Synthetic studies towards the octahydro-1H-benzo[f]pyrrolo[3,2,1-ij]quinolines: Enantioselective synthesis of (2R,3S)-2-[(1S)-3-(benzyloxy)-1-(tert-butyldimethyl-silyloxymethyl) propyl]-3-phenylhexahydropyridine

Article abstract of DOI:10.1016/S0040-4039(01)01260-6

(2R,3S)-2-[(1S)-3-(Benzyloxy)-1-(tert-butyldimethylsilyloxymethyl)propyl]-3- phenylhexahydropyridine - a key synthetic intermediate for the preparation of (6aS,11bS,11cS)-1H-benzo[f]pyrrolo[3,2,1-ij]quinoline - was synthesized in nine steps. The synthetic approach uses both enantiomers of the chiral auxiliary trans-2,5-bis(methoxymethoxymethyl)pyrrolidine, employed in different stages of the synthesis for the construction of two out of the three stereogenic centers in the above mentioned intermediate.

Full text of DOI:10.1016/S0040-4039(01)01260-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 6393–6396
Synthetic studies towards the
octahydro-1H-benzo[ f ]pyrrolo[3,2,1-ij]quinolines:
enantioselective synthesis of
(2R,3S)-2-[(1S)-3-(benzyloxy)-1-(tert-butyldimethyl-
silyloxymethyl)propyl]-3-phenylhexahydropyridine
George S. Zaponakis and Haralambos E. Katerinopoulos*
Division of Organic Chemistry, Department of Chemistry, University of Crete, Heraklion 71 409, Crete, Greece
Received 14 May 2001; accepted 12 July 2001
Abstract—(2R,3S)-2-[(1S)-3-(Benzyloxy)-1-(tert-butyldimethylsilyloxymethyl)propyl]-3-phenylhexahydropyridine—a key synthetic
intermediate for the preparation of (6aS,11bS,11cS)-1H-benzo[ f ]pyrrolo[3,2,1-ij]quinoline—was synthesized in nine steps. The
synthetic approach uses both enantiomers of the chiral auxiliary trans-2,5-bis(methoxymethoxymethyl)pyrrolidine, employed in
different stages of the synthesis for the construction of two out of the three stereogenic centers in the above mentioned
intermediate. © 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
(methoxymethoxymethyl)pyrrolidine4 with a twofold
excess of phenylacetyl chloride and triethylamine
(CH2Cl2, 0°C) gave amide 6 in 75% yield. The reaction
of 6 with 1 equiv. LDA (THF, −78°C) led to an
enolate, which according to the literature5–7 should be
As part of an ongoing project for the preparation of
dopaminergic compounds,1–3 a number of compounds
were designed, based on the structure of 1H-
benzo[ f ]pyrrolo[3,2,1-ij]quinoline (1), which incorpo-
rates the phenethylamine moiety in its rigid framework.
Initial attempts focused on the synthesis of the
(6aS,11bS,11cS) enantiomer, which maintains a trans/
trans fusion of the three rings, keeping the phenethyl-
amine moiety in an antiperiplanar conformation, a
requirement for dopaminergic activity.1
2
1
3
11
8
NH
11a
11c
N
10
9
4
11b
5
OBn
6a
7a
6
7
OTBS
1
2
Retrosynthetic ring opening in 1 (Scheme 1) leads to
(2R,3S)-2-[(1S)-3-(benzyloxy)-1-(tert-butyldimethylsilyl-
OBn
OH
oxymethyl)propyl]-3-phenylhexahydropyridine
(2)
which, upon disconnection of a carbonꢀnitrogen bond
leads to methyl (2S,3S,4S)-2-(2-benzyloxyethyl)-3-
hydroxy-4-phenyl-6-heptenoate (3), which in turn can
be further disconnected to (S)-2-phenyl-4-pentenoic
acid (4) and (2S,5S)-N-(4-benzyloxybutyryl)-2,5-bis-
(methoxymethoxymethyl)pyrrolidine (5).
OMe
O
3
OMOM
N
COOH
BnO
The synthesis of (S)-2-phenyl-4-pentenoic acid (4) is
depicted in Scheme 2. Reaction of (2R,5R)-bis-
O
OMOM
4
5
Scheme 1. Retrosynthetic approach to (6aS,11bS,11cS)-1H-
benzo[ f ]pyrrolo[3,2,1-ij]quinoline (1).
* Corresponding author. Tel.: +30-81-393-626; fax: +30-81-393-601;
e-mail: kater@chemistry.uoc.gr
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01260-6

6394
G. S. Zaponakis, H. E. Katerinopoulos / Tetrahedron Letters 42 (2001) 6393–6396
acylating agent.11 This procedure appeared to be prob-
OMOM
N
lematic in our case given the length of the preparation
process of (S)-8 and the practical difficulties in the
addition of the lithium enolate of 5 to the reaction
mixture. The problem was overcome by addition of 1.1
equiv. of (S)-8 in one portion to the lithium enolate of
5 (THF, −78°C), brief stirring (1 min) of the mixture
and quenching by addition of water (Scheme 3). The
desired diastereomer 9 was isolated in 84% yield (83%
conversion). The diastereoselectivity of the reaction was
ca. 97%; a second minor (2%) diastereomer was isolated
via column chromatography. The proposed (S)-stereo-
chemistry at the acylated carbon is based on the notion
that the acylation occurs in the same sense as the
alkylation. Although one would expect that the new
stereogenic center in b-dicarbonyl compound 9 would
easily racemize via enolization, the compound showed a
striking stability in weakly acidic and basic solutions.
Previous studies on pyrrolidine and oxazolidone b-keto
amides report a similar stability of their respective
stereogenic centers.7 This observation could be
explained by the fact that the methine hydrogen of the
new stereogenic center in 9 lies nearly orthogonal to the
p-system of the adjacent imide carbonyl function, due
to the minimization of the nonbonding interactions
between the chiral auxiliary, and the other two sub-
stituents of the stereogenic center.7
a
b
COCl
OMOM
O
6
OMOM
N
d
c
COOH
OMOM
O
4
7
COCl
8
Scheme 2. Reagents and conditions: (a) TEA, rt, 12 h,
CH2Cl2, 78%; (b) allyl bromide, LDA, −78°C, rt, 3 h, 76%; (c)
HCl 3N, 3 h, reflux, 72%, 99% ee; (d) SOCl2, rt, 3 h, 100%.
of Z-stereochemistry. Addition of allyl bromide and
subsequent hydrolysis (3N HCl, reflux, 3 h) of the
resulting amide 7 gave (S)-2-phenyl-4-pentenoic acid (4)
in 55% yield from 6 and 99% ee as deduced from an
NMR study of the diastereomeric salts of 4 with (S)-a-
phenethylamine. The proposed (S) stereochemistry in 4,
induced by the presence of the (2S,5S)-auxiliary reagent
is in agreement with earlier studies using this chiral
reagent for the preparation of similarly a-substituted
carboxylic acids.5 Given the low chemical yield of 4 by
this procedure and the need of larger amounts of the
reagent, a more efficient alternative synthesis was also
used. Alkylation of phenylacetonitrile with allyl
bromide8 (NaOH 12.5N, BnNEt3Cl, 50°C, 3 h) and
subsequent hydrolysis (NaOH 12.5N, MeOH, reflux, 8
h) of the nitrile moiety gave racemic 4 in 60% overall
yield. Optical resolution of the racemic mixture using
(S)-a-phenethylamine gave (S)-4 whose optical activity
was identical to that of a purified sample from the first
preparation, as well as to that reported in the literature
([h]D=+102.5 (c=1, acetone)).9,10 The corresponding
acid chloride 8 was prepared quantitatively upon treat-
ment with thionyl chloride (rt, 3 h).
The next crucial step of the synthesis is the diastereose-
lective reduction of the keto group to the anti hydroxy
amide 10. The use of reagents that may form complexes
with the carbonyl moieties would involve a six-mem-
bered transition state leading to the undesired syn
product.12 Indeed, the use of reagents such as
13
Zn(BH4)2 led to a single diastereomer with a syn
arrangement of the three groups as indicated by its
NMR spectrum. Alternative use of non-complexing,
bulky reagents such as KEt3BH14 furnished the desired
OMOM
a
COCl
N
+
BnO
O
OMOM
8
5
OBn
OMOM
OBn
OMOM
N
c
b
N
Reaction of 4-benzyloxybutyric acid with (2S,5S)-bis-
(methoxymethoxymethyl)pyrrolidine4 (DCC, CH2Cl2,
0°C, 12 h), gave amide 5 in 81% yield. The DCC-
induced acid-auxiliary amine coupling was used as a
more efficient alternative to a direct reaction of the
corresponding acid chloride with the amine, since 4-
benzyloxybutyryl chloride spontaneously cyclized to
butyrolactone.
OH
O
OMOM
O
O
OMOM
9
10
OBn
OMe
OH
O
3
Coupling of (S)-8 with (2S,5S)-N-(4-benzyloxybu-
tyryl)-2,5-bis(methoxymethoxymethyl)pyrrolidine
(5)
(LDA THF, −78°C, 1 min) in basic media presented a
challenge in this synthesis. According to literature data
on related asymmetric acylations, affordable yields were
obtained only upon addition of an enolate to the
Scheme 3. Reagents and conditions: (a) LDA, −78°C, 5 min,
70% >97% ee; (b) K(Et)3BH, 1 h, ether, 0°C, 78%; (c) conc.
HCl, MeOH, then NaOH 3N, 60°C, 4 h then TMSCl, MeOH,
rt, 15 h, 73%.

G. S. Zaponakis, H. E. Katerinopoulos / Tetrahedron Letters 42 (2001) 6393–6396
6395
anti alcohol in 78% yield. The pyrrolidine auxiliary
was then removed in two steps by (a) acidic (MeOH,
trace of conc. HCl) cleavage of the MOM protective
groups and (b) subsequent basic hydrolysis (3N
Verification of the trans relationship of the H-2 and
H-3 of the piperidine ring was established via an H
1
NMR study of the N-benzylated derivative of 2. The
N-benzyl methylene protons appear as a pair of dou-
blets with a large chemical shift difference, whereas
the same signal in the cis isomer should appear as a
broad singlet.16 Indicative of the relative position of
the three methine protons is that COSY-NMR experi-
ments displayed a splitting of the H-2 resonance only
by its adjacent endocyclic methine proton (J=10.3
Hz) and that NOESY-NMR showed no interaction
between H-2 and H-3.
NaOH, 80°C,
4
h) of the 2,5-bis(hydroxy-
methyl)pyrrolidine. The resulting acid was esterified
(TMSCl, MeOH) without further purification. The
white crystalline product 3 was isolated from its
diastereomeric byproducts, by flash chromatography,
in 73% yield from 10 indicating that in this acid/base
removal of the chiral auxiliary, the presence of base
did not induce extensive racemization in 10.15
In conclusion, we have synthesized compound 2—a
key synthetic intermediate for the preparation of
(6aS,11bS,11cS) - 1H - benzo[ f ]pyrrolo[3,2,1 - ij]quino-
line (1) in nine steps and 7.7% overall yield from 4.17
The advantage of the synthetic approach chosen, is
the employment of both enantiomers of the chiral
auxiliary trans-2,5-bis(methoxymethoxymethyl)pyrro-
lidine, used in different stages in the synthesis for the
construction of two out of the three stereogenic cen-
ters in 2. The fact that both (2R,5R) and (2S,5S)
enantiomers of this auxiliary, are readily available by
the same synthetic route,4 makes these compounds
attractive agents for the construction of stereogenic
centers with highly predictable stereochemistry.
Reduction of the ester group in 3 with LiAlH4 (THF,
rt, 3 h) gave diol 11 in 98% yield (Scheme 4). The
primary alcohol group was selectively protected as the
TBDMS ether (TBDMSCl, imidazole, CH2Cl2, 3 h
93%) and the double bond in 12 was then hydrobo-
rated (BMS, then 3N NaOH, H2O2, 2 h, 70%) to give
diol 13. Mesylation of this material (MsCl, py,
CH2Cl2, 12 h, 77%) yielded the bis methanosulfonate
ester, which was transformed to the azide 14 (LiN3,
DMSO, rt, 6 h) in 60% yield by selective displace-
ment of the primary mesylate by the azide anion.
Reduction of the azide was achieved by hydrogena-
tion over 10% platinum on carbon, to avoid
hydrogenolysis of the benzyl ether. Under the reac-
tion conditions (H2, 10% Pt/C, EtOH) the intermedi-
ate amine partially cyclized to the piperidine
derivative 2, via an SN2 mechanism, with inversion of
configuration of the stereogenic center. The reaction
was driven to completion by treatment of the mixture
with K2CO3 in CH2Cl2.
Acknowledgements
This work was partially supported by an EPEAEK
grant from the EU and the Greek Ministry of
Education.
OBn
OBn
References
b
OMe
a
1. Katerinopoulos, H. E.; Kouvarakis, A. Synth. Commun.
1995, 25, 3035–3044.
OH OH
O
OH
2. Tagmatarchis, N.; Thermos, K.; Katerinopoulos, H. E. J.
3
11
Med. Chem. 1998, 41, 4165–4170.
3. Thermos, K.; Froudakis, G. E.; Tagmatarchis, N.;
Katerinopoulos, H. E. Bioorg. Med. Chem. Lett. 2001,
11, 883–886.
4. Yamamoto, Y.; Hoshino, J.; Fujimoto, Y.; Ohmoto, J.;
Sawada, S. Synthesis 1993, 298–302.
5. Kawanami, Y.; Ito, Y.; Kitagawa, T.; Taniguchi, Y.;
Katsuki, T.; Yamaguchi, Y. Tetrahedron Lett. 1984, 25,
857–860.
HO
OBn
OBn
c
d, e
OH OTBS
OTBS
OH
13
12
N3
6. Evans, D. A.; Takacs, J. M. Tetrahedron Lett. 1980, 21,
4233–4236.
OBn
NH
f
7. Evans, D. A.; Ennis, M. D.; Lee, T. J. Am. Chem. Soc.
1984, 106, 1154–1156.
8. Organic Syntheses; Wiley & Sons: New York, 1988.
OBn
OMs OTBS
OTBS
Collect. Vol. VI, pp. 897–900.
2
14
9. Veldstra, H.; Van De Westeringh, C. Rec. Trav. Chim.
1951, 70, 1113–1226.
10. Fredga, A.; Westman, L. Ark. Kemi 1954, 193.
11. Fujita, M.; Hiyama, T. Org. Synth. 1990, 69, 44–54.
12. Ito, Y.; Katsuki, T.; Yamaguchi, M. Tetrahedron Lett.
1984, 25, 6015–6016.
Scheme 4. Reagents and conditions: (a) LiAlH4, THF, 3 h,
rt, 98%; (b) TBDSCl, imidazole, CH2Cl2, 3 h, 93%; (c)
BMS, 3N NaOH, H2O2, 2 h, 65%; (d) pyridine, MsCl,
CH2Cl2, 12 h, 75%; (e) DMSO, LiN3, rt, 6 h, 60%; (f)
Pt/C/H2 then CH2Cl2, K2CO3, 55%.

6396
G. S. Zaponakis, H. E. Katerinopoulos / Tetrahedron Letters 42 (2001) 6393–6396
13. Gensler, W. J.; Johnson, F.; Sloan, D. J. Am. Chem.
Soc. 1960, 82, 6074–6079.
14. Ito, Y.; Katsuki, T.; Yamaguchi, M. Tetrahedron Lett.
1985, 26, 4643–4646.
15. Myers, A. G.; Cleason, J. L.; Yoon, T. J. Am. Chem.
Soc. 1995, 117, 8488–8489.
(m, 4H), 3.35–3.5 (m, 2H), 3.42 (s, 3H), 3.89 (ddd,
J1=4.7 Hz, J2=4.7 Hz, J3=8 Hz, 1H), 4.43 (s, 2H),
4.9 (d, J=10 Hz, 1H), 4.99 (d, J=17 Hz, 1H), 5.59
(ddt, J1=17 Hz, J2=10 Hz, J3=7 Hz, 1H), 7.17–7.32
(m, 10H). IR w (cm−1) 701, 739, 1169, 1729, 3500.
Compound 2. 1H NMR (500 MHz) l 0.0 (s, 6H), 0.89
(s, 9H), 1.5–1.55 (m, 1H), 1.55–1.75 (m, 3H), 1.8–1.87
(m, 2H), 1.94–1.99 (m, 1H), 2.6–2.65 (m, 1H), 2.75 (dt,
J1=2.6 Hz, J2=11.7 Hz, 1H), 3.03 (d, J=10.3 Hz,
1H), 3.18 (d, J=11.8 Hz, 1H), 3.37 (dd, J1=7.2 Hz,
J2=9 Hz, 1H), 3.42–3.46 (m, 1H), 3.5 (dd, J1=3.9 Hz,
J2=10.3 Hz, 1H), 3.61 (dd, J1=5.9 Hz, J2=10.2 Hz,
1H), 4.38 (d, J=12 Hz, 1H), 4.47 (d, J=12 Hz, 1H),
7.15–7.35 (m, 10H). 13C NMR (125 MHz) l 1.0, 18.13,
24.9, 25.9, 25.9, 34.39, 38.07, 46.5, 46.86, 64.53, 65.5,
68.9, 72.6, 126.3, 127.4, 127.6, 127.8, 128.3, 128.5,
138.4.
16. Walsh, D. A.; Smissman, E. E. J. Org. Chem. 1974, 39,
3705–3708.
17. Selected spectra: Compound 5. 1H NMR (250 MHz) l
1.85–2.14 (m, 6H), 2.41–2.49 (m, 2H), 3.29 (s, 3H), 3.34
(s, 3H), 3.41–3.55 (m, 4H), 3.24–3.3 (m, 1H), 3.69 (dd,
J1=3 Hz, J2=9.2 Hz, 1H), 4.01 (ddd, J1=7.7 Hz, J2=
7.7 Hz, J3=4.2 Hz, 1H), 4.17–4.23 (m, 1H), 4.46 (d,
J=1 Hz, 2H), 4.53 (s, 2H), 4.55 (d, J=7.5 Hz, 1H),
4.57 (d, J=7.5 Hz, 1H), 7.22–7.33 (m, 5H). IR w (cm−1
)
700, 740, 1044, 1110, 1641, 2824–2933. Compound 3.
1H NMR (300 MHz) l 1.71–2.04 (m, 2H), 2.42–2.77
.