Biomimetic epoxide-opening cascades of oxasqualenoids triggered by hydrolysis of the terminal epoxide

Article abstract of DOI:10.1021/ol401081e

The biomimetic epoxide-opening cascades from squalene polyepoxides 4-6 to triterpene polyethers (oxasqualenoids) teurilene (1), glabrescol (2), and omaezakianol (3), respectively, were reproduced in a single event by chemical reaction. These cascades proceeded through the 5-exo tandem cyclization triggered by Bronsted acid-catalyzed hydrolysis of the terminal epoxide, mimicking the direct hydrolysis mechanism of epoxide hydrolases.

Full text of DOI:10.1021/ol401081e

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Biomimetic Epoxide-Opening Cascades of
Oxasqualenoids Triggered by Hydrolysis
of the Terminal Epoxide
Yoshiki Morimoto,* Eriko Takeuchi, Hitomi Kambara, Takeshi Kodama,
Yoshimitsu Tachi, and Keisuke Nishikawa
Department of Chemistry, Graduate School of Science, Osaka City University,
Sumiyoshi-ku, Osaka 558-8585, Japan
morimoto@sci.osaka-cu.ac.jp
Received April 22, 2013
ABSTRACT
The biomimetic epoxide-opening cascades from squalene polyepoxides 4ꢀ6 to triterpene polyethers (oxasqualenoids) teurilene (1), glabrescol
(2), and omaezakianol (3), respectively, were reproduced in a single event by chemical reaction. These cascades proceeded through the 5-exo
tandem cyclization triggered by Brønsted acid-catalyzed hydrolysis of the terminal epoxide, mimicking the direct hydrolysis mechanism of
epoxide hydrolases.
Organisms exquisitely produce biologically significant,
complex, and diverse natural products. It is well-known
that steroidal skeletons such as lanosterol are biosynthe-
Scheme 1. Two Mechanisms for Epoxide Hydrolases
sized in a single event from 2,3-oxidosqualene via the
protosterol cation by cyclases.1 Such cascade cyclization
is also invoked in neurotoxic ladderlike polyethers such as
brevetoxin B as a biogenetic hypothesis.2 The epoxide-
opening cascade biogenesis was originally proposed for
ionophoric polyether antibiotics as in the Caneꢀ
CelmerꢀWestley hypothesis in 1983,3 and now it has been
thought that it is universal for natural polyethers including
marine toxins, acetogenins, and triterpenoids.4 The epox-
ide-opening cascades have also attracted great interest
of chemists from the viewpoint of chemical synthesis
and have been utilized as a method to rapidly construct
polyether frameworks.5 However, the existence of enzymes
catalyzing the epoxide-opening cascade reaction has been a
puzzle for a long time.
Recently, direct experimental evidence on enzymatic
epoxide-opening reactions was reported by Oikawa et al.6
(1) Christianson, D. W. Chem. Rev. 2006, 106, 3412.
(2) (a) Nakanishi, K. Toxicon 1985, 23, 473. (b) Yamazaki, M.;
Izumikawa, M.; Tachibana, K.; Satake, M.; Itoh, Y.; Hashimoto, M.
J. Org. Chem. 2012, 77, 4902.
(3) Cane, D. E.; Celmer, W. D.; Westley, J. W. J. Am. Chem. Soc.
1983, 105, 3594.
ꢀ
(4) (a) Gallimore, A. G. Nat. Prod. Rep. 2009, 26, 266. (b) Fernandez,
J. J.; Souto, M. L.; Norte, M. Nat. Prod. Rep. 2000, 17, 235.
(5) Vilotijevic, I.; Jamison, T. F. Angew. Chem., Int. Ed. 2009, 48, 5250.
(6) Shichijo, Y.; Migita, A.; Oguri, H.; Watanabe, M.; Tokiwano, T.;
Watanabe, K.; Oikawa, H. J. Am. Chem. Soc. 2008, 130, 12230.
r
10.1021/ol401081e
XXXX American Chemical Society

Scheme 2. Hypothetical Biogenesis of Some Oxasqualenoids 1ꢀ3 Based on Epoxide-Opening Cascades Triggered by Direct Hydrolysis
of the Terminal Epoxide and Previous Work by Xiong and Corey12
Transformation of the prelasalocid diepoxide to lasalocid
A was achieved by an epoxide hydrolase Lsd19 in the final
stage of the biosynthesis. Lsd19 belongs to epoxide hydro-
lases, and the amino acid residues composing the active
site are similar to those of epoxide hydrolases catalyzing
the reaction that proceeds not with stepwise mechanism
by way of covalent bond ester intermediates between
substrates and the enzyme, but with a direct hydrolysis
mechanism (Scheme 1).7 In many examples of epoxide-
opening cascades, it is almost always the case that intra-
molecular nucleophiles such as π-electrons, hydroxy and
carboxy groups, and so on trigger the ring-opening of the
neighboring epoxide. In this contribution, we report bio-
mimetic and chemical epoxide-opening cascades triggered
by an intermolecular nucleophilic attack of water to the
terminal epoxide that mimics an intrinsic role of epoxide
hydrolases catalyzing direct hydrolysis.
glabrescol (2),9 and omaezakianol (3),10 we became inter-
ested in the possibility of their epoxide-opening cascade
biogenesis. The hypothetical biogenesis of 1ꢀ3, isolated
from marine and terrestrial plants,11 is shown in Scheme 2.
Considering the epoxide-opening cascade triggered by an
intermolecular nucleophilic attack of water to the terminal
epoxide that mimics an intrinsic role of epoxide hydrolases
catalyzing direct hydrolysis, compounds 1 and 3 could stereo-
specifically be derived from chiral tetraepoxide 48a,9a and
pentaepoxide 6,10b respectively, in a single event with inversion
of configuration upon regioselective opening of each epoxide.
On the other hand, optically active glabrescol (2) could
be provided from meso hexaepoxide 5 in the same manner,
if the enantiotopic terminal epoxides were differentiated.9a,c
Previously, Xiong and Corey reported the chemical epoxide-
opening cascade of substrate 8 related to our hypothesis;
however, it was from pentaepoxy diol 8 to non-natural
pentaTHF compound 7 initiated by a nucleophilic attack
of the intramolecular hydroxy group to the neighboring
epoxide.12 Although it has been unknown whether there are
epoxide hydrolases catalyzing our hypothetical biogenesis
triggered by direct hydrolysis or not, here we show that the
biomimetic epoxide-opening cascades from squalene poly-
epoxides 4ꢀ6 to oxasqualenoids 1ꢀ3, respectively, can be
reproduced in a single event by chemical reaction.
During the course of our and other chemical syntheses
of some oxasqualenoids such as cytotoxic teurilene (1),8
(7) (a) Minami, A.; Migita, A.; Inada, D.; Hotta, K.; Watanabe, K.;
Oguri, H.; Oikawa, H. Org. Lett. 2011, 13, 1638. (b) Hotta, K.; Chen, X.;
Paton, R. S.; Minami, A.; Li, H.; Swaminathan, K.; Mathews, I. I.;
Watanabe, K.; Oikawa, H.; Houk, K. N.; Kim, C.-Y. Nature 2012, 483,
355.
(8) (a) Hashimoto, M.; Harigaya, H.; Yanagiya, M.; Shirahama, H.
J. Org. Chem. 1991, 56, 2299. (b) Morimoto, Y.; Iwai, T.; Kinoshita, T.
The stereoselective syntheses of key substrates 4ꢀ6
ꢀ
ꢀ
J. Am. Chem. Soc. 1999, 121, 6792. (c) Rodrıguez-Lopez, J.; Crisostomo,
are depicted in Scheme 3. Preparation of C2 symmetric
ꢀ
F. P.; Ortega, N.; Lopez-Rodrıguez, M.; Martın, V. S.; Martın, T.
Angew. Chem., Int. Ed. 2013, 52, 3659.
(9) (a) Morimoto, Y.; Iwai, T.; Kinoshita, T. J. Am. Chem. Soc. 2000,
122, 7124. (b) Xiong, Z.; Corey, E. J. J. Am. Chem. Soc. 2000, 122, 9328.
(c) Transformation to (ꢀ)-glabrescol (2) from a C2 symmetric squalene
hexaepoxide, diastereomeric to meso hexaepoxide 5, by a base-promoted
middle-to-terminal double epoxide-opening cascade is reported. See:
Yang, P.; Li, P.-F.; Qu, J.; Tang, L.-F. Org. Lett. 2012, 14, 3932.
(10) (a) Morimoto, Y.; Okita, T.; Kambara, H. Angew. Chem., Int.
Ed. 2009, 48, 2538. (b) Xiong, Z.; Busch, R.; Corey, E. J. Org. Lett. 2010,
12, 1512.
(11) (a) Suzuki, T.; Suzuki, M.; Furusaki, A.; Matsumoto, T.; Kato,
A.; Imanaka, Y.; Kurosawa, E. Tetrahedron Lett. 1985, 26, 1329. (b)
Morita, H.; Kishi, E.; Takeya, K.; Itokawa, H.; Iitaka, Y. Phytochem-
istry 1993, 34, 765. (c) Harding, W. W.; Lewis, P. A.; Jacobs, H.;
McLean, S.; Reynolds, W. F.; Tay, L.-L.; Yang, J.-P. Tetrahedron Lett.
1995, 36, 9137. (d) Matsuo, Y.; Suzuki, M.; Masuda, M.; Iwai, T.;
Morimoto, Y. Helv. Chim. Acta 2008, 91, 1261.
(12) Xiong, Z.; Corey, E. J. J. Am. Chem. Soc. 2000, 122, 4831.
B
Org. Lett., Vol. XX, No. XX, XXXX

Scheme 3. Stereoselective Syntheses of Squalene Polyepoxides 4ꢀ6
tetraepoxide 4 began with epoxidation of commercially
available trans,trans-farnesyl acetate (9). Hydrolysis of
the terminal epoxide followed by oxidative cleavage of
the resulting vicinal diol afforded aldehyde 10. After
closely examining protective groups of the aldehyde, we
found a TBS ether of cyanohydrin13 was suitable for the
latter conversion. Deacetylation and subsequent Sharpless
asymmetric epoxidation (AE) of the resultant allylic alcohol
provided an epoxy alcohol as a 1:1 mixture at the cyanohy-
drin moiety. AE of the remaining alkene with Shi’s L-ketone
1214 then gave diepoxy alcohol 13. ParikhꢀDoering oxida-
tion of the alcohol 13 and Wittig methylenation of the
resulting aldehyde furnished alkene 14. The dimerization
of 14 was carried out in 78% yield by olefin metathesis
with the Grubbs catalyst 15, and diimide reduction of the
product provided tetraepoxide 16. The TBS ether 16 was
subjected to desilylation with TBAF to immediately afford
The cross-metathesis16 of known monoepoxy triene 1812
with triepoxy alkene 1710a in the presence of the Grubbs
catalyst 15 afforded the metathesis product. Shi AE of
the triene proceeded in a chemoselective manner at the
trisubstituted alkenes to provide a hexaepoxy alkene,
diimide reduction of which furnished meso hexaepoxide 5.
Pentaepoxide 6 was prepared from easily available squa-
lene monoepoxide 1917 by the same methodology as that
of 4 using the TBS ether of cyanohydrin. The exhaustive
epoxidation10b,12 of pentaene 20 was achieved with a large
excess of Shi’s L-ketone 12 to afford pentaepoxide 21 in
64% yield. Regeneration of the terminal trisubstituted
alkene was carried out by the aforementioned method,
providing the squalene pentaepoxide 6.
After investigating the epoxide-opening cascades of
polyepoxides 4ꢀ6 under various reaction conditions
(Brønsted and Lewis acid catalysts, solvent, temperature,
time, and so forth), we found that TfOH (0.2ꢀ0.3 equiv) in
THF/H2O (9:1) at rt for 10 min to 24 h afforded the best
yields in the reactions (Scheme 4). Thus, optically active
tetraepoxide 4 and pentaepoxide 6 could be converted into
mesoteurilene(1) and (þ)-omaezakianol (3) in 28and 33%
yields, respectively. On the other hand, meso hexaepoxide
5 was transformed into (()-glabrescol (2) in less than
10% yield.18 Respective 1H and 13C NMR spectra of the
ꢀ
a labile cyanohydrin, which upon treatment with Fetizon’s
reagent (Ag2CO3 on Celite) in toluene regenerated the
aldehyde functionality. The Wittig olefination of the dia-
ldehyde with an excess of Ph3PdCMe2 gave the requisite
tetraepoxide 4.15
(13) Yamamoto, K.; Watanabe, N.; Matsuda, H.; Oohara, K.;
Araya, T.; Hashimoto, M.; Miyairi, K.; Okazaki, I.; Saito, M.; Shimizu,
T.; Kato, H.; Okuno, T. Bioorg. Med. Chem. Lett. 2005, 15, 4932.
(14) (a) Wang, Z.-X.; Tu, Y.; Frohn, M.; Zhang, J.-R.; Shi, Y. J. Am.
Chem. Soc. 1997, 119, 11224. (b) Zhao, M.-X.; Shi, Y. J. Org. Chem.
2006, 71, 5377.
(15) A synthesis of 4 by the other method is reported. See:Ujihara, K.
Ph.D. Dissertation, The University of Tokyo, Tokyo, Japan, 2004.
(16) Chatterjee, A. K.; Choi, T.-L.; Sanders, D. P.; Grubbs, R. H.
J. Am. Chem. Soc. 2003, 125, 11360.
(17) van Tamelen, E. E.; Curphey, T. J. Tetrahedron Lett. 1962, 3,
121.
(18) In these reactions starting materials 4ꢀ6 were completely con-
sumed, and many products other than 1ꢀ3 were also observed; however,
we could not isolate identifiable ones.
Org. Lett., Vol. XX, No. XX, XXXX
C

Scheme 4. Epoxide-Opening Cascades of Squalene Polyepoxides 4ꢀ6 to Oxasqualenoids 1ꢀ3, Respectively, Triggered by Hydrolysis of
the Terminal Epoxide
synthetic compounds 1ꢀ3 were identical to those of pre-
viously reported natural and synthetic ones.8ꢀ10,11a,11c,11d
On the basis of the cascade cyclizations of similar sub-
strates reported by Corey10b,12 and McDonald,19 we dis-
cuss mechanistic aspects on the epoxide-opening cascades
of polyepoxides 4ꢀ6. In the transformation from tetraep-
oxide 4 to teurilene (1), the first step would be initiated by
hydrolysis of the terminal epoxide, to which water seems
more accessible, with inversion of configuration at the
more substituted epoxide carbon to afford diol intermedi-
ate 22 (Scheme 4). The type of 2,5-linked polyTHF such as
1 could be formed from the diol 22 via the kinetically
favored 5-exo20 cascade cyclizations similar to those of 8 in
Scheme 2.10b,12 Such an intramolecular attack of the neigh-
boring epoxide oxygen to the less substituted carbon of the
terminal epoxide as shown in 4 of Scheme 2 should not be
the first step, because under acidic conditions the intra-
molecular attack occurs at the more substituted carbon of
the terminal epoxide, and subsequently, polyepoxide sub-
strates such as 4 undergo 6-endo20 cascade cyclizations to
ladderlike fused polycyclic ethers via an epoxonium ion
intermediate proposed by McDonald et al.19 Whichever
terminal epoxide is hydrolyzed, only meso teurilene (1) is
produced due to C2 symmetric 4. Similarly, the racemic
glabrescol (()-2 was obtained from the meso hexaepoxide 5
in low yield along with many other complex mixtures.9c In
the initial hydrolysis of pentaepoxide 6, the right and left
terminal epoxides need to be precisely differentiated be-
cause if the epoxide-opening cascade were initiated from
the left terminal epoxide, an undesired compound ent-3,
antipodal to the natural 3, would be formed. Practically, the
initial hydrolysis took place at the right terminal epoxide
which was thought to be less hindered, because the optical
rotation of the synthetic (þ)-omaezakianol (3), [R]29
D
þ18.1 (c 0.20, CHCl3), was in good agreement with those
of the natural product, [R]20D þ15.8 (c 0.59, CHCl3),11d and
other synthetic (þ)-3’s, [R]29D þ17.7 (c 0.57, CHCl3)10a and
[R]23 þ17.1 (c 1.0, CHCl3).10b Without the aforemen-
D
tioned epoxide-opening cascade mechanism initiated by
hydrolysis of the terminal epoxide, the stereochemical
relationship between substrates 4ꢀ6 and products 1ꢀ3,
respectively, could not reasonably be explained.
In conclusion, we have realized the biomimetic and che-
mical epoxide-opening cascades of squalene polyepoxides
4ꢀ6 to oxasqualenoids 1ꢀ3, respectively, through the 5-exo
cascade cyclizations triggered by the protic acid-catalyzed
hydrolysis of the terminal epoxide. Because epoxide hydro-
lase Lsd19 is responsible for the epoxide-opening reaction in
the final polyether formation of lasalocid A,6,7 some epoxide
hydrolases with a direct hydrolysis mechanism might also
be implicated in the enzymatic epoxide-opening cascades
of squalene polyepoxides 4ꢀ6. In particular, to generate
the optically active glabrescol (2) from meso hexaepoxide 5
enzymatic participation would be so essential that the
(R)-epoxide of the two enantiotopic terminal ones has to
be selectively hydrolyzed (Scheme 4). Details of the reaction
mechanism and the possibility of the epoxide hydrolase-
triggered epoxide-opening cascades are under investigation.
Acknowledgment. We are grateful to Professors H.
Oikawa (Hokkaido University) and M. Hashimoto
(Hirosaki University) for valuable discussions. This work
was financially supported by the JSPS, the MEXT, and the
Asahi Glass Foundation.
Supporting Information Available. Full experimental
details and spectra are provided. This material is available
free of charge via the Internet at http://pubs.acs.org.
(19) McDonald, F. E.; Bravo, F.; Wang, X.; Wei, X.; Toganoh, M.;
Rodrıˆguez, J. R.; Do, B.; Neiwert, W. A.; Hardcastle, K. I. J. Org. Chem.
2002, 67, 2515.
(20) Baldwin, J. E. J. Chem. Soc., Chem. Commun. 1976, 734.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX