Biomimetic synthesis of trans,syn,trans-fused polyoxepanes: Remarkable substituent effects on the endo-regioselective oxacyclization of polyepoxides

Article abstract of DOI:10.1021/ja050013i

The biogenesis of trans,syn,trans-fused polycyclic ether substructures of neurotoxic natural products of the brevetoxin-ciguatoxin family may involve regio- and stereoselective oxacyclizations of polyepoxides. We report that a biomimetic Lewis acid-promot

Full text of DOI:10.1021/ja050013i

Published on Web 03/10/2005
Biomimetic Synthesis of trans,syn,trans-Fused Polyoxepanes: Remarkable
Substituent Effects on the endo-Regioselective Oxacyclization of
Polyepoxides
Jason C. Valentine, Frank E. McDonald,* Wade A. Neiwert,† and Kenneth I. Hardcastle†
Department of Chemistry, Emory UniVersity, Atlanta, Georgia 30322
Received January 3, 2005; E-mail: fmcdona@emory.edu
Scheme 1. 2,3-Epoxysilane Oxacyclizationa
trans,syn,trans-Fused polycyclic ether structures are found in
many biologically active marine natural products.1 These com-
pounds have received much attention from the synthetic community
and many elegant methods toward their synthesis have been
described.2 However, these syntheses generally rely on iterative
strategies in which each cyclic ether is formed over a stepwise series
of operations. Our laboratory has previously described the bio-
mimetic synthesis of trans,syn,trans-fused polyoxepanes3 and
polypyrans4 via Lewis acid-intitiated endo-regioselective tandem
oxacyclization of polyepoxides.
a Conditions: (a) BF3-OEt2, CH2Cl2, -40 °C.
Scheme 2. 2,3-Epoxysilane Oxacyclization/Protiodesilylationa
Our mechanistic hypothesis (Figure 1) for tandem endo-oxa-
cyclization involves Lewis acid activation 1 followed by nucleo-
philic addition of 6,7-epoxide to give fused bicyclic epoxonium
ion 2. Subsequent epoxide additions occur as shown for 2 f 3,
until the oxacyclization cascade is terminated by a tethered carbonyl
nucleophile, to provide all-trans,syn,trans-polycyclic ether product,
i.e., 4.
a Conditions: (a) 11, BF3-OEt2, CH2Cl2, -40 °C; Ac2O, pyridine. (b)
12, BF3-OEt2, CH2Cl2, -40 °C; Ac2O, pyridine; CF3CO2H. (c) Bu4NF,
THF; Ac2O, pyridine.
Scheme 3. 6,7-Epoxysilane Oxacyclizationa
Figure 1. Proposed mechanism of Lewis acid-initiated oxacyclization.
From previous work by Coxon and others5 on regioselective
hydroxyepoxide oxacyclizations, we anticipated that Lewis acid-
initiated oxacyclization of disubstituted epoxides would not yield
the endo-regiochemistry required for trans,syn,trans-fused poly-
cyclic ethers, but would instead give the undesired exo-oxacycliza-
tion product. For this reason all previous examples from our
laboratory had utilized alkyl substitution (i.e., R′ ) R′′ ) Me)3,4
to overcome the kinetic and stereoelectronic bias toward exo-
oxacyclization. However, alkyl substitution is not found at all ring
junctions in the naturally occurring fused polycyclic ethers. Herein
we report the extension of our Lewis acid-initiated oxacyclization
reaction to include 6,7-disubstituted epoxides (R′ ) Me, R′′ ) H),
and further disclose the application of 3-silyl epoxides (R′ ) SiMe3,
R′′ ) Me) to achieve endo-regiochemistry.
Preliminary experiments with the diepoxide 5 from 3-desmethyl-
geraniol (Scheme 1) confirmed the preference for exo-cyclization
predicted by Coxon5 to give 7 in low yield.6 However further
experimentation revealed that 3-silyl diepoxide 87,8 favored regio-
selective tandem endo-oxacyclization to afford 99 in good yield.
We next investigated 3-silyl triepoxides 11 and 126-8 (Scheme
2), which upon treatment with freshly distilled BF3-OEt2 at
a Conditions: (a) BF3-OEt2, CH2Cl2, -40 °C. (b) Ac2O, pyridine.
-40 °C provided the trans,syn,trans-fused dioxepane product 13.
The more nucleophilic dimethylcarbamate consistently gives su-
perior yields over the tert-butyl carbonate in these oxacyclization
reactions (vide infra). Treatment of 13 with TBAF followed by
peracylation gave ring-opened dioxepane 14, corresponding to
protiodesilylation at C-3 accompanied by cyclic carbonate removal.9
On the basis of Coxon’s observations,5 we expected silyl sub-
stitution at the 6,7-epoxide would also be required to achieve endo-
oxacyclization (Figure 1, R′ ) Me, R′′ ) SiMe3). The reactions of
both 15 and 16 with BF3-OEt2 gave the bisoxepane-carbonate 17
as the major product (Scheme 3).6 However, 7-silyl substituted
dioxepane 17 could not be protiodesilylated or functionalized under
a variety of conditions, apparently as a neighboring hydroxyl group
was not available to assist silicon migration from carbon.
These results led us to directly investigate the cyclization
regioselectivity of 6,7-disubstituted epoxides (Figure 1, R′ ) Me,
R′′ ) H; Scheme 4). In contrast to expectations,5 both substrates
18 and 1910 provided the all-endo oxacyclization product best
characterized as the ester derivative 20.11 The structure of 20 was
confirmed by single-crystal X-ray analysis.6 Likewise, oxacycliza-
† Emory University X-ray Crystallography Laboratory.
9
4586
J. AM. CHEM. SOC. 2005, 127, 4586-4587
10.1021/ja050013i CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Scheme 4. Oxacyclizations of Disubstituted Epoxide Substratesa
of polyepoxides. We note the reaction yields for oxacyclization of
substrates 21 and 22, 20% and 25% respectively, are similar to
that of the geranylgeraniol-derived tetraepoxide substrate previously
reported (27% yield3b). We have further demonstrated the utility
of 2,3-epoxysilanes (8, 11, 12) to serve as a regioselectivity-
directing surrogate in this tandem oxacyclization reaction, which
can be efficiently removed. These findings greatly expand the scope
of biomimetic oxacyclization methodology so that naturally oc-
curring, non-terpene-derived polycyclic ethers can now be ef-
ficiently prepared by our approach.
Acknowledgment. This research was supported by the National
Science Foundation (Grant No. CHE-9982400). We also acknowl-
edge the use of shared instrumentation (NMR spectroscopy, X-ray
diffractometry, polarimetry) provided by grants from the National
Institutes of Health, the National Science Foundation, the Georgia
Research Alliance, and the University Research Committee of
Emory University. We thank Dr. Shaoxiong Wu of the Emory
University NMR facility for his assistance with NOESY spectros-
copy experiments. J.C.V. also acknowledges a GAANN fellowship
(2000-2002) and ARCS Foundation fellowship (2004-2005).
a Conditions: (a) BF3-OEt2, CH2Cl2, -40 °C. (b) Ac2O, pyridine.
tion of substrates 21 and 2210 bearing two “internal” disubstituted
epoxides also resulted in the formation of trans,syn,trans-fused
tetracyclic product 23.12
The general endo-regioselectivity of substrates 18, 19, 21, and
22 is consistent with the idea of nucleophile-driven regiochemical
control (Figure 2)3 where the nature of the nucleophile (i.e., epoxide
vs carbonyl) drives the regioselectivity of oxacyclization with
disubstituted epoxide electrophiles. In the case of the 2,3-epoxide
nucleophile addition to either C6 or C7 of epoxonium ion 24, endo-
cyclization is observed due to formation of fused bicyclo[4.1.0]
intermediate 25 rather than bicyclo[3.1.0] 26 which would have
arisen from exo addition. We speculate that 25 is favored due to
minimization of ring strain in formation of [4.1.0] epoxonium ion
25 relative to 26.
Supporting Information Available: Experimental procedures and
spectroscopic data for the preparation of all polyepoxide substrates and
their cyclization products; thermal ellipsoid figures and essential data
for crystal structures of a derivative of compound 9, and compound
20. X-ray crystallographic data in CIF format. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Marine Toxins: Origin, Structure and Molecular Pharmacology; Hall, S.,
Strichartz, G., Eds.; ACS Symposium Series 418; American Chemical
Society: Washington, DC, 1990.
(2) (a) Inoue, M. Org. Biomol. Chem. 2004, 2, 1811. (b) Marmsa¨ter, F. P.;
West, F. G. Chem. Eur. J. 2002, 8, 4347. (c) Evans, P. A.; Delouvrie´, B.
Curr. Opin. Drug DiscoVery 2002, 5, 986.
(3) (a) McDonald, F. E.; Wang, X.; Do, B.; Hardcastle, K. I. Org. Lett. 2000,
2, 2917. (b) McDonald, F. E.; Bravo, F.; Wang, X.; Wei, X.; Toganoh,
M.; Rodriguez, J. R.; Do, B.; Neiwert, W. A.; Hardcastle, K. I. J. Org.
Chem. 2002, 67, 2515. (c) Bravo, F.; McDonald, F. E.; Neiwert, W. A.;
Hardcastle, K. I. Org. Lett. 2004, 6, 4487.
(4) (a) Bravo, F.; McDonald, F. E.; Neiwert, W. A.; Do, B.; Hardcastle, K.
I. Org. Lett. 2003, 5, 2123. For other tandem cyclization syntheses of
fused polycyclic ethers, see: (b) Tokiwano, T.; Fujiwara, K.; Murai, A.
Synlett 2000, 335. (c) Fujiwara, K.; Hayashi, N.; Tokiwano, T.; Murai,
A. Heterocycles 1999, 50, 561. (d) Zakarian, A.; Batch, A.; Holton, R.
A. J. Am. Chem. Soc. 2003, 125, 7822.
(5) (a) Janda, K. D.; Shevlin, C. G.; Lerner, R. A. J. Am. Chem. Soc. 1995,
117, 2659. (b) Coxon, J. M.; Thorpe, A. J. J. Am. Chem. Soc. 1999, 121,
10955 and references therein.
Figure 2. Rationale for regioselectivity directed by epoxide nucleophile.
With carbonyl nucleophiles (i.e., tert-butyl carbonate and N,N-
dimethylcarbamate) there is little ring strain associated with either
intermediate 29 or 30,3a,b and the kinetically anticipated exo product
arising from 28 (H at C-3) predominates in substrates with 2,3-
disubstituted epoxide electrophiles. To achieve the endo-oxa-
cyclization to 29 via 27 in the terminal cyclization, either an alkyl
substituent (R′ ) Me) or removable surrogate such as R′ ) SiMe3
is essential (Figure 3).
(6) See Supporting Information for experimental details.
(7) Lead reference for construction of polyene: (a) Lipshutz, B. H.; Bulow,
G.; Fernandez-Lazaro, F.; Kim, S.-K.; Lowe, R.; Mollard, P.; Stevens,
K. L. J. Am. Chem. Soc. 1999, 121, 11664. (b) For regio- and
stereoselective introduction of silicon, see: Kim, K. D.; Magriotis, P. A.
Tetrahedron Lett. 1990, 31, 6137. Lead references for asymmetric
epoxidations: (c) Gao, Y.; Hanson, R. M.; Klunder J. M.; Ko, S. Y.;
Masamune, H.; Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765. (d)
Wang, Z.; Tu, Y.; Frohn, M.; Zhang, J.; Shi, Y. J. Am. Chem. Soc. 1997,
119, 11224.
In conclusion, we have discovered that “internal” disubstituted
epoxides are viable substrates in Lewis acid-initiated oxacyclization
(8) (a) For an excellent review of R,â-epoxysilane chemistry: Hudrlik, P.
F.; Hudrlik, A. M. R,â-Epoxysilanes. In AdVances in Silicon Chemistry;
Larson, G. L., Ed.; JAI Press: Greenwich, CT, 1993; Vol. 2, pp 1-89.
For conceptually similar oxacyclizations of R,â-epoxysilanes: (b) Adiwid-
jaja, G.; Flo¨rke, H.; Kirschning, A.; Schaumann, E. Tetrahedron Lett. 1995,
36, 8771. (c) Heffron, T. P.; Jamison, T. F. Org. Lett. 2003, 5, 2339.
(9) Compound 9 also undergoes efficient protiodesilylation (Supporting
Information). Careful studies of the protiodesilylation of 9 and 13 reveal
that carbonate opening precedes protiodesilylation. The liberated primary
alcohol then assists in protiodesilylation via Brook-type silicon migration.
(10) The polyalkene precursors were generated by iterative ortho ester Claisen
rearrangement: Johnson, W. S.; Werthemann, L.; Bartlett, W. R.;
Brocksom, T. J.; Li, T.; Faulkner, D. J.; Petersen, M. R. J. Am. Chem.
Soc. 1970, 92, 741.
(11) The structure of compound 20 was confirmed by single-crystal X-ray
analysis.
(12) The structure of compound 23 was confirmed by NOESY spectroscopy.
Figure 3. Rationale for regioselectivity in terminal cyclization step.
JA050013I
9
J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4587