LETTER
1103
Synthetic Study on Ecteinascidin 743 Starting from D-Glucose
Atsushi Endo, Toshiyuki Kann, Tohru Fukuyama*
Graduate School of Pharmaceutical Sciences, University of Tokyo, CREST, The Japan Science and Technology Corporation (ST),
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
Fax+81-3-5802-8694; E-mail: fukuyama@mol.f.u-tokyo.ac.jp
Received 14 April 1999
selective synthesis of this key intermediate, employing the
incorporation of two nitrogen atoms into D-glucose and
the addition of phenol to an imine.
Abstract: During the course of synthetic study on ecteinascidin 743
(1), key intermediate 5 was prepared. Stereocontrolled synthesis of
5 from D-glucose was accomplished using incorporation of two ni-
trogen atoms and stereoselective addition of a phenol to an imine as
key steps.
Our synthesis began with epoxide 6, available on multi-
gram scale from inexpensive D-glucose in five steps
(Scheme 2).8 Upon treatment with p-toluenesulfonamide
and Cs2CO3, epoxide 6 underwent smooth ring opening
followed by mesylation to give 7. Acidic hydrolysis of ac-
etonide 7 afforded a 3:2 mixture of diastereomers at the
anomeric position. Subsequent treatment with SnCl4 fur-
nished the desired a-isomer 8.9 Base-induced ring closure
of mesylate 8 and subsequent protection as a TBS ether
provided aziridine 9.
Key words: ecteinascidin 743, D-glucose, bicyclo [3.3.1] system,
imine
The ecteinascidins were isolated by Rinehart and co-
workers from the marine tunicate Ecteinascidia turbina-
ta.1,2 Their novel structure and extremely potent antitumor
activities have promoted extensive studies.3,4,5 The first of
these compounds to advance to clinical trials was ectein-
ascidin 743 (1). In this communication, we report our syn-
thetic approach to 1. We plan to construct this compound
from the pentacyclic framework 2, which is common to all
ecteinascidins (Scheme 1). An intramolecular ortho-addi-
tion of the phenol to an aldehyde will provide the B-ring
of 2.6 The C-ring should be accessible from amino alde-
hyde 3, since a similar ring closure was demonstrated in
our total synthesis of saframycin A.7 The dialdehyde 3
will be prepared by oxidative cleavage of the diol 4, which
should be readily derived from 5. The pentacycle 5 bears
four chiral centers at C1, C3, C11, and C13 as well as the ar-
omatic A- and E-rings of 1. Described herein is a stereo-
The Grignard reagent 13, precursor for the E-ring, was
prepared from 3-methylcatechol (10) in six steps (Scheme
3). Selective protection of 10 by TsCl, bromination of the
resulting phenol, and methylation gave bromide 11. After
the phenol protecting group of 11 was changed from the
tosylate to a MOM ether, the bromide 12 was converted to
Grignard reagent 13 by treatment with Mg in THF.
A copper-catalyzed ring opening of the aziridine 9 by
freshly prepared 13 gave 14 (Scheme 4). Following pro-
tection of sulfonamide 14 with the Boc group and cleav-
age of the TBS ether, the second nitrogen atom was
incorporated by conversion of the alcohol to a triflate and
Synlett 1999, No. 07, 1103–1105 ISSN 0936-5214 © Thieme Stuttgart · New York