Dimerization of butenolide structures. A biomimetic approach to the dimeric sesquiterpene lactones (±)-biatractylolide and (±)- biepiasterolide

Article abstract of DOI:10.1021/jo0488053

The biomimetic synthesis of the bisesquiterpene lactones (±)-biatractylolide 1 and (±)-biepiasterolide 2 via dimerization of the captodative stabilized radical 8 is reported. Atractylon 7 has also been shown to be a possible intermediate during the biosynthesis of biatractylolide 1, biepiasterolide 2, atractylolide 3, and hydroxyatractylolide 4.

Full text of DOI:10.1021/jo0488053

Dimerization of Butenolide Structures. A Biomimetic Approach to
the Dimeric Sesquiterpene Lactones (()-Biatractylolide and
(()-Biepiasterolide
Sharanjeet K. Bagal, Robert M. Adlington, Jack E. Baldwin,* and Rodolfo Marquez†
Chemistry Research Laboratory, University of Oxford, Mansfield Road, Oxford OX1 3TA, U.K.
jack.baldwin@chem.ox.ac.uk
Received July 14, 2004
The biomimetic synthesis of the bisesquiterpene lactones (()-biatractylolide 1 and (()-biepiasterolide
2
via dimerization of the captodative stabilized radical 8 is reported. Atractylon 7 has also been
shown to be a possible intermediate during the biosynthesis of biatractylolide 1, biepiasterolide 2,
atractylolide 3, and hydroxyatractylolide 4.
As part of our work on the biomimetic synthesis of
bisesquiterpene lactones, we have recently described the
total synthesis of biatractylolide and biepiasterolide
through biomimetic radical dimerizations.1 We now
report further studies on these biomimetic transforma-
tions that shed new light on the proposed biosynthesis
of these very interesting compounds.
naturally occurring constituents of A. macrocephala
(Figure 1).6
Our proposal for the biomimetic synthesis of both
biatractylolide 1 and biepiasterolide 2 involved the fusion
of two units of atractylolide 3 or hydroxyatractylolide 4
through a key dimerization step that takes place via the
captodative stabilized radical 87 (or its equivalent).
Moreover, atractylolides 3 and 4 could emanate in vivo
from atractylon 7, given that Hikino has reported the
aerial autoxidation of atractylon 7 into atractylolides 3
and 4 (Scheme 1).9
8
Biatractylolide 1 and biepiasterolide 2 are novel bi-
sesquiterpene lactones that have been isolated from the
Chinese medicinal plant Atractylodes macrocephala.2
-4
Biologically, biatractylolide 1 is a potential blood pressure
5
lowering agent, with significant negative inotropic and
As a model system, we originally explored the dimer-
ization aptitude of the model butenolide 12, which
incorporated critical features of atractylolide 3. Buteno-
lide 12 was efficiently synthesized through minor modi-
chronotropic effects.
The structure of biatractylolide 1 has been elucidated
by X-ray analysis, while that of biepiasterolide 2 has been
1
0
determined by extensive NMR studies.2
-4
fication of the method of Minato and Nagasaki. Pyrro-
Both com-
lidine enamine 9 was initially alkylated to produce ester
pounds have been shown to be dimeric bisesquiterpenoids
joined at the C8-C8a bridgehead positions. The struc-
tural similarity of 1 and 2 with the naturally occurring
sesquiterpene lactones atractylolide 3 and hydroxy-
atractylolide 4 has prompted speculation about the
biosynthetic origin of biatractylolide 1 and biepiasterolide
1
0, followed by hydrolysis to give crude acid 11 and
cyclization to afford butenolide 12 in reasonable yield
(
Scheme 2).
After initial failed attempts to dimerize butenolide 12,1
di-tert-butyl peroxide (DTBP) was considered as a likely
dimerization agent. DTBP has been shown to be an
effective model for reactive radicals that promote hydro-
2
. Atractylolide 3 and hydroxyatractylolide 4, together
with the closely related sesquiterpenes peroxy-
11
gen atom abstraction reactions in biological systems and
is known to effect dehydrodimerizations of alcohols,
atractylolide 5, atractylenolide 6, and atractylon 7, are
1
2
amides, polyhaloalkanes, esters, and ethers. Finally,
†
Present address: School of Life Sciences, University of Dundee,
Nethergate, Dundee DD1 4HN, U.K.
(
1) (a) Bagal, S. K.; Adlington, R. M.; Marquez, R.; Baldwin, J. E.;
(6) (a) Wang, Y. S.; Chang, J. C.; Li, K. K.; Wu, C. H.; Tin, K.; Liu,
Y. H. Shaanxi Xinyiyao 1980, 9, 47. (b) Zhang, Q. F.; Luo, S. D.; Wang,
H. Y. Chin. Chem. Lett. 1998, 9, 1097. (c) Chen, Z. Planta Med. 1987,
53, 493.
(7) Janousek, Z.; Merenyi, R.; Viehe, H. G. Acc. Chem. Res. 1985,
18, 148.
(8) For example, a pinacol-type dimerization of a keto form of 4
followed by bislactonization.
Cowley, A. Tetrahedron Lett. 2003, 44, 4993. (b) Bagal, S. K.; Adlington,
R. M.; Marquez, R.; Baldwin, J. E.; Cowley, A. Org. Lett. 2003, 5, 3049.
(
2) Lin, Y.; Jin, T.; Wu, X.; Huang, Z.; Fan, J. J. Nat. Prod. 1997,
6
0, 27.
3) Wang, B. D.; Yu, Y. H.; Teng, N. N.; Jiang, S. H.; Zhu, D. Y.
Huaxue Xuebao 1999, 57, 1022.
(
(
4) Huang, Z.; Lin, Y.; Liu, S.; Chan, W. Indian J. Chem., Sect. B:
Org. Chem. Incl. Med. Chem. 1999, 38B, 106. Note: In this article,
the absolute stereochemistry at the C5 (C5a) center of the trans-C5
(9) (a) Hikino, H.; Hikino, Y.; Yosioka, I. Chem. Pharm. Bull. 1964,
12, 755. (b) Hikino, H.; Hikino, Y.; Yosioka, I. Chem. Pharm. Bull. 1962,
10, 641.
(
C5a)-C10 (C10a) fused decalin ring system appears incorrectly
assigned (according to the CIP system nomenclature) and should be
R (5aR).
5) Pu, H. L.; Wang, Z. L.; Huang, Q. J.; Xu, S. B.; Lin, Y. C.; Wu,
X. Y. Zhongguo Yaolixue Tongbao 2000, 16, 60.
(10) (a) Minato, H.; Nagasaki, T. J. Chem. Soc. C 1966, 1866. (b)
Minato, H.; Nagasaki, T. J. Chem. Soc. C 1966, 377.
(11) Tanko, J. M.; Friedline, R.; Suleman, N. K.; Castagnoli, N. J.
Am. Chem. Soc. 2001, 123, 5808.
5
(
10.1021/jo0488053 CCC: $27.50 © 2004 American Chemical Society
9100
J. Org. Chem. 2004, 69, 9100-9108
Published on Web 11/20/2004

Dimerization of Butenolide Structures
FIGURE 1. (+)-Biatractylolide 1, biepiasterolide 2, (+)-atractylolide 3, (+)-hydroxyatractylolide 4, (+)-peroxyatractylolide 5,
(
+)-atractylenolide 6, and (+)-atractylon 7.
Based on literature precedent,15 the synthesis of halo-
lactone 15 began with the reduction of lactone 12 followed
by silylation of the resulting furan 16 to afford silylfuran
SCHEME 1. Biosynthetic Proposal for
Biatractylolide 1 and Biepiasterolide 2
1
7 in reasonable yield (25%). The desired chloro lactone
9 was then generated in good overall yield by efficiently
1
oxidizing silylfuran 17 with singlet oxygen to afford
hydroxybutenolide 18, which was then halogenated with
excess thionyl chloride (Scheme 4).
SCHEME 2
Unfortunately, zinc, copper bronze, or activated copper
bronze (Vogel’s method)16 failed to dimerize chloro lactone
1
9 in higher than 2% yield. However, freshly prepared
17
Co(PPh
3
)
3
Cl afforded the dimerized adduct 14 at room
temperature and in good yield (57%) through the general
method of Yamada.18 Diastereomers 14a (dl) and 14b
(
meso) were produced in a ratio greater than 98:2,
respectively, and could be separated by crystallization
(
Scheme 5 and Table 1).14
The relative stereochemistry at both C8 and C8a in
the major diastereomer dl-bis-butenolide 14a was un-
equivocally shown to be the same as that which is
characteristic of biatractylolide 1 and biepiasterolide 2
by X-ray analysis. Furthermore, the X-ray structure
indicated that the C8-C8a bond is the longest in the
molecule at 1.57 Å, which compares favorably with the
corresponding bond length reported for biatractylolide 1
after extensive experimentation including the use of
photochemical conditions,13 an oxygen-free mixture of
butenolide 12, acetone, and DTBP (0.5 equiv) in a sealed
tube at 140 °C afforded the desired dimers 14a and 14b,
in an isolated yield of 34%. The dimerization presumably
takes place through the captodative stabilized radical 13
(
1.59 Å; Figure 2).4
Upon completion of the dimerization studies of the
model system, we began the synthesis of the complete
monomer unit atractylolide 3 based on the work of
Minato and Nagasaki.10 Thus, reduction of the com-
mercially available tetralone 20 to alcohol 21 proceeded
cleanly over two steps. Enol ether 21 was then converted
to ketone 22 by Oppenauer oxidation followed by cuprate
addition and subsequent fluoride-mediated desilylation
of the TMS enol ether intermediate. Finally, the ketone
(
Scheme 3).
Model dimer 14 was obtained as a mixture of two
diastereomers in a 6:1 ratio (as determined by NMR
integration), with the major diastereomer 14a (dl) being
separable by recrystallization from a 20% mixture of
14
2
EtOAc in Et O. Although DTBP was a successful dimer-
ization agent for butenolide 12, the reaction yield could
not be further optimized. At this point, we began explor-
ing the possibility of dimerizing a halolactone, 15, using
metals and metal salts that have been shown to couple
allylic and benzylic halides (Scheme 3).
19
23 was generated by Wittig olefination and methyl
ether hydrolysis (Scheme 6).
Alkylation of ketone 23 then afforded the keto ester
24 as a mixture of diastereomers through the corre-
SCHEME 3. Dimerization of Model Butenolide 12
J. Org. Chem, Vol. 69, No. 26, 2004 9101

Bagal et al.
(0.5 equiv of DTBP, acetone, 140 °C, 18 h).1 Use of
stronger reaction conditions (0.5-2.0 equiv of DTBP,
acetone, 120-170 °C, 18 h), however, afforded polymeric
material, with the amount of polymeric material formed
increasing as the severity of the conditions increased. We
originally attributed the lack of dimerization to the
increased functionality present in atractylolide 3 com-
pared to that in the model system 12 (i.e., the axial
methyl group and the exocyclic double bond). However,
removal of the exocyclic double bond through hydrogena-
SCHEME 4
9
tion generated butenolide 26 as a mixture of diastere-
omers, which under the DTBP dimerization conditions
failed to dimerize (Scheme 7).
SCHEME 5. Cobalt-Mediated Dimerization of
Chloro Lactone 19
Based upon these results, a possible explanation for
the observed lack of DTBP-mediated dimerization is that
t
•
the steric hindrance between the BuO radical (or pos-
•
sibly Me radical) conducting the hydrogen atom abstrac-
tion and the axial methyl group precludes the key
abstraction (radical generation) process. These results
corroborate the work of Tanko, in which he reports that
TABLE 1. Metal-Mediated Dimerization of Chloro
Lactone 19 To Generate Dimer 14
t
•
BuO radicals require a relatively ordered transition
temperature reaction
yield of 14
(%)
state during hydrogen abstraction.11 Thus, it is feasible
that the hindrance caused by the axial methyl group
coupling agent
(°C)
time (h) solvent
copper
90-100
90-100
20-25
20-25
1
1
24
2
benzene
benzene
EtOAc
2
2
t
•
disrupts the trajectory of approach of the BuO radical
to atractylolide 3 so that the “ordered transition state”
necessary for hydrogen atom abstraction is not possible
activated copper
zinc
no reaction
Co(PPh3)3Cl
benzene 57
(
Figure 3).
FIGURE 3. Proposed inhibition of hydrogen atom abstraction.
At this junction, our synthetic approach switched to
1
our other promising model dimerization study, in which
FIGURE 2. X-ray crystal structure of (()-14a.
radical 8 could be accessed from hydroxyatractylolide 4
via the corresponding chloroatractylolide 27 (Scheme 8).
Because the conversion of model butenolide 12 to
hydroxybutenolide 18 by conventional methods required
a four-step sequence together with a tedious chromato-
graphic purification of silylfuran 17 from parent furan
16 (Scheme 4), a more efficient method to generate
sponding pyrrolidine enamine. The target monomer
atractylolide 3 was easily obtained in good yield by ester
hydrolysis followed by lactonization of the corresponding
keto acid 25.
Regrettably, none of the desired dimers 1 and 2 were
produced when atractylolide 3 was treated with our
previously optimized thermal DTBP reaction conditions
hydroxyatractylolide
4 was sought. To this end,
atractylolide 3 was first transformed into silyloxy furan
20
28, which was then cautiously oxidized to afford the
SCHEME 6
9102 J. Org. Chem., Vol. 69, No. 26, 2004

Dimerization of Butenolide Structures
SCHEME 7
SCHEME 8. Proposed Synthesis of 1 and 2 from
Hydroxyatractylolide 4
ring of butenolide 6, thereby generating an enol, which
can be transformed into the corresponding hydroxy
lactone 4 via a keto acid intermediate.
Suitable halogenation conditions capable of generating
the desired chloroatractylolide 27 in good yield (ca. 50%)
were eventually developed by the addition of pyridine and
operation at low temperature (-78 °C). However, unlike
chloro lactone 19, chloroatractylolide 27 could not be
purified by flash chromatography because this led to the
elimination of HCl. The most efficient chlorination of
hydroxyatractylolide 4 generated a mixture of chloroat-
ractylolide 27 and atractylenolide 6 in a 3:1 ratio,
respectively (as determined by NMR integration). This
inseparable mixture was then used in the subsequent
coupling step without further purification. The stereo-
chemistry of the newly introduced chlorine on the face
naturally occurring sesquiterpene hydroxyatractylolide
6
a
4
(Scheme 9).
The structure of hydroxyatractylolide 4 was corrobo-
rated by X-ray crystallography, which indicated that the
hydroxyl group is located in an axial position (Figure 4).21
This assignment is in agreement with the observations
9
of Hikino et al., in which hydroxyatractylolide 4 was
concluded to possess a 1,3-diaxial relationship between
the angular methyl group and the hydroxyl group (Scheme
9
), and presumably arises because of the equatorial
N
opposite of the C-methyl group is based upon an S 2-
like attack from the less-hindered face of activated
butenolide 4 (Scheme 11).
attack of the carboxyl oxygen onto the ketone functional-
ity.
The critical dimerization step was then attempted by
treating chloroatractylolide 27 with our previously de-
veloped dimerization conditions [Co(PPh
2
3
)
3
Cl, PhH, rt,
h]. This process proceeded cleanly to generate three
dimeric adducts from which (()-biatractylolide 1, (()-
1
FIGURE 4. X-ray analysis of (()-hydroxyatractylolide 4.
Chloroatractylolide 27 could not be obtained from
hydroxyatractylolide 4 under our previous halogenation
1
conditions (SOCl
2
, CH
2
Cl
2
, rt, 18 h), although the natural
product atractylenolide 6 was produced (most likely
through an E1 mechanism). Interestingly, hydroxy-
atractylolide 4 was regenerated in good yields by stirring
atractylenolide 6 with an aqueous base (Scheme 10). It
is mechanistically plausible that this conversion involves
nucleophilic attack by a hydroxide ion to open the lactone
FIGURE 5. X-ray analysis of (()-biatractylolide 1.
biepiasterolide 2, and another diastereoisomer (()-29
were isolated (Scheme 11).2 The structures of synthetic
(()-biatractylolide 1 and (()-biepiasterolide 2 were first
2,23
SCHEME 9
J. Org. Chem, Vol. 69, No. 26, 2004 9103

Bagal et al.
TABLE 2. Comparison of H and 13C NMR Data from Synthetic (()-1 and (()-2 with the Literature Data2,3
1
literature (+)-1a 1H; 13
C
synthetic (()-1b 1H; 13
C
literature 2c 1H; 13
C
synthetic (()-2b 1H; 13
C
position
1
(ppm)
(ppm)
(ppm)
(ppm)
1.23 (1H, ddd),
1.22 (1H, td),
1.59-1.61 (1H, m); 42.1
1.63-1.65 (2H, m); 22.3
1.40-1.65 (2H, m); 42.3
1.38-1.50 (1H, m),
d
1
.61 (1H, m); 42.5
1.53-1.59 (1H, m); 42.2
1.38-1.50 (1H, m),
d
2
3
1.65 (2H, m); 22.3
1.40-1.65 (2H, m); 22.9
1.60-1.66 (1H, m); 22.8
1.94 (1H, br q),
1.90-1.99 (1H, m),
2.35 (1H, br d); 35.8
147.9
2.08 (1H, ddm),
2.34 (1H, dm); 36.7
148.2
2.05-2.13 (1H, m),
2.37 (1H, dm); 36.8
148.1
2.35 (1H, br d); 35.8
4
5
6
147.8
1.72 (1H, dd); 52.7
2.65 (1H, dd),
1.71-1.75 (1H, m); 52.9
2.65 (1H, dd),
2.78 (1H, t); 27.9
164.5
2.81 (1H, dd); 42.8
2.18 (1H, ddq),
2.63 (1H, dd); 24.9
163.4
2.82 (1H, dd); 42.7
2.19 (1H, ddd),
2.65 (1H, dd); 24.8
163.3
2.72 (1H, t); 27.8
7
8
9
164.3
89.2
1.42 (1H, d),
89.3
89.3
89.3
1.42 (1H, d),
2.82 (1H, d); 49.7
36.9
2.00 (1H, d),
2.31 (1H, d); 47.2
36.9
2.01 (1H, d),
2.32 (1H, d); 47.0
36.6
2.82 (1H, d); 49.6
1
1
1
1
1
1
0
1
2
3
4
5
36.9
124.3
124.4
172.0
1.75 (3H, s); 8.4
1.13 (3H, s); 17.1
4.65 (1H, s),
4.86 (1H, s); 107.3
125.5
172.7
1.77 (3H, d); 8.7
0.49 (3H, s); 19.1
4.62 (1H, d),
4.86 (1H, d); 107.2
125.4
172.6
1.79 (3H, d); 8.6
0.50 (3H, s); 19.0
4.64 (1H, s),
4.87 (1H, s); 107.1
171.7
1.75 (3H, s); 8.3
1.13 (3H, s); 17.1
4.65 (1H, br s),
4.86 (1H, br s); 107.2
a
NMR spectra were recorded in CDCl3 at 400 MHz ( H) and 22.5 MHz ( C). b NMR spectra were recorded in CDCl3 at 500 MHz ( H)
1
13
1
1
3
c
1
13
d
and 125.7 MHz ( C). NMR spectra were recorded in CDCl3 at 400 MHz ( H) and 100.6 MHz ( C). Overlapping.
SCHEME 10
the 5R,10S,8R-8aR,10aR,5aS isomer (plus its enanti-
omer). This was confirmed by X-ray spectroscopy (Figure
; see the Supporting Information). Attempts to optimize
7
corroborated by comparison of the NMR data to the
reported literature data (Table 2)2,3 and then unambigu-
ously assigned using HRMS and X-ray spectroscopy
(
Figures 5 and 6).21
FIGURE 7. X-ray analysis of (()-dimer 29 with crystal-
lographic numbering system.
the dimerization yield by in situ conversion of chloroat-
ractylolide 27 into the corresponding iodoatractylolide
failed to improve the dimerization yield and product
distribution.
The high diastereoselectivity observed during the C8-
C8a bond formation [dl . meso] in the model dimeriza-
FIGURE 6. X-ray analysis of (()-biepiasterolide 2.
Furthermore, 1- and 2-D NMR analysis and HRMS
suggest that the remaining diastereoisomer (()-29 was
SCHEME 1123
9104 J. Org. Chem., Vol. 69, No. 26, 2004

Dimerization of Butenolide Structures
SCHEME 1224
SCHEME 13
tion to bis-butenolide 14 through either the DTBP
(
3 3
Scheme 3) or Co(PPh ) Cl (Scheme 5) method does not
have an obvious origin. However, it should be noted that
an analogous high diastereoselectivity during C8-C8a
bond formation in the real system was likewise observed,
because (()-1, (()-2, and (()-29 all constitute “dl” forms
(
Scheme 11).
SCHEME 1424
Once having successfully completed the total synthesis
of both biatractylolide 1 and biepiasterolide 2 through
cobalt-mediated radical dimerizations, we decided to
expand our butenolide dimerization investigation studies
in order to be able to include a fresh look at the aerial
9
autoxidation studies of Hikino and co-workers. In his
work, Hikino demonstrated that atractylon 7 undergoes
aerial autoxidation when allowed to stand in the presence
of air and methanol to generate atractylolide 3 and
hydroxyatractylolide 4. It is possible to speculate that the
observed aerial autoxidation might itself proceed via a
radical intermediate because triplet dioxygen is a diradi-
cal species. The radical intermediate formed could then
be capable of dimerization. Hence, the aerial autoxidation
of atractylon 7 might play a more crucial role in the
biosynthesis of biatractylolide 1 and biepiasterolide 2
than originally proposed.
ably more forceful conditions.25 In their oxidation proce-
dure, neat menthofuran 31 was converted to hydroxy-
butenolide 32 in low yield in an oxygen atmosphere at
1
00 °C (Scheme 13). Although these conditions were sig-
nificantly different from those employed in Hikino’s
studies, we felt that they might provide us with further
insight into the biosynthetic pathway of the observed
dimerizations.
Interestingly, treatment of atractylon 7 under Wood-
ward’s conditions generated hydroxyatractylolide 4, as
well as dimerization adducts 1, 2, and 29 in reasonable
yield.22 However, atractylolide 3 and its isomer 30 were
never detected in the reaction mixture (Scheme 14).
Modification of the reaction conditions (i.e., through
the use of tert-butyl alcohol as a solvent combined with
a slight decrease of the reaction temperature to 80 °C
for 1 h) increased the yields of both hydroxyatractylolide
Atractylon 7 was efficiently generated by reduction of
atractylolide 3 and then subjected to Hikino’s aerial
9
autoxidation conditions. Oxidation of a methanolic solu-
tion of atractylon 7 in an open flask at room temperature
for 35 days afforded a mixture of products from which
atractylolide 3, hydroxyatractylolide 4, the novel buteno-
lide 30, and all three dimers (()-1, (()-2, and (()-29, were
isolated (Scheme 12).2
2,24
Woodward and co-workers had also previously reported
oxygen-mediated furan oxidations, albeit under consider-
4
(27%) and the dimerization products 1, 2, and 29 (12%,
22,24
diastereomeric ratio ca. 1:1:1),
while once again failing
(
12) Paquette, L. A. Encyclopedia of Reagents for Organic Synthesis;
Wiley: Chichester, U.K., 1995; Vol. 3, p 1616.
13) (a) Knowles, H. S.; Hunt, K.; Parsons, A. F. Tetrahedron 2001,
7, 8115. (b) Beaujean, M.; Mignani, S.; Mer e´ nyi, R.; Jancusek, Z.;
Viehe, H. G.; Kirch, M.; Lehn, J. M. Tetrahedron 1984, 40, 4395.
to produce any atractylolide 3 or isomer 30.
(
In a separate experiment, treatment of atractylolide 3
5
25
under the original Woodward reaction conditions as a
(
14) The diastereomeric ratio was determined by NMR analysis
conducted after a two-phase workup and careful purification by flash
chromatography.
(21) The atomic coordinates for 4 (deposition number CCDC 235995),
1 (deposition number CCDC 207880), 2 (deposition number CCDC
207879), and 29 (deposition number CCDC 254587) are available upon
request from the Cambridge Crystallographic Data Centre, University
Chemical Laboratory, Lensfield Road, Cambridge CB2 1EW, U.K. The
crystallographic numbering system differs from that used in the test;
therefore, any request should be accompanied by the full literature
citation of this paper.
(22) The dimer diastereomeric ratios were determined after careful
purification by flash chromatography and preparative TLC by NMR
analysis. Likewise, the diastereomeric ratio of butenolide 30 was
determined by NMR analysis after purification.
(23) Yields quoted are based upon recovered hydroxyatractylolide
4.
(24) Yields quoted are based upon recovered atractylon 7.
(25) Woodward, R. B.; Eastman, R. H. J. Am. Chem. Soc. 1950, 72,
399.
(
15) (a) Wu, H. J.; Yen, C. H.; Chuang, C. T. J. Org. Chem. 1998,
6
6
3, 5064. (b) Shikari, R.; Tadano, K.; Ogawa, S. J. Org. Chem. 1996,
1, 2845. (c) R o¨ hrig, S.; Hennig, L.; Findeisen, M.; Welzel, P. Tetra-
hedron 1998, 54, 3439.
(
16) Vogel, A. I. Vogel’s Textbook of Practical Organic Chemistry,
5
th ed.; Wiley: New York, 1989; p 426.
(
(
17) Aresta, M.; Rossi, M.; Sacco, A. Inorg. Chim. Acta 1969, 3, 227.
18) ) (a) Yamada, Y.; Momose, D.; Iguchi, K.; Sugiyama, T.
Tetrahedron Lett. 1983, 24, 921. (b) Yamada, Y.; Momose, D. Chem.
Lett. 1981, 1277. (c) Yamada, Y.; Momose, D.; Iguchi, K.; Sugiyama,
T. Chem. Pharm. Bull. 1984, 32, 1840.
(
19) Miller, R. B.; Nash, R. D. J. Org. Chem. 1973, 38, 4424.
(
20) (a) Jefford, C. W.; Sledeski, A. W.; Rossier, J.; Boukouvalas, J.
Tetrahedron Lett. 1990, 31, 5741. (b) Maltais, F.; Lachance, N.;
Boukouvalas, J. Tetrahedron Lett. 1994, 35, 7897.
J. Org. Chem, Vol. 69, No. 26, 2004 9105

Bagal et al.
SCHEME 15
4. Atractylolide 3 and the novel butenolide 30 are likely
to be autoxidation products of atractylon 7 derived
formally from oxygen atom addition (Scheme 17).
5
. If the autoxidation of atractylon 7 can be linked to
our own cobalt-mediated dimerization study of chloro-
atractylolide 27 [in which a captodative stabilized radical
8
can be proposed as a precursor of biatractylolide 1,
biepiasterolide 2, and the dimer 29 (Scheme 11)], then
the autoxidation of atractylon 7 would also need to
proceed by formal oxygen atom addition and hydrogen
atom abstraction in order to generate the same radical
precursor 8 which leads to the observed dimeric adducts
1
, 2, and 29 (Scheme 18).
SCHEME 16. Formal Dioxygen Addition to
Atractylon 7 To Generate Hydroxyatractylolide 4
Conclusions
In conclusion, we have completed the synthesis of the
bisesquiterpene lactones biatractylolide 1 and biepia-
sterolide 2 through a biomimetic approach. During this
process, we have correlated the work of Hikino with our
own results and demonstrated that, under chemical
conditions, atractylon 7 is a possible intermediate during
the synthesis of (+)-biatractylolide 1, biepiasterolide
control failed to generate either the hydroxyatractylolide
4
or any of the dimeric structures 1, 2, and 29 and led
only to the recovery of unreacted atractylolide 3 (Scheme
5).
2
, (+)-atractylolide 3, and (+)-hydroxyatractylolide 4.
1
We have also shown that, under such conditions,
atractylolide 3 and hydroxyatractylolide 4 are competing
products to the dimerization reaction and not intermedi-
ate species en route to (()-biatractylolide 1 and (()-
biepiasterolide 2.
Similarly, treatment of hydroxyatractylolide 4 under
Hikino’s conditions (open to air, methanol, rt, 35 days)9
as a control failed to generate any of the dimeric
structures 1, 2, and 29 and led only to the recovery of
unreacted hydroxyatractylolide 4.
By considering these results together, it is possible to
form the following tentative conclusions regarding the
synthesis of this family of terpenes:
Experimental Section
3
,3′-Dimethyl-4,5,6,7,4′,5′,6′,7′-octahydro[7a,7′a]bi-
1
. Atractylolide 3 and hydroxyatractylolide 4 are
benzofuranyl-2,2′-dione (14). Caution! Reactions in sealed
tubes and ampules must be performed behind a safety shield,
and all organic peroxides should be handled with due care.
unlikely to be direct biosynthetic precursors of bi-
atractylolide 1 and biepiasterolide 2.
2. Biatractylolide 1, biepiasterolide 2, atractylolide 3,
1
. DTBP-Mediated Dimerization. A solution of butenolide
hydroxyatractylolide 4, dimer 29, and the novel buteno-
lide 30 are all likely to be autoxidation products of
atractylon 7 generated by the oxidation of atractylon 7
to higher oxidation levels. The degree of oxidation can
vary, and the product formed depends on the oxidation
level that atractylon 7 is promoted to. Each oxidation
level leads to different oxidation products.
1
2 (200 mg, 1.31 mmol, 1 equiv), di-tert-butyl peroxide (121
µL, 0.66 mmol, 0.5 equiv), and acetone (0.5 mL) was placed in
an ampule, which was degassed with three freeze-pump-
thaw cycles and then sealed. After heating to 140 °C for 18 h
overnight, the ampule was opened at room temperature under
an atmosphere of argon. The residue was diluted with Et
washed with water, dried (Na SO ), filtered, and concentrated
2
O,
2
4
under reduced pressure. The residue was placed under high
vacuum at 50 °C overnight and then purified by flash chro-
matography (CH Cl ) to yield the title compound 14 (66.6 mg,
3
. Hydroxyatractylolide 4 is likely to be an autoxidation
product of atractylon 7 derived from formal dioxygen
addition (Scheme 16).
2
2
34%) as a white solid in a 6:1 dl/meso mixture of diastereomers.
SCHEME 17. Formal Oxygen Atom Addition to Atractylon 7 To Generate Atractylolide 3 and Isomer 31
SCHEME 18. Formal Oxygen Atom Addition and Hydrogen Atom Abstraction from Atractylon 7
9106 J. Org. Chem., Vol. 69, No. 26, 2004

Dimerization of Butenolide Structures
Recrystallization from Et
major dl diastereomer to be isolated.
. Co(PPh Cl-Mediated Dimerization. Chloro lactone
9 (100 mg, 0.54 mmol, 1 equiv) dissolved in benzene (1 mL)
was added to a suspension of CoCl(PPh (570 mg, 0.65 mmol,
.2 equiv) in benzene (4 mL) under argon, and the resulting
mixture was stirred at room temperature for 2 h under argon.
The reaction mixture was diluted with Et O and filtered, and
2
O/EtOAc (4:1) allowed the pure,
24.5, 36.0, 36.6, 41.2, 51.1, 51.6, 103.5, 106.7, 122.0, 148.4,
+
160.9, 172.3; MS (CI) m/z 249 [MH , 20%], 231 [100]; HRMS
+
2
)
3 3
calcd for C15
H
21
O
3
(MH ) 249.1491, found 249.1497.
1
(()-3,8a,3′,8′a-Tetramethyl-5,5′-dimethylene-4,4a,5,6,7,8,-
8a,9,4′,4′a,5′,6′,7′,8′,8′a,9′-hexadecahydro[9a,9′a]bi[naphtho-
3 3
)
2-4
1
[2,3-b]furanyl]-2,2′-dione (1, 2, 29, and 30). To a solution
of hydroxyatractylolide 4 (500 mg, 2.0 mmol, 1 equiv) and
pyridine (1.63 mL, 20 mmol, 10 equiv) in THF (25 mL) at -78
°C and under argon was added freshly distilled thionyl chloride
(0.73 mL, 10 mmol, 5 equiv), and the resulting solution was
2
the filtrate was washed with water and concentrated to
dryness in vacuo. Purification by flash chromatography [pe-
troleum ether 30-40/Et
2
O (7:3)] gave the title compound 14
stirred for 30 min. Et O (30 mL) and 0.1 M aqueous HCl (10
2
as a white solid (46 mg, 57%), and mainly as the dl diastere-
omer. Only trace meso diastereomer was detected.
mL) were then added, and the solution was allowed to warm
to room temperature. The organic layer was separated and
3
. Data for (()-dl Diastereomer 14a: mp 188-189 °C
2 4
washed with water, dried (Na SO ), filtered, and concentrated
-
1
(
from 4:1 Et
2
O/EtOAc); R
f
0.3 (CH
2
Cl
2
); IR (film) νmax(cm
)
in vacuo at room temperature to give a mixture of chloroat-
2
947s, 2864m, 1747s, 1672m, 1447m, 1266s, 1018s, 737s; MS
ractylolide 27 and atractylenolide 6 in a 3:1 ratio, respectively,
+
-1
(
CI) m/z 303 [MH , 25%], 152 [100]; HRMS calcd for C18
H
26
-
as a colorless oil: IR (film) νmax (cm ) 2933s, 1778s, 1691m,
+
NO
4
(MNH
. NMR Data for (()-dl Diastereomer 14a: δH/2 (400
MHz, CDCl ) 1.10-1.30 (1H, m), 1.36 (1H, td, J ) 14.0 and
.0 Hz), 1.64-1.73 (1H, m), 1.75 (3H, s), 2.03-2.23 (3H, m),
.63 (1H, dm, J ) 13.5 Hz), 2.87 (1H, dm, J ) 14.0 Hz); δC/2
) 8.4, 21.4, 26.8, 27.6, 37.2, 88.2, 122.5,
4
) 320.1862, found 320.1855
1651m, 1265s, 1098m, 738s; NMR δHmax (400 MHz, CDCl
3
) 1.86
(100.6 MHz, CDCl
98.3 (C-Cl). The crude chloroatractylolide 27 was dissolved
in benzene (8 mL) and added to a suspension of CoCl(PPh
4
(3H, s), 4.62 (1H, s), 4.88 (1H, s); NMR δ
C
3
)
3
5
2
(
1
3 3
)
(2.1 g, 2.4 mmol, 1.2 equiv) in benzene (12 mL) under argon,
and the resulting mixture was stirred at room temperature
for 2 h under argon. The reaction mixture was filtered, and
the filtrate was washed with water and concentrated to
dryness in vacuo. Purification by flash chromatography [pe-
100.6 MHz, CDCl
3
66.1, 172.9.
. NMR Data for (()-meso Diastereomer 14b. The meso
diastereomer was not isolated in pure form: δH/2 (400 MHz,
CDCl ) 1.10-1.30 (1H, m), 1.40-1.50 (1H, m), 1.50-1.62 (1H,
m), 1.80-2.00 (1H, m), 1.90 (3H, s), 1.96-2.04 (1H, m), 2.03-
.23 (1H, m), 2.44-2.52 (1H, m), 2.66-2.84 (1H, m); δC/2 (100.6
) 8.4, 22.5, 26.2, 27.7, 36.7, 87.8, 122.5, 166.3,
5
2 2 2
troleum ether 30-40/CH Cl /Et O (8:1:1)] gave recovered
3
hydroxyatractylolide 4 (43.0 mg) and the dimer compounds as
a white solid (164 mg) and a mixture of three diastereomers
in a 1:3:4 ratio. Further purification by preparative chroma-
2
MHz, CDCl
3
tography [petroleum ether 30-40/CH
()-biatractylolide 1 (14.0 mg, 3%), (()-biepiasterolide 2 (43.0
mg, 10%), and diastereomer (()-29 (57.0 mg, 13%).
2 2 2
Cl /Et O (8:1:1)] gave
1
72.9.
a-Chloro-3-methyl-5,6,7,7a-tetrahydro-4H-benzofuran-
-one (19). To a solution of hydroxybutenolide 18 (50.0 mg,
.3 mmol, 1 equiv) in CH Cl (2.2 mL) under argon at room
(
7
2
0
2
,4
(()-Biatractylolide (1): white solid (14.0 mg, 3%); mp
2,4
2
2
temperature was added freshly distilled thionyl chloride (110
µL, 1.5 mmol, 5 equiv), and the mixture was stirred for 18 h.
The solvent and excess thionyl chloride were evaporated under
reduced pressure, and the residue was purified by flash
210-211 °C (from EtOH, lit. (+)-1 210-212 °C); R
[petroleum ether 30-40/CH Cl /Et O (8:1:1)]; IR (KBr, disk)
max (cm ) 3436br m, 2930s, 1757s, 1670m, 1648m, 1439m,
1105m, 1012m, 986m, 888m, 732m; NMR δH/2 (500 MHz,
CDCl ) 1.13 (3H, s), 1.22 (1H, td, J ) 12.0 and 6.0 Hz), 1.42
f
0.3
2
2
2
-
1
ν
chromatography [petroleum ether 30-40/CH
2
Cl
2
(9:1)] to give
0.25
3
the title compound 19 as a clear oil (29 mg, 52%): R
f
(1H, d, J ) 14.5 Hz), 1.59-1.61 (1H, m), 1.63-1.65 (2H, m),
1.71-1.75 (1H, m), 1.75 (3H, s), 1.90-1.99 (1H, m), 2.35 (1H,
br d, J ) 13.0 Hz), 2.65 (1H, dd, J ) 13.0, 3.5 Hz), 2.78 (1H,
t, J ) 13.0 Hz), 2.82 (1H, d, J ) 14.5 Hz), 4.65 (1H, s), 4.86
-1
[
petroleum ether 30-40/Et
2
O/CH
2
Cl
2
(8:1:1)]; IR (film) νmax(cm )
(500
3
) 1.24-1.35 (1H, m), 1.69-1.76 (1H, m), 1.79-
2
915s, 2866m, 1778s, 1689m, 1440m, 1286m; NMR δ
H
MHz, CDCl
1
)
.89 (2H, m), 1.84 (3H, s), 2.02-2.08 (1H, m), 2.46 (1H, tm, J
(1H, s); NMR δC/2 (125.7 MHz, CDCl
3
) 8.4, 17.1, 22.3, 27.9, 35.8,
14.0 Hz), 2.66 (1H, dm, J ) 13.5 Hz), 2.76 (1H, dm, J )
(125.7 MHz, CDCl ) 8.2, 21.9, 24.2, 25.8,
1.2, 99.9, 120.6, 161.7, 170.7; MS (CI) m/z 187 [MH , 2%],
36.9, 42.1, 49.7, 52.9, 89.3, 107.3, 124.4, 147.9, 164.5, 172.0;
+
+
1
4
1
4.0 Hz); NMR δ
C
3
MS (EI) m/z 485 [MNa , 100%], 463 [MH , 36%]; HRMS calcd
+
+
for C30H
42NO
4
(MNH
4
) 480.3114, found 480.3109.
3
5
+
3
51 [100]. HRMS calcd for C
9
H
15 ClNO
2
(MNH
4
) 204.0791,
(()-Biepiasterolide (2): white solid (43.0 mg, 10%); mp
found 204.0787.
205-208 °C (from EtOH, lit.
ether 30-40/CH Cl /Et
3
203-205 °C); R 0.3 [petroleum
f
()-Hydroxyatractylolide (4).9,26 To a stirred solution of
-1
(
2
2
2
O (8:1:1)]; IR (KBr, disk) νmax (cm
)
furan 28 (69.0 mg, 0.2 mmol, 1 equiv) in CH Cl (2 mL) at
room temperature was added mCPBA (70% purity, 49.0 mg,
2
2
2925s, 1763s, 1670m, 1641m, 1432m, 1018m, 882m; NMR δH/2
(500 MHz, CDCl ) 0.50 (3H, s), 1.38-1.50 (2H, m), 1.53-1.59
3
0
3
.2 mmol, 1 equiv), and the reaction mixture was stirred for
0 min under argon. Five percent aqueous Na (1 mL)
(1H, m), 1.60-1.66 (1H, m), 1.79 (3H, d, J ) 2.0 Hz), 2.01 (1H,
d, J ) 15.0 Hz), 2.05-2.13 (1H, m), 2.19 (1H, ddd, J ) 17.5,
8.0, and 2.0 Hz), 2.32 (1H, d, J ) 15.0 Hz), 2.37 (1H, dm, J )
13.5 Hz), 2.65 (1H, dd, J ) 17.5 and 10.0 Hz), 2.82 (1H, dd, J
) 10.0 and 8.0 Hz), 4.64 (1H, s), 4.87 (1H, s); NMR δC/2 (125.7
2 2 3
S O
was added, and stirring continued for a further 10 min. The
aqueous layer was extracted with CHCl , and the organic
layers were combined, washed with saturated aqueous NaH-
CO and brine, dried (Na SO ), filtered, and concentrated in
vacuo. Purification by flash chromatography [petroleum ether
0-40/Et O (4:1)] gave the title compound 4 (27.0 mg, 55%)
as a white solid: mp 196-198 °C [lit. (+)-4 197-197.5 °C];
0.35 (7:3 petroleum ether 30-40/EtOAc); IR (KBr, disk) νmax
3
3
2
4
MHz, CDCl
3
) 8.6, 19.0, 22.8, 24.8, 36.6, 36.8, 42.2, 42.7, 47.0,
+
89.3, 107.1, 125.4, 148.1, 163.3, 172.6; MS (EI) m/z 485 [MNa ,
+
+
3
2
100%], 463 [MH , 20%]; MS (CI) m/z 463 [MH , 40%], 232
[100], 231 [55]; HRMS calcd for C30
found 480.3117.
9
+
4 4
H42NO (MNH ) 480.3114,
R
f
-
1
(
cm ) 3344br m, 1739s, 1697m, 1426m, 940m, 897m; NMR
(
()-Dimer 29. White solid (57.0 mg, 13%); np 240-242 °C
δ
1
1
H 3
(500 MHz, CDCl ) 1.04 (3H, s), 1.20-1.30 (1H, m), 1.54-
(
(
from EtOAc); R
f
0.3 [petroleum ether 30-40/CH
2
Cl
2
/Et
2
O
.62 (1H, m), 1.55 (1H, d, J ) 14.0 Hz), 1.60-1.75 (2H, m),
.82 (3H, s), 1.83-1.87 (1H, m), 1.93-2.00 (1H, m), 2.27 (1H,
-1
8:1:1)]; IR (KBr, disc) νmax (cm ) 3433 br m, 2927s, 1754s,
1
667m, 1645m, 1440m, 1261m, 1099m, 1018m, 894m; δ
MHz, CDCl ) 0.51 (3H, s), 1.16 (3H, s), 1.20 (1H, dd, J 13.0,
.5 Hz), 1.36-1.66 (7H, m), 1.45 (1H, d, J 14.5 Hz), 1.68-1.72
1H, m), 1.76 (3H, s), 1.78 (3H, d, J 2.0 Hz), 1.94 (1H, br q, J
H
(500
d, J ) 14.0 Hz), 2.36 (1H, br d, J ) 13.0 Hz), 2.45 (1H, t, J )
3.0 Hz), 2.63 (1H, d, J ) 13.0 Hz), 3.54 (1H, br s), 4.61 (1H,
3
1
4
s), 4.87 (1H, s); NMR δ (125.7 MHz, CDCl ) 8.1, 16.5, 22.2,
C
3
(
1
3.0 Hz), 1.98 (1H, d, J 15.0 Hz), 2.06 (1H, br t, J 13.0 Hz),
.34 (1H, br d, J 13.0 Hz), 2.34-2.38 (1H, m), 2.38-2.43 (1H,
m), 2.42 (1H, d, J 15.0 Hz), 2.48 (1H, t, J 12.5 Hz), 2.57-2.64
2
(
26) Huang, B. S.; Sun, J. S.; Chen, Z. L. Zhiwu Xuebao 1992, 34,
6
14.
J. Org. Chem, Vol. 69, No. 26, 2004 9107

Bagal et al.
(
1H, m), 2.61 (1H, dd, J 12.5, 3.5 Hz), 2.69 (1H, dd, J 10.0, 8.0
Hz), 2.77 (1H, d, J 14.5 Hz), 4.60 (1H, s), 4.66 (1H, s), 4.84
1H, s), 4.87 (1H, s); δ (125.7 MHz, CDCl ) 8.5 (2C’s), 18.2,
CDCl
3
) 14.1, 17.3, 21.4, 23.2, 36.3, 36.9, 37.9, 40.5, 41.5, 45.0,
107.3, 113.9, 147.2, 148.7, 180.2.
(
C
3
Further purification by preparative chromatography [pe-
1
4
1
1
8.4, 22.4, 22.7, 25.6, 27.5, 35.6, 36.4, 36.5, 36.9, 42.3 (2C’s),
troleum ether 30-40/CH Cl /Et O (8:1:1)] gave (()-biatracty-
2
2
2
3.2, 46.8, 50.8, 54.0, 88.0, 90.2, 106.8, 107.2, 124.0, 126.1,
lolide 1 (2.1 mg, 5.5%), (()-biepiasterolide 2 (1.4 mg, 3.5%),
and diastereomer (()-29 (1.8 mg, 4.5%).
+
47.9, 148.0, 163.9, 165.7, 172.3, 172.5; MS (CI) m/z 463 [MH ,
+
3%], 233[100]; HRMS calcd for C30
H
39
O
4
(MH ) 463.2848,
The alkaline layer was adjusted to pH 7 with 1 M aqueous
found 463.2841.
2
HCl, and the aqueous layer was extracted with Et O. The
Autoxidation of (()-Atractylon (7). All yields quoted are
based upon recovered atractylon 7, and all diastereomeric
ratios were measured after purification.
2 4
organic extracts were washed with water, dried (Na SO ),
filtered, and concentrated in vacuo. Purification by flash
chromatography [petroleum ether 30-40/EtOAc (7:3)] gave
Procedure 1: Aerial Oxidation under Hikino’s Condi-
hydroxyatractylolide 4 as a white solid (14.9 mg, 35%).
9
25
tions. A mixture of (()-atractylon 7 (50.0 mg, 0.23 mmol) in
methanol (1 mL) was left standing open to air for 35 days.
The solvent was removed under reduced pressure. The reaction
Procedure 2: Woodward Conditions.
Neat (()-
atractylon 7 (100 mg, 0.46 mmol) was heated to 100 °C for 1
h with magnetic stirring while oxygen was bubbled through
the flask. The reaction mixture was cooled to room tempera-
mixture was diluted with Et
organic layer was dried (Na
in vacuo. Purification by flash chromatography [petroleum
ether 30-40/CH Cl (9:1)] gave recovered atractylon 7 (13 mg),
atractylolide 3 (4.6 mg, 11.5%), and atractylolide isomer 30
5.4 mg, 13.5%) as a mixture of diastereomers in a 1:1 ratio.
2
O and 1 M aqueous NaOH. The
2
SO ), filtered, and concentrated
4
ture and diluted with Et O and 1 M aqueous NaOH. The
2
organic layer was dried (Na SO ), filtered, and concentrated
2
4
2
2
in vacuo. Purification by flash chromatography [petroleum
ether 30-40/CH Cl (9:1)] gave recovered atractylon 7 (13.8
2
2
(
mg). Further purification by preparative chromatography
The diastereomers of isomer 30 were not separated further,
although NMR data for the individual compounds could be
obtained by using COSY, HMQC, and HMBC NMR analysis.
[petroleum ether 30-40/CH Cl /Et O (8:1:1)] gave (()-biatrac-
2
2
2
tylolide 1 (2.5 mg, 2.5%), (()-biepiasterolide 2 (3.4 mg, 3.5%),
and diastereomer (()-29 (3.4 mg, 3.5%).
(
()-3,8a-Dimethyl-5-methylene-4a,5,6,7,8,8a,9,9a-oc-
The alkaline layer was adjusted to pH 7 with 1 M aqueous
27
tahydro-3H-naphtho[2,3-b]furan-2-one (30): colorless oil
obtained as a mixture of diastereomers in a 1:1 ratio; R 0.4
O (8:1:1)]; IR (film) νmax
cm ) 3400w, 2964s, 2930s, 1798s, 1718m, 1643m, 1262s,
HCl, and the aqueous layer was extracted with Et
2
O. The
SO ),
f
organic extracts were washed with water, dried (Na
2
4
[
(
2 2 2
petroleum ether 30-40/CH Cl /Et
1
filtered, and concentrated in vacuo. Purification by flash
chromatography [petroleum ether 30-40/EtOAc (7:3)] gave
hydroxyatractylolide 4 as a white solid (15.4 mg, 15.5%).
-
+
1
076s, 1015s, 799m; MS (CI) m/z 233 [MH , 100%], 250
+
+
[
MNH
found 233.1537.
. NMR Data for Diastereomer 30a: δ
4
, 38%]; HRMS calcd for C15
H
21
O
2
(MH ) 233.1542,
Acknowledgment. We thank Oxford University for
funding for S.K.B. and Dr. A. R. Cowley for X-ray crystal
structure determinations.
1
H
(500 MHz, CDCl )
3
0
2
3
.81 (3H, s), 1.34 (3H, d, J ) 6.0 Hz), 1.43-1.71 (4H, m), 2.00-
.15 (5H, m), 2.18-2.25 (1H, m), 2.40 (1H, dm, J ) 13.0 Hz),
.10-3.18 (1H, m), 4.61 (1H, s), 4.85 (1H, s); δ
C
(125.7 MHz,
) 14.2, 17.4, 21.6, 23.2, 36.4, 36.9, 37.9, 41.1, 41.5, 45.3,
07.4, 114.0, 147.3, 148.8, 180.3.
H 3
. NMR Data for Diastereomer 30b: δ (500 MHz, CDCl )
CDCl
1
3
Supporting Information Available: General experimen-
tal methods; experimental procedures and spectroscopic data
for compounds 3, 6, 7, 12, 16-18, 21-24, 26, 28, and CoCl-
2
(
3 3
PPh ) ; copies of NMR spectra for compounds 1, 2, 3, 4, 7,
0
2
3
.81 (3H, s), 1.33 (3H, d, J ) 6.0 Hz), 1.43-1.71 (4H, m), 2.00-
1
4a/14b, 19, 29, and 30; and X-ray crystallographic data for
.15 (5H, m), 2.18-2.25 (1H, m), 2.40 (1H, dm, J ) 13.0 Hz),
compounds 1, 2, 4, and 14a. This material is available free of
charge via the Internet at http://pubs.acs.org.
C
.10-3.18 (1H, m), 4.61 (1H, s), 4.85 (1H, s); δ (125.7 MHz,
(
27) Bohlmann, F.; Dutta, L. N.; Knauf, W.; Robinson, H.; King, R.
M. Phytochemistry 1980, 19, 433.
JO0488053
9108 J. Org. Chem., Vol. 69, No. 26, 2004