Tandem photocycloaddition - Retro-Mannich fragmentation of enaminones. A route to spiropyrrolines and the tetracyclic core of koumine

Article abstract of DOI:10.1021/ol052955y

Intramolecular [2 + 2] photocycloaddition of β-aminoalkylidene malonates gives transiently a cyclobutane which undergoes retro-Mannich fragmentation to a Δ1-pyrroline. The tandem sequence, exemplified in two series based on tryptamine and aminoethyl-1,4-c

Full text of DOI:10.1021/ol052955y

ORGANIC
LETTERS
2006
Vol. 8, No. 6
1081-1084
Tandem
Photocycloaddition−Retro-Mannich
Fragmentation of Enaminones. A Route
to Spiropyrrolines and the Tetracyclic
Core of Koumine
James D. White* and David C. Ihle
Department of Chemistry, Oregon State UniVersity, CorVallis, Oregon 97331-4003
james.white@oregonstate.edu
Received December 6, 2005
ABSTRACT
Intramolecular [2
fragmentation to a
leads to a spiroindolopyrroline skeleton and to the nonindolenine portion of koumine.
+
∆
2] photocycloaddition of
â-aminoalkylidene malonates gives transiently a cyclobutane which undergoes retro-Mannich
1-pyrroline. The tandem sequence, exemplified in two series based on tryptamine and aminoethyl-1,4-cyclohexadiene,
Enaminones (vinylogous amides) are photochemically reac-
tive, and when irradiated they undergo [2 + 2] cycloaddition
with an alkene to generate a cyclobutane.1,2 The intramo-
lecular variant of this photocycloaddition (Scheme 1) is also
Mannich reaction, with the result that a ∆1-pyrroline (3) is
formed in a stereodefined manner.
The overall conversion 1 f 3 is the intramolecular
nitrogen counterpart of a process first described by De Mayo,
in which an enolic â-diketone undergoes [2 + 2] photocy-
cloaddition with an alkene followed by retro-aldol fragmen-
tation of the intermediate cyclobutane to produce a 1,5-
diketone.6 An early example of the process shown in Scheme
1 was reported by Schell, who showed that irradiation of 4
through Corex glass led initially to tetracycle 5 and then to
the cyclooctanone 6 (Scheme 2).7 Application of this tandem
sequence has been made by Winkler in syntheses of several
alkaloids, including mesembrine,8 perhydrohistrionicotoxin,9
vindorozine,10 and manzamine A,11 and Swindell has used
the method for construction of the taxane BC subunit.12
Scheme 1
known.3-5 Ring strain generated in the conversion of 1 to 2
can lead to fragmentation of the cyclobutane via a retro-
(6) De Mayo, P. Acc. Chem. Res. 1971, 4, 41.
(7) Schell, F. M.; Cook, P. M. J. Org. Chem. 1984, 49, 4067.
(8) Winkler, J. D.; Muller, C. L.; Scott, R. D. J. Am. Chem. Soc. 1988,
110, 4831.
(1) Bohme, E. H.; Valenta, Z,; Wiesner, K. Tetrahedron Lett. 1965, 2441.
(2) Cantrell, T. S. Tetrahedron 1971, 27, 1227.
(3) Wiesner, K.; Musil, V.; Wiesner, K. J. Tetrahedron Lett. 1968, 5643.
(4) Tamura, Y.; Ishibashi, H.; Hirai, M.; Kita, Y.; Ikeda, M. J. Org.
Chem. 1975, 40, 2702.
(5) Schell, F. M.; Cook, P. M.; Hawkinson, S. W.; Cassady, R. E.;
Thiessen, W. E. J. Org. Chem. 1979, 44, 1380.
(9) Winkler, J. D.; Hershberger, P. M. J. Am. Chem. Soc. 1989, 111,
4852.
(10) Winkler, J. D.; Scott, R. D.; Williard, P. G. J. Am. Chem. Soc. 1990,
112, 8971.
(11) Winkler, J. D.; Axten, J. M. J. Am. Chem. Soc. 1998, 120, 6425.
10.1021/ol052955y CCC: $33.50
© 2006 American Chemical Society
Published on Web 02/21/2006

Scheme 2
Scheme 3
However, the scope of the intramolecular photocycloaddi-
tion-retro-Mannich process has received less attention than
the De Mayo reaction despite the opportunities it presents
for preparing a diverse set of nitrogen heterocycles with good
stereocontrol.
We now report the results of a study which demonstrate
that the intramolecular [2 + 2] photocycloaddition-retro-
Mannich construct can be applied to substrates based on
tryptamine and â-phenethylamine templates. Among other
attributes, this chemistry has provided entry to the core
structure of the indolenine alkaloid koumine (7).13
Our initial studies focused on the intramolecular photo-
cycloaddition of â-amino-substituted alkylidene malonates
on the premise that a sequence akin to that in Scheme 1
would yield a product in which the ester groups could be
differentiated and therefore modified in selective fashion.
The protected tryptamine 8 was condensed with diethyl
â-ethoxymethylidene malonate (9) under basic conditions to
afford 10,14 which was irradiated through Corex glass with
a 450 W Hanovia mercury lamp (Scheme 3). After 7 h,
reaction was complete and the spiropyrroline 11 was isolated
in high yield. The intermediate tetracycle 12 could not be
detected in this reaction, presumably because retro-Mannich
fragmentation occurred as soon as 12 was formed, but a
subsequent experiment provided circumstantial evidence that
12 was indeed the initial product from 10 (vide infra). Only
a single stereoisomer of 11 was produced from 10, reflecting
the uniquely defined configuration around the tetrasubstituted
cyclobutane of 12. This result appears to eliminate a stepwise
radical mechanism initiated by attack of the photoexcited
alkylidene malonate at the indole 3-position.
methyl substituent into 11 was accomplished by N-alkylation
with methyl iodide followed by reduction of the resulting
iminium iodide with sodium borohydride to furnish 14.
Removal of the Boc protection from 14 with trifluoroacetic
acid afforded a good yield of 15, but also led to a minor
amount of spiroimine 16 in which the malonate residue was
absent. The latter product is apparently the result of a further
retro-Mannich fragmentation of 15.
A useful extension of the sequence shown in Scheme 3
would be incorporation of an alkyl substituent at C2 of the
spiropyrrolidine 15 since this could potentially offer a new
route to the family of Strychnos alkaloids17 that includes,
for example, akuammicine (17).18 To that end, protected
tryptamine 8 was condensed with alkylidene malonate 18 to
give 19 (Scheme 4). Irradiation of 19 through Pyrex again
afforded a single spiroimine 20, presumably via cyclobutane
21. Reduction of 20 with sodium cyanoborohydride yielded
22 as a single epimer resulting from delivery of hydride to
the less hindered face of the imine.
Alkaloids of the spiroindolopyrrolidine family,15 such as
the oxindole coerulescine 13,16 generally bear a methyl
substituent on the pyrrolidine nitrogen. Introduction of this
(12) Swindell, C. S.; Patel, B. P.; DeSolms, S. J.; Springer, J. P. J. Org.
Chem. 1987, 52, 2346.
(13) (a) Liu, C.; Wang, Q.; Wang, C. J. Am. Chem. Soc. 1981, 103,
4634. (b) Khuong-huu, F.; Chiaroni, A.; Riche, C. Tetrahedron Lett. 1981,
22, 733. (c) Zhujin, L.; Qianshong, Y. Youji Huaxu 1986, 1, 36.
(14) Momose, T.; Tanaka, T.; Yokota, T.; Nagamoto, N.; Yamada, K.
Chem. Pharm. Bull. 1978, 26, 2224.
Our strategy for assembling the nonindolenine portion 23
of koumine (7) envisioned cyclization of octahydroisoquino-
(15) For a review, see: Marti, C.; Carreira, E. M. Eur. J. Org. Chem.
2003, 2209.
(16) Anderton, N.; Cockrum, P. A.; Colegate, S. M.; Edgar, J. A.; Flower,
K.; Vit, I.; Willing, R. I. Phytochemistry 1998, 48, 437.
(17) For a review of Strychnos alkaloids: Bosch, J.; Bonjoch, J.; Amat,
M. The Strychnos Alkaloids; Academic: New York, 1996; pp 75-189.
(18) (a) Millson, P.; Robinson, R.; Thomas, A. F. Experientia 1953, 9,
89. (b) Edwards, P. N.; Smith, G. F. J. Chem. Soc. 1961, 152.
1082
Org. Lett., Vol. 8, No. 6, 2006

Scheme 4
Scheme 6
vinyl substituent needed for 24. The basic pyrrolidine
nitrogen of 31 was alkylated with methyl iodide in the
expectation that the resultant quaternary iodide 32 would
exist in equilibrium with acyliminium iodide 33 (Scheme
7). In practice, reduction with sodium borohydride in situ
after treatment of 31 with methyl iodide afforded tertiary
amine 34 in excellent yield. Conversion of 34 to its N-oxide
line 24, which we believed could be constructed via a
photocycloaddition-retro-Mannich approach similar to that
developed with tryptamine (Scheme 5). In this case, our
Scheme 5
Scheme 7
starting point was the Birch reduction product 2519 of
â-phenethylamine, which was condensed with 9 to furnish
26 (Scheme 6). Irradiation of 26, as expected, produced
spiroimine 27 (61% yield), but it was found that the yield
of this process could be improved if 26 was first converted
to its Boc derivative 28. When the latter was irradiated
through Corex rather than Pyrex glass, the cyclobutane 29
could be isolated in nearly quantitative yield. Subsequent
removal of the Boc protection from 29 caused spontaneous
retro-Mannich fragmentation to 27. To activate 27 for
cyclization to a precursor for 24, the imine nitrogen was first
methylated and the resultant methiodide 30 was treated with
methylamine in warm toluene. The product 31, obtained in
high yield, was the result of initial formation of an aminal
followed by intramolecular acylation by one of the two ethyl
esters to give a δ-lactam.
With 31, there was now an opportunity to employ the fused
pyrrolidine of this tricycle as the progenitor of the angular
(19) Wimalasena, K.; Alliston, K. R. J. Am. Chem. Soc. 1995, 117, 1220.
Org. Lett., Vol. 8, No. 6, 2006
1083

It was now necessary to invert the configuration of the
ester carbon of 36 in order to set in place an endo primary
alcohol that could be used to construct the tricyclic ether
23. This was accomplished by saponification of 36, conver-
sion of the resulting carboxylic acid to its acyl chloride, and
reduction with diisobutylaluminum hydride, a sequence that
avoided unwanted reduction of the lactam carbonyl (Scheme
8). The mixture of primary alcohols 38 produced in this
manner22 was isomerized to endo alcohol 39 via its dianion
followed by reprotonation at low temperature. Treatment of
39 with dimethyldioxirane23 afforded exo epoxide 40 with
only a trace of product resulting from epoxidation of the
vinyl group, and exposure of 40 to camphorsulfonic acid
resulted in clean cyclization to cylic ether 41. This alcohol
was oxidized24 in situ to ketone 42, which now stands ready
for homologation of the lactam carbonyl, Fischer indoliza-
tion, and final intramolecular alkylation at the indole â carbon
to reach koumine.25,26
Scheme 8
Acknowledgment. Financial support from the National
Institute of General Medical Sciences through Grant No.
GM058889 is gratefully acknowledged.
Supporting Information Available: Experimental pro-
cedures, characterization data, and representative 1H and 13
C
NMR spectra for new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
OL052955Y
(22) The stereochemistry of the resulting alcohols was determined by
cyclization with mercury(II) acetate. The exo primary alcohol underwent
intramolecular oxymercuration exlusively at the vinyl group, whereas 39
yielded the product of oxymercuration at the cyclohexene double bond.
(23) Murray, R. W.; Singh, M. Org. Synth. 1997, 91.
(24) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(25) For a total synthesis of (+)-koumine, see: (a) Magnus, P.; Mugrage,
B.; Deluca, M.; Cain, G. A. J. Am Chem. Soc. 1989, 111, 786. (b) Magnus,
P.; Mugrage, B.; Deluca, M.; Cain, G. A. J. Am. Chem. Soc. 1990, 112,
5220.
35 was straightforward, but Cope elimination20 of 35 in hot
DMF yielded isoquinoline 36 accompanied by a significant
quantity of 34 resulting from deoxygenation.21 The latter was
recycled through 35 to augment the yield of 36.
(20) Cope, A. C.; Trumbull, E. R. Org. React. 1960, 11, 317.
(21) Crabb, T. A.; Wilkinson, J. R. J. Chem. Soc., Perkin Trans. 1 1975,
1465.
(26) For a formal synthesis of (+)-koumine, see: (a) Bailey, P. D.;
McLay, N. R. Tetrahedron Lett. 1991, 32, 3895. (b) Bailey, P. D.; McLay,
N. R. J. Chem. Soc., Perkin Trans. 1 1993, 441.
1084
Org. Lett., Vol. 8, No. 6, 2006