Enantioselective total synthesis of fumonisin B2

Full text of DOI:10.1021/jo9711347

5666
J . Org. Chem. 1997, 62, 5666-5667
En a n tioselective Tota l Syn th esis of
F u m on isin B2
Yan Shi, Lee F. Peng, and Yoshito Kishi*
Department of Chemistry and Chemical Biology,
Harvard University, Cambridge, Massachusetts 02138
Received J une 23, 1997
The work on the assignment of the stereochemistry of
AAL toxins/fumonisins1 and maitotoxin2 from this labo-
ratory has yielded a vast volume of experimental data
on the structural properties of fatty acids and related
classes of compounds. These data show that the struc-
tural properties of a compound in question are inherent
to the specific stereochemical arrangements of (small)
substituents on its carbon backbone and are independent
from the rest of the molecule. We then recognized the
possibility that fatty acids and related compounds bear-
ing (small) substituents on their backbones have the
capacity of creating unique structural motifs, carrying
specific information, and serving as functional materials.2a
In this context, we are intrigued with the fumonisin
was subsequently installed on to the backbone. The
benzyl group was chosen to protect the C2 amino, the
C3 and C5 hydroxyl,6 and the TCA carboxylic groups so
that all the protecting groups could be removed in a
single step at the end of synthesis.
The synthesis of the left segment 2 began with coupling
of the chiral alkyne 57 with the triflate 68 to give the
alkyne 7 (Scheme 1). The alkyne 7 was converted to the
trans-alkene acid 8 via (1) site-selective osmylation, (2)
Pb(OAc)4 cleavage of the resultant diol, followed by
NaBH4 reduction, (3) Na/NH3 reduction of the alkyne to
class of natural products. Fumonisin B2 (FB2, 1),3
a
a trans-alkene, and (4) Swern9 and then NaClO2 oxida-
10
representative member of this class of mycotoxins, is
known to exhibit a wide range of biological activities.4
Structurally, FB2 possesses two distinct halves, each
containing clustered chiral centers on its backbone, which
are separated by six methylene units. Thus, according
to our hypothesis, FB2 contains two structural motifs, and
each of them could be linked to a specific biological
event(s) independent from the other. This class of
natural products might therefore present an interesting
opportunity to test our hypothesis experimentally. In
this communication, we report an enantioselective total
synthesis of FB2 that is flexible and effective for the
preparation of the remote diastereomers of FB2 and other
analogs required for the proposed experimental work.
A convergent approach to FB2 was adopted, with the
molecule being divided into three fragmentssthe left
segment 2, the right segment 3, and the tricarballylic acid
(TCA) segment 4.5 A Wittig reaction between 2 and 3
was used to form the backbone, and the TCA segment 4
tions of the primary alcohol to the acid. The vicinal
hydroxyl groups at C14 and C15 were stereoselectively
introduced on the backbone of 8 in three steps: (1)
iodolactonization of 8 under equilibrium conditions in
CH3CN at -30 °C to give the iodo lactone 9 in 84% yield,
with a diastereomeric ratio greater than 20:1,11 (2) ring
opening of the lactone with PhCH2ONa to yield the C14-
C15 epoxide benzyl ester, and (3) deprotection of the
resultant benzyl ester, with concomitant epoxide ring
opening, to furnish the lactone alcohol with the desired
stereochemistry at both C14 and C15. The lactone
alcohol was reduced to a triol, the two vicinal hydroxyl
groups were protected as an acetonide, and Swern
oxidation of the resultant primary alcohol furnished the
left segment 2.12
The synthesis of the right segment 3 is outlined in
Scheme 2. Allylation of R-amino aldehyde 1013,14 with
(6) The numbering system adopted in this paper corresponds to that
of FB2, cf. the structure 1.
(1) (a) For the stereochemistry assignment of the AAL toxins and
fumonisins from this and other groups up to the beginning of 1995,
see: Boyle, C. D.; Kishi, Y. Tetrahedron Lett. 1995, 36, 5695 and
references cited therein. (b) For more recent work on this subject, see:
Blackwell, B. A.; Edwards, O. E.; Fruchier, A.; ApSimon, J . W.; Miller,
J . D. Fumonisins in Food; J ackson, L., et al., Eds.; Plenum Press: New
York, 1996; pp 75-91 and references cited therein.
(2) For the stereochemistry assignment of maitotoxin, see: (a)
Zheng, W.; Demattei, J . A.; Wu, J .-P.; Duan, J . J .-W.; Cook, L. R.;
Oinuma, H.; Kishi, Y. J . Am. Chem. Soc. 1996, 118, 7946. (b)
Nonomura, T.; Sasaki, M.; Matsumori, N.; Murata, M.; Tachibana, K.;
Yasumoto, T. Angew. Chem., Int. Ed. Engl. 1996, 35, 1675 and
references cited therein.
(3) Bezuidenhout, S. C.; Gelderblom, W. C. A.; Gorst-Allman, C. P.;
Horak, R. M.; Marasas, W. F. O.; Spiteller, G.; Vleggaar, R. J . Chem.
Soc., Chem. Commun. 1988, 743.
(4) For examples, see: (a) Mycopathologia 1992, 117, pp 1-124 for
18 reviews regarding various aspects of biological activity of fumonisins
and AAL toxins. (b) Merrill, A. H., J r.; Wang, E.; Gilchrist, D. G.; Riley,
R. T. Adv. Lipid Res. 1993, 26, 215.
(5) The TCA segment 4 was synthesized by using the asymmetric
Michael reaction reported by Hanessian [Hanessian, S.; Gomtsyan, A.;
Payne, A.; Herve´, Y.; Beaudoin, S. J . Org. Chem. 1993, 58, 5032]; the
antipode of compound 8 in the Hanessian paper was subjected to (1)
O3, then J ones oxidation, (2) BnOH, EDCI, DMAP, and (3) TFA to give
(-)-4 in approximately 29% overall yield. Its optical purity was
estimated to be greater than 90% ee from the 1H NMR spectrum of its
(-)-menthol ester, and its absolute configuration was established by
a chemical correlation with a known compound [Boyle, C. D.; Kishi,
Y. Tetrahedron Lett. 1995, 36, 4579].
(7) The alkyne 5 was synthesized in 65% overall yield from (R)-2-
methyl-1-hexanol [Myers, A. G.; Yang, B. H.; Chen, H.; Gleason, J . L.
J . Am. Chem. Soc. 1994, 116, 9361] via a Swern oxidation,9 followed
by a Corey-Fuchs protocol [Corey, E. J .; Fuchs, P. L. Tetrahedron Lett.
1972, 3769]. The optical purity of the alcohol was estimated to be
greater than 95% ee from the 1H NMR spectrum of its Mosher ester.
(8) The triflate 6 was synthesized from (S)-2,5-dimethyl-4-hexen-
1-ol, obtained by using the pseudoephedrine-based asymmetric alky-
lation developed by Myers.7 The optical purity of the alcohol was
estimated to be greater than 95% ee from the 1H NMR spectrum of its
Mosher ester.
(9) Mancuso, A.; Huang, S.-L.; Swern, D. J . Org. Chem. 1978, 43,
2480.
(10) Kraus, G. A.; Taschner, M. J . J . Org. Chem. 1980, 45, 1175.
(11) For reviews on this subject, see: (a) Cardillo, G.; Orena, M.
Tetrahedron 1990, 46, 3321. (b) Bartlett, P. A.; Richardson, D. P.;
Myerson, J . Tetrahedron 1984, 40, 2317.
(12) Previous work in this laboratory showed that the installation
of the C14 and C15 hydroxyl groups via dihydroxylation of a cis-olefin
without the use of a chiral ligand resulted in a 1:3 ratio of desired to
undesired diols, with the best result being a 1:1 ratio of desired to
undesired diols using the Sharpless DHQ-IND ligand. See: (a) the
supporting information of Boyle, C. D.; Harmange, J .-C.; Kishi, Y. J .
Am. Chem. Soc. 1994, 116, 4995. (b) Boyle, C. D. Ph.D. Thesis, Harvard
University, 1995.
(13) Fehrentz, J -A.; Castro, B. Synthesis 1983, 676.
(14) The optical purity of 10 was estimated to be greater than 96%
ee from the 1H NMR spectrum of the Mosher ester prepared from 10
via NaBH4 reduction followed by esterification.
S0022-3263(97)01134-1 CCC: $14.00 © 1997 American Chemical Society

Communications
J . Org. Chem., Vol. 62, No. 17, 1997 5667
Sch em e 1
Sch em e 3
Wittig reaction of the ylide generated from the phospho-
nium salt 3 with the aldehyde 2 in 80% yield. Trifluo-
roacetic acid treatment to remove the acetonide and
acylation of the resultant diol with the (-)-TCA segment
45 gave the fully protected FB2 15 in 90% yield. Hydro-
genation of the alkene and hydrogenolysis of all the
benzyl protecting groups were accomplished under the
conditions of H2 (1 atm)/Pearlman’s catalyst/HCl (3
equiv)/t-BuOH18-THF/rt, to furnish the synthetic FB2 in
60% yield. On comparison of spectroscopic and chro-
matographic data, the synthetic FB2 was found to be
identical with naturally occurring FB2.19,20
Sch em e 2
In conclusion, the first total synthesis of FB2 was
accomplished in a convergent manner from the three
building blocks 2, 3, and 4. This synthesis was designed
in such a way that the remote diastereomers of FB2 and
their analogs, with different tether lengths or different
tethers between the two structural motifs, can also be
synthesized for the purpose of probing our hypothesis
experimentally.21
Ack n ow led gm en t. Financial support from the Na-
tional Institutes of Health (NS 12108) and the National
Science Foundation (CHE 89-09762) is gratefully ac-
knowledged. We thank the National Institutes of Health
for a Postdoctoral Fellowship to Y.S. (GM 17540).
Brown’s chiral (-)-B-allyldiisopinocampheylborane15 gave
syn-amino alcohol 11 in 75% yield with a diastereoselec-
tivity of 94%.16 Protection of 11 as an acetonide and
ozonolysis of the alkene with a dimethyl sulfide workup
provided an aldehyde, which was then reacted with (+)-
B-allyldiisopinocampheylborane15 to give the anti-alcohol
12 with a diastereomeric ratio of about 10:1 in 65%
overall yield.17 Acetonide deprotection of 12 followed by
benzyl group protection provided an alkene, which was
transformed into an aldehyde by ozonolysis with a
dimethyl sulfide workup. A two-carbon chain elongation
in two steps, i.e., (1) Horner-Wadsworth-Emmons ole-
fination and (2) hydrogenation (Lindlar catalyst), fur-
nished the ester 13 in 70% overall yield from 12.
Removal of the BOC group, protection of the resultant
amine, and DIBAL reduction of the methyl ester resulted
in an alcohol, which was transformed into an alkyl iodide
and then treated with PPh3 in acetonitrile to give the
right-half segment 3 in 75% overall yield from 13.
The completion of FB2 synthesis is depicted in Scheme
3. As planned, the backbone 14 was installed through a
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for the syntheses reported and 1H NMR spectra of FB2
(synthetic and natural) and its diastereomer at the TCA moiety
(17 pages).
J O9711347
(18) The choice of solvents for hydrogenation/hydrogenolysis is
important. For example, conducting the reduction in methanol led to
the formation of methyl esters.
(19) Fumonisin B2 was purchased from the South African Research
Council, Tygerberg, South Africa, and was isolated from the Fusarium
moniliforme strain MRC 826.
(20) There is a disagreement on the absolute configuration of the
TCA moieties of fumonisins; see (a) Shier, W. T.; Abbas, H. K.; Badria,
F. A. Tetrahedron Lett. 1995, 36, 1571 and (b) Boyle, C. D.; Kishi, Y.
Tetrahedron Lett. 1995, 36, 5695. Therefore, the diastereomer of FB2
bearing (S)-TCA was also synthesized from 14 and the antipode of 4.
It exhibited a 1H NMR spectrum distinctly different from that of the
natural and synthetic FB2 (see the spectra attached in Supporting
Information), thereby confirming our previous conclusion.
(21) By combination of the left segment 2 or its antipode with the
right segment 3 or its antipode, we have recently completed the
synthesis of all four remote diastereomers of the FB2 aminopolyol
backbone and also all the possible (+)- and (-)-4 attached FB2 remote
diastereomers. In addition, the olefin 12 provides easy access to analogs
with different tether lengths or different tethers between the two
structural motifs. We will study the biological behaviors of these
analogs, in reference to the eicosane derivatives bearing only the left-
half or right-half functional groups.
(15) Brown, H. C.; Bhat, K. S.; Randad, R. S. J . J . Org. Chem. 1989,
54, 1570.
(16) When (+)-B-allyldiisopinocampheylborane was employed, the
anti-amino alcohol corresponding to 12 was obtained in 94% de.
(17) When (-)-B-allyldiisopinocampheylborane was used, the syn-
alcohol corresponding to 13 was obtained in greater than 18:1 ratio.