Convergent, enantioselective syntheses of guanacastepenes A and E featuring a selective cyclobutane fragmentation

Article abstract of DOI:10.1021/ja060847g

The evolution of a convergent strategy that led to efficient, enantioselective syntheses of both natural (+)- and unnatural (-)-guanacastepene E and formal total syntheses of (+)- and (-)-guanacastepene A is described. A union of five- and six-membered ri

Full text of DOI:10.1021/ja060847g

Published on Web 05/10/2006
Convergent, Enantioselective Syntheses of Guanacastepenes
A and E Featuring a Selective Cyclobutane Fragmentation1
William D. Shipe† and Erik J. Sorensen*
Contribution from the Frick Chemical Laboratory, Princeton UniVersity,
Princeton, New Jersey 08544-1009
Received February 4, 2006; E-mail: ejs@princeton.edu
Abstract: The evolution of a convergent strategy that led to efficient, enantioselective syntheses of both
natural (+)- and unnatural (-)-guanacastepene E and formal total syntheses of (+)- and (-)-guanacastepene
A is described. A union of five- and six-membered ring intermediates by an efficient π-allyl Stille cross-
coupling reaction was followed by an intramolecular enone-olefin [2 + 2] photocycloaddition and a
stereoelectronically controlled, reductive fragmentation of the resulting cyclobutyl ketone. The latter two
transformations enabled controlled formation of the C-11 quaternary stereocenter and the central seven-
membered ring of the guanacastepenes. An enantiospecific synthesis of the functionalized five-membered
ring vinyl stannane from the monoterpene R-(-)-carvone featuring a carbon-carbon bond forming ring
contraction was also developed.
Introduction
C-C bond is oriented to permit very little overlap with the
π-system of the carbonyl group, whereas the external C-C bond
enjoys excellent overlap. The external bond is thus predisposed
toward fragmentation; it is the bond that fragments selectively
in the presence of reducing metals.
Our laboratory was drawn to this and related examples of
stereoelectronically controlled strained ring fragmentations5,6 and
to the possibility that conjugated cyclobutyl ketones might
undergo analogous reductive ring openings7,8 as we began to
contemplate a strategy for synthesizing the unique molecular
Strain-releasing fragmentations of small rings were among
the earliest reports of free radical rearrangements in organic
chemistry. In 1950, two laboratories proposed a free radical
chain mechanism and a selective fragmentation of a cyclobutyl
carbinyl radical to explain the reaction of â-pinene with
trichloromethyl radical (Scheme 1).2 In this transformation, the
strain intrinsic to a four-membered ring drives the conversion
of one tertiary radical intermediate into another.3,4
Strained rings flanked by carbonyl groups can also be
fragmented under reducing conditions.5 The tactic of generating
angular methyl groups by selective reductive cleavage of the
cyclopropane ring of conjugated cyclopropyl ketones has been
known for more than 40 years and is often utilized in organic
synthesis.6 Interesting reductive ring openings, such as the one
shown in Scheme 2, led to the generalization that the cyclo-
propane bond that is cleaved is the one possessing the maximum
overlap with the π-orbital system of the carbonyl group.6d,e The
rigid structure of carone is such that the internal cyclopropane
(6) For examples of selective, reductive openings of conjugated cyclopropyl
ketones, see: (a) Wehrli, H.; Heller, M. S.; Schaffner, K.; Jeger, O. HelV.
Chim. Acta 1961, 44, 2162-2173. (b) Heller, M. S.; Wehrli, H.; Schaffner,
K.; Jeger, O. HelV. Chim. Acta 1962, 45, 1261-1274. (c) Norin, T. Acta
Chem. Scand. 1963, 17, 738-748. (d) Norin, T. Acta Chem. Scand. 1965,
19, 1289-1292. (e) Dauben, W. G.; Deviny, E. J. J. Org. Chem. 1966, 31,
3794-3798. (f) Fraisse-Jullien, R.; Frejaville, C.; Toure, V. Bull. Chim.
Fr. 1966, 12, 3725-3727. (g) Bellamy, A. J.; Whitham, G. H. Tetrahedron
1968, 24, 247-254. (h) House, H. O.; Blankley, C. J. J. Org. Chem. 1968,
33, 47-53. (i) Hill, R. K.; Morgan, J. W. J. Org. Chem. 1968, 33, 927-
928. (j) Piers, E.; De Wall, W.; Britton, R. W. Can. J. Chem. 1969, 47,
4299-4306. (k) Piers, E.; Britton, R. W.; De Wall, W. Can. J. Chem. 1969,
47, 4307-4312. (l) Monti, S. A.; Bucheck, D. J.; Shepard, J. C. J. Org.
Chem. 1969, 34, 3080-3084. (m) Stork, G.; Uyeo, S.; Wakamatsu, T.;
Grieco, P.; Labovitz, J. J. Am. Chem. Soc. 1971, 93, 4945-4947. (n) Corey,
E. J.; Virgil, S. C. J. Am. Chem. Soc. 1990, 112, 6429-6431. (o) Batey,
R. A.; Motherwell, W. B. Tetrahedron Lett. 1991, 32, 6649-6652. (p)
Kirschberg, T.; Mattay, J. Tetrahedron Lett. 1994, 35, 7217-7220. (q)
Ziegler, F. E.; Zheng, Z. J. Org. Chem. 1990, 55, 1416-1418.
(7) For examples of metal-mediated cleavages of conjugated cyclobutyl ke-
tones, see: (a) Baker, W. R.; Senter, P. D.; Coates, R. M. J. Chem. Soc.,
Chem. Commun. 1980, 21, 1011-1012. (b) Comins, D. L.; Zhang, X.-M.;
Zheng, X. J. Chem. Soc., Chem. Commun. 1998, 22, 2509-2510. (c)
Comins, D. L.; Zheng, X.; Goehring, R. R. Org. Lett. 2002, 4, 1611-
1613.
(8) For examples of selective fragmentations of cyclobutyl carbinyl radicals
in synthesis, see: (a) Crimmins, M. T.; Mascarella, S. W. Tetrahedron
Lett. 1987, 28, 5063-5066. (b) Ziegler, F. E.; Zheng, Z. J. Org. Chem.
1990, 55, 1416-1418. (c) Cossy, J.; Aclinou, P.; Bellosta, V.; Furet, N.;
Baranne-Lafont, J.; Sparfel, D.; Souchaud, C. Tetrahedron Lett. 1991, 32,
1315-1316. (d) Crimmins, M. T.; Dudek, C. M.; Cheung, A. W.-H.
Tetrahedron Lett. 1992, 33, 181-184. (e) Crimmins, M. T.; Wang, Z.;
McKerlie, L. A. Tetrahedron Lett. 1996, 37, 8703-8706. (f) Ziegler, F.
E.; Kover, R. X.; Yee, N. N. K. Tetrahedron Lett. 2000, 41, 5155-5159.
(g) Kakiuchi, K.; Minato, K.; Tsutsumi, K.; Morimoto, T.; Kurosawa, H.
Tetrahedron Lett. 2003, 44, 1963-1966.
† Current address: Merck Research Laboratories, West Point, Penn-
sylvania.
(1) This manuscript is based on the Ph.D. thesis of William D. Shipe, The
Scripps Research Institute, 2004.
(2) (a) Oldroyd, D. M.; Fisher, G. S.; Goldblatt, L. A. J. Am. Chem. Soc. 1950,
72, 2407-2410. (b) Dupont, G.; Dulou, R.; Cle´ment, G. Bull. Soc. Chim.
Fr. 1950, 1056-1057. (c) Dupont, G.; Dulou, R.; Cle´ment, G. Bull. Soc.
Chim. Fr. 1950, 1115-1120.
(3) For selected reviews and discussions of free radical rearrangements, see:
(a) Beckwith, A. L. J. Chem. Soc. Special Pub. No. 24 1970, 239-269.
(b) Yet, L. Tetrahedron 1999, 55, 9349-9403. (c) Newcomb, M.
Tetrahedron 1993, 49, 1151-1176.
(4) For discussions of molecular strain and reactivity in organic chemistry,
see: (a) Wiberg, K. B. Angew. Chem., Int. Ed. Engl. 1986, 25, 312-322.
(b) Wiberg, K. B. Foundations of Chemistry; Kluwer Academic Publish-
ers: Netherlands, 2004; Vol. 6, pp 65-80.
(5) For discussions, see: (a) Staley, S. W. In SelectiVe Organic Transforma-
tions; Thyagarajan, B. S., Ed.; Wiley-Interscience: New York, 1972; Vol.
2, pp 309-348. (b) Wong, H. N. C.; Hon, M.-Y.; Tse, C.-W.; Yip, Y.-C.;
Tanko, J.; Hudlicky, T. Chem. ReV. 1989, 89, 165-198.
9
10.1021/ja060847g CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 7025-7035
7025

A R T I C L E S
Shipe and Sorensen
Figure 1. Selected members of the guanacastepene family of natural products.
Scheme 1. An Early Example of a Strain-Releasing Fragmentation of the Pinene System
Scheme 2. Lithium Metal Reduction of Carone to Carvomenthone (+ Its Methyl Epimer) by the Groups of Norin6d and Dauben;6e the Bond
that Cleaves Is the One Having the Maximum Overlap with the π-Orbital System of the Carbonyl Group
architecture of the guanacastepene diterpenes. The guanacaste-
penes are produced by a previously unknown endophytic fungus
that was discovered in the Guanacaste Conservation Area in
Costa Rica.9 Using NMR spectroscopic and X-ray crystal-
lographic methods, Clardy and co-workers revealed the diverse
and complex structures of the guanacastepenes. These com-
pounds represented a new diterpene structural type and were
shown to possess an interesting 5-7-6 tricyclic carbon skeleton
with an oxidized, polar face and a hydrophobic face. Figure 1
shows the structures of four members of this family of natural
products.
distinct designs for synthesizing the guanacastepenes have been
reported.15 Recently, a synthesis of racemic guanacastepene C
(2) was described by Mehta and co-workers,16 and an enantio-
selective route to guanacastepene A was published by the
Danishefsky group.17
Our laboratory viewed the guanacastepene architecture as a
challenging context for the development of a concise synthesis
of the 5-7 hydroazulane carbon framework based on a selective
cyclobutane fragmentation.15h In this article, we describe how
this approach to the seven-membered ring problem paved the
way for convergent, asymmetric syntheses of (+)-guanacaste-
The realization that the guanacastane diterpene architecture
was novel, combined with the discovery that two members of
the family, guanacastepenes A (1) and I (4), are potent inhibitors
of drug resistant strains of Staphylococcus aureus and Entero-
coccus faecalis,10,11 stimulated great interest in these natural
products as objectives for chemical synthesis. The reaction of
the chemical community to research opportunities provided by
the guanacastepenes was seemingly undiminished by the report
that guanacastepenes A (1) and I (4), the two members of the
class with promising antibacterial properties, also damage the
delicate membranes of human red blood cells.10 In addition to
the creative synthesis of guanacastepene A in racemic form by
Danishefsky and co-workers12 and the subsequent formal total
syntheses by the groups of Snider13 and Hanna,14 more than 10
(15) (a) Snider, B. B.; Hawryluk, N. A. Org. Lett. 2001, 3, 569-572. (b) Magnus,
P.; Waring, M. J.; Ollivier, C.; Lynch, V. Tetrahedron Lett. 2001, 42, 4947-
4950. (c) Dudley, G. B.; Danishefsky, S. J. Org. Lett. 2001, 3, 2399-
2402. (d) Dudley, G. B.; Tan, D. S.; Kim, G.; Tanski, J. M.; Danishefsky,
S. J. Tetrahedron Lett. 2001, 42, 6789-6791. (e) Snider, B. B.; Shi, B.
Tetrahedron Lett. 2001, 42, 9123-9126. (f) Mehta, G.; Umarye, J. D. Org.
Lett. 2002, 4, 1063-1066. (g) Gradl, S. N.; Kennedy-Smith, J. J.; Kim, J.;
Trauner, D. Synlett 2002, 3, 411-414. (h) Shipe, W. D.; Sorensen, E. J.
Org. Lett. 2002, 4, 2063-2066. (i) Nguyen, T. M.; Lee, D. Tetrahedron
Lett. 2002, 43, 4033-4036. (j) Dudley, G. B.; Danishefsky, S. J.; Sukenick,
G. Tetrahedron Lett. 2002, 43, 5605-5606. (k) Nakazaki, A.; Sharma, U.;
Tius, M. A. Org. Lett. 2002, 4, 3363-3366. (l) Mehta, G.; Umarye, J. D.;
Gagliardini, V. Tetrahedron Lett. 2002, 43, 6975-6978. (m) Boyer, F.-
D.; Hanna, I. Tetrahedron Lett. 2002, 43, 7469-7472. (n) Magnus, P.;
Ollivier, C. Tetrahedron Lett. 2002, 43, 9605-9609. (o) Nguyen, T. M.;
Seifert, R. J.; Mowrey, D. R.; Lee, D. Org. Lett. 2002, 4, 3959-3962. (p)
Mehta, G.; Umarye, J. D.; Srinivas, K. Tetrahedron Lett. 2003, 44, 4233-
4237. (q) Du, X.; Chu, H. V.; Kwon, O. Org. Lett. 2003, 5, 1923-1926.
(r) Brummond, K. M.; Gao, D. Org. Lett. 2003, 5, 3491-3494. (s) Hughes,
C. C.; Kennedy-Smith, J. J.; Trauner, D. Org. Lett. 2003, 5, 4113-4115.
(t) Sarpong, R.; Su, J. T.; Stoltz, B. M. J. Am. Chem. Soc. 2003, 125,
13624-13625. (u) Srikrishna, A.; Dethe, D. H. Org. Lett. 2004, 6, 165-
168. (v) Chiu, P.; Li, S. Org. Lett. 2004, 6, 613-616. (w) Mandal, M.;
Danishefsky, S. J. Tetrahedron Lett. 2004, 45, 3827-3829. (x) Mandal,
M.; Danishefsky, S. J. Tetrahedron Lett. 2004, 45, 3831-3833. (y) Du,
X.; Chu, H. V.; Kwon, O. Tetrahedron Lett. 2004, 45, 8843-8846. (z)
Hughes, C. C.; Miller, A. K.; Trauner, D. Org. Lett. 2005, 7, 3425-3428.
(aa) Li, C.-C.; Liang, S.; Zhang, X.-H.; Xie, Z.-X.; Chen, J.-H.; Wu, Y.-
D.; Yang, Z. Org. Lett. 2005, 7, 3709-3712. (bb) Yun, H.; Danishefsky,
S. J. Tetrahedron Lett. 2005, 46, 3879-3882.
(9) (a) Brady, S. F.; Singh, M. P.; Janso, J. E.; Clardy, J. J. Am. Chem. Soc.
2000, 122, 2116-2117. (b) Brady, S. F.; Bondi, S. M.; Clardy, J. J. Am.
Chem. Soc. 2001, 123, 9900-9901.
(10) Singh, M. P.; Janso, J. E.; Luckman, S. W.; Brady, S. F.; Clardy, J.;
Greenstein, M.; Maiese, W. M. J. Antibiot. 2000, 53, 256-261.
(11) Clardy, J.; Walsh, C. Nature 2004, 432, 829-837.
(12) (a) Tan, D. S.; Dudley, G. B.; Danishefsky, S. J. Angew. Chem., Int. Ed.
2002, 41, 2185-2188. (b) Lin, S.; Dudley, G. B.; Tan, D. S., Danishefsky,
S. J. Angew. Chem., Int. Ed. 2002, 41, 2188-2191.
(13) Shi, B.; Hawryluk, N. A.; Snider, B. B. J. Org. Chem. 2003, 68, 1030-
(16) Mehta, G.; Pallavi, K.; Umarye, J. D. Chem. Commun. 2005, 4456-4458.
(17) Mandal, M.; Yun, H.; Dudley, G. B.; Lin, S.; Tan, D. S.; Danishefsky, S.
J. J. Org. Chem. 2005, 70, 10619-10637.
1042.
(14) Boyer, F.-D.; Hanna, I.; Ricard, L. Org. Lett. 2004, 6, 1817-1820.
9
7026 J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006

Syntheses of Guanacastepenes A and E
A R T I C L E S
Scheme 3. An Approach Initiated by an Intermolecular Carbonyl Addition Reactiona
a X ) a carbonyl surrogate; Y ) metal; P ) unspecified protecting group. The metal counterions in intermediates 9 and 10 are not shown.
Scheme 4. Formation of Tetracycle 16 from Compound 12
pene E (3) and its enantiomer as well as the direct precursor of
naturally occurring (+)-guanacastepene A (1).
be a stereoelectronically controlled cyclobutane ring cleavage
in the context of a ketyl radical anion such as 9. If successful,
this selective ring fragmentation would leave in its wake a
reactive species 10 having the desired 5-7-6 tricyclic skeleton
and a â-alkoxy enolate system, which would be expected to
collapse to a guanacastepene-like enone. We hoped that this
â-elimination reaction and a reduction of the putative secondary
carbon radical intermediate would afford the advanced guana-
castepene structural type 11. From this vantage point, the
prospects for achieving syntheses of various members of the
guanacastepene family seemed excellent.
In the course of investigating this strategy, we synthesized
the complex pentacyclic ketone 12 (Scheme 4) by a reaction
sequence involving an intermolecular carbonyl addition reaction
and an intramolecular enone olefin [2 + 2] photocycloaddition.15h
While this structure lacked the isopropyl group that is a
conspicuous feature of many of the guanacastepenes, it seemed
to provide an ideal context for probing the feasibility of some
of the key ideas outlined in Scheme 3. When a solution of ketone
12 in THF at room temperature was treated with 3 equiv of the
powerful reducing agent samarium diiodide, a new, structurally
Results and Discussion
1. An Approach for Synthesis Based on a Selective
Cyclobutane Ring Cleavage. To achieve a convergent synthesis
of the guanacastane molecular architecture, we would attempt
a carbonyl addition of a cyclopentenyl organometallic reagent
such as 5, wherein Y is a metal and X is a suitable surrogate
for a ketone carbonyl group, to the aldehyde function of 6. This
union would establish a needed carbon-carbon bond and create
a favorable setting for a key intramolecular enone-olefin [2 +
2] photocycloaddition.18 A diastereoface-selective [2 + 2]
cycloaddition of the ethenyl side chain of 7 to the underside of
the cyclopentenone ring would generate a rigid, cyclobutyl
ketone of type 8 with the desired configuration at the methyl-
bearing C-11 quaternary stereocenter (Scheme 3).
An analysis of a molecular model of tetracycle 8 revealed
that the cyclobutane bond that is exocyclic to the five-membered
ring is nearly parallel to the π-orbital system of the adjacent
carbonyl group. Owing to this circumstance, we reasoned that
this bond would fragment selectively in the course of a one-
electron reduction of the keto group in 8. The pivotal transfor-
mation in our approach to the guanacastane architecture would
(18) For a comprehensive review of enone olefin [2 + 2] photochemical
cycloadditions, see: Crimmins, M. T.; Reinhold, T. L. Org. React. 1993,
44, 297-588.
9
J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006 7027

A R T I C L E S
Shipe and Sorensen
Scheme 5. An Approach Initiated by a Stille Cross-Coupling Reactiona
a P ) Unspecified protecting group.
complex cyclobutane-containing compound having the structure
16 was formed.19 This outcome was encouraging because it
suggested that the strained cyclobutane ring of compound 12
did, in fact, fragment in the desired and expected fashion.
Moreover, the dihydrofuran ring was opened, almost certainly
via the anticipated â-elimination. However, the supposition that
the putative carbon radical arising from the reductive cleavage
of 12 would simply convert itself to a methylene group and not
bring about undesired chemistry proved to be the weakness of
our initial plan. We assume that the guanacastepene-like enone
system in intermediate 15 exists only transiently in this process
and undergoes a transannular addition of either a neutral
secondary carbon radical or a samarium(III) carbanion as
implied in Scheme 4. Fortunately, this outcome suggested a
means by which to alter the synthesis in order to obtain the
desired 5-7-6 tricyclic ring system. The oxidation state at
C-2 simply had to be reduced to prevent â-elimination, since
that event led to the irreversible and undesired transannular
carbon-carbon bond formation. The manner in which this was
accomplished is described below.
2. A Strategy Featuring an Efficient π-Allyl Palladium
Stille Cross-Coupling Reaction. Having demonstrated the
susceptibility of the “exocyclic” cyclobutane bond toward a
reductive fragmentation, we felt comfortable with the foundation
of our approach to the synthesis of the guanacastepenes. We
next directed our attention to the problem of synthesizing a
fragmentation substrate, such as 20, wherein a â-elimination at
C-2 during the reductive fragmentation of the cyclobutane ring
would not be a possibility (Scheme 5).
crucial fragment coupling held at least one important advantage
over the carbonyl addition strategy: there would be no
complications arising from the production of diastereoisomeric
secondary alcohols at C-2. A synthesis of a compound of type
19 would be followed by an intramolecular [2 + 2] photocy-
cloaddition reaction to fashion conjugated cyclobutyl ketone 20.
Once again, we hoped to effect a reductive fragmentation of
only one of the cyclobutyl bonds in 20 to yield the seven-
membered B ring characteristic of the guanacastepenes. One
of the nice features of this fragmentation strategy is that the
putative samarium(III) enolate formed in the process would be
regiodefined. It would thus be possible, at least in principle, to
make use of that regiodefined enolate ion in a controlled
synthesis of the needed double bond between C-1 and C-2. Our
new goal was to carry out a tandem reductive fragmentation/
enolate trapping sequence, followed by a heteroatom elimination
reaction to produce the guanacastepene architecture 22.
In a previous publication,15h we described the successful
application of this general design to a racemic synthesis of the
5-7-6 ring framework of the guanacastepenes lacking a C-12
isopropyl group. It was now required that we construct and use
an A ring domain containing the required isopropyl group and
that the overall synthesis be rendered enantioselective. Since
our synthesis would employ a convergent Stille cross-coupling
step, we needed to obtain the A ring and C ring building blocks
in enantiomerically pure form.
3. Asymmetric Synthesis of an A Ring Fragment. After
considering several ideas for mastering the relative and absolute
stereochemistry of the rather densely substituted five-membered
ring domain of guanacastepene A (1) (colored in red, Scheme
6), we decided to turn to a useful building block from the chiral
To reach compound 20, we envisioned a reaction sequence
wherein two building blocks, a vinylstannane 17 and an allylic
acetate 18 representing the A and C rings of the guanacaste-
penes, respectively, would be joined through a π-allyl Stille
cross-coupling reaction.15h,20,21 This method for achieving the
(20) For an excellent review of the Stille reaction, see: Farina, V.; Krishna-
murthy, V.; Scott, W. J. Org. React. 1997, 50, 1-652.
(21) For an example of a π-allyl Stille coupling reaction to achieve a merger of
complex fragments from our laboratory, see: Vanderwal, C. D.; Vosburg,
D. A.; Weiler, S.; Sorensen, E. J. J. Am. Chem. Soc. 2004, 125, 5393-
5407.
(19) For excellent discussions of the chemistry of samarium diiodide, see:
Molander, G. A. Org. React. 1994, 46, 211-367.
9
7028 J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006

Syntheses of Guanacastepenes A and E
A R T I C L E S
Scheme 6. Strategy for an Asymmetric Synthesis of an A-Ring Fragment
Scheme 7. Synthesis of Keto Enol 25a
a (a) 0.2 mol % PtO2, H2, rt, 23 h, 100%. (b) LDA, THF, -78 °C f 0 °C; then MeI, 0 °C f rt, 17 h, 96% as a 5.5:1 mixture of diastereoisomers. (c)
O3, EtOAc, -78 °C, 90 min; then H2, Pd/C, rt, 17 h, 48-54%. (d) NaCN, p-TsOH, THF‚H2O, rt, 14 h, 99%. (e) EDCI, 0 °C f rt, CH2Cl2, 7 h, 79%. (f)
3.0 equiv of LiHMDS, THF, rt, 2 h; then 1 N aq. HCl, 50-58%. LDA ) lithium diisopropylamide; p-TsOH ) para-toluenesulfonic acid; EDCI ) 1-[3-
(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride; rt ) room temperature; LiHMDS ) LiN(SiMe3)2.
Scheme 8. Proposed Mechanism of the Base-Induced Ring Contraction of Compound 24a
a LiHMDS ) LiN(SiMe3)2; rt ) room temperature.
pool. The choice of (S)-(+)-carvone (23) as a starting material
arose in part from the fact that it already contains an isopropenyl
substituent with the same stereochemical configuration as the
isopropyl group of the guanacastepenes; a simple hydrogenation
of carvone’s isopropenyl group would afford the isopropyl
function. Of course, a commitment to (S)-(+)-carvone meant
that we would, at some point, be required to effect a one-atom
ring contraction to access the five-membered guanacastepene
A ring. This objective led us to consider the suitability of
R-acyloxy nitrile 24 (a cyanohydrin lactone) as an advanced
intermediate en route to A-ring surrogate 25. Even if compound
24 was constructed as a mixture of diastereoisomers, we were
9
J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006 7029

A R T I C L E S
Shipe and Sorensen
Scheme 9. Synthesis of Vinylstannane 17a
a (a) Et3N, NfF, CH2Cl2, rt, 14 h, 94%. (b) Pd(dppf)Cl2, Me3SnSnMe3, NMP, 60 °C, 8 h, 63% of 17, 11% of 34, 16% of 25. NMP ) 1-methyl-2-
pyrrolidinone; rt ) room temperature; dppf ) 1,1′-bis(diphenylphosphino)ferrocene.
Table 1. Selected Reaction Conditions Screened for the
drawn to the possibility that this substance might undergo
conversion to a single 1,2-diketone (or its enol tautomer 25) on
treatment with a strong, non-nucleophilic base. It was always
our intention to attempt an introduction of the A-ring acetoxy
group at a late stage in the synthesis.
Wulff-Stille Synthesis of Vinylstannane 17a
From (S)-(+)-carvone (23), the synthesis began with the
selective hydrogenation of the less hindered olefin (Scheme 7).
This known transformation could be carried out with either
platinum(IV) oxide (Adams’ catalyst)22 or tris(triphenylphos-
phine)rhodium chloride (Wilkinson’s catalyst);23 however, we
found that both the platinum-catalyzed hydrogenation was more
reliable and the isolation of the product in that case was much
simpler. The resulting dihydrocarvone did not require purifica-
tion and was directly employed in the base-induced R-methy-
lation;24 this sequence afforded compound 26 in excellent yield
after distillation as a 5.5:1 mixture of diastereoisomers. This
mixture was carried forward, as the configuration of the newly
formed stereocenter would be of no consequence. Reductive
ozonolysis22c of enone 26 resulted in the excision of two carbon
atoms and yielded the linear aldehyde carboxylic acid 27, which
was easily converted to the cyanohydrin 28 by treatment with
sodium cyanide and p-toluenesulfonic acid. Lactonization of the
mixture of cyanohydrin carboxylic acids with EDCI yielded the
R-acyloxy nitriles 24 as a mixture of four diastereoisomers.
When this mixture was treated with lithium bis(trimethylsilyl)-
amide (LiHMDS), keto enol 25 emerged as a single compound
from the reaction mixture in 50-58% yield.25-27
temp
C)
time
(h)
entry
−
E
catalyst
solvent
(
°
% yield
1
2
3
-OTf
-ONf
-ONf
Pd(dppf)Cl2‚CH2Cl2
Pd(dppf)Cl2‚CH2Cl2
Pd(dppf)Cl2‚CH2Cl2
THF
THF
NMP
60
60
60
20
20
8
34
44
63b
a NMP ) 1-methyl-2-pyrrolidinone; dppf ) 1,1′-bis(diphenylphos-
phino)ferrocene; dba ) trans,trans-dibenzylideneacetone; rt ) room
temperature. b 474 mg of 17 produced.
Table 2. Wulff-Stille Coupling of Hexamethylditin and Nonaflate
35
temp
entry
catalyst
additives
solvent
(
°
C)
time (h)
% yield
1
2
3
Pd2dba3
Pd2dba3
Pd2dba3
dppf
dppf
THF
THF
NMP
65
70
rt
36
8
24
90
83
85
One possible mechanism by which this interesting and
potentially general cyclic 1,2-diketone synthesis could occur is
shown in Scheme 8. Deprotonation of 24 with LiHMDS could
generate the nitrile-stabilized anion 29, which might undergo
intramolecular addition to the lactone carbonyl. The interme-
diacy of epoxy alkoxide ion 30 would be brief, for it would be
expected to rapidly open to the isomeric alkoxide ion 31.
Collapse of the latter species with expulsion of cyanide ion
would then give rise to the five-membered ring 1,2-diketone
32. 1,2-Dicarbonyl compounds frequently exist as their enol
tautomers. However, in the case of compound 32, the tautomer-
ization could take place in two directions. In early experiments,
a single equivalent of base was used to effect the contraction
of compound 24. In these instances, comparable yields of the
ring-contracted carbocyclic products were attained, but the less
substituted tautomer 33 was found to be a minor product in
these reactions. Increasing the amount of base to 3.0 equiv
altogether prevented the formation of this undesired tautomer.
The conversion of keto enol 25 to the desired A-ring
vinylstannane 17 was expected to be straightforward (Scheme
9). In fact, the formation of vinyl nonafluorobutanesulfonate
(22) (a) Kitahara, T.; Kurata, H.; Matsuoka, T.; Mori, K. Tetrahedron 1985,
41, 5475-5485. (b) Paquette, L. A.; Dahnke, K.; Doyon, J.; He, W.; Wyant,
K.; Friedrich, D. J. Org. Chem. 1991, 56, 6199-6205. (c) Deslongchamps,
P.; et al. Can. J. Chem. 1990, 68, 127-152.
(23) Ireland, R. E.; Bey, P. Organic Syntheses; Wiley: New York, 1988; Collect.
Vol. VI, p 459.
(24) Hua, D. H.; Takasu, K.; Huang, X.; Millward, G. S.; Chen, Y.; Fan, J.
Tetrahedron 2000, 56, 7389-7398.
(25) For studies on the use of protected cyanohydrins as acyl carbanion
equivalents, see: (a) Zervos, M.; Wartski, L.; Seyden-Penne, J. Tetrahedron
1986, 42, 4963-4973. (b) Albright, J. D. Tetrahedron 1983, 39, 3207-
3233. (c) Ferrino, S. A.; Maldonado, L. A. Synth. Commun. 1980, 10, 717-
723. (d) Stork, G.; Maldonado, L. J. Am. Chem. Soc. 1974, 96, 5272-
5274. (e) Stork, G.; Maldonado, L. J. Am. Chem. Soc. 1971, 93, 5286-
5287. (f) Jacobson, R. M.; Lahm, G. P.; Clader, J. W. J. Org. Chem. 1980,
45, 395-405.
(26) For an investigation of the trapping of R-acyloxy nitrile carbanions with
aldehydes and Michael acceptors, see: Kraus, G. A.; Dneprovskaia, E.
Tetrahedron Lett. 2000, 41, 21-24.
(27) Recently, a method has been described for the synthesis of R-keto esters
involving the rearrangement of aromatic cyanohydrin carbonate esters using
LDA: Thasana, N.; Prachyawarakorn, V.; Tontoolarug, S.; Ruchirawat,
S. Tetrahedron 2003, 44, 1019-1021.
9
7030 J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006

Syntheses of Guanacastepenes A and E
A R T I C L E S
Scheme 10. Classical Resolution of the C-Ring Alcohol 38a
a (a) 37, DMAP, DCC, CH2Cl2, 0 f rt, 90 min, 98%. DMAP ) 4-dimethylaminopyridine; DCC ) dicyclohexylcarbodiimide; PMP ) p-methoxyphenyl.
Scheme 11. Synthesis of Enantiopure Allylic Acetate 41a
a (a) K2CO3, MeOH, rt, 30 min, 97%. (b) Ac2O, DMAP, pyridine, rt, 15 min, 100%. DMAP ) 4-dimethylaminopyridine; rt ) room temperature; PMP
) p-methoxyphenyl.
Table 3. Some Conditions Screened for the π-Allyl Stille
Cross-Coupling of 17 and 41
(nonaflate) 34 from 25 was quite smooth;28 however, efforts to
transform the nonaflate to vinylstannane 17 repeatedly failed.29
It was not until many sets of reaction conditions were tested,30
three of which are shown in Table 1, that an acceptable yield
of stannane 17 was attained. The critical modification was the
use of [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium-
(II) as the catalyst. Attempts to form this catalyst in situ from
tris(dibenzylideneacetone)dipalladium(0) and 1,1′-bis(diphen-
ylphosphino)ferrocene ligand never led to success.29 Some 34
and 25 were always recovered from the reaction mixture along
with the product stannane and were recycled.
It is not entirely clear why we faced so much resistance in
the conversion of vinyl sulfonates (nonaflates or triflates) to
vinylstannane 17. It seemed that the mere addition of an
isopropyl group on the far side of the five-membered ring had
temp
C)
time
(h)
shut down the reaction under nearly all conditions tested.29 To
properly test that notion, we synthesized compound 35 (Table
2), a nonaflate lacking an isopropyl group, from the correspond-
ing commercially available 1,2-diketone. The Wulff-Stille
reaction consistently succeeded under conditions identical to
those which had failed in the isopropyl-containing system,
without the need to resort to the bis(diphenylphosphino)-
ferrocene dichloropalladium catalyst.31 This observation indi-
cates that it is the isopropyl group that necessitates the use of
a more active catalyst.
entry
catalyst
additives
solvent
(
°
% yield
1
2
3
Pd2dba3
Pd(PPh3)4
Pd(PPh3)4
AsPh3
CuCl, LiCl
CuCl, LiCl
NMP
DMSO
DMSO
150
rt, 60
rt, 60
18
1, 20
1, 15
0
91a
87b
a 86 mg of 42 produced. b 221 mg of 42 produced.
4. Classical Resolution of the C-Ring Fragment. The
relative stereochemistry of the cyclohexenyl domain of the
guanacastepenes was established at an early stage of the project
via an efficient intermolecular Diels-Alder reaction between
dimethyl acetylenedicarboxylate and a cross-conjugated silyl
enol ether.15h Seeing no simple way to render this process
asymmetric, we resorted to a classical resolution as a means to
obtain an enantiopure C-ring precursor that could be elaborated
to an allylic acetate of the type 18 (Scheme 5). To this end, we
chose an approach that involved the formation of an ester with
one of our alcohol intermediates and the O-acetate of inexpen-
sive (S)-(+)-mandelic acid (compound 37, Scheme 10).32
Racemic allylic alcohol 38,33 an advanced intermediate in our
synthesis, was reacted with O-acetyl (S)-(+)-mandelic acid 37
in the presence of a dehydrating agent; this union afforded a
(28) Commercially available nonafluorobutanesulfonyl fluoride is a stable and
easily handled liquid. Obtained from the electrochemical fluorination of
2,5-dihydrothiophene 1,1-dioxide, this compound is cheaper than triflic
anhydride, see: Lyapkalo, I. M.; Webel, M.; Reissig, H.-U. Eur. J. Org.
Chem. 2002, 1015-1025.
(29) See Table 1 in the Supporting Information for a summary of our
observations on the synthesis of compound 17.
(30) (a) Wulff, W. D.; Peterson, G. A.; Bauta, W. E.; Chan, K.-S.; Faron, K.
L.; Gilbertson, S. R.; Kaesler, R. W.; Yang, D. C.; Murray, C. K. J. Org.
Chem. 1986, 51, 277-279. (b) Scott, W. J.; Stille, J. K. J. Am. Chem. Soc.
1986, 108, 3033-3040.
(31) We knew from previous chemistry that the Wulff-Stille reaction to convert
a vinyl iodide to a vinylstannane in the absence of an isopropyl group was
a smooth process (see ref 15h). However, this new two-step synthesis of
vinylstannane 36 was actually an improvement on the R-iodination/Wulff-
Stille sequence developed earlier in terms of both yield and ease of
execution.
9
J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006 7031

A R T I C L E S
Shipe and Sorensen
Scheme 12. A Palladium-Catalyzed Coupling of Vinyl Stannane 17 with Allylic Mandelate Ester 39a
a (a) LiCl, CuCl, Pd(PPh3)4, DMSO, rt f 60 °C, 78%. PMP ) p-methoxyphenyl.
Scheme 13. Synthesis of Tricyclic Dienone 45a
a (a) hν , i-Pr2NEt (0.5 equiv), Et2O, 3 h, 82%. (b) SmI2 (2.5 equiv), HMPA (10 equiv), THF, rt, 15 min; then PhSeBr, 50%. (c) mCPBA, CH2Cl2, -78
°C, 10 min, 86%. HMPA ) hexamethylphosphoramide; mCPBA ) 3-chloroperoxybenzoic acid; PMP ) p-methoxyphenyl.
1:1 mixture of diastereoisomers 39 and 40 that could be sep-
arated by flash column chromatography on silica gel.34
In our synthesis of a guanacastepene-like tricycle, racemic
allylic acetate 41 performed very well in π-allyl Stille cross-
coupling reactions with five-membered ring vinyl stannanes.15h
Owing to this experience, we opted to transform mandelate ester
39 to allylic alcohol 38, after which a simple acetylation of the
hydroxyl group afforded enantiopure allylic acetate 41 in 97%
overall yield (Scheme 11).
5. An Enantioselective Formal Synthesis of Guanacaste-
pene A. With routes to optically active ester 41 and the chiral
pool derivative 17 in place, we were in a position to test the
feasibility of the general strategy outlined previously in Scheme
5. The first hurdle was the projected π-allyl Stille coupling,15h,20,21
and having already faced a steep challenge in the formation of
vinylstannane 17 by a Wulff-Stille coupling, we cautiously
proceeded using the reaction conditions that had worked well
in our prior model study. To our great dismay, the desired
fragment coupling failed. All aspects of the new reaction were
identical to the model, except for the presence of an isopropyl
group on the vinyl stannane component. We once again set about
surveying conditions described in the literature for hindered
Stille cross-couplings; three of the reactions we carried out and
their outcomes are shown in Table 3. With this problem, there
was no such thing as a partial success. Only when we imple-
mented the outstanding procedure developed by the Corey
laboratory for achieving cuprous-chloride-accelerated Stille cou-
plings (entries 2 and 3) could this critical merger be achieved.35,36
As advertised, this procedure can be uniquely effective at joining
sterically crowded vinyl stannanes with organic electrophiles.
We also discovered that allylic mandelate ester 39, which
was synthesized for the sole purpose of achieving a resolution
of stereoisomers, is a good substrate for the π-allyl Stille
coupling (Scheme 12). This approach obviated the transforma-
tions shown in Scheme 11, and the overall yield of the three-
step process (∼84%) only slightly exceeded the 78% yield
achieved in this single reaction.
(32) For classical resolution of mandelic esters, see: Whitesell, J. K.; Reynolds,
D. J. Org. Chem. 1983, 48, 3548-3551. For classical resolution of O-acetyl
mandelic esters, see: Breitholle, E. G.; Stammer, C. H. J. Org. Chem. 1974,
39, 1311-1312.
(33) A Diels-Alder cycloaddition between dimethylacetylene dicarboxylate and
the cross-conjugated trimethylsilyl enol ether derived from 3-methyl
cyclohexenone and a regiospecific Baeyer-Villiger oxidation were key
steps in our previously published synthesis of racemic allylic alcohol 38
(see Supporting Information and ref 15h). The configuration of the
p-methoxyphenyl-bearing stereogenic center in optically active 38 (see
Scheme 11) was determined by an NOE experiment.
Having now been thwarted twice by the presence of the
isopropyl group during key reactions, it was with some anxiety
that we embarked on the construction of the C-11 quaternary
carbon center and seven-membered B ring. Fortunately, these
fears proved groundless, as irradiation of compound 42 with a
450 W Hanovia mercury vapor lamp effected a [2 + 2]
photocycloaddition that was more facile and efficient than in
(34) Over 20 solvent systems were screened before one was found that
satisfactorily separated the mandelic esters on TLC (49% CH2Cl2/49%
toluene/2% Et2O). Without recourse to HP silica gel, the diastereoisomers
could be separated with an efficiency of 40-60% (60-40% remained a
mixture and was recycled).
(35) Han, X.; Stoltz, B. M.; Corey, E. J. J. Am. Chem. Soc. 1999, 121, 7600-
7605.
(36) See Table 3 in the Supporting Information for a summary of our
observations on the Stille coupling of compounds 17 and 41.
9
7032 J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006

Syntheses of Guanacastepenes A and E
A R T I C L E S
Scheme 14. Reductive Fragmentations Mediated by Lithium Metala
a (a) PPTS, cat. MeOH, 2,2-dimethoxypropane, 60 °C, 17 h, 75%. (b) Li°, NH3(l), t-BuOH, THF, -78 °C; then trapping agent. Trapping agent, -R,
yield: a -MeOH, -H, 63%. b -TMSCl, -H, 64%. c -PhSeBr, -SePh, 46%. PMP ) p-methoxyphenyl.
Scheme 15. Synthesis of Crystalline Diester 48a
a (a) H2NOMe‚HCl, pyridine, MeOH, rt, 3 h, 77%. (b) 4-Bromobenzoyl chloride, DMAP, pyridine, rt, 2 h, 65%. DMAP ) 4-dimethylamino pyridine;
PMP ) p-methoxyphenyl.
Scheme 16. Synthesis of Known Compound 49 and Completion
of a Formal Synthesis of (+)-Guanacastepene Aa
Scheme 17. Rubottom-Type R-Acetoxylationa
a (a) PPTS (0.2 equiv), 2,2-dimethoxypropane, 60 °C, 20 h, 77%. PPTS
) pyridinium p-toluenesulfonate; PMP ) p-methoxyphenyl.
any model system (Scheme 13). Previously, we had observed
outstanding diastereofacial selectivity in model systems with
this intramolecular reaction.37 The presence of the isopropyl
group on the cyclopentenone ring reinforced that selectivity.
Treatment of ketone 43 with samarium diiodide19 accomplished
a selective, reductive fragmentation of the cyclobutane ring, after
which trapping of the putative samarium enolate with phenyl-
selenenyl bromide gave rise to organoselenide 44 as a mixture
of diastereoisomers. Elimination of the selenoxide38 formed by
oxidation of 44 with mCPBA generated the complete tricyclic
[5-7-6] ring system of the guanacastepenes.
In an attempt to improve the yield of the samarium-diiodide-
mediated reductive fragmentation/enolate trapping sequence, we
explored reductions with lithium metal in liquid ammonia
(Scheme 14). The aromatic benzylidene acetal protecting group
was clearly incompatible with these conditions, so we replaced
it with an isopropylidene ketal in a single reaction (43 f 46).
In the fragmentation reaction, attempted enolate trapping with
chlorotrimethylsilane39 gave the same product as the methanol
quench control reaction after workup. Trapping with phenylse-
a (a) Et3N, Et3SiOTf, CH2Cl2, -78 °C, 15 min. (b) mCPBA, CH2Cl2,
-78 °C, 5 min. (c) Ac2O, DMAP, pyridine, rt, 30 min, 45%, 3 steps. DMAP
) 4-dimethylamino pyridine; PMP ) p-methoxyphenyl.
lenenyl bromide afforded 47c in 46% yield, which was not an
improvement over the samarium diiodide approach.
At this point in our work, we wished to confirm that the
relative and absolute stereochemistry were as we believed them
to be. To that end, we sought a crystalline intermediate for
an X-ray crystal structure determination. Previously, we had
managed to obtain a crystal from a rigid, pentacyclic methoxime
derivative.15h We therefore chose to derivatize cyclobutyl ke-
tone 43 as a methoxime (Scheme 15). Under the conditions of
this reaction, the benzylidene acetal was also cleaved, and it
was possible to advance the resulting diol to the crystalline
diester 48 by straightforward acylations. An X-ray crystal-
lographic analysis of this compound confirmed its complex
structure.40
(38) (a) Sharpless, K. B.; Lauer, R. F. J. Am. Chem. Soc. 1973, 95, 2697-
2699. (b) Reich, H. J.; Reich, I. L.; Renga, J. M. J. Am. Chem. Soc. 1973,
95, 5813-5815. (c) Nicolaou, K. C.; Petasis, N. A. Selenium in Natural
Product Synthesis; CIS, Inc.: Philadelphia, PA, 1984.
(37) Inspection of molecular models showed a significant steric interaction
between the C16 methyl group and the vinyl C17 methyl group in the
transition state that would lead to addition on the undesired face.
9
J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006 7033

A R T I C L E S
Shipe and Sorensen
Scheme 18. Synthesis of (+)-Guanacastepene E (3)a
a (a) PPTS (0.25 equiv), MeOH, 70 °C, 30 min, 88%. (b) SiO2, CH2Cl2, rt, 36 h, 78%. PPTS ) pyridinium p-toluenesulfonate.
Scheme 19. Synthesis of Unnatural (-)-Guanacastepene E
Proof that the reductive ring fragmentation had actually
accomplished cleavage of the desired cyclobutyl bond was also
sought. This was done in a single experiment by synthesizing
an advanced intermediate in the Danishefsky12,17 and Snider13
routes to (()-guanacastepene A. Warming of a solution of
compound 45 in 2,2-dimethoxypropane containing 0.2 equiv
of pyridinium p-toluenesulfonate transformed the benzylidene
acetal ring to the isopropylidene ketal of 49 (Scheme 16). The
1H NMR spectrum of this substance exactly matched the one
published by the Snider group.13 With this result, we too could
claim a formal total synthesis of guanacastepene A.
instability of R-hydroxy ketone 51, we immediately produced
acetate ester 52 by acetylation of the hydroxyl group. Our H
1
NMR spectrum of 52 showed a vicinal coupling constant of
6.0 Hz between the methine protons on the five-membered ring
(C-12 and C-13), indicating a cis relationship between them
and, thus, a highly stereoselective R-hydroxylation. This is in
accord with observations on coupling constants made by
Danishefsky’s group on the acetoxylation of their acetonide-
containing system.43
From compound 52, the paths to guanacastepenes A and E
were direct (Scheme 18). Acid-catalyzed methanolysis of the
benzylidene acetal in 52 afforded diol 53 in 88% yield. The
latter substance was previously converted to guanacastepene A
(1) by a selective oxidation of the primary alcohol,12,17 although
we found this transformation to be challenging owing to the
proclivity of 53 toward an intramolecular conjugate addition
reaction.15h Nevertheless, this conjugate addition reaction was
welcome, for it gave rise to (+)-guanacastepene E (3) as a single
diastereoisomer. We were pleased with this result, since no
synthesis of the tetracyclic natural product 3 has been reported.
The formation of compound 3 in this manner was best achieved
by stirring a solution of diol 53 and silica gel in diethyl ether
6. A Synthesis of (+)-Guanacastepene E. After several
original strategies for installing the C-13 acetoxy group found
in guanacastepenes A and C were observed to be unsatisfactory,
we resorted to a three-step R-acetoxylation approach analogous
to the one employed by the Danishefsky group on their
acetonide-containing intermediate 49.41 Formation of the tri-
ethylsilyl enol ether 50 from 45 occurred smoothly, and the
crude material was subsequently treated with mCPBA to effect
a net R-hydroxylation (Scheme 17).42 Wary of the potential
(39) (a) Stork, G.; Uyeo, S.; Wakamatsu, T.; Grieco, P.; Labovitz, J. J. Am.
Chem. Soc. 1971, 93, 4945-4947. (b) Batey, R. A.; Motherwell, W. B.
Tetrahedron Lett. 1991, 32, 6649-6652.
(42) We noted that the two-step yield was somewhat low; thus dimethyldioxirane
was employed as an oxidant. Danishefsky’s group observed a high-yielding
conversion to the R-hydroxylated compound using this reagent. In our case,
however, with a benzylidene acetal as a protective group, much decomposi-
tion was quickly seen by TLC and the yield of 51 was very low.
(43) The stereoselectivity of the oxidation of 50 may be sensitive to temperature;
in one reaction run at 0 °C (rather than -78 °C), the cis/trans ratio was
nearly 1:1.
(40) We are grateful to Dr. Doug Ho, director of the X-crystallographic facility
in the Chemistry Department at Princeton University, for performing the
X-ray crystallographic analysis of compound 48.
(41) We sincerely thank Professor Samuel Danishefsky and members of his
group for providing us with unpublished procedures and data for the late
stages of their synthesis.
9
7034 J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006

Syntheses of Guanacastepenes A and E
A R T I C L E S
1
for 36 h at room temperature.44 The H NMR spectrum of our
synthetic (+)-guanacastepene E (3) was identical to that supplied
by the Clardy group.9b
a novel ring contraction that affords cyclic 1,2-diketones. The
synthesis of the C-ring cyclohexene domain featured a Diels-
Alder/Baeyer-Villiger/methanolysis sequence15h that provided
efficient access to the desired relative stereochemistry. The
convergent strategy described herein gives rise to guanacaste-
pene E in 28 steps with a longest linear sequence of 20 steps.
7. Synthesis of Unnatural (-)-Guanacastepene E. Natu-
rally, half of the material that had been elaborated to the stage
of the classical resolution had the incorrect absolute stereo-
chemistry (compound 40, Scheme 19) for a synthesis of
natural (+)-guanacastepene E. However, we were able to
make productive use of this material in a synthesis of unnatural
(-)-guanacastepene E. When coupled with vinylstannane 54
derived in eight steps from (R)-(-)-carvone, we had, in a
sense, a completely reliable model system since it differed
from the real system only in absolute configuration. In fact,
much of the chemistry described above was developed using
this material.
Acknowledgment. We dedicate this paper to our friend
Professor Albert Eschenmoser on the occasion of his 80th
birthday. For invaluable technical assistance during the course
of this project we thank Dr. Dee-Hua Huang and Dr. Laura
Pasternack (NMR, TSRI), Dr. Istvan Pelczer and Dr. Carlos
Pacheco (NMR, Princeton), Dr. Gary Siuzdak (Mass Spectrom-
etry, TSRI), Dr. Dorothy Little (Mass Spectrometry, Princeton),
Dr. Raj Chadha (X-ray, TSRI), and Dr. Douglas Ho (X-ray,
Princeton). This work was supported by The Skaggs Institute
for Chemical Biology at TSRI, Princeton University, National
Institute for General Medical Sciences (GM065483), Merck
Research Laboratories, a Lilly Grantee Award, a Beckman
Young Investigator Award, a Bristol-Myers Squibb unrestricted
research grant in organic synthesis, and a predoctoral fellowship
from the Skaggs Institute (W.D.S.).
Conclusions
Three reactions, an efficient π-allyl Stille cross-coupling, an
intramolecular [2 + 2] photocycloaddition, and a stereoelec-
tronically controlled reductive fragmentation of a conjugated
cyclobutyl ketone, enabled a convergent synthesis of the
complex, tricyclic guanacastane molecular architecture and
asymmetric total syntheses of both (+)- and (-)-guanacastepene
E. Formal asymmetric syntheses of (+)- and (-)-guanacastepene
A were also achieved. Our synthesis of the A-ring cyclopen-
tenone building block from carvone led to the development of
Supporting Information Available: Experimental procedures
and characterization data for all compounds and a complete
listing of authors for ref 22c. This material is available free of
charge via the Internet at http://pubs.acs.org.
(44) We observed an analogous cyclization in a model system lacking the C13
acetoxy and C12 isopropyl groups (see ref 15h).
JA060847G
9
J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006 7035