Total syntheses of (+)- and (-)-pestalotiopsin A

Article abstract of DOI:10.1021/jo9012546

(Chemical Equation Presented) An enantioselective total synthesis of both enantiomers of caryophyllene-type sesquiterpenoid pestalotiopsin A has been achieved, thereby establishing the absolute stereochemistry of natural (+)-pestalotiopsin A. Highlights o

Full text of DOI:10.1021/jo9012546

pubs.acs.org/joc
Total Syntheses of (þ)- and (-)-Pestalotiopsin A
Ken-ichi Takao,* Nobuhiko Hayakawa, Reo Yamada, Taro Yamaguchi, Hiroshi Saegusa,
Masatoshi Uchida, Suguru Samejima, and Kin-ichi Tadano*
Department of Applied Chemistry, Keio University, Hiyoshi, Kohoku-ku, Yokohama 223-8522, Japan
takao@applc.keio.ac.jp; tadano@applc.keio.ac.jp
Received June 12, 2009
An enantioselective total synthesis of both enantiomers of caryophyllene-type sesquiterpenoid pestalo-
tiopsin A has been achieved, thereby establishing the absolute stereochemistry of natural (þ)-pestalo-
tiopsin A. Highlights of the synthesis include a [2 þ 2] cycloaddition of N-propioloyl Oppolzer’s
camphorsultam and ketene dialkyl acetal and subsequent highly stereoselective 1,4-hydride addition/
protonation, an aldol reaction of functionalized bicyclic lactone with aldehyde, an efficient intramole-
cular Nozaki-Hiyama-Kishi (NHK) reaction for the construction of the highly strained (E)-cyclono-
nene ring, and a palladium-catalyzed reduction of allylic mesylate with retention of the E configuration.
Introduction
The secondary metabolites of endophytic microorganisms
have attracted considerable attention in the scientific commu-
nity since the paclitaxel (Taxol)-producing fungus was isolated
as an endophyte from the Pacific yew (Taxus brevifolia).1 In
1996, Sugawara and co-workers reported the isolation and
structural determination of two new caryophyllene-type ses-
quiterpenoids, (þ)-pestalotiopsin A (1) and (-)-pestalotiop-
sin B (2) (Figure 1), from Pestalotiopsis sp., an endophytic
fungus associated with the bark and leaves of Taxus brevifo-
lia.2 Later, (þ)-pestalotiopsin C (3), a congener of 1 and 2, was
also isolated from the same fungus.3 The structures of 1-3
were determined mainly by the analysis of 1D and 2D NMR
data and FAB-MS. The relative stereochemistries and con-
formations of 1 and 2 were elucidated through single-crystal
X-ray diffraction analysis. The unveiled structure of 1 is the
hitherto unknown oxatricyclic skeleton that is composed of a
cyclobutane ring fused with both an (E)-cyclononene ring and
FIGURE 1. Structures of (þ)-pestalotiopsin A and related com-
pounds.
a γ-lactol unit. From the biological perspective, compound
1 shows immunosuppressive activity in the mixed lymphocyte
reaction and cytotoxicity,2 but the details have not been
reported. In 2003, structurally related natural products pesta-
lotiopsolide A (4), taedolidol (5), and 6-epitaedolidol (6) were
discovered in cultures of Pestalotiopsis sp. obtained from a
(1) Stierle, A.; Strobel, G.; Stierle, D. Science 1993, 260, 214–216.
(2) Pulici, M.; Sugawara, F.; Koshino, H.; Uzawa, J.; Yoshida, S.;
Lobkovsky, E.; Clardy, J. J. Org. Chem. 1996, 61, 2122-2124. Correction:
J. Org. Chem. 1997, 62, 1564.
(3) Pulici, M.; Sugawara, F.; Koshino, H.; Okada, G.; Esumi, Y.; Uzawa,
J.; Yoshida, S. Phytochemistry 1997, 46, 313–319.
6452 J. Org. Chem. 2009, 74, 6452–6461
Published on Web 08/06/2009
DOI: 10.1021/jo9012546
r
2009 American Chemical Society

Takao et al.
JOCFeatured Article
SCHEME 1. Retrosynthetic Analysis
synthesis of (()-caryophyllene by using a ring-contraction
of 13-membered lactam sulfoxide for the construction of
the (E)-cyclononene.11 More recently, an intramolecular Tsu-
ji-Trost reaction was employed for the synthesis of the
structural relative caryophyllene, (-)-antheliolide A, in the
laboratory of Corey.12 To develop new synthetic methods and
strategies for the caryophyllenes, we have focused on two
subjects in studies directed toward the total synthesis
of pestalotiopsin A, i.e., (1) an efficient formation of the
highly oxygenated (E)-cyclononene skeleton and (2) an
asymmetric synthesis of the multisubstituted cyclobutane
derivatives. Our retrosynthetic analysis of pestalotiopsin
A is shown in Scheme 1. As the absolute stereochemistry
was undetermined at the outset of our work on pestalotiop-
sin A, enantiomer 7 was arbitrarily selected as the target.
Since embarking on this research, we have investigated
an efficient method for the direct closure of the (E)-cyclono-
nene skeleton of pestalotiopsin A. However, the following
early attempts were unsuccessful: (a) intramolecular pinacol
coupling reaction for C1-C2 or C4-C5 bond formation,
(b) Dieckmann condensation, intramolecular SmI2-medi-
ated reductive cyclization, and intramolecular R-sulfonyl
anion or protected cyanohydrin anion cyclization for
C2-C3 bond formation, and (c) ring-closing metathesis for
C4-C5 double-bond formation.13 After these experiments,
we assumed that a Nozaki-Hiyama-Kishi (NHK) reaction
approach14 would enable efficient (E)-cyclononene ring forma-
tion. Therefore, we designed a synthetic intermediate 8 carrying
a hydroxy group at C3, which would be synthesized by using
the intramolecular NHK reaction of aldehyde/alkenyl iodide 9.
Deoxygenation of the hydroxy group at C3 in 8 was considered
to be involved in the end game. The assembly of 9 could be
achieved by the aldol reaction of functionalized bicyclic lactone
10 and γ-iodo-β,γ-unsaturated aldehyde 11. The bicyclic
lactone structure of 10 would offer potential advantages in
terms of stereoselectivity in the aldol reaction. This bicyclic
lactone 10, in turn, would be prepared from a multisubstituted
cyclobutane 12. For the preparation of intermediary cyclo-
butane derivative 12 in an enantioenriched form, a program
was launched to develop a new methodology for the asymmetric
synthesis of cyclobutane compounds,15 featuring the [2 þ 2]
cycloaddition of chiral propiolate 13 and ketene acetal 14.
Asymmetric Synthesis of Cyclobutane 17. It has been
reported that some propiolic acid esters feasibly undergo
[2 þ 2] cycloaddition with ketene silyl acetals to provide
cyclobutene derivatives, which were reduced to cyclobutane
derivatives.16 We anticipated that asymmetric induction in
Pinaceae plant, Pinus taeda.4 They could be biosynthesized
from a precursor, such as pestalotiopsins. In the context
described above, the research groups of Procter,5 Paquette,6
8
and Marko, in addition to our group, have been interested in
ꢀ 7
pestalotiopsin A (1) as a synthetic target, especially due to its
formidable structure. Recently, we reported the first total
synthesis of (-)-pestalotiopsin A (7, Scheme 1), thereby
establishing the absolute stereochemistry of natural 1.9 Here-
in, we describe a full account of the total synthesis of pesta-
lotiopsin A, unnatural 7, and finally, natural enantiomer 1.
The evaluation of their biological activities is also presented.
Results and Discussion
Synthetic Strategy. In the landmark total synthesis of
(()-caryophyllene by Corey and co-workers,10 the strained
(E)-cyclononene framework was constructed by Grob frag-
mentation. Ohtsuka, Oishi, and co-workers achieved the total
(4) Magnani, R. F.; Rodrigues-Fo, E.; Daolio, C.; Ferreira, A. G.; de
Souza, A. Q. L. Z. Naturforsch., C: J. Biosci. 2003, 58c, 319–324.
(5) (a) Johnston, D.; Francon, N.; Edmonds, D. J.; Procter, D. J. Org.
ꢀ
Lett. 2001, 3, 2001–2004. (b) Johnston, D.; Couche, E.; Edmonds, D. J.;
Muir, K. W.; Procter, D. J. Org. Biomol. Chem. 2003, 1, 328–337.
(c) Edmonds, D. J.; Muir, K. W.; Procter, D. J. J. Org. Chem. 2003, 68,
3190–3198. (d) Baker, T. M.; Edmonds, D. J.; Hamilton, D.; O’Brien, C. J.;
Procter, D. J. Angew. Chem., Int. Ed. 2008, 47, 5631–5633.
(11) Ohtsuka, Y.; Niitsuma, S.; Tadokoro, H.; Hayashi, T.; Oishi, T.
J. Org. Chem. 1984, 49, 2326–2332.
ꢁ
(6) (a) Paquette, L. A.; Cuniere, N. Org. Lett. 2002, 4, 1927–1929.
(12) Mushti, C. S.; Kim, J.-H.; Corey, E. J. J. Am. Chem. Soc. 2006, 128,
14050–14052.
(13) The Paquette group also reported that construction of the
(E)-cyclononene of pestalotiopsin A via ring-closing metathesis proved to
be difficult. See ref 6d.
(b) Paquette, L. A.; Kang, H.-J. Tetrahedron 2004, 60, 1353–1358.
(c) Paquette, L. A.; Parker, G. D.; Tei, T.; Dong, S. J. Org. Chem. 2007, 72,
7125-7134. Correction: J. Org. Chem. 2009, 74, 1812. (d) Paquette, L. A.;
Dong, S.; Parker, G. D. J. Org. Chem. 2007, 72, 7135–7147. (e) Dong, S.;
Paquette, L. A. Heterocycles 2007, 72, 111–114.
(14) For recent reviews on the synthetic utility of the NHK reaction, see:
€
(a) Wessjohann, L. A.; Scheid, G. Synthesis 1999, 1–36. (b) Furstner, A.
ꢀ
(7) Maulide, N.; Marko, I. E. Chem. Commun. 2006, 1200–1202.
(8) Takao, K.; Saegusa, H.; Tsujita, T.; Washizawa, T.; Tadano, K.
Tetrahedron Lett. 2005, 46, 5815–5818.
Chem. Rev. 1999, 99, 991–1045.
(15) For reviews on the synthesis of optically active cyclobutane deriva-
tives, see: (a) Lee-Ruff, E.; Mladenova, G. Chem. Rev. 2003, 103, 1449–1483.
(9) Takao, K.; Hayakawa, N.; Yamada, R.; Yamaguchi, T.; Morita, U.;
Kawasaki, S.; Tadano, K. Angew. Chem., Int. Ed. 2008, 47, 3426–3429.
(10) (a) Corey, E. J.; Mitra, R. B.; Uda, H. J. Am. Chem. Soc. 1963, 85,
362–363. (b) Corey, E. J.; Mitra, R. B.; Uda, H. J. Am. Chem. Soc. 1964, 86,
485–492. (c) Recently, the enantioselective total synthesis of (-)-caryophyl-
lene was also achieved. See: Larionov, O. V.; Corey, E. J. J. Am. Chem. Soc.
2008, 130, 2954–2955.
€
(b) Butenschon, H. Angew. Chem., Int. Ed. 2008, 47, 3492–3495. For a review
on the application of cyclobutane derivatives in organic synthesis, see:
(c) Namyslo, J. C.; Kaufmann, D. E. Chem. Rev. 2003, 103, 1485–1537.
(16) (a) Quendo, A.; Rousseau, G. Tetrahedron Lett. 1988, 29, 6443–6446.
(b) Mitani, M.; Sudoh, T.; Koyama, K. Bull. Chem. Soc. Jpn. 1995, 68, 1683–
1687. (c) Miesch, M.; Wendling, F. Eur. J. Org. Chem. 2000, 3381–3392.
J. Org. Chem. Vol. 74, No. 17, 2009 6453

Takao et al.
JOCFeatured Article
TABLE 1. Reduction of Cyclobutenes 15 Followed by Removal of the Chiral Auxiliary
yieldsb (%)
entry
cyclo-butene
conditionsa
cis-17
trans-17
Xc-H
ee (%) of 17c
config of 17d
1
2
3
4
5
15aa
15aa
15ab
15ab
15bb
A
B
A
B
B
36
33
51
54
38
14
20
20
38
37
63
69
80
100
28
20
32
50
98
13
R
R
R
S
S
aSee Scheme 2. bIsolated yields from 15aa, 15ab, or 15bb. cDetermined by chiral HPLC analysis after conversion of cis-17 into the corresponding
monotosylate cis-18. In entry 4, the ee of trans-17 was also determined by chiral HPLC analysis and confirmed to be identical to the ee of cis-17.
dAbsolute configuration of C4 in the major enantiomer.
SCHEME 2. Asymmetric Synthesis of Cyclobutanols, cis-17
and trans-17
cases of 15aa under conditions A and B, unsatisfactory
stereoselectivities of the obtained cyclobutane 16aa were
observed (Table 1, entries 1 and 2). Due to a concern that
the asymmetric center at the acetal carbon in 15aa might
cause low stereoselectivity, we next used a symmetrical
ketene, such as bis(trimethylsilyl) acetal 14b,20 as a partner
in the [2 þ 2] cycloaddition. The reaction of 13a and bis-silyl
acetal 14b gave the sole adduct 15ab. Although the hydro-
genation (conditions A) of 15ab showed a moderate stereo-
induction (entry 3), the 1,4-hydride addition/protonation
(conditions B) realized an excellent level to provide cyclo-
butane 16ab in an almost diastereomerically single form
(entry 4). In contrast, the use of Evans’ oxazolidinone as
the chiral auxiliary gave a lower level of asymmetric induc-
tion. The cycloadduct 15bb was reduced under conditions
B to cyclobutane 16bb with only 13% de (entry 5). To obtain
the cyclobutane derivative 12 in Scheme 1, the chiral aux-
iliaries in 16aa, 16ab, or 16bb were removed with lithium
aluminum hydride. These reactions, however, induced the
simultaneous cleavage of the silyl acetal and the subsequent
reduction of the resulting cyclobutanone. In the case of entry
4, two diastereomers of cyclobutanols, cis-17 and trans-17,
were obtained with 98% ee.21 Oppolzer’s camphorsultam
(Xc-H) was quantitatively recovered.
To explain the observed high level of stereoselection in the
tandem 1,4-hydride addition/protonation in the case of
15ab, we reasoned from the well-recognized transition states
argument by Oppolzer et al.22 (Scheme 3). In the more
favorable conformation of 15ab, the carbonyl group directs
anti to the SO2 group and adopts s-trans conformation to the
R,β-unsaturated bond as depicted to avoid a steric repulsion
occurring between bis(trimethylsilyloxy) and the SO2 groups
in the s-cis conformer. Then, the 1,4-hydride addition to the
s-trans conformer generates the Z-enolate intermediate,
which is reorganized to the more stable lithium-chelated
conformer as depicted. A proton approaches predominantly
from the side opposite the bulky auxiliary (from the front
side), giving 16ab almost exclusively.
First-Generation Synthesis of Bicyclic Lactone 24. Having
established the method for the asymmetric synthesis of the
multisubstituted cyclobutanes cis-17 and trans-17, we turned
our attention to the construction of the bicyclic lactone 24
(Scheme 4). Toward this end, we examined one-carbon
homologation of the hydroxymethyl group at C4 in cis-17
these reaction sequences would be realized by using a chiral
auxiliary strategy. Therefore, two chiral propiolic acid deri-
vatives, Oppolzer’s (1S)-camphorsultam 13a and Evans’
oxazolidinone 13b, both incorporating an N-propioloyl
group, were prepared according to the procedure described
in the literature17 (Scheme 2). Rousseau and Quendo re-
ported that ethyl propiolate reacted with dimethylketene
methyl trimethylsilyl acetal (14a) in the presence of a cata-
lytic amount of ZrCl4 in dichloromethane to provide a
cyclobutene derivative as a result of [2 þ 2] cycloaddition.16a
Under the same conditions, the reaction of 13a with ketene
silyl acetal 14a proceeded to produce cyclobutene 15aa as a
mixture of diastereomers (ca. 3:2). Asymmetric induction at
C(R) in the reduction of a C-C double bond of the resulting
cyclobutenecarbonyl amide 15aa was examined by the addi-
tion of hydrogen in the presence of palladium on carbon
(conditions A)18 or 1,4-hydride addition with lithium tri-sec-
butyl borohydride (L-Selectride) in toluene and subsequent
protonation of the resulting enolate (conditions B).19 In the
(20) (a) Ainsworth, C.; Kuo, Y.-N. J. Organomet. Chem. 1972, 46, 73–87.
(b) Schulz, W. J., Jr.; Speier, J. L. Synthesis 1989, 163–166.
(21) For assignment of the absolute stereochemistry at C4 in the en-
antioenriched (98% ee) cis-17 and trans-17, see the Supporting Information.
(22) Oppolzer, W.; Poli, G.; Starkemann, C.; Bernardinelli, G. Tetrahe-
dron Lett. 1988, 29, 3559–3562.
ꢁ
(17) Fonquerna, S.; Moyano, A.; Pericas, M. A.; Riera, A. Tetrahedron:
Asymmetry 1997, 8, 1685–1691.
ꢀ
(18) Oppolzer, W.; Mills, R. J.; Reglier, M. Tetrahedron Lett. 1986, 27,
183–186.
(19) Oppolzer, W.; Poli, G. Tetrahedron Lett. 1986, 27, 4717–4720.
6454 J. Org. Chem. Vol. 74, No. 17, 2009

Takao et al.
JOCFeatured Article
SCHEME 3. Plausible Mechanism for the 1,4-Hydride Addi-
tion/Protonation of 15ab
SCHEME 4. First-Generation Synthesis of Bicyclic Lactone 24
and trans-17. Selective tosylation of the primary hydroxy
group in cis-17 was easily achieved, whereas the selectivity on
trans-17 was modest. Treatment of monotosylate cis-18 or
trans-18 with potassium cyanide in hot dimethyl sulfoxide
unexpectedly caused a ring-opening reaction involving the
elimination of the tosyloxy group, i.e., a Grob fragmenta-
tion, to form 2,2-dimethylpent-4-en-1-al, which was at-
tacked by a cyanide ion to produce the corresponding
cyanohydrin in significant yields of 30-40%. Therefore,
the secondary hydroxy group in cis-18 and trans-18 was
temporally protected as ethoxyethyl (EE) ether. The result-
ing cis-19 and trans-19 were mixed at this stage and moved
forward together. The tosyloxy group in the mixture 19 was
replaced by a cyano group to provide nitrile 20. Brief
exposure of 20 to hydrochloric acid hydrolyzed the ethoxy-
ethyl ether to regenerate secondary alcohol 21, which was
oxidized with Dess-Martin periodinane23 to cyclobutanone
22. The Grignard reaction of 22 with vinylmagnesium bro-
mide in toluene occurred as anticipated from the more
accessible β-face to provide vinyl-adduct 23 as a single
isomer. Acid hydrolysis of the nitrile group accompanied
by γ-lactone formation provided bicyclic lactone 24. The
synthesis arrived at bicyclic lactone 24, but our first-genera-
tion synthetic route was, rather, a multistep process includ-
ing (1) the formation of the two diastereomers cis-17 and
trans-17 and (2) the protection of the secondary hydroxy
group in cis-18 and trans-18 for the homologation. These
disadvantages were originally caused by the high reactivity of
the bis(trimethylsilyl) acetal moiety in 16ab to LiAlH4 under
the conditions used for the removal of the camphorsultam.
Second-Generation Strategy. To circumvent a long synth-
esis, an improved protocol was developed by using ketene
dialkyl acetal 14c as the partner for the [2 þ 2] cycloaddition
of 13a (Scheme 5). Compound 14c was prepared in batches
of 25 g from methacrolein diethyl acetal (25) by olefin
isomerization with lithium/ethylenediamine.24 The Lewis
acid-catalyzed [2 þ 2] cycloaddition of 13a with 14c afforded
a single regioisomeric adduct 26 in a slightly higher yield
(85%). The 1,4-hydride addition/protonation of cyclobutene
26 with L-Selectride showed extremely high stereoselectivity
again, providing cyclobutane derivative 27 quantitatively.
Treatment of 27 with lithium aluminum hydride pro-
vided cyclobutanemethanol 12 (R=Et in Scheme 1) in 94%
yield for two steps with an excellent enantiomeric excess
of >95%,25 and camphorsultam was recovered efficiently
in a multigram-scale experiment. Conversion of 12 into
nitrile 28 through the tosylate and subsequent acid hydro-
lysis of the dialkyl acetal moiety in 28 provided cyclo-
butanone 22, which was subjected to the same two-step
sequence with the first-generation synthesis to arrive at
bicyclic lactone 24. The second-generation approach enables
more convenient access to the key intermediate 24 (29%
overall yield for 8 steps from 13a) than the previous approach
(16% overall yield for 12 steps).
With bicyclic lactone 24 in hand, we turned to the dihy-
droxylation of the vinyl group. We initially assumed that
Sharpless asymmetric dihydroxylation26 of 24 would provide
the desired (R)-diastereomer (Scheme 6). However, treat-
ment of 24 with AD-mix-R preferentially produced the
undesired (S)-isomer. After selective O-silylation of the
primary hydroxy group in the dihydroxylation products,
the diastereomeric products 29 and 30 were obtained in a
ratio of 1:4. This stereochemical propensity was also ob-
served in the case of using AD-mix-β (29:30 = 1:2) or achiral
reagents OsO4/NMO (29:30 = 1:4). We, therefore, explored
(23) (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155–4156.
(b) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277–7287.
(c) Ireland, R. E.; Liu, L. J. Org. Chem. 1993, 58, 2899.
(25) The ee of 12 was determined by 1H NMR analysis of the Mosher
esters derived from 12.
(24) For the conventional method for the synthesis of 14c from triethyl
orthoisobutyrate, see: McElvain, S. M.; Davie, W. R. J. Am. Chem. Soc.
1951, 73, 1400–1402.
(26) For a review on Sharpless asymmetric dihydroxylation, see: Kolb,
H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev. 1994, 94, 2483–
2547.
J. Org. Chem. Vol. 74, No. 17, 2009 6455

Takao et al.
JOCFeatured Article
SCHEME 5. Second-Generation Synthesis of Bicyclic Lactone
24
SCHEME 6. Synthesis of Lactone 10
SCHEME 7. Synthesis of Aldehyde 11
a method for the reversal of the stereochemistry in the major
isomer 30. Fortunately, (S)-isomer 30 was converted into the
desired 29 by an oxidation/reduction strategy. Thus, the
Dess-Martin oxidation of 30, followed by the sodium
borohydride reduction of the resulting ketone, provided
(R)-isomer 29 in a high overall yield of 90%. The substrate
for the planned aldol was protected with (4-methoxyphenyl)-
methyl (MPM) imidate to provide the functionalized bicyclic
lactone 10.
Synthesis of (E)-Cyclononene Skeleton. The γ-iodo-β,γ-
unsaturated aldehyde 11, the coupling partner for the aldol
reaction of 10, was synthesized from D-glyceradehyde acet-
onide (31) (Scheme 7). Although the additions of metal
acetylide reagents to 31 have been studied,27 stereochemical
control is still difficult. Therefore, we adopted Takano’s
protocol28 to prepare the chiral propargylic alcohol. Treat-
ment of the known epoxy chloride 32 prepared from 31 with
3 equiv of n-butyllithium resulted in the formation of the
intermediary dianion, which was trapped with iodomethane
in the presence of hexamethylphosphoramide (HMPA) to
afford the C,O-methylated propargyl alcohol 33. Hydrolysis
of the acetonide group in 33, followed by a palladium-
catalyzed hydrostannation of the resulting diol 34 using
Bu3SnH/Pd(PPh3)2Cl2,29 gave the terminally stannylated
olefin 35 and its regioisomer 36 in ratio of 3.4:1. The
regioselectivity of 35:36 has now been greatly improved by
the stannylcupration methodology identified in a later ex-
periment for the synthesis of 1 (vide infra). The metal-
halogen exchange of 35 with iodine provided alkenyl iodide
37, which was subjected to an oxidative cleavage using silica
gel-supported sodium metaperiodate30 to provide essentially
pure aldehyde 11 in a good yield.
With bicyclic lactone 10 and aldehyde 11 in hand, we next
focused on the coupling of these partners by an aldol reaction
(Scheme 8). On the basis of our previous experiment per-
formed by using structurally analogous coupling partners,8
the aldol reaction of 10 with 11 was examined by use of
sodium bis(trimethylsilyl)amide (NaHMDS) as a base. As
expected, this reaction provided the anti-aldol 38 predomi-
nantly (38:39 = 3:1), securing the two contiguous stereo-
genic centers required for the target molecule. The approach
of aldehyde 11 occurred exclusively from the convex face of
the bicyclic skeleton of 10. Other bases were screened for the
generation of the enolate (LiHMDS and KHMDS), but no
better result was found.
(27) (a) Mulzer, J.; Greifenberg, S.; Beckstett, A.; Gottwald, M. Liebigs
Ann. Chem. 1992, 1131–1135. (b) Mulzer, J.; Funk, G. Synthesis 1995, 101–
112.
(28) Takano, S.; Samizu, K.; Sugihara, T.; Ogasawara, K. J. Chem. Soc.,
Chem. Commun. 1989, 1344–1345.
ꢀ
(29) (a) Zhang, H. X.; Guibe, F.; Balavoine, G. J. Org. Chem. 1990, 55,
1857–1867. (b) For a review on the metal-catalyzed hydrostannation, see:
Smith, N. D.; Mancuso, J.; Lautens, M. Chem. Rev. 2000, 100, 3257–3282.
(30) Zhong, Y.-L.; Shing, T. K. M. J. Org. Chem. 1997, 62, 2622–2624.
6456 J. Org. Chem. Vol. 74, No. 17, 2009

Takao et al.
JOCFeatured Article
SCHEME 8. Aldol Reaction of Lactone 10 and Aldehyde 11
SCHEME 9. Intramolecular NHK Reaction of 45
The next stage was to construct the formidable (E)-cyclo-
nonene ring. The initial route was designed to form the
γ-lactol moiety prior to the NHK reaction (Scheme 9).
Protection of the secondary alcohol 38 as methoxymethyl
(MOM) ether, followed by reduction of the resulting 40 with
diisobutylaluminum hydride (DIBAL-H), provided γ-lactol
41 as an anomeric mixture (ca. 2:1). Acetalization of 41 with
trimethyl orthoformate and a catalytic amount of pyridi-
nium p-toluenesulfonate (PPTS) in methanol gave R-isomer
42 and β-isomer 43 in 75% and 10% yields, respectively. The
predominantly obtained isomer 42 was treated with tetra-
butylammonium fluoride (TBAF) to provide the desilylated
alcohol 44, which was oxidized to aldehyde 45 with Dess-
Martin periodinane. Treatment of a 0.005 M solution of 45 in
dimethyl sulfoxide with CrCl2 and catalytic NiCl2 effected an
intramolecular NHK reaction that provided the cyclized
product 46 in 50% yield as a single diastereomer. The yield
of the cyclized product 46 was moderate because the inter-
mediary metalated olefin underwent competitive proton-
ation during the NHK reaction, thereby forming an
uncyclized product 47 in 26% yield. The stereochemistry
and conformation of 46 were determined by NOE experi-
ments as shown. The NOE relations suggest that compound
46 exists in solution as a βR-conformer,31 similarly to the
natural product 1. The ββ-conformational isomer was not
observed in NMR analysis of 46 at room temperature.
In an effort to optimize the ring closure process, we next
examined the intramolecular NHK reaction of the γ-lactone
corresponding to 45 (Scheme 10). The requisite substrate
for the reaction, aldehyde/alkenyl iodide 9, was available
via a desilylation of 40 and subsequent oxidation of the
resulting 48. We were pleased to find that the intramolecular
NHK reaction of 9 provided the cyclized product 8 in 92%
yield as a single diastereomer and as a single atropisomer
(βR-conformer), as determined by NOE analysis. Consider-
ing the highly strained structure of (E)-cyclononene, the high
yield of 8 was remarkable.
obtained; the geometrical isomerization of the trisubstituted
(E)-olefin moiety occurred. For example, when the methyl
xanthate ester of 8 was treated with tributyltin hydride in
refluxing toluene, the desired deoxygenation proceeded,
accompanied by a complete isomerization of the olefin
geometry, resulting in the quantitative formation of the
Z-isomer. Moreover, the MOM group at C7-OH was
difficult to remove under acidic conditions in the final step.33
We were, therefore, forced to explore an alternative strategy
that could avoid the isomerization of the (E)-olefin moiety as
well as an easily removable protecting group on C7-OH.
These encountered difficulties were solved by using a palla-
dium-catalyzed reduction strategy34 and by switching the
MOM group to the triethylsilyl (TES) group, as shown in
Scheme 11. Leaving group investigation in the palladium-
catalyzed reduction, a mesylate was found to be suitable in
Completion of the Synthesis of (-)-Pestalotiopsin A (7).
The final stage of the synthesis required the removal of the
extra hydroxy group introduced in 8. Preliminarily, we
attempted the radical-induced deoxygenation of 8 utilizing
the Barton method.32 The anticipated product was not
(31) Using the terminology of Shirahama et al., see: Shirahama, H.;
Osawa, E.; Chhabra, B. R.; Shimokawa, T.; Yokono, T.; Kanaiwa, T.;
Amiya, T.; Matsumoto, T. Tetrahedron Lett. 1981, 22, 1527–1528.
(32) (a) Barton, D. H. R.; McCombie, S. W. J. Chem. Soc., Perkin Trans.
1 1975, 1574–1585. (b) For a review on the radical deoxygenation of alcohols,
see: Hartwig, W. Tetrahedron 1983, 39, 2609–2645.
(33) It was reported that the caryophyllene skeleton is somewhat sensitive
ꢀ
´
as-Sanchez, A. J. Nat. Prod. Rep.
to acids: Collado, I. G.; Hanson, J. R.; Macı
1998, 15, 187–204.
(34) For a review on palladium-catalyzed hydrogenolysis, see: Tsuji, J.;
Mandai, T. Synthesis 1996, 1–24.
J. Org. Chem. Vol. 74, No. 17, 2009 6457

Takao et al.
JOCFeatured Article
SCHEME 10. Intramolecular NHK Reaction of 9
SCHEME 11. Completion of the Synthesis of (-)-Pestalotiop-
sin A (7)
this case. Thus, the cyclized product 8 was converted into
allylic mesyl ester 49. Deprotection of the MOM group in 49
was proved successful by use of standard acidic hydrolysis
conditions. Silylation of the resulting alcohol 50, followed by
cleavage of the MPM ether in 51, afforded the TES ether 52.
Treatment of 52 with a combination system of Pd2(dba)3/
nBu3P and NaBH435 provided the desired deoxygenated pro-
duct 53 in a practical manner,36 accompanied by a small
amount (ca. 8%) of the C3-C4 olefin isomer. Use of
2-epi-pestalotiopsin A (64) (Scheme 12). As described in
Scheme 6, diastereomer 30 was obtained as the major isomer
in the step for the dihydroxylation of 24. We considered that
a similar reaction sequence used for the total synthesis of
7 would lead to 64 starting with compound 30. This was
eventually accomplished as shown in Scheme 12. The aldol
reaction of the bicyclic lactone 56 derived from 30 and the
aforementioned aldehyde 11 using NaHMDS provided a
diastereomeric mixture of aldol adducts, which were pro-
tected as the MOM ether to afford anti-isomer 57 and syn-
isomer in a ratio of 2:1. Removal of the silyl group in 57 and
subsequent oxidation led to aldehyde/alkenyl iodide 58. The
intramolecular NHK reaction of 58 required heating at 50 °C
to effect complete conversion. In this case, two diastereo-
mers, 59 and 60, were obtained as cyclized products in 73%
and 14% yields, respectively.38 The major isomer 59 was
converted into allylic mesylate 61, which was subjected to the
palladium-catalyzed reduction to provide the deoxygenated
product 62. DIBAL-H reduction of 62 showed low conver-
sion, but the unreacted 62 was recovered well as a reusable
material. After acetylation, diacetate 63 was obtained in 83%
yield based on the recovered starting material. Unlike the
previous synthesis for 7, attempted deprotection of 63 failed
under acidic hydrolysis conditions. We therefore turned to a
two-step sequence for the final deprotection. In the first step,
the TES group was removed with TBAF, and subsequent
37
HCO2NH4 as a hydride source in place of NaBH4 in this
reaction resulted in the preferential formation of the undesired
C3-C4 olefin isomer. DIBAL-H reduction of 53 provided γ-
lactol 54 as a sole R-anomer, while the C3-C4 olefinic isomer
obtained in the former reaction was removed at this stage.
Acetylation of 54 provided diacetate 55, and the mild acid
hydrolysis allowed clean deprotection of the TES group and
the acetyl group in the γ-lactol moiety to provide (-)-pesta-
lotiopsin A (7). The synthetic sample was identical in all
1
respects (mp, TLC, IR, H and 13C NMR, and HRMS) to
those of natural pestalotiopsin A, except for the sign of optical
rotation [[R]21D -74.7 (c 0.535, MeOH) for 7; lit.2 [R]22D þ76.8
(c 1, MeOH) for the natural sample]. This fact verified that we
had synthesized the antipode of natural pestalotiopsin A.
Synthesis of 2-epi-Pestalotiopsin A (64). We next directed
our efforts to the synthesis of the unnatural congener
(35) Hutchins, R. O.; Learn, K.; Fulton, R. P. Tetrahedron Lett. 1980, 21,
27–30.
(36) Deoxygenation of 51 under the same conditions used for 52 provided
a complex mixture. The hydroxy group at C2 in 52 might play an important
role in the observed high regioselectivity of the hydride attack. For similar
assistance of the neighboring hydroxy group in the palladium-catalyzed
displacement of allylic substrates by hydride, see: (a) Oshima, M.; Yamazaki,
H.; Shimizu, I.; Nisar, M.; Tsuji, J. J. Am. Chem. Soc. 1989, 111, 6280–6287.
(b) Ono, N.; Hamamoto, I.; Kamimura, A.; Kaji, A. J. Org. Chem. 1986, 51,
3734–3736.
(37) Takahashi, T.; Nakagawa, N.; Minoshima, T.; Yamada, H.; Tsuji, J.
Tetrahedron Lett. 1990, 31, 4333–4336.
(38) The stereochemistries of 59 and 60 were determined by NOE experi-
ments; see: Supporting Information.
6458 J. Org. Chem. Vol. 74, No. 17, 2009

Takao et al.
JOCFeatured Article
SCHEME 12. Synthesis of 2-epi-Pestalotiopsin A (64)
SCHEME 13. Synthesis of (þ)-Pestalotiopsin A (1)
stannylcupration/protonation protocol was found to be
highly efficient. Although the moderate regioselectivity of
3.4:1 was observed in our early experiment, as shown in
Scheme 7, treatment of ent-34 with (n-Bu)3Sn(n-Bu)Cu-
CNLi240 in THF/MeOH exhibited a perfect regiocontrol to
produce the desired ent-35 exclusively. All other reaction
sequences proceeded uneventfully, and we ultimately
achieved the total synthesis of (þ)-pestalotiopsin A (1) with
an optical rotation [[R]22.5D þ75 (c 0.26, MeOH)], consistent
with that reported for the natural sample.
Biological Activities. We conducted biological studies with
synthetic samples (þ)-pestalotiopsin A (1), (-)-pestalotiop-
sin A (7), and 2-epi-pestalotiopsin A (64). In the assay for
activity against P388 murine leukemia cells in vitro, natural
1 and unnatural 7 were both found to show cytotoxicity with
IC50s (μg/mL) of 1.3 for 1 and 1.6 for 7. Compound 64
showed slightly weaker activity with an IC50 value of 2.7 μg/
mL. These results suggest that the cyctotoxic property of
pestalotiopsin A is nonenantiospecific.
treatment with 0.5 equiv of sodium methoxide in methanol
provided 2-epi-pestalotiopsin A (64) as an anomeric mixture
of 5:1; the compound had a specific rotation of [R]24 þ14
D
(c 0.35, MeOH) and showed spectroscopic data (1H and 13
C
Conclusions
NMR) that were distinguishable from those of pestalotiop-
sin A.
We have achieved the enantioselective total syntheses of
both (þ)- and (-)-pestalotiopsin A by using the correspond-
ing enantiomers of Oppolzer’s camphorsultam. Key features
of this synthetic venture include (1) a [2 þ 2] cycloaddition of
N-propioloyl camphorsultam 13a and ketene dialkyl acetal
14c followed by stereoselective 1,4-hydride addition/pro-
tonation to provide highly enantioenriched cyclobutane
derivative 12, (2) the aldol reaction of bicyclic lactone 10
with aldehyde 11 to assemble all the skeletal carbons, (3)
the intramolecular NHK reaction of 9 for the construction
of the highly strained (E)-cyclononene ring, and (4) the
Synthesis of (þ)-Pestalotiopsin A (1). With a successful
synthesis of the enantiomer of the natural product com-
pleted, we finally turned to the synthesis of natural (þ)-
pestalotiopsin A (1) (Scheme 13). We repeated all of the steps
of the synthesis, starting from (1R)-camphorsultam (in place
of the 1S-isomer) and L-glyceraldehyde acetonide (ent-31),
both of which were conveniently available.39 Upon opt-
mization for the hydrostannation of alkyne ent-34, the
(39) (1R)-Camphorsultam was prepared from (1R)-camphorsulfonic
acid. L-Glyceraldehyde acetonide (ent-31) was prepared from L-ascorbic
acid; see (a) Hubschwerlen, C.; Specklin, J.-L.; Higelin, J. Org. Synth.
1995, 72, 1-5 (Coll. Vol. IX, 1998, 454). (b) Jung, M. E.; Shaw, T. J. J. Am.
Chem. Soc. 1980, 102, 6304–6311.
(40) Betzer, J.-F.; Delaloge, F.; Muller, B.; Pancrazi, A.; Prunet, J. J. Org.
Chem. 1997, 62, 7768–7780.
J. Org. Chem. Vol. 74, No. 17, 2009 6459

Takao et al.
JOCFeatured Article
palladium-catalyzed reduction of allylic mesylate 52 without
destruction of the E configuration. This work has established
the absolute stereochemistry of natural (þ)-pestalotiopsin
A (1). In addition, our synthetic venture has provided
opportunities to access a related analogue and to investigate
the biological activities of pestalotiopsins.
CDCl3) δ 19.1, 23.1, 25.0, 26.9 (3C), 28.6, 34.2, 35.7, 41.0, 45.2,
55.2, 56.7, 61.5, 70.8, 71.8, 79.4, 80.5, 92.4, 99.5, 113.6 (2C),
127.9 (4C), 129.2 (2C), 130.0 (2C), 130.1, 132.5, 133.0, 135.6
(2C), 135.7 (2C), 139.2, 159.1, 180.3; HRMS (EI) calcd for
C36H42O7SiI (Mþ - t-C4H9) m/z 741.1745, found 741.1754.
Intramolecular NHK Reaction of 9: Synthesis of 8. The
following reaction was carried out under Ar. To a stirred
solution of NiCl2 (2.5 mg, 0.019 mmol) and CrCl2 (285 mg,
2.32 mmol) in degassed DMSO (56 mL) was added a solution of
9 (184 mg, 0.305 mmol) in degassed DMSO (5 mL). The mixture
was stirred for 19 h, diluted with saturated aqueous NH4Cl
(60 mL), and extracted with EtOAc (50 mL ꢀ 3). The combined
extracts were washed with H2O (30 mL ꢀ 2) and saturated brine
(20 mL), dried, and concentrated under reduced pressure. The
residue was purified by column chromatography on silica gel
(EtOAc/hexane, 1:2) to provide 134 mg (92%) of 8 as a colorless
Experimental Section
[2 þ 2] Cycloaddition of 13a and 14c: Synthesis of 26. To a stirred
solution of 13a (8.88 g, 33.2 mmol) in CH2Cl2 (120 mL) were added
ZrCl4 (387 mg, 1.66 mmol) and 14c (4.8 mL, 33 mmol). The mixture
was refluxed while 14c (4.8 mL, 33 mmol) was added six times over a
period of 6 days. The mixture was diluted with EtOAc (300 mL) and
washed with saturated aqueous NaHCO3 (200 mL ꢀ 3). The
organic layer was dried and concentrated under reduced pressure.
The residue was purified by column chromatography on silica gel
(EtOAc/hexane, 1:8) to provide 11.6 g (85%) of 26 as white
oil: TLC Rf 0.28 (EtOAc:hexane, 1:1); [R]20 -118 (c 0.740,
D
CHCl3); IR 3460, 2940, 1780 cm-1 1H NMR (300 MHz,
;
CDCl3) δ 1.11 (s, 3H), 1.33 (s, 3H), 1.40 (dd, 1H, J = 5.1,
12.6 Hz), 1.91 (br s, 3H), 2.16 (dd, 1H, J = 10.2, 12.6 Hz), 2.37
(d, 1H, J = 1.9 Hz, OH), 2.72 (m, 1H), 2.97 (m, 1H), 3.27 (s, 3H),
3.35 (s, 3H), 3.79 (s, 3H), 3.98 (d, 1H, J = 9.0 Hz), 4.07 (dd, 1H,
J = 5.6, 11.6 Hz), 4.25 (dd, 1H, J = 2.4, 5.6 Hz), 4.62 (br d, 1H,
J = 9.0 Hz), 4.65 (d, 1H, J = 11.1 Hz), 4.72 (d, 1H, J = 6.8 Hz),
4.74 (d, 1H, J = 11.1 Hz), 4.83 (d, 1H, J = 6.8 Hz), 5.26 (br d,
1H, J = 11.6 Hz), 6.87 (d, 2H, J = 8.1 Hz), 7.27 (d, 2H, J = 8.1
Hz); 13C NMR (68 MHz, CDCl3) δ 12.9, 24.7, 27.1, 34.4, 41.8,
42.8, 54.9, 55.3, 55.6, 56.4, 69.0, 75.0, 80.2, 81.3, 84.3, 92.6, 97.0,
113.9 (2C), 123.5, 128.7 (2C), 130.6, 140.9, 159.1, 178.7; HRMS
(EI) calcd for C26H36O8 (Mþ) m/z 476.2410, found 476.2412.
Palladium-Catalyzed Reduction of 52: Synthesis of 53. The
following reaction was carried out under Ar. A solution of
catalyst was prepared by mixing Pd2(dba)3 and n-Bu3P in
degassed 1,4-dioxane at 40 °C for 10 min. To a stirred solution
of 52 (12.8 mg, 0.0252 mmol) in degassed 1,4-dioxane (1 mL)
were added NaBH4 (15.5 mg, 0.410 mmol) and a solution of the
premixed catalyst (2.5 μmol for Pd2(dba)3, 11 μmol for n-Bu3P)
in 1,4-dioxane (0.22 mL). The mixture was stirred for 6 h,
quenched with H2O (0.1 mL), and filtered through a pad of
Celite. The filtrate was diluted with saturated brine (10 mL) and
extracted with EtOAc (5 mL ꢀ 3). The combined extracts were
dried and concentrated under reduced pressure. The residue was
purified by column chromatography on silica gel (EtOAc/
hexane, 1:6) to provide 9.0 mg (87%) of an inseparable mixture
(ca. 10:1) of 53 and its regioisomer as white solids: TLC Rf 0.44
(EtOAc/hexane, 1:2); IR 3390, 2960, 1760 cm-1; 1H NMR (300
MHz, CDCl3) δ 0.56-0.67 (m, 6H), 0.92-0.97 (m, 9H), 1.08 (s,
3H), 1.32 (s, 3H), 1.34 (dd, 1H, J = 5.7, 12.6 Hz), 1.53 (d, 1H,
J = 11.1 Hz, OH), 1.87 (d, 3H, J = 1.1 Hz), 2.16 (dd, 1H, J =
9.6, 12.6 Hz), 2.38 (dd, 1H, J = 11.1, 11.2 Hz), 2.59 (dd, 1H, J =
5.6, 11.2 Hz), 2.66 (dd, 1H, J = 2.6, 2.8 Hz), 2.76 (ddd, 1H, J =
2.6, 5.7, 9.6 Hz), 3.21 (s, 3H), 3.78 (dd, 1H, J = 5.6, 11.8 Hz),
4.23 (dt, 1H, J = 5.6, 11.1 Hz), 4.30 (dd, 1H, J = 2.8, 5.6 Hz),
5.02 (qd, 1H, J = 1.1, 11.8 Hz); 13C NMR (68 MHz, CDCl3) δ
4.7 (3C), 6.7 (3C), 18.2, 24.6, 26.6, 33.5, 42.4, 42.7, 45.6, 56.1,
59.2, 69.7, 78.8, 83.4, 93.9, 125.2, 137.5, 178.7; HRMS (EI) calcd
for C22H38O5Si (Mþ) m/z 410.2489, found 410.2493.
crystals: mp 79-83 °C; TLC Rf 0.59 (EtOAc/hexane, 1:3); [R]24
D
-23.6 (c 0.290, CHCl3); IR 2980, 1680 cm-1; 1H NMR (300 MHz,
CDCl3) δ 0.98 (s, 3H), 1.18 (t, 3H, J = 7.1 Hz), 1.19 (s, 3H), 1.23 (s,
3H), 1.24 (t, 3H, J = 7.1 Hz), 1.25 (s, 3H), 1.34-1.43 (m, 2H),
1.86-1.93 (m, 3H), 2.03-2.05 (m, 2H), 3.41 (d, 1H, J = 13.7 Hz),
3.51 (d, 1H, J = 13.7 Hz), 3.52-3.75 (m, 4H), 4.01 (t, 1H, J =
6.1 Hz), 7.16 (s, 1H); 13C NMR (75 MHz, CDCl3) δ 15.1, 15.5,
19.8, 21.0, 21.4, 21.9, 26.4, 33.0, 38.5, 45.0, 47.6, 48.1, 53.5, 53.8,
59.1, 60.5, 65.5, 105.5, 136.6, 159.3, 161.9; HRMS (EI) calcd for
C21H33NO5S (Mþ) m/z 411.2079, found 411.2078.
Aldol Reaction of 10 and 11: Synthesis of 38 and 39. The
following reaction was carried out under Ar. To a cooled
(-78 °C) stirred solution of 10 (364 mg, 0.650 mmol) in THF
(5 mL) was added NaHMDS (1.0 M solution in THF, 2.4 mL,
2.4 mmol). The mixture was stirred at -78 °C for 15 min, and a
solution of 11 (393 mg, 1.63 mmol) in THF (4 mL) was added.
After being stirred at -78 °C for 15 min, the mixture was
warmed to ambient temperature over 5 min, diluted with
saturated aqueous NH4Cl (60 mL), and extracted with EtOAc
(30 mL ꢀ 3). The combined extracts were dried and concen-
trated under reduced pressure. The residue was purified by
column chromatography on silica gel (EtOAc/toluene, 1:70) to
provide 266 mg (51%) of 38, 90.7 mg (17%) of 39, and 42.6 mg
(12%) of recovered 10. Compound 38 was obtained as a color-
less oil: TLC Rf 0.74 (EtOAc/toluene, 1:8); [R]23D -45.1 (c 1.35,
CHCl3); IR 3490, 2940, 1770 cm-1 1H NMR (300 MHz,
;
CDCl3) δ 0.87 (s, 3H), 1.05 (s, 9H), 1.23 (s, 3H), 1.39 (dd, 1H,
J = 6.8, 12.0 Hz), 1.98 (dd, 1H, J = 9.2, 12.0 Hz), 2.40 (br, 1H,
OH), 2.65 (d, 3H, J = 1.3 Hz), 2.93-2.98 (m, 2H), 3.34 (s, 3H),
3.47 (br d, 1H, J = 9.1 Hz), 3.82 (s, 3H), 3.88 (dd, 1H, J = 8.3,
10.7 Hz), 3.96 (dd, 1H, J = 2.1, 10.7 Hz), 4.07 (dd, 1H, J = 2.1,
8.3 Hz), 4.30 (dd, 1H, J = 9.1, 9.6 Hz), 4.69 (d, 1H, J = 11.1 Hz),
4.98 (d, 1H, J = 11.1 Hz), 5.91 (qd, 1H, J = 1.3, 9.6 Hz), 6.88 (d,
2H, J = 8.5 Hz), 7.28 (d, 2H, J = 8.5 Hz), 7.31-7.43 (m, 6H),
7.69-7.73 (m, 4H); 13C NMR (68 MHz, CDCl3) δ 19.1, 24.5,
26.1, 26.9 (3C), 29.1, 36.3, 39.0, 41.5, 50.0, 55.3, 56.5, 64.5, 73.9,
74.4, 78.7, 82.9, 92.4, 103.3, 113.5 (2C), 127.6 (2C), 127.7 (2C),
128.9 (2C), 129.5, 129.6 (2C), 131.7, 133.4, 135.7 (4C), 137.1,
158.8, 177.4; HRMS (EI) calcd for C36H42O7SiI (Mþ - t-C4H9)
m/z 741.1745, found 741.1755. Compound 39 was obtained as a
colorless oil: TLC Rf 0.54 (EtOAc/toluene, 1:8); [R]23.5D -9.6 (c
1.15, CHCl3); IR 3480, 2930, 1750 cm-1; 1H NMR (300 MHz,
CDCl3) δ 1.02 (s, 3H), 1.09 (s, 9H), 1.11 (s, 3H), 1.69-1.72 (m,
2H), 2.46 (d, 3H, J = 1.5 Hz), 3.03-3.10 (m, 2H), 3.25 (s, 3H),
3.61 (dd, 1H, J = 5.5, 6.2 Hz), 3.71-3.81 (m, 3H), 3.78 (s, 3H),
3.94 (dd, 1H, J = 5.3, 8.8 Hz), 4.04 (d, 1H, J = 11.3 Hz), 4.30
(d, 1H, J = 11.3 Hz), 4.23 (d, 1H, J = 2.1 Hz), 6.00 (qd, 1H,
J = 1.5, 8.8 Hz), 6.80 (d, 2H, J = 8.5 Hz), 7.04 (d, 2H, J = 8.5
Hz), 7.40-7.48 (m, 6H), 7.65-7.74 (m, 4H); 13C NMR (68 MHz,
Synthesis of (-)-Pestalotiopsin A (7). A solution of 55 (6.8 mg,
0.014 mmol) in THF (0.4 mL), H2O (0.4 mL), and AcOH
(0.4 mL) was stirred for 3 h. The mixture was diluted with
saturated aqueous NaHCO3 (8 mL) and extracted with EtOAc
(4 mL ꢀ 3). The combined extracts were dried and concentrated
under reduced pressure. The residue was purified by column
chromatography on silica gel (EtOAc/hexane, 2:1) to provide
4.4 mg (96%) of 7 as white crystals: mp 204-205 °C; TLC Rf
0.16 (EtOAc/hexane, 2:1), 0.21 (MeOH:CHCl3, 2:98), 0.46
(acetone:toluene, 1:1); [R]21 -74.7 (c 0.535, MeOH); IR
D
3370, 2940, 1710 cm-1
;
1H NMR (300 MHz, CDCl3/
6460 J. Org. Chem. Vol. 74, No. 17, 2009

Takao et al.
JOCFeatured Article
CD3OD=10/1) δ 1.04 (s, 3H), 1.17 (s, 3H), 1.65 (dd, 1H, J =
6.4, 12.2 Hz), 1.93 (d, 3H, J = 0.9 Hz), 2.01 (dd, 1H, J = 9.6,
12.2 Hz), 2.07 (s, 3H), 2.42 (m, 1H), 2.43 (dd, 1H, J = 10.5, 10.7
Hz), 2.55 (dd, 1H, J = 5.3, 10.5 Hz), 2.63 (ddd, 1H, J = 1.5, 6.4,
9.6 Hz), 3.32 (s, 3H), 3.88 (dd, 1H, J = 6.2, 11.5 Hz), 3.96 (dd,
1H, J = 1.9, 6.2 Hz), 5.10 (br d, 1H, J = 11.5 Hz), 5.25 (dd, 1H,
J = 5.3, 10.7 Hz), 5.84 (d, 1H, J = 2.8 Hz); 13C NMR (68 MHz,
CDCl3/CD3OD=10/1) δ 17.6, 21.4, 23.9, 27.2, 38.4, 39.6, 41.1,
42.4, 56.1, 64.8, 73.6, 77.3, 83.0, 98.3, 108.8, 123.9, 137.8, 170.4;
HRMS (EI) calcd for C18H28O6 (Mþ) m/z 340.1886, found
340.1887.
Synthesis of (þ)-Pestalotiopsin A (1). As described for the
preparation of 7 from 55, compound ent-55 (7.9 mg, 0.016
mmol) was converted to 5.2 mg (96%) of 1: mp 204-205 °C;
[R]22.5D þ75 (c 0.26, MeOH).
Acknowledgment. We thank Prof. Fumio Sugawara
(Tokyo University of Science) for providing a sample and
spectral copies of natural (þ)-pestalotiopsin A. We also
thank Dr. Kaoru Yamada (Keio University) for performing
the biological assays. This work was supported in part by a
Grant-in-Aid for Young Scientists (B) from MEXT and the
Astellas Pharma Award in Synthetic Organic Chemistry,
Japan.
Supporting Information Available: Experimental proce-
dures, full characterization data, and copies of 1H and 13C
NMR spectra for all new compounds described herein. This
material is available free of charge via the Internet at http://
pubs.acs.org.
J. Org. Chem. Vol. 74, No. 17, 2009 6461