Preparation of hexahydrobenzo[f]isoquinolines using a vinylogous Pictet-Spengler cyclization

Article abstract of DOI:10.1021/ol006533u

(matrix presented) A vinylogous Pictet-Spengler cyclization has been carried out using activated aldehydes, ketones, and alkynes to prepare a variety of substituted hexahydrobenzo[f]isoquinolines. A unique set of conditions was utilized to effect efficien

Full text of DOI:10.1021/ol006533u

ORGANIC
LETTERS
2000
Vol. 2, No. 23
3635-3638
Preparation of
Hexahydrobenzo[f]isoquinolines Using a
Vinylogous Pictet−Spengler Cyclization
Richard R. Cesati, III, and John A. Katzenellenbogen*
Department of Chemistry, UniVersity of Illinois at Urbana-Champaign,
Urbana, Illinois 61801
jkatzene@uiuc.edu
Received August 31, 2000
ABSTRACT
A vinylogous Pictet−Spengler cyclization has been carried out using activated aldehydes, ketones, and alkynes to prepare a variety of substituted
hexahydrobenzo[f]isoquinolines. A unique set of conditions was utilized to effect efficient cyclization with acid-sensitive electrophiles.
The Pictet-Spengler cyclization has been extensively studied
since its discovery in 1911.1 Over the years, the reaction
has been applied in the context of several heterocyclic
syntheses,2 and it continues to be a significant focus of
research. Early applications of the Pictet-Spengler reaction
were for the preparation of complex heterocycles such as
aza-steroids3 and morphine alkaloids.4 However, most of
the current literature is focused on the development of
diastereo- and enatioselective variants of the Pictet-Spengler
reaction, especially for preparing substituted tetrahydro-
isoquinolines5 and tetrahydro-â-carbolines.6 Recent mecha-
nistic investigations have provided some insight into
the stereochemical course of tetrahydro-â-carboline forma-
tion.7
For some time, we have been interested in the synthesis
of receptor-targeted breast tumor imaging agents, which
integrate metallic radionuclides, for use in single photon
emission computed tomography (SPECT) and positron
emission tomography (PET).8 We have recently designed a
novel series of compounds that possess steroidal backbones
(Figure 1). Our approach to the ligands for these complexes
evolved around the use of a vinylogous Pictet-Spengler
reaction9 in a manner similar to that used previously for the
preparation of a variety of 13-aza-steroids.10 In the course
of our studies, we surveyed several electrophiles (aldehydes,
ketones, and alkynes) and examined different reaction
conditions.
(1) Pictet, A.; Spengler, T. Ber. Dtsch. Chem. Ges. 1911, 44, 2030.
(2) (a) Bailey, P. D.; Morgan, K. M. Chem. Commun. 1996, 1479. (b)
Corey, E. J.; Gin, D. Y.; Kania, R. S. J. Am. Chem. Soc. 1996, 118, 9202.
(c) Hino, T.; Nakagawa, M. Heterocycles 1997, 46, 673.
(3) (a) Dijkink, J.; Speckamp, W. N. Tetrahedron Lett. 1977, 935. (b)
Kano, S.; Yuasa, Y. Heterocycles 1983, 20, 857. (c) Schleigh, W. R.; Catala,
A.; Popp, F. D. J. Chem. Soc. 1965, 379. (d) Zunnebeld, W. A.; Speckamp,
W. N. Tetrahedron 1975, 31, 1717.
(4) Douglas, J. L.; Meunier, J. Can. J. Chem. 1975, 53, 3681.
(5) (a) Czarnocke, A.; Arazby, Z. Heterocycles 1999, 44, 2871. (b)
Rozwadowska, M. D. Heterocycles 1994, 39, 903.
(6) (a) Cox, E. D.; Cook, J. M. Chem. ReV. 1995, 95, 1797. (b) Soe, T.;
Kawate, T.; Fukui, N.; Hino, T.; Nakagawa, M. Heterocycles 1996, 50,
1033. (c) Waldmann, H.; Schmidt, G. Tetrahedron 1994, 50, 11865.
Figure 1. Receptor-targeted breast tumor imaging agents that
possess steroidal backbones.
10.1021/ol006533u CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/21/2000

A generic diagram of the vinylogous Pictet-Spengler
reaction that we proposed to study for preparing the 13-aza-
steroid analogues is depicted in Scheme 1. The dihydronaph-
aldehydes and with cyclic ketones. The details of these
investigations are given in Table 1. Ethyl glyoxylate and
Table 1. Cyclization with Aldehydes and Ketones
Scheme 1
thylamine (1) has been reported in the literature and can be
prepared in a number of ways.11 For our purposes, we found
that homologation of 6-methoxy-1-tetralone via a Horner-
Emmons reaction with diethyl cyanomethylphosphonate,12
followed by Lewis acid mediated reduction (LiAlH4/AlCl3),
worked well. By this method, we were able to obtain the
∆
8(9) isomer (steroid numbering) exclusively in 70% overall
yield after recrystallization of the corresponding hydrochlo-
ride salt.
In our initial investigation of this reaction, we utilized a
traditional Pictet-Spengler synthetic protocol of heating the
dihydronaphthylamine hydrochloride (1‚HCl) and the car-
bonyl compound in an alcoholic solvent. We found that
although propionaldehyde cyclized readily, acetone did not.
In other cases, the products were isolated as mixtures of the
a Ar ) p-OCH3C6H4. b Time at reflux in butanol. c Compound 8 exists
as a 9:1 mixture of olefin isomers.
∆
8(9) and ∆9(11) olefin isomers (steroid numbering), and only
upon extended heating, or in those cases where there was
additional functionalization, were we able to obtain products
consisting only of the ∆8(9) isomer (see below). If butanol
rather than methanol was used as the solvent, reaction times
were shortened and product yields improved.13
diethyl ketomalonate both reacted smoothly with 1 in 24 h
to yield the corresponding hexahydrobenzo[f]isoquinolines
in 84% and 97% yield, respectively, as single olefin isomers.
Preparation of the spiro-fused compounds 4 and 5 using
tetrahydro-4H-pyran-4-one and cyclopentanone, respectively,
required longer reaction times (72 h). These extended
reaction times were needed not to improve reaction yield
but rather to isomerize the intermediate mixture of isomers
exclusively to the ∆8(9) isomer.
The preparation of cyclic ketone 8 deserves some ad-
ditional comment. In our approach to target molecule II we
desired to synthesize a compound that bore a dithiol
functionality of suitable orientation to form a tripartite chelate
(S,N,S) for technetium in a [3 + 1] fashion. However, when
we attempted to cyclize an acyclic version of ketone 6
(specifically, thioacetic acid S-(3-acetylsulfanyl-2-oxo-pro-
pyl) ester (7)), we were unable to obtain any of the tricyclic
product. Hence, we decided to try ketone 6, which could be
readily prepared in two steps from 4-methoxybenzaldehyde
and ethyl 2-mercaptoacetate, according to a literature method.14
We were pleased that compound 6 indeed underwent smooth
cyclization under conditions similar to those described above
for compound 4. However, despite an extended reaction time,
we could only obtain compound 8 as a 9:1 mixture of isomers
favoring the desired ∆8(9) isomer.
Despite these complications, we were pleased to find that
the cyclization worked very well with more electrophilic
(7) (a) Cox, E. D.; Hamaker, L. K.; Li, J.; Yu, P.; Czerwinski, K. M.;
Deng, L.; Bennet, D. W.; Cook, J. M. J. Org. Chem. 1997, 62, 44. (b)
Ungemach, F.; DiPierro, M.; Weber, R.; Cook, J. M. J. Org. Chem. 1981,
46, 164. (c) Soerens, D.; Sandrin, J.; Ungemach, F.; Mokry, P.; Wu, G. S.;
Yamanaka, E.; Hutchins, L.; Dipierro, M.; Cook, J. M. J. Org. Chem. 1979,
44, 535. (d) Jackson, A. H.; Smith, A. E. Tetrahedron 1968, 24, 403. (e)
Jackson, A. H.; Naidoo, B.; Smith, P. Tetrahedron 1968, 24, 6119. (f)
Kawate, T.; Nakagawa, M.; Ogata, K.; Hino, T. Heterocycles 1992, 33,
801. (g) Ungemach, F.; Cook, J. M. Heterocycles 1978, 9, 1089 and
references cited therein.
(8) (a) Chi, D. Y.; Katzenellenbogen, J. A. J. Am. Chem. Soc. 1993,
115, 7045. (b) Chi, D. Y.; O’Neil, J. P.; Anderson, C. J.; Welch, M. J.;
Katzenellenbogen, J. A. J. Med. Chem. 1994, 37, 928. (c) Hom, R. K.;
Katzenellenbogen, J. A. Nucl. Med. Biol. 1997, 24, 485. (d) Skaddan, M.
B.; Katzenellenbogen, J. A. Bioconjugate Chem. 1999, 10, 119. (e) Skaddan,
M. B.; Wust, F. R.; Katzenellenbogen, J. A. J. Org. Chem. 1999, 64, 8108.
(9) If one considers that a p-methoxystyrene is a trienol system and that
reactions of enols with imines are termed Mannich reactions, then the
reaction described in this report could be considered a vinylogous Mannich
reaction. However, this term is currently employed by Martin to illustrate
the nucleophilic addition of 2-trialkylsiloxy furans to cyclic iminium ions;
cf. Martin, S. F. et al. J. Am. Chem. Soc. 1996, 118, 3299 and Martin, S.
F. et al. J. Am. Chem. Soc. 1999, 121, 6990. It should also be noted that
Overman has described a related intramolecular Mannich reaction; cf.
Loegers, M. et al. J. Am. Chem. Soc. 1995, 117, 9139 and references cited
therein.
(10) Refer to ref 3b-d.
(11) Refer to ref 4.
(12) Geier, M.; Hesse, M. Synthesis 1990, 56.
(13) Similar temperature affects have been observed in the Pictet-
Spengler reaction of tryptamines carried out in nonacidic, aprotic media;
cf. ref 7c.
Robust substrates were needed to withstand cyclization
under the rather harsh conditions described above. In fact,
(14) Luettrinhaus, A. Justus. Liebigs Ann. Chem. 1963, 661, 84.
Org. Lett., Vol. 2, No. 23, 2000
3636

several attempts to achieve cyclization using activated
ketones under these conditions were unsuccessful. Specifi-
cally, when we attempted the reaction with either a â-
ketoester or â-keto-γ-lactone, these substrates were degraded
during the reaction. Therefore, we sought to develop a milder
protocol for use with these acid-sensitive functionalities.
We were aware of previous work in the tetrahydro-â-
carboline literature, which showed that indeed esters, ketones,
and amides with a â-carbonyl functionality could be cyclized
by acid catalysts (Scheme 2).15 The two-step protocol
Table 2. Cyclization with Activated Ketones and Alkynesa
Scheme 2. Two-Step Synthesis of Tetrahydro-â-carbolines
a The reactions shown in this table were performed using a two-step
protocol similar to that shown in Scheme 2. The enamine intermediates
were used directly in the subsequent cyclization step, but could be isolated
by column chromatography if desired. b The time refers to the duration with
which enamine formation was allowed to occur. A reaction time of 30
minutes was used for all subsequent cyclizations; neat TFA, 0 °C. c Yields
are combined for the two-step process. Yields in parentheses are for the
enamine formation.
reported in this work consisted of initial enamine formation
at room temperature, followed by treatment with acid at low
temperature. With this excellent precedent in hand, we
proceeded to submit the substrates we desired to the
conditions described above. The results of these studies are
summarized in Table 2.
results of the one-step protocol reported above, under these
milder two-step conditions, the desired ∆8(9) isomer was
produced exclusively. We also found that anhydrous TFA
was required for this procedure, because small amounts of
water in the reagent caused extensive hydrolysis of the
enamines during the cyclization. Unlike what happened in
our earlier attempts, we observed no ring opening of the
lactone during the formation of the spiro compound (11) by
the two-step procedure.
Amines are known to undergo conjugate addition with
activated alkynes to yield the corresponding enamines.19 In
fact, the use of alkynes in the Pictet-Spengler reaction has
been previously documented in the synthesis of several
tetrahydro-â-carbolines.20 Upon the basis of these precedents,
we reasoned that the alkynes methyl propiolate and dimethyl
acetylenedicarboxylate (DMAD) would be suitable substrates
for the vinylogous Pictet-Spengler reaction. In the case of
methyl propiolate, we found enamine formation to be
exceedingly slow. In fact, reasonable yields of the conjugate
product were only obtained after 120 h at room temperature.
The enamine was isolated as a mixture of Z:E isomers in a
3:1 ratio, determined by integration of the signals of the vinyl
hydrogen atoms. Although the yield was modest, this
enamine underwent smooth cyclization upon treatment with
TFA to afford the hexahydrobenzo[f]isoquinoline (12) in 24%
overall yield. As expected, conjugate addition with the more
When a solution of 1‚HCl and triethylamine in DMF is
treated with diethyl 1,3-acetonedicarboxylate at room tem-
perature for 12 h, the desired enamine could be isolated as
a single isomer. Although we did not rigorously determine
the geometry of the intermediate enamine for compounds 9,
12, and 13 (see below), we believe that it is Z, on the basis
of earlier reports.16 In this geometry, the enamino ester
system would be stabilized via intramolecular hydrogen
bonding, and there is evidence for such an interaction by
the extreme downfield shift of the amine proton in the NMR
(δ 8.58, 7.83, and 8.09, respectively). In contrast, in the case
of the intermediate enamine resulting from reaction with
dihydropyran-2,4-dione,17 the cyclic geometry aminodihy-
dropyran requires that the geometry of the enamine be E,
which is supported by the upfield location (δ 4.56) of the
hydrogen atom.
The cyclization of these enamines was effected by treat-
ment of the neat oils with excess trifluoroacetic acid at 0 °C
for 30 min.18 The yields of these compounds are given in
Table 2. We were gratified to find that, in contrast to the
(15) Kirkpatrick, A.; Maclaren, J. A. Aust. J. Chem. 1983, 36, 833.
(16) (a) Barillier, D.; Strobel, M. P.; Morin, L.; Paquer, D. Tetrahedron
1983, 39, 767. (b) Kozerski, L.; Kawecki, R.; Hansen, P. E. Magn. Reson.
Chem. 1994, 32, 517.
(17) D’Angelo, J.; Gomez-Pardo, D. Tetrahedron Lett. 1991, 32, 3063.
(18) TFA is known to cause C(1)-N(2) bond cleavage in methoxy-
substituted tetrahydro-â-carbolines through a carbocation-mediated mech-
anism; cf. ref 7a and Reedy, S. M.; Cook, J. M. Tetrahedron Lett. 1994,
35, 5413.
(19) Henin, J.; Vercauteren, J.; Mangenot, C.; Henin, B.; Nuzillard, J.
M.; Guilhem, J. Tetrahedron 1999, 55, 9817.
(20) (a) Bailey, P. D.; Hollinshead, S. P.; Dauter, Z. J. Chem. Soc., Chem.
Commun. 1985, 21, 1507. (b) Vercauteren, J.; Lavaud, C.; Levy, J.; Massiot,
G.; Fac. Pharm, R. F. J. Org. Chem. 1984, 49, 2278.
Org. Lett., Vol. 2, No. 23, 2000
3637

reactive DMAD was complete after only 24 h. The enamine
product was of single configuration, and when submitted to
the cyclization conditions, provided the desired tricyclic
product (13) in 35% overall yield.
Spengler cyclization literature provides for further extension
of this reaction through a rational choice of electrophiles.
Acknowledgment. We are grateful for support of this
research through grants from the Department of the Army
(DAMD17-97-1-7292), National Cancer Institute DHHS
(PHS 5 T32 CA09067), and the Department of Energy (DE
FG02 86ER60401).
The results of this brief investigation of a vinylogous
Pictet-Spengler reaction have been encouraging. Not un-
expectedly, both aldehydes and ketones can be cyclized using
standard Pictet-Spengler conditions with relative ease, and
we found that high yields are obtained only when cyclic
substrates are used. The harsh conditions required to cyclize
unactivated ketones were not suitable for more acid-sensitive
substrates, but a milder, two-step protocol was developed
for these cases. Suitable enamine cyclization precursors could
be prepared by condensation with activated ketones or by
conjugate addition with activated alkynes. The vast Pictet-
Supporting Information Available: Complete experi-
mental procedures and spectral data for compounds 1-9 and
11-13. This material is available free of charge via the
Internet at http://pubs.acs.org
OL006533U
3638
Org. Lett., Vol. 2, No. 23, 2000