3 (720 µL, 1.9 mmol), and a few crystals
of 2,6-di-tert-butyl-4-methylphenol were suspended in DMF (5
mL). The mixture was stirred at 145 °C for 3 h. A second
portion of (2-methoxyphenyl)SnBu3 (720 µL, 1.9 mmol) was
added, and the heating was continued at 145 °C for an
additional 9 h. The volatiles were evaporated, and the residue
was partioned between 10% aqueous NaHCO3 and CH2Cl2. The
combined organic extracts were dried (K2CO3), filtered, and
concentrated. The residue was chromatographed [alumina
(deactivated with 7.5% H2O); EtOAc/hexanes (5:1)]. Two
presumed atropisomers13 were separated. The main isomer
was converted into the hydrochloride, which was recrystallized
from acetonitrile/ether to give 300 mg (49%) of pure 15‚HCl.
15: 1H NMR (CDCl3) δ 2.27 (s, 3 H), 2.50-2.67 (m, 2 H),
2.60 (s, 3 H), 2.80 (dd, 1 H), 3.08 (ddd, 1 H), 3.12-3.28 (m, 3
H), 7.07-7.65 (m, 10 H); 13C NMR (CDCl3) δ 29.2, 35.2, 35.7,
43.9, 52.7, 53.0, 54.9, 55.5, 62.0, 62.2, 111.1, 111.7, 120.5, 121.3,
124.1, 125.0, 125.2, 126.1, 126.8, 126.9, 127.1, 127.2, 128.3,
128.5, 130.3, 130.4, 131.3, 131.5, 131.7, 132.0, 132.2, 132.6,
133.1, 134.2, 134.6, 135.4, 135.4, 135.6, 136.0, 137.2, 155.8,
156.8.
(-)-(R)-N-Ben zyl-11-m eth oxyn or a p or p h in e (20). Com-
pound 20 was synthesized from 1915 (5.37 g, 0.0109 mol) using
the procedure described for the preparation of 18. The reaction
time was 18 h. The crude mixture was concentrated in vacuo,
and the residue was partitioned between ether and 10%
aqueous Na2CO3. The organic layer was dried (Na2SO4),
filtered, and concentrated. Excess Bu3N was removed by
vacuum distillation at 80 °C and 2 mmHg. The crude mixture
was chromatographed [aluminum oxide; ether/hexanes (1:8-
1:3) and silica gel; toluene to toluene/MeCN (20:1)] to give 3.60
g of pure 20 and 0.16 g of the oxidized analogue N-benzyl-
6a,7-didehydro-11-methoxynoraporphine (26). The amine 20
was converted into the hydrochloride salt, which was recrys-
tallized from MeCN/ether to give 3.52 g (85%) of pure
20‚HCl: 1H NMR (CD3OD) δ 2.97-3.14 (m, 2 H), 3.18-3.65
(m, 2 H), 3.64-3.82 (m, 2 H), 3.88 (s, 3 H), 4.42 (m, 2 H), 5.07
(br d, 1 H), 7.07 (app d, 1 H), 7.09 (app d, 1 H), 7.15 (app d, 1
H), 7.31 (dd, 1 H), 7.34 (dd, 1 H), 7.49-7.56 (m, 3 H), 7.58-
7.65 (m, 2 H), 8.21 (app d, 1 H); 13C NMR (CD3OD) δ 26.9,
33.2, 49.5, 56.2, 58.9, 62.4, 112.8, 122.1, 123.1, 128.5, 128.9,
129.1, 129.6, 130.2, 130.4, 130.5, 130.7, 131.4, 132.8, 133.3,
135.7, 158.3.
Compound 26 was chromatographed [aluminum oxide;
ether/hexanes (1:8-1:1)] and recrystallized (ether) to give 0.14
g (3.8%) of pure 26: mp 126-127 °C; 1H NMR (CDCl3) δ 3.29
(app dd, 2 H), 3.47 (app dd, 2 H), 4.05 (s, 3 H), 4.65 (s, 2 H),
6.72 (app s, 1 H), 6.83 (dd, 1 H), 7.12-7.41 (m, 8 H), 7.53 (dd,
1 H), 9.57 (app d, 1 H);
(-)-(R)-11-(2-Hyd r oxyp h en yl)a p or p h in e (16). A mix-
ture of 15‚HCl (90 mg, 0.24 mmol) and 48% aqueous HBr (10
mL) kept under nitrogen was stirred at 110 °C. After 2 h the
volatiles were evaporated, and the residue was partitioned
between CH2Cl2 and 10% aqueous NaHCO3. The combined
organic extracts were dried (MgSO4), filtered, and concen-
trated. The residue was chromatographed [silica gel; CHCl3/
MeOH (9:1)], and the amine was converted into the hydro-
chloride, which was recrystallized from 2-PrOH/ether to give
73 mg (97%) of pure 16‚HCl.
13C NMR (CDCl3) δ 31.3, 48.1, 55.6,
55.9, 102.9, 104.7, 115.2, 119.8, 123.7, 124.3, 125.9, 126.5,
126.6, 126.9, 127.3, 128.6, 131.5, 132.2, 136.2, 138.4, 142.5,
158.7. Anal. (C24H21NO) C, H, N.
(-)-(R)-N-Ben zyl-11-h yd r oxyn or a p or p h in e (21). Com-
pound 21 was synthesized from 20‚HCl (0.60 g, 1.6 mmol)
using the procedure described for the preparation of 2‚HCl.
The reaction time was 6 h. Crude 21‚HCl was partitioned
between CH2Cl2/EtOH (9:1) and 10% aqueous NaHCO3. The
organic layer was dried (Na2SO4), filtered, and concentrated.
16: 1H NMR (CDCl3) δ 2.44-2.80 (m, 3 H), 2.54 (s, 3 H),
2.97 (dd, 1 H), 3.03-3.30 (m, 3 H), 6.58-7.50 (m, 10 H); 13C
NMR (CDCl3) δ 29.0, 35.3, 35.6, 44.0, 52.76, 52.79, 61.9, 62.0,
115.9, 120.5, 121.1, 124.8, 125.6, 125.8, 126.8, 127.2, 127.5,

11-Substituted (R)-Aporphines
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3511
The residue was chromatographed [silica gel; CHCl3 to CHCl3/
MeOH (39:1)] to give 0.050 g of recovered 20 and pure 21. The
amine 21 was converted into the hydrochloride salt, which was
recrystallized from MeOH to give 0.39 g (total yield 77%) of
pure 21‚HCl: 1H NMR (CD3OD, 50 °C) δ 2.94-3.44 (m, 4 H),
3.64-3.83 (m, 2 H), 4.36-4.51 (m, 2 H), 5.06 (br d, 1 H), 6.90
(app d, 1 H), 6.93 (app d, 1 H), 7.13 (app d, 1 H), 7.14 (dd, 1
H), 7.36 (dd, 1 H), 7.50-7.64 (m, 5 H), 8.41 (app d, 1 H); 13C
NMR (CD3OD, 50 °C) δ 27.0, 33.5, 50.0, 59.3, 62.9, 117.5, 120.8,
128.2, 129.0, 129.3, 130.15, 130.24, 130.5, 130.7, 131.6, 132.7,
134.0, 135.5, 156.4.
(-)-(R)-11-Meth oxyn or a p or p h in e (22). A mixture of
20‚HCl (1.12 g, 2.96 mmol) and Pd(C) (10%, 0.20 g) in HOAc
(30 mL) was stirred under H2 (1 atm). After stirring for 4 h,
the volatiles were removed in vacuo. Crude 22‚HCl was
purified by recrystallization from MeCN and MeOH/ether to
give 0.75 g (88%) of pure 22‚HCl: 1H NMR (CD3OD) δ 2.93
(dd, 1 H), 3.04-3.19 (m, 2 H), 3.25-3.48 (m, 2 H), 3.70-3.79
(m, 1 H), 3.90 (s, 3 H), 4.39 (dd, 1 H), 6.96 (app d, 1 H), 7.06
(app d, 1 H), 7.18 (app d, 1 H), 7.27 (dd, 1 H), 7.34 (dd, 1 H),
8.24 (app d, 1 H); 13C NMR (CD3OD) δ 26.3, 35.0, 42.5, 54.3,
56.2, 112.8, 121.9, 123.2, 128.8, 128.9, 129.1, 129.6, 130.3,
130.8, 133.0, 135.6, 158.6.
(-)-(R)-11-Hyd r oxyn or a p or p h in e (23). Compound 23
was prepared from 21‚HCl (0.17 g, 0.47 mmol) using the
procedure for the synthesis of 22‚HCl. Absolute EtOH (40 mL)
was added to dissolve the starting material. The reaction time
was 6 h. Crude 23‚HCl was purified by recrystallization from
MeOH/ether to give 0.12 g (94%) of pure 23‚HCl: 1H NMR
(CD3OD) δ 2.92 (dd, 1 H), 3.03-3.17 (m, 2 H), 3.13-3.48 (m,
2 H), 3.69-3.81 (m, 1 H), 4.39 (dd, 1 H), 6.82 (app d, 1 H),
6.86 (app d, 1 H), 7.10 (dd, 1 H), 7.15 (app d, 1 H), 7.34 (dd, 1
H), 8.40 (app d, 1 H); 13C NMR (CD3OD) δ 26.4, 35.2, 42.7,
54.6, 117.4, 120.7, 121.5, 128.4, 128.9, 129.2, 129.3, 130.0,
130.7, 133.7, 135.6, 156.5.
(-)-(R)-N-Isop r op yl-11-m et h oxyn or a p or p h in e (24).
Compound 24 was synthesized from 22 according to a method
previously described by Liu et al.24 A suspension of 22 (0.31
g, 1.2 mmol), K2CO3 (0.19 g, 1.4 mmol), diisopropylethylamine
(0.23 mL, 1.4 mmol), and 2-iodopropane (0.14 mL, 1.4 mmol)
in MeCN (1.5 mL) was stirred at 70 °C for 89 h. The crude
mixture was diluted with CH2Cl2, filtered, and concentrated.
The residue was chromatographed [aluminum oxide; ether/
hexanes (1:8)], and the amine was converted into the hydro-
chloride salt. Recrystallization from MeCN/ether gave 0.35 g
(86%) of pure 24‚HCl: 1H NMR (CD3OD) δ 1.33 (d, 3 H), 1.55
(d, 3 H), 2.93 (dd, 1 H), 3.08-3.33 (m, 2 H), 3.33-3.52 (m, 2
H), 3.75-3.88 (m, 1 H), 3.88 (s, 3 H), 4.30 (sept, 1 H), 4.49
(dd, 1 H), 7.01 (app d, 1 H), 7.05 (app d, 1 H), 7.18 (app d, 1
H), 7.27 (dd, 1 H), 7.34 (dd, 1 H), 8.21 (app d, 1 H); 13C NMR
(CD3OD) δ 14.2, 18.7, 27.3, 32.6, 43.5, 54.6, 56.2, 60.0, 112.8,
122.2, 123.2, 128.6, 128.9, 129.0, 130.1, 130.3, 131.1, 133.5,
135.6, 158.3.
SO4), filtered, and concentrated. The residue was suspended
in a mixture of absolute EtOH and 1 M aqueous H2SO4 (1:3).
Zink dust (0.35 g, 5.4 mmol) was added, and the mixture was
stirred at 75 °C for 42 h. Crude 27 was purified by preparative
TLC [silica gel; toluene/MeCN (10:1)]. The amine was con-
verted into the hydrochloride salt, which was recrystallized
from MeOH/ether to give 18 mg of recovered 26 and 58 mg of
pure 27‚HCl (total yield 91%): mp 232-236 °C. Anal. (C24H24
-
ClNO) C, H, N. Racemic 27‚HCl was used for determination
of the optical purity of 22 as described below.
Deter m in a tion of Op tica l P u r ity of Com p ou n d 22. The
optical purity of 22 was determined as follows: 22‚HCl (32
mg, 85 mmol) was partitioned between CH2Cl2 and 10%
aqueous Na2CO3, and the organic layer was dried (K2CO3),
filtered, and concentrated. The residue was stirred with (R)-
O-methylmandelic chloride as described previously20 to give
29. Compound 28 was synthesized from 27‚HCl (25 mg, 66
mmol) using the procedure for the synthesis of 22. Treatment
of 28 with (R)-O-methylmandelic chloride as described previ-
ously20 afforded a 1:1 mixture of 29 and 30 (capillary GC). The
enantiomeric excess of 22 was determined to be >99.8% ee,
which was shown by comparison with a sample of 29 to which
0.2% of the 1:1 mixture of 29 and 30 had been added (capillary
GC).
Com p u ta tion a l Meth od s. MacMimic version 2.1 (InStar
Software, IDEON Research Park, S-233 70 Lund, Sweden) was
used as graphical interface for MM2(91).22 Torsion angle
drivers on aromatic and N-alkyl substituents were performed
from 0° to 360° with a 10° increment and full energy minimi-
zation except for the dihedral angles used as driving angle(s).
The resulting local minima were built and subjected to
unrestricted energy minimization. The most stable conforma-
tion of the aporphine skeleton was used having a pseudoequa-
torial N-alkyl substituent and adopting a half-chair confor-
mation of the tetrahydropyridine ring with P-helicity. Con-
formations with P-helicity and a pseudoaxial N-substituent,
M-helicity and a pseudoequatorial N-substituent, and M-
helicity with a pseudoaxial N-substituent have relative steric
energies of 0.8, 3.1, and 3.6 kcal/mol, respectively.15 Com-
pounds with a C11-OH-substituent were modeled with C10-
C11-O-H in an eclipsed orientation as selected previously
for H-bonding to active site residues.3
Active Site Defin ition . The definition of the 5-HT1A
receptor active site has been described elsewhere.15 Amino
acids having atoms within a 10 Å sphere around the centroid
of the CR atoms of Asp114, Ser193, and Phe390 were consid-
ered in the initial active site definition for the D2A receptor
model. A systematic search around ø(1) and ø(2) of the Asp
residue and around ø(1) of the Ser residue was performed in
Sybyl 6.0.3 to probe allowed orientations of these functional
groups (resolution 15°, default VDW cut-off values). The
resulting sites were filtered to retain those that were compat-
ible with the distance between the ligand -OH and -N+-H
functional groups. The ligand N+-H was therefore extended
to a distance of 2.6 Å, and the ligand OH was extended to 2.75
Å. The distance between these extended points in 1 was 10.16
Å, and the sites selected from the systematic search output
were those that had the Asp-O and the Ser-O at distances
between 9.91 and 10.41 Å. Compound 1 was fitted to the
Asp-O and Ser-O using the N+-H and OH extended points
defined above. For the third fitting point a normal was defined
through the A-ring extending 6 Å toward the aromatic ring of
Phe390. The end point of this normal was fitted onto the
centroid of Phe390. The weight of this last fitting point was
1/10th of the weight of the other two points. The rms value of
the fit was used together with the common volume of ligand
and active site residues to select complexes for futher optimi-
zation.
(-)-(R)-11-H yd r oxy-N-isop r op yln or a p or p h in e
(25).
Compound 25 was synthesized from 24‚HCl (0.17 g, 0.52
mmol) using the procedure for the synthesis of 2‚HCl. The
reaction time was 4 h and 30 min. Crude 25‚HBr was
partitioned between CH2Cl2/EtOH (5:1) and 10% aqueous
NaHCO3. The organic layer was dried (Na2SO4), filtered, and
concentrated. The amine was converted into the hydrochloride
salt, which was recrystallized from MeCN/MeOH followed by
MeOH/ether to give 0.15 g (92%) of pure 25‚HCl: 1H NMR
(CD3OD) δ 1.35 (d, 3 H), 1.54 (d, 3 H), 2.87 (dd, 1 H), 3.13-
3.40 (m, 3 H), 3.46 (dd, 1 H), 3.78-3.89 (m, 1 H), 4.32 (sept, 1
H), 4.59 (dd, 1 H), 6.88 (app d, 2 H), 7.12 (dd, 1 H), 7.18 (app
d, 1 H), 7.36 (dd, 1 H), 8.40 (app d, 1 H); 13C NMR (CD3OD) δ
14.1, 18.7, 27.3, 32.7, 43.6, 54.5, 60.1, 117.2, 120.8, 121.2, 128.2,
128.7, 129.1, 129.8, 130.0, 130.9, 134.1, 135.4, 156.1.
(()-N-Ben zyl-11-m eth oxyn or a p or p h in e (27).1a Com-
pound 27 was prepared from 26 according to a method
previously described by Gerecke et al.19 A suspension of 26
(71 mg, 0.21 mmol) and zink (0.20 g, 3.1 mmol) in 1 M aqueous
H2SO4 (5 mL) was stirred at 75 °C for 29 h. Due to lack of
further progress of the reaction, it was restarted using the
following procedure: The mixture was partitioned between
CHCl3 and 10% NaHCO3, and the organic layer was dried (Na2-
Min ir ecep tor Mod elin g. Active site optimization was
performed in Yak version 3.9 (SIAT Biographics Laboratory,
Missionsstrasse 60, CH-4055 Basel, Switzerland). Atomic
charges of ligands transferred to Yak were obtained from a
single-point MNDO calculation (MOPAC version 5.0, imple-
mented in Sybyl 6.0.3, Tripos Associates Inc., 1699 S. Hanley
Rd., Suite 303, St. Louis, MO). The necessary free energies
of binding and ligand solvation were obtained from the Ki

3512 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Hedberg et al.
600-607. (f) Gao, Y.; Baldessarini, R. J .; Kula, N. S.; Neumeyer,
J . L. Synthesis and dopamine receptor affinities of enantiomers
of 2-substituted apomorphines and their N-n-propyl analogues.
J . Med. Chem. 1990, 33, 1800-1805.
values and the program ESOLV (supplied with Yak and based
on an algorithm developed by Still et al.40), respectively.
Minimizations started with 25 steps of the steepest-descent
minimizer followed by 25 steps of the conjugate gradient
minimizer. When convergence was not achieved after these
50 steps, a new 50-step minimization was performed.
(2) (a) Martin, L. P.; Cox, R. F.; Waszczak, B. L. Efficacy and potency
comparisons among aporphine enantiomers: Effects on dopam-
ine neurons in Substantia Nigra of rat. Neuropharmacology
1990, 29, 135-143. (b) Baldessarini, R. J .; Kula, N. S.; Zong,
R.; Neumeyer, J . L. Receptor affinities of aporphine enantiomers
in rat brain tissue. Eur. J . Pharmacol. 1994, 254, 199-203.
(3) Hedberg, M. H.; J ohansson, A. M.; Nordvall, G.; Yliniemela¨, A.;
Li, H. B.; Martin, A. R.; Hjorth, S.; Unelius, L.; Sundell, S.;
Hacksell, U. (R)-11-Hydroxy- and (R)-11-hydroxy-10-methyl-
aporphine: synthesis, pharmacology, and modeling of D2A and
5-HT1A receptor interactions. J . Med. Chem. 1995, 38, 647-658.
(4) Saari, W. S.; King, S. W.; Lotti, V. J .; Scriabine, A. Synthesis
and biological activity of some aporphine derivatives related to
apomorphine. J . Med. Chem. 1974, 17, 1086-1089.
(5) (a) Cannon, J . G.; Mohan, P.; Bojarski, J .; Long, J . P.; Bhatnagar,
R. K.; Leonard, P. A.; Flynn, J . R.; Chatterjee, T. K. (R)-(-)-
10-Methyl-11-hydroxyaporphine: a highly selective serotonergic
agonist. J . Med. Chem. 1988, 31, 313-318. (b) Cannon, J . G.;
Moe, S. T.; Long, J . P. Enantiomers of 11-hydroxy-10-methyl-
aporphine having opposing pharmacological effects at 5-HT1A
receptors. Chirality 1991, 3, 19-23. (c) Cannon, J . G.; Raghupati,
R.; Moe, S. T.; J ohnson, A. K.; Long, J . P. Preparation and
pharmacological evaluation of enantiomers of certain nonoxy-
genated aporphines: (+)- and (-)-aporphine (+)- and (-)-10-
methylaporphine. J . Med. Chem. 1993, 36, 1316-1318. (d)
Cannon, J . G.; Flaherty, P. T.; Ozkutlu, U.; Long, J . P. A-ring
ortho-disubstituted aporphine derivatives as potential agonists
or antagonists at serotonergic 5-HT1A receptors. J . Med. Chem.
1995, 38, 1841-1845.
(6) Cacchi, S.; Ciattini, P. G.; Morera, E.; Ortar, G. Palladium-
catalyzed triethylammonium formate reduction of aryl triflates.
A selective method for the deoxygenation of phenols. Tetrahedron
Lett. 1986, 27, 5541-5144.
(7) Saa´, J . M.; Martorell, G.; Garc´ıa-Raso, A. Palladium-catalyzed
cross-coupling reactions of highly hindered, electron-rich phenol
triflate and organostannanes. J . Org. Chem. 1992, 57, 678-685.
(8) Gadamer, J .; Oberlin, M.; Schoeler, A. Die Synthese des Apor-
phins. (Synthesis of apomorphine.) Arch. Pharm. (Weinheim,
Ger.) 1925, 263, 81-99.
(9) Saa´, J . M.; Dopico, M.; Martorell, G.; Garc´ıa-Raso, A. Deoxy-
genation of highly hindered phenols. J . Org. Chem. 1990, 55,
991-995.
(10) Chambers, M. R. I.; Widdowson, D. A. Nickel catalysed conver-
sion of phenol triflates into aromatic nitriles and acids. J . Chem.
Soc., Perkin Trans. I 1989, 1365-1366.
(11) Farina, V.; Krishnan, B.; Marshall, D. R.; Roth, G. P. Palladium-
catalyzed coupling of arylstannanes with organic sulfonates: a
comprehensive study. J . Org. Chem. 1993, 58, 5334-5444.
(12) Saa´, J . M.; Martorell, G. Palladium-catalyzed cross-coupling
synthesis of hindered biaryls and terphenyls. Cocatalysis by
copper(I) salts. J . Org. Chem. 1993, 58, 1963-1966.
(13) Oki, M. Recent advances in atropisomerism. Top. Stereochem.
1980, 14, 1-81.
(14) Ram, V. J .; Neumeyer, J . L. Aporphines. 42. Synthesis of (R)-
(-)-11-hydroxyaporphines from morphine. J . Org. Chem. 1982,
47, 4372-4374.
(15) Hedberg, M. H.; J ansen, J . M.; Nordvall, G.; Hjorth, S.; Unelius,
L.; J ohansson, A. M. 10-Substituted 11-Oxygenated (R)-Apor-
phines: Synthesis, Pharmacology, and Modeling of 5-HT1A
Receptor Interactions. J . Med. Chem. 1996, 39, 3491-3502.
(16) (a) Cava, M. P.; Edie, D. L.; Saa´, J . M. Direct dehydrogenation
of aporphine alkaloids. J . Org. Chem. 1975, 40, 3601-3602. (b)
Davis, P. J .; Seyhan, S.; Soine, W.; Smith, R. V. Convenient
synthesis of (S)-(+)-apomorphine from (R)-(-)-apomorphine. J .
Pharm. Sci. 1980, 69, 1056-1058.
(17) (a) Murahashi, S.-I.; Hirano, T.; Yano, T. Palladium catalyzed
amine exchange reaction of tertiary amines. Insertion of pal-
ladium(0) into carbon-hydrogen bonds. J . Am. Chem. Soc. 1978,
100, 348-350. (b) Murahashi, S.-I.; Watanabe, T. Palladium
catalyzed hydrolysis of tertiary amines with water. J . Am. Chem.
Soc. 1979, 101, 7429-7430.
An active site model for the D2A receptor was derived from
the systematically generated sites in a stepwise manner using
the high-affinity ligands 1 and 2. Prior to optimization, two
spheres were defined around the ligand of interest by selecting
residues that had atoms within 5 and 7.5 Å of ligand atoms,
respectively. All residues within the 7.5 Å sphere were
transferred to Yak where the residues in the inner 5 Å sphere
were optimized; the additional residues in the outer shell
served to maintain the overall shape of the binding site. The
active site complex resulting from Yak optimization of 1 was
used as the starting point for docking 2 by fitting that ligand
onto the optimized position of 1 (using the nitrogen, the C11-
oxygen and the A-ring centroid). The model was obtained by
optimization of 1 and 2 simultaneously using 3 and 5 Å
spheres, respectively. The 5-HT1A15 and D2A models were used
to dock the different conformations of selected compounds,
grouped into sets that differ only in the substitution on one
position of the aporphine skeleton. The 5 and 3 Å spheres were
defined around the whole set of those ligands fitted onto 1 (D2A
)
and either 5 or 14 (5-HT1A). The resulting site was used to
optimize the orientation of the individual ligands together with
the torsional degrees of freedom of the amino acids in the 3 Å
sphere in Yak.
P h a r m a cology. Recep tor Bin d in g Assa ys. The sero-
3,15
41
42
tonin 5-HT1A
and the dopamine D2A and D1 receptor
binding assays were performed as described previously, using
[3H]-8-OH-DPAT‚HBr, [3H]raclopride, and [3H]SCH23390,
respectively, as radioligands.
Bioch em istr y a n d Beh a vior . The monitoring and scoring
of the behavior produced by the compounds and the estimation
of their effects on central 5-HT, DA, and NA synthesis rates
by measuring the accumulation of 5-hydroxytryptophan (5-
HTP) and 3,4-dihydroxyphenylalanine (DOPA), respectively,
were performed in reserpinized rats as previously described.15,34b
Ack n ow led gm en t. Professor Uli Hacksell is ac-
knowledged for helpful discussions and constructive
criticism of the manuscript. Skillful technical as-
sistance by Anne-Marie Eriksson and Gun Torell-
Svantesson in the binding assays, by Gerd Leonsson in
the biochemical studies, and by Reza Karimi in the
synthetic work is gratefully acknowledged. Financial
support was obtained from the Swedish National Board
for Industrial and Technical Development (NUTEK),
Astra Arcus AB, and the Swedish Medical (No. 7486)
Research Council.
Su p p or tin g In for m a tion Ava ila ble: Assigned 1H NMR
data (4 pages). Ordering information is given on any current
masthead page.
Refer en ces
(1) (a) Neumeyer, J . L.; Granchelli, F. E.; Fuxe, K.; Ungerstedt, U.;
Corrodi, H. Aporphines. 11. Synthesis and dopaminergic activity
of monohydroxyaporphines. Total synthesis of (()-11-hydroxyapor-
phine, (()-11-hydroxynoraporphine, and (()-11-hydroxy-N-n-
propylnoraporphine. J . Med. Chem. 1974, 17, 1090-1095. (b)
Neumeyer, J . L.; Baldessarini, R. J .; Arana, G. W.; Campbell,
A. Aporphines as dopamine agonists and antagonists at central
dopamine receptors. New Methods Drug Res. 1985, 1, 153-166.
(c) Gao, Y.; Zong, R.; Campbell, A.; Kula, N. S.; Baldessarini, R.
J .; Neumeyer, J . L. Synthesis and dopamine agonist and
antagonist effects of (R)-(-)- and (S)-(+)-11-hydroxy-N-n-pro-
pylnoraporphine. J . Med. Chem. 1988, 31, 1392-1396. (d) Gao,
Y.; Ram, V. J .; Campbell, A.; Kula, N. S.; Baldessarini, R. J .;
Neumeyer, J . L. Synthesis and structural requirements of
N-substituted norapomorphines for affinity and activity at
dopamine D-1, D-2, and agonist receptor sites in rat brain. J .
Med. Chem. 1990, 33, 39-44. (e) Schaus, J . M.; Titus, R. D.;
Foreman, M. M.; Mason, N. R.; Truex, L. L. Aporphines as
antagonists of dopamine D-1 receptors. J . Med. Chem. 1990, 33,
(18) Cava, M. P.; Havlicek, S. C.; Lindert, A.; Spangler, R. J . A
photochemical aporphine synthesis. Tetrahedron Lett. 1966, 7,
2937-2939.
(19) Gerecke, M.; Borer, R.; Brossi, A. Conversion of natural (S)-
bulbocapnine into two (ring-A)-substituted derivatives of (R)-
(-)-apomorphine. Helv. Chim. Acta 1979, 62, 1543-1548.
(20) J ohansson, A. M.; Fredriksson, K.; Hacksell, U.; Grol, C. J .;
Svensson, K.; Carlsson, A.; Sundell, S. Synthesis and pharma-
cology of the enantiomers of cis-7-hydroxy-3-methyl-2-(dipropy-
lamino)tetralin. J . Med. Chem. 1990, 33, 2925-2929; corrections
J . Med. Chem. 1991, 34, 1517.
(21) Hedberg, M. H.; J ohansson, A. M.; Hacksell, U. Facile syntheses
of aporphine derivatives. J . Chem. Soc., Chem. Commun. 1992,
845-846.

11-Substituted (R)-Aporphines
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3513
(22) (a) Allinger, N. L.; Yuh, Y. Quantum Chemistry Program
Exchange 1980, 12, 395. (b) Burkert, U.; Allinger, N. L. Molec-
ular Mechanics; American Chemical Society: Washington, DC,
1982.
(23) Malmberg, A° .; Nordvall, G.; J ohansson, A. M.; Mohell, N.;
Hacksell, U. A molecular basis for the binding of 2-aminotetra-
lines to human dopamine D2A and D3 receptors. Mol. Pharmacol.
1994, 46, 299-312.
(24) Liu, Y.; Yu, H.; Mohell, N.; Nordvall, G.; Lewander, T.; Hacksell,
U. Derivatives of cis-2-amino-8-hydroxy-1-methyltetralin: Mixed
5-HT1A-receptor agonists and dopamine D2-receptor antagonists.
J . Med. Chem. 1995, 38, 150-160.
(25) Nordvall, G.; Hacksell, U. Binding-site modeling of the musca-
rinic m1 receptor: A combination of homology-based and indirect
approaches. J . Med. Chem. 1993, 36, 967-976.
(26) (a) Vedani, A.; Zbinden, P.; Snyder, J . P. Pseudo-receptor
modeling: a new concept for the three-dimensional construction
of receptor binding sites. J . Rec. Res. 1993, 13, 163-177. (b)
Snyder, J . P.; Rao, S. N.; Koehler, K. F.; Vedani, A. Minireceptors
and pseudoreceptors. In 3D QSAR in drug design; Kubinyi, H.,
Ed.; ESCOM Science Publishers B.V.: Leiden, The Netherlands,
1993; pp 336-354. (c) Vedani, A.; Zbinden, P.; Snyder, J . P.;
Greenidge, P. A. Pseudoreceptor modeling: The construction of
three-dimensional receptor surrogates. J . Am. Chem. Soc. 1995,
117, 4987-4994.
(27) (a) Vedani, A.; Dunitz, J . D. Lone-pair directionality in hydrogen
bond potential functions for molecular mechanics calculations:
the inhibition of human carbonic anhydrase II by sulfonamides.
J . Am. Chem. Soc. 1985, 107, 7653-7658. (b) Vedani, A.; Huhta,
D. W. A new force field for modeling metalloproteins. J . Am.
Chem. Soc. 1990, 112, 4759-4767.
(28) (a) Ande´n, N.-E.; Carlsson, A.; Ha¨ggendal, J . Adrenergic mech-
anisms. Annu. Rev. Pharmacol. 1969, 9, 119-134. (b) Aghaja-
nian, G. K. Feedback regulation of central monoaminergic
neurons: evidences from single cell recording studies. In Essays
in neurochemistry and neuropharmacology; Youdim,, M. B. H.,
Lovenberg, W., Sharman, D. F., Lagnado, J . R., Eds.; J ohn Wiley
and Sons: New York, 1978; pp 2-32. (c) Hjorth, S.; Magnusson,
T. The 5-HT1A agonist 8-OH-DPAT preferentially activates cell
body 5-HT autoreceptors in rat brain in vivo. Naunyn-Schmiede-
berg’s Arch. Pharmacol. 1988, 338, 463-471. (d) Neckers, L. M.;
Neff, N. H.; Wyatt, R. J . Increased serotonin turnover in corpus
striatum following an injection of kainic acid: evidence for
neuronal feedback regulation of synthesis. Naunyn-Schmiede-
berg’s Arch. Pharmacol. 1979, 306, 173-177.
droxylase activities in brain in vivo using an inhibitor of the
aromatic amino acid decarboxylase. Naunyn-Schmiedeberg’s
Arch. Pharmacol. 1972, 275, 153-168.
(32) Arvidsson, L.-E.; Hacksell, U.; Nilsson, J . L. G.; Hjorth, S.;
Carlsson, A.; Lindberg, P.; Sanchez, D.; Wikstro¨m, H. 8-Hydroxy-
2-(dipropylamino)tetralin, a new centrally acting 5-hydroxy-
tryptamine receptor agonist. J . Med. Chem. 1981, 24, 921-923.
(33) Hjorth, S.; Carlsson, A.; Lindberg, P.; Sanchez, D.; Wikstro¨m,
H.; Arvidsson, L.-E.; Hacksell, U.; Nilsson, J . L. G. 8-Hydroxy-
2-(dipropylamino)tetralin, 8-OH-DPAT, a potent and selective
ergot congener with central 5-HT-receptor stimulating activity.
J . Neural Transm. 1982, 55, 169-188.
(34) (a) Tricklebank, M.; Forler, C.; Fozard, J . R. The involvement
of subtypes of the 5-HT1 receptor and of catecholaminergic
systems in the behavioural response to 8-hydroxy-2-(di-n-pro-
pylamino)tetralin in the rat. Eur. J . Pharmacol. 1984, 106, 271-
282. (b) Hjorth, S.; Sharp, T. Mixed agonist/antagonist properties
of NAN-190 at 5-HT1A receptors: behavioural and in vivo brain
microdialysis studies. Life Sci. 1990, 46, 955-963.
(35) O¨ gren, S. O.; Hall, H.; Ko¨hler, C.; Magnusson, O.; Lindbom, L.
O.; A¨ ngeby, K.; Florvall, L. Remoxipride,
a new potential
antipsychotic compound with selective antidopaminergic actions
in the rat brain. Eur. J . Pharmacol. 1984, 102, 459-474.
(36) Walters, J . R.; Bergstro¨m, D. A.; Carlsson, J . H.; Chase, T. N.;
Braun, A. R. D1 dopamine receptor activation required for
postsynaptic expression of D2 agonist effects. Science 1987, 236,
719-722.
(37) Liu, Y.; Yu, H.; Svensson, B. E.; Cortizo, L.; Lewander, T.;
Hacksell, U. Derivatives of 2-(dipropylamino)tetralin: Effect of
the C8-substituent on the interaction with 5-HT1A receptors. J .
Med. Chem. 1993, 36, 4221-4229.
(38) Yu, H. Thesis, Uppsala University, Uppsala, Sweden, 1996;
ISBN 91-544-3765-6.
(39) (a) Glaser, R.; Bernstein, M. A. 1H and 13C NMR studies on the
conformation of N-methyl diastereomers of (+)-glaucine hydro-
trifluoroacetate, an aporphine alkaloid salt. J . Chem. Soc.,
Perkin Trans. 2 1991, 2047-2053. (b) Maryanoff, B. E.; McCom-
sey, D. F.; Gardocki, J . F.; Shank, R. P.; Costanzo, M. J .; Nortey,
S. O.; Schneider, C. R.; Setler, P. E. Pyrroloisoquinoline anti-
depressants. 2. In-depth exploration of structure-activity rela-
tionships. J . Med. Chem. 1987, 30, 1433-1454.
(40) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T.
Semianalytical treatment of solvation for molecular mechanics
and dynamics. J . Am. Chem. Soc. 1990, 112, 6127-6129.
(41) Malmberg, A° .; J ackson, D. M.; Eriksson, A.; Mohell, N. Unique
binding characteristics of antipsychotic agents interacting with
human dopamine D2A, D2B, and D3 receptors. Mol. Pharmacol.
1993, 43, 749-754.
(29) Carlsson, A.; Lindqvist, M. Effect of ethanol on the hydroxylation
of tyrosine and tryptophan in rat brain in vivo. J . Pharm.
Pharmacol. 1973, 25, 437-440.
(30) Shum, A.; Sole, M. J .; van Loon, G. R. Simultaneous measure-
ment of 5-hydroxytryptophan and L-dihydroxyphenylalanine by
high performance liquid chromatography with electrochemical
detection. Measurement of serotonin and catecholamine turnover
in discrete brain regions. J . Chromatogr. 1982, 228, 123-130.
(31) Carlsson, A.; Davis, J . N.; Kehr, W.; Lindqvist, M.; Atack, C. V.
Simultaneous measurement of tyrosine and tryptophan hy-
(42) Hall, H.; Sa¨llemark, M.; J erning, E. Effects of remoxipride and
some related new substituted salicylamides on rat brain recep-
tors. Acta Pharmacol. Toxicol. 1986, 58, 61-70.
J M960189I