Cyclohexyne cycloinsertion in the divergent synthesis of guanacastepenes

Article abstract of DOI:10.1002/chem.201202222

The guanacastepenes are a family of 15 diterpenes that share a common 5-6-7 tricyclic core, which is decorated with quaternary centers, unsaturation, hydroxyl and carbonyl groups. Some of these natural products show interesting antimicrobial potency. Their collective structural and biological features have stirred up vibrant activity among organic chemists. Herein, we disclose an account of our studies toward the synthesis of a number of guanacastepenes. The synthetic strategy relies on the use of cyclohexyne in a cycloinsertion reaction to rapidly construct the guanacastepene core. Isolation of a cyclobutenol as intermediate in the cyclohexyne cycloinsertion provided us with the possibility to study further the reactivity of this metastable compound, and we uncovered novel rearrangements and ring-opening reactions. Stereoselective, late-stage oxidative diversification of the carbon scaffold allowed the synthesis of guanacastepenes N and O and paved the way for the synthesis of guanacastepenes H and D. No strain, no gain! Studies toward the synthesis of the guanacastepene family of diterpenes are presented. The synthetic strategy relies on the use of cyclohexyne as the guanacastepene C-ring, which undergoes cycloinsertion to form the tricyclic carbon scaffold. A host of procedures developed for the diversification of the carbon core allowed the synthesis of guanacastepenes N and O and paved the way for the synthesis of guanacastepenes D and H. Copyright

Full text of DOI:10.1002/chem.201202222

FULL PAPER
DOI: 10.1002/chem.201202222
Cyclohexyne Cycloinsertion in the Divergent Synthesis of Guanacastepenes
Christian M. Gampe and Erick M. Carreira*[a]
Abstract: The guanacastepenes are a
family of 15 diterpenes that share a
common 5-6-7 tricyclic core, which is
decorated with quaternary centers, un-
saturation, hydroxyl and carbonyl
groups. Some of these natural products
show interesting antimicrobial potency.
Their collective structural and biologi-
cal features have stirred up vibrant ac-
tivity among organic chemists. Herein,
we disclose an account of our studies
toward the synthesis of a number of
guanacastepenes. The synthetic strategy
relies on the use of cyclohexyne in a
cycloinsertion reaction to rapidly con-
struct the guanacastepene core. Isola-
tion of a cyclobutenol as intermediate
in the cyclohexyne cycloinsertion pro-
vided us with the possibility to study
further the reactivity of this metastable
compound, and we uncovered novel re-
arrangements and ring-opening reac-
tions. Stereoselective, late-stage oxida-
tive diversification of the carbon scaf-
fold allowed the synthesis of guanacas-
tepenes N and O and paved the way
for the synthesis of guanacastepenes H
and D.
Keywords: cyclohexyne
· cycloin-
sertion · guanacastepene · natural
products · total synthesis
Introduction
The continued emergence of drug-resistant strains of patho-
genic bacteria compels researchers to maintain a steady
supply of new antibiotic lead structures.[1] In this respect,
natural products have been a valuable source of inspiration
for the development of antibiotic drugs.[2] In the early 2000s,
Clardy set out to identify new antimicrobial metabolites of
endophytic fungi—an underexplored source of natural prod-
ucts.[3] The Clardy group found that extracts from a previ-
ously undescribed basidomycete found in Daphnopsis ameri-
cana in the Guanacaste Conservation area in Costa Rica ex-
hibited pronounced antibacterial properties, as assessed by
agar diffusion assays.[4] Resolution of the crude extracts led
to the identification of a tricyclic diterpene as active antimi-
crobial ingredient, called guanacastepene A (1, Figure 1).
The biological activity of this isoprenoid was related to its
ability to cause rupture of biological membranes in concen-
trations of 12–40 mm, as evidenced by rapid lysis of red
blood cells and induction of K+ efflux from E. coli imp.[4c,5]
Closer examination of the extracts obtained led to identifi-
cation of 14 additional, structurally related guanacastepe-
nes.[4b]
Figure 1. Representative members of the guanacastepene family of diter-
penes.
previously and is thought to arise biosynthetically from cyc-
lization and rearrangement of
a dolabellane skeleton.
These diterpenes share a common tricyclic carbon core
comprising annealed 5-,7-, and 6-membered rings (A, B, and
C ring) and two quaternary stereogenic centers at C11 and
C8 (Figure 1). This terpene scaffold had not been described
Whereas the lower part (C7–C12) is unfunctionalized
throughout the series, the concave rim of the carbon scaffold
(C13&C14, C1–C6) is subject to oxidative diversification,
which introduces carbonyl and hydroxyl functions as well as
unsaturation in the hydrocarbon backbone. In some mem-
bers of the guanacastepenes, C15 is attached to C2 to form a
lactam (e.g., D, H), a dihydrofuran (e.g., E, K), or a lactone
(e.g., L, N, O). Further structural complexity is found in
guanacastepenes D (3) and H (4), in which the lactam unit
is itself part of an oxazepine hemiacetal. Salient structural
feature of guanacastepenes K (6) and L (7) is a norbornane
unit, formed by cyclization of the isopropyl group onto C1,
[a] Dr. C. M. Gampe, Prof. Dr. E. M. Carreira
Laboratorium fꢀr Organische Chemie, ETH Zꢀrich
Wolfgang-Pauli-Str. 10, 8093 Zꢀrich (Switzerland)
Fax : (+41)44-632-13-28
E-mail: carreira@org.chem.ethz.ch
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202222.
Chem. Eur. J. 2012, 18, 15761 – 15771
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
15761

and a tetrahydrofuran, formally formed by linking the C12
isopropyl to the C13 hydroxyl, respectively.
which the C ring, in the form of cyclohexyne, was inserted
into an A–B’ pentalene to create the tricyclic terpene scaf-
fold (strategy III, Scheme 1).[8] Access to the common gua-
nacastepene core allowed for a late-stage oxidative diversifi-
cation, reminiscent of the proposed biosynthetic pathway.
This strategy gave access to the lactone-guanacastepenes N
(8) and O (9) and provides a common precursor to the
lactam-guanacastepenes D (3) and H (4).
Collectively, the structural and biological properties of the
guanacastepenes have piqued the interest of numerous re-
search groups, and a host of synthetic studies toward the
guanacastepenes were published shortly after the initial iso-
lation had been described.[6] Despite the tremendous re-
search efforts, to date only a subset of the guanacastepenes
has been accessible through chemical synthesis,[7] underscor-
ing the synthetic challenges posed by these terpenes. The
synthesis of the 5-7-6-ring framework with its quaternary
stereogenic centers has proven to be non-trivial and formi-
dable as a test of a number of cyclization strategies and re-
actions.[6,7]
Pioneering work in this field has been carried out by Dan-
ishefsky, who disclosed the total synthesis of guanacastepene
A (1) in 2002.[7a,b] The effective strategy for assembly of the
tricyclic carbon core involved functionalization of an A-ring
building block followed by alkylation and cyclization to con-
struct the A–B hydroazulene (Scheme 1, strategy I). In a
Results and Discussion
Synthetic plan: The cycloinsertion strategy as a means of
generating the tricyclic guanacastepene core was based on
our previous finding that cyclohexyne undergoes smooth in-
ꢀ
sertion into the (CO) Ca bond of n-cyclic ketones to create
[n+2.6.0] bicyclic enones.[8a] We surmised that cyclohexyne
cycloinsertion into pentalenone 11 would give rise to the tri-
cyclic guanacastepene scaffold 13 (Scheme 2). The tetrasub-
stituted enone in 13 would then provide a suitable handle
for installation of the C8-methyl group, and dehydrogena-
tion of the resulting ketone would provide 14. Enone 14 was
envisioned as point of divergence in the synthesis, that is
late stage oxidative diversification of the common carbon
scaffold 14 would provide access to a variety of guanacaste-
penes.
Synthesis of pentalenone 11: First task in the synthesis was
the development of a scalable synthesis of cyclopentenone
10, the precursor to pentalenone 11 (Scheme 3). Conjugate
addition of isopropenyl–MgBr to commercially available
enoate 15 proceeded smoothly under copper catalysis
(CuBr·SMe2), with TMSCl as the Lewis-acidic activator and
HMPA as additive, to give ester 16 in 98% yield (>20 g).[9]
It was found that dimethylimidazolidinone (DMI)[10] was
equally effective in promoting the conjugate addition obvi-
ating the use of HMPA on large scale. Ring closure of ester
16 could be effected by Friedel–Crafts acylation in meta-
phosphoric acid at 1258C;[11] however, a 2:1 mixture of de-
sired enone 10 and its regioisomer 19 was obtained in 79%
combined yield. We surmised that this side product had
formed from intermediate cation 18, which could undergo
[1,2]-H shift (Wagner–Meerwein shift) prior to deprotona-
tion.[12] To avoid this side reaction ester 16 was first convert-
ed into the corresponding acid chloride 17, which could be
cyclized at lower reaction temperatures under assistance of
appropriate Lewis acids. A variety of Lewis acids were
screened (TiCl4, SnCl4, AlCl3, BBr3), and it was found that
treatment of 17 with BBr3 at ꢀ788C induced cyclization,
Scheme 1. Strategies for construction of the tricyclic guanacastepene
core.
second alkylation/cyclization sequence, the C8 quaternary
center was installed and the C-ring was annulated. A similar
strategy was chosen by Mehta in his synthesis of the struc-
turally related guanacastepene C (2).[7c] The research groups
of Sorensen,[7d] Trauner,[7e] and Overman[7f] disclosed the
total syntheses of guanacastepenes E (5), and N (8) by judi-
cious selection of convergent strategies for the assembly of
the tricyclic scaffold (strategy II). Thus, the A-ring was teth-
ered to the C-ring and subsequent linking of these rings gen-
erated ring B.
Herein, we document a detailed account of our efforts to
implement a conceptually distinct cycloinsertion approach
for the construction of the tricyclic guanacastepene core, in
Scheme 2. Synthetic planning for the divergent assembly of both lactone- and lactam-guanacastepenes.
15762
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15761 – 15771

Divergent Synthesis of Guanacastepenes
FULL PAPER
and enone 10 was isolated in an 4:1 mixture with 19 in 62–
68% yield. Thus, the overall yield of 10 could be improved;
however, enones 10 and 19 could not efficiently be separat-
ed by column chromatography on SiO2 or distillation, conse-
quently we were compelled to identify reaction conditions
that would selectively generate enone 10. We decided to
carry out an anionic intramolecular allylation to avoid the
interference of the Wagner–Meerwein H-shift observed in
the cationic allylation reaction. To this end, silane 20 was
prepared in 96% yield from enoate 15 by conjugate addition
of the appropriate Grignard reagent under conditions simi-
lar to the ones described above.[13] When allylsilane 20 was
treated with CsF in DMSO smooth allylation occurred in
3 h at 238C to selectively yield enone 10 (64%) along with
protiodesilylated ester 16. Both LiF and KF exhibited signif-
icantly reduced solubility in DMSO and did not induce the
desired reaction.
formed from Lipshutzꢂs cuprate and lithium 4-butene,[19] al-
lowed addition of the organic residue to enone 10 in the
presence of 2 equiv of BF3·OEt2. Under optimized condi-
tions ketone 25 was isolated after 18 h at ꢀ788C in 51%
yield along with 28% starting material, which was resubject-
ed to the reaction conditions. Notably, ketone 25 was ob-
tained as single diastereomer, as judged by analysis of
1H NMR spectra.
Scheme 4. a) Bromobut-4-ene (1.5 equiv), tBuLi (3.0 equiv), Et2O,
ꢀ788C, then: 10 (0.25m in THF, 1.55 equiv), BF3·OEt2 (2.0 equiv),
ꢀ708C, 18 h, 25: 51%, 10: 28%; b) O3, MeOH, Sudan III, ꢀ788C, then:
Me2S, 33–63%; c) OsO4, NMR, THF/H2O, then: NaIO4·SiO2, CH2Cl2,
238C, 88%; d) KOH (1.1 equiv), MeOH, 238C, 20 min, 86%; e) Jonesꢂ
reagent, acetone/H2O, 0 8C quant.; f) ethylene glycol, triethylorthofor-
mate (5 equiv), pTsOH (2 mol%), 238C, 50 min; g) DMSO (2.2 equiv),
oxalylchloride (1.1 equiv), NEt3 (5 equiv), CH2Cl2, ꢀ788C to 238C, 71–
78% over two steps.
Scheme 3. a) Isopropenyl magnesium bromide (1.4 equiv), CuBr·SMe2
(5 mol%), DMI (2.0 equiv), TMSCl (2.0 equiv), THF, ꢀ40 ! 08C, 0.5 h,
98%; b) KOH, MeOH, H2O, 98%; c) SO2Cl2, CH2Cl2, quant; d) 16,
meta-polyphosphoric acid, 1258C, 1 h, 79 %, 10/19 2:1; e) 17, BBr3
(1.05 equiv), CH2Cl2, 2 h, ꢀ78 ! 238C, 62–68%, 10/19 8:2; f) (3-(trime-
thylsilyl)prop-1-en-2-yl)magnesium bromide (2.0 equiv), CuBr·SMe2
(5 mol%), HMPA (2.5 equiv), TMSCl (2.0 equiv), THF, ꢀ408C ! RT,
96%; g) CsF (1.05 equiv), DMSO, 238C, 3 h, 64%.
Oxidative cleavage of the double bond in 25 by ozonoly-
sis, using Sudan III as indicator,[20] provided aldehyde 26,
albeit in drastically varying yields (33–63%). In contrast, di-
hydroxylation with catalytic OsO4 followed by diol cleavage
with NaIO4 on SiO2 gave aldehyde 26 in reliable 88%
yield.[21] Treatment of 26 with KOH in MeOH led to rapid
formation of the epimeric pentalenes 27 and 28 along with
bridged side product 29. In order to verify the assigned
structures, in particular the configuration of the quaternary
center, the isomers were separated by chromatography on
silica gel, and alcohols 27 and 28 were separately oxidized
under Jonesꢂ conditions to give diketone 30.[22] Pentalene-
dione 30 readily crystallized from Et2O, and X-ray crystallo-
graphic analysis showed that the desired syn-pentalene scaf-
fold had been formed and that the C11 stereogenic center
had been constructed with the appropriate guanacastepene
configuration, that is syn to the iPr group.[23] The unpurified
reaction mixture obtained from aldol addition (27, 28, and
29) could directly be subjected to acetalization with ethylene
glycol, and ketone 11 was obtained after subsequent Swern-
oxidation in 71–78% yield over 2 steps and in 21% yield
over seven steps from enoate 15 (>10 g).[24]
With enone 10 in hand we focused on the introduction of
the C11-quaternary center,[14] which could be installed by
1,4-addition of
a
three-carbon chain to enone 10
(Scheme 4). Incorporation of an appropriate chain would fa-
cilitate cyclization to C1 to form the pentalene scaffold by
an intramolecular aldol addition. For this purpose a number
of three-carbon nucleophiles were tested, however, as re-
ported previously for 10 and similar substrates,[7e,f] function-
alization of enone 10 by conjugate addition proved difficult.
For instance, despite extensive screening of conditions, diox-
olane 22 could only be obtained in 11% yield (using a Lip-
shutz cuprate[15] and BF3·OEt2). Generation of acetylene 23
was thwarted by the low reactivity of enone 10 toward ace-
tylene alane,[16] -alanate[17] or cuprate reagents,[18] and 1,4-ad-
dition of an allyl moiety to obtain 24 was likewise not possi-
ble using allyl-SiMe3, diallyl zinc, or allyl cuprate reagents.
In all cases enone 10 did not react or suffered exclusive 1,2-
addition. Gratifyingly, the use of a homoallylic cuprate,
Chem. Eur. J. 2012, 18, 15761 – 15771
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15763

E. M. Carreira and C. M. Gampe
Cyclohexyne cycloinsertion: Rapid access to pentalenone 11
provided us with sufficient material to explore the cycloin-
sertion strategy (Scheme 5). Treatment of 11 with KOCEt3
(3.0 equiv) and iodonium salt 31 (2.4 equiv), as previously
described,[8a,b] rapidly generated a single reaction product
that slowly underwent conversion into a 2:1 mixture of
enones 34 and 35 upon prolonged stirring at room tempera-
ture (ca. 2 h). If the cycloinsertion reaction was allowed to
warm from ꢀ788C to room temperature and quenched after
20 min, cyclobutenol 32 was isolated as single diastereomer
in 70–74% yield along with 13% starting material.
Functionalization of cyclobutenol 32: The tricyclic core of
the guanacastepenes was accessible in only eight chemical
steps from commercially available starting material, namely
enoate 15. However, in light of the impending installation of
the C8 quaternary center, generation of a mixture of 34 and
35 seemed disadvantageous. Although 34 could potentially
be subjected to tautomerization to give enone 13
(Scheme 2), a key intermediate in the synthetic plan, forma-
tion of side product 35 and introduction of an isomerization
step would significantly reduce the efficiency of the process.
Therefore, we decided to explore further the chemical reac-
tivity of cyclobutenol 32.
We surmised that the intrinsic reactivity of the strained
double bond or the allylic tertiary alcohol could be har-
nessed to directly install the required C8-quaternary center
in 32 (Scheme 6). Consecutive ring opening of the resulting
cyclobutane would generate ketone 36. We first investigated
radical additions to the strained double bond. Although
photochemically generated carbon radicals of methanol,
acetonitrile, or 1,3-dioxolane are known to add to enones
and styrenes,[28] cyclobutene 32 remained untouched under
these conditions. Attempts to react cyclobutenol 32 in a
Claisen- or Ireland–Claisen rearrangement were not met
with success, and starting material was reisolated.[29] Like-
wise, neither SN2’ substitution of the allylic acetate in Ac-
32,[30] nor cyclopropanation of 32, or its TMS protected de-
rivative, could be effected under a variety of conditions
(Furukawa-[31] or Shi-Furukawa;[32] CH2N2 or TMSCHN2
photochemical[33] or metal catalyzed[34]). In these cases
either no reaction or degradation of the starting material
was observed. Consequently, we abandoned the strategy to
functionalize C8 in 32 and set the focus on exploration of
novel ring opening reactions of cyclobutene 32, which would
pave the way for installation of the C8 methyl group on the
tricyclic scaffold.
Scheme 5. a) KOCEt3 (3.0 equiv), 31 (2.4 equiv), THF, ꢀ788C ! RT,
20 min; b)Ac2O (40 equiv), DMAP (3 equiv), iPrNEt2 (48 equiv), tol-
uene, 1208C, 16 h, 40% (11% s.m.).
The structure of 32 was tentatively assigned by 1D- and
2D-NMR spectroscopy, but we aimed to validate the struc-
ture by x-ray crystallography. Whereas 32 was crystallized as
thin needles unsuitable for X-ray scattering, acylation of 32
under forcing conditions (Ac2O (40 equiv), DMAP
(3 equiv), iPrNEt2 (48 equiv), toluene, 1208C, 16 h) yielded
acetate Ac-32 in 40% yield, which crystallized readily from
Et2O. X-ray crystallographic analysis of Ac-32 validated the
initial structural assignment and revealed the intriguing
structural properties of the cyclobutene unit.[25] Most nota-
bly, the ring strain of the cyclobutene of about 32 kcalmolꢀ1
Scheme 6. Attempted functionalization of cyclobutenes 32 and Ac-32 at
C8.
manifests itself in a stretched C2–C9 single bond (1.574 ꢃ)
[26]
ꢀ
and a shortened C C double bond (1.329 ꢃ).
The isolation of 32 implicates it as an intermediate in the
cyclohexyne cycloinsertion cascade. This hypothesis was sup-
ported by our observation that treatment of 32 with KOtBu
in THF resulted in formation of a 2:1 mixture of enones 34
and 35. In agreement with our previously devised working
hypothesis for cyclohexyne cycloinsertion we surmise that
the initially formed 32-alkoxide first rearranges to enolate
33 by torquoselective electrocyclic ring opening.[8a] This in-
termediate can follow two reaction pathways: Isomerization
of the trans-double bond and protonation gives enone 34,
whereas tautomerization of the dienolate to C1, followed by
elimination of the dioxolane, furnishes 35.[27]
Rearrangements of cyclobutenol 32: We had observed that
electrocyclic ring opening of cyclobutene 32 occurs readily
under alkaline conditions, that is, via the potassium alkoxide
of 32 (see above). These conditions, however, gave rise to a
mixture of products, and we set out to identify conditions
that would allow ring opening to occur more selectively. Sur-
prisingly, attempts to induce thermal cyclobutene opening in
32 under neutral conditions were not met with success. Cy-
clobutene 32 proved remarkably stable and could be heated
to 1608C in xylene without undergoing any noticeable reac-
tion after 12 h. The fact that the corresponding potassium
15764
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15761 – 15771

Divergent Synthesis of Guanacastepenes
FULL PAPER
alkoxide undergoes cyclobutene cleavage at room tempera-
ture certainly underscores the previously observed accelerat-
ing effect of alkoxide substituents on electrocyclic reac-
tions.[35]
After exploring isohypsic transformations of cyclobutenol
32, the possibility of inducing oxidative ring rupture was ex-
amined (Scheme 8). Exposure of 32 to diacetoxyiodoben-
[43]
zene/I2 or PdACTHNUTRGNEUNG
(OAc)2/O2/pyridine[44] led to decomposition
The Woodward–Hoffmann rules predict thermal electro-
cyclic ring opening of cyclobutenes to proceed in a conrota-
tory fashion,[36] and we hypothesized that the activation bar-
rier to formation of a strained cis-trans-cycloheptadiene was
too high for this process to occur under neutral, that is unca-
talyzed, thermal conditions. Hence, we turned our attention
to photochemical ring opening, which was envisioned to
proceed more easily, since this electrocyclic reaction would
occur in a disrotatory fashion to directly give a cis,cis-cyclo-
heptadiene. However, 32 was reluctant to undergo ring
opening in various solvents upon irradiation with a Hg
medium-pressure lamp for 6 h.[37] Metal catalysts (AgI, CuI,
Fe0, PdII, RhI) have likewise been shown to effect 4-electron
electrocyclic reactions in a disrotatory mode and were
tested next for their ability to induce selective cyclobutene
ring opening.[38] Surprisingly, treatment of 32 with AgBF4
led to formal [1,3]-carbon shift of C1 to C8 to give cyclobu-
tanone 37 in 65% yield (Scheme 7). Pettit and co-workers
of the starting material. However, when 32 was treated with
ꢀ
Pb
N
smoothly and diene 41 (80%) was isolated along with ace-
tate 40 (8%).[45] Both products were crystalline and their
structure could unambiguously be validated by X-ray crys-
tallography.[46] We hypothesize that acetate 40 and diene 41
are formed from the same intermediate allyl cation 39 by
acetate addition and deprotonation, respectively.
Scheme 8. PbACTHUNGTREN(NUGN OAc)4 (1.05 equiv), CH2Cl2, 238C, 45 min, 40: 8%, 41:
80%.
Construction of the C8 quaternary center: Scrutiny of the
reactivity of cyclobutenol 32 rewarded us with a number of
intriguing reactions. In particular, enone 13 and diene 41
were highly valuable substrates for the installation of the C8
quaternary center and the missing C4-carboxyl unit. Availa-
bility of diene 41 offered the possibility to first add a car-
ꢀ
boxyl surrogate to the C3 C4 enone to afterwards generate
ꢀ
a C3 C8 enone (cf. 44, Scheme 9), in which the C8 quater-
nary center could be installed. A nitrile group seemed per-
fectly suited as carboxyl equivalent to access both the
lactam and lactone series of the guanacastepenes. Treatment
of 41 with 1.05 equiv of Nagataꢂs reagent (Et2AlCN) in
CH2Cl2 at room temperature led to formation of nitrile 42
in a 1:1 mixture of diastereomers (39%) along with 24% of
a byproduct, identified as norbornane 43.[47,48] This unantici-
Scheme 7. a) AgBF4 (1.0 equiv), THF, 508C, 2.5 h, 65%; b) [Fe2(CO)9]
(2.0 equiv), PhH, 20 h, 958C, then addition of DBU (20 equiv), 8 h, 51%.
first demonstrated that coordination of Fe(CO)4 to simple
cyclobutenes leads to ring opening to give iron diolefin com-
plexes.[39] Formation of the initial Fe(CO)4-cyclobutene com-
plex was thereby achieved by treatment of the substrate
with [Fe2(CO)9], which disproportionates under thermal
conditions and transfers a Fe(CO)4 to the olefin. We were
pleased to find that treatment of 32 with this diironcarbonyl
complex in benzene at 958C indeed led to formation of de-
conjugated enone 34. Addition of DBU to the reaction in-
duced demetallation as well as tautomerization to give
enone 13 in 51% yield.[40] Alternative procedures to transfer
the iron carbonyl unit to cyclobutenol 32, that is under pho-
tochemical conditions, using sonication,[41] or under Schiff-
base catalysis,[42] furnished enone 13 in only 10–35% yield.
Scheme 9. a) Et2AlCN (2.0 equiv), CH2Cl2, 08C, 5 min, 96%, d.r. >95:5;
b) DBU (0.75 equiv), CH2Cl2, 238C, 6 h, 88%.
Chem. Eur. J. 2012, 18, 15761 – 15771
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15765

E. M. Carreira and C. M. Gampe
pated bond formation between C9 and C14 can be rational-
ized considering the bent structure of diene 41, which brings
C9 and C14 in unexpectedly close proximity (ca. 3.7 ꢃ,
Scheme 8). We hypothesized that ring contraction could
occur in an intermediate aluminum enolate, formed from 41
and Nagataꢂs reagent, upon activation of the acetal by the
Lewis-acidic reagent. Gratifyingly, we found that formation
of the side product could be avoided if 41 was exposed to
2 equiv of Nagataꢂs reagent at 08C, which led to formation
of nitrile 42 in 96% as single diastereomer after only 5 min.
The C4-a-configuration was determined by analysis of the
ꢀ
C3 C4 proton coupling constants and is in agreement with
Scheme 10. a) DIBAL-H (1.1 equiv), nBuLi (1.1 equiv), PhMe, ꢀ788C,
30 min, 65%, (18% s.m.); b) ZnEt2 (1m in hexane, 4 equiv), TFA
(4 equiv), CH2I2 (4 equiv), CH2Cl2, 08C, 15 min, 80–83%; c) PDC
(1.1 equiv), Ac2O (2.5 equiv), CH2Cl2, 238C, 5.5 h, 85–95%; d) Li in NH3,
ꢀ408C, 20 min, then: CH2Cl2, SiO2, O2, 70–79%; e) Li in NH3, ꢀ408C,
20 min, then: HCl/H2O or NaOMe/MeOH, 95%; f) Me2S (20 equiv), ace-
tone, 238C, 15 h, 68–79% over steps d and f; g) SOCl2 (2.8 equiv), pyri-
dine, 08C, 5 min, 85–95%.
earlier reports by Nagata and co-workers of the primary
adduct being formed by axial attack of the nitrile reagent.
For installation of the C8 quaternary center enone 42 was
isomerized to 44 by the action of DBU (Scheme 9). With 44
in hand we tested multiple conditions for direct addition of
a methyl nucleophile to 44 and were disappointed to find
that the quaternary center could not be installed using
methyl-cuprates,[49] -zincates,[50] or aluminates.[51] Enone 44
proved furthermore reluctant to cyclopropanation with
sulfur ylides[52] or with ZnEt2/CH2I2 or ZnEt2/TFA/
CH2I2.[31,32] Thus our attention was drawn to Daubenꢂs pro-
tocol for installation of axial methyl groups in steroid scaf-
folds.[53] This procedure involves the reduction of an enone
to an allylic alcohol, which is used in a directed cyclopropa-
nation of the double bond. After oxidation to the corre-
sponding cyclopropyl ketone the methyl group can then be
generated from the cyclopropane under dissolving metal
conditions. Reduction of enone 44 was plagued by formation
of multiple products and the desired allylic alcohol could
not be cleanly obtained under various conditions (DIBALH,
NH3(l) cyclopropyl cleavage was smoothly effected. Surpris-
ingly, after chromatographic purification of the reaction
mixture on SiO2, hydroperoxide 49 was obtained instead of
the expected ketone 48. It was hypothesized that 49 was
formed by oxidation of an intermediate enol in air. In fact,
such facile oxidation of metastable enols has been described
previously, for instance by Kuwajima and co-workers in
their synthesis of taxusin.[56] Treatment of the reaction mix-
ture obtained from cyclopropane cleavage with HCl/THF or
NaOMe/MeOH suppressed formation of 49 and provided
ketone 48 in up to 95% yield, presumably by facilitating
tautomerization of the intermediate enol. In contrast, if the
unpurified product mixture was dissolved in CH2Cl2 and stir-
red in air, or under an atmosphere of O2, formation of 49
was observed;[17] however, only if SiO2 was added. This pro-
cedure provided hydroperoxide 49 in 70–79% yield from 47.
With the C8 quaternary center installed we turned our at-
tention to incorporation of the last missing carbon at C4. To
this end the plan was to access enone 14, which would un-
dergo 1,4-addition by a suitable C1-carboxyl surrogate. Oxi-
dation of ketone 48 to enone 14 proved surprisingly difficult
and could not be effected under various conditions (i.e.,
IBX, LDA/PhSeBr, DDQ, benzylselenic acid),[57] most likely
due to steric shielding of C3. The ability to selectively gener-
ate hydroperoxide 49, which was initially dismissed as side
DIBAL-H·nBuLi, HAlACHTNUTRGNE(UGN OtBu)3Li, NaBH4). We thus reas-
sessed our synthetic strategy and focused on first installing
the C8 methyl group in enone 13 before addition of the C4
substituent.
We subjected enone 13 to Daubenꢂs procedure and were
pleased to find that 1,2-reduction of enone 13 proceeded
smoothly, using a 1:1 mixture of DIBAL-H and nBuLi, to
give exclusively the a-configured allylic alcohol 46 in 65%
yield along with 18% starting material (Scheme 10).[54] The
directing effect of the allylic hydroxyl under Furukawaꢂs cy-
clopropanation conditions[31] (ZnEt2/CH2I2) was low, and a
1:1 mixture of diastereomeric cyclopropane products was
obtained. Gratifyingly, we observed excellent substrate in-
duced diastereoselectivity in the cyclopropanation reaction
when Shiꢂs conditions were employed.[32] Thus, ZnEt2 was
first reacted with 1 equiv of TFA followed by CH2I2 to gen-
erate the highly electrophilic cyclopropanating reagent
TFAZnCH2I, to which allylic alcohol 46 was added. After
consecutive oxidation of the secondary alcohol, cyclopropyl
ketone 47 was obtained in 51% over three steps, and its
structure was unambiguously confirmed by X-ray crystallog-
raphy.[8b,55]
ꢀ
product, now proved beneficial. The O O bond in 49 was
cleaved by action of Me2S and the resulting tertiary alcohol
was treated with SOCl2 to give enone 14 in 85–95% yield
over two steps. Enone 14 was envisioned as branching point
in the synthesis. Introduction of appropriate carboxyl and
amide surrogates at C4 would give access to the lactone and
lactam series of the guanacastepenes, respectively.
Formation of the lactam and lactone guanacastepene scaf-
fold: We first investigated installation of the lactam and lac-
tone rings using cyanide as precursor to both. To this end a
nitrile at C4 was installed using Nagataꢂs reagent,[47] to
X-ray crystallographic analysis of 47 showed that the con-
figuration at C8 had been set with the targeted guanacaste-
pene configuration.[8b] Upon treatment of 47 with Li in
15766
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15761 – 15771

Divergent Synthesis of Guanacastepenes
FULL PAPER
obtain 50 in 82% yield and as 2:1 mixture of epimers at C4
(Scheme 11). Treatment of 50 with HClO4 (35% aq.) in di-
oxane at room temperature effected dioxolane and nitrile
hydrolysis followed by condensation of the amide group
with the C2-ketone to give lactam 51 in 20–30% yield. Sur-
prisingly, dinitrile 52 was also obtained in 30–40% yield. Its
structure was established by X-ray crystallography,[58] and its
formation can be rationalized by elimination of HCN from
50 followed by condensation of HCN with the C14 keto
group. We could not identify reaction conditions, neither
acidic nor basic, that would prevent this undesired elimina-
tion reaction. Furthermore, it is interesting to note that
lactam 51 is stable to further hydrolysis to the corresponding
lactone, and, while this sequence provides the lactam-guana-
castepene scaffold, we sought a procedure to introduce a
C4-carboxylate that would give access to a precursor to the
lactone-guanacastepenes.
Scheme 12. a) 53 (4.0 equiv), Et2O, 23 8C, 45 min, then: NaOMe/MeOH,
10 min, 78–82% yield; b) RuCl3 (6 mol%), oxone (7.7 equiv), NaHCO3
(17 equiv), MeCN, H2O, EtOAc, 4.5 h, 55% after 1ꢄrecycling; c) THF/
HCl (1.0m), 578C, 4 h; then: (COCl)2 (1.0 equiv), pyridine (10 equiv),
CH2Cl2, 08C, 15 min, 79%.
intermediate, acetylenes are first converted into their corre-
sponding a-diketones. Whereas diol cleavage is smoothly ef-
ꢀ
fected under these conditions, rupture of the C C bond in
an a-diketone requires harsher conditions. In this respect, it
has been documented that RuO4 is capable of cleaving ace-
tylenes, albeit with the drawback that oxidation even of un-
activated CH groups is often observed as side reaction. We
found that acetylene 54 did not react with OsO4/NMO in
acetone/H2O or KMnO4/NaIO4 in tBuOH/H2O.[63] The use
of H2O2 in combination with a molybdophosphate catalyst[64]
or C6F5IACHTUNGRTNEUNG(O2CCF3)2 led to degradation of the starting materi-
al and we were left with RuO4 as oxidant.[65] In a biphasic
solvent system, as described by Sharpless and co-workers,[66]
the action of catalytic amounts of RuO4 led to multiple reac-
tion products and acid 55 was only detected in traces. The
working group of Yang recently prescribed the use of RuO4
along with oxone as nucleophilic oxidant to effect cleavage
of the a-diketone intermediate.[67] Application of this proce-
dure furnished acid 55 in 34% yield along with 22% of the
diketone intermediate. It was found that best results were
obtained if the reaction was stopped after about 50% con-
version of the starting material. Acid 55 was then isolated
and the remaining mixture of acetylene 54 and diketone was
resubjected to the reaction conditions. This procedure gave
access to acid 55 in 55% yield from acetylene 54. Treatment
of acid 55 with aqueous HCl in THF hydrolyzed the dioxo-
lane, and consecutive treatment with oxalylchloride/pyridine
effected lactonization after epimerization at C4.
Scheme 11. a) Et2AlCN (1.0m in PhMe, 4.0 equiv), CH2Cl2, 08C, 2 h,
82%, d.r. 2:1; b) dioxane, HClO4 (35% aq.), 238C, 21 h; 51: 20–30%; 52:
30–40%.
We investigated the use of a number of carboxyl surro-
gates that have been documented in the literature.[59] How-
ever, it was found that enone 14 did not participate in a Ru-
catalyzed hydroformylation[60] reaction; it also failed to par-
take in radical addition reactions using 1,3-dioxolane.[28]
Conjugate addition of bromoform could likewise not be ef-
fected,[61] and addition of furan or benzofuran using the cor-
responding dimethylalane reagents was also not possible.[62]
We surmised that conjugate addition of the alane reagents
to the rigid carbon core of enone 14 was disfavored due to
1,3-allylic strain in an intermediate enolate. Since introduc-
tion of a nitrile group had proceeded smoothly we investi-
gated the use of an acetylene as carboxyl surrogate. Gratify-
ingly, as 14 was treated with alane 53 (4 equiv) in Et2O
smooth addition occurred, and acetylene 54 was obtained in
78–82% yield (Scheme 12). It should be noted that the enol
intermediate needed to be tautomerized with NaOMe/
MeOH to avoid its oxidation in air, as seen in case of
ketone 47.
Progress toward guanacastepenes D and H: Access to
lactam 51 enabled us to explore the oxidative functionaliza-
tion of the carbon core en route to guanacastepenes D and
H (Scheme 13). In scouting experiments we found that ex-
posure of lactam 51 to excess NBS (13 equiv) in CH2Cl2 led
to dehydrogenation, presumably via the C4-bromide, to give
57 in 40% yield. Reaction of lactam 51 with NCS
(2.0 equiv) in THF furnished allylic chloride 58 in 95%
yield. We assume that this reaction follows a radical path-
way via the C4-chloride, which eliminates HCl and suffers
consecutive allylic chlorination.[68] A similar dehydrogena-
tion/halogenation sequence has previously been described
by Overman.[7f] A different source of electrophilic bromide,
that is Me3NPhBr3, displayed entirely different reactivity
and induced bromination at C13 of lactones 51 and 57 to
give lactones 59 (d.r. 1:1) and 60 (d.r. a/b 4:1), respectively.
Next in the synthesis was the transformation of the acety-
lene into a carboxyl unit. However, oxidative cleavage of in-
ternal acetylenes is challenging. In analogy to oxidative
cleavage of an olefin with strong oxidizing reagents
(KMnO4, NaIO4, PbACTHNUGTRNE(NUG OAc)4), which proceeds through a diol
Chem. Eur. J. 2012, 18, 15761 – 15771
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15767

E. M. Carreira and C. M. Gampe
With these facile procedures for dehydrogenation and func-
tionalization at C13 and C5 at our disposal we are in a posi-
tion to examine the introduction of the oxazepine unit and
conversion of the bromide and chloride substituents into a
hydroxyl and ketone, respectively.
lowed by Davis’ oxaziridine. Therefore, we changed the
order of functionalization to first install the C13-OH before
introducing the C3/C4 unsaturation. We found that lactone
56 was easily converted into its bis-TBS enol ether by expo-
sure to excess TBSOTf/NEt3. Treatment of the unpurified si-
loxane with OsO4/NMO in aqueous acetone/HOtBu induced
dehydrogenation at C3/C4 along with hydroxylation at C13,
and ketone 63 was obtained along with alcohol 64 in 69%
combined yield. Most notably, hydroxylation had occurred
from the a-face of the molecule to give 64 in a d.r. of 9:1.
Acylation of 64 provided 66, which showed a 1H NMR
proton coupling constant of the C13 proton of 11.8 Hz. This
value is consistent with a C13-a-configuration.[7a] Notably,
the stereoinduction in the hydroxylation step is contrary to
that obtained in procedures previously described for the in-
troduction of the C13 hydroxyl by epoxidation of a similar
silyl enol ether, which furnished the b-hydroxide in d.r. of
9:1.[7a,d,f] The use of other silyl groups (TMS, TES) or em-
ployment of a mixed C14-TES/C15-TBS-enol ether resulted
in a faster reaction but diminished yields. Interestingly, ex-
posure of the bis-TBS-enol ether to stoichiometric amounts
of OsO4 in anhydrous THF, as prescribed by Kishi and co-
workers,[69] led to a sluggish reaction and ketone 63 was
formed exclusively.
Scheme 13. a) NBS (13 equiv), CH2Cl2, 24 h, 238C, 40%; b) NCS
(2.0 equiv), THF, 12 h, 238C, 95%, d.r. >95:5; c) Me3NPhBr3 (10 equiv),
THF, 74 h, 238C, 90%, d.r. 1:1; d) Me3NPhBr3 (2.5 equiv), THF, 238C,
3.5 h, 90%, d.r. a/b 4:1.
This OsO4-induced oxidation cascade proved as valuable
point of bifurcation in the synthesis, since treatment of
Syntheses of guanacastepenes N and O: For the final oxida-
tive diversification of the lactone-guanacastepene carbon
core 56 we envisioned installation of the C3/C4 unsaturation
and a C5-bromide under radical conditions, in analogy to
work by Overman (Scheme 14).[7f] We were pleased to find
that treatment of 56 with NBS (2.3 equiv) and (PhCO2)2
(0.1 equiv) in refluxing CCl4 generated the desired bromide
61 in 40% yield along with the C4-brominated lactone 62.
The C13-hydroxyl was planned to be installed in 61 by oxi-
dation of an intermediate enol ether. Surprisingly, the C13/
C14-silyl enol ether could not be derived from 61 by either
treatment with LDA/TMSCl, KHMDS/TMSCl, TMSOTf/
NEt3. Likewise 61 did not react upon exposure to LDA fol-
ketone 63 with MnACTHNGUTERNNU(G OAc)3 furnished the epimeric acetate 65
in 68% yield with 4:1 preference for b-configuration at
C13.[70] Stereoselective access to both acetate epimers 65
and 66 enabled the synthesis of both guanacastepenes N and
O, respectively. Allylic radical bromination of 65 at C5 gen-
erated the last synthetic intermediate in Overmanꢂs synthe-
sis of guanacastepene N (8),[7f] which can be transformed
into the target structure by a radical Br/OH substitution,
pioneered by Nakamura and co-workers.[71] A similar se-
quence disclosed the first total synthesis of guanacastepene
O (9). The diastereoselectivities obtained in this final radical
1
functionalization are remarkable (>95:5 by H NMR) and
can be explained by the concept of torsional steering:[72] The
Scheme 14. a) NBS (2.3 equiv), (PhCO2)2 (0.1 equiv), CCl4, 1 h, 858C, 40%; b) TBSOTf (4.0 equiv), NEt3 (20 equiv), CH2Cl2, ꢀ788C to 238C, 30 min;
then: OsO4 (0.2 equiv), NMO (2.0 equiv), MeSO2NH2 (1.0 equiv), acetone/H2O/HOtBu, 08C, 21 h, 63: 40%; 64: 29% (d.r. 9:1); c) Mn(OAc)3
ACHTUNGTRENNUNG
(6.5 equiv), molecular sieves 3 ꢃ, PhH, 858C, 3 h, 68%; d) NBS (1.1 equiv), (PhCO2)2 (0.1 equiv), CCl4, 858C, 1 h, 58% (d.r. >95:5); e) ref. [7f]; f)Ac2O
(5 equiv), NEt3 (20 equiv), DMAP (1 equiv), CH2Cl2, 238C, 87%; g) NBS (1.1 equiv), (PhCO2)2 (0.1 equiv), CCl4, 858C, 1 h, 76% (d.r. >95:5);
h) HSnBu3 (10 equiv), PhMe, air, 5 h, then: PPh3 in CHCl3, 1 h, 9: 72% (d.r. >95:5), 66: 27%.
15768
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15761 – 15771

Divergent Synthesis of Guanacastepenes
FULL PAPER
Luckman, S. F. Brady, J. Clardy, M. Greenstein, W. M. Maiese, J.
Antibiot. 2000, 53, 256–261.
observed pseudo-axial attack of O2 on the intermediate allyl
radical allows the guanacastepene C-ring to rotate directly
into a favorable chair conformation, whereas equatorial
attack would lead to a less-favorable twist-boat conformer.
[5] B. Sampson, R. Misra, S. Benson, Genetics 1989, 122, 491–501.
[6] Synthetic studies toward the guanacastepenes: a) B. B. Snider, N. A.
Hawryluk, Org. Lett. 2001, 3, 569–572; b) P. Magnus, M. J. Waring,
C. Ollivier, V. Lynch, Tetrahedron Lett. 2001, 42, 4947–4950, erra-
tum: P. Magnus, M. J. Waring, C. Ollivier, V. Lynch, Tetrahedron
Lett. 2001, 42, 6977; c) G. B. Dudley, S. J. Danishefsky, Org. Lett.
2001, 3, 2399–2402; d) G. B. Dudley, D. S. Tan, G. Kim, J. M. Tanski,
S. J. Danishefsky, Tetrahedron Lett. 2001, 42, 6789–6791; e) B. B.
Snider, B. Shi, Tetrahedron Lett. 2001, 42, 9123–9126; f) G. Mehta,
J. D. Umarye, Org. Lett. 2002, 4, 1063–1066; g) S. N. Gradl, J. J.
Kennedy-Smith, J. Kim, D. Trauner, Synlett 2002, 411–414; h) T. M.
Nguyen, D. Lee, Tetrahedron Lett. 2002, 43, 4033–4036; i) G. Mehta,
J. D. Umarye, V. Gagliardini, Tetrahedron Lett. 2002, 43, 6975–6978;
j) A. Nakazaki, U. Sharma, M. A. Tius, Org. Lett. 2002, 4, 3363–
3366; k) F. D. Boyer, I. Hanna, Tetrahedron Lett. 2002, 43, 7469–
7472; l) P. Magnus, C. Ollivier, Tetrahedron Lett. 2002, 43, 9605–
9609; m) W. D. Shipe, E. J. Sorensen, Org. Lett. 2002, 4, 2063–2066;
n) T. M. Nguyen, R. J. Seifert, D. R. Mowrey, D. Lee, Org. Lett.
2002, 4, 3959–3962; o) G. Mehta, J. D. Umarye, K. Srinivas, Tetrahe-
dron Lett. 2003, 44, 4233–4237; p) X. Du, H. V. Chu, O. Kwon, Org.
Lett. 2003, 5, 1923–1926; q) K. M. Brummond, D. Gao, Org. Lett.
2003, 5, 3491–3494; r) C. C. Hughes, J. J. Kennedy-Smith, D. Trau-
ner, Org. Lett. 2003, 5, 4113–4115; s) R. Sarpong, J. T. Su, B. M.
Stoltz, J. Am. Chem. Soc. 2003, 125, 13624–13625; t) S. Srikrishna,
D. H. Dethe, Org. Lett. 2004, 6, 165–168; u) P. Chiu, S. Li, Org. Lett.
2004, 6, 613–616; v) X. Du, H. V. Chu, O. Kwon, Tetrahedron Lett.
2004, 45, 8843–8846; w) C. C. Hughes, A. K. Miller, D. Trauner,
Org. Lett. 2005, 7, 3425–3428; x) C.-C. Li, X.-H. Zhang, Z.-X. Xie,
J.-H. Chen, Z. Yang, Org. Lett. 2005, 7, 3709–3712; y) C.-C. Li, C.-
H. Wang, B. Liang, X.-H. Zhang, L.-J. Deng, S. Liang, J.-H. Chen,
Y.-D. Wu, Z. Yang, J. Org. Chem. 2006, 71, 6892–6897; z) C. A.
McGowan, A.-K. Schmieder, L. Roberts, M. F. Greaney, Org.
Biomol. Chem. 2007, 5, 1522–1524; aa) K. Michalak, M. Michalak,
J. Wicha, Tetrahedron Lett. 2008, 49, 6807–6809; for formal synthe-
sis of guanacastepene A, see: ab) B. Shi, N. A. Hawryluk, B. B.
Snider, J. Org. Chem. 2003, 68, 1030–1042; ac) F.-D. Boyer, I.
Hanna, L. Ricard, Org. Lett. 2004, 6, 1817–1820; ad) ref. [7d];
ae) K. Michalak, M. Michalak, J. Wicha, Tetrahedron Lett. 2010, 51,
4344–4346; for a review of the synthetic studies, see: af) M. Hierse-
mann, H. Helmboldt, Top. Curr. Chem. 2005, 243, 73–136; ag) S. V.
Maifeld, D. Lee, Synlett 2006, 1623–1644.
[7] Total synthesis of guanacastepene A: a) D. S. Tan, G. B. Dudley, S.
Danishefsky, Angew. Chem. 2002, 114, 2289–2292; Angew. Chem.
Int. Ed. 2002, 41, 2185–2188; b) S. N. Lin, G. B. Dudley, D. S. Tan,
S. J. Danishefsky, Angew. Chem. 2002, 114, 2292–2295; Angew.
Chem. Int. Ed. 2002, 41, 2188–2191; guanacastepene C: c) G.
Mehta, K. Palavi, J. D. Umarye, Chem. Commun. 2005, 4456–4458;
guanacastepene E: d) W. D. Shipe, E. J. Sorensen, J. Am. Chem. Soc.
2006, 128, 7025–7035; e) A. K. Miller, C. C. Hughes, J. J. Kennedy-
Smith, S. N. Gradl, D. Trauner, J. Am. Chem. Soc. 2006, 128, 17057–
17062; guanacastepene N: f) S. Iimura, L. E. Overman, R. Paulini,
A. Zakarian, J. Am. Chem. Soc. 2006, 128, 13095–13101.
[8] a) C. M. Gampe, S. Boulos, E. M. Carreira, Angew. Chem. 2010, 122,
4186–4189; Angew. Chem. Int. Ed. 2010, 49, 4092–4095; b) C. M.
Gampe, E. M. Carreira, Angew. Chem. 2011, 123, 3019–3022;
Angew. Chem. Int. Ed. 2011, 50, 2962–2965; c) for a recent review
of the use of arynes and cyclohexyne in natural product synthesis,
see: C. M. Gampe, E. M. Carreira, Angew. Chem. 2012, 124, 3829–
3842; Angew. Chem. Int. Ed. 2012, 51, 3766–3778.
Conclusion
In summary, we have reported our studies concerning the di-
vergent synthesis of a variety of guanacastepenes using an
unprecedented cycloinsertion strategy to generate the tricy-
clic guanacastepene core. A concise and scalable synthetic
route provided a pentalenone, which was subjected to cyclo-
hexyne cycloinsertion, furnishing the guanacastepene tricy-
clic scaffold in eight chemical steps. Interruption of the cy-
cloinsertion cascade at the stage of the initially formed cy-
clobutenol allowed us to investigate the reactivity of this
synthetic intermediate, and a number of intriguing rear-
rangements and ring opening reactions were unraveled. The
iron(0)-promoted electrocyclic ring opening and the Pb-
ACHTUNGTRENNUNG(OAc)4-induced facile oxidative cyclobutenol rupture are
most notable, providing the guanacastepene core in different
oxidation states. Introduction of the second quaternary
center was achieved by substrate controlled diastereoselec-
tive cyclopropanation followed by formal hydroperoxide
opening of the cyclopropane. Access to the lactam and lac-
tone guanacastepene scaffold was accomplished by the use
of cyanide and a TMS-acetylene as carbonyl surrogate, re-
spectively. We developed various procedures for the diver-
gent oxidation of the lactam-scaffold, which will be useful in
the synthesis of guanacastepenes D and H. Furthermore, we
disclosed the synthesis of guanacastepene N and the first
total synthesis of guanacastepene O by a late-stage oxida-
tion of the lactone-scaffold using MnACHTUNRTGENUNG(OAc)3 and OsO4, re-
spectively. Salient feature of the synthesis presented herein
is its remarkable chemo- and stereoselectivity, evidenced by
the use of one protecting group and the high substrate-in-
duced diastereomeric ratios observed throughout the syn-
thetic sequence.
Acknowledgements
This research was funded by an ETH grant (2-77234-08) and the Swiss
National Science Foundation (200020-119838). C.M.G. acknowledges the
Fonds der Chemischen Industrie for a Kekulꢅ scholarship. We are grate-
ful to Dr. W. B. Schweizer for X-ray crystallographic analyses.
[1] a) C. A. Arias, B. E. Murray, New Engl. J. Med. 2009, 360, 439–443;
b) M. L. Cohen, Nature 2000, 406, 762–767.
[2] a) D. J. Newman, G. M. Cragg, K. M. Snader, J. Nat. Prod. 2003, 66,
1022–1037; b) M. S. Butler, J. Nat. Prod. 2004, 67, 2141–2153;
c) D. J. Newman, G. M. Cragg, J. Nat. Prod. 2007, 70, 461–477.
[3] a) D. L. Hawksworth, Mycol. Res. 1991, 95, 641–655; b) H. Yu, L.
Zhang, L. Li, C. Zheng, L. Guo, W. Li, P. Sun, L. Qin, Microbiol.
Res. 2010, 165, 437–449.
[9] a) Y. Horiguchi, S. Matsuzawa, E. Nakamura, I. Kuwajima, Tetrahe-
dron Lett. 1986, 27, 4025–4028; b) S. Matsuzawa, Y. Horiguchi, E.
Nakamura, I. Kuwajima, Tetrahedron 1989, 45, 349–362; c) H.
Sakata, Y. Aoki, I. Kuwajima, Tetrahedron Lett. 1990, 31, 1161–
1164.
[10] T. Mukhopadhyay, D. Seebach, Helv. Chim. Acta 1982, 65, 385–391.
[11] For a review on the Friedel–Crafts acylation of alkenes, see: J. K.
Groves, Chem. Soc. Rev. 1972, 1, 73–97.
[4] a) S. F. Brady, M. P. Singh, J. E. Janso, J. Clardy, J. Am. Chem. Soc.
2000, 122, 2116–2117; b) S. F. Brady, S. M. Bondi, J. Clardy, J. Am.
Chem. Soc. 2001, 123, 9900–9901; c) M. P. Singh, J. E. Janso, S. W.
Chem. Eur. J. 2012, 18, 15761 – 15771
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15769

E. M. Carreira and C. M. Gampe
[12] I a different scenario, isomerization of 10 to 19 could also occur via
[1,5]-H shift in the enol tautomer of 10.
717–718; c) W. Grimme, H. G. Kçser, J. Am. Chem. Soc. 1981, 103,
5919–5920.
[40] a) M. Nitta, H. Miyano, T. Kobayashi, Heterocycles 1986, 24, 77–85;
b) S. Araki, E. Bonfantini, P. Vogel, Helv. Chim. Acta 1988, 71,
1354–1366.
[13] B. M. Trost, T. A. Grese, D. M. T. Chan, J. Am. Chem. Soc. 1991,
113, 7350–7362.
[14] Guanacastepene labelling is used.
[41] S. V. Ley, C. M. R. Low, A. D. White, J. Organomet. Chem. 1986,
302, C13–C16.
[15] B. H. Lipshutz, M. Koerner, D. A. Parker, Tetrahedron Lett. 1987,
28, 945–948.
[42] a) H.-J. Knçlker, G. Baum, N. Foitzik, H. Goesmann, T. Gonser,
P. G. Jones, H. Rçttele, Eur. J. Inorg. Chem. 1998, 993–1007; b) H.-J.
Knçlker, B. Ahrens, P. Gonser, M. Heiniger, P. G. Jones, Tetrahedron
2000, 56, 2259–2271.
[16] a) R. T. Hansen, D. B. Carr, J. Schwartz, J. Am. Chem. Soc. 1978,
100, 2244–2245; b) J. Schwartz, D. B. Carr, R. T. Hansen, F. M.
Dayrit, J. Org. Chem. 1980, 45, 3053–3061.
[17] S. Kim, J. H. Park, Synlett 1995, 163–164.
[43] a) M. Ochiai, S. Iwaki, T. Ukita, Y. Nagao, Chem. Lett. 1987, 133–
134; b) D. Le Goanvic, M. A. Tius, J. Org. Chem. 2006, 71, 7800–
7804.
[44] T. Nishimura, K. Ohe, S. Uemura, J. Am. Chem. Soc. 1999, 121,
2645–2646.
[18] M. Eriksson, T. Iliefski, M. Nilsson, T. Olsson, J. Org. Chem. 1997,
62, 182–187.
[19] W. F. Bailey, J. D. Brubaker, K. P. Jordan, J. Organomet. Chem.
2003, 681, 210–214.
[20] T. Veysoglu, L. A. Mitscher, J. K. Swayze, Synthesis 1980, 807–810.
[21] Y.-L. Zhong, T. K. M. Shing, J. Org. Chem. 1997, 62, 2622–2624.
[22] K. Bowden, I. M. Heilbron, E. R. H. Jones, B. C. L. Weedon, J.
Chem. Soc. 1946, 39–45.
ˇ
ˇ
´
´
´
´
[45] a) M. Lj. Mihailovic, M. Jakovljevic, Z. Cekovic, Tetrahedron 1969,
ˇ ´
´
25, 2269–2273; b) M. Lj. Mihailovic, L. Lorenc, M. Gasic, M. Rogic,
´
A. Melera, M. Stafanovic, Tetrahedron 1966, 22, 2345–2358.
[46] CCDC 808456 and -808453 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[47] W. Nagata, M. Yoshioka, S. Hirai, J. Am. Chem. Soc. 1972, 94,
4635–4643.
[48] CCDC 808452 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[49] Y. Yamamoto, S. Yamamoto, H. Yatagai, Y. Ishihara, K. Maruyama,
J. Org. Chem. 1982, 47, 119–126.
[50] a) C. Petrier, J. C. De Souza Barbosa, C. Dupuy, J.-L. Luche, J. Org.
Chem. 1985, 50, 5761–5765; b) A. E. Greene, J.-P. Lansard, J.-L.
Luche, C. Petrier, J. Org. Chem. 1984, 49, 931–932.
[51] a) E. C. Ashby, G. Heinsohn, J. Org. Chem. 1974, 39, 3297–3299;
b) S. Flemming, J. Kabbara, K. Nickisch, H. Neh, J. Westermann,
Synthesis 1995, 317–320.
[52] a) E. J. Corey, M. Chaykovsky, J. Am. Chem. Soc. 1965, 87, 1353–
1364; for a reviews, see: b) A.-H. Li, L.-X. Dai, V. K. Aggarwal,
Chem. Rev. 1997, 97, 2341–2372; c) Y. G. Gololobov, A. N. Nes-
meyano, V. P. Lysenko, I. E. Boldeskul, Tetrahedron 1987, 43, 2609–
2651.
[53] W. G. Dauben, E. J. Deviny, Org. Chem. 1966, 31, 3794–3798.
[54] S. Kim, K. H. Ahn, J. Org. Chem. 1984, 49, 1717–1724.
[55] CCDC 785390 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[56] R. Hara, T. Furukawa, H. Kashima, H. Kusama, Y. Horiguchi, I. Ku-
wajima, J. Am. Chem. Soc. 1999, 121, 3072–3082.
[57] For a comprehensive overview of methods to oxidized adjacent to
C=O bonds, see: D. R. Buckle, I. L. Pinto in Comprehensive Organic
Synthesis Volume 7 (Eds.: B. M. Trost, I. Fleming), Pergamon Press,
Oxford, 1991, pp. 119–149.
[58] CCDC 808454 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[59] For a review of acyl anion equivalents, see: R. A. Aitken, A. W.
Thomas, Adv. Heterocycl. Chem. 2001, 79, 89–114.
[23] CCDC 808455 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[24] S. L. Huang, D. Swern, J. Org. Chem. 1978, 43, 4537–4538; b) S. L.
Huang, D. Swern, Tetrahedron 1978, 34, 1651–1660.
[25] CCDC 808451 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[26] P. R. Khoury, J. D. Goddard, W. Tam, Tetrahedron 2004, 60, 8103–
8112.
[27] Recent theoretical calculations revealed an additional, non-electro-
cyclic ring opening pathway, see: C. A. Sader, K. N. Houk, J. Org.
Chem. 2012, 77, 4939–4948.
[28] a) Z. Benko, B. Fraser-Reid, J. Org. Chem. 1988, 53, 2066–2072;
b) T. Yamashita, J. Org. Chem. 1996, 61, 6438–6441; for a review,
see: c) G. S. C. Srikanth, S. L. Castle, Tetrahedron 2005, 61, 10377–
10441.
[29] For a comprehensive review on [3,3]-sigmatropic rearrangements,
see: A. M. Martꢆn Castro, Chem. Rev. 2004, 104, 2939–3002.
[30] a) H. L. Goering, C. C. Tseng, J. Org. Chem. 1983, 48, 3986–3990;
b) H. L. Goering, S. S. Kanter, C. C. Tseng, J. Org. Chem. 1983, 48,
715–721.
[31] a) J. Furukawa, N. Kawabata, J. Nishimura, Tetrahedron Lett. 1966,
7, 3353–3354; b) J. Furukawa, N. Kawabata, J. Nishimura, Tetrahe-
dron 1968, 24, 53–58.
[32] a) Z. Yang, J. C. Lorenz, Y. Shi, Tetrahedron Lett. 1998, 39, 8621–
8624; b) J. C. Lorenz, J. Long, Z. Yang, S. Xue, Y. Xie, Y. Shi, J.
Org. Chem. 2004, 69, 327–334.
[33] W. von E. Doering, P. LaFlamme, J. Am. Chem. Soc. 1956, 78,
5447–5448.
[34] a) T. Aoyama, Y. Iwamoto, S. Nishigaki, T. Shioiri, Chem. Pharm.
Bull. 1989, 37, 253–256; b) E. Vogel, W. Wiedmann, H. Kiefer, W. F.
Harrison, Tetrahedron Lett. 1963, 4, 673–678; c) R. E. Pincock, J. I.
Wells, J. Org. Chem. 1964, 29, 965–967.
[35] a) N. G. Rondan, K. N. Houk, J. Am. Chem. Soc. 1985, 107, 2099–
2111, and ref. therein; b) B. K. Carpenter, Tetrahedron 1978, 34,
1877–1884; c) S. Niwayama, E. A. Kallel, D. C. Spellmeyer, C. Sheu,
K. N. Houk, J. Org. Chem. 1996, 61, 2813–2825.
[36] R. B. Woodward, R. Hoffmann, Angew. Chem. 1969, 81, 797–869;
Angew. Chem. Int. Ed. Engl. 1969, 8, 781–853.
[60] S. Ko, Y. Na, S. Chang, J. Am. Chem. Soc. 2002, 124, 750–751.
[61] E. Kaspar, R. Wiechert, Chem. Ber. 1958, 91, 2664–2670.
[62] J. Hooz, R. B. Layton, J. Am. Chem. Soc. 1971, 93, 7320–7322.
[63] P. A. Jacobi, S. Murphree, F. Ruprecht, W. Zheng, J. Org. Chem.
1996, 61, 2413–2427.
[64] F. P. Ballistreri, S. Failla, E. Spina, G. A. Tomaselli, J. Org. Chem.
1989, 54, 947–949.
[65] R. M. Moriarty, R. Penmasta, A. K. Awasthi, I. Prakash, J. Org.
Chem. 1988, 53, 6124–6125.
[37] For a review, see: W. J. Leigh, Chem. Rev. 1993, 93, 487–505.
[38] a) K. C. Bishop III, Chem. Rev. 1976, 76, 461–486; b) F. D. Mango,
Coord. Chem. Rev. 1975, 15, 109–205; c) R. Pettit, H. Sugahara, J.
Wristers, W. Merk, Discuss. Faraday Soc. 1969, 47, 71–78.
[39] a) W. Slegeir, R. Case, J. S. McKennis, R. Pettit, J. Am. Chem. Soc.
1974, 96, 287–288; see also: b) W. Grimme, E. Schneider, Angew.
Chem. 1977, 89, 754–756; Angew. Chem. Int. Ed. Engl. 1977, 16,
15770
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15761 – 15771

Divergent Synthesis of Guanacastepenes
FULL PAPER
[66] P. H. J. Carlsen, T. Katsuki, V. S. Martin, K. B. Sharpless, J. Org.
Chem. 1981, 46, 3936–3938.
[69] J. K. Cha, W. J. Christ, Y. Kishi, Tetrahedron Lett. 1983, 24, 3943–
3946.
[67] a) D. Yang, F. Chen, Z.-M. Dong, D.-W. Zhang, J. Org. Chem. 2004,
69, 2221–2223; b) X.-W. Chang, D.-W. Zhang, F. Chen, Z.-M. Dong,
D. Yang, Synlett 2009, 3159–3162.
[68] We favor a radical mechanism over an ionic reaction pathway due
to observation of an extended (5–12 h) initiation period, in which no
reaction can be observed on TLC. After this initiation, the reaction
is complete in about 30 min. See the Supporting Information for de-
tails.
[70] N. K. Dunlap, M. R. Sabol, D. S. Watt, Tetrahedron Lett. 1984, 25,
5839–5842.
[71] E. Nakamura, T. Inubushi, S. Aoki, D. Machii, J. Am. Chem. Soc.
1991, 113, 8980–8982.
[72] P. Caramella, N. G. Rondan, M. N. Paddon-Row, K. N. Houk, J. Am.
Chem. Soc. 1981, 103, 2438–2440.
Received: June 22, 2012
Published online: October 18, 2012
Chem. Eur. J. 2012, 18, 15761 – 15771
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15771