COMMUNICATIONS
[10] G. Dyker, Angew. Chem. 1994, 106, 117; Angew. Chem. Int. Ed. Engl.
1994, 33, 103 ± 105.
systems, little is known about the biosynthesis of sponge
metabolites.[2] One class of cytotoxic sponge metabolites
which have recently fascinated organic chemists are the
manzamine alkaloids. The first member of this class, man-
zamine A (1, Figure 1), was isolated in 1986 by Higa et al.[3]
[11] a) R. W. Hoffmann, Dehydrobenzene and Cycloalkynes, Academic
Press, New York, 1967; b) H. Hart in The Chemistry of Functional
Groups, Suppl. C2: The Chemistry of Triple-Bonded Functional
Groups (Ed.: S. Patai), Wiley, Chichester, 1994, pp. 1017 ± 1134.
[12] Y. Himeshima, T. Sonoda, H. Kobayashi, Chem. Lett. 1983, 1211 ±
1214.
[13] Prepared in our laboratory by a published procedure: ªPalladium in
Organic Synthesisº: L. S. Hegedus in Organometallics in Synthesis: A
Manual (Ed.: M. Schlosser), Wiley, 1994. Commercially available
material gave poorer yields.
[14] a) L. S. Meriwether, E. C. Colthup, G. W. Kinnerly, R. N. Reusch, J.
Org. Chem. 1961, 26, 5155 ± 5163; b) Y. Wakatsuki, O. Nomura, K.
Kitaura, K. Morokuma, H. Yamazaki, J. Am. Chem. Soc. 1983, 105,
1907 ± 1912.
N
H
N
H
N
H
N
H
N
N
H
OH
O
N
N
N
H
NH+Cl–
NH
[15] a) N. Boden, R. J. Bushby, A. N. Cammidge, S. Duckworth, G.
Headdock, J. Mater. Chem. 1997, 7, 601 ± 605; b) K. Praefcke, A.
Eckert, D. Blunk, Liq. Cryst. 1997, 22, 113 ± 119.
1
2
3
Figure 1. Manzamine A (1), B (2), and C (3).
[16] Triphenylene and the substituted derivatives gave correct analytical
and spectroscopic data. 3: M.p. 1948C (literature value[6a] 1988C). 7:
1H NMR ([D8]THF) d 8.42 (t, J 10.3 Hz, 6H); MS: m/z (%): 336
(100), 168 (16); HR-MS: calcd for C18H6F6: 336.0374, found: 336.0364.
and recently synthesized.[12] The unprecedented structure led
the authors to the conclusion that ªno obvious biogenetic
pathº could be envisaged leading to 1. Manzamines B (2) and
C (3) were subsequently isolated from the same sponge.[4]
In 1992, we put forward a biogenetic hypothesis for the
formation of the manzamines.[5] We proposed that each
structure could be reduced into four building blocks: ammo-
nia, a C10 unit (a symmetrical dialdehyde), tryptophan, and a
C3 unit (an acrolein equivalent), shown in Scheme 1 for
manzamine B (2). The key step in the proposal is the
intramolecular endo Diels ± Alder cycloaddition of the bis-
dihydropyridine 4.[6] To date it is not known whether a
ªDiels ± Alderaseº exists.[7]
1
11: M.p. 1888C; H NMR (CDCl3) d 9.04 (dd, J 8.5, 0.8 Hz, 1H),
8.10 (d, J 8.1 Hz, 1H), 8.00 (d, J 8.1 Hz, 1H), 7.48 (m, 3H), 4.04 (s,
3H), 3.97 (s, 6H); 13C NMR (CDCl3) d 158.2, 157.9, 157.0, 133.0,
132.6, 131.9, 127.3, 127.0, 126.8, 121.1, 118.8, 118.4, 118.1, 115.9, 115.0,
110.4, 108.7, 108.5, 56.1, 55.8, 55.7; MS: m/z (%): 318 (100), 303 (25),
288 (15); HR-MS: calcd for C21H18O3: 318.1256, found: 318.1268. 12:
1H NMR (CDCl3) d 9.00 (dd, J 8.5, 1.2 Hz, 3H), 7.48 (t, J 8.2 Hz,
3H), 7.14 (dd, J 8.1, 1.0 Hz, 3H), 4.02 (s, 9H); MS: m/z (%): 318
(100), 303 (22), 288 (30).
Since the publication of the hypothesis a large number of
manzamine and related alkaloids have been isolated from
various species of sponge worldwide.[8] Despite the lack of
experimental evidence, the proposal has been applied repeat-
edly to explain the biogenetic origin of the manzamine and
related alkaloids. One related alkaloid is keramaphidin B (5,
Scheme 1), which was isolated independently by both the
Kobayashi and Andersen groups.[9] Structurally 5 is simply the
reduced form of the proposed cycloadduct 6 (Scheme 1).
Herein, we report the biomimetic synthesis of 5, the first in
vitro chemical evidence for this proposal.
The synthesis of 7 was first communicated in 1996,[6e] but we
found the route unsatisfactory because of moderate yields
(7% overall) and the instability of one intermediate. Since
then we have modified the synthesis (Scheme 2) with
significant improvements (37% overall yield). Hydroxyphos-
phonium salt 8 was masked as its tetrahydropyranyl (THP)
derivative 9 (93%). Olefin 10 was obtained from the ylide
generated from 9 and 3-(3-pyridyl)propanal in 83% yield.
Acid-mediated deprotection gave the alcohol 11 (94%),
which was treated with p-toluenesulfonyl chloride to give 12
(95%). A one-pot Finkelstein reaction, dimerization and
macrocyclization was effected by the slow addition of 12 into a
mixture of NaI in 2-butanone under reflux. The crude product
was reduced to give bis-tetrahydropyridine 13 in 56% yield
over the two steps. Oxidation of 13 with 3-chloroperbenzoic
acid (mCPBA) furnished diastereomeric N-oxides (98%),
which could be treated with trifluoroacetic anhydride to give
bis-dihydropyridine 7 (100%).
Investigations into the Manzamine Alkaloid
Biosynthetic Hypothesis**
Jack E. Baldwin,* Timothy D. W. Claridge,
Andrew J. Culshaw, Florian A. Heupel, Victor Lee,
David R. Spring, Roger C. Whitehead,
Robert J. Boughtflower, Ian M. Mutton, and
Richard J. Upton
Over the past decade there has been an upsurge in the
discovery of biologically active natural products from marine
sponges.[1] In comparison to terrestrial plant and microbial
[*] Prof. J. E. Baldwin, Dr. T. D. W. Claridge, Dr. A. J. Culshaw,
Dr. F. A. Heupel, Dr. V. Lee, D. R. Spring, Dr. R. C. Whitehead
The Dyson Perrins Laboratory
University of Oxford
South Parks Road, Oxford OX1 3QY (UK)
Fax: (44)1865-275-632
R. J. Boughtflower, I. M. Mutton, Dr. R. J. Upton
Glaxo Wellcome Research and Development
Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY (UK)
[**] This work was supported financially by the EPSRC (A.J.C and D.R.S)
and the Rhodes Trust (F.A.H). We are indebted to Prof. Raymond J.
Andersen (Department of Chemistry, University of British Columbia,
Canada) and Prof. Junꢁichi Kobayashi (Faculty of Pharmaceutical
Sciences, Hokkaido University, Japan), for their generous gifts of
keramaphidin B. We also thank the EPSRC mass spectrometry
service (Swansea) for high resolution mass measurements.
Angew. Chem. Int. Ed. 1998, 37, No. 19
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3719-2661 $ 17.50+.50/0
2661