J. Am. Chem. Soc. 1997, 119, 11325-11326
The Absolute Configuration of Ciguatoxin
11325
Masayuki Satake,† Akio Morohashi,† Hiroki Oguri,‡
Tohru Oishi,‡ Masahiro Hirama,*,‡ Nobuyuki Harada,§ and
Takeshi Yasumoto*,†
Faculty of Agriculture, Tohoku UniVersity
Tsutsumidori-Amamiya, Aoba-ku, Sendai 981, Japan
Graduate School of Science, Tohoku UniVersity
Aramaki Azaaoba, Aoba-ku, Sendai 980-77, Japan
Institute for Chemical Reaction Science
Tohoku UniVersity
Katahira, Aoba-ku, Sendai 980-77, Japan
Figure 1. Structures of CTX (1), CTX4A (2), CTX4B (3), p-
bromobenzoate of CTX4A AB ring fragment (4a), and CTX4A AB
ring fragment (4b).
ReceiVed July 22, 1997
Ciguatera is the name of human intoxication that is caused
by eating certain tropical reef fishes.1 Ciguatoxin (CTX) is the
principal toxin which was first isolated in Hawaii from moray
eel livers. Its structure (1), except for the absolute configuration
and C2 stereochemistry, was later determined by the Yasumoto
group.2 Two ciguatoxin precursors coded CTX4A (2) and
CTX4B (formerly GT4b, 3), were also isolated from epiphytic
dinoflagellate microalgae Gambierdiscus toxicus, and their
relative stereostructures were elucidated (see Figure 1 for
structures).3 Further chemical studies to determine the absolute
configuration of these intriguing toxins were hampered because
of extremely limited availability of the toxins: 0.35 mg of CTX
from 4000 kg of fish and 0.45 mg of 2 from 1100 L of
dinoflagellate cultures. We had assigned R configuration to C5
of CTX4B by comparing its CD spectrum with that of a
synthetic fragment composed of the butadienyl side chain and
AB rings of CTX4B.4 However, uncertainty remained due to
the small Cotton effects observed.
The 2S configuration assigned on the basis of the CD exciton
chirality data of tetrakis-p-bromobenzoates of 1 and tris-p-
bromobenzoates of AB fragments also needed further confirma-
tion because of the difficulty to assign the position of tetrakis-
p-bromobenzoates of 1.5 In view of the importance of knowing
the absolute configuration, we renewed our efforts to eliminate
the above ambiguities. When a new chiral fluorescent reagent
was used in conjunction with the CD exciton chirality method,
the absolute configurations of CTX and CTX4A were success-
fully determined with very small amounts of toxins.
Scheme 1
Scheme 2
As shown in Scheme 1, hydroxyl groups of 1 (5 µg) were
protected as (benzyloxy)methyl (BOM) ethers (5). Compound
5 was cleaved at the C3, C4-double bond with OsO4/NaIO4,
and the resultant aldehyde (6) was immediately reduced with
NaBH4 yielding a glycerol derivative (7).5b Alcohol 7 was
esterified with chiral reagent (S)-2-tert-butyl-2-methyl-1,3-
benzodioxole-4-carboxylic acid [(S)-TBMB-carboxylic acid]
resulting in a fluorescent derivative (8).6 Reference (2S)- and
(2R)-TBMB esters (8) were prepared from (R)- and (S)-2,2-
dimethyl-1,3-dioxolane-4-methanol (9) as shown in Scheme 2.
Structures of the fluorescent esters (8) thus prepared were
confirmed by FABMS (MNa+, m/z 573) and NMR spectra. The
two diastereomers have clearly different retention times on both
normal and reversed phase columns.7 The retention times
(normal phase, 17 min; reversed phase, 54 min) of the derivative
of 1 agreed well with those of the TBMB ester having a 2R
configuration derived from the 2S-standard. Consequently, C2
configuration of 1 was unambiguously determined to be S.
The configuration of C5 in 2 was determined by comparing
CD spectra of 11-p-bromobenzoate of the AB ring fragment of
CTX4A having 5R configuration (4a) and 11,32,47-tris(p-
bromobenzoyl)-CTX4A. p-Bromobenzoate 4a was stereose-
lectively synthesized from 125 as shown in Scheme 3. The CD
spectrum of 4a clearly exhibits a split Cotton effect with a
positive followed by a negative extremum, which is caused by
the expected interaction between the 1,3-diene and p-bromoben-
zoate (MeOH, λext 242 nm, ∆ꢀ +25; λext 225 nm, ∆ꢀ -14).8
(1) Scheuer, P. J.; Takahashi, W.; Tsutsumi, J.; Yoshida, T. Science 1976,
155, 1267-1268.
(2) Murata, M.; Legrand, A.-M.; Ishibashi, Y.; Fukui, M.; Yasumoto,
T. J. Am. Chem. Soc. 1990, 112, 4380-4386.
(7) Normal phase HPLC analysis was done under the following condi-
tions: column, Develosil 60-3 (4.6 × 250 mm, Nomura Chemical) solvent,
hexane/t-BuOH 150:1; flow, 1.7 mL/min; detection, excitation 310 nm and
emission 370 nm T ) 25 °C Reversed phase HPLC was done under the
following conditions: column, Cosmosil 5C18 AR (4.6 × 250 mm, Nakarai
Tesch); solvent, MeOH/H2O 75:25; flow, 1.1 ml/min; detection, excitation
310 nm and emission 370 nm T ) 40 °C.
(8) Unlike that of p-bromobenzoate 4a exhibiting intense dual Cotton
effects, the CD spectrum of alcohol 4b shows a single weak Cotton effect:
4b, CD (MeCN) λext 224 nm (∆ꢀ +6.8). Therefore, as a first approximation,
the contribution of 1,3-diene helicity can be neglected. See also the CD
spectra of related compounds in ref 4.
(3) Satake, M.; Ishibashi, Y.; Legrand, A.-M.; Yasumoto, T. Biosci.
Biochem. Biotech. 1996, 60, 2103-2105.
(4) Suzuki, T.; Sato, O.; Hirama, M.; Yamamoto, Y.; Murata, M.;
Yasumoto, T.; Harada, N. Tetrahedron Lett. 1991, 32, 4505-4508.
(5) (a) Oguri, H.; Hishiyama, S.; Oishi, T.; Hirama, M. Synlett 1995,
1252-1254. (b) Oguri, H.; Hishiyama, S.; Sato, O.; Oishi, T.; Hirama, M.;
Murata, M.; Yasumoto, T.; Harada, N. Tetrahedron 1997, 53, 3057-3072.
(6) (a) Nishida, Y.; Ohrui, H.; Meguro, H. Tetrahedron Lett. 1989, 30,
5277-5280. (b) Kim, J.-H.; Uzawa, H.; Nishida, Y.; Ohrui, H.; Meguro,
H. J. Chromatogr. 1994, 677, 35-43.
S0002-7863(97)02482-7 CCC: $14.00 © 1997 American Chemical Society
Satake
