Total synthesis of (-)-pseudolaric acid B

Article abstract of DOI:10.1021/ja806724x

We report a full account of our work toward the total synthesis of pseudolaric acid B (1a), a diterpene acid isolated from the bark of Pseudolarix kaempferi Gordon (pinaceae). Compound 1a is an antifungal and antifertility agent. Furthermore, its capacity

Full text of DOI:10.1021/ja806724x

Published on Web 11/08/2008
Total Synthesis of (-)-Pseudolaric Acid B
Barry M. Trost,* Jerome Waser, and Arndt Meyer
Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
Received August 28, 2008; E-mail: bmtrost@stanford.edu
Abstract: We report a full account of our work toward the total synthesis of pseudolaric acid B (1a), a
diterpene acid isolated from the bark of Pseudolarix kaempferi Gordon (pinaceae). Compound 1a is an
antifungal and antifertility agent. Furthermore, its capacity for inhibiting tubulin polymerization makes it a
potential lead for cancer therapy. Herein, we describe the use of a Ru- or Rh-catalyzed [5 + 2] intramolecular
cycloaddition reaction of an alkyne and a vinylcyclopropane for the construction of the polyhydroazulene
core of the molecule. Our first unsuccessful strategy for the introduction of the quaternary center based on
an epoxide opening with cyanide led to the discovery of a new TBAF-mediated isomerization of a 1,4-
diene to a 1,3-diene and a vinylogous eliminative opening of an epoxide to form a dienol. Our second
strategy, based on the cyclization of an alkoxycarbonyl radical upon a diene system, succeeded in forming
the quaternary center. Detailed studies showed the dependence of this underutilized approach for the
synthesis of lactones on substrate structure and reaction conditions. In the late stage of the synthesis, the
unique capacity of cerium organometallic reagents to add to a sensitive, sterically hindered ketone was
demonstrated. The easy formation of an oxo-bridged derivative was the major hurdle to the completion of
the synthesis and showcased the intriguing reactivity of the complex core of the pseudolaric acids.
pathways in human melanoma cells.4c-g Finally, 1a inhibits the
Introduction
polymerization of tubulin in multidrug-resistant cancer cell lines
Pseudolaric acids A-C (Figure 1) are diterpene acids isolated
from the bark of Pseudolarix kaempferi Gordon (pinaceae).1
The extract of the root bark of P. kaempferi is a Chinese herbal
medicine called tujinpi that is used against fungal infections of
the skin and nails.1 Pseudolaric acid B (1a) has been identified
as a potent antifungal,2a antifertility,2b,c and cytotoxic agent1d,2d
and displays much higher activity than pseudolaric acid A (1b)
and C (1c). More recently, several other members of the
pseudolaric acid family have been isolated, but none displayed
significant activity.3 Pseudolaric acid B has been shown to be
an agonist for transcriptional activation of peroxisome prolif-
erator-activated receptors (PPARs), which are important for the
membrane integrity and lipid metabolism of Candida species.4b
Furthermore, 1a inhibits angiogenesis by diminishing the
secretion of vascular endothelial growth factor (VEGF) in tumor
cells4a,b and has been shown to induce apoptosis through several
and thus constitutes a potential lead for new cancer therapy.4h,i
Because of these remarkable biological properties and the low
toxicities of these compounds, the pseudolaric acids have
attracted significant interest among the scientific community.
The compact tricyclic core of 1a constitutes a challenge for
modern synthetic chemistry. The four contiguous stereocenters,
one of them quaternary (C10), and the unusual trans-substituted
fused [5-7] ring system (polyhydroazulene) contribute to
making any stereoselective approach toward the pseudolaric
acids a daunting task. The many unsuccessful attempts toward
(4) (a) Li, M. H.; Miao, Z. H.; Tan, W. F.; Yue, J. M.; Chao, Z.; Lin,
L. P.; Zhang, X. W.; Jian, D. Clin. Cancer Res. 2004, 10, 8266. (b)
Tan, W. F.; Zhang, X. W.; Li, M. H.; Yue, J. M.; Chen, Y.; Lin,
L. P.; Ding, J. Eur. J. Pharmacol. 2004, 499, 219. (c) Gong, X. F.;
Wang, M. W.; Tashiro, S.; Onodera, S.; Ikejima, T. Arch. Pharm.
Res. 2005, 28, 68. (d) Gong, X. F.; Wang, M. W.; Wu, Z.; Tashiro,
S.; Onodera, S.; Ikejima, T. Exp. Mol. Med. 2004, 36, 551. (e) Gong,
X. F.; Wang, M. W.; Tashiro, S. I.; Onodera, S.; Ikejima, T. Exp.
Mol. Med. 2006, 38, 428. (f) Wu, W. Y.; Guo, H. Z.; Qu, G. Q.; Han,
J.; Guo, D. A. Am. J. Chin. Med. 2006, 34, 887. (g) Ko, J. K. S.;
Leung, W. C.; Ho, W. K.; Chiu, P. Eur. J. Pharmacol. 2007, 559, 1.
(h) Wong, V. K. W.; Chiu, P.; Chung, S. S. M.; Chow, L. M. C.;
Zhao, Y. Z.; Yang, B. B.; Ko, B. C. B. Clin. Cancer Res. 2005, 11,
6002. (i) Liu, B.; Chen, H.; Lei, Z. Y.; Yu, P. F.; Xiong, B. J. Asian
Nat. Prod. Res. 2006, 8, 241.
(1) (a) Li, Z. L.; Pan, D. J.; Hu, C. Q.; Wu, Q. L.; Yang, S. S.; Xu, G. Y.
Acta Chim. Sin. 1982, 40, 447. (b) Li, Z. L.; Pan, D. J.; Wu, Q. L.;
Xu, G. Y. Acta Chim. Sin. 1982, 40, 757. (c) Zhou, B. N.; Ying, B. P.;
Song, G. Q.; Chen, Z. X.; Han, J.; Yan, Y. F. Planta Med. 1983, 47,
35. (d) Hamburger, M.; Shieh, H.; Pezzuto, J.; Cordell, G. A.; Zhou,
B. N. Planta Med. 1988, 54, 579. (e) Hamburger, M. O.; Shieh, H. L.;
Zhou, B. N.; Pezzuto, J. M.; Cordell, G. A. Magn. Reson. Chem. 1989,
27, 1025.
(2) (a) Li, E. G.; Clark, A. M.; Hufford, C. D. J. Nat. Prod. 1995, 58, 57.
(b) Wang, W. C.; Lu, R. F.; Zhao, S. X.; Gu, Z. P. Acta Pharmacol.
Sin. 1988, 9, 445. (c) Zhang, Y. L.; Lu, R. Z.; Yan, A. L. Acta
Pharmacol. Sin. 1990, 11, 60. (d) Pan, D. J.; Li, Z. L.; Hu, C. Q.;
Chen, K.; Chang, J. J.; Lee, K. H. Planta Med. 1990, 56, 383.
(3) (a) Li, Z. L.; Chen, K.; Pan, D. J.; Xu, G. Y. Acta Chim. Sin. 1989,
47, 258. (b) Yue, J.; Yang, S.; Ding, J.; Xiao, D.; Yuan, S. U.S. Patent
6,887,895, 2002. (c) Yang, S. P.; Dong, L.; Wang, Y.; Wu, Y.; Yue,
J. M. Bioorg. Med. Chem. 2003, 11, 4577. (d) Peng, L.; Guo, H.;
Wang, W.; Zhang, J.; Li, N.; Han, J.; Zhou, N.; Hu, Y.; Zhang, T.;
Liu, Z.; Guo, D. J. Nat. Prod. 2007, 70, 533.
(5) (a) Pan, B. C.; Chang, H. Y.; Cai, G. L.; Guo, Y. S. Pure Appl. Chem.
1989, 61, 389. (b) Higuchi, R. I. Ph.D. Thesis, Stanford University,
Stanford, CA, 1995. (c) Chiu, P.; Chen, B.; Cheng, K. F. Tetrahedron
Lett. 1998, 39, 9229. (d) Hu, Y. H.; Ou, L. G.; Wang, X. L.; Bai,
D. L. Chin. Chem. Lett. 1999, 10, 281. (e) Jiang, X. T.; Ou, L. G.;
Han, D. M.; Zhai, Y. F.; Bai, D. L. Chin. Chem. Lett. 2001, 12, 113.
(f) Chiu, P.; Chen, B.; Cheng, K. F. Org. Lett. 2001, 3, 1721. (g)
Chen, B.; Ko, R. Y. Y.; Yuen, M. S. M.; Cheng, K. F.; Chiu, P. J.
Org. Chem. 2003, 68, 4195. (h) Wu, B. G.; Karle, J. M.; Watkins,
E. B.; Avery, M. A. Tetrahedron Lett. 2002, 43, 4095. (i) Geng, Z.;
Chen, B.; Chiu, P. Angew. Chem., Int. Ed. 2006, 45, 6197.
9
16424 J. AM. CHEM. SOC. 2008, 130, 16424–16434
10.1021/ja806724x CCC: $40.75
2008 American Chemical Society

Total Synthesis of (-)-Pseudolaric Acid B
A R T I C L E S
substitution at C10, probably because of the high steric
hindrance. Consequently, it was necessary to introduce the
quaternary center after the [5 + 2] cycloaddition had been
performed. When considering the trans substitution of the
oxygen at C4 and the ester at C10, we considered the possibility
of cyanide opening of a tetrasubstituted epoxide derived from
conjugated diene 5 (Scheme 1B). The thermodynamically more
stable conjugated diene 5 should be accessible from 1,4-diene
6 produced by the Ru-catalyzed [5 + 2] cycloaddition of 7.
[5 + 2] Cycloaddition and Epoxide-Opening Strategy. We
first concentrated our efforts toward the synthesis of an adequate
precursor 7 for the [5 + 2] cycloaddition. Iodide 11 and
aldehyde 14 were considered to be adequate precursors to access
7 via a homologation-olefination sequence (Scheme 2).
The synthesis of iodide 11 began with the highly selective
Noyori reduction of 2-acetyl butyrolactone (8) to generate the
two adjacent stereocenters (Scheme 2A).9 After TBS protection
of the secondary alcohol 9, the lactone was reduced with
DIBAL-H. The resulting lactol was directly reacted with
TMSCHN2/LDA followed by a TMSCl quench to give the
corresponding protected alkyne 10. For iodination of alcohol
10, a one-step procedure (PPh3/I2) was used in small-scale
syntheses, but a two-step procedure (MsCl then NaI) was more
convenient for larger-scale work. Aldehyde 14 was obtained in
three steps and 84% overall yield starting from cis-butenediol
(12) via monoprotection with TBDPSCl, Charette modification
of the Simmons-Smith cyclopropanation,10 and Moffat-Swern
oxidation (Scheme 2B).
Figure 1. Structure of the Pseudolaric Acids.
a total synthesis of these natural products showcase these
difficulties.5a-h In 2006, the first successful total synthesis of
pseudolaric acid A was reported by Chiu.5i Key to this success
was the development of an efficient carbene cyclization cy-
cloaddition cascade to form the polyhydroazulene core of the
molecule. Despite intensive optimization, the diastereoselectivity
of the cyclization reaction never exceeded 1.6:1. Herein, we
describe a full account of our work,6 which highlights the
challenges and surprises associated with the selective synthesis
of such a densely functionalized polycyclic compound.
Results and Discussion
Retrosynthesis. The polyhydroazulene constituting the core
of the pseudolaric acids can be envisioned to be made by the
metal-catalyzed [5 + 2] intramolecular cycloaddition of a
vinylcyclopropane and an alkyne introduced by the Wender7
(with Rh catalysts) and Trost8 (with Ru catalysts) groups. One
of the major challenges for this approach is the stereoselective
elaboration of the quaternary center at C10.
We intended to introduce the lactone ring and the fatty acid
chain at a late stage of the synthesis in order to focus first on
the challenging formation of the polyhydroazulene. These
disconnections reveal 2 as a potential precursor (Scheme 1A).
In our first approach, we envisaged accessing 2 using a Ru-
catalyzed [5 + 2] cycloaddition reaction of the linear precursor
3 via simultaneous C4-C10 and C5-C6 bond formation. This
strategy would allow formation of the polyhydroazulene and
quaternary center in a single synthetic transformation.
Reaction of iodide 11 with triphenylphosphoniummethylide
provides the homologated phosphonium salt as a nonisolated
intermediate. In situ conversion of this new phosphonium salt
to an ylide using the Schlosser modification of the Wittig
olefination with aldehyde 14 gave, in this one-pot operation,
the desired vinyl cyclopropane 15 in 58% overall yield after
TMS deprotection with good E selectivity (>10:1) (eq 1).11 This
high selectivity was important, as the Z olefins proved to be
unreactive in the [5 + 2] cycloaddition.
In the course of model studies (not shown), it rapidly became
apparent that the Ru catalyst could not accommodate any
(6) Trost, B. M.; Waser, J.; Meyer, A. J. Am. Chem. Soc. 2007, 129, 14556.
(7) (a) Wender, P. A.; Takahashi, H.; Witulski, B. J. Am. Chem. Soc.
1995, 117, 4720. (b) Wender, P. A.; Rieck, H.; Fuji, M. J. Am. Chem.
Soc. 1998, 120, 10976. (c) Wender, P. A.; Sperandio, D. J. Org. Chem.
1998, 63, 4164. (d) Wender, P. A.; Husfeld, C. O.; Langkopf, E.; Love,
J. A.; Pleuss, N. Tetrahedron 1998, 54, 7203. (e) Wender, P. A.;
Husfeld, C. O.; Langkopf, E.; Love, J. A. J. Am. Chem. Soc. 1998,
120, 1940. (f) Wender, P. A.; Dyckman, A. J. Org. Lett. 1999, 1,
2089. (g) Wender, P. A.; Fuji, M.; Husfeld, C. O.; Love, J. A. Org.
Lett. 1999, 1, 137. (h) Wender, P. A.; Dyckman, A. J.; Husfeld, C. O.;
Kadereit, D.; Love, J. A.; Rieck, H. J. Am. Chem. Soc. 1999, 121,
10442. (i) Wender, P. A.; Glorius, F.; Husfeld, C. O.; Langkopf, E.;
Love, J. A. J. Am. Chem. Soc. 1999, 121, 5348. (j) Wender, P. A.;
Zhang, L. Org. Lett. 2000, 2, 2323. (k) Wender, P. A.; Dyckman,
A. J.; Husfeld, C. O.; Scanio, M. J. C. Org. Lett. 2000, 2, 1609. (l)
Wender, P. A.; Gamber, G. G.; Scanio, M. J. C. Angew. Chem., Int.
Ed. 2001, 40, 3895. (m) Wender, P. A.; Barzilay, C. M.; Dyckman,
A. J. J. Am. Chem. Soc. 2001, 123, 179. (n) Wender, P. A.; Pedersen,
T. M.; Scanio, M. J. C. J. Am. Chem. Soc. 2002, 124, 15154. (o)
Wender, P. A.; Williams, T. J. Angew. Chem., Int. Ed. 2002, 41, 4550.
(p) Yu, Z. X.; Wender, P. A.; Houk, K. N. J. Am. Chem. Soc. 2004,
126, 9154.
The cyclization of 15 using [CpRu(CH3CN)3]+PF6- (17) as
the catalyst proceeded only sluggishly, and full conversion was
never observed even with 20 mol % catalyst. In contrast to this
result, substrate 16 bearing a free acetylene was more reactive
and was fully consumed under these reaction conditions,
although a 20 mol % catalyst loading was still required.
(9) Mashima, K.; Kusano, K. H.; Sato, N.; Matsumura, Y.; Nozaki, K.;
Kumobayashi, H.; Sayo, N.; Hori, Y.; Ishizaki, T.; Akutagawa, S.;
Takaya, H. J. Org. Chem. 1994, 59, 3064.
(8) (a) Trost, B. M.; Toste, F. D.; Shen, H. J. Am. Chem. Soc. 2000, 122,
2379. (b) Trost, B. M.; Shen, H. C. Org. Lett. 2000, 2, 2523. (c) Trost,
B. M.; Shen, H. C. Angew. Chem., Int. Ed. 2001, 40, 2313. (d) Trost,
B. M.; Shen, H. C.; Schulz, T.; Koradin, C.; Schirok, H. Org. Lett.
2003, 5, 4149. (e) Trost, B. M.; Shen, H. C.; Horne, D. B.; Toste,
E. D.; Steinmetz, B. G.; Koradin, C. Chem.sEur. J. 2005, 11, 2577.
(10) Charette, A. B.; Juteau, H.; Lebel, H.; Molinaro, C. J. Am. Chem.
Soc. 1998, 120, 11943.
(11) Wang, Q.; Deredas, D.; Huynh, C.; Schlosser, M. Chem.sEur. J. 2003,
9, 570.
9
J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008 16425

A R T I C L E S
Trost et al.
Scheme 1. Retrosyntheses of Pseudolaric Acid B (1a)
Initial experiments in methylene chloride led to the formation
of two major products in a 1:1 ratio. Subsequently, one of them
was identified as the desired [5 + 2] cyclization product 19,
with a diastereoselectivity of 9:1 to 15:1, while the other proved
to be conjugated diene 20 (eq 2).12 The stereochemistry of the
major diastereoisomer 19 was assigned on the basis of the
general selectivity of the Ru-catalyzed cycloaddition reaction
and further confirmed via nuclear Overhauser effect (NOE)
experiments, although the weakness of the observed signals did
not allow a certain assignment. The subsequent behavior of 19
in the synthesis, however, further corroborated this assignment
(see below).
17 was not observed. Consequently, activation of the C-H bond
had to occur from an intermediate in the catalytic cycle. For
Rh catalysts, C-H activation is not so facile. When Wender’s
catalyst [(C8H10)Rh(COD)]+SbF6- (18)7o was used, the desired
product 19 was obtained exclusively in 88% isolated yield and
15:1 diastereoselectivity (18 g scale), although the reaction still
required an unusually high catalyst loading (10 mol %) (eq 2).
We then focused on the isomerization of diene 19 and its
further elaboration. At the beginning, several attempts to
isomerize 19 to a conjugated diene were unsuccessful, as either
no reaction (HCl, THF; PPTS, toluene, 110 °C; CSA, toluene,
110 °C) or decomposition (TFA, 0 °C) was observed under
acidic conditions and decomposition occurred under common
basic conditions (KOtBu, DMSO; KOtBu, THF; LDA, THF).
No reaction of the diene system occurred in the presence of Pd
catalysts ([Pd(CH3CN)2]2+(BF4)-2, CH3CN; Pd/C, H2, benzene).
Finally, a surprising solution to the isomerization problem was
found when attempting the deprotection of the two silyl ether
groups of 19. As the removal of the silyl groups was slow with
TBAF only, activated molecular sieves (3 Å, 0.50 g/mmol in
TBAF) were added. This resulted in a fast deprotection, but
the obtained product was in fact a 1:1 mixture of the expected
product 21 and the conjugated diene 22 (eq 3). It was soon
found that the amount of molecular sieves was crucial in
determining the 21/22 ratio, and using 1.0 g of molecular sieves/
mmol of TBAF led to the exclusive formation of conjugated
diene 22. It is interesting to note that submitting cyclization
product 20 to the same reaction conditions simply resulted in a
clean cleavage of the silyl groups without isomerization, and
the corresponding diastereoisomer of 19 with the hydrogen at
C10 and the hydroxymethyl group at C7 in trans positions gave
a mixture of elimination and simple deprotection products.
Again, no isomerization product was obtained. An explanation
for these observations would be an intramolecular deprotonation
of the bisallylic hydrogen via a “naked” alkoxide generated
during the deprotection (A in eq 3). Such a process would be
possibleonlywithrigorousexclusionofwater.Theisomerization-
deprotection sequence worked well up to a 3.2 mmol (1.8 g)
scale. On larger scales, resubmitting the mixture to KOtBu in
THF was necessary to achieve full conversion. The fact that an
anhydrous base is able to mediate the isomerization of 21 to 22
whereas no reaction was observed with 19 further supports the
proposed mechanism. It is interesting to note that the Ru- and
Fortunately, the 19/20 product ratio could be increased to
3:1 by using acetone as the solvent. In this case, a combined
yield of 63% was obtained. With more polar DMF as the
solvent, dehydration reactions became competitive and triene
side products were dominant.
Following these preliminary results, several additives were
examined in the [5 + 2] cycloaddition reaction. It was
envisioned that an equimolar amount of a coordinating additive
could bind to the Ru center and modulate its selectivity without
hampering reactivity, as two free coordination sites on the Ru
center should be sufficient for an efficient cycloaddition reaction.
However, we were not able to identify any additive that could
further increase the 19/20 ratio. Weak ligands (cyclopentanone,
acetic acid, sulfolane, 2,4-diacetylpyridine, dimethylcarbonate,
phosphate esters) simply showed no effects, and stronger ligands
(pyridine, bipyridine, triphenylphosphine oxide, triphenylphos-
phine sulfide, arylphosphines, DMPU, cyclooctadiene, excess
acetonitrile) led to low turnover numbers, even if only 1 equiv
of ligand with respect to the Ru catalyst was used. A particular
behavior was observed for DMF and DMA, which led to
elimination products, while nitromethane favored the formation
of 20 (19/20 ) 1:1.4).
To explain the formation of 20, we speculated that the
insertion of a Ru intermediate into the labile bisallylic C-H
bond leads to a Ru-pentadienyl complex, which then deacti-
vates the catalyst. In fact, the low turnover observed for the
reaction in the case of Ru catalysts seems to be related to the
formation of 20, as the isolated yield of 20 was in the same
range as the catalyst loading. While not unexpected with Ru
catalysts in general, this behavior was not observed in other
Ru-catalyzed [5 + 2] cycloaddition reactions. Protonation of
the intermediate Ru-pentadienyl complex then results in
conjugated diene 20. It is important to note that in a separate
experiment, conversion of 19 to 20 in the presence of Ru catalyst
(12) The exact structure of 21 was difficult to ascertain because of the
overlap of numerous signals in the NMR spectra. The drawn structure
would be in accordance with the data available to date. Removal of
the silyl protecting group of 20 led to a solid compound, but repeated
attempts to recrystallize this product have not been successful to date.
9
16426 J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008

Total Synthesis of (-)-Pseudolaric Acid B
A R T I C L E S
Scheme 2. Synthesis of Iodide 11 and Aldehyde 14
Scheme 3. Synthesis of Epoxide 23 and its Rearrangement to Ketone 24
Scheme 4. Revised Retrosynthesis
TBAF-mediated isomerizations proceed to give completely
different diene systems because they involve different isomer-
ization mechanisms.
intermediate, was obtained in high yield (Scheme 3).14 Unfor-
tunately, this ring-contraction reaction appeared to be highly
favored with any strong Lewis acid, even at low temperatures.
No conversion was observed when nucleophilic activation of
cyanide or weaker Lewis acids was used. Nevertheless, this
sequence does provide a convenient diastereocontrolled entry
into such spiro compounds, which are also a common motif in
natural products.
Revised Retrosynthesis. At this point, it had become clear
that the epoxide-opening strategy was not viable for introducing
the quaternary center. As intermolecular addition of a nucleo-
phile to form the quaternary center had proven to be very
difficult, the synthesis strategy was modified to include an
intramolecular bond formation (Scheme 4). A radical cyclization
was envisioned to introduce the quaternary center at C10
stereoselectively. An alkoxycarbonyl selenide (X ) O, Y )
SePh in 26) was chosen as a radical precursor, as there were
rare examples of such an intermediate for radical lactonization
reactions in the literature.15 A dienoate (as in 27) was envisioned
as the radical acceptor, which in turn would derive from the
dienol 28. The side chain would be introduced via the reaction
of an organometallic reagent with a ketone at C11. In principle,
side-chain introduction could be envisaged before or after
formation of the quaternary center.
The research next was focused on the functionalization of
diene 22. In order to obtain the desired diastereoselectivity for
the epoxidation of 22, it was necessary to protect the two
hydroxy groups (Scheme 3). The subsequent epoxidation
1
proceeded with good selectivity (10:1 by H NMR), but the
obtained vinyl epoxide 23 proved to be unstable. The best results
were obtained when the crude reaction mixture was directly
filtered over deactivated (NEt3) silica gel and used immediately.
The key formation of the quaternary center was first attempted
via cyanide addition. Et2AlCN appeared to be the reagent of
choice, as it has been reported to open similar 6,6-bicyclic
tetrasubstituted vinyl epoxides with high regio- and diastereo-
selectivity.13 When epoxide 23 was treated with this reagent,
however, the desired product was not observed. Instead, ketone
24, resulting from a pinacol-like rearrangement via a cationic
(14) For similar rearrangements of vinyl epoxides, see: (a) Jung, M. E.;
Damico, D. C. J. Am. Chem. Soc. 1995, 117, 7379. (b) Deng, X. M.;
Sun, X. L.; Tang, Y. J. Org. Chem. 2005, 70, 6537.
(15) (a) Singh, A. K.; Bakshi, R. K.; Corey, E. J. J. Am. Chem. Soc. 1987,
109, 6187. (b) Bachi, M. D.; Bosch, E. J. Org. Chem. 1992, 57, 4696.
(13) Neef, G.; Eckle, E.; Muller-Fahrnow, A. Tetrahedron 1993, 49, 833.
9
J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008 16427

A R T I C L E S
Trost et al.
Scheme 5. Synthesis of Dienonate 27
Scheme 6. Synthesis and Cyclization Attempts of Tertiary Alkoxycarbonyl Selenide 33
Radical-Based Strategy. While the addition of any carbon
nucleophile to vinyl epoxide 23 appeared to be a daunting task,
an alternative approach to convert this compound into a more
useful intermediate for the synthesis of pseudolaric acid B (1a)
was sought. Examination of a molecular model of vinyl epoxide
23 revealed that the oxygen atom of the epoxide and the allylic
hydrogen of the alkene were in close proximity to each other.
It was envisioned that opening of the epoxide to form a dienol
should be possible via coordination of the oxygen atom and
directed deprotonation (A in Scheme 5). Indeed, treatment of
the epoxide formed from diene 2916 with a large excess of
LDA17 resulted in the formation of dienol 28 (R ) TES)
(Scheme 5). Furthermore, only the major diastereoisomer of the
epoxide reacts, greatly simplifying the purification of the
obtained product. As such, it was found to be most convenient
to perform the epoxidation and opening in one reaction sequence
to give dienol 28 (R ) TES) in 72% overall yield.
The transformation of diene 29 (which is readily available
by silylation of diol 22 with TESCl in 82% yield) into dienol
28 (R ) TES) constitutes an important step toward 1a, as it
introduced the tertiary hydroxy group at C4 with the correct
stereochemistry. It also introduced the C7-C8 double bond and
a double bond at C9-C10 that can serve as a handle for further
functionalization. The methyl ester was introduced in a selective
two-step oxidation sequence at the allylic position (Scheme 5).
Finally, deprotection of the secondary TES group led to diol
27.
in high yield form diol 27 via oxidation with Dess-Martin
periodinane (DMP) (eq 4). Introduction of the side chain was
then attempted using different organometallic acetylide reagents,
but clean conversion could not be obtained using free alcohol
30.
Acylation of 30 proceeded smoothly to give acetate 31 in
quantitative yield (Scheme 6). Although good conversion for
acetylide addition could not be obtained with lithium or Grignard
reagents, cerium acetylide gave full conversion of the starting
material. Unfortunately, the obtained tertiary propargylic alcohol
was unstable, decomposing both on standing and during column
chromatography.
Obviously, the free propargylic alcohol was not a viable
intermediate, so we decided to transform it in situ into an
alkoxycarbonyl imidazole intermediate by quenching the addi-
tion reaction with carbonyldiimidazole (CDI). Gratifyingly, this
approach proved to be highly successful, and 32 was obtained
in 91% yield and good diastereoselectivity (Scheme 6, dr )
8:1 to 12:1). Alkoxycarbonyl imidazole 32 was sufficiently
stable to allow easy purification by column chromatography.
The conversion of 32 to alkoxycarbonyl selenide 33 pro-
ceeded smoothly in 76% yield. Unfortunately, when 33 was
submitted to standard conditions for radical generation, no
cyclization was observed. A complex mixture was obtained, in
which the major product appeared to be the decarboxylation
product 34, as the phenyl group was absent but no changes had
The most direct approach is to introduce the side chain prior
to cyclization. To achieve this goal, ketone 30 was synthesized
(16) In a second synthesis campaign, the TBS group was replaced by a
TES group, as it had become apparent that removal of the bulkier
TBS group would require the use of harsh conditions that are not
compatible with sensitive substrates.
(17) The amount of LDA was 5-7 equiv. Although the reaction also
proceeded when a smaller excess was used, this resulted in a longer
reaction time that was detrimental to the yield.
1
occurred in the structure of the diene, as shown by H NMR
spectroscopy. All further attempts were also unsuccessful.
Initiation at room temperature using BEt3/O2 resulted in a
complex mixture or reisolation of the starting material. An
9
16428 J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008

Total Synthesis of (-)-Pseudolaric Acid B
A R T I C L E S
Scheme 7. Synthesis and Cyclization of Secondary Alkoxycarbonyl Selenide 35
attempt to activate the conjugated system with Yb(OTf)3 resulted
in the elimination of HOAc and formation of triene products.
From the studies on substrate 33, we concluded that frag-
mentation with loss of CO2 to form a tertiary radical was favored
over the desired cyclization reaction. As an alternate strategy,
we synthesized the secondary alkoxycarbonyl selenide 35
derived from 27 (Scheme 7).
the complete tricyclic core of 1a. We first attempted to use basic
conditions to promote ꢀ-elimination to the conjugated ester.19
At low temperature, no reaction was observed with either NaH,
LiHMDS or LDA, and higher temperature led to decomposition.
In the case of LDA, racemization of the C7 center next to the
methyl ester was observed, indicating that enolate formation
was not the problem. An attempt to generate a silyl enolate prior
to elimination only resulted in silylation of the propargylic
alcohol.
When alkoxycarbonyl selenide 35 was submitted to the
standard reaction conditions for radical cyclization, a mixture
containing three major products was obtained in 90% yield after
extensive optimization. For successful cyclization, a 0.6:2.2:1
AIBN/Bu3SnH/35 ratio and slow addition of azobisisobutyroni-
trile (AIBN) were essential.18 The best results were obtained
with a relatively high concentration of Bu3SnH, showing that
the main problem was not (as is often the case) direct reduction
of the carbonyl radical but rather decomposition of the formed
allylic instead of the desired hydride abstraction. This also
explains the high quantity of initiator needed for good conver-
sion. Apart from the desired product 36, two other compounds
were isolated: one of them was tentatively assigned as isomer
37, whereas the other, lacking the methyl ester group, could
have been bislactone 38, resulting from a subsequent intramo-
lecular lactonization. Separation of this mixture was not
successful, and it was directly submitted to the next step.
It was our hope that the mixture resulting from the radical
cyclization could be transformed to a single compound under
hydrolysis conditions followed by methylation of the resulting
carboxylates. Indeed, when the mixture of 36-38 was subjected
to forcing hydrolysis conditions (1 M NaOH, 100 °C, 24 h)
followed by methylation with TMSCHN2, a new compound was
obtained as a diastereomeric mixture (Scheme 7). The formed
compound was identified as 39, which is diastereomeric at C7
and resulted from an intramolecular attack of the tertiary alcohol
at C4 onto the conjugate ester. We were never able to achieve
the opening of the lactone(s) ring(s) without concomitant
formation of the oxo bridge.
We then turned to acid activation. Lewis acid either resulted
in no reaction (BF3 ·OEt2, Ti(OiPr)4) or in elimination at the
propargylic position (TMSOTf, TsOH). Finally, an attempt was
made to promote lactonization using Otera’s catalyst20 prior to
opening of the oxo bridge; however, no reaction was observed
under standard conditions, and higher temperature resulted in
decomposition. Examination of a molecular model showed that
the formation of the oxo bridge “pulls away” the ester at C10
and the alcohol at C11, making lactonization unfavorable.
From these experiments, we concluded that protection of the
C4 alcohol could not be avoided for a successful synthesis of
1a. Because of the strong basic conditions needed for lactone
opening, we needed a protecting group that was stable to base
but could be removed under mild neutral conditions. For this
reason, a PMB group was finally selected. Introduction of the
PMB group was achieved in 94% yield using activation of the
imidate with Sc(OTf)3 (Scheme 8). For this reaction, low
temperature and catalyst loading were essential to achieve good
yields, as elimination of the tertiary alcohol was observed
otherwise.
The protection of the alcohol seemed to slow down the radical
cyclization, however, and and when the standard conditions
developed for the free alcohol 35 were used, only a small
amount of conversion was observed in the case of PMB-
protected 41 (entry 1 of Table 1). It was found that a higher
initiator loading (1.2 equiv) was necessary to achieve full
conversion (entry 2). Under these conditions, a mixture of the
double-bond isomers 43 and 44 was observed along with several
unidentified impurities. Addition of DBU to the reaction mixture
directly after completion of the cyclization allowed clean
conversion of 44 to the desired product 43, which could be
isolated in 56% yield and ∼90% purity (entry 3). To improve
this yield, we decided to use azobis(dicyclohexylcarbonitrile)
(42) as the initiator, as it is thermally more stable and allows
In principle, the oxo bridge could be used as a protecting
group for the tertiary alcohol, provided it could be opened at a
later stage. To test this hypothesis, the secondary alcohol of 39
was oxidized with DMP to give the corresponding ketone.
Cerium acetylide addition proceeded in 76% yield to give the
stable propargylic alcohol 40.
The next task was to perform the opening of the oxo bridge
of 40 and the lactonization of the propargylic alcohol to give
(18) In the optimized procedure, a reaction mixture containing 1.4 equiv
of Bu3SnH and 0.5 equiv of AIBN was heated at 90 °C, after which
a solution containing 0.7 equiv of Bu3SnH and 0.5 equiv AIBN was
added over 15 min.
(19) For an example of similar eliminative opening, see: Jotterand, N.;
Vogel, P. Synlett 1999, 1883.
(20) Otera, J.; Danoh, N.; Nozaki, H. J. Org. Chem. 1991, 56, 5307.
9
J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008 16429

A R T I C L E S
Trost et al.
Scheme 8. Completion of the Synthesis
Table 1. Radical Cyclization of 41
entry
conditions
products
1
2
3
4
5
2.2 equiv Bu3SnH, 0.6 equiv AIBN, benzene, 80 °C
3 equiv Bu3SnH, 1.2 equiv AIBN, benzene, 80 °C
3 equiv Bu3SnH, 1.2 equiv AIBN, benzene, 80 °C, then DBU
1.4 equiv Bu3SnH, 0.2 equiv 42, toluene, 110 °C
<20% conversion
5-15% 43, 30-50% 44, several impurities
56% 43, several impurities
5-15% 43, 30-50% 44, 20-30% impurity
85% 43, 92% purity
3.5 equiv Bu3SnH, 2.0 equiv 42, benzene, 70 °C, then DBU
for cleaner, continuous generation of radicals.21 When the
reaction was conducted in refluxing toluene with only 0.2 equiv
of 42, full conversion was indeed achieved, and the reaction
was cleaner (entry 4). Unfortunately, the formation of other side
products was favored, apparently via reaction of the benzylic
position of the PMB protecting group, as shown by the
disappearance of the characteristic two doublets of the PMB
group in the 1H NMR spectrum. We found after further
optimization that this side reaction was minimized at lower
temperature. In this case, a higher loading of 42 was needed
for full conversion. Use of 2 equiv of 42 and 3.5 equiv of
Bu3SnH in benzene at 70 °C for 15 h followed by isomerization
with DBU allowed the isolation of 43 in 85% yield and 92%
purity (entry 5). As the radical generation is controlled by the
low rate of decomposition of 42 and not the rate of addition,
all of the reagents could be mixed together at the beginning,
leading to an operationally simpler procedure.
led to removal of the PMB group, mostly with isolation of
bislactone 38. We concluded that the main problem was the
acidic workup conditions and the subsequent methylation
reaction. For a successful ring opening, direct methylation of
the formed carboxylate salt was required; this precluded the
use of excess base in water for the hydrolysis. We finally chose
KOTMS as the reagent, as the anhydrous salt is readily soluble
in organic solvents. Furthermore, potassium was expected to
give a dicarboxylate salt that was more soluble than the barium
salt, for example, allowing for an easier subsequent methylation.
Indeed, reaction with KOTMS followed by methylation with
dimethyl sulfate in the presence of bicarbonate resulted in
successful formation of the desired alcohol in 50% yield together
with isolated starting material (10%) and several side products,
one of them resulting from methylation of the secondary alcohol
at C11. Unfortunately, this protocol was difficult to reproduce
on a larger scale, leading to lower yields (<30%).
The next task on the way toward completion of the synthesis
was to open the lactone of 43. Several attempts using hydrolysis
with base in water followed by extraction from an acidic solution
and methylation with TMSCHN2 were unsuccessful and mainly
We hypothesized that use of a soluble buffer system after
ring opening with KOTMS should be more efficient for
controlling the pH of the reaction. Optimization studies led to
the use of 6 equiv of TsOH and 12 equiv of Hu¨nig’s base in
methanol as the buffer. These conditions were designed to
completely protonate any alkoxide present and leave only the
(21) Overberger, C. G.; Biletch, H.; Finestone, A. B.; Lilker, J.; Herbert,
J. J. Am. Chem. Soc. 1953, 75, 2078.
9
16430 J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008

Total Synthesis of (-)-Pseudolaric Acid B
A R T I C L E S
Scheme 9. Selective Cyclization of Diol 47
Figure 2. NOE studies on 48 and a model for acetylide addition to ketone
46.
carboxylates as the best nucleophiles in the reaction mixture.
Indeed, the desired alcohol was formed in 65% yield (75% brsm)
under these conditions, and no methylation of the alcohol was
observed (Scheme 8).
DMP was then used to oxidize the alcohol to the correspond-
ing ketone 45. Buffering with NaHCO3 and decreasing the
temperature to 0 °C were important to prevent decomposition
during the oxidation step. The problem of relactonization was
easier to control when column chromatography was omitted
prior to oxidation. In that case, ketone 45 was obtained in 59%
(73% brsm) over two steps.
studies (Figure 2). The strong NOE signal observed between
methyl group 13 and H10 was especially crucial for the
determination of the configuration of the propargylic center. The
high selectivity of the addition step is explained by a Felkin-Ahn
approach on ketone 46, where the methyl ester group blocks
the other face of the carbonyl group.
The deprotection of the terminal acetylene proceeded un-
eventfully in 87% yield using TBAF in THF (Scheme 8). The
acylation of the tertiary alcohol was attempted next. The use of
DMAP and AcCl reported for the synthesis of pseudolaric acid
A (1b) led to a very slow reaction (∼20% in 24 h).5i Fortunately,
freshly dried Sc(OTf)3 in acetic anhydride solvent at 0 °C
resulted in clean acylation of the tertiary alcohol in 98% yield.23
At this point, we chose a Stille coupling to complete the
synthesis, as it allows the use of free acids for the coupling
reaction. The iodide 51 needed for the Stille reaction is a known
compound and was synthesized in two steps from diethyl methyl
malonate using a literature procedure.24
The hydrostannylation of the free alkyne in 49 proceeded in
good yield with good regio- and trans selectivity, as shown by
1H NMR, but the formed stannane 50 was surprisingly unstable.
Consequently, it was always prepared immediately prior to Stille
coupling and purified over a short plug of deactivated SiO2.
First attempts at the Stille coupling were made with the most
often used conditions, namely, ligandless Pd(II) salts in DMF.25
Both Pd(CH3CN)Cl2 and Pd(PhCN)2Cl2 led to low yields
(<30%) of pseudolaric acid B (1a). In both cases, protodestan-
nylation and isomerization of the diene were serious issues.
There is a single report in the literature describing Stille coupling
of an acid under basic conditions using Pd2(dba)3 as the catalyst
and NMP as the solvent.26 Gratifyingly, these conditions were
efficient in shutting down both the destannylation and isomer-
ization side reactions, and 1a was obtained in 62% yield after
HPLC purification. All of the physical and spectroscopic data
(melting point, optical rotation, and 1H and 13C NMR, IR, and
mass spectra) were in agreement with the published values for
natural pseudolaric acid B.1e,2a
Cerium acetylide addition was not compatible with the PMB
protecting group, as ꢀ-alkoxide elimination was a major
problem. For this reason, the PMB group was first removed
with DDQ in 76% yield.
A few attempts to acylate the tertiary alcohol of 46 were
made. However, a considerable amount of elimination to the
R,ꢀ-unsaturated ketone was observed, so we decided to focus
on 46 as the substrate for the cerium acetylide addition.
The first attempts at cerium acetylide addition using the
standard procedure were disappointing, as they resulted in a
complex mixture. In a single case, however, we obtained a clean
reaction to the desired product 47 in nearly quantitative yield!
Thus, it appeared that this reaction was extremely sensitive to
the quality of the cerium reagent, which led to major difficulties
in reproducing this result. At this point, we become interested
in a recent report of Knochel and co-workers22 regarding the
use of soluble cerium-lithium choride reagents for addition
reactions. Although we still encountered difficulties when using
large quantitites of Ce reagents, we found that on a small scale
(<1 mmol cerium), simply codrying CeCl3 ·7H2O with 2 equiv
of LiCl was successful in giving a clear solution of the reagent
in THF. Importantly, the cerium acetylide addition was suc-
cessful every time the clear solution was used as the reagent
and gave the desired addition product as a single diastereoisomer
in 80-90% yield with 20-10% isolated starting material
(Scheme 8). The optimal ratio of reagent to substrate was found
to be 8:1, as smaller ratios led to lower conversions. Inverse
quenching of the reaction mixture with bicarbonate solution was
important to prevent intramolecular attack of the tertiary alcohol
onto the conjugate ester.
Summary
We have reported the successful total synthesis of the highly
biologically active and structurally intriguing compound pseudola-
ric acid B (1a). Vinylcyclopropane 16 was easily accessed in
seven linear steps and 33% overall yield from 2-acetylbutyro-
lactone (8). The intramolecular cycloaddition of 16 constituted
an efficient access to the polyhydroazulene core of 1a. In this
particular case, Rh catalyst 18 was superior to Ru catalyst 17,
The lactonization of 47 was examined next. Basic conditions
could not be used, as they led to exclusive formation of oxo-
bridge compound 40 (Scheme 9). Fortunately, the use of Otera’s
catalyst20 in toluene (30 min at 130 °C under microwave
irradiation) resulted in formation of the desired lactone 48 in
94% yield. This perfect orthogonality between basic conditions
and Otera’s catalyst is noteworthy and probably results from
the unique ability of Otera’s catalyst to activate both the methyl
ester and the alcohol. At this point, the stereochemistry of the
propargylic position could be assigned by one-dimensional NOE
(23) Ishihara, K.; Kubota, M.; Kurihara, H.; Yamamoto, H. J. Am. Chem.
Soc. 1995, 117, 4413.
(24) Baker, R.; Castro, J. L. J. Chem. Soc., Perkin Trans. 1 1990, 47.
(25) Smith, A. B.; Ott, G. R. J. Am. Chem. Soc. 1998, 120, 3935.
(26) Song, H. Y.; Joo, J. M.; Kang, J. W.; Kim, D. S.; Jung, C. K.; Kwak,
H. S.; Park, J. H.; Lee, E.; Hong, C. Y.; Jeong, S.; Jeon, K.; Park,
J. H. J. Org. Chem. 2003, 68, 8080.
(22) Krasovskiy, A.; Kopp, F.; Knochel, P. Angew. Chem., Int. Ed. 2006,
45, 497.
9
J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008 16431

A R T I C L E S
Trost et al.
2.07-2.01 (m, 1H), 1.68-1.56 (m, 2H), 1.38-1.21 (m, 1H), 1.12
(d, J ) 6.2 Hz, 3H), 1.05 (s, 9H), 0.84 (s, 9H,), 0.02 (s, 3H), -0.02
(s, 3H). 13C NMR (125 MHz, CDCl3): δ 150.4, 135.6, 133.9, 132.1,
130.0, 129.5, 127.6, 119.7, 69.6, 68.6, 52.0, 43.7, 40.4, 34.0, 28.8,
26.8, 25.8, 24.4, 22.3, 19.3, 18.0, -4.2, -4.8. IR ν (cm-1): 3070
(w), 3051 (w), 2999 (w), 2955 (s), 2930 (s), 2896 (m), 2858 (s),
1958 (w), 1888 (w), 1823 (w), 1770 (w), 1728 (w), 1694 (w), 1660
(w), 1590 (w), 1487 (w), 1472 (m), 1463 (m), 1446 (w), 1428 (m),
1390 (w), 1370 (m), 1361 (m), 1255 (m), 1187 (w), 1154 (m), 1112
(s), 1042 (m), 1007 (m), 954 (w), 939 (w), 910 (w), 882 (w), 834
(s), 807 (m), 774 (m), 739 (m), 701 (s), 667 (w), 613 (m), 506
(m), 487 (m). HRMS (EI): calcd for C35H52O2Si2, m/z 560.3506
(M+); found, 560.3508. Anal. Calcd for C35H52O2Si2: C, 74.94; H,
9.34. Found: C, 74.78; H, 9.43. Additional data: 1H COSY, HSQC,
NOE (see the Supporting Information).
as it displayed a lower tendency toward C-H insertion and
subsequent isomerization of the diene product. Further elabora-
tion of the polyhydroazulene core via a novel isomerization of
a 1,4-diene to a 1,3-diene using TBAF and an interesting
vinylogous opening of a vinyl epoxide with LDA as key steps
resulted in alkoxycarbonyl selenides 33, 35, and 41, but only
the secondary alkoxycarbonyl selenides 35 and 41 were efficient
precursors for the introduction of the quaternary center via a
radical cyclization reaction. On the way to pseudolaric acid B,
a novel procedure for the opening of lactones and a highly
selective cerium acetylide addition to a sensitive ketone were
developed. Interesting observations were made on the orthogo-
nality of cyclization conditions for generating diverse polycyclic
ring systems from the polyhydroazulene core: intramolecular
conjugate addition was preferred under basic conditions, whereas
lactonization was observed exclusively using Otera’s catalyst.
This work culminated in the total synthesis of 1a in 28 overall
steps and 1.4% overall yield. The strength of our synthesis is
the very high selectivity observed for the installation of the four
stereocenters of the molecule via Noyori reduction, epoxidation,
radical cyclization, and cerium acetylide addition, which together
with the [5 + 2] cycloaddition contributes to its efficiency.
Furthermore, the late-stage introduction of the acid side chain
is also interesting from the point of view of analogue synthesis.
It should be noted that a new strategy for the synthesis of
[6.5]spiro fused bicycles also emerged. Thus, the [5 + 2]
cycloaddition can also initiate formation of other isomeric
bicyclic skeletons that are also found among natural and other
bioactive targets.
Data for 20. Rf: 0.40 (3:1 pentane/DCM, anisaldehyde). [R]D (25
°C): -80.0 (CHCl3, c ) 1.0). 1H NMR (500 MHz, CDCl3): δ
7.70-7.68 (m, 4H), 7.46-7.34 (m, 6H), 5.47 (m, 1H), 5.42 (m,
1H), 3.78-3.75 (m, 1H), 3.56 (d, J ) 4.8 Hz, 2H), 2.32-2.14 (m,
5H), 2.08-1.96 (m, 4H), 1.61-1.55 (m, 1H), 1.11 (d, J ) 6.1 Hz,
3H), 1.06 (s, 9H), 0.84 (s, 9H), 0.09 (s, 3H), 0.07 (s, 3H). 13C
NMR (125 MHz, CDCl3): δ 135.6, 135.3, 133.9, 133.3, 129.5,
127.6, 124.1, 122.3, 68.4, 67.8, 48.1, 37.6, 34.5, 29.2, 26.8, 25.9,
23.8, 22.6, 22.2, 19.3, 18.1, -3.7, -4.8. IR ν (cm-1): 3074 (w),
3050 (w), 3029 (w), 2957 (m), 2929 (s), 2895 (m), 2857 (s), 1487
(w), 1472 (m), 1463 (w), 1428 (m), 1390 (w), 1375 (w), 1361 (w),
1264 (m), 1191 (w), 1112 (s), 1073 (m), 1018 (w), 1006 (m), 985
(w), 966(w), 938 (w), 914 (w), 887 (w), 835 (s), 811 (m), 793 (m),
773 (m), 739 (m), 701 (s), 614 (w), 504 (m). HRMS (EI): calcd
for C35H52O2Si2, m/z 560.3506 (M+); found, 560.3488. Additional
1
data: H COSY, HSQC (see the Supporting Information).
(R)-1-((1R,6R)-6-Hydroxymethyl-1,2,3,6,7,8-hexahydroazulen-1-
yl)ethanol (22). TBAF (1 M in THF, 19 mL, 19 mmol, 6.0 equiv)
was added over 10 min to a suspension of 3 Å molecular sieves
(flame-dried 4× in HV, 19 g) and 19 (1.82 g, 3.24 mmol, 1.00
equiv) in THF (190 mL) at 0 °C. After the mixture was stirred for
4 h at 23 °C, the reaction was quenched with methanol (20 mL);
the mixture was filtered, and the residues were washed with
methanol (10 mL) and Et2O (300 mL). The reaction mixture was
washed with saturated NH4Cl (3 × 200 mL) and brine (100 mL),
and the combined water layers were extracted with Et2O (3 × 300
mL). The combined organic layers were dried over MgSO4, and
the solvent was removed under reduced pressure. The crude product
was purified by flash column chromatography (1:1-0:1 PET/Et2O)
to yield 22 (636 mg, 3.05 mmol, 94%) as a colorless oil. On a
larger scale (no full conversion), TBAF (1 M in THF, 100 mL,
100 mmol, 3.3 equiv) was added over 10 min to a suspension of 3
Å molecular sieves (dried 15 h at 200 °C in an oven in vacuo,
150 g) and 19 (16.8 g, 30.0 mmol, 1.00 equiv) in THF (1 L) at 0
°C. After the reaction mixture was stirred for 16 h at 23 °C, the
reaction was quenched with methanol (200 mL); the mixture was
filtered, and the residues were washed with methanol (500 mL).
The solvent were removed under reduced pressure, and Et2O (600
mL) was added. The reaction mixture was washed with saturated
NH4Cl (3 × 200 mL) and brine (100 mL), and the combined water
layers were extracted with Et2O (3 × 300 mL). The combined
organic layers were dried over MgSO4, and the solvent was removed
under reduced pressure. The crude product was purified by flash
column chromatography (1:1-0:1 PET/Et2O) to yield 22 (5.35 g,
25.7 mmol, 86%, 6:1 mixture of double-bond isomers) as a colorless
oil. The mixture was dissolved in THF (140 mL) at 0 °C, and a
solution of KOtBu (8.0 g, 71 mmol, 2.8 equiv) in THF (40 mL)
was added dropwise. After 50 min, the reaction mixture was poured
onto saturated NaHCO3 (200 mL) and extracted with Et2O (3 ×
200 mL). The combined organic layers were washed with brine
(100 mL) and dried over MgSO4, and the solvent was removed
under reduced pressure to give pure 22 (5.15 g, 24.7 mmol, 96%,
82% from 19) as a colorless oil. Rf: 0.30 (Et2O, anisaldehyde). [R]D
Experimental Section
(1R,3aS,6R)-1-[(R)-1-(tert-Butyldimethylsilanyloxy)ethyl]-6-(tert-
butyldiphenylsilanyloxymethyl)-1,2,3,3a,6,7-hexahydroazulene (19)
and (1R,6S)-1-[(R)-1-(tert-Butyldimethylsilanyloxy)ethyl]-6-(tert-
butyldiphenylsilanyloxymethyl)-1,2,3,5,6,7-hexahydroazulene (20).
-
[CpRu(CH3CN)3]+PF6 (17) (1.9 g, 4.4 mmol, 0.23 equiv) was
added in portions over 20 min to a solution of 16 (10.7 g, 19.1
mmol, 1.00 equiv) in acetone (distilled over Drierite, 200 mL) at
0 °C. After 1 h at 23 °C, the reaction mixture was poured into
PET (1 L) and filtered over SiO2; the plug was washed with 10:1
PET/Et2O, and the solvent was removed under reduced pressure.
The crude product was purified by flash column chromatography
(30:1-25:1-20:1-15:1-10:1-3:1 pentane/DCM, 1 kg SiO2, col-
lected fraction: 200 mL) to yield a first fraction (fractions 32-91)
of diastereomerically pure 19 (4.59 g, 8.18 mmol, 43%) and a
second fraction (fractions 92-105) containing 19 together with the
trans diastereomer [534 mg, 0.952 mmol, 5%, dr(cis:trans) ) 1:1.6]
(total 19: 5.13 g, 9.15 mmol, 48%, dr ) 15:1); the third fraction
(fractions 106-122) contained 20 (1.60 g, 2.85 mmol, 15%), whose
structure has tentatively been assigned as shown in eq 2. All of the
isolated products were colorless, air-sensitive oils.
Synthesis of 19 Using Rh Catalyst. Rh catalyst 18 (2.0 g, 3.5
mmol, 0.11 equiv) was added in four portions of 500 mg every 10
min to a solution of 16 (18.2 g, 32.4 mmol, 1.00 equiv) in DCE
(distilled over CaH2, 320 mL) at 0 °C. After 1 h at 23 °C, the
reaction mixture was poured into PET (300 mL) and filtered over
SiO2; the plug was washed with 1:1 PET/Et2O (1.5 L), and the
solvent was removed under reduced pressure. The crude product
was purified by flash column chromatography (50:1-20:1-15:
1-10:1-3:1 pentane/DCM, 1 kg SiO2) to yield 19 (16.1 g, 28.7
mmol, 88%, dr ) 15:1) as a colorless oil.
Data for 19. Rf: 0.50 (3:1 pentane/DCM, anisaldehyde). [R]D (25
°C): -12.9 (CHCl3, c ) 1.0). 1H NMR (500 MHz, CDCl3): δ
7.68-7.65 (m, 4H), 7.44-7.36 (m, 6H), 5.71-5.68 (m, 1H),
5.49-5.42 (m, 2H), 4.00-3.96 (m, 1H), 3.54-3.45 (m, 2H), 3.29
(br m, 1H), 2.47 (br m, 1H), 2.38 (br m, 1H), 2.23-2.19 (m, 2H),
1
(25 °C): +107.5 (CHCl3, c ) 1.0). H NMR (500 MHz, CDCl3):
9
16432 J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008

Total Synthesis of (-)-Pseudolaric Acid B
A R T I C L E S
δ 5.82 (d, J ) 11.6 Hz, 1H), 5.67 (dd, J ) 11.5, 4.9 Hz, 1H), 4.08
(qd, J ) 6.4, 2.3 Hz, 1H), 3.61 (dd, J ) 10.5, 5.8 Hz, 1H), 3.58
(dd, J ) 10.5, 7.2 Hz, 1H), 2.68 (br m, 1H), 2.58 (br m, 1H),
2.47-2.42 (m, 3H), 2.38-2.10 (m, 2H), 1.86-1.77 (m, 4H), 1.65
(br s, 1H), 1.15 (d, J ) 6.4 Hz, 3H). 13C NMR (125 MHz, CDCl3):
δ 140.9, 134.9, 132.6, 125.9, 67.4, 66.0, 57.9, 43.3, 37.2, 27.5,
26.9, 21.2, 20.5. IR ν (cm-1): 3384 (s), 3008 (w), 2925 (s), 2871
(s), 1644 (w), 1633 (w), 1626 (w), 1445 (m), 1434 (m), 1372 (m),
1350 (w), 1308 (w), 1261 (w), 1148 (w), 1117 (w), 1079 (m), 1029
(s), 949 (w), 913 (w), 861 (w), 834 (w), 737 (m), 666 (w), 645
(w). HRMS (EI): calcd for C13H20O2, m/z 208.1463 (M+); found,
208.1467. Anal. Calcd for C13H20O2: C, 74.96; H, 9.68. Found: C,
was heated at 70 °C for 15 h, it was cooled to 23 °C. TLC (2:1
PET/AcOEt, CAN) showed the formation of two major products,
identified as the desired 43 as well as its double-bond isomer 44.27
DBU (12 mL) was added, and the reaction was monitored by TLC.
After 5 h, isomerization to the desired 43 was complete; the reaction
mixture was diluted with Et2O (100 mL) and washed with 1 M
NaHSO4 (2 × 100 mL), and the combined aqueous layers were
extracted with Et2O (2 × 100 mL). The combined organic layers
were washed with saturated NaHCO3 solution (50 mL) and brine
(50 mL) and dried over MgSO4, and the solvent was removed under
reduced pressure. The crude product was purified by flash column
chromatography (50 g SiO2 and 5 g KF,28 10:1-5:1-2:1-1:1 PET/
AcOEt) to yield 43 (447 mg, 1.09 mmol, 85%, >92% pure by 1H
NMR) as a colorless foam. Rf: 0.30 (2:1 PET/AcOEt, CAN). [R]D
1
75.02; H, 9.52. Additional data: H COSY, HSQC, NOE (see the
Supporting Information).
1
(25 °C): 101.9 (CHCl3, c ) 1.0). H NMR (500 MHz, CDCl3): δ
(3S,3aS)-3-((R)-1-Triethylsilanyloxyethyl)-6-triethylsilanyloxym-
ethyl-2,3,4,5-tetrahydro-1H-azulen-3a-ol 28 (R ) TES). A suspen-
sion of 29 (2.60 g, 5.45 mmol, 1.00 equiv) and NaHCO3 (1.2 g, 14
mmol, 2.4 equiv) in DCM (65 mL) was cooled to -20 °C, and
m-CPBA (95%, washed with pH 7 buffer and dried in HV before
titration, 1.4 g, 7.9 mmol, 1.3 equiv) was added. After 30 min, the
cool solution was directly filtered through a plug of silica gel
(deactivated with NEt3); the plug was washed with 10:1 PET/Et2O,
and the solvent was removed under reduced pressure to give nearly
pure epoxide 23b (R ) TES). Epoxide 23b (R ) TES) was used
at once in the next step. Rf: 0.60 (10:1 PET/Et2O, anisaldehyde).
1H NMR (400 MHz, benzene-d6): δ 5.76 (dd, J ) 11.9, 2.4 Hz,
1H), 5.65 (dd, J ) 11.9, 5.5 Hz, 1H), 4.00 (qd, J ) 6.3, 1.8 Hz,
1H), 3.59-3.52 (m, 2H), 2.63 (br m, 1H), 2.43-2.36 (m, 1H),
2.27-2.24 (m, 1H), 2.04-1.86 (m, 4H), 1.68-1.62 (m, 1H),
7.26 (d, J ) 8.7 Hz, 2H), 7.23-7.18 (m, 1H), 6.89 (d, J ) 8.7 Hz,
2H), 4.56 (qd, J ) 6.4, 2.3 Hz, 1H), 4.39 (d, J ) 10.4 Hz, 1H),
4.32 (d, J ) 10.4 Hz, 1H), 3.81 (s, 3H), 3.69 (s, 3H), 2.83 (dd, J
) 14.9, 6.4 Hz, 1H), 2.76 (ddd, J ) 14.5, 4.2, 1.6 Hz, 1H), 2.55
(dd, J ) 14.5, 8.8 Hz, 1H), 2.52-2.41 (m, 2H), 2.34 (dd, J )
14.6, 5.9 Hz, 1H), 2.12-2.00 (m, 1H), 1.93-1.80 (m, 3H), 1.57
(ddd, J ) 14.3, 12.4, 1.5 Hz, 1H), 1.35 (d, J ) 6.5 Hz, 3H). 13C
NMR (125 MHz, CDCl3): δ 174.8, 168.3, 159.1, 143.2, 134.3,
130.0, 128.7, 113.9, 83.6, 75.1, 62.1, 56.6, 55.3, 51.9, 45.7, 35.9,
27.1, 26.2, 19.1, 18.8, 18.8. IR ν (cm-1): 3062 (w), 2950 (m), 2870
(w), 1732 (s), 1714 (s), 1644 (w), 1614 (m), 1586 (w), 1514 (s),
1440 (m), 1384 (w), 1361 (w), 1324 (w), 1301 (m), 1271 (m), 1251
(s), 1204 (m), 1171 (s), 1123 (m), 1094 (m), 1070 (m), 1028 (m),
912 (w), 822 (w), 770 (w), 732 (m), 648 (w), 574 (w). HRMS
(ESI): calcd for C23H28O6Na, m/z 423.1784 ([M + Na]+); found,
423.1798.
1.46-1.27 (m, 2H), 1.06-0.95 (m, 21H), 0.65-0.63 (m, 12H). 13
C
NMR (125 MHz, benzene-d6): δ 136.3, 126.9, 75.3, 69.0, 68.8,
64.4, 53.0, 43.7, 33.9, 24.1, 22.2, 22.1, 21.3, 7.3, 7.3, 7.1, 5.8, 5.4,
4.8, 4.7.
(1S,3aR,8aS)-8a-Hydroxy-1-((S)-1-hydroxy-1-methyl-3-trimeth-
ylsilanylprop-2-ynyl)-2,3,4,7,8,8a-hexahydro-1H-azulene-3a,6-di-
carboxylic Acid Dimethyl Ester (47). CeCl3 ·7H2O (0.30 g, 0.81
mmol, 9.0 equiv) and lithium chloride (68 mg, 1.6 mmol, 18 equiv)
were dried under HV (<0.05 Torr was essential for good drying),29
warmed from 23 to 150 °C over 3 h (T was increased by ∼10 °C
every 15 min), and then maintained at 150 °C for 2 h. During this
whole process, good stirring was essential to keep the mixture as
a light homogeneous powder. After the mixture was cooled to 23
°C, THF (6 mL) was added, and the suspension was vigorously
stirred at 23 °C. After 4 h, a clear solution was obtained, and the
reaction mixture was cooled to -78 °C. In a separate flask, nBuLi
(2.5 M in hexane, 0.28 mL, 0.70 mmol, 8.0 equiv) was added to a
solution of trimethylsilyl acetylene (distilled, 0.12 mL, 0.85 mmol,
9.8 equiv) in THF (2 mL) at -78 °C. After 30 min, this solution
was transferred via canula to the CeCl3 ·2LiCl solution at -78 °C.
After 1 h, a solution of ketone 46 (27 mg, 0.087 mmol, 1.0 equiv)
in cool [-78 °C (important!)] THF (2 mL) was added via canula.
After 1 h, the reaction mixture was poured into saturated NaHCO3
solution (10 mL) and extracted with Et2O (3 × 10 mL). The
combined organic layers were washed with brine (5 mL) and dried
over MgSO4, and the solvent was removed under reduced pressure.
The crude product was purified by flash column chromatography
(1:0-20:1-10:1-5:1-2:1 DCM/AcOEt) to yield 47 (31 mg, 0.076
mmol, 87%, dr >20:1) as a colorless oil and reisolated 46 (3 mg,
0.01 mmol, 11%). Rf: 0.50 (10:1 DCM/AcOEt, CAN/KMnO4). [R]D
LDA [1 M in THF, freshly prepared from diisopropyl amine
(7.4 mL, 41 mmol) and nBuLi (2.5 M in hexane, 24 mL, 41 mmol)
in THF (36 mL), 41 mmol, 7.0 equiv] was added over 20 min to
a solution of crude epoxide 23b in THF (120 mL) at -10 °C. The
reaction was monitored by TLC (10:1 PET/Et2O). After 40 min,
no 23b (R ) TES) was left, and the reaction was quenched with
saturated NH4Cl solution (100 mL) and extracted with Et2O (3 ×
150 mL). The combined organic layers were washed with brine
(50 mL) and dried over MgSO4, and the solvent was removed under
reduced pressure. The crude product was purified by flash column
chromatography (10:1-3:1 PET/Et2O) to yield 28 (1.93 g, 4.26
mmol, 72% over two steps) as a colorless oil. Compound 28 (R )
TES) proved to be unstable in pure form, especially on larger scales,
and had to be kept frozen in benzene to prevent decomposition.
Rf: 0.20 (10:1 PET/Et2O, anisaldehyde). [R]D (25 °C): -19.1
1
(CHCl3, c ) 1.0). H NMR (500 MHz, CDCl3): δ 5.82 (dm, J )
7.5 Hz, 1H), 5.64 (dm, J ) 7.6 Hz, 1H), 4.11 (s, 2H), 4.07 (qd, J
) 6.3, 3.3 Hz, 1H), 2.58-2.50 (m, 2H), 2.48-2.39 (m, 1H), 2.28
(dt, J ) 18.1, 4.1 Hz, 1H), 2.08-1.88 (m, 2H), 1.91-1.88 (m,
1H), 1.81-1.74 (m, 3H), 1.16 (d, J ) 6.3 Hz, 3H), 0.97 (t, J ) 8.1
Hz, 9H), 0.91 (t, J ) 8.0 Hz, 9H), 0.62 (q, J ) 7.8 Hz, 6H), 0.52
(q, J ) 8.0 Hz, 6H). 13C NMR (125 MHz, CDCl3): δ 152.2, 142.3,
118.7, 117.2, 80.8, 68.2, 67.5, 59.4, 32.9, 30.2, 25.0, 23.4, 22.9,
6.9, 6.8, 5.3, 4.5. IR ν (cm-1): 3363 (w), 2956 (s), 2912 (m), 2877
(m), 1738 (w), 1661 (w), 1634 (w), 1456 (m), 1434 (w), 1416 (m),
1372 (m), 1316 (w), 1261 (m), 1240 (m), 1192 (w), 1133 (m), 1112
(m), 1071 (s), 1016 (s), 971 (m), 921 (m), 894 (w), 870 (m), 802
(m), 740 (s), 673 (w). HRMS (EI): calcd for C25H46O2Si2, m/z
434.3036 ([M - H2O]+); found, 434.3031.
1
(25 °C): -1.4 (CHCl3, c ) 1.0). H NMR (500 MHz, CDCl3): δ
6.97 (td, J ) 5.9, 1.4 Hz, 1H), 4.80 (s, 1H), 3.70 (s, 3H), 3.65 (s,
3H), 2.76 (d, J ) 6.4 Hz, 1H), 2.70 (ddd, J ) 14.4, 9.8, 4.5 Hz,
1H), 2.62-2.50 (m, 2H), 2.47 (dd, J ) 9.9, 9.2 Hz, 1H), 2.21-2.12
(1R,7S,8S,9R)-7-(4-Methoxybenzyloxy)-9-methyl-11-oxo-10-
oxatricyclo[6.3.2.01,7]tridec-3-ene-4-carboxylic Acid Methyl Ester
(43). Tributyltin hydride (1.2 mL, 4.5 mmol, 3.5 equiv) was added
to a solution of acyl selenium 41 (708 mg, 1.28 mmol, 1.00 equiv)
and 1,1′-azobis(cyclohexanecarbonitrile) (42) (0.63 g, 2.6 mmol,
2.0 equiv) in benzene (120 mL), and the reaction mixture was
degassed with two freeze-thaw cycles. After the reaction mixture
(27) At this point, removal of the solvent and column chromatography (10:1
SiO2/KF, 10:1-5:1-2:1-1:1 PET/AcOEt) allowed the isolation of
the desired product 43, but the major product was the nonconjugated
ester 44. Direct in situ isomerization of the double bond was more
convenient, however.
(28) Harrowven, D. C.; Guy, I. L. Chem. Commun. 2004, 1968.
(29) A slightly modified version of the procedure given in ref 22 was
used.
9
J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008 16433

A R T I C L E S
Trost et al.
(m, 1H), 2.12-2.06 (m, 1H), 1.97-1.85 (m, 2H), 1.85-1.76 (m,
1H), 1.67 (s, 3H), 1.67 (br s, 1H), 0.12 (s, 9H). 13C NMR (125
MHz, CDCl3): δ 177.0, 168.2, 141.1, 133.9, 111.0, 85.9, 84.1, 69.6,
64.1, 60.0, 52.5, 52.0, 34.7, 33.1, 31.4, 31.1, 26.9, 20.8, 0.0. IR ν
(cm-1): 3436 (m), 2954 (m), 2166 (w), 1714 (s), 1641 (w), 1455
(m), 1436 (m), 1372 (w), 1320 (m), 1276 (m), 1249 (s), 1200 (s),
1102 (m), 1065 (w), 987 (w), 933 (m), 865 (m), 843 (s), 757 (m),
700 (w), 625 (w). HRMS (ESI): calcd for C21H32O6NaSi, m/z
431.1866 ([M + Na]+); found, 431.1881.
0.28 equiv) was added. After an additional 5 h, TLC showed that
only traces of stannane 51 remained, and the reaction mixture was
diluted with AcOEt (10 mL) and washed with 1 M NaHSO4 (2 ×
5 mL). The combined aqueous layers were extracted with AcOEt
(2 × 10 mL). The combined organic layers were washed with brine
(5 mL) and dried over MgSO4, and the solvent was removed under
reduced pressure. The crude product was purified by flash column
chromatography (20:1-5:1-3:1-1:1 PET/AcOEt, 0.5% AcOH) to
yield crude pseudolaric acid B (1a) as a yellowish solid. Purification
via preparative HPLC [10 mL/min flow rate, detection at 265 nm,
PET/AcOEt (0.5% AcOH in AcOEt) from 85:15 to 1:1 over 20
min, 20 min at 1:1, from 50:50 to 5:95 over 20 min, tR ) 31.1
min] finally gave pure 1a (7.2 mg, 0.017 mmol, 62%) as a colorless
solid.
(1R,7S,8R,9S)-7-Hydroxy-9-methyl-11-oxo-9-trimethylsilanyl-
ethynyl-10-oxatricyclo[6.3.2.01,7]tridec-3-ene-4-carboxylic Acid Meth-
yl Ester (48). A solution of diol 47 (58 mg, 0.14 mmol, 1.0 equiv)
and Otera’s catalyst20 (17 mg, 0.028 mmol, 0.20 equiv) in toluene
(4 mL) was heated at 130 °C in a microwave oven for 30 min. The
reaction mixture was poured directly onto a chromatography column
(SiO2, 1:0-10:1-5:1 DCM/AcOEt) to yield lactone 48 (48 mg,
0.13 mmol, 94%) as a colorless solid. Rf: 0.60 (10:1 DCM/AcOEt,
CAN/KMnO4). [R]D (25 °C): 44.4 (CHCl3, c ) 1.0) Mp: 143-145
Rf: 0.25 (1:1 PET/AcOEt, 0.5% AcOH, UV). [R]D (24 °C): -22.5
(MeOH, c ) 0.9) (lit. -25.9 (MeOH, c ) 0.9). Mp: 143-144 °C
1
(lit. 139-141 °C, 145-146 °C). H NMR (500 MHz, CDCl3): δ
7.26 (d, J ) 11.2 Hz, 1H), 7.24-7.18 (m, 1H), 6.55 (dd, J ) 15.0,
11.5 Hz, 1H), 5.92 (d, J ) 14.9 Hz, 1H), 3.72 (s, 3H), 3.31 (d, J
) 4.70 Hz, 1H), 3.08 (dd, J ) 13.8, 6.5 Hz, 1H), 2.90 (dd, J )
15.0, 6.3 Hz, 1H), 2.75 (dd, J ) 15.2, 8.9 Hz, 1H), 2.61 (dd, J )
15.5, 4.3 Hz, 1H), 2.16-2.10 (m, 1H), 2.13 (s, 3H), 1.96 (s, 3H),
1.92-1.67 (m, 5H), 1.60 (s, 3H). 13C NMR (125 MHz, CDCl3): δ
173.1, 172.8, 169.4, 168.0, 144.5, 141.7, 138.7, 134.4, 127.7, 121.7,
90.1, 83.7, 55.2, 52.0, 49.2, 33.3, 30.7, 28.4, 27.7, 24.3, 21.8, 20.1,
12.6. IR ν (cm-1): 3500-2500 (br m), 2955 (m), 2924 (m), 1740
(s), 1709 (s), 1680 (m), 1642 (m), 1611 (w), 1438 (m), 1415 (m),
1386 (w), 1278 (s), 1259 (s), 1208 (s), 1165 (s), 1072 (m), 1030
(m), 983 (w), 952 (w), 801 (w) 742 (w) (lit. 3500-2700, 1740,
1720, 1682, 1635, 1605). HRMS (ESI): calcd for C23H28O8Na, m/z
455.1682 ([M + Na]+); found, 455.1664.
1
°C. H NMR (500 MHz, CDCl3): δ 7.23-7.17 (m, 1H), 3.70 (s,
3H), 2.85 (ddt, J ) 15.4, 6.2, 1.6 Hz, 1H), 2.66 (dd, J ) 14.7, 8.6
Hz, 1H), 2.59-2.45 (m, 3H), 2.28-2.18 (m, 1H), 2.17 (d, J ) 6.4
Hz, 1H), 2.08-1.98 (m, 2H), 1.91-1.79 (m, 2H), 1.79 (s, 3H),
1.75 (s, 1H), 0.17 (s, 9H). 13C NMR (125 MHz, CDCl3): δ 173.6,
168.2, 142.7, 134.1, 105.0, 92.1, 80.3, 79.5, 55.0, 54.4, 52.0, 35.0,
33.5, 30.3, 27.2, 26.0, 19.7, -0.3. IR ν (cm-1): 3436 (m), 2957
(m), 2173 (w), 1715 (s), 1650 (w), 1436 (m), 1380 (w), 1347 (w),
1273 (s), 1252 (s), 1205 (m), 1176 (m), 1120 (m), 1076 (m), 1042
(w), 943 (w), 905 (w), 859 (s), 844 (s), 770 (w), 734 (w), 602 (w).
HRMS (ESI): calcd for C20H28O5NaSi, m/z 399.1604 ([M + Na]+);
found, 399.1600.
Pseudolaric Acid B (1a). A solution of acetylene 49 (9.0 mg, 26
µmol, 1.0 equiv) and Pd(PPh3)2Cl2 (1.0 mg, 1.4 µmol, 0.050 equiv)
in THF (0.3 mL) was degassed with two freeze-thaw cycles.
Bu3SnH (distilled, 12 µL, 45 µmol, 1.7 equiv) was added dropwise,
during which the color changed from yellow to orange. After 15
min, the solvent was removed under HV, and the crude product
was purified via short column chromatography (SiO2, deactivated
with NEt3, 1:0-1:1-0:1 PET/DCM) to give crude stannane 50 (15
mg, 23 µmol, 90%, 90% pure by 1H NMR), which was used
immediately in the next step.
Acknowledgment. We thank the NSF and the NIH (GM13598)
for their generous support of our programs. J.W. thanks the Swiss
National Science Foundation and the Roche Research Foundation
for postdoctoral fellowships. A.M. thanks the German Academic
Exchange Service (DAAD) for a postdoctoral fellowship. Mass
spectra were provided by the Mass Spectrometry Regional Center
of the University of California, San Francisco, which is supported
by the NIH Division of Research Resources. We thank Johnson
Matthey for a generous supply of palladium and ruthenium salts.
In accord with a reported procedure,26 stannane 50 (17 mg, 27
µmol, 1.0 equiv) was dissolved in NMP (distilled, 0.4 mL), and
Hunig’s base (distilled, 30 µL, 0.17 mmol, 6.3 equiv) was added,
followed by iodide 51 (12 mg, 57 µmol, 2.1 equiv). The reaction
mixture was degassed with two freeze-thaw cycles, and Pd2dba3
(3.5 mg, 3.8 µmol, 0.28 equiv) was added. The reaction was
monitored via TLC (5:1 PET/AcOEt, 0.5% AcOH). After 17 h, no
further conversion was observed, and Pd2dba3 (3.5 mg, 3.8 µmol,
Supporting Information Available: Experimental procedures
and spectral data for all of the new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org
JA806724X
9
16434 J. AM. CHEM. SOC. VOL. 130, NO. 48, 2008