Total synthesis of (+/-)-isoschizogamine.

Article abstract of DOI:10.1021/ol9902632

[formula: see text] Isoschizogamine has been prepared for the first time. The synthesis requires eight steps from a readily available ketone starting material and features an aminal-forming cyclization that is based on a proposed biosynthetic transformati

Full text of DOI:10.1021/ol9902632

ORGANIC
LETTERS
1999
Vol. 1, No. 8
1315-1317
Total Synthesis of (±)-Isoschizogamine
Jed L. Hubbs and Clayton H. Heathcock*
Department of Chemistry, UniVersity of California, Berkeley, California 94720
chh@steroid.cchem.berkeley.edu
Received September 1, 1999
ABSTRACT
Isoschizogamine has been prepared for the first time. The synthesis requires eight steps from a readily available ketone starting material and
features an aminal-forming cyclization that is based on a proposed biosynthetic transformation.
Isoschizogamine (1), a member of the schizozygane family
of indole alkaloids, was recently isolated by Hajicek and co-
workers from the shrub Schizozygia caffaeoides.1,2 Its
structure, which differs from the ring system typically found
in the schizozygane alkaloids (e.g., 2),3 was elucidated by
NMR techniques. The most striking difference is the presence
of the aminal motif. Hajicek and co-workers proposed a
partial biosynthesis to explain the formation of the unusual,
aminal-containing hexacyclic ring system from an intermedi-
ate with the more typical schizozygane skeleton, 2 (Scheme
1). Attack of the indoline nitrogen on the immonium carbon
to mimic the proposed biosynthetic formation of the aminal
function by modification of compounds we had previously
employed for our total synthesis of vallesamidine.4 In the
vallesamidine synthesis, lactam 5 was prepared in four steps
from 2-ethylcyclopentanone (Scheme 2). It can be seen that
Scheme 2
Scheme 1
would lead to the aziridinium intermediate 3, which might
be opened reductively and the lactam ring closed to provide
isoschizogamine (1).
We saw the novel structure of isoschizogamine as an
interesting synthetic target and thought that we might be able
(1) Ha´j´ıcek, J.; Taimr, J.; Budes´ınsky, M. Tetrahedron Lett. 1998, 39,
505.
(2) Renner, U. Lloydia 1964, 27, 406.
protonation of the enamide double bond in 5 would give an
acyl immonium species that is similar to intermediate 2 in
(3) Kompis, I.; Hesse, M.; Schmid, H. Lloydia 1971, 34, 269.
10.1021/ol9902632 CCC: $18.00 © 1999 American Chemical Society
Published on Web 09/18/1999

the proposed biosynthesis. The only major difference be-
tween the ring system in such an acyl immonium intermedi-
ate and biosynthetic intermediate 2 is that the five-membered
indoline ring is closed in 2. Addition of the aniline nitrogen
in 5 to the immonium carbon formed by protonation of the
enamide double bond would yield the isoschizogamine
framework.
This idea was used to model a key transformation in the
total synthesis of isoschizogamine. Because enamide 5 did
not undergo the desired cyclization, it was reduced with
lithium aluminum hydride to tertiary enamine 6. Upon
workup,5 enamine 6 was protonated to give immonium ion
7, which cyclized to form diastereomeric aminals 8 and 9.
The formation of the two observed diastereomers can be
explained by protonation of the two diastereotopic faces of
the double bond in enamine 6. The stereochemistry indicated
in structure 8 was established by single-crystal X-ray
diffraction. Treatment of either isomer with acetic acid
resulted in an 85:15 ratio of 8 to 9.
3-azidopropanal7 in an aldol reaction. There is disagreement
in the literature as to whether 2-substituted cyclopentanones
can be deprotonted regioselectivly at the more substituted
R-position.8 However, it was found that the thermodynamic
enolate could be formed with high selectivity by slowly
adding a slight excess of KHMDS to 13 over 2 h. To obtain
high regioselectivity and diastereoselectivity in the aldol
reaction, the enolate of a metal other than potassium was
needed. Consequently, a number of Lewis acids were
examined for trapping the potassium enolate. Eventually we
found that treatment of the potassium enolate with Bu2BOTf
and subsequent aldol reaction gave exclusively syn aldol 14.
The use of i-Bu2AlCl gave a slightly higher yield of adduct
14, but with this Lewis acid there was produced a mixture
of diastereomers that was difficult to separate. Hydrogenation
of the azide group in aldol 14 gave imine 12.
Unfortunately, the Michael addition reaction of imine 12
to R,â-unsaturated acid 11 which was directly analogous to
that used in our vallesamidine synthesis (Scheme 3) gave
none of the desired product. Changing the Michael acceptor
to the acid chloride,9,10 azide,9,11 anhydride,9 and ester also
all failed to give the desired addition-cyclization product.
We then turned our attention to R,â-unsaturated dicarbonyl
compounds as the Michael acceptors in this reaction. A
search of the literature indicated that Meldrum’s acid
derivatives of aldehydes such as 16 (Scheme 5) were
excellent Michael acceptors and also had the advantage that
the adducts formed from the Michael addition to these
substrates could serve as good acylating agents.12 Thus, R,â-
unsaturated diester 16 was prepared by Knovenagel conden-
sation of 2-nitroveratraldehyde with Meldrum’s acid (Scheme
5). We then utilized this Michael acceptor in our imine
addition-cyclization reaction. Imine 12 underwent Michael
addition to 16 at -40 °C to give an intermediate which, upon
heating, underwent cyclization with concomitant loss of
acetone and carbon dioxide, providing a mixture of diaster-
eomeric lactams 17 and 18 in a ratio of 88:12 and a total
yield of 74% from aldol 14. Dehydration of 17 with Martin’s
sulfurane13 gave alkene 19.
With confidence that the basic aminal-forming reaction
would work, we set out to prepare isoschizogamine via an
appropriately functionalized version of lactam 5 such as 10
(Scheme 3). Lactam 10 was envisioned as coming from the
Scheme 3
Michael addition of the enamine tautomer of imine 12 to
cinnamic acid 11 in a reaction similar to that used in our
vallesamidine synthesis.4
As shown in Scheme 4, the thermodynamic enolate of the
readily available ketone 136 was allowed to react with
The reduction-cyclization sequence was carried out under
conditions similar to those used in the model system (Scheme
2). The aromatic nitro group in lactam 19 was first reduced
using NaBH4 and catalytic Cu(acac)2 (Scheme 6) to give
compound 20. The air and acid sensitivity of 20 made it
14
Scheme 4
1316
Org. Lett., Vol. 1, No. 8, 1999

Scheme 5
Scheme 7
crude product was then oxidized with pyridinium dichro-
mate15 to give (()-isoschizogamine (1) in 27% yield for the
four steps from 19. None of the other possible diastereomer
was observed in the NMR spectrum of the crude oxidation
product.
In summary, isoschizogamine was prepared from ketone
13 in eight steps and 7% overall yield.
Acknowledgment. This research was supported by a
research grant from the National Institutes of Health
(GM46057).
Scheme 6
Supporting Information Available: Experimental pro-
cedures and full characterization for compounds 8, 9, 12,
14-19, and 1; 1H NMR spectra for compounds 8, 9, and 1;
13C NMR spectrum for compound 1. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL9902632
(4) (a) Dickman, D. A.; Heathcock, C. H. J. Am. Chem. Soc. 1989, 111,
1528. (b) Heathcock, C. H.; Norman, M. H.; Dickman, D. A. J. Org. Chem.
1990, 55, 798.
(5) Feiser, L. F.; Feiser, M. Reagents for Organic Synthesis, Vol. 1;
Wiley: New York, 1967; p 584.
(6) Canole, G.; Mears, R. J.; DeSilva, H.; Whiting, A. J. Chem. Soc.,
Perkin Trans. 1 1995, 14, 1825.
(7) Davies, A. J.; Donald, S. R.; Marks, R. E. J. Chem. Soc. C 1967,
2109.
(8) (a) Stork, G.; Hudrlik, P. F. J. Am. Chem. Soc. 1968, 90, 4462. (b)
House, H. O.; Kramar, V. J. Org. Chem. 1963, 28, 3362. (c) House, H. O.;
Czuba, L. J.; Gall, M.; Olmstead, H. D. J. Org. Chem. 1969, 34, 2324. (d)
Negishi, E.; Chatterjee, S. Tetrahedron Lett. 1983, 24, 1341.
(9) Paulvannan, K.; Stille, J. R.; J. Org. Chem. 1992, 57, 5319.
(10) Cado, F.; Bazureau, J. P.; Hamelin, Bull. Soc. Chim. Belg. 1996,
105, 273.
(11) Danieli, B.; Lesma, G.; Palmisano, G.; Tollari, S. Synthesis 1984,
353.
(12) (a) McNab, H. Chem. Soc. ReV. 1978, 7, 345. (b) Quiroga, J.;
Hormaza, A.; Insuasty, B.; Ma´rquez, M. J. Heterocycl. Chem. 1998, 35,
409.
(13) Martin, J. C.; Arhart, R. J. J. Am. Chem. Soc. 1971, 93, 4327.
(14) Murphy, J. A.; Rasheed, F.; Gastaldi, S.; Ravishanker, T.; Lewis,
N. J. Chem. Soc., Perkin Trans. 1 1997, 1549.
(15) Takanos, S.; Yonaga, M.; Morimoto, M.; Ogasawara, K. J. Chem.
Soc., Perkin Trans. 1 1985, 305.
difficult to handle and purify so it was used without
purification in the next step. The lactam carbonyl was
reduced with LiAlH4, which upon workup gave aminal 21
as a single diastereomer. Treatment of this aminal with acetic
acid and water under conditions insufficient to hydrolyze
the dioxolane acetal gave a 3:7 ratio of diastereomer 21 to
22. The stereochemistry was determined in these cases by
comparison of the H NMR spectra of 21 and 22 to the
spectra of aminals 8 and 9.
Treatment of 21 with aqueous acetic acid under more
forcing conditions provided hemiaminal 23 (Scheme 7). This
1
Org. Lett., Vol. 1, No. 8, 1999
1317