Formal total synthesis of testudinariol A, a triterpene with C2 symmetry

Article abstract of DOI:10.1246/cl.2001.898

The alcohol corresponding to half of the molecule of testudinariol A, a triterpene with a symmetric C₂ structure isolated from Pleurobrancus testudinarius, was enantioselectively synthesized. The alcohol has been converted to testudinariol A by Mori et al.

Full text of DOI:10.1246/cl.2001.898

898
Chemistry Letters 2001
Formal Total Synthesis of Testudinariol A, a Triterpene with C2 Symmetry
Hideaki Hioki, Masayo Hamano, Miwa Kubo, Tomoko Uno, and Mitsuaki Kodama*
Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima 770-8514
(Received April 5, 2001; CL-010307)
The alcohol corresponding to half of the molecule of testu-
prenylated product 7a. Silyl ether was then removed and the
resulting alcohol was oxidized to aldehyde 8a. The minor prod-
uct 5b was also transformed into the aldehyde 8b using the
same sequence of reactions (Scheme 1).
dinariol A, a triterpene with a symmetric C2 structure isolated
from Pleurobrancus testudinarius, was enantioselectively syn-
thesized. The alcohol has been converted to testudinariol A by
Mori et al.
Testudinariol A (1) is a triterpene ether isolated from the
skin and mucus of Pleurobrancus testudinarius, together with
its diastereomer, testudinariol B (2).1 A characteristic feature of
1 is its symmetric C2 structure and partially cyclized squalene
skeleton.2 The biologic role of 1 in mollusks is not known, but
the substance is considered to act as a defensive allomone
because it is ichthyotoxic against Gambusia affinis. As a part
of synthetic studies of natural products using baker’s yeast
reduction of α-hydroxyketone,3 an excellent method for chirali-
ty induction in terpenoid synthesis, we attempted enantioselec-
tive total synthesis of testudinariol A (1). Because 1 has a sym-
metric C2 structure, dimerization of half of the molecule was
the most effective strategy towards synthesizing 1. Thus, a
compound such as 3 is a potential synthetic intermediate in the
present synthesis. Recent publication of the total synthesis of 1
by Mori and coworkers4 prompted us to report our efforts to
synthesize testudinariol A.5 Herein, we report synthesis of the
alcohol 3 (R = TBS), which was successfully converted to the
natural testudinariol A by Mori’s group.
The ene reaction9 of 8a using diethylaluminum chloride
afforded two cyclopentanol derivatives 9 and 10, while under
the same conditions, 8b gave three products 11, 12, and 13
(Scheme 2). The relative stereochemistries around the
cyclopentane ring in each product were deduced on the basis of
careful analyses of their NOESY spectra. At this stage, howev-
er, it was still difficult to determine the stereochemical relation-
ship at C6 and C7. Comparison of the 1H and 13C NMR spectra
In the course of our total synthesis of hippospongic acid
A,3,6 a triterpene possessing inhibitory activity for gastrulation
of starfish embryos, we synthesized the dihydropyran derivative
4 in an optically pure form by the baker’s yeast reduction of α-
hydroxyketone derived from myrcene. This compound is a
good precursor for the synthesis of 3 because introduction of a
prenyl unit and subsequent cyclopentane ring formation provide
the desired key intermediate 3.
Hydroboration of 4 using 9-BBN afforded the epimeric
alcohols 5a and 5b in ca. 3:1 ratio. As described below, the
major product 5a was suggested to have a (7R)-configuration.
Because the stereochemistry at C7 could not be determined at
this stage, we proceeded with the synthesis of the key intermedi-
ate 3. The alcohol 5a was sulfenylated using Hata’s method7 to
give the thioether 6a. Oxidation of 6a with oxone,8 prenylation
of the resulting sulfone, followed by desulfonylation, yielded the
Copyright © 2001 The Chemical Society of Japan

Chemistry Letters 2001
899
4). Thus, 3 (R = TBS) was converted to the bromide 16, which
11
was treated with Ni(cod)2 to yield a mixture in moderate
yield, whose NMR spectra obviously showed that the mixture
consisted of 17 and 18 (ca. 1:1). But separation of these iso-
mers was totally unsuccessful. The difficulty in separation was
not improved by deprotecting to the alcohol mixture (1 and 19).
1
The detailed comparison of the H and 13C NMR spectra12 of
to those of natural 1 (Table 1) revealed that only 11 has the
chemical shifts consistent with those of natural product [∆δ <
0.04 ppm (1H); ∆δ < 0.2 ppm (13C)] suggesting that the isomer
11 was the desired compound, although the stereochemistry has
not been established yet. Five of eight possible diastereomers
were formed and the isopropenyl group and the hydroxy group
in the products tend to have cis disposition. But clear explana-
tion of the stereoselectivity of the ene reaction was difficult. To
utilize the diastereomer formed in the hydroboration of 4, 5a
was transformed into the thioether 6b in 96% overall yield
(Scheme 3).
the obtained mixture with those of natural testudinariol A clear-
ly revealed the presence of testudinariol A (1) in the mixture.
We are very grateful to Professor K. Mori, Science
University of Tokyo, for the identification of the key compound
and useful discussion. This work was partially supported by a
Grant-in-Aid for Scientific Research (No. 11672132) from the
Ministry of Education, Science, Sports and Culture of Japan.
References and Notes
1
A. Spinera, E. Mollo, E. Trivellone, and G. Cimino, Tetrahedron,
53, 16891 (1997).
For a recent review of marine polyether triterpenes, see J. J.
Fernándes, M. L. Souto, and M. Norte, Nat. Prod. Rep., 17, 235
(2000).
For our previous synthesis using baker’s yeast reduction, see H.
Hioki, H. Ooi, M. Hamano, Y. Mimura, S. Yoshio, M. Kodama, S.
Ohta, M. Yanai, and S. Ikegami, Tetrahedron, 57, 1235 (2001) and
references cited therein.
2
3
4
5
H. Takikawa, M. Yoshida, and K. Mori, Tetrahedron Lett., 42, 1527
(2001).
The results were presented previously. M. Kodama, H. Hioki, S.
Yoshio, M. Matsushita, M. Hamano, C. Kanehara, Y. Ohnishi, Y.
Umemori, M. Kubo, and H. Sakai, 42nd Symposium on the
Chemistry of Natural Products, Okinawa, November, 2000, Abstr.,
No. 40.
6
H. Hioki, H. Ooi, Y. Mimura, S. Yoshio, and M. Kodama, Synlett,
1998, 729.
It was necessary to improve the yield of 11 in the ene reac-
tion of 5b. In preliminary experiments using model compounds,
we found that scandium triflate,10 Sc(OTf)3, is a better catalyst
of the present ene reaction and we obtained the desired 11 in
49% yield together with 37% of the isomer mixture using 0.1
mol equivalent of Sc(OTf)3. Thus, the obtained 11 was con-
verted to the key synthetic intermediate 3 (R = TBS) by simple
modification of the protective groups (Scheme 4).
7
8
9
I. Nagawa and T. Hata, Tetrahedron Lett., 1975, 1409.
B. M. Trost and D. P. Curran, Tetrahedron Lett., 22, 1287 (1981).
For reviews, see: B. B. Snider, in “Comprehensive Organic
Synthesis,” ed. by B. M. Trost and I. Fleming, Pergamon, Oxford
(1991), Vol. 2, pp 527–567 and Vol. 5, pp 1–27.
10 V. K. Aggarwal, G. P. Vennall, P. N. Davey, and C. Newman,
Tetrahedron Lett., 39, 1409 (1998).
11 P. J. M. Reijnders, H. R. Fransen, and H. M. Buck, Recl. Trav.
Chim. Pays-Bas., 98, 511 (1979).
12 Selected NMR chemical shifts of 1; H NMR δ (CDCl3) 1.84 (6H,
1
Although successful synthesis of 3 (R = TBS) constitutes
the formal synthesis of testudinariol A (1) because the alcohol
thus obtained was identical to the compound prepared by Mori
et al.,3 we independently attempted the synthesis of 1 (Scheme
s), 2.40 (2H, ddd, J = 5.8, 5.8, 11.6 Hz), 3.19 (2H, dd, J = 8.7, 8.7
Hz), 3.73 (2H, d, J = 12.9 Hz), 4.18 (2H, m), 4.59 (2H, d, J = 12.9
Hz), 4.82 (2H, s), 4.97 (2H, s), 5.17 (2H, t-like, J = 6.3 Hz); 13C
NMR δ (CDCl3) 23.3, 26.7, 27.2, 27.3, 32.0, 33.0, 52.0, 53.1, 66.7,
74.8, 80.7, 112.2, 123.4, 134.3, 144.4.