Notes and references
1 Isolation: (a) Ascididemin: J. Kobayashi, J. Cheng, H. Nakamura, Y.
Ohizumi, Y. Hirata, T. Sasaki, T. Ohta and S. Nozoe, Tetrahedron Lett.,
1988, 29, 1177; (b) 2-bromoascididemin (2-bromoleptoclidinininone):
J. J. Bloor and F. J. Schmitz, J. Am. Chem. Soc. 1987, 109, 6134;
11-Hidroxyascididemin (c): F. Schmitz, F. S. DeGuzman, M. B.
Hossain and D. van der Helm, J. Org. Chem., 1991, 56, 804.
2 Reviews: T. F. Molinski, Chem. Rev., 1993, 93, 1825; B. S. Davidson,
Chem. Rev. 1993, 93, 1771; A. M. Echavarren, in Advances in Nitrogen
Heterocycles, ed. C. J. Moody, JAI Press, Greenwich, 1996, vol. 2. ch.
5.
3 Indeed, we have found that analogue 2b is highly cytotoxic in vitro to
mouse lymphoma (P-388), human lung carcinoma (A-549), human
colon carcinoma (HT-29), and human melanoma (MEL-28) (IC50
values of 0.01, 0.0012, 0.005 and 0.0025 mg ml21, respectively).
4 A. M. Echavarren, J. Org. Chem., 1990, 55, 4255.
5 J. M. Cuerva and A. M. Echavarren, Synlett, 1997, 173; M. C. Carreño,
J. M. Cuerva, Ribagorda and A. M. Echavarren, Angew. Chem., Int. Ed.,
1999, 38, 1449.
6 The first step of a synthesis of the alkaloid meridine is a Diels–Alder
cycloaddition of the nitro analogue of 3 which proceeds in low yield
(6%): Y. Kitahara, F. Tamura and A. Kubo, Chem. Pharm. Bull. 1994,
42, 1363; S. Nakahara, Y. Tanaka and A. Kubo, Heterocycles, 1996, 43,
2113; a similar reaction was used for the synthesis of cistodamine: Y.
Kitahara, F. Tamura and A. Kubo, Tetrahedron Lett. 1997, 38, 4441.
Alternative synthesis of aromatized adducts: P. Molina, A. Pastor and
M. Villaplana, Tetrahedron 1995, 51, 1265.
7 Previous synthesis of 2a: J. R. Peterson, J. K. Zjawiony, S. Liu, C. D.
Hupfford, A. M. Clark and R. D. Rogers, J. Med. Chem., 1992, 35,
4069.
8 Synthesis of another regioisomer of 1a and 2b: E. Gómez-Bengoa and
A. M. Echavarren, J. Org. Chem., 1991, 56, 3497.
9 Intramolecular cycloaddition of 1-dimethylamino-1-azadienes: N.
Bushby, C. J. Moody, D. A. Riddick and I. R. Waldron, Chem.
Commun., 1999, 793; R. E. Dolle, W. P. Armstrong, A. N. Shaw and R.
Novelli, Tetrahedron Lett., 1988, 29, 6349.
10 P. J. McCarthy, T. P. Pitts, G. P. Gunawardana, M. Kelly-Borges and
S. A. Pomponi, J. Nat. Prod., 1992, 55, 1664.
Scheme 3
Much to our surprise, 6a was cleanly transformed into
tetracycle 7a14 by heating with acid (1+1 10% aq HCl–
1,4-dioxane, reflux; 94%). Quinone 7a could also be obtained
directly from 5a by thermolysis (xylene, reflux; 55%) or by
treatment with TFA (23 °C; 68%).15 Under these latter
conditions, heterocyclic quinone 7b was similarly obtained
from 5b (66%).13 Scheme 3 suggests a possible mechanism for
these remarkable transformations in which the dimethylhy-
drazine and two carbons are lost under relatively mild
conditions. Accordingly, a 6-endo-trig cyclization of the
trifluoroacetamide (from 6a,b) or the amine (via 6c,d formed in
situ from 5a,b) onto the quinone double bond with concomitant
(or subsequent) transacylation would lead to intermediates 8,
which may undergo aromatization to form a pyridine ring by a
Grob-type fragmentation to give acetylene and the hydrazine
derivative. The first step of this process, nucleophilic addition–
elimination on a quinone, is somewhat reminiscent of the so-
called mitomycin rearrangement.16
According to the hypothetical equilibrium of Scheme 3,
oxidation of intermediate 8 could yield a precursor of
tetracycles related to the ascididemins (1). In the event,
treatment of 6a,b with excess MnO2 in CH2Cl2 at 23 °C
smoothly led to tetracycles 9a,b in quantitative yield as single
isomers. The rearrangement also takes place in the presence of
DDQ or CAN as the oxidants. A trans configuration was
assigned for the alkenyl portion of these derivatives on the basis
of a vicinal coupling constant of 15.5 Hz. However, the
configuration around the enamine nitrogen was not rigorously
assigned. Reaction of 9a,b with NH4Cl and NaOAc in EtOH
under refluxing conditions gave pentacyclic 2a4 (93%) and 2b
(84%).13 After examining several hydrazone and oxime deriva-
tives,17 we found that thermolysis of phenylhydrazone 5c
directly furnished 2a (refluxing xylene, 40%), probably through
intermediate 10a. In fact, related tetracyclic derivative 10b
could be isolated in 54% yield by thermolysis of 5b in
xylene.
11 Hydrazones 5c–e were obtained in three steps by: (i) reaction of the
quinolinequinone with o-aminocinnamol (CeCl3·7H2O, MeOH, 23 °C);
(ii) oxidation of the allyl alcohol to the aldehyde with PCC or MnO2;
(iii) condensation of the aldehyde with the hydrazine.
12 By using the reaction pathway outlined in Scheme 2, adducts 5d–e could
be converted into 11-hydroxyascididemin (1c).
13 Yield based on unrecovered starting material.
14 A. Ettienne and A. Staehelin, Bull. Soc. Chim. Fr., 1954, 748; V. Zanker
and F. Mader, Chem. Ber., 1960, 93, 850. 2-Chloro derivative: M. Prato,
G. Scorrano, M. Stivanello, P. Tecilla and V. Lucchini, Gazz. Chim.
Ital., 1987, 117, 325.
15 Quinone 5a was quantitatively converted into 7a and N,N-dimethylhy-
drazine at 50–55 °C (3+1 benzene-d6–TFA-d solution, sealed NMR
tube). In addition, a small signal at 1.8 ppm attributable to dissolved
acetylene was also observed.
16 M. Kono, Y. Saitoh, K. Shirahata, Y. Arai and S. Ishii, J. Am. Chem.
Soc., 1987, 109, 7224; T. Fukuyama and L. Yang, J. Am. Chem. Soc.,
1987, 109, 7881; T. Fukuyama and L. Yang, J. Am. Chem. Soc., 1989,
111, 8303.
17 The methoxime and diphenylhydrazone analogues of 5a afforded 7a
after being treated with TFA.
In summary, a highly concise synthesis of heterocycles 2a,b
has been developed based on an intramolecular aza-Diels–Alder
cycloaddition followed by a novel rearrangement in which the
aminoaryl formally undergoes a 1,2-shift.
This work was supported by the DGES (Project PB97-
0002-C2-02). We acknowledge the Instituto Biomar, S.A., for
the data in ref. 3.
Communication 9/05234H
1722
Chem. Commun., 1999, 1721–1722