Total synthesis of ciguatoxin and 51-hydroxyCTX3C

Article abstract of DOI:10.1021/ja063041p

Ciguatoxins, the principal causative toxins of ciguatera seafood poisoning, are large ladder-like polycyclic ethers with the 13 ether rings ranging from five- to nine-membered. In this paper, we describe the total synthesis of the two most toxic members o

Full text of DOI:10.1021/ja063041p

Published on Web 07/01/2006
Total Synthesis of Ciguatoxin and 51-HydroxyCTX3C
Masayuki Inoue,*,†,‡ Keisuke Miyazaki,† Yuuki Ishihara,† Atsushi Tatami,† Yuyu Ohnuma,†
Yuuya Kawada,† Kazuo Komano,† Shuji Yamashita,† Nayoung Lee,† and Masahiro Hirama*,†
Department of Chemistry, Graduate School of Science, Tohoku UniVersity, and Research and Analytical Center for
Giant Molecules, Graduate School of Science, Tohoku UniVersity, Sendai 980-8578, Japan
Received May 2, 2006; E-mail: inoue@ykbsc.chem.tohoku.ac.jp; hirama@ykbsc.chem.tohoku.ac.jp
Ciguatera poisoning is an important medical issue in the tropical
and subtropical Pacific and Indian Ocean regions and in the tropical
Caribbean.1 As reef fish are increasingly exported to other areas, it
has become a worldwide health problem. Ciguatoxins, the principal
causative agents of ciguatera, are produced by an epiphytic
dinoflagellate, Gambierdiscus toxicus, and transferred to herbivo-
rous and carnivorous fish through the aquatic food chain,2 at the
end of which humans are exposed. Ingestion of affected fish leads
to gastrointestinal, cardiovascular, and neurological disorders which
may last for weeks or even years. The lethal potency of ciguatoxin
[a.k.a. CTX1B (1, Scheme 1)] by intraperitoneal injection into mice
was reported to be 280 times greater than that of tetrodotoxin.3
The low fatality rate of ciguatera is due solely to the minute
concentration of ciguatoxins in fish flesh.
Since the seminal structural elucidation by Yasumoto in 1989
using 0.35 mg of ciguatoxin 1 extracted from 4000 kg of moray
eels (Gymnothorax jaVanicus),3b,c more than 20 ciguatoxin conge-
ners have been structurally determined.4 G. toxicus produces less
oxidized ciguatoxins that undergo varying extents of oxidations as
they pass through the food chain. Interestingly, the more oxygenated
forms 1 and 25 are typically 10 times more toxic than the ciguatoxin
congeners produced by G. toxicus (e.g., CTX3C6).7 This increase
in potency amplifies the risk of intoxication in humans.
The very limited supply of ciguatoxins from natural sources has
prevented detailed biological studies as well as development of
therapeutic methods for ciguatera. Their intriguing and challenging
molecular architecture, coupled with this fact, led us to launch a
synthetic study, which resulted in the total synthesis of CTX3C in
20018 and its second generation synthesis in 2003.9 Herein, we
describe the first total synthesis of the two most toxic members
of the ciguatoxin family, ciguatoxin 1 and 51-hydroxyCTX3C 2,
based on our second generation strategy in combination with a
newly developed stereo- and chemoselective radical cyclization
process.
Ciguatoxins are huge ladder-like polycyclic ethers with the 13
ether rings ranging from five- to nine-membered.10,11 Despite
structural differences in the carbon backbones and their varying
oxidation levels, the ciguatoxin congeners share common FG-ring
structures.4 Therefore, our unified convergent strategy was designed
to be applicable to all congeners alike: the entire structures were
retrosynthetically disconnected at the central portions to afford
comparably complex A-E- and H-M-ring systems. In the synthetic
direction, it was envisioned that direct formation of key O,S-acetals
from the two halves of the molecules, subsequent FG-ring construc-
tion, and global deprotection would deliver the oxidized ciguatoxins
1 and 2. Synthesis of the requisite left and right wing fragments
(3,12 4,13 1414) has been reported previously.
As shown in Scheme 1, total synthesis of 51-hydroxyCTX3C 2
started with assembly of 3 and 4. Installation of an R-chloride in
H-M-ring sulfide 4 was carried out using NCS, leading to
R-chlorosulfide 5. The resulting solution of unstable 5 was directly
treated with AgOTf and DTBMP in the presence of A-E-ring
alcohol 3, giving rise to undecacyclic O,S-acetal 6 in 70% yield.15,16
The TIPS group of the key intermediate 6 was then removed with
TBAF to give secondary alcohol 7, which was converted to methyl
acrylate 8a using methyl propiolate and NMM.17 By subjecting 8a
to radical cyclization in the presence of n-Bu3SnH and AIBN, the
G-ring of 9a was constructed via 7-exo cyclization with complete
control of the two stereogenic centers at C26 and C27. However,
the yield of 9a was modest (40%) due to the concomitant formation
of byproduct 10a via 6-exo cyclization at C23 (20% yield).
The stereo- and chemoselectivity of the radical reaction was
reproduced in a model experiment using 23a (Table 1, entry 1),
which represents the central region of the molecule. The exclusive
formation of the desired stereoisomer via 7-exo cyclization is
explained by the relative stability of the conformation of the
corresponding radical intermediate: the C-C bond formation based
on the most stable conformation, 25, leads to the desired product
26 among the four possible stereoisomers.18 It is possible that the
7-exo/6-exo chemoselectivity of the reaction originates from a
balance between SOMO/LUMO interactions and entropic factors.
Although there is a stronger interaction between the SOMO of the
nucleophilic R-oxy radical and the LUMO of the electron-deficient
acrylate than that of the terminal olefin,19 6-exo cyclization is
entropically favored over its 7-exo counterpart.20
The only way to bypass the intrinsic preference for the undesired
6-exo path is to enhance the SOMO/LUMO interaction of the
acrylate. Accordingly, the methyl group was replaced with a series
of electron-withdrawing functional groups (Table 1, entries 2-4).21,22
The preference for 26 was found to improve in proportion to the
number of fluorides on the phenoxy ring. Remarkably, the simple
replacement of Me with pentafluorophenyl enables a 6-fold increase
in 7-exo selectivity without affecting stereoselectivity (entry 4).
The new chemoselective radical cyclization was successfully
applied to the actual system (Scheme 1). Acrylate 8d was prepared
from 7 using pentafluorophenyl propiolate and PMe3.23 The radical
reaction of 8d exhibited significantly higher selectivity for 7-exo
cyclization (9e:10d ) 10.6:1). As the pentafluorophenyl ester was
hydrolyzed in the workup, the product was isolated as carboxylic
acid 9e (74% yield),24 which was methylated to afford 9a.
With the 12 ether rings in place, the final steps were construction
of the F-ring by ring-closing metathesis (RCM) reaction and
deprotection.25 DIBAL reduction of ester 9a to the aldehyde,
followed by Wittig methylenation, produced pentaene 11. Upon
treatment of 11 with catalytic first-generation Grubbs reagent 13,26
only the two terminal olefins were activated, resulting in successful
closure of the nine-membered F-ring of 12 in 93% yield. Last,
† Department of Chemistry.
‡ Research and Analytical Center for Giant Molecules.
9
9352
J. AM. CHEM. SOC. 2006, 128, 9352-9354
10.1021/ja063041p CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Scheme 1 a
a Reagents and conditions: (a) N-chlorosuccinimide (NCS) (1.0 equiv), CH2Cl2/CCl4 (1:6), rt; (b) 3 (1.0 equiv), 5 (1.3 equiv), AgOTf (2.5 equiv), di-
tert-butylmethylpyridine (DTBMP) (5.0 equiv), 4 Å molecular sieves, CH2Cl2/CCl4 (5:1), -70 to 0 °C, 70% from 3; (c) TBAF, THF, 35 °C, 100%; (d)
methyl propiolate, N-methylmorpholine (NMM), CH2Cl2, rt, 100%; (e) pentafluorophenyl propiolate, PMe3, CH2Cl2, rt, 95%; (f) 8a, n-Bu3SnH, 2,2′-
azobisisobutyronitrile (AIBN), toluene, 85 °C, 9a, 40%; 10a, 20%; 8d, n-Bu3SnH, AIBN, toluene, 85 °C, 9e, 74%; 10d, 7%; (g) TMSCHN2, MeOH/benzene
(2:5), rt, 87%; (h) DIBAL, CH2Cl2, -90 °C; (i) Ph3PCH3Br, t-BuOK, THF, 0 °C, 98% (2 steps); (j) 13 (0.2 equiv), CH2Cl2, 40 °C, 93%; (k) 2,3-dichloro-
5,6-dicyano-1,4-benzoquinone (DDQ) (10 equiv), CH2Cl2/H2O (2:1), rt, 59%; (l) 14 (1.0 equiv), 5 (1.5 equiv), AgOTf (2.5 equiv), DTBMP (5.0 equiv), 4
Å molecular sieves, CH2Cl2/CCl4 (5:1), -70 to 0 °C, 63% from 14; (m) TBAF, THF, 35 °C, 92%; (n) pentafluorophenyl propiolate, PMe3, CH2Cl2, rt, 94%;
(o) n-Bu3SnH, AIBN, toluene, 85 °C, 18e, 59%; 19d, 22%; (p) TMSCHN2, MeOH/benzene (2:5), 84%; (q) DIBAL, CH2Cl2, -90 °C; (r) Ph3PCH3Br,
t-BuOK, THF, 0 °C, 77% (2 steps); (s) 13 (0.3 equiv), CH2Cl2, 40 °C, 78%; (t) DDQ (30 equiv), CH2Cl2/H2O (1:1), rt; (u) 1 N HCl, MeOH, rt, 30% (2
steps).
oxidative removal of the four 2-naphthylmethyl (NAP) groups of
12 using DDQ gave rise to 51-hydroxyCTX3C 2 in 59% yield.27
Having established the synthetic route to 2, our focus next turned
to the total synthesis of ciguatoxin 1. Ciguatoxin 1 not only contains
an additional dihydroxybutenyl side chain, but it also possesses a
seven-membered E-ring rather than the eight-membered ring of 2.
The synthetic challenge presented by 1 is especially heightened by
the presence of the acid/base/oxidant-sensitive bisallylic C5-ether.
A-E-ring alcohol 14 was coupled with R-chlorosulfide 5 by the
action of AgOTf in the presence of DTBMP, generating O,S-acetal
15 in 63% yield. The subsequent two steps exchanged the TIPS
group of 15 for the pentafluorophenyl acrylate of 17d. For closure
of the G-ring, 17d was treated with AIBN and n-Bu3SnH,
selectively producing oxepane 18e (59% yield) with the desired
stereochemistry, along with 19d as a minor compound (18e:19d )
2.7:1). The observed 7-exo selectivity was lower than that of
9e presumably due to the more favorable orientation of the allyl
group of the seven-membered E-ring in 17d for the 6-exo
cyclization. Nevertheless, as clearly demonstrated in the model
experiment using 24 (Table 1, entries 5 and 6), the pentafluoro-
phenyl group functioned as a selectivity control element regardless
of the size of the E-ring, and improved the 7-exo/6-exo chemose-
lectivity.
the high chemoselectivity of the Grubbs catalyst 13 in this highly
complex system, affording the fully protected ciguatoxin 21 in 78%
yield. DDQ was then used to remove five of the six NAP groups
of 21 and simultaneously converted one NAP ether to the
2-naphthylidene acetal at the C1,2-diol, resulting in tetraol 22.
Finally, careful treatment of acetal 22 with acidic conditions
delivered ciguatoxin 1 in 30% yield (two steps).
In summary, ciguatoxin 1 and 51-hydroxyCTX3C 2 were
synthesized for the first time from the corresponding left and right
wing fragments in only 10 and 9 steps, respectively. Synthetic
ciguatoxins were determined to be identical in all respects to the
naturally occurring form by thin-layer chromatography, HPLC, 1H
NMR, and MS.28 This work proved the high applicability of our
unified strategy for construction of highly complex structures with
chemically labile functionalities, such as the A-ring side chain of
ciguatoxin. Salient methodologies employed in our successful
campaign include (i) direct construction of the O,S-acetal; (ii) radical
cyclization to form the G-ring; (iii) RCM reaction to form the
F-ring; and (iv) an efficient protective group strategy using the NAP
group, which can be readily removed under the oxidative conditions.
In addition, the discovery that pentafluorophenyl acrylates controlled
the reaction path of the radical cyclizations should have wider
applications beyond this synthesis. Ciguatoxins prepared here will
accelerate biological studies as well as the development of strategies
for controlling ciguatera seafood poisoning.
Three-step functional group manipulations converted carboxylic
acid 18e to hexaene 20. The RCM reaction of 20 again demonstrated
9
J. AM. CHEM. SOC. VOL. 128, NO. 29, 2006 9353

C O M M U N I C A T I O N S
Table 1. Development of Stereo- and Chemoselective Radical
Cyclization
(3) (a) Scheuer, P. J.; Takahashi, W.; Tsutsumi, J.; Yoshida, T. Science 1967,
155, 1267. (b) Murata, M.; Legrand, A. M.; Ishibashi, Y.; Yasumoto, T.
J. Am. Chem. Soc. 1989, 111, 8929. (c) Murata, M.; Legrand, A.-M.;
Ishibashi, Y.; Fukui, M.; Yasumoto, T. J. Am. Chem. Soc. 1990, 112,
4380. (d) Satake, M.; Morohashi, A.; Oguri, H.; Oishi, T.; Hirama, M.;
Harada, N.; Yasumoto, T. J. Am. Chem. Soc. 1997, 119, 11325.
(4) Yasumoto, T.; Igarashi, T.; Legrand, A.-M.; Cruchet, P.; Chinain, M.;
Fujita, T.; Naoki, H. J. Am. Chem. Soc. 2000, 122, 4988.
(5) Satake, M.; Fukui, M.; Legrand, A.-M.; Cruchet, P.; Yasumoto, T.
Tetrahedron Lett. 1998, 39, 1197.
(6) Satake, M.; Murata, M.; Yasumoto, T. Tetrahedron Lett. 1993, 34, 1975.
(7) Dechraoui, M.-Y.; Naar, J.; Pauillac, S.; Legrand, A.-M. Toxicon 1999,
37, 125.
(8) (a) Hirama, M.; Oishi, T.; Uehara, H.; Inoue, M.; Maruyama, M.; Oguri,
H.; Satake, M. Science 2001, 294, 1904. (b) Inoue, M.; Uehara, H.;
Maruyama, M.; Hirama, M. Org. Lett. 2002, 4, 4551.
(9) (a) Inoue, M.; Miyazaki, K.; Uehara, H.; Maruyama, M.; Hirama, M. Proc.
Natl. Acad. Sci. U.S.A. 2004, 101, 12013. (b) Inoue, M.; Hirama, M. Acc.
Chem. Res. 2004, 37, 961. (c) Hirama, M. Chem. Rec. 2005, 5, 240.
(10) For recent reviews on syntheses of polycyclic ethers, see: (a) Nakata, T.
Chem. ReV. 2005, 105, 4314. (b) Inoue, M. Chem. ReV. 2005, 105, 4379.
(11) For recent synthetic studies from other groups, see: (a) Clark, J. S.;
Hamelin, O. Angew. Chem., Int. Ed. 2000, 39, 372. (b) Kira, K.; Isobe,
M. Tetrahedron Lett. 2001, 42, 2821. (c) Takakura, H.; Sasaki, M.; Honda,
S.; Tachibana, K. Org. Lett. 2002, 4, 2771. (d) Fuwa, H.; Fujikawa, S.;
Tachibana, K.; Takakura, H.; Sasaki, M. Tetrahedron Lett. 2004, 45, 4795.
(e) Fujiwara, K.; Goto, A.; Sato, D.; Ohtaniuchi, Y.; Tanaka, H.; Murai,
A.; Kawai, H.; Suzuki, T. Tetrahedron Lett. 2004, 45, 7011. (f) Domon,
D.; Fujiwara, K.; Murai, A.; Kawai, H.; Suzuki, T. Tetrahedron Lett. 2005,
46, 8285. (g) Hamajima, A.; Isobe, M. Org. Lett. 2006, 8, 1205 and
references cited therein.
(12) (a) Maruyama, M.; Inoue, M.; Oishi, T.; Oguri, H.; Ogasawara, Y.; Shindo,
Y.; Hirama, M. Tetrahedron 2002, 58, 1835. (b) Kobayashi, S.; Takahashi,
Y.; Komano, K.; Alizadeh, B. H.; Kawada, Y.; Oishi, T.; Tanaka, S.;
Ogasawara, Y.; Sasaki, S.; Hirama, M. Tetrahedron 2004, 60, 8375.
(13) (a) Tatami, A.; Inoue, M.; Uehara, H.; Hirama, M. Tetrahedron Lett. 2003,
44, 5229. (b) Inoue, M.; Yamashita, S.; Tatami, A.; Miyazaki, K.; Hirama,
M. J. Org. Chem. 2004, 69, 2797.
(14) Kobayashi, S.; Alizadeh, B. H.; Sasaki, S.; Oguri, H.; Hirama, M. Org.
Lett. 2004, 6, 751.
(15) (a) Inoue, M.; Wang, G. X.; Wang, J.; Hirama, M. Org. Lett. 2002, 4,
3439. (b) Inoue, M.; Wang, J.; Wang, G. X.; Ogasawara, Y.; Hirama, M.
Tetrahedron 2003, 59, 5645.
(16) (a) Mukaiyama, T.; Sugaya, T.; Marui, S.; Nakatsuka, T. Chem. Lett. 1982,
1555. (b) McAuliffe, J. C.; Hindsgaul, O. Synlett 1998, 307.
(17) Lee, E.; Tae, J. S.; Lee, C.; Park, C. M. Tetrahedron Lett. 1993, 34, 4831.
(18) (a) Sasaki, M.; Inoue, M.; Noguchi, T.; Takeichi, A.; Tachibana, K.
Tetrahedron Lett. 1998, 39, 2783. (b) Imai, H.; Uehara, H.; Inoue, M.;
Oguri, H.; Oishi, T.; Hirama, M. Tetrahedron Lett. 2001, 42, 6219.
(19) Giese, B. Angew. Chem., Int. Ed. Engl. 1983, 22, 753.
(20) Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95.
(21) By analogy to pericyclic reactions, it was expected that Lewis acids
strongly enhance the orbital interaction. However, application of Lewis
acids [MgBr2, Me3Al, Sc(OTf)3] to the acrylate failed to increase the yield
of the 7-exo product due to the acid-sensitive nature of the O,S-acetal.
(22) (a) Kan, T.; Ohfune, Y. Tetrahedron Lett. 1995, 36, 943. (b) Ohfune, Y.;
Kan, T.; Nakajima, T. Tetrahedron 1998, 54, 5207.
(23) Inanaga, J.; Baba, Y.; Hanamoto, T. Chem. Lett. 1993, 241.
(24) Pentafluorophenyl esters 10d and 19d were more stable to hydrolysis than
the 7-exo products presumably because of their vinylogous carbonate
structures.
(25) For reviews, see: (a) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001,
34, 18. (b) Fu¨rstner, A. Angew. Chem., Int. Ed. 2000, 39, 3012. (c) Deiters,
A.; Martin, S. F. Chem. ReV. 2004, 104, 2199. (d) Nicolaou, K. C.; Bulger,
P. G.; Sarlah, D. Angew. Chem., Int. Ed. 2005, 44, 4490.
Acknowledgment. This work was supported financially by
SORST, Japan Science and Technology Agency (JST), and a Grant-
in-Aid for Scientific Research (S) from the Japan Society for the
Promotion of Science (JSPS). A fellowship to K.M. from JSPS is
gratefully acknowledged. We thank Professors Michio Murata
(Osaka University) and Masayuki Satake (University of North
Carolina) for providing NMR spectra of ciguatoxins.
(26) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996, 118,
100.
(27) NAP is known to be oxidatively removable similarly to p-methoxybenzyl,
but is more acid-stable. For the development of our protective group
strategy, see ref 8b. For NAP ether references, see: (a) Gaunt, M. J.; Yu,
J.; Spencer, J. B. J. Org. Chem. 1998, 63, 4172. (b) Wright, J. A.; Yu, J.;
Spencer, J. B. Tetrahedron Lett. 2001, 42, 4033. (c) Xia, J.; Alderfer, J.
L.; Piskorz, C. F.; Matta, K. L. Chem.sEur. J. 2001, 7, 356.
(28) A preliminary toxicity study of the synthetic ciguatoxins was carried out
in mice. Acute toxicity on median lethal dose (LD50, intraperitoneal
injection) of 1 and 2 was determined to be 1.64 and 0.31 µg/kg,
respectively.
Supporting Information Available: Experimental procedures and
spectroscopic data for synthetic compounds (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Scheuer, P. J. Tetrahedron 1994, 50, 3. (b) Lewis, R. J. Toxicon 2001,
39, 97. (c) Yasumoto, T.; Murata, M. Chem. ReV. 1993, 93, 1897.
(2) Yasumoto, T.; Nakajima, I.; Bagnis, R.; Adachi, R. Bull. Jpn. Soc. Sci.
Fish. 1977, 43, 1015.
JA063041P
9
9354 J. AM. CHEM. SOC. VOL. 128, NO. 29, 2006