Synthesis of supposed enone prodrugs of apomorphine and N-propyl-norapomorphine

Article abstract of DOI:10.1016/j.tet.2007.04.103

We have previously demonstrated that the enone prodrug GMC-6650 acts as a highly efficient dopaminergic agonist. In vivo, this compound is bioactivated to its corresponding catecholamine, TL-334. The goal here was to investigate if this bioactivation also occurs for the supposed enone prodrug of apomorphine. We describe the 12-step synthesis of this supposed prodrug, 6-alkyl-5,6,6a,8,9,10-hexahydro-4H-dibenzo[de,g]quinolin-11(7H)-one (R=Me, n-Pr).

Full text of DOI:10.1016/j.tet.2007.04.103

Tetrahedron 63 (2007) 7264–7270
Synthesis of supposed enone prodrugs of apomorphine
and N-propyl-norapomorphine
*
Danyang Liu, Bastiaan J. Venhuis, H a˚ kan V. Wikstr €o m and Durk Dijkstra
Department of Medicinal Chemistry, University Centre for Pharmacy, University of Groningen,
Antonius Deusinglaan 1, NL-9713 AV Groningen, The Netherlands
Received 11 December 2006; revised 22 March 2007; accepted 12 April 2007
Available online 22 May 2007
Abstract—We have previously demonstrated that the enone prodrug GMC-6650 acts as a highly efficient dopaminergic agonist. In vivo, this
compound is bioactivated to its corresponding catecholamine, TL-334. The goal here was to investigate if this bioactivation also occurs for the
supposed enone prodrug of apomorphine. We describe the 12-step synthesis of this supposed prodrug, 6-alkyl-5,6,6a,8,9,10-hexahydro-4H-
dibenzo[de,g]quinolin-11(7H)-one (R¼Me, n-Pr).
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
R
N
Apomorphine (1) was first synthesized from morphine in
869. In the past, it was used as an emetic compound and
1
HO
1
OH
it is now clinically used in Parkinson’s disease (PD), in par-
ticular as a rescue compound at the ‘off’ state during the use
of L-dopa. Apomorphine is known as a dopamine D /D
2
1 R=CH
3
1
2
2
R=CH2CH2CH3
agonist. N-Propyl-norapomorphine (NPA, 2) is also a D1/
D agonist with a more potent D agonistic effect than apo-
morphine (Fig. 1).
2
2
Figure 1. The structures of apomorphine (APO, 1) and N-propyl-norapo-
morphine (NPA, 2).
3
The catecholamine moiety of apomorphine and NPA oxi-
dizes rapidly in vitro and in vivo, resulting in poor pharma-
cokinetic properties. Many efforts have been made to find
new apomorphine structures and analogs that circumvent
be interesting prodrugs with clinical potential against PD.
Here, we report the synthesis of enone compounds 21a,b.
4
–6
these bioavailability problems.
We have developed
a new type of prodrug, which contains a cyclohexenone–
ethylamine moiety as the basic structural element. In vivo,
the enone structure can be bioactivated to the corresponding
catecholamine. Figure 2 shows a typical example of a do-
pamine agonist S-(ꢀ)-5,6-di-OH-DPAT (3) and its corre-
sponding enone prodrug PD148903 (4). We found that 3 is
formed in vivo after the administration of 4. GMC-6650
2
. Results and discussion
In order to prepare this type of challenging compound, retro-
synthetic analysis of the aporphines was done to design
a synthetic route (Scheme 1).
7
,8
In this strategy, an important transformation will be alkyl-
ation at the a-position of ketone (14). This tricyclic ring
system (B, C and D rings) will be the key intermediate in
the synthetic route, as the functional carbonyl group allows
alkylation at the a-position. These benzo[de]quinolines can
be prepared from 1-methyl-isoquinolines, starting from
phenyl ethylamine.
(5) is a related enone prodrug and its corresponding cate-
cholamine TL-334 (6) is an extremely potent dopamine
D /D agonist. The structural difference between apomor-
1
2
phine and TL-334 is only one aromatic ring. Due to the
potent dopamine agonistic effects observed in vivo with
these enone prodrugs, the corresponding enone derivatives
(
21a and 21b, respectively) of apomorphine and NPA might
The synthetic procedure followed is displayed in Schemes 2
and 3, starting from phenyl ethylamine 7. Acetylation of 7
gave the acetamide 8, which underwent a high temperature
Keywords: Enone-prodrug; Apomorphine; Retro-synthetic; Ring opening.
*
Corresponding author. Tel.: +31 50 3633310; fax: +31 50 3636908;
e-mail: durk.dijkstra@rug.nl
9
cyclization with polyphosphoric acid (PPA) yielding 9 in
0
040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.04.103

D. Liu et al. / Tetrahedron 63 (2007) 7264–7270
7265
N
N
N
N
HO
HO
OH
OH
O
O
6
3
4
5
Figure 2. Structures of enone prodrugs and their catecholamines: S-(ꢀ)-5,6-di-OH-DPAT (3) and PD148903 (4); GMC-6650 (5) and TL-334 (6).
R
N
C
R
N
C
R
N
C
HO
A
B
A
B
B
O
HO
O
D
D
D
O
1
4a, R = CH3
2
1a, R = CH3
14b, R = CH2CH2CH3
21b, R = CH2CH2CH3
R
N
C
R
N
C
O
R
N
C
O
H2N
C
EtO
B
D
D
D
D
7
Scheme 1. Retro-synthetic analysis of 21a,b.
NH2
H
N
a
b
O
7
8
d
c
+
N
N
N
R
R
9
1
0a R = CH3
11a R = CH
3
10b R = CH2CH2CH3
11b R = CH CH CH
2
2
3
R
N
e
f
g
+
N
N
R
R
O
O
OEt
O
OEt
1
4a R = CH3
14b R = CH2CH2CH3
1
2a R = CH3
1
3a R = CH3
12b R = CH2CH2CH3
13b R = CH2CH2CH3
ꢁ
ꢁ
ꢁ
Scheme 2. Reagents and conditions: (a) Ac
1%; R¼CH CH CH : toluene, CH CH CH
R¼CH , over three steps), 69% (R¼CH CH
2
O, pyridine, 90 C, 2 h, 77%; (b) PPA, 130 C, 30 min, 200 C, 3 h, 53%; (c) R¼CH
3
: benzene, CH
3
I, rt, overnight,
, EtOH, rt, overnight, 71%
ꢁ
6
(
2
2
3
3
2
2
I, reflux, overnight, 85%; (d) 4 N NaOH, 0 C; (e) toluene, BrCH
2
COOEt; (f) NaBH
4
ꢁ
3
2
3
CH
3
, over three steps); (g) PPA, 140 C, 1 h, 54% (R¼CH
3
2 3 3
), 72% (R¼CH CH CH ).
moderate yield. The methylation and propylation of the iso-
quinoline 9 were performed under slightly different condi-
were formed. The last step in the preparation of the key in-
termediate 14 was a ring closure with PPA.
1
0
tions due to the different reactivity of methyl iodide (rt in
benzene) and propyl iodide (reflux in toluene). In the pres-
ence of 4 M NaOH, the iodide salts of the imines (10a and
10b) were converted to the enamines (11a and 11b), which
were next allowed to react with ethyl bromoacetate to yield
1
1
The next critical step was the alkylation at the a-position of
the ketone 14. Since the direct introduction of the ethyl bu-
1
2
tyrate group might be difficult to realize, we have chosen
a two-step procedure. First, we did an ethoxycarbonylation
with diethyl carbonate using NaH as base to produce the
the 1-iso-quinolinepropionate derivatives (12a and 12b). Af-
ter the reduction of these imines with NaBH , 13a and 13b
1
3
anion, yielding 15a and 15b. The procedure published in
4

7266
D. Liu et al. / Tetrahedron 63 (2007) 7264–7270
OEt
O
O
R
N
O
O
R
N
O
O
R
N
EtO
EtO
EtO
h
i
1
4a R = CH3
HO
14b R = CH CH CH
3
2
2
1
5a R = CH3
5b R = CH CH CH
16a R = CH3
17a R = CH3
1
16b R = CH2CH2CH3
17b R = CH CH CH
3
2
2
3
2
2
OH
OH
O
R
N
O
R
N
R
N
j
k
l
O
HO
O
1
9a R=CH3
20a R=CH3
20b R=CH2CH2CH3
21a R=CH3
21b R=CH2CH2CH3
19b R=CH2CH2CH3
Scheme 3. Reagents and conditions: (h) CO(OC
2
H
5
)
THF/DMF, reflux, overnight or NaH, DME, reflux, overnight; (j) concd HCl/CH
2
, NaH, benzene, reflux, overnight, 62% (R¼CH
3
), 66% (R¼CH
COOH, reflux, 6 h, 75% (R¼CH , over two steps), 75% (R¼CH
, over two steps), 35% (R¼CH CH CH , over two steps).
2
CH
3
CH
3
); (i) Br(CH
2
)
3
COOC
2
H
5
, NaH,
3
3
2
CH CH
3
3
,
ꢁ
over two steps); (k) NaBH
4
, NaHCO
3
, rt, overnight; (l) PPA, 65 C, 1 h, 34% (R¼CH
3
2
3
3
the literature (sodium ethoxide in ethanol) was not success-
the crude acids 19a and 19b were purified by column chro-
matography. The reduction of 19a and 19b with NaBH was
performed in aq NaHCO solution to yield 20a and 20b. Af-
3
1
4
ful in our hands. Compounds 15a and 15b existed in equi-
librium between the keto and the enol form. The NMR
spectroscopic experiments indicated that in solution, 15a
and 15b existed as the enols 16a and 16b, respectively. It
may be the case that in solution 15a and 15b are susceptible
for ketone–enol tautomerism and the equilibrium shifts in
4
ter the solution was acidified to pH <2, the solution was
evaporated. Without further purification, PPA was added
ꢁ
17
and the mixture was heated to 65 C for 1 h.
1
5
solution to the enol form. The alkylation of 15a and 15b
with ethyl-4-bromobutyrate resulted in 17a and 17b, which
existed as a mixture of diastereomers. Apparently, the keto
form is more reactive because the reaction with ethyl-4-
bromobutylate gave mainly the desired di-ester. Due to the
presence of enol form of 16, ether formation of 18a and
After this cyclization, it was found that the ring-opened (ring
C) by-product 22 was present in the final products (Fig. 4).
Aromatization of the B-ring is probably the driving force
for this ring opening. The cyclization of 20a,b in PPA was
crucial in this case. There seems to be an optimum between
the formation of product and ring-opened by-product. In-
complete reaction was found with shorter reaction time
(0.5 h), however, only ring-opened by-product was found
with the extended reaction time (2 h). Therefore, 1 h reac-
1
8b (see Fig. 3) were observed as by-products, which ex-
plained that after hydrolysis, 14a and 14b were regenerated
from 18a and 18b.
ꢁ
tion at 60 C was found to be the optimal time for the balance
of obtaining product and keeping a minimal amount of
by-product.
1
6
Subsequent de-ethoxycarbonylation of compounds 17a
and 17b was performed in concd HCl and acetic acid
(
1:1). The ester group in 17a and 17b was hydrolyzed to
the corresponding acids 19a and 19b. Although these two
compounds have two stereogenic carbon atoms, due to the
ketone–enol tautomerism, NMR spectroscopic data did not
show the evidence of being diastereomers. After work-up,
The racemic mixtures of 21a and 21b were resolved into two
enantiomers by preparative chiral HPLC with a Chiralpack
AD column. Figure 5 shows the result of the single crystal
X-ray analysis of (ꢀ)-21b. From the X-ray structure, the ab-
solute configuration on the fused C-ring of this compound
was shown to be (R)-(ꢀ). From the torsion angles data,
ꢁ
C5–C6 has a torsion angle of 174.3 with C7–N; the torsion
O
O
R
N
O
R
N
C
R
N
O
O
A
B
1
8a R = CH3
D
O
O
18b R = CH2CH2CH3
21
22
Figure 3. The structure of ether by-product 18 formed in step (i) in
Scheme 2.
Figure 4. The structure of enone 21 and ring opening by-product 22.

D. Liu et al. / Tetrahedron 63 (2007) 7264–7270
7267
PD148903 (4) and GMC-6650 (5), has not been observed
for these apoenones. Therefore, it can be concluded that
21a and 21b are not suitable to become clinically effective
dopaminergic agonists.
4. Experimental section
4.1. General
Melting points were determined in glass capillaries on an
electro thermal digital melting point apparatus and are
uncorrected. H NMR spectra were recorded at 300 MHz
1
1
3
on a Varian-VXR 300 spectrometer and C NMR spectra
were recorded at 50.3 MHz on a Varian Gemini 200 spec-
trometer. The chemical shifts are given in parts per million
Figure 5. The structure of R-(ꢀ)-21b.
(
ppm) relative to the solvent; the splitting patterns are desig-
ꢁ
C7 and C11–C10 is 141.6 . The large torsion angles may be
the reason for the instability of this type of compound, since
the aromatization of the B-ring is the driving force to elim-
inate the large torsion.
angle between N–C7 and C11–C12 is 160.3 ; between C6–
ꢁ
nated as follows: s (singlet), d (doublet), t (triplet), q (quar-
tet), m (multiplet). Coupling constants are given in hertz
(Hz). The spectra recorded were consistent with the pro-
posed structures of intermediates and final compounds. Elec-
tronic ionization (EI) mass spectra were obtained on
Shimadzu QP5000 GC/MS system equipped with a 17A
GC, an AOC-20i auto-injector, and the GC–MS solution
software 1.10 was used. Chemical ionization (CI) mass spec-
tra were recorded by the Mass Spectrometry Unit of the Uni-
versity of Groningen. High-resolution mass spectrometry
(HRMS) was analyzed on a JEOL MS route JMS-600H by
the Department of Chemistry, University of Groningen.
For identification of the metabolites of 21a,b, racemic apo-
morphine (1) and NPA (2) were used as standard. A multiple
reaction monitoring (MRM) experiment was performed,
which is accomplished by specifying the parent mass of the
compound for MS/MS fragmentation and then specifically
monitoring for a single fragmentation. The specific
experiment is known as a ‘transition’, which can be written
as parent mass/fragment mass. The MRM transition
4.1.1. 1-Methyl-3,4-dihydroisoquinoline (9). Phenyl ethyl-
amine (7) (32.3 g, 0.27 mol) was dissolved in pyridine
(25 mL, 0.30 mol) and acetic anhydride (29 mL, 0.31 mol)
was added dropwise to the solution. The solution was heated
2
68/219 was chosen to identify apomorphine. The re-
tention time is 7.40 min. After the injection of brain and
plasma samples of apoenone (21a) to LC/MS/MS system,
there was no peak found at the retention time of 7.4 min
with MRM transition 268/219. The multiple reaction mon-
itoring (MRM) transition 296/219 was chosen to identify
NPA (2) and the retention time is 7.96 min. After the injec-
tion of brain and plasma samples of NPA-enone (21b) to
LC/MS/MS system, the MRM transition 296/219 was cho-
sen and there was no peak found at the retention time of
ꢁ
at 90 C for 2 h. After cooling to rt, the volatiles were re-
moved in vacuo. The residue was partitioned between ethyl
acetate (200 mL) and 4 M HCl (40 mL). The organic layer
was washed with 1 M NaOH (3ꢂ40 mL), brine (3ꢂ
20 mL), and dried over MgSO . The solvent was filtered
4
and evaporated in vacuo. Crude N-phenethylacetamide 8
(34 g, 77% yield) was obtained as a light yellow solid, which
was used for the next step without further purification.
7
.96 min either. Therefore, in our preliminary experiments,
apomorphine (1) or NPA (2) were found neither in blood
nor in brain samples, which indicated that no conversion of
these enone compounds occurred in vivo or probably the for-
mation of the aromatized products are faster. Meanwhile, the
apomorphine enone and the NPA-enone by-products (ring-
opened, 22) were found in blood as well as in brain samples.
PPA (180 g) was added to 8 (34 g, 0.21 mol) and the syrup
was heated at 130 C for 30 min, then 200 C for 3 h. The re-
ꢁ
action was cooled to 40 C and water (900 mL) was added.
ꢁ
ꢁ
The solution was adjusted to pH¼9 with NH (25% water so-
3
lution) and extracted with CH Cl (3ꢂ300 mL). The organic
2
2
phase was washed with brine (3ꢂ50 mL) and dried over
MgSO . After filtration and evaporation of the solvent, a
4
3
. Conclusion
black oil was obtained. Purification by column chromato-
graphy (silica, CH Cl /MeOH¼30:1) afforded 9 as a light
2
2
1
We succeeded in synthesizing these challenging prodrugs of
apomorphine and NPA (21a and 21b) via a 12-steps route in
an overall yield of 1%.
yellow oil (16 g, 53% yield). H NMR (CDCl ) d 7.11–7.45
3
1
3
(m, 4H), 3.63 (m, 2H), 2.66 (m, 2H), 2.35 (s, 3H) ppm;
C
NMR (CDCl ) d 162.7, 135.8, 129.0, 128.0, 125.9, 125.3,
3
+
123.7, 45.4, 24.5, 21.8 ppm; MS (EI) m/z 145 (M ).
The result of identification of the metabolites of 21a,b with
the MRM experiment showed no conversion of these enone
compounds occurred in vivo or probably the formation of the
aromatized products are faster.
4.1.2. 1-Methyl-3,4-dihydroisoquinoline (11a). The iso-
quinoline 9 (10 g, 69.0 mmol) was dissolved in benzene
(100 mL) and methyl iodide (30 g, 211.2 mmol) was intro-
ducedtothesolutionatrtunder N . Afteraddition, thesolution
2
So far, the preliminary pharmacological results showed that
the expected bioactivation mechanism, found for both
turned to a suspension and was stirred overnight. After filtra-
tion, the solid was washed with benzene and then with ether;

7268
D. Liu et al. / Tetrahedron 63 (2007) 7264–7270
10a iodide salt was obtained as a yellow solid, 12 g (61%
yield). The analytic sample was re-crystallized from metha-
nol/ether to yield light yellow needles, mp: 132–134 C.
which was purified by column chromatography (silica
treated with NH , CH Cl /MeOH¼40:1) yielding 13a as
3
2
2
ꢁ
1
a yellow oil (7.3 g, 71% yield over two steps). H NMR
(
CDCl ) d 7.12 (m, 4H), 4.09 (q, 2H, J¼7.1 Hz), 3.53 (m,
3
The iso-quinolinium salt 10a (12 g, 41.8 mmol) was dis-
solved in H O (50 mL) and the solution was basified with
1H), 3.05 (m, 1H), 2.48–2.82 (m, 3H), 2.44 (s, 3H), 2.20–
2.42 (m, 1H), 2.10 (m, 3H), 1.23 (t, 3H, J¼7.1 Hz) ppm;
2
ꢁ
13
4
M NaOH at 0 C. This solution was extracted with toluene
C NMR (CDCl ) d 172.7, 135.9, 133.8, 127.1, 125.5,
3
(
(
2ꢂ100 mL). The organic layer was washed with brine
124.4, 124.3, 61.3, 58.6, 47.3, 41.6, 28.5, 28.2, 25.3,
12.7 ppm; MS (EI) m/z 202 (Mꢀ45).
3ꢂ50 mL) and dried over Na SO . After filtration, 11a
2
4
was used in toluene solution directly for the next step.
4
.1.7. Ethyl 3-(2-propyl-1,2,3,4-tetrahydro-1-isoquinoli-
4
.1.3. 1-Propyl-3,4-dihydroisoquinoline (11b). To a stirred
solution of 1-methyl-3,4-dihydroisoquinoline 9 (13 g,
9.7 mmol) in toluene (30 mL) was added propyl iodide
nyl)propyl ether (13b). This compound was prepared
from 12b (from 11b, 9.9 g, 52.9 mmol) using the same pro-
cedure as described for 13a. After purification, 13b was ob-
8
(
9
14.2 mL, 145.0 mmol) at rt. The mixture was heated at
0 C for 18 h. After the suspension was filtrated, the formed
tained as a yellow oil (10 g, 36.3 mmol, 69% yield over two
steps). H NMR (CDCl ) d 7.03–7.26 (m, 4H), 4.11 (dd, 2H,
ꢁ
solid was washed with CH Cl /ether (1:2) to yield 10b as
1
3
J¼7.1, 7.3 Hz), 3.59 (m, 1H), 3.17 (m, 1H), 2.69–2.95 (m,
2
2
a yellow solid, 24 g (85% yield). The analytic sample was
crystallized from methanol/ether as light yellow crystals,
mp: 194–197 C.
2H), 2.27–2.60 (m, 5H), 1.21 (m, 2H), 1.41–1.56 (dd, 2H,
J¼7.1, 7.3 Hz), 1.26 (t, 3H, J¼7.3 Hz), 0.90 (t, 3H,
ꢁ
13
J¼7.3 Hz) ppm; C NMR (CDCl ) d 172.7, 136.8, 133.6,
3
1
27.3, 126.3, 124.3, 124.2, 58.7, 58.6, 54.1, 42.3, 29.6,
The iso-quinolinium salt 10b (24 g, 76.4 mmol) was dis-
solved in H O (100 mL) and 4 M NaOH was added to the
23.0, 19.6, 12.7, 10.3 ppm; MS (EI) m/z 230 (Mꢀ45).
2
ꢁ
solution at 0 C. The solution was extracted with toluene
2ꢂ150 mL) and the combined organic layers were dried
4.1.8. 1-Methyl-1,2,3,8,9,9a-hexahydro-7H-benzo[de]-
quinolin-7-one (14a). A mixture of 13a (5 g, 20.2 mmol)
(
over Na SO . Filtration and evaporation of the solvent
ꢁ
to 60 C, the mixture was poured into water (800 mL), and
and PPA (100 g) was stirred at 140 C for 1 h. After cooling
ꢁ
2
4
1
yielded 11b as a red oil (9.9 g, 70% yield). H NMR
CDCl ) d 7.74 (m, 1H), 7.15 (m, 3H), 4.48 (s, 1H), 3.86 (s,
(
the solution was adjusted to pH¼9 by the addition of NH
3
3
1
H), 3.24 (m, 4H), 2.89 (t, 2H, J¼5.8 Hz), 1.69 (q, 2H,
(25% water solution). The solution was extracted with
CH Cl (4ꢂ150 mL). The combined organic layers were
1
3
J¼7.3 Hz), 0.97 (t, 3H, J¼7.3 Hz) ppm; C NMR (CDCl )
3
2
2
d 145.0, 133.3, 131.3, 126.5, 125.8, 124.7, 123.4, 75.4,
5
washed with H O (3ꢂ100 mL) and brine (3ꢂ50 mL), and
2
+
2.7, 46.2, 29.1, 17.3, 10.3 ppm; MS (EI) m/z 187 (M ).
dried over Na SO . After filtration and evaporation of the
2
4
solvent, a black oil was obtained, which was purified by col-
umn chromatography (silica treated with NH , CH Cl /
4
.1.4. Ethyl 3-[2-methyl-3,4-dihydro-1(2H)-isoquinolinyl-
3
2
2
idene]propanoate (12a). A solution of ethyl bromoacetate
6 mL, 53.0 mmol) in toluene (20 mL) was added slowly
at rt to a stirred solution of 11a (from 12 g 10a salt) in tolu-
MeOH¼50:1). Compound 14a was obtained as a yellow
1
(
oil (2.2 g, 54% yield). H NMR (CDCl ) d 7.86 (m, 1H),
3
7.27 (m, 2H), 3.05–3.35 (m, 3H), 2.49–2.86 (m, 4H), 2.49
(s, 3H), 1.77–1.95 (m, 2H) ppm; C NMR (CDCl3)
1
3
ene under N at rt in 45 min. The reaction was stirred as such
2
overnight. A red syrup 12a was formed during the reaction.
The toluene was removed and the red syrup was washed with
toluene (4ꢂ30 mL). Ethanol (100 mL) was added to the
syrup and this solution was used without further purification
in the next step.
d 196.2, 139.0, 132.6, 132.5, 129.4, 125.3, 123.8, 61.2,
52.3, 41.5, 36.1, 26.9, 26.6 ppm; MS (EI) m/z 201 (M ).
HRMS 201.1166 (obsd). Calcd for C H NO 201.1154.
13 15
+
4.1.9. 1-Propyl-1,2,3,8,9,9a-hexahydro-7H-benzo[de]qui-
nolin-7-one (14b). This compound was prepared from 13b
(10 g, 36.4 mmol) using the same procedure as described
4.1.5. Ethyl 3-[2-propyl-3,4-dihydro-1(2H)-isoquinolinyl-
idene]propanoate (12b). Ethyl bromoacetate (35 mL,
for 14a. Purification of the residue resulted in 14b as a yellow
oil (6 g, 72% yield). H NMR (CDCl ) d 7.86 (m, 1H), 7.27
1
3
5
15.0 mmol) was added slowly to the red oil 11b (9.9 g,
2.9 mmol) while stirring. A red syrup was formed after the
3
(m, 2H), 3.56 (m, 1H), 3.02–3.28 (m, 2H), 2.81 (m, 3H),
2.40–2.69 (m, 4H), 1.86 (m, 1H), 1.52–1.74 (m, 2H), 0.93
ꢁ
addition. The reactionwas heatedat60 C for1.5 h. The extra
ethyl bromoacetate was removed and the red syrup was
washed withtoluene (4ꢂ50 mL). Ethanol(80 mL) was added
to the red syrup and this solution was used for the next step.
1
3
(t, 3H, J¼7.3 Hz) ppm; C NMR (CDCl ) d 196.4, 139.8,
3
133.0, 132.5, 129.5, 125.2, 123.8, 58.9, 53.5, 48.2, 36.3,
27.1, 26.7, 17.7, 10.5 ppm; MS (EI) m/z 229 (M ). HRMS
+
2
29.1471 (obsd). Calcd for C H NO 229.1467.
15 19
4.1.6. Ethyl 3-(2-methyl-1,2,3,4-tetrahydro-1-isoquinoli-
nyl)propanoate (13a). To the stirred ethanol solution of
4.1.10. 1-Ethyl-7-hydroxy-1-methyl-2,3,9,9a-tetrahydro-
1H-benzo[de]quinoline-8-carboxylate (16a). To a suspen-
sion of NaH (60%, 250 mg, 6.2 mmol) in benzene (5 mL)
was added dropwise a solution of diethyl carbonate
(353 mg, 3.0 mmol) in benzene (5 mL), followed by the
dropwise addition of 14a (500 mg, 2.5 mmol) in benzene
(5 mL). The reaction was heated at reflux overnight. After
cooling, the mixture was poured into ice. The water layer
was extracted with ether (3ꢂ30 mL). The combined organic
1
2a (from 10a salt, 12 g, 41.8 mmol) was added NaBH
ꢁ
4
(
3.1 g, 81.6 mmol) in small portions at 0 C. The mixture
was allowed to reach to rt and continued to stir overnight.
After removal of the solvent in vacuo, water (50 mL) was
added to the residue. The mixture was extracted with
CH Cl (3ꢂ100 mL). The organic layer was washed with
2
2
brine (3ꢂ30 mL), dried over Na SO . After filtration and
2
4
evaporation of the solvent, the crude product was obtained,

D. Liu et al. / Tetrahedron 63 (2007) 7264–7270
7269
layers were washed with brine (3ꢂ10 mL), dried over
MeOH¼15:1), yielding 19a as a white solid (205 mg, 75%
1
Na SO . After filtration and evaporation, the obtained resi-
4
yield). H NMR (CD OD) d 7.93 (d, 1H, J¼7.1 Hz), 7.50
2
3
due was purified by column chromatography (silica treated
with NH , CH Cl /MeOH¼60:1) and yielded 16a as a yel-
(m, 2H), 3.24–3.86 (m, 8H), 3.06 (s, 3H), 2.75 (m, 2H),
1
3
2.11–2.35 (m, 2H), 1.73 (m, 2H) ppm; C NMR (CDCl3)
d 195.7, 168.1, 132.4, 132.1, 129.9, 129.6, 127.2, 124.8,
59.5, 50.8, 36.5, 32.7, 29.3, 27.4, 22.8, 20.4, 15.5 ppm;
MS (CI) m/z 288.2 (M+1). HRMS 287.1550 (obsd). Calcd
for C H NO 287.1521.
3
2
2
1
low oil (420 mg, 62% yield). H NMR (CDCl ) d 12.46 (s,
3
1
3
H), 7.64 (m, 1H), 7.23 (m, 2H), 4.30 (q, 2H, J¼7.1 Hz),
.00–3.27 (m, 4H), 2.56–2.77 (m, 2H), 2.51 (s, 3H), 2.45
1
3
(
m, 1H), 1.36 (t, 3H, J¼7.1 Hz) ppm; C NMR (CDCl3)
d 171.0, 163.2, 133.8, 131.9, 129.7, 127.2, 125.0, 120.8,
4.0, 59.7, 59.2, 52.0, 42.3, 27.3, 24.2, 12.9 ppm; MS (CI)
17
21
3
9
4.1.15. 4-(1-Propyl-7-oxo-2,3,7,8,9,9a-hexahydro-1H-
benzo[de]quinolin-8-yl)butanoic acid (19b). This com-
pound was prepared from 17b (300 mg, 0.72 mmol) using
the same procedure as described for 19a. After work-up,
+
m/z 274 (M +1).
4
1
.1.11. Ethyl 7-hydroxy-1-propyl-2,3,9,9a-tetrahydro-
H-benzo[de]quinoline-8-carboxylate (16b). This com-
1
19b was obtained as a white solid (170 mg, 75% yield). H
NMR (CD OD) d 7.77 (d, 1H, J¼6.9 Hz), 7.31 (m, 2H),
pound was prepared from 14b (1 g, 4.4 mmol) with diethyl
carbonate (620 mg, 5.2 mmol) using the same procedure
as described for 16a. After work-up, 16b was obtained as
3
4.34 (d, 1H, J¼9.5 Hz), 3.57 (d, 1H, J¼7.6 Hz), 2.82–3.20
13
(m, 7H), 2.56 (dd, 1H, J¼3.3 Hz), 2.21 (m, 2H), 1.43–1.98
1
a yellow oil (870 mg, 66% yield). H NMR (CDCl3)
d 12.38 (s, 1H), 7.57 (m, 1H), 7.17 (m, 2H), 4.26 (q, 2H,
J¼7.3 Hz), 3.43 (m, 1H), 2.96–3.16 (m, 3H), 2.64–2.88
(m, 6H), 0.92 (t, 3H, J¼7.3 Hz) ppm; C NMR (CD OD)
3
d 196.5, 175.5, 134.7, 132.2, 130.8, 129.9, 126.4, 124.3,
58.4, 51.1, 47.2, 46.4, 32.9, 30.0, 27.5, 23.7, 20.5, 16.0,
8.7 ppm; MS (CI) m/z 316.2 (M+1). HRMS 315.1834
(obsd). Calcd for C H NO 315.1834.
(
m, 2H), 2.39 (m, 2H), 2.10 (t, 1H, J¼4.6 Hz), 1.53 (m,
2
H), 1.31 (t, 3H, J¼7.3 Hz), 0.91 (t, 3H, J¼7.3 Hz) ppm;
1
9
25
3
1
3
C NMR (CDCl ), d 171.1, 163.2, 134.6, 132.3, 129.7,
3
1
2
27.5, 125.5, 124.8, 120.7, 94.2, 59.2, 57.1, 54.4, 47.7,
7.4, 24.1, 17.6, 12.9, 10.5 ppm; MS (EI) m/z 228 (Mꢀ73).
4.1.16. 4-(7-Hydroxy-1-methyl-2,3,7,8,9,9a-hexahydro-
1H-benzo[de]quinolin-8-yl)butanoic acid (20a). Com-
pound 19a (170 mg, 0.59 mmol) was dissolved in water
(3 mL) containing NaHCO3 (110 mg, 1.31 mmol) at rt.
4
2
.1.12. Ethyl 8-(4-ethoxy-4-oxobutyl)-1-methyl-7-oxo-
,3,7,8,9, 9a-hexahydro-1H-benzo[de]quinoline-8-car-
NaBH (33 mg, 0.87 mmol) was added to this solution por-
4
boxylate (17a). To a suspension of NaH (60%, 79 mg,
.98 mmol) in THF (10 mL) and DMF (1 mL) was added
a solution of 16a (430 mg, 1.57 mmol) in THF (10 mL) at
tionwise and the mixture was stirred at rt overnight. The
solution was adjusted with 4 N HCl to pH¼2 and the vola-
tiles were removed in vacuo. The residue 20a was used in
the next step without further purification.
1
ꢁ
0
C. After stirring at rt for 2 h, a solution of ethyl-4-bromo-
butylate (480 mg, 2.46 mmol) in THF (10 mL) was added
dropwise, and the resulting mixture was heated under reflux
for 20 h. After cooling, the solvent was evaporated and fol-
4.1.17. 4-(7-Hydroxy-1-methyl-2,3,7,8,9,9a-hexahydro-
1H-benzo[de]quinolin-8-yl)butanoic acid (20b). This
compound was prepared from 19b (170 mg, 0.54 mmol)
using the same procedure described for 20a and the obtained
20b was used in the next step without further purification.
lowed by the addition of H O (3 mL), extracted with ether
2
(
Na SO . Purification by column chromatography (silica,
3ꢂ30 mL), washed with brine (3ꢂ10 mL), and dried over
2
4
ethyl acetate) yielded 17a as a light yellow oil (220 mg).
NMR and GC indicated the mixture of diastereomers of
the product, also the by-product 18a was involved, which
is 20% of the total amount according to GC. The mixture
was used directly in the next step.
4.1.18. 6-Methyl-5,6,6a,8,9,10-hexahydro-4H-dibenzo-
[de,g]quinolin-11(7H)-one (21a). To PPA (3.7 g) was added
20a (from 0.59 mmol 19a) in one pot and the syrup was
ꢁ
stirred for 1 h at 65 C. The resulting dark syrup was cooled
ꢁ
to 40 C, followed by the addition of water (30 mL). The so-
lution was adjusted to pH¼9 with NH (25% water solution).
4
2
.1.13. Ethyl 8-(4-ethoxy-4-oxobutyl)-1-propyl-7-oxo-
,3,7,8,9,9a-hexahydro-1H-benzo[de]quinoline-8-car-
3
The solution was extracted with CH Cl (3ꢂ10 mL),
2
2
boxylate (17b). This compound was prepared from 16b
(870 mg, 2.9 mmol) in ethylene glycol dimethyl ether (DME)
washed with water (3ꢂ10 mL) and brine (3ꢂ10 mL), and
dried over Na SO . After filtration and evaporation of the
2
4
(
15 mL) using the same procedure as described for 17a. Pu-
solvent, a dark syrup was obtained, which was purified by
column chromatography (silica treated with NH , CH Cl /
rification by column chromatography (silica, hexane/ethyl
acetate, 3:1) resulted in 17b as a light yellow oil (630 mg).
It was clearly shown that the product existed as a mixture
of diastereomers based on NMR and GC, also the by-product
18b was involved, which is 20% of the total amount accord-
ing to GC. The mixture was used directly in the next step.
3
2
2
MeOH¼60:1), yielded 21a (50 mg, 34% yield over two
1
steps). H NMR (CDCl ) d 7.83 (d, 1H, J¼7.8 Hz), 7.20
3
(m, 1H), 7.01 (d, 1H, J¼7.6 Hz), 2.97–3.20 (m, 3H), 2.25–
1
3
2.75 (m, 11H), 2.04 (m, 2H) ppm; C NMR (CDCl )
3
d 195.3, 155.8, 130.9, 130.8, 129.1, 128.2, 126.1, 124.7,
123.4, 58.8, 51.8, 42.6, 37.7, 34.1, 30.7, 27.5, 20.2 ppm;
MS (CI) m/z 254.2 (M+1).
4
.1.14. 4-(1-Methyl-7-oxo-2,3,7,8,9,9a-hexahydro-1H-
benzo[de]quinolin-8-yl)butanoic acid (19a). The com-
pound 17a (300 mg, 0.78 mmol) was dissolved in concd
HCl and acetic acid (3 mL: 3 mL), and the solution was
1
The ring opening by-product is 22a (40 mg, 27%). H NMR
(CDCl ) d 9.27 (d, 1H, J¼8.8 Hz), 8.20 (d, 1H, J¼8.8 Hz),
3
heated under reflux for 6 h under N . After cooling to rt,
2
7.26–7.55 (m, 3H), 3.44 (s, 1H), 3.29 (t, 2H, J¼7.3 Hz),
3.07 (t, 2H, J¼6.1 Hz), 2.95 (m, 2H), 2.76 (m, 2H), 2.46
the solvent was evaporated in vacuo. The obtained residue
was purified by column chromatography (silica, CH Cl /
1
3
(s, 3H), 2.16 (p, 2H, J¼6.6, 6.1 Hz) ppm; C NMR
2
2

7270
D. Liu et al. / Tetrahedron 63 (2007) 7264–7270
(
1
2
CDCl ) d 199.0, 144.7, 134.1, 130.4, 129.7, 127.9, 126.8,
3
26.5, 125.5, 125.3, 124.0, 51.0, 39.6, 34.4, 31.6, 29.9,
1.4 ppm; MS (CI) m/z 254.2 (M+1).
ethanol until a diffuse cloud formed in the border formed
by the two phases. The two layers were kept standing over-
night, and platelet, colorless crystals were formed. The crys-
tallographic data for the structure 21b in this paper have
been deposited with the Cambridge Crystallographic Data
Centre as supplementary publication no. CCDC 281494.
Copies of the data can be obtained free of charge at
www.ccdc.cam.ac.uk/conts/retrieving.html or form the
Cambridge Crystallographic Data Centre (CCDC), 12 Union
Road, Cambridge CB2 1EZ, UK [fax: +44 1223 336033;
e-mail: deposit@ccdc.cam.ac.uk].
4
[
.1.19. 6-Propyl-5,6,6a,8,9,10-hexahydro-4H-dibenzo-
de,g]quinolin-11(7H)-one (21b). This compound was
prepared from 20b (from 0.54 mmol 19b) using the same
procedure as described for 21a. Purification of the residue
yielded 21b (60 mg, 35% yield over two steps). H NMR
(
8
(
1
(
1
CDCl ) d 7.75 (d, 1H, J¼7.7 Hz), 7.11 (t, 1H, J¼7.7,
3
.1 Hz), 6.95 (d, 1H, J¼7.7 Hz), 3.34 (m, 1H), 2.95–3.13
m, 2H), 2.22–2.80 (m, 10H), 1.90–2.14 (m, 2H), 1.46–
.63 (m, 2H), 0.91 (t, 3H, J¼7.3 Hz) ppm; C NMR
3
1
3
CDCl ) d 195.3, 156.1, 131.5, 131.3, 129.2, 128.5, 126.0,
Acknowledgements
1
1
24.5, 123.3, 56.2, 54.9, 47.6, 37.7, 34.2, 30.7, 27.6, 20.3,
7.9, 10.6 ppm; MS (CI) m/z 282 (M+1).
We gratefully acknowledge Dr. Auke Meetsma (Crystal
Structure Center, Chemical Physics Materials Science
Center, University of Groningen) for X-ray analysis and
Lundbeck A/S (Denmark) for the financial support.
1
The ring-opened product is 22b (60 mg, 35%). H NMR
(
CDCl ) d 9.27 (d, 1H, J¼8.5 Hz), 8.20 (d, 1H, J¼8.8 Hz),
3
7
(
2
.53 (m, 1H), 7.31 (m, 2H), 3.27 (t, 2H, J¼7.3 Hz), 3.09
dd, 2H, J¼6.1, 5.9 Hz), 2.96 (dd, 2H, J¼6.3, 7.1 Hz),
.77 (dd, 2H, J¼6.1, 7.1 Hz), 2.60 (t, 2H, J¼7.3 Hz), 2.16
m, 2H), 1.80 (s, 1H), 1.51 (dd, 2H, J¼7.1, 7.3 Hz), 0.89
Supplementary data
(
(
1
1
1
3
t, 3H, J¼7.3 Hz) ppm; C NMR (CDCl ) d 199.0, 144.6,
Supplementary data associated with this article can be found
in the online version, at doi:10.1016/j.tet.2007.04.103.
3
34.6, 130.4, 129.7, 128.1, 126.8, 126.5, 125.3, 125.2,
23.8, 50.3, 49.3, 39.6, 32.3, 29.9, 21.6, 21.4, 10.3 ppm;
MS (CI) m/z 282 (M+1).
References and notes
4
.2. General procedure for chiral separation of 21a
1
. Stewart, J. J. J. Toxicol. Clin. Toxicol. 1983, 20, 199–253.
A 30 mg/mL solution of racemic 21a in 2-propanol was in-
jected into a HPLC system using a Water 510 HPLC pump
with a Chiralpack AD preparative column (250ꢂ10 mm).
Mobile phase was a mixture produced by an ISCO Model
2. Chen, J. J.; Obering, C. Clin. Ther. 2005, 27, 1710–1724.
3. Mouradian, M. M.; Heuser, I. J.; Baronti, F.; Giuffra, M.;
Conant, K.; Davis, T. L.; Chase, T. N. J. Neurol. Neurosurg.
Psychiatry 1991, 54, 401–405.
2
[
360 gradient programmer and consisted of 93% n-hexane
containing 0.1% (w/w) triethyl amine] and 7% 2-propanol/
4. Borgman, R. J.; McPhillips, J. J.; Stitzel, R. E.; Goodman, I. J.
J. Med. Chem. 1973, 16, 630–633.
n-hexane [1/1 (w/w) containing 0.1% triethyl amine]. Flow
of the mobile phase was 4.0 mL/min. The peaks were de-
tected by a Water 486 Millipore tunable absorbance detector
5. Horn, A. S.; Dijkstra, D.; Feenstra, M. G. P.; Grol, C. J.;
Rollema, H.; Westerink, B. H. C. Eur. J. Med. Chem. 1980,
15, 387–392.
(
l¼254 nm, AUFS¼2.0) and were recorded on paper using
6. Wikstr €o m, H.; Lindberg, P.; Martinson, P.; Hjorth, S.; Carlsson,
A. J. Med. Chem. 1978, 21, 864–867.
7. Venhuis, B. J.; Dijkstra, D.; Wustrow, D. J.; Meltzer, L. T.;
Wise, L. D.; Johnson, S. J.; Heffner, T. G.; Wikstrom, H. V.
J. Med. Chem. 2003, 46, 584–590.
8. Liu, D.; Wikstrom, H. V.; Dijkstra, D.; de Vries, J. B.; Venhuis,
B. J. J. Med. Chem. 2006, 49, 1494–1498.
9. Snyder, H. R.; Werber, F. X. J. Am. Chem. Soc. 1950, 72, 2962–
a Kipp & Zonen flatbed recorder (chart speed 5 mm/min).
After evaporation of the mobile phase, the optical rotation
of the two fractions was determined using a Perkin–Elmer
2
nol), second fraction: [a] +312 (c 0.55, methanol). Both en-
antiomers (ee >95%) were analyzed for their purity using
analytical HPLC (Chiralpack AD column, 250ꢂ4.6 mm).
2
3
41 polarimeter. First fraction: [a] ꢀ316 (c 0.65, metha-
D
2
3
D
2
10. Berney, D.; Schuh, K. Helv. Chim. Acta 1982, 65, 1304–1309.
965.
4
.3. Procedure for chiral separation of 21b
1
1. Bremner, J. B.; Coban, B.; Griffith, R.; Groenewoud, K. M.;
Yates, B. F. Bioorg. Med. Chem. 2000, 8, 201–214.
Similar as for 21a, the solution of racemic compound was
0 mg/mL, and the mobile phase consisted of 95% n-hexane
containing 0.1% (w/w) triethyl amine] and 5% 2-propanol/
n-hexane [1/1(w/w), containing 0.1% triethyl amine]. First
2
[
12. Wada, A.; Sakai, M.; Kinumi, T.; Tsujimoto, K.; Yamauchi,
M.; Ito, M. J. Org. Chem. 1994, 59, 6922–6927.
13. Miyashi,T.;Nishizawa, Y.;Fujii,Y.;Yamakawa,K.;Kamata,M.;
Akao, S.; Mukai, T. J. Am. Chem. Soc. 1986, 108, 1617–1632.
14. Adams, R.; Loewe, S.; Theobald, C. W.; Smith, C. M. J. Am.
Chem. Soc. 1942, 2653–2655.
2
3
fraction: [a] ꢀ254 (c 0.365, methanol), second fraction:
D
2
3
[
a] +235 (c 0.405, methanol). Both enantiomers (ee
D
>
95%) were analyzed for their purity using analytical
HPLC (Chiralpack AD column, 250ꢂ4.6 mm).
15. Suzuki, K.; Tatsuoka, T.; Ishihara, T.; Ohno, T.; Aisaka, K.;
Ogino, R.; Kuroki, M.; Satoh, F.; Miyano, S.; Sumoto, K.
Chem. Pharm. Bull. 1996, 44, 132–138.
4
.4. X-ray crystallographic data of 21b
1
6. Huisgen, V. R.; Pawellek, D. Liebigs Ann. Chem. 1961, 641,
71–75.
17. Kulkarni, S. B.; Dev, S. Tetrahedron 1968, 24, 553–560.
Compound (ꢀ)-21b (10 mg, as its HCl salt) was dissolved in
abs EtOH (0.5 mL) at rt. Ether was gently added on top of