Ga(III)-catalyzed cycloisomerization strategy for the synthesis of icetexane diterpenoids: Total synthesis of (±)-salviasperanol

Article abstract of DOI:10.1021/ol061037c

A general approach to the tricyclic core of the icetexane natural products via the cycloisomerization of alkynyl indenes using GaCl₃ is presented. This strategy provides an efficient synthesis of the natural product salviasperanol and sets the

Full text of DOI:10.1021/ol061037c

ORGANIC
LETTERS
2006
Vol. 8, No. 13
2883-2886
Ga(III)-Catalyzed Cycloisomerization
Strategy for the Synthesis of Icetexane
Diterpenoids: Total Synthesis of
(±)-Salviasperanol
Eric M. Simmons and Richmond Sarpong*
Department of Chemistry, UniVersity of California, Berkeley, California 94720
rsarpong@berkeley.edu
Received April 28, 2006
ABSTRACT
A general approach to the tricyclic core of the icetexane natural products via the cycloisomerization of alkynyl indenes using GaCl3 is presented.
This strategy provides an efficient synthesis of the natural product salviasperanol and sets the stage for access to other members of this
family of diterpenoids.
The icetexane diterpenoids encompass a variety of bioactive
and structurally interesting compounds (Scheme 1). For
trypanocidal activity against Trypanosoma cruzi, the parasite
responsible for Chagas’ disease, which is prevalent in Central
and South America.2 Although the biological activity of
salviasperanol (1), isolated from the roots of SalVia aspera,
has not been studied in detail, a number of related compounds
derived from this genus show significant activity against
Mycobacterium tuberculosis, the causative agent of tuber-
culosis.3
Scheme 1. Unified Approach to the Icetexanes
Because of their interesting bioactivity and structure, these
compounds have begun to garner attention from the synthetic
community, which has recently resulted in one total synthesis
of komaroviquinone (2).4
As part of our ongoing interest in the synthesis of seven-
membered ring-containing natural products, we envisioned
a general and modular entry to this class of compounds via
a common cycloheptadiene intermediate (4), which could
(2) (a) Uchiyama, N.; Kabututu, Z.; Kubata, B. K.; Kiuchi, F.; Ito, M.;
Nakajima-Shimada, J.; Aoki, T.; Ohkubo, K.; Fukuzumi, S.; Martin, S. K.;
Honda, G.; Urade, Y. Antimicrob. Agents Chemother. 2005, 49, 5123-
5126. (b) Bastien, J. W. The Kiss of Death, Chagas’ Disease in the
Americas; The University of Utah Press: Salt Lake City, 1998.
(3) Esquivel, B.; Flores, M.; Hernandez-Ortega, S.; Toscano, R. A.;
Ramamoorthy, T. P. Phytochemistry 1995, 39, 139-143.
(4) (a) Sengupta, S.; Drew, M. G. B.; Mukhopadhyay, R.; Achari, B.;
Banerjee, A. K. J. Org. Chem. 2005, 70, 7694-7700. For a synthetic
approach, see: (b) Padwa, A.; Boonsombat, J.; Rashatasakhon, P.; Willis,
J. Org. Lett. 2005, 7, 3725-3727.
example, komaroviquinone (2) and komarovispirone (3) were
isolated from the Uzbekistani perennial semishrub Draco-
cephalum komaroVi.1 These compounds have shown in vitro
(1) (a) Uchiyama, N.; Kiuchi, F.; Ito, M.; Honda, G.; Takeda, Y.;
Khodzhimatov, O. K.; Ashurmetov, O. A. J. Nat. Prod. 2003, 66, 128-
131. (b) Uchiyama, N.; Ito, M.; Kiuchi, F.; Honda, G.; Takeda, Y.;
Khodzhimatov, O. K.; Ashurmetov, O. A. Tetrahedron Lett. 2004, 45, 531-
533.
10.1021/ol061037c CCC: $33.50
© 2006 American Chemical Society
Published on Web 06/02/2006

in turn arise from alkynyl indene 5. Salviasperanol was
chosen as an initial synthetic target with the expectation that
it could serve as a precursor to komaroviquinone (2) via
sequential diastereoselective hydrogenation and oxygenation.
Komarovispirone (3) may in turn be synthetically or bio-
genetically derived from komaroviquinone via a formal
ring contraction rearrangement as has been proposed by
Uchiyama et al.1b
To the best of our knowledge, there are no reports of enyne
cycloisomerizations involving indenes to produce cyclohep-
tadienes. As a result, our initial investigations employed the
simple model 7,5 with which we screened a variety of
conditions known to catalyze enyne cycloisomerizations.
Several metal complex/additive combinations, including
the Grubbs I and Grubbs II alkylidenes,6 Pt(PPh3)2Cl2/PhIO,7
[Rh(CO)2Cl]2, Rh(PPh3)3Cl, Rh(PPh3)3Cl/AgBF4, and [Ru-
(CO)3Cl2]2/AgBF4, gave either no reaction or only trace
conversion. Other complexes (Table 1), including PtCl2,8
Chatani and Murai to effect a skeletal reorganization of a
variety of enynes under very mild conditions.12 Gratifyingly,
upon exposure of 7 to GaCl3 (10 mol %) for 1 h at 23 °C
(entry 8), cycloheptadiene 8 was obtained as the sole product,
with no detectable formation of 9.
With conditions for the key conversion of indenyl alkynes
to cycloheptadienes defined using 7 as a model, we embarked
on a synthesis of the more complex substrate 5. Our synthetic
efforts commenced with the preparation of indanone 6 as
outlined in Scheme 2. Formylation of isopropyl veratrole
Scheme 2. Synthesis of the Indene Precursor 5
Table 1. Screen of Cycloisomerization Complexes
entry
catalyst
temp (°C) concn (M)a ratio (8/9)b
(10)13 followed by Wittig reaction of the resulting aldehyde
with the stabilized carbethoxymethylidene triphenylphos-
phorane ylide provided enoate 11. Hydrogenation with
Adam’s catalyst followed by saponification of the ethyl ester
then afforded acid 12 in excellent yield over the two steps.
At this stage, Friedel-Crafts acylation of the corresponding
acid chloride gave indanone 6 in good yield.
1
2
3
4
5
6
7
8d
PtCl2
PtCl2
PtCl2
PtCl2
PtCl4
[Ru(CO)2Cl2]2
[Rh(CO)2Cl]2/AgBF4
GaCl3
50
80
80
80
50
80
23
23
0.05
0.14
0.05
0.025
0.05
0.05
0.05
0.05
no reaction
0.8:1
1.1:1
2:1c
0.6:1
2.5:1
3:1
1:0
a Concentration of 7. b Ratios based on integration of 1H NMR signals.
c Reaction was incomplete after 4 h; the ratio is based on converted material.
d Reaction was judged complete (by TLC) after 1 h.
Direct alkylation of indanone 6 was pursued using iodide
13, which is readily available from the corresponding
alcohol.14 However, this was complicated by competitive
overalkylation, which was obviated by an initial Claisen
reaction with Mander’s reagent to install a carbomethoxy
group followed by alkylation to afford â-ketoester 14.
Saponification of the methyl ester proceeded with subsequent
decarboxylation upon workup. Reduction of the resulting
carbonyl followed by a net dehydration (Ms2O, Et3N) yielded
alkynyl indene 5 as a single alkene regioisomer in good yield.
With the fully functionalized indene 5 in hand, our attention
turned to the viability of the formal enyne metathesis reaction
on this more elaborate substrate.
which has been extensively exploited by Fu¨rstner for related
purposes,9 did promote the desired cycloisomerization (en-
tries 2-7) but gave a mixture of the cycloheptadiene product
810 along with the structural isomer 9 as an inseparable
mixture.11
Following this screen of transition-metal complexes, we
turned our attention to GaCl3, which has been shown by
(5) For full synthesis details of 7, see the Supporting Information.
(6) For a review on enyne metathesis using the Grubbs ruthenium
alkylidene complexes, see: Diver, S. T.; Giessert, A. J. Chem. ReV. 2004,
104, 1317-1382.
(7) Bhanu Prasad, B. A.; Yoshimoto, F. K.; Sarpong, R. J. Am. Chem.
Soc. 2005, 127, 12468-12469.
(8) Chatani, N.; Furukawa, N.; Sakurai, H.; Murai, S. Organometallics
1996, 15, 901-903.
(9) For applications in natural product synthesis, see: (a) Fu¨rstner, A.;
Szillat, H.; Gabor, B.; Mynott, R. J. Am. Chem. Soc. 1998, 120, 8305-
8314. (b) Fu¨rstner, A.; Szillat, H.; Stelzer, F. J. Am. Chem. Soc. 2000, 122,
6785-6786. (c) For a seminal report, see: Chatani, N.; Kataoka, K.; Murai,
S.; Furukawa, N.; Seki, Y. J. Am. Chem. Soc. 1998, 120, 9104-9105.
(10) Tricycle 8 has been reported previously; see: Paquette, L. A.;
Chamot, E.; Browne, A. R. J. Am. Chem. Soc. 1980, 102, 637-643.
(11) The reason for the observed dependency of product ratios on
concentration (entries 2-4) is under active investigation.
Indene 5 was found to react slowly under the initially
established cycloisomerization conditions. Our studies have
revealed that the sluggish rate of reactivity of 5 is attributable
to the steric congestion of the gem-dimethyl substitution
adjacent to the alkyne.15 Substrates devoid of an adjacent
(12) Chatani, N.; Inoue, H.; Kotsuma, T.; Murai, S. J. Am. Chem. Soc.
2002, 124, 10294-10295.
(13) Majetich, G. A.; Liu, S. Synth. Commun. 1993, 23, 2331-2335.
(14) Malezcka, R. E.; Gallagher, W. P. Org. Lett. 2001, 3, 4173-4176.
For preparation of 13, see the Supporting Information.
(15) For related prior observations, see: Trost, B. M.; Chang, V. K.
Synthesis 1993, 824-832.
2884
Org. Lett., Vol. 8, No. 13, 2006

quaternary carbon center (e.g., 7) readily isomerized in high
yield to the corresponding cycloheptadienes. However,
increasing the reaction temperature beyond 50 °C in an
attempt to accelerate the rate of reaction of 5 led to significant
formation of 15, which presumably arises from isomerization
of the indene double bond prior to cycloisomerization.
After a screen of various additives,16 an optimal set of
conditions was identified (20 mol % of GaCl3, 4 Å MS, 0.04
M in PhH, 40 °C, 24 h) that converted 5 to cycloheptadiene
4 in excellent yield (Scheme 3).17
Alternatively, reversible isomerization of 16 via cyclo-
propanes 18 and 19 may culminate in the formation of
cycloheptadiene 8.20 Pathway B also begins with the common
nonclassical zwitterion 16, which may interconvert with 20,
which contains a spiro quaternary carbon center and a
benzylic carbocation. Attack of the alkene onto the benzylic
carbocation would lead to zwitterion 21, which may be in
equilibrium with cyclobutene 22.21 At this stage, formal retro
4π electrocyclization of 22 should yield cycloheptadiene 8.22
Although pathway A clearly accounts for the formation of
products such as 9, the genesis of the cycloheptadiene
products (i.e., 8) is less obvious. The isolation of cyclobutene
intermediates in the GaCl3-catalyzed enyne cycloisomeriza-
tions reported by Murai and Chatani gives strong support to
pathway B for the formation of cycloheptadienes such as
8.23,24 Furthermore, our preliminary studies have shown an
increase in the rate of formation of cycloheptadiene products
(e.g., 8) with increasing electron density on the aromatic core
(by substitution of electron-donating alkoxy groups for the
phenyl hydrogens).25 This observation is most readily
explained by increased stabilization of the benzylic carbo-
cation intermediate 20 (pathway B) relative to 17 (pathway
A), lending further support to pathway B for the formation
of the cycloheptadiene products. Additionally, a related
precedent by Trost26 and Fu¨rstner9b suggests pathway B may
be the primary reaction path and a conversion of 20 to 18
may provide the pathway to the observed byproduct 9. This
interconversion may result because canonical nonclassical
resonance structures such as 16 and intermediates such as
18 may experience added stabilization from transition
metals.27 This characteristic most likely dictates the superior
selectivity in the gallium-catalyzed cycloisomerization reac-
tions (where intermediates such as 18 are less stabilized) as
compared to the transition-metal-mediated processes.
Scheme 3. Cycloisomerization of 5
Mechanistically, the cycloisomerization of alkynyl indene
substrates such as 7 (eq 1) may proceed via two alternate
pathways (Scheme 4).18 These pathways are distinguished
Scheme 4. Potential Cycloisomerization Mechanisms
(19) Nonclassical intermediates such as 16 have been previously
proposed; see: refs 9b and 12.
(20) For an initial proposal of cyclopropylmetallocarbenoid intermediates,
see ref 9c.
(21) Zwitterion intermediate 21 may also lead to product; see: Fu¨rstner,
A.; Davies, P. W.; Gress, T. J. Am. Chem. Soc. 2005, 127, 8244-8245.
(22) A direct thermal 4π electrocyclic ring opening of 22, which should
proceed via conrotation, is unlikely as a concerted process because of the
constraints of the trans-alkene geometry that would result in the formation
of the cycloheptadiene product. See: Baldwin, J. E.; Gallagher, S. S.; Leber,
P. A.; Raghavan, A. Org. Lett. 2004, 6, 1457-1460.
(23) For example, cycloisomerization of i yielded cyclobutene ii. See
ref 12.
by the intermediate zwitterions (17 and 18 or 20) that may
arise from an initially formed nonclassical zwitterion inter-
mediate, 16.19 In pathway A, conversion of 16 to 17 followed
by proton transfer and isomerization yields tricycle 9.
(24) Cyclobutenes have also been observed as intermediates in other
enyne cycloisomerization reactions. See, for example, ref 15.
(25) Example substrates such as iii displayed an accelerated rate of
reaction (i.e., shorter times to reach completion) compared to 7.
(16) Additives were explored to curtail catalyst decomposition due to
adventitious water and resulting acid-catalyzed product decomposition.
(17) Interestingly, GaI3, GaBr3, and InCl3 are also effective albeit at
higher temperatures (60-80 °C) and provide 4 along with 15. On the other
hand, Ga(OTf)3 is completely ineffective. For recent examples of In(III)-
catalyzed enyne cycloisomerizations, see: Miyanohana, Y.; Chatani, N. Org.
Lett. 2006, 8, 2155-2158.
(18) For a recent discussion of possible enyne cycloisomerization
pathways of transition and main group metal complexes, see: Bruneau, C.
Angew. Chem., Int. Ed. 2005, 44, 2328-2334.
(26) Trost, B. M.; Tanoury, G. J. J. Am. Chem. Soc. 1988, 110, 1636-
1638.
(27) We would like to thank a referee for this suggestion.
Org. Lett., Vol. 8, No. 13, 2006
2885

With tricycle 4 in hand, several methods were then
investigated for the oxygenation of the cycloheptadiene
moiety. In the end, the most efficient strategy entailed
selective epoxidation to obtain 23 (Scheme 5), which, upon
In summary, we report the first total synthesis of the
icetexane diterpenoid salviasperanol via a unique synthetic
strategy, which employs a cycloisomerization of alkynyl
indenes to access a key cycloheptadiene intermediate. The
modular and efficient synthesis of the icetexane core reported
herein provides a platform for the syntheses of the related
natural products komaroviquinone and komarovispirone.
Furthermore, a late-stage epoxidation to access salviasperanol
presents the possibility of an enantioselective synthesis of
this class of natural products, which is currently under
investigation in our laboratory.
Scheme 5. Completion of Salviasperanol
Acknowledgment. The authors are grateful to UC
Berkeley, GlaxoSmithKline, and the donors of the ACS-
Petroleum Research Fund (ACS-PRF No. 43546-G1) for
generous financial support.
treatment with catalytic trifluoroacetic acid, isomerized to
dihydrofuran 24 in high yield. Conversion of dimethylsalvi-
asperanol (24) to salviasperanol (1) was achieved by bis-
methyl ether cleavage using sodium ethanethiolate. Spectral
data for synthetic salviasperanol (1H, 13C NMR, IR, and MS)
were identical to those reported for the natural sample.
Supporting Information Available: Experimental details
and characterization data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL061037C
2886
Org. Lett., Vol. 8, No. 13, 2006