Tandem oxidative cyclization with rhenium oxide. Total synthesis of 17,18-bisepi-goniocin

Full text of DOI:10.1021/ja964273z

12014
J. Am. Chem. Soc. 1997, 119, 12014-12015
Scheme 1
Tandem Oxidative Cyclization with Rhenium Oxide.
Total Synthesis of 17,18-bisepi-Goniocin
Subhash C. Sinha,*,† Anjana Sinha,† Santosh C. Sinha,‡ and
Ehud Keinan†,‡
Department of Molecular Biology and Skaggs Institute for
Chemical Biology, The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, California 92037
Department of Chemistry, Technion-Israel Institute of
Technology, Technion City, Haifa 32000, Israel
Scheme 2. Total Synthesis of 2, First Approacha
ReceiVed December 12, 1996
Many of the acetogenins isolated from the Annonaceae plants1
have shown remarkable cytotoxic, antitumor, antimalarial,
immunosuppressive, pesticidal, and antifeedant activities.2
Classification of these fatty acid derivatives into three subgroups
is based on the number and relative positioning of the tetrahy-
drofuran moieties within the molecule: the mono-THF, the
adjacent bis-THF, and the nonadjacent bis-THF acetogenins.1
We have recently shown that many acetogenins of the first and
second subgroups, including solamin, reticulatacin, asimicin,
bullatacin, trilobacin, and trilobin, can be efficiently synthesized3
either by a convergent approach or via the “naked” carbon
skeleton strategy,4,5 combining the Sharpless asymmetric dihy-
droxylation (AD) reaction6 with the Kennedy oxidative cycliza-
tion reaction.7
Goniocin, which has been recently isolated from Goniothala-
mus giganteus,8 possesses three adjacent THF rings and,
therefore, represents the first example of a new subclass of
Annonaceous acetogenins. Structure 1 was proposed for
1
goniocin on the basis of its MS and H and 13C NMR data.8
Clearly, construction of the tris-trans-THF fragment I with the
appropriate configuration of the seven stereogenic carbinol
centers represents the main challenge in the synthesis of 1. Our
retrosynthetic analysis (Scheme 1) was based on previous
findings that two consecutive oxidative cyclizations with 4,8-
dienols can be carried out in a single step to produce bis-THF
derivatives.9 We reasoned that I could be synthesized from a
4,8,12-trienol substrate using the tandem oxidative cyclization
methodology. Coupling of I with the butenolide fragment II
could lead to an efficient total synthesis of 1.
Here, we report that all trans-4,8,12-trienol substrates indeed
undergo a highly stereospecific triple oxidative cyclization
reaction in the presence of a rhenium(VII) reagent to produce
a single stereoisomer of a tris-THF product. Surprisingly,
however, the product’s stereochemistry is not trans-threo-trans-
threo-trans-threo as expected, but trans-threo-cis-threo-cis-
a (a) Ti(OiPr)4, (-)-DIPT, TBHP, powdered Molecular Sieves 4A,
-20 °C, 6 h. (b) Red-Al, THF, 0 °C, 4 h. (c) TBDPSCl, diisopropyl-
ethylamine, CH2Cl2, room temperature (rt) 16 h. (d) Re2O7, TFAA,
THF, rt, 1 h, concentration under vacuum and washing with cold
pentane, then alcohol 6, TFAA, CH2Cl2, 0 °C to rt, 6 h. (e) i. MOMCl,
diisopropylethylamine, CH2Cl2, 0 °C to rt, 16 h; ii. TBAF, THF, 0 °C
to rt, 2 h; iii. I2, PPh3, 0 °C to rt, 2 h; iv. PPh3, NaHCO3, CH3CN, 45
°C, 48 h. (f) BuLi, THF, 0 °C, then aldehyde II. (g) i. H2, Wilkinson’s
catalyst (20%, w/w), C6H6-EtOH (4:1), rt, 4 h; ii. 4% AcCl in MeOH/
CH2Cl2 (1:1, v/v), rt, 16 h.
threo. Consequently, we have synthesized 17,18-bisepi-
goniocin (2) rather than 1.
The key intermediate in our synthesis (Scheme 2) is the
“naked” carbon skeleton (6) which is easily prepared from (E,E)-
ethyl heneicosa-4,8-dienoate3a (see the Supporting Information).
Asymmetric epoxidation10 of 3 using Ti(OPr)4 and (-)-DIPT
produces epoxy alcohol 4 in more than 95% ee. Reductive
cleavage of the epoxide ring using Red-Al affords the 1,3-diol
5,11 which is then monoprotected at the primary position to give
the silyl ether 6.
We planned to use the single stereogenic center in 6 as the
only source of chirality at the tris-THF fragment and achieve
the other six stereogenic carbinol centers by a tandem oxidative
cyclization reaction using a Re(VII) reagent. We found that
both reagents originally used by Kennedy, i.e. Re2O7/lutidine
and Re2O7/H5IO6 in dichloromethane, are useful for monocy-
clization with simple substrates possessing one double bond.
However, for double cyclization with substrates containing two
double bonds, the more reactive mixture, Re2O7/H5IO6, was
† The Scripps Research Institute.
‡ Technion.
(1) Zeng, L.; Ye, Q.; Oberlies, N. H.; Shi, G.; Gu, Z.-M.; He, K.;
McLaughlin, J. L. Nat. Prod. Rep. 1996, 275.
(2) Gu, Z.-M.; Zhao, G.-X.; Oberlies, N. H.; Zeng, L.; McLaughlin, J.
L. In Recent AdVances in Phytochemistry; Arnason, J. T., Mata, R., Romeo,
J. T., Eds.; Plenum Press: New York, 1995; Vol. 29, pp 249-310.
(3) (a) Sinha, S. C; Keinan, E. J. Am. Chem. Soc. 1993, 115, 4891. (b)
Sinha, S. C.; Sinha-Bagchi, A.; Yazbak, A.; Keinan, E. Tetrahedron Lett.
1995, 36, 9257. (c) Sinha, S. C.; Sinha, A.; Yazbak, A.; Keinan, E. J. Org.
Chem. 1996, 61, 7640. (d) Sinha, A.; Sinha, S. C.; Keinan, E. Submitted.
(4) Keinan, E.; Sinha, A.; Yazbak, A.; Sinha, S. C.; Sinha, S. C. Pure
Appl. Chem. 1997, 69, 423.
(5) (a) Sinha, S. C; Keinan, E. J. Org. Chem. 1994, 59, 949. (b) Sinha,
S. C; Keinan, E. J. Org. Chem. 1997, 62, 377. (c) Sinha, S. C; Sinha-
Bagchi, A.; Keinan, E. J. Org. Chem. 1993, 58, 7789.
(6) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. ReV.
1994, 94, 2483.
(7) (a) Tang, S.; Kennedy, R. M. Tetrahedron Lett. 1992, 33, 3729. (b)
Tang, S.; Kennedy, R. M. Tetrahedron Lett. 1992, 33, 5299. (c) Tang, S.;
Kennedy, R. M. Tetrahedron Lett. 1992, 33, 5303. (d) Boyce, R. S.;
Kennedy, R. M. Tetrahedron Lett. 1994, 35, 5133.
(8) Gu, Z.-M.; Fang, X.-P.; Zeng, L.; McLaughlin, J. L. Tetrahedron
Lett. 1994, 35, 5367.
(9) Sinha, S. C; Sinha-Bagchi, A.; Keinan, E. J. Am. Chem. Soc. 1995,
117, 1447.
(10) (a) Johnson, R. A.; Sharpless, K. B. In Catalytic Asymmetric
Synthesis; Ojima, I., Ed.; VCH Publishers Inc.: New York, 1993; p 103.
(b) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.; Masamune, H.;
Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765.
(11) (a) Finan, J. M.; Kishi, Y. Tetrahedron Lett. 1982, 23, 2719. (b)
Viti, S. M. Tetrahedron Lett. 1982, 23, 4541.
S0002-7863(96)04273-4 CCC: $14.00 © 1997 American Chemical Society

Communications to the Editor
J. Am. Chem. Soc., Vol. 119, No. 49, 1997 12015
Scheme 3. Total Synthesis of 2, Second Approacha
found to be more useful in terms of both reaction rate and yield.9
Yet, for triple cyclization, although this reagent does produce a
tris-THF product, the reaction conditions are too acidic to be
compatible with the silyl protecting group. Therefore, we
examined various alternative perrhenate reagents, including the
Wilkinson’s mononuclear perrhenate complex12 (trifluoro-
acetyl)perrhenate (CF3CO2ReO3), which has been recently used
by McDonald.13 We found that a mixture of CF3CO2ReO3 and
trifluoroacetic anhydride in dichloromethane is a better reagent
for oxidative polycyclizations. Thus, when trienol 6 was treated
with a mixture of CF3CO2ReO3 and trifluoroacetic anhydride,
a stereochemically pure tris-THF product was isolated in 48%
a (a) i. MOMCl, diisopropylethylamine, CH2Cl2, 0 °C to rt, 16 h; ii.
TBAF, THF, 0 °C to rt, 2 h; iii. I2, PPh3, 0 °C to rt, 2 h; iv. PPh3,
NaHCO3, CH3CN, 45 °C, 48 h. (b) i. BuLi, THF, 0 °C, then aldehyde
II, 0.5 h; ii. TMSBr, CH2Cl2, -30 °C, 1 h. (c) Re2O7, TFAA, THF, rt,
1 h, concentration under vacuum and washing with cold pentane, then
alcohol 11, CH2Cl2, TFAA, 0 °C to rt, 3 h. (d) i. H2, Wilkinson’s catalyst
(20%, w/w), C6H6-EtOH (4:1), rt, 4 h; ii. 4% AcCl in MeOH, CH2Cl2
(1:1, v/v), rt, 16 h.
1
yield.14,15 Surprisingly, H and 13C NMR data of this product
did not match the expected characteristics of the trans-threo-
trans-threo-trans-threo stereoisomer 1. We assigned the ster-
eochemistry of 7 as trans-threo-cis-threo-cis-threo on the basis
of 2D-NMR experiments (1H-1H COSY, TOCSY, and ROE-
SY)16 as well as by independent synthesis.17
The tris-THF derivative 7 was converted to the phosphonium
salt 8 in a four-step sequence: we protected the secondary
alcohol in the form of a MOM ether, deprotected the primary
alcohol by desilylation, converted it to the corresponding iodide,
and then substituted with triphenylphosphine to produce 8.
Wittig reaction of the latter with aldehyde II (for the synthesis
of II, see the Supporting Information) afforded alkene 9.
Finally, hydrogenation and removal of the protecting groups
afforded 2.
In the above-described synthesis of 2, we placed the triple-
cyclization step at an early stage of the synthetic scheme, before
the attachment of the butenolide fragment. Alternatively, since
the tandem oxidative cyclization reaction is compatible with
many functional groups, it could be carried out in a much later
stage, after the completion of the molecular carbon skeleton.
To check this possibility, we converted trienol 6 to the
phosphonium salt 10 and then coupled it with aldehyde II
(Scheme 3). The resultant tetraenol (11) represents an interest-
ing substrate for the tandem oxidative cyclization reaction
because it is far more complex than substrate 6 in terms of both
molecular size and number of functional groups that could
potentially bind the rhenium metal. Moreover, substrate 11
provides an interesting opportunity to examine the relative
reactivity of the rhenium reagent toward homoallylic and bis-
homoallylic alcohols. In principle, as the secondary carbinol
center in 11 is both homoallylic and bis-homoallylic, competition
between the two oxidative cyclization alternatives might have
occurred. Nevertheless, cyclization with trifluoroacetyl per-
rhenate and trifluoroacetic anhydride took place exclusively at
the bis-homoallylic site, producing the tris-THF intermediate
12 in 50% yield. This is a remarkable observation, considering
the fact that six new stereogenic centers are formed in a single
transformation with very high diastereoselectivity, producing a
compound that is only two steps away from the final target
molecule. Indeed, hydrogenation and deprotection of 12
afforded 2.
In conclusion, we have shown that the use of trifluoroacetyl
perrhenate for oxidative polycyclization is a general reaction,
compatible with complex molecular structures and a variety of
functional groups, including a homoallylic alcohol, a silyl ether,
and an R,â-unsaturated γ-lactone. The very high regio- and
diastereoselectivity of this reaction make it a powerful tool for
synthesis of polyoxygenated carbon skeletons and for the
Annonaceous acetogenins in particular. Thus, the asymmetric
total synthesis of 2 has been achieved via two alternative routes
(in 11 steps with 4.97% yield or in 12 steps with 8.07% yield),
starting from the achiral hydrocarbon 3. All asymmetric centers
in the main fragment have been introduced in two synthetic
steps: asymmetric epoxidation and tandem oxidative cyclization
reaction.
(12) (a) Edwards, P. Wilkinson, G. J. Chem. Soc., Dalton Trans. 1984,
2695. (b) Herrmann, W. A.; Thiel, W. R.; Ku¨hn, F. E.; Fischer, R. W.;
Kleine, M.; Herdtweck, E.; Scherer, W.; Mink, J. Inorg. Chem. 1993, 32,
5188.
(13) McDonald, F. E.; Towne, B. T. J. Org. Chem. 1995, 60, 5750.
(14) The general procedure used for the tricyclization reaction with CF3-
CO2ReO3 was similar to the one reported in ref 13: Re2O7 (1.94 g, 4 mmol)
was placed in a dry flask under argon. Dry THF (40 mL) was added followed
by addition of TFAA (0.75 mL, 5.3 mmol), and the mixture was stirred at
room temperature (rt) for 1 h. Solvents were removed under reduced pressure
at 0 °C, and the residue was washed twice with cold, dry pentane. Dry
CH2Cl2 (40 mL) was added followed by TFAA (0.75 mL, 5.3 mmol) at 0
°C. A solution of trienol 6 (644 mg, 1 mmol) in dry CH2Cl2 (5 mL) was
added, and the mixture was stirred at 0 °C to rt for 6 h and then was worked
up by slow addition of saturated aqueous NaHCO3 (10 mL) and H2O2 (2
mL) and extraction with ether. Purification by column chromatography
(neutral alumina above silica gel; ethyl acetate:hexane, 1:4) afforded pure
7 (335 mg, 48%) as the only isolatable product, [R]D ) -3.7 (c ) 5.46,
CHCl3).
(15) Triple-oxidative cyclization with this reagent has been recently
reported independent of our work: Towne, B. T.; McDonald, F. E. J. Am.
Chem. Soc. 1997, 119, 6022.
Acknowledgment. We thank the U.S.-Israel Binational Science
Foundation, the Israel Cancer Research Fund, PharMore Biotechnolo-
gies, and the Skaggs Institute for Chemical Biology for financial
support. We thank Prof. J. L. McLaughlin for copies of NMR spectra
of goniocin.
(16) Compound 7 (which was obtained from the reaction with the
rhenium reagent) was desilylated and the resultant diol was converted to
the corresponding bis(4-nitrobenzoate) derivative 7a. The latter was analyzed
1
by 2D H-1H COSY, TOCSY, and ROESY NMR (Bruker Avance DRX
600) to assign the hydrogens (for consistency, we use here the goniocin
numbering) attached to the oxycarbons H-8 (δ 4.45, m), H-10 (4.16, m),
H-13 (4.01, q), H-14 (3.81, q), H-17 (3.92, q), H-18 (3.85, q), H-21 (4.07,
q) and H-22 (5.19, q). The ROESY spectrum shows strong NOE correlation
between H-8 and H-10, between H-14 and H-17, and between H-18 and
H-21. It is evident from the latter two correlations that the stereochemistry
of rings B and C is cis. Two weak NOE correlations are observed between
H-10 and H-14 and between H-13 and H-22. The absence of an NOE
correlation between H-10 and H-13 confirms the trans stereochemistry of
ring A.
Supporting Information Available: 1H NMR spectra of com-
pounds 2-4, 6, 8, 9, 11 and 12, 13C NMR spectra of compounds 2-4,
8, and 9, NMR spectra of two samples of 7 obtained from oxidative
cyclization and from the independent synthesis, 1H-1H COSY, TOCSY,
and ROESY spectra of 7a, and information about the independent
synthesis of 7 and synthesis of aldehyde II (28 pages). See any current
masthead page for ordering and Internet access instructions.
(17) For the synthesis of compound 7 using the AD reaction, see
Supporting Information and Sinha, S. C.; Sinha, S. C.; Keinan, E. Submitted.
JA964273Z