Structure elucidation of arteannuin O, a novel cadinane diol from Artemisia annua, and the synthesis of arteannuins K, L, M and O

Article abstract of DOI:10.1016/S0040-4020(01)00633-0

The novel cadinane diol, arteannuin O (1), has been obtained from Artemisia annua and its structure has been established by 2D NMR and X-ray crystallography. A reconstructive synthesis of arteannuin O from artemisinin is described, which also yields the natural products arteannuin K and arteannuin L. Mechanistic considerations have led to the conclusion that the stereochemistry of the 5-hydroxyl group was wrongly assigned when arteannuins K, L and M were first reported as natural products. This was confirmed by derivatization of synthetic arteannuins K, L and M as their Mosher esters.

Full text of DOI:10.1016/S0040-4020(01)00633-0

TETRAHEDRON
Pergamon
Tetrahedron 57 -2001) 8481±8493
Structure elucidation of arteannuin O, a novel cadinane diol from
Artemisia annua, and the synthesis of arteannuins K, L, M and O
Lai-King Sy, Kung-Kai Cheung, Nian-Yong Zhu and Geoffrey D. Brownp
Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, People's Republic of China
Received 7 March 2001; revised 15 May 2001; accepted 7 June 2001
AbstractÐThe novel cadinane diol, arteannuin O -1), has been obtained from Artemisia annua and its structure has been established by 2D
NMR and X-ray crystallography. A reconstructive synthesis of arteannuin O from artemisinin is described, which also yields the natural
products arteannuin K and arteannuin L. Mechanistic considerations have led to the conclusion that the stereochemistry of the 5-hydroxyl
group was wrongly assigned when arteannuins K, L and M were ®rst reported as natural products. This was con®rmed by derivatization of
synthetic arteannuins K, L and M as their Mosher esters. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
artemisinin. We herein report a re-investigation of this A.
annua extract from which we have obtained the novel
cadinane diol arteannuin O -1), which is a diastereoisomer
of arteannuin M. Synthesis of arteannuin O from dihydro-
epi-deoxyarteannuin B -3) has led us to propose a structure
revision of the stereochemistry claimed for the 5-OH group
in arteannuins K, L and M.
Following the discovery of the important anti-malarial
amorphane sesquiterpene artemisinin -qinghaosu) -2)1
from Artemisia annua in the 1970s, there have been exten-
sive phytochemical investigations of this species, resulting
in the description of some 35 amorphane/cadinane
sesquiterpenes at the time of writing.2,3 In 1998, we reported
three sesquiterpenes,2 arteannuins K, L and M from A.
annua, with structures apparently derived from dihydro-
epi-deoxyarteannuin B -3) -Fig. 1). These natural products
are of some interest from both biogenetic and mechanistic
perspectives, as it has been proposed that lactone diols such
as arteannuin M might undergo Grob fragmentation4 at C-4/
C-5 -the tertiary hydroxyl group behaving as an electron
donor and the ®ve-membered lactone behaving as a good
leaving group), either in vivo or in vitro, ultimately yielding
2. Results and discussion
HPLC separation of a polar fraction from column chromato-
graphy of the CH2Cl2 extract of A. annua2 has resulted in the
isolation of a novel cadinane diol, arteannuin O -1), in addi-
tion to its known diastereoisomer, arteannuin M. The planar
structure of 1 was rigorously established by 2D NMR
experiments such as HSQC, HMBC and 1H±1H COSY
Figure 1. Structures of natural products arteannuins K, L and M as reported in the literature.2 The stereochemistry of the 5-OH group, which was reported as
depicted above, is believed to have been wrongly assigned in all cases. Structures of dihydro-epi-deoxyarteannuin B -3)2,5 and artemisinin -2)1 are also shown.
Keywords: arteannuin O; arteannuin K; arteannuin L; arteannuin M; artemisinin; dihydro-epi-deoxyarteannuin B; cadinane; Artemisia annua; 2D NMR; X-ray
crystallography; Mosher ester.
Corresponding author. Tel.: 1852-2859-7915; fax: 1852-2857-1586; e-mail: gdbrown@hkucc.hku.hk
p
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020-01)00633-0

8482
L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
Table 1. Structure elucidation of the natural product arteannuin O -1) by 2D NMR
Position
2and 3-bond correlations
from13C to 1H in HMBC
Correlations
from 1H to 1H in
1H±1H COSY
Correlations from 1H to 1H in
NOESY
a,b
dC
a
dH
1
2a
2b
3a
3b
4
5
6
7
8a
8b
9a
9b
10
41.7 -CH)
20.2 -CH2)
1.48
1.76
1.47
1.70
1.70
±
3.56, 0.93
±
±
1.70, 1.47
1.76
1.76
1.76
±
±
±
3.09, 1.74, 1.13
2.59, 1.13
2.59, 1.74
1.63
1.06
0.93
2.59, 1.14
±
3.09
1.06, 0.93
1.47, 0.93
1.76, 0.93
1.23c
33.9 -CH2)
3.56, 1.23
1.23c
72.4 -C)
72.7 -CH)
87.7 -C)
38.8 -CH)
23.9 -CH2)
3.56, 1.23
±
3.56
3.09, 1.14
3.09
±
3.56
±
3.09, 2.59, 1.23
±
2.59
1.74
1.13
1.06
1.63
1.37
3.09
3.09, 1.14
1.14
0.93
1.23
3.56, 3.09, 1.74, 1.14
2.59, 1.63, 1.14, 1.13, 1.06
1.74, 1.63, 1.37
1.74, 1.63, 1.48, 0.93
1.74, 1.37, 1.13, 1.06, 0.93
1.63, 1.13, 0.93
3.56, 2.59, 1.14
32.1 -CH2)
0.93
30.2-CH)
38.9 -CH)
±
9.3 -CH3)
20.0 -CH3)
26.7 -CH3)
0.93
1.14
±
3.09
±
11
12177.9 -C)
13
14
15
3.09, 2.59, 1.74
1.76, 1.63, 1.48, 1.47, 1.37, 1.06
3.56, 1.70
1.37
±
±
a
1H connected to 13C by a single bond determined from correlations observed in HSQC.
Multiplicity in 13C determined by DEPT.
Ambiguity as to which proton is involved in nOe with H-15 as chemical shifts are the same for the 3a- and 3b-positions.
b
c
which allowed the assignment of all protons and carbons in
the molecule -Table 1). However, the stereochemistry of the
vicinal diol remained ambiguous due to the possibility of the
A-ring adopting more than one conformation, which in turn
allows two alternative diastereoisomeric con®gurations of
the diol at C-4 and C-5 in 1 to be proposed, both of which
would be consistent with the observed NOESY correlations
for this natural product -Fig. 2). Unlike other chemical shifts
around the vicinal diol, the chemical shift of the 4-OH
proton -dH 3.37) was almost unaffected by the concentration
of the sample used in recording NMR spectra, and we
propose that this group is therefore axial -4b-OH) and
involved in intramolecular hydrogen bonding with the car-
bonyl group -i.e. the ®rst conformation appearing in Fig. 2).
Fortunately, arteannuin O could be crystallized and the
relative stereochemistry of the vicinal diol was then
unambiguously determined by X-ray crystallography as
4b-OH, 5a-OH -4S,5R) -Fig. 3).
Artemisinin -2) was converted into the cis-lactone 6 in two
steps. The ®rst step, conversion of 2 being ring-opened to
give methyl ester 4, is best effected by use of strongly acidic
conditions -otherwise the 1,2,4-trioxane ring is not com-
pletely opened)6,7 for a short period of time -in order to
prevent epimerization at the 1- and 7-positions).8,9 In the
second step, Robinson annulation of 4 in the presence of
barium hydroxide octahydrate was accompanied by
cleavage of the methyl ester group.10 If the acidic work-up
of this reaction is performed in a controlled way, then it is
possible to isolate the decalenone acid 5 in reasonable
yield.8 Work-up under harsher conditions induces intra-
molecular nucleophilic attack of the carboxylic acid at the
a,b-unsaturated ketone functionality, resulting in lactone 68
in good yield -Scheme 1).
Grignard reaction of 6 with methyl iodide then provided the
desired intermediate, compound 3; physical properties of 3
obtained by synthesis were identical to those reported for
the natural product dihydro-epi-deoxyarteannuin B.2,5
Presumably 3 is formed in situ by dehydration of the initial
products of Grignard reaction at the ketone group of 6,
compounds 7/8 -Fig. 4). One remarkable feature of this
reaction is that only the desired D4-dehydration product 3
was obtained, with no detectable formation of the
In order to obtain larger amounts of arteannuin O -1) for
studies of its reactivity under Grob fragmentation
conditions -as part of our on-going investigations into the
biogenetic origins of artemisinin) we decided to synthesize
1 from artemisinin -2) via dihydro-epi-deoxyarteannuin B
-3).5
Figure 2. Two possible con®gurations for the vicinal diol in 1, which would be consistent with observed nOe data for this compound, depending on the
1
conformation adopted by the A-ring. Critical NOESY correlations are shown by double headed arrows from H to 1H.

L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
8483
Figure 3. ORTEP diagrams of the natural product arteannuin O -1) and the synthetic product dihydro-epi-arteannuin B -14).
alternative D3 regio-isomer. This was contrary to our expec-
tation, based on recent precedents in the literature11,12 for
very closely related compounds -amorphanes and cadinanes
containing a tertiary hydroxyl group at the 4-position) which
are reported to undergo dehydration to give nearly equal
amounts of both double bond regio-isomers. One possible
explanation for this unexpected selectivity is the participa-
tion of the lactone group, which may be able to undergo

8484
L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
Scheme 1. Synthesis of arteannuins O, K and L -1, 15 and 16, respectively) from artemisinin -2) via dihydro-epi-deoxyarteannuin B -3). The correct
stereochemistry is shown at the 5-position for arteannuins K -15) and L -16), as determined by the results of Mosher ester studies.
reversible ring-opening under the conditions of the
reaction as shown in Fig. 4, thereby assisting formation of
the D4-isomer. Such selectivity is clearly desirable in the
context of this synthesis, and it acts to offset the main
dif®culty associated with this reaction, which is that the
optimized yield of 3 from 6 was quite low -26%). The
reasons for the low yield are believed to be the tendency
for the lactone ring to open -as in the formation of 5) or
to participate in Grignard reaction itself, as in the
formation of compounds 9±12 and 13 -Fig. 5, Tables 1
and 2), which have incorporated 2 and 3 equiv. of the
methyl Grignard reagent respectively. -N.B. The use of
fewer equivalents of Grignard reagent resulted only in the
recovery of starting material and lactone ring-opened
product 5).
nantly in the desired trans diol 1, possessing identical physi-
cal properties to the natural product arteannuin O. Also
separable from the crude reaction product, were small quan-
tities of the allylic alcohol 15, together with smaller
amounts of its regio-isomer 16 -Scheme 1) and trace quan-
tities of the chlorinated product 17 -Fig. 5; Tables 2and 3).
Although no molecular ion was seen for 17 in HREIMS, the
presence of chlorine and the relative stereochemistry at the
4- and 5-positions could be con®rmed by X-ray crystallo-
graphy -structure not shown). Formation of all four products
probably involves opening of the epoxide ring to yield a
tertiary carbocation at C-4, which is then either quenched
by addition of water/chloride -as in compounds 1/17) or
which eliminates a proton from H-3/H-15 -as in compounds
15/16).
Treatment of 3 with m-chloroperoxybenzoic acid13,14 gave a
single epoxide product, compound 14 -Scheme 1). Oxygen
was delivered to the less hindered a-face of the double bond
in 3 as shown by X-ray crystallography of 14 -Fig. 3). We
have recently isolated a novel natural product as a minor
constituent of A. annua -unpublished results) with NMR
spectra identical to those of synthetic dihydro-epi-
arteannuin B -14) reported in Tables 2and 3. Hydrolysis
of epoxide 14 under acidic conditions resulted predomi-
The NMR spectra of synthetic products 15 and 16 matched
those previously published for the natural products
arteannuins K and L, respectively.2 This was somewhat
surprising as direct opening of the a-epoxide group in 14
would be expected to yield allylic alcohol products contain-
ing a 5a-hydroxyl group, rather than the 5b-hydroxyl
functionality which has been claimed for these natural
products -see Fig. 1). -N.B. none of the products 1, 15, 16
or 17 underwent interconversion with one another under the
Figure 4. A possible explanation for the unexpected regio-selectivity observed in the conversion of 6 to only the D4-isomer 3 via dehydration of Grignard
addition products 7/8.

L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
8485
Figure 5. Side-products from reactions shown in Schemes 1 and 2.
Table 2. 13C NMR data -d, ppm) for novel compounds reported in Schemes 1 and 2and Fig. 4
Position
7
8
9a
10a
11a
12
13
14
17
27
28
29
1
2
3
4
5
6
7
8
49.0
23.3
40.5
71.1
48.4
84.9
43.7
23.9
32.5
30.0
39.8
178.0
9.4
49.1
21.1
39.1
69.9
46.0
86.1
43.3
23.8
32.4
30.3
39.8
10156.10
9.3
50.5
21.0
39.8
70.8
46.6
85.8
51.5
21.3
39.0
70.3
46.3
83.9
50.4
21.1
39.6
71.270.3
46.0
83.4
44.0
23.0
39.3
70.9
47.7
86.6
52.1
21.4
38.7
39.6
20.1
27.2
42.3
20.9
36.6
44.7
22.0
37.6
52.4
21.5
42.7
50.0
19.4
36.7
63.1
67.6
73.3
77.6 .8 72
210.7
88.4
45.7
73.4
40.7
22.5
37.3
31.8
40.5
178.6
60.6
84.6
39.5
74.0
85.7
41.1
79.7
88.0
38.6 .7 42
20.1
27.2
28.9
38.3
175.9
84.3
70.5
45.253.246.6
51.9
56.6
23.3
34.5
30.5
44.1
2106.7
9.3
19.8
30.4
29.2
±
21.0
23.5
34.2
30.9
45.9
20.8
26.5
28.9
106.1
179.3
23.4
31.7
29.5
39.1
23.9
32.4
29.9
39.0
23.5
32.3
30.6
40.7
23.3
35.0
30.6
37.2
9
10
11
12179.0
13
14
15
16
35.9
31.5
44.6
144.1
17.8
19.9
30.9
30.5
±
74.3
180.4
177.8
10.5
19.8
30.4
26.7
10.219.5
9.4
9.4
10.5
9.1
15.2
19.9
28.8
±
19.9
30.2
±
21.1
30.1
11.3
20.1
31.0
33.5b
20.0
24.2
±
20.0
31.4
±
21.0
26.4
±
19.9
25.6
±
19.6
27.2
±
b
17
±
±
±
±
2
±5.9
±
±
±
±
Assigned by the same 2D NMR techniques shown in Table 1.
Assigned as a mixture.
Interchangeable.
a
b
²
conditions of the reaction, and 15 and 16 are thus presumed
to be formed directly from the epoxide). Unfortunately,
neither synthetic compound 15 nor 16 could be obtained
as crystals for X-ray crystallographic determination of the
true stereochemistry at the 5-position of arteannuins K and
L. However, the secondary alcohol group in both
compounds did react cleanly with the R--2) and S--1)
forms of a-methoxy-a--tri¯uoromethyl)phenylacetyl chlo-
ride -MTPCl)15,16 yielding either the S- or R-a-methoxy-a-
-tri¯uoromethyl)phenylacetate esters -OMTP) 18/19 -from
15) and 20/21 -from 16). The complete 1H -and 13C) NMR
chemical shift assignments for all four derivatives were then
made by 2D NMR -Table 4), allowing the con®guration of
the secondary hydroxyl group to be reliably determined as
5R -5a-OH) for both 15 and 16 -by analysis of 1H chemical
shift differences between the diastereoisomeric ester pairs
made by subtracting the 1H chemical shifts for the R-OMTP
²
The R/S nomenclature is reversed when the acid chloride derivatizing
agent is converted into the Mosher ester product.

8486
L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
Table 3. 1H NMR data -d, ppm) for novel compounds reported in Schemes 1 and 2and Fig. 4
Position
7
8
9a
10a
11a
12
13
14
17
27
28
29
1
1.06
1.89
1.44
1.521.46
1.87
1.521.41
2.14
2.04
1.69
1.13
1.021.03
1.65
1.35
3.08
1.13
0.920.94
1.36
±
1.01
1.84
1.58
1.38
1.87
0.93
1.75
1.55
1.31
1.78
0.74
1.66
1.55
0.87
±
±
1.020.77
1.721.67
1.59
1.40
1.59
1.39
1.70
1.11
1.76
1.19
1.921.64
1.79
0.85
1.48
b
2a
2b
3a
3b
5a
5b
7
8a
8b
9a
9b
10
11
13
14
15
16
17
b
1.49
1.75
1.45
1.821.51
1.521.621.41
1.98
1.33
1.38
1.32 1.91
1.85
1.74
1
1.77
1.78
1.87
±
1.95
±
3.26
2.55
2.27
±
±
2.89
1.96
3.78
±
1.15
1.15
1.2 1.121.14
2.38
1.70
±
3.83
2.62
1.721.58
1.17
2.17
1.99
1.69
1.15
1.97
1.65
1.90
1.50
1.99
1.28
2.20
2.20
1.50
1.25
2.27
1.32
1.80
1.62
3.04
2.43
1.75
1.20
1.83
b
±
±
1.65
1.28
1.821.74
b
1.43
1.12
1.42
0.98
0.98
0.94
0.88
0.99
0.92 1.04
1.78
1.04
1.13
1.021.01
1.69
1.58
1.66
1.36
3.11
1.14
1.65
1.37
2.56
1.01
1.74
1.56
3.02
1.25
1.621.61
1.31
2.80
0.98
0.94
1.17
1.40
±
1.61
0.88
1.64
1.48
1.77
1.64
1.46
±
1.53
1.60
2.27
0.95
1.33
3.25
1.18
1.421.51
3.02
1.14
0.98
1.63
±
3.60
1.15
2.80
1.12
2.79
1.37
0.90
0.85
0.86
0.94
0.98
0.88
1.19
±
±
1.14
1.55
±
1.19
2.21
±
1.14
1.71
±
1.21
1.33
±
1.23
±
±
1.54
±
±
1.34
±
±
1.27c
1.18c
±
±
±
Assigned by the same 2D NMR techniques shown in Table 1.
Assigned as a mixture.
a
b
Not resolved.
Interchangeable.
c
ester 19/21 from those for the corresponding ester S-OMTP
ester 18/20). Inspection of Fig. 6 shows that all differences
to the left of the newly formed Mosher ester groups are
positive, while all values to the right are negative: such a
clear division in up®eld and down®eld shifts about the
OMTP functionality is generally considered as good
evidence that the Mosher ester is adopting the expected
conformation -shown in Fig. 6), which can then be used in
reliably assigning the absolute stereochemistry for a
secondary hydroxyl group.17
1
Table 4. H and 13C NMR data -d, ppm) for the S-OMTP esters of compounds 15, 16 and 22 -18, 20 and 23/25, respectively) and the R-OMTP esters of
compounds 15, 16 and 22 -19, 21 and 24/26, respectively) fully assigned by 2D NMR
dH
dC
Position
18
19
20
21
23
24
25
26
18
19
20
21
23
24
1
1.38
2.31
1.98
5.81
±
±
5.33
±
1.75
1.53
0.99
0.70
1.53
1.25
3.16
1.39
2.31
2.00
5.78
±
±
5.31
±
1.87
1.64
1.07
0.81
1.621.55
1.32
3.26
±
1.09
0.91
1.58
±
1.36
1.83
1.46
2.15
2.33
±
5.55
±
1.92
1.60
1.06
0.77
1.43
1.83
1.43
2.06
2.25
±
1.20
1.78
1.32
1.49
1.66
±
5.04
±
1.73
1.50
1.06
0.74
1.59
1.28
3.11
1.22
1.82
1.40
1.51
1.70
±
5.05
±
1.88
1.59
1.11.211
0.81
1.16.262
1.33
3.20
1.57
1.83
1.40
1.94
2.43
±
3.93
±
2.63
1.71
1.57
1.82
1.38
1.91
2.47
±
3.94
±
2.61
1.71
39.5
27.3
39.8
27.3
43.0
24.7
43.2
24.8
43.0
22.1
43.4
22.2
2a
2b
3a
3b
4
5
6
7
129.9
129.5
29.8
29.5
35.2
35.7
126.8
72.1
83.5
38.2
24.2
126.9
72.8
83.4
38.6
24.2
140.8
75.3
84.3
38.7
24.2
140.6
75.9
84.4
39.0
24.2
71.7
76.6
84.7
39.4
23.9
71.5
77.2
84.6
39.8
23.8
5.49
±
2.02
1.67
1.121.00
0.88
1.51
1.32
3.18
±
1.13
0.90
5.10, 5.08
±
8a
8b
9a
9b
10
11
12±
13
14
15
10
1.05
1.05
31.6
31.8
32.1
32.3
32.1
32.3
1.61
1.28
3.17
1.35
3.06
1.36
3.06
31.7
38.9
178.8
31.7
38.9
178.4
9.3
19.7
20.8
30.6
38.7
178.3
30.7
38.7
178.1
29.9
38.8
178.0
29.9
38.7
±
±
±
±
±
178.8
1.121.12 9.2 9.2 9.3
0.92
1.74
±
±
±
7.51
7.44
7.44
1.09
0.88
1.66
±
±
±
7.60
7.43
7.43
3.57
1.13
0.87
5.15, 5.12
±
±
±
7.49
7.43
7.43
1.08
0.88
1.43
±
±
±
7.67
7.44
7.44
1.09
0.90
1.40
±
±
±
7.61
7.47
7.47
9.3
9.3
0.92
1.75
±
±
±
7.52
7.43
7.43
55.4
19.6
21.2
166.5
84.5
131.8
127.2
128.6
130.1
55.3
19.3
118.0
165.7
84.5
131.9
127.3
128.6
129.9
55.7
20.0
118.1
165.4
84.8
131.6
127.5
128.6
129.8
55.6
19.9
26.4
166.6
84.7
131.8
127.6
128.7
130.0
20.0
25.6
166.0
84.8
131.5
127.5
128.7
130.0
166.5
84.7
20
30
±
±
7.60
7.44
7.44
3.53
±
±
131.4
127.5
128.7
129.9
55.4
40/80
50/70
60
7.52
7.44
7.44
3.59
20-OMe
3.523.51
3.55 3.35.452
55.6
The S-OMTP esters 18, 20 and 23/25 are formed from reaction of R-MTPCl with compounds 15, 16 and 22, respectively; the R-OMTP esters 19, 21 and 24/26
are formed from reaction of S-MTPCl with compounds 15, 16 and 22, respectively; the CF3 group in the Mosher ester was generally not observed due to 19F-13C
coupling.

L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
8487
Figure 6. Dd1H NMR values -S-OMTP±R-OMTP; see Table 4) for Mosher ester derivatives 18 and 19, which were used in determining the absolute
stereochemistry at the 5-position of compound 15; and Dd H NMR values -S-OMTP±R-OMTP; see Table 4) for Mosher ester derivatives 20 and 21, which
1
were used in determining the absolute stereochemistry at the 5-position for compound 16. The analysis of these results assume the conformation of the Mosher
ester which is depicted.
Hence, based on both mechanistic considerations and the
results of Mosher ester studies, we propose that the stereo-
chemistry at the 5-position for natural products arteannuins
K and L, which was originally assigned as 5S -5b-OH)
should be revised to 5R -5a-OH). Given that the 5-hydroxyl
stereochemistry of arteannuins K and L appears to have
been wrongly assigned, we next turned our attention to the
natural product arteannuin M, which was also isolated from
A. annua in the same study.2 Treatment of synthetic
dihydro-epi-deoxyarteannuin B -3) with osmium tetroxide18
resulted predominantly in the cadinane diol 22 -Scheme 2),
with NMR spectra matching those reported for arteannuin
M.
S-MTPCl with the secondary hydroxyl group in 22, which
yielded the S- and R-OMTP esters 23 and 24, respectively.
1H chemical shifts differences between the two Mosher ester
derivatives did indeed suggest that the stereochemistry
which was originally proposed at the 5-position for
arteannuin M as 5S -b-OH) should also be revised to 5R
-5a-OH) -Table 4, Fig. 7). However, the Mosher ester
results for compound 22 are slightly less convincing than
those for compounds 15 and 16, with the assignment of
absolute stereochemistry at the secondary hydroxyl group
resting on only a single positive chemical shift difference at
the 15-position -all other differences were negative). This
less satisfactory result may be due to hydrogen bonding
between the oxygen substituents at the 4- and 5-positions
in 23 and 24, which distorts the normally preferred confor-
mation of the Mosher ester, and perturbs the expected
pattern of negative/positive chemical shift differences.
Evidence for hydrogen bonding of the secondary hydroxyl
Because of the mechanism of the osmylation reaction, the
vicinal diol in 22 ought to have cis stereochemistry. Steric
considerations would suggest that osmium tetroxide is more
likely to attack the double bond in 3 from the less hindered
a-face -cf. a-epoxidation of 3 to 14 in Scheme 1), which in
turn leads to the conclusion that the true stereochemistry of
the vicinal diol in the natural product arteannuin M is 4R, 5R
-4a-OH, 5a-OH): in the original report, the stereochemistry
at the 4-position of arteannuin M was unde®ned, but the
5-position was assigned as 5S -5b-OH).
1
group to the oxygen at C-4 was seen in the H NMR
spectrum of 22, in which the H chemical shift for the
1
5-OH group underwent comparatively smaller concentra-
tion-dependent changes than was the case for the 4-OH
group. The yields of both Mosher esters from 22 were
extremely low -see Section 3), and this might also be
ascribed to a reduced reactivity of the secondary hydroxyl
group as a result of hydrogen bonding with the tertiary
hydroxyl group.
The need for structure revision of arteannuin M, as well as
arteannuins K and L, was con®rmed by the reacion of R- and
Scheme 2. Synthesis of arteannuin M -22) from dihydro-epi-deoxyarteannuin B -3). The correct stereochemistry is shown for the 5-position of arteannuin M,
as determined by the results of Mosher ester studies with 22. The stereochemistry at the 4-position is more tentatively assigned based on the expected cis
mechanism for the osmylation reaction. Cadinane diol 27 was obtained as a minor product.

8488
L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
Figure 7. Dd 1H NMR values -S-OMTP±R-OMTP: see Table 4) for Mosher ester derivatives 23 and 24 used in tentatively assigning the absolute stereo-
chemistry at the 5-position of compound 22 assuming the conformation of the Mosher ester is as depicted; and Dd H NMR values -S-OMTP±R-OMTP: see
1
Table 4) for Mosher ester derivatives 25 and 26 used in a failed attempt to determine the stereochemistry at the 4-position for compound 22Ðthe ester
probably adopts no single well-de®ned conformation in this case.
Rather unexpectedly, we were also able to obtain the S- and
R-OMTP esters of the tertiary hydroxyl group in 22,
compounds 25 and 26, respectively. Analysis of 1H
chemical shift differences between 25 and 26 -Table 4,
Fig. 7) was inconclusive in assigning the absolute stereo-
chemistry at the 4-positionÐchemical shift differences
were small and there is no clear pattern of positive and
negative values. Most probably, the Mosher esters of the
tertiary hydroxyl group in 22 adopt no single well-de®ned
conformation Ð indeed we are unaware of any previous
attempts to study the Mosher esters of tertiary alcohols. In
summary, the results of Mosher ester studies with 22,
although lending some support to the revision of the stereo-
chemistry at the 5-position of arteannuin M to 5R -5a-OH),
do not shed any light on the stereochemistry of the tertiary
hydroxyl group at C-4, and this is assumed to be 4R on
mechanistic grounds only. Further synthetic studies in
order to establish more rigorously the stereochemistry of
the vicinal diol in arteannuin M, particularly at the 4-posi-
tion, would be most welcome.
arteannuin O -1a), arteannuin K -15a) and arteannuin L
-16a) -as well as the chloro-compound 17a) have been
³
synthesised by substituting [15-13C]-dihydro-epi-deoxy-
arteannuin B -3a), obtained by use of 13CH3MgI in the
Grignard reaction with 6, for isotopically normal 3 in the
epoxidation and acid hydrolysis reactions shown in Scheme
1. These labeled compounds will be used in future feeding
experiments with A. annua in order to establish the status of
arteannuins K, L and O as potential precursors in the
biogenesis of artemisinin -2) in vivo.
3. Experimental
3.1. General
Chemical shifts are expressed in ppm -d) relative to TMS as
internal standard. Proton chemical shifts, multiplicities,
coupling constants and integrals reported in this section
are those which are clearly resolved in 1D 1H NMR without
recourse to 2D NMR analysis -see Tables 1±4 for full
assignments by 2D NMR). All NMR experiments were
run on a Bruker DRX 500 instrument. HSQC, HMBC,
1H±1H COSY and NOESY spectra were recorded with
1024 data points in F2 and 256 data points in F1. High-
resolution MS were recorded in EI mode at 70 eV on a
Finnigan-MAT 95 MS spectrometer. IR spectra were
recorded in CHCl3 on a Shimadzu FTIR-8201 PC instru-
ment. Column chromatography -CC) was performed using
silica gel 60±200 mm -Merck). HPLC separations were
performed using a Varian chromatograph equipped with
RI star 9040 and UV 9050 detectors and a YMC diol
2 0 mm£25 cm column, ¯ow rate 8 ml min21. Melting
points were recorded by a Perkin±Elmer differential scan-
ning calorimeter 7 -DSC7). Optical rotations were measured
by a Perkin±Elmer 343 Polarimeter with polarized light -Na
A second cadinane diol, compound 27, was also obtained as
a minor product from the osmylation reaction. The NMR
spectra of 27 are different from those of both 1 and 22, and
this compound has been assigned as 4S, 5S -4b-OH, 5b-
OH), on the basis of correlations observed in NOESY and
mechanistic considerations, which require a cis-diol.
Compound 27 is the third of four diastereoisomers which
are possible from di-hydroxylation of the double bond in
dihydro-epi-deoxyarteannuin B; because 27 was only
isolated in small amounts, presumably as a consequence
of steric hindrance to osmylation of the b-face of 3, we
have been unable to obtain further evidence to corroborate
the proposed stereochemistry. The structures of two
further minor products of the osmylation reaction,
compounds 28 and 29, are shown in Fig. 5 -see also Tables
2and 3).
³
The suf®x `a' is used to indicate that the isotopically normal [15-CH3]
group has been replaced by [15-13CH3].
[15-13C]-Isotopomers of all of the natural products,

L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
8489
589 nm). [a]D Values are given in 1021 deg cm2 g21 and
CHCl3 was used as a solvent.
NMR -CDCl3): 207.0 -C-4), 178.0 -C-12), 85.9 -C-6),
50.7 -C-5), 48.1 -C-1), 43.3 -C-7), 40.3 -C-3), 39.9
-C-11), 32.5 -C-9), 31.2 -C-10), 24.4 -C-8), 24.0 -C-2),
19.9 -C-14), 9.2-C-13); HREIMS m/z -rel. int.) 236.1412
[M1, C14H20O3 requires 236.1415] -93), 208 -100), 180
-75), 179 -72), 163 -67), 134 -35).
3.2. Isolation of the natural products, arteannuins M
and O
Leaves of A. annua were collected from the mountains in
You Yang Country in Sichuan Province, Southern China, in
July and dried in the sun for two weeks before being sent to
Hong Kong for extraction. The dried leaves -1 kg) were
pulverized to a ®ne powder under liquid N2, repetitively
extracted with CH2Cl2 -AR grade), dried -MgSO4) and
solvent was removed under reduced pressure to yield a
dark green gum -97 g; 9.7% w/w). A portion of the extract
-25 g) was subjected to gradient column chromatography
-developing solvent: 100% n-hexane to 100% EtOAc) and
crude fractions eluted by 40±50% EtOAc/n-hexane were
further puri®ed by CC and HPLC -38% EtOAc/n-hexane).
Arteannuin M -158 mg, Rt 25.4 min)Ðsee Ref. 2 for physi-
cal properties. Arteannuin O -1): Solid. -182mg, Rt
21.4 min). [a]Dˆ286.9 -c 0.64, CHCl3); IR nmax -CHCl3):
3574, 2937, 1773, 1460 cm21; 1H NMR -CDCl3): 3.56 -1H,
s), 3.37 -1H, br s, 4-OH), 3.09 -1H, dq, Jˆ6.8, 7.1 Hz), 2.59
-1H, ddd, Jˆ11.0, 6.8, 5.4 Hz), 1.23 -3H, s), 1.14 -3H, d,
Jˆ7.1 Hz), 0.93 -3H, d, Jˆ6.4 Hz) Ð see Table 1 for full
assignments; 13C NMR -CDCl3): see Table 1; HREIMS m/z
-rel. int.) 268.1675 [M1, C15H24O4 requires 268.1675] -1),
250 -5), 232 -8), 195 -100), 192 -28), 179 -45), 177 -35).
3.4. Preparation of 3 by reaction of 6 with a methyl
Grignard reagent
To small Mg chips -1.33 g, 54.7 mmol) in anhyd. Et2O
-250 ml) was added
a solution of MeI -4.16 ml,
66.8 mmol) in anhyd. Et2O -100 ml) and the reaction
mixture was re¯uxed for 1.5 h. A solution of lactone 6
-6.5 g, 27.5 mmol) in anhyd. Et2O -250 ml) was added drop-
wise and the reaction was allowed to re¯ux for a further
2.5 h. Completion of the reaction was determined by TLC
and the reaction mixture was cooled in an ice bath and H2O
-250 ml) was added to destroy the excess Grignard reagent.
The mixture was extracted with Et2O -5£100 ml) and the
combined organic layers for this `neutral extract' were
washed by brine -3£50 ml), dried -MgSO4) and solvent
was removed under reduced pressure to yield a crude
product -4.2g; 65% w/w) consisting of compounds 7±13
which were separated by HPLC -30% EtOAc/n-hexane).
The aqueous layer left behind after the `neutral extraction'
was acidi®ed by dropwise addition of HCl -3 M) to pH 2and
then extracted again by Et2O -3£100 ml). The combined
organic layers from this `acidic extract' were washed with
brine -3£50 ml), dried -MgSO4) and solvent was removed
under reduced pressure to yield a crude mixture -2.1 g; 32%
w/w) consisting mostly of dihydro-epi-deoxyarteannuin B
-3) together with a little of decalenone acid 5, which could
be separated by CC -10% EtOAc/n-hexane). Compound 3
-1.65 g, 7.05 mmol; 26%)Ðsee also Refs. 2 and 5 for physi-
cal properties: Oil. [a]Dˆ1104.1 -c 0.32, CHCl3); IR nmax
3.3. Conversion of artemisinin /2) to lactone 6 by acid
degradation and Robinson annulation
To a cooled solution of conc. H2SO4 -120 ml) in MeOH
-180 ml) in an ice bath was added artemisinin -2) -12.0 g,
42.6 mmol). The mixture was stirred for 10 min -a colour
change to pale yellow was noted) after which iced water
-250 ml) was added to quench the reaction. The mixture
was ®ltered and extracted with CHCl3 -5£100 ml) and the
combined organic layers were washed with brine -3£50 ml)
and dried -MgSO4), then solvent was removed under
reduced pressure to yield a crude product -9.05 g; 75%
w/w) consisting predominantly of methyl ester 4Ðsee
Ref. 8 for physical properties. BaOH2´8H2O in MeOH
-8 g; 150 ml) was added to the crude product -9.05 g) and
the reaction mixture was stirred at room temperature for 3 h,
then ®ltered and neutralized -to pH 7) by dropwise addition
of CH3COOH, under ice bath cooling. The volume of the
solvent was reduced in vacuo -to ca. 30 ml), and the mixture
was then gradually acidi®ed by dropwise addition of HCl
-3 M) to pH 2, while stirring was maintained for 15 min in
an ice bath. The mixture was extracted by CHCl3
-3£100 ml), washed with brine -3£20 ml), dried -MgSO4)
and the solvent was removed under reduced pressure to
yield a crude product -8.0 g; 88% w/w) consisting mostly
of lactone 6 -6.8 g, 28.8 mmol; 68%) which was puri®ed by
recrystallization from EtOAc -to remove impurities of
uncyclized compound 5). Compound 6Ðsee Ref. 8 for
physical properties: Solid. Mp 171±1728C; [a]Dˆ250.9
-c 3.0, CHCl3); IR nmax -CHCl3): 3024, 2934, 2881, 1771,
-CHCl3): 3020, 2939, 2876, 1751, 1668, 1454 cm21; H
1
NMR -CDCl3): 5.64 -1H, d, Jˆ1.4 Hz), 3.14 -1H, dq,
Jˆ7.0, 7.2Hz), 1.69 -3H, s), 1.14 -3H, d, Jˆ7.2Hz), 0.94
-3H, d, Jˆ6.6 Hz); 13C NMR -CDCl3): 179.4 -C-12), 142.2
-C-4), 121.9 -C-5), 83.2 -C-6), 46.6 -C-1), 42.9 -C-7), 39.6
-C-11), 32.5 -C-9), 30.9 -C-3), 29.7 -C-10), 23.7 -C-15),
23.4 -C-8), 21.1 -C-2), 19.6 -C-14), 9.4 -C-13); HREIMS
m/z -rel. int.) 234.1616 [M1, C15H22O2 requires 234.1620]
-10), 219 -5), 190 -100), 175 -20), 161 -85). Compound 5
-0.20 g, 0.85 mmol; 3%)Ðsee Ref. 8 for physical proper-
ties. Compound 7: Oil. -Rt 34.6 min, 1.00 g, 3.97 mmol;
15%); [a]Dˆ268.5 -c 0.78, CHCl3); IR nmax -CHCl3):
1
3587, 3009, 2974, 2934, 2880, 1765, 1456 cm21; H NMR
-CDCl3): 3.08 -1H, dq, Jˆ6.9, 7.1 Hz), 2.14 -1H, dd,
Jˆ13.9, 2.1 Hz), 2.04 -1H, ddd, Jˆ11.7, 6.9, 5.3 Hz), 1.52
-1H, d, Jˆ13.5 Hz), 1.44 -1H, dddd, Jˆ11.9, 11.9, 11.9,
2.3 Hz), 1.36 -3H, s), 1.13 -3H, d, Jˆ7.1 Hz), 0.92-3H, d,
Jˆ6.6 Hz) - see Table 3 for full assignments; 13C NMR: see
Table 2; HREIMS m/z -rel. int.) 252.1722 [M1, C15H24O3
requires 252.1725] -6), 234 -38), 206 -17), 179 -36), 178
-60), 161 -100), 160 -68), 151 -54). Compound 8: Oil. -Rt
15.6 min, 0.79 g, 3.13 mmol; 11%); [a]Dˆ288.6 -c 0.61,
CHCl3); IR nmax -CHCl3): 3568, 3007, 2976, 2936, 2880,
1
1
1719, 1456 cm21; H NMR -CDCl3): 3.02-1H, dq Jˆ6.9,
1767, 1456 cm21; H NMR -CDCl3): 3.31 -1H, br s, OH),
7.1 Hz), 2.79 -1H, dd, Jˆ14.9, 2.0 Hz), 2.54 -1H, ddd,
Jˆ15.4, 2.5, 2.5 Hz), 2.35 -1H, d, Jˆ14.9 Hz), 2.33 -1H,
m), 1.13 -3H, d, Jˆ7.1 Hz), 1.00 -3H, d, Jˆ6.5 Hz); 13C
3.11 -1H, dq, Jˆ6.8, 7.1 Hz), 2.17 -1H, dd, Jˆ14.5, 2.3 Hz),
1.99 -1H, ddd, Jˆ11.0, 6.8, 6.3 Hz), 1.58 -1H, dddd,
Jˆ12.5, 12.5, 12.5, 3.3 Hz), 1.46 -1H, dd, Jˆ13.4,

8490
L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
4.2Hz), 1.41 -1H, d, Jˆ14.5 Hz), 1.36 -1H, m), 1.19 -3H,
s), 1.14 -3H, d, Jˆ7.1 Hz), 0.94 -3H, d, Jˆ6.6 Hz)Ðsee
Table 3 for full assignments; 13C NMR: see Table 2;
HREIMS m/z -rel. int.) 252.1730 [M1, C15H24O3 requires
252.1725] -20), 234 -30), 219 -5), 191 -20), 179 -65), 178
-100), 161 -75). Compounds 9±11 were isolated as an
inseparable mixture by HPLC -Rt 17.5 min, 0.98 g,
3.66 mmol; 13%): Oil. IR nmax -CHCl3): 3429 -br), 2932,
2874, 2849, 1688, 1454 cm21; HREIMS m/z -rel. int.)
268.2040 [M1, C16H28O3 requires 268.2038] -3), 250
-100), 232 -20), 217 -17), 178 -30), 160 -55). The planar
structures of all three compounds were determined, as a
mixture, by correlations observed in the 2D NMR experi-
3.4.1. Reaction of 6 with 1 equiv. of Grignard reagent.
When the same procedures described in Section 3.4 were
repeated with 1 equiv. of Grignard reagent being added to
the lactone 6 -50 mg, 0.22 mmol), the starting material was
recovered from the `neutral fraction' -30 mg, 0.13 mmol;
60%) while the `acidic fraction' contained only compound
5 -15 mg, 0.06 mmol; 30%).
3.5. Epoxidation of dihydro-epi-deoxyarteannuin B /3) to
dihydro-epi-arteannuin B /14)
To a solution of compound 3 -1.6 g, 6.83 mmol) in CHCl3
-100 ml) was added m-chloroperoxybenzoic acid -mCPBA,
1.77 g, 50±55%). The reaction mixture was stirred over-
night at room temperature and when completion had been
determined by TLC, H2O -100 ml) was added and the reac-
tion mixture was extracted with CHCl3 -3£100 ml). The
combined organic extracts were washed successively with
NaHCO3 -30%; 3£50 ml) and brine -3£50 ml), dried
-MgSO4) and the solvent was removed under reduced
pressure to yield a crude product -1.59 g; 99% w/w)
which was puri®ed by column chromatography -15%
EtOAc/n-hexane) to afford epoxide 14 which was further
puri®ed by recrystallization from n-hexane. Compound 14
-1.52g, 6.08 mmol; 89%): Solid. Mp 133±135 8C;
[a]Dˆ212.0 -c 6.4, CHCl3); IR nm1ax -CHCl3): 3024,
3013, 2932, 2878, 1767, 1456 cm21; H NMR -CDCl3):
3.25 -1H, dq, Jˆ6.9, 7.1 Hz), 3.04 -1H, s), 2.43 -1H, ddd,
Jˆ10.4, 6.9, 6.5 Hz), 1.33 -3H, s), 1.18 -3H, d, Jˆ7.1 Hz),
0.88 -3H, d, Jˆ6.5 Hz)Ðsee Table 3 for full assignments;
13C NMR: see Table 2; HREIMS m/z -rel. int.) 250.1557
[M1, C15H22O3 requires 250.1568] -2), 232 -1), 222 -12),
193 -15), 180 -37), 179 -100).
1
ments HSQC, HMBC and H±1H COSY -Tables 2and 3).
Positive cross-peaks observed in NOESY -most obviously
2.56/3.02/2.80 for H-11; 1.97/1.99/2.38 for H-5b; 1.55/
2.21/1.40 for H-16; 1.01/1.25/0.98 for H-13 in 9, 10 and
11 respectively) indicated that all three compounds were
involved in chemical exchange with one another -see Fig.
5)Ðhence the observation that they cannot be separated
chromatographically. The stereochemistry of each of
compounds 9, 10 and 11 has been tentatively assigned
from negative correlations observed in the same NOESY
spectrum -the exchange process causes some ambiguity in
interpreting correlations due to dipole±dipole interactions).
1
Compound 9 -ca. 50% of the mixture by NMR): H NMR
-CDCl3): 4.93 -1H, br s, OH), 4.54 -1H, br s, OH), 2.56 -1H,
dq, Jˆ6.9, 7.1 Hz), 1.97 -1H, dd, Jˆ13.8, 2.5 Hz), 1.55 -3H,
s), 1.14 -3H, s), 1.01 -3H, d, Jˆ7.1 Hz), 0.90 -3H, d,
Jˆ6.6 Hz)Ðsee Table 3 for full assignments; 13C NMR:
see Table 2. Compound 10 -ca. 30% of the mixture by
1
NMR): H NMR -CDCl3): 4.73 -1H, br s, OH), 3.02-1H,
dq, Jˆ2.4, 7.3 Hz), 2.28 -1H, br s, OH), 2.21 -3H, s), 1.99
-1H, dd, Jˆ13.6, 2.5 Hz), 1.19 -3H, s), 1.25 -3H, d,
Jˆ7.3 Hz), 0.85 -3H, d, Jˆ6.6 Hz), 0.74 -1H, ddd,
Jˆ12.0, 12.0, 3.4 Hz)Ðsee Table 3 for full assignments;
13C NMR: see Table 2. Compound 11 -ca. 20% of the
mixture by NMR): 1H NMR -CDCl3): 4.84 -1H, br s,
OH), 4.64 -1H, br s, OH), 2.80 -1H, dq, Jˆ6.9, 7.3 Hz),
2.38 -1H, dd, Jˆ14.0, 2.5 Hz), 1.40 -3H, s), 1.17 -3H, s),
0.98 -3H, d, Jˆ7.3 Hz), 0.88 -3H, d, Jˆ6.7 Hz)Ðsee Table
3 for full assignments; 13C NMR: see Table 2. Compound 12
-Rt 9.2min, 0.42g, 1.68 mmol; 6%): Oil. [ a]Dˆ24.9 -c
0.32, CHCl3); IR nmax -CHCl3): 3471, 2930, 2872, 2856,
3.6. Preparation of compounds 1, 15, 16 and 17 from the
treatment of dihydro-epi-arteannuin B /14) with acid
To a solution of epoxide 14 -1.45 g, 5.80 mmol) in Et2O
-100 ml) was added HCl -3 M, 12ml) and the reaction
mixture was stirred overnight at room temperature. When
completion had been determined by TLC, H2O -50 ml) was
added and the mixture was extracted with Et2O -3£100 ml).
The combined organic layers were washed with brine
-3£30 ml), dried -MgSO4) and the solvent was removed
under reduced pressure to yield a crude product -1.43 g;
99% w/w), which was separated by HPLC -20% EtOAc/n-
hexane/1% CH3COOH) to afford compound 1 -Rt 51.7 min)
as the predominant product -1.23 g, 4.59 mmol; 79%).
Compound 1 was further puri®ed by recrystallization from
EtOAc and its structure was con®rmed to be identical with
the natural product arteannuin O by X-ray crystallography.
Compound 1: physical properties of 1 were similar to
arteannuin O. NMR spectra in CDCl3 solution for natural
and synthetic arteannuin O were identical within the resolu-
tion of the instrument -^0.01 ppm for 1H and ^0.1 ppm for
13C) when recorded at the same concentration -7.2mg/
0.6 ml CDCl3). Chemical shifts for some resonances in the
vicinity of the oxygen-containing functional groups showed
concentration-dependent changes -e.g. C-6 and C-12- DdC
0.2and 0.3 ppm, respectively) and H-5, 4-OH and 5-OH
-DdH 0.02, 0.03 and 0.46 ppm, respectively)) when more
concentrated solutions were studied. Molecular modeling
showed that the 4-OH group is involved in intramolecular
1
1717, 1456 cm21; H NMR -CDCl3): 4.59 -1H, br s, OH),
2.20 -1H, m), 2.00 -1H, dd, Jˆ14.2, 2.8 Hz), 1.71 -3H, s),
1.53 -3H, d, Jˆ0.9 Hz), 1.14 -3H, s), 0.94 -3H, d,
Jˆ6.8 Hz)Ðsee Table 3 for full assignments; 13C NMR:
see Table 2; HREIMS m/z -rel. int.) 250.1935 [M1,
C16H26O2 requires 250.1933] -100), 232 -12), 217 -24),
189 -20), 161 -36), 160 -46). Compound 13 -Rt 18.5 min,
0.27 g, 0.95 mmol; 4%): Oil. [a]Dˆ210.9 -c 0.22, CHCl3);
IR nmax -CHCl3): 3321 -br), 2970, 2930, 2872, 2849,
1
1456 cm21; H NMR -CDCl3): 4.68 -1H, br s, OH), 2.27
-1H, dq, Jˆ1.8, 7.3 Hz), 2.27 -1H, d, Jˆ14.4, 2.5 Hz),
1.27 -3H, s), 1.21 -3H, s), 1.18 -3H, s), 0.95 -3H, d,
Jˆ7.3 Hz), 0.92-3H, dddd, Jˆ12.5, 12.5, 12.5, 3.9 Hz),
0.86 -3H, d, Jˆ6.6 Hz), 0.77 -1H, ddd, Jˆ14.0, 10.7,
3.7 Hz)Ðsee Table 3 for full assignments; 13C NMR:
see Table 2; HREIMS m/z -rel. int.) 284.2350 [M1,
C17H32O3 requires 284.2351] -6), 266 -17), 248 -60), 233
-42), 230 -80), 215 -50), 193 -42), 190 -47), 180 -75), 161
-100).

L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
8491
hydrogen-bonding with the lactone functional group, whilst
the 5-OH group does not participate in intramolecular
hydrogen bonding, and these changes in chemical shift are
thus consistent with the expected changes in the extent of
hydrogen bonding of the 4-OH and 5-OH groups as the
concentration is changed. Solid. Mp 197±1988C; [a]Dˆ
2106.6 -c 0.4, CHCl3); IR nmax -CHCl3): 3568, 3410 -br),
IR nmax -CHCl3): 3422 -br), 3013, 2934, 2876, 1763,
1
1452cm 21; H NMR -CDCl3): 3.83 -1H, d, Jˆ5.8 Hz),
3.02-1H, dq, Jˆ6.8, 7.1 Hz), 2.62 -1H, ddd, Jˆ11.4, 6.8,
5.0 Hz), 2.29 -1H, d, Jˆ5.8 Hz, OH), 1.63 -3H, s), 1.14 -3H,
d, Jˆ7.1 Hz), 0.94 -3H, d, Jˆ6.3 Hz) - see Table 3 for full
assignments; 13C NMR -CDCl3): see Table 2; HREIMS m/z
-rel. int.) 250.1574 [M12HCl, C15H22O3 requires 250.1569]
-20), 232 -10), 222 -35), 207 -20), 192 -18), 180 -48), 179
-100), 167 -65).
1
3024, 2932, 2862, 1773, 1456 cm21; H NMR -7.2mg/
0.6 ml CDCl3): 3.55 -1H, s), 3.38 -1H, br s, 4-OH), 3.09
-1H, dq, Jˆ6.9, 7.1 Hz), 2.59 -1H, ddd, Jˆ10.9, 6.9,
5.0 Hz), 2.29 -1H, br s, 5-OH), 1.23 -3H, s), 1.14 -3H, d,
Jˆ7.1 Hz), 0.94 -3H, d, Jˆ6.4 Hz); 13C NMR -CDCl3):
178.0 -C-12), 87.7 -C-6), 72.7 -C-5), 72.5 -C-4), 41.7 -C-
1), 38.9 -C-7), 38.8 -C-11), 33.9 -C-3), 32.1 -C-9), 30.2 -C-
10), 26.7 -C-15), 23.9 -C-8), 20.2 -C-2), 20.0 -C-14), 9.2 -C-
13); HREIMS m/z -rel. int.) 250.1563 [M12H2O, C15H22O3
requires 250.1569] -10), 232 -15), 222 -10), 207 -12), 195
-100), 192-35), 179 -55), 177 -60). Compound 15 -Rt
28.0 min, 80 mg, 0.32 mmol; 6%): physical properties as
for arteannuin KÐsee Ref. 2. NMR spectra in CDCl3 solu-
tion for natural and synthetic arteannuin K were identical
3.6.1. Derivitization of 15 and 16 as Mosher esters. To a
solution of 15 -21.8 mg, 0.087 mmol) in pyridine -0.25 ml)
was added R--2)-a-methoxy-a--tri¯uoromethyl)phenyl-
acetyl chloride -0.05 ml, 0.29 mmol) and the solution was
allowed to stand overnight at room temperature. N,N-diiso-
propylethylamine -0.03 ml, 0.17 mmol) was added, and
after 10 min the solvent was evaporated to yield a crude
residue -40 mg) which was puri®ed by HPLC -13%
EtOAc/n-hexane) to afford the S-OMTP ester, compound
1
18 -Rt 15.6 min, 12.1 mg, 0.026 mmol; 30%): H NMR
-CDCl3): 7.60 -2H, m), 7.43 -3H, m), 5.81 -1H, br), 5.33
-1H, s), 3.57 -3H, s), 3.16 -1H, dq, Jˆ6.9, 7.1 Hz), 2.31 -1H,
m), 1.98 -1H, m), 1.75 -1H, ddd, Jˆ12.1, 6.9, 5.5 Hz), 1.66
-3H, s), 1.38 -1H, ddd, Jˆ11.0, 11.0, 5.3 Hz), 1.09 -3H, d,
Jˆ7.1 Hz), 0.88 -3H, d, Jˆ6.6 Hz), 0.70 -1H, ddd, Jˆ12.1,
1
within the resolution of the instrument -^0.01 ppm for H
and ^0.1 ppm for 13C) when recorded at the same concen-
tration -5.7 mg/0.6 ml CDCl3). The appearance of the H-5
resonance -dH 3.70 ppm) varied with concentration, appear-
ing as either a singlet, a broad singlet or a doublet
-Jˆ7.5 Hz). Molecular modeling showed that the 5-OH
group does not participate in intramolecular hydrogen bond-
ing, and the observed changes in multiplicity for the H-5
proton may be related to the rate of exchange of the 5-OH
proton to which it is coupled, as the concentration is
changed. Oil. [a]Dˆ2148.7 -c 0.62, CHCl3); IR nmax
1
12.1, 12.1 Hz)Ðsee Table 4 for full H and 13C NMR
assignments. The same procedure was applied to compound
15 with S--1)-a-methoxy-a--tri¯uoromethyl) phenylacetyl
chloride to yield a crude residue -42mg) which was puri®ed
by HPLC -13% EtOAc/n-hexane) to afford the R-OMTP
ester, compound 19 -Rt 16.9 min, 13.2mg, 0.028 mmol;
1
32%): H NMR -CDCl3): 7.60 -2H, m), 7.44 -3H, m),
-CHCl3): 3591, 3447 -br), 3028, 2928, 1759, 1456 cm21
;
5.78 -1H, br), 5.31 -1H, s), 3.53 -3H, s), 3.26 -1H, dq,
Jˆ6.9, 7.3 Hz), 2.31 -1H, m), 1.87 -1H, ddd, Jˆ11.9, 6.9,
5.5 Hz), 1.58 -3H, d, Jˆ1.8 Hz), 1.39 -1H, ddd, Jˆ11.0,
11.0, 5.2Hz), 1.09 -3H, d, Jˆ7.2Hz), 0.91 -3H, d,
1H NMR -5.7 mg/0.6 ml; CDCl3): 5.66 -1H, q, Jˆ1.7 Hz),
3.70 -1H, d, Jˆ7.5 Hz -coupling to OH)), 3.11 -1H, dq,
Jˆ6.9, 7.2Hz), 2.72-1H, ddd, Jˆ12.1, 6.9, 4.7 Hz), 2.27
-1H, m), 1.79 -3H, dd, Jˆ2.4, 1.7 Hz), 1.14 -3H, d,
Jˆ7.2Hz), 0.94 -3H, d, Jˆ6.4 Hz); 13C NMR -CDCl3):
179.5 -C-12), 131.2 -C-4), 126.4 -C-3), 85.2 -C-6), 70.1
-C-5), 39.0 -C-11), 38.8 -C-1), 38.0 -C-7), 32.1 -C-9),
31.7 -C-10), 27.2 -C-2), 24.4 -C-8), 21.2 -C-15), 19.8 -C-
14), 9.3 -C-13); HREIMS m/z -rel. int.) 250.1564 [M1,
C15H22O3 requires 250.1569] -3), 167 -100), 151 -20).
Compound 16 -Rt 26.9 min, 35 mg, 0.14 mmol; 2%): physi-
cal properties as for arteannuin LÐsee Ref. 2. NMR spectra
in CDCl3 solution for natural and synthetic arteannuin L
were identical within the resolution of the instrument
1
Jˆ6.4 Hz)Ðsee Table 4 for full H and 13C NMR assign-
ments.
To a solution of 16 -22 mg, 0.087 mmol) in pyridine
-0.25 ml) was added R--2)-a-methoxy-a--tri¯uoromethyl)-
phenylacetyl chloride -0.05 ml, 0.29 mmol) and the solution
was allowed to stand overnight at room temperature. N,N-
diisopropylethylamine -0.03 ml, 0.17 mmol) was added,
and after 10 min the solvent was evaporated to yield a
crude residue -42mg) which was puri®ed by HPLC -13%
EtOAc/n-hexane) to yield the S-OMTP ester, compound 20
-Rt 15.6 min, 13 mg, 0.028 mmol; 32%): 1H NMR -CDCl3):
7.49 -2H, m), 7.43 -3H, m), 5.55 -1H, s), 5.15 -1H, s), 5.12
-1H, s), 3.52-3H, s), 3.17 -1H, dq, Jˆ6.6, 7.1 Hz), 1.92-1H,
ddd, Jˆ12.1, 6.6, 5.4 Hz), 1.36 -1H, ddd, Jˆ10.7, 10.7,
3.2Hz), 1.13 -3H, d, Jˆ7.1 Hz), 0.87 -3H, d,
1
-^0.01 ppm for H and ^0.1 ppm for 13C) when recorded
at the same concentration -7.0 mg/0.6 ml CDCl3). No
signi®cant changes were noted in the NMR spectra when
the concentration was changed. Oil. [a]Dˆ248.8 -c 0.77,
CHCl3); IR nmax -CHCl3): 3400 -br), 3026, 2931, 2873,
1
1
1763, 1460 cm21; H NMR -CDCl3): 4.93 -1H, dd, Jˆ1.5,
Jˆ6.4 Hz)Ðsee Table 4 for full H and 13C NMR assign-
1.5 Hz), 4.91 -1H, dd, Jˆ1.5, 1.5 Hz), 4.10 -1H, s), 3.11
-1H, dq, Jˆ6.9, 7.2Hz), 2.60 -1H, ddd, Jˆ11.3, 6.9,
5.5 Hz), 1.14 -3H, d, Jˆ7.2Hz), 0.94 -3H, d, Jˆ6.5 Hz);
13C NMR -CDCl3): 179.2-C-12), 146.2-C-4), 114.5 -C-15),
86.2-C-6), 73.5 -C-5), 41.8 -C-1), 38.8 -C-11), 38.4 -C-7),
32.4 -C-9), 30.7 -C-10), 29.2 -C-3), 25.2 -C-2), 24.3 -C-8),
20.1 -C-14), 9.3 -C-13); HREIMS m/z -rel. int.) 250.1565
[M1, C15H22O3 requires 250.1569] -5), 232 -10), 222 -6),
204 -6), 179 -95), 177 -100). Compound 17: -Rt 34.9 min,
50 mg, 0.17 mmol; 3%) Oil. [a]Dˆ274.2- c 1.3, CHCl3);
ments. The same procedure was applied to compound 16
with
S--1)-a-methoxy-a--tri¯uoromethyl)phenylacetyl
chloride to yield a crude residue -39 mg) which was puri®ed
by HPLC -13% EtOAc/n-hexane) to afford the R-OMTP
ester, compound 21 -Rt 16.9 min, 11 mg, 0.024 mmol;
1
27%): H NMR -CDCl3): 7.52-H2 , m), 7.44 -3H, m),
5.49 -1H, s), 5.10 -1H, s), 5.08 -1H, s), 3.51 -3H, s), 3.18
-1H, dq, Jˆ6.6, 7.1 Hz), 1.13 -3H, d, Jˆ7.1 Hz), 0.90 -3H,
d, Jˆ6.6 Hz)Ðsee Table 4 for full 1H and 13C NMR assign-
ments.

8492
L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
3.7. cis-Hydroxylation of dihydro-epi-deoxyarteannuin
B /3) by OsO4
dddd, Jˆ14.2, 3.7, 3.4, 3.2 Hz), 1.54 -3H, s), 1.19 -1H,
ddd, Jˆ11.4, 10.7, 3.4 Hz), 1.12-3H, d, Jˆ7.1 Hz), 0.98
-3H, d, Jˆ6.6 Hz)Ðsee Table 3 for full assignments; 13C
NMR: see Table 2; HREIMS m/z -rel. int.) 266.1522 [M1,
C15H22O4 requires 266.1518] -13), 248 -6), 238 -90), 220
-26), 210 -42), 205 -18), 164 -34), 152 -100). Compound 29
-Rt 16.3 min, 1.4 mg, 0.005 mmol; 3%): Oil. [a]Dˆ222.8
-c 0.14, CHCl3); IR nmax -CHCl3): 3545, 2955, 2930, 2874,
To a solution of compound 3 -47 mg, 0.20 mmol) in
t-BuOH/H2O -1.5 ml/1.5 ml) was added K3Fe-CN)6
-0.198 g, 0.60 mmol), K2CO3 -0.083 g, 0.60 mmol) and a
small portion -0.05 ml) of a solution of OsO4 in t-BuOH,
-prepared from 1 g, 3.9 mmol, OsO4 in 80 ml t-BuOH,
incorporating several drops of t-BuOOH -70%)). The reac-
tion mixture was stirred overnight at room temperature and
completion was determined by TLC. Na2SO3 -0.05 g,
0.40 mmol) was added and stirring continued for a further
2h before the solution was concentrated to dryness under
reduced pressure and extracted with Et2O -3£20 ml). The
combined organic layers were washed with brine -3£5 ml),
dried -MgSO4) and solvent removed under reduced pressure
to yield a crude product -45 mg; 96% w/w) which was
separated by HPLC -40% EtOAc/n-hexane). Compound
22 -Rt 23.0 min, 31 mg, 0.12 mmol; 58%): physical proper-
ties as for arteannuin MÐsee Ref. 2. NMR spectra in CDCl3
solution for natural and synthetic arteannuin M were iden-
tical within the resolution of the instrument -^0.01 ppm for
1H and ^0.1 ppm for 13C) when recorded at the same
concentration -7.1 mg/0.6 ml CDCl3). Chemical shifts for
some resonances in the vicinity of the oxygen-containing
functional groups showed concentration-dependent changes
-e.g. C-4, C-5, C-6, C-12and C-15 - DdC 0.3, 0.2, 0.3, 0.3
and 0.3 ppm, respectively) and H-15, 4-OH and 5-OH -DdH
0.03, 0.73 and 0.38 ppm, respectively)) when more concen-
trated solutions were studied. Molecular modeling showed
that the 5-OH group is involved in an intramolecular hydro-
gen-bond to the oxygen of the 4-OH group, and these
changes in chemical shift are thus consistent with the
expected changes in the extent of hydrogen bonding of the
4-OH and 5-OH groups as the concentration is changed. Oil.
[a]Dˆ253.0 -c 0.88, CHCl3); IR nmax -CHCl3): 3510, 3421
1
1717, 1456 cm21; H NMR -CDCl3): 4.34 -1H, s, 6-OH),
3.78 -1H, s), 2.79 -1H, dq, Jˆ7.2, 7.5 Hz), 2.53 -1H, br s,
4-OH), 1.37 -3H, d, Jˆ7.5 Hz), 1.34 -3H, s), 0.88 -3H, d,
Jˆ6.6 Hz)Ðsee Table 3 for full assignments; 13C NMR: see
Table 2; HREIMS m/z -rel. int.) 250.1571 [M12H2O,
C15H22O3 requires 250.1569] -21), 232 -30), 195 -60), 192
-100), 179 -32), 177 -37).
3.7.1. Derivitization of 22 as Mosher esters 23±26. To a
solution of 22 in pyridine -13.5 mg, 0.050 mmol; 0.25 ml)
was added R--2)-a-methoxy-a--tri¯uoromethyl)phenyl-
acetyl chloride -0.2ml, 1.15 mmol) and the solution was
allowed to stand overnight at room temperature. N,N-diiso-
propylethylamine -0.2ml, 1.15 mmol) was added, and after
10 min the solvent was evaporated to yield a crude residue
-25 mg) which was puri®ed by HPLC -20% EtOAc/n-
hexane) to afford the S-OMTP ester of the secondary
hydroxyl group, compound 23 -Rt 20.8 min, 0.8 mg,
1
0.0017 mmol; 3%): Oil. H NMR -CDCl3): 7.67 -2H, m),
7.44 -3H, m), 5.04 -1H, s), 3.59 -3H, s), 3.11 -1H, dq, Jˆ6.9,
6.6 Hz), 1.78 -1H, dddd, Jˆ13.9, 3.7, 3.7, 3.7 Hz), 1.73 -1H,
ddd, Jˆ11.1, 6.9, 4.5 Hz), 1.43 -3H, s), 1.08 -3H, d,
Jˆ6.6 Hz), 0.88 -3H, d, Jˆ6.4 Hz), 0.74 -1H, dddd,
1
Jˆ12.8, 12.8, 12.8, 2.5 Hz)Ðsee Table 4 for full H and
13C NMR assignments. The same procedure was applied to
compound 22 with S--1)-a-methoxy-a--tri¯uoromethyl)
phenylacetyl chloride to yield a crude residue -28 mg)
which was puri®ed by HPLC -20% EtOAc/n-hexane) to
afford the R-OMTP ester of the secondary hydroxyl
group, compound 24 -Rt 20.4 min, 0.5 mg, 0.0011 mmol;
;
-br), 3024, 2941, 2876, 1763, 1456 cm21 1H NMR
-CDCl3): 3.45 -1H, s), 3.18 -1H, br s, 5-OH), 3.09 -1H,
dq, Jˆ6.9, 7.1 Hz), 2.65 -1H, ddd, Jˆ10.8, 6.9, 5.6 Hz),
1.52-1H, ddd, Jˆ12.5, 11.1, 3.3 Hz), 1.39 -3H, s), 1.34
-1H, dddd, Jˆ12.7, 12.7, 12.7, 3.7 Hz), 1.13 -3H, d,
Jˆ7.1 Hz), 0.92-3H, d, Jˆ6.4 Hz); 13C NMR -CDCl3):
179.2-C-1)2, 86.2-C-6), 74.2-C-5), 7.27 -C-4), 41.7
1
2%): Oil. H NMR -CDCl3): 7.61 -2H, m), 7.47 -3H, m),
5.05 -1H, s), 3.55 -3H, s), 3.20 -1H, dq, Jˆ6.8, 7.1 Hz), 1.88
-1H, ddd, Jˆ11.7, 6.8, 4.5 Hz), 1.82-1H, dddd, Jˆ13.5, 3.4,
3.4, 3.4 Hz), 1.40 -3H, s), 1.09 -3H, d, Jˆ7.1 Hz), 0.90 -3H,
d, Jˆ6.4 Hz)Ðsee Table 4 for full 1H and 13C NMR assign-
ments. Also isolated were the S-OMTP and R-OMTP esters
of the tertiary hydroxyl group in 22, compounds 25 and 26,
respectively. Compound 25 -Rt 17.1 min, 0.7 mg,
-C-1), 39.1 -C-7), 38.8 -C-11), 34.2-C-3), 3.23 -C-9),
29.9 -C-10), 26.6 -C-15), 23.9 -C-8), 22.1 -C-2), 20.1
-C-14), 9.4 -C-13); HREIMS m/z -rel. int.) 268.1676 [M1,
C15H24O4 requires 268.1675] -1), 250 -8), 222 -18), 195
-84), 179 -100). Compound 27 -Rt 18.4 min, 2.2 mg,
0.008 mmol; 4%): Oil. [a]Dˆ224.5 -c 0.22, CHCl3); IR
1
0.0015 mmol; 3%): Oil. H NMR -CDCl3): 7.51 -2H, m),
7.44 -3H, m), 3.93 -1H, s), 3.52-3H, s), 3.06 -1H, dq, Jˆ6.9,
7.3 Hz), 2.63 -1H, ddd, Jˆ10.9, 6.9, 5.6 Hz), 2.43 -1H, m),
2 .33 -1H, br s, OH), 1.94 -1H, ddd,Jˆ13.3, 13.3, 4.8 Hz),
nmax -CHCl3): 3553, 3026, 2936, 2878, 1763, 1456 cm21
;
1H NMR -CDCl3): 3.60 -1H, dq, Jˆ6.9, 7.5 Hz), 3.26 -1H,
d, Jˆ9.4 Hz), 2.69 -1H, d, Jˆ9.4 Hz, 5-OH), 2.65 -1H, br s,
4-OH), 2.55 -1H, ddd, Jˆ10.5, 6.9, 4.8 Hz), 1.23 -3H, s),
1.15 -3H, d, Jˆ7.5 Hz), 0.98 -3H, d, Jˆ6.6 Hz)Ðsee Table
3 for full assignments; 13C NMR: see Table 2; HREIMS m/z
-rel. int.) 250.1570 [M12H2O, C15H22O3 requires
250.1569] -7), 232 -15), 222 -5), 195 -100), 192 -48), 179
-38), 177 -38), 159 -35). Compound 28 -Rt 11.7 min,
3.0 mg, 0.011 mmol; 6%): Oil. [a]Dˆ269.0 -c 0.3,
CHCl3); IR nmax -CHCl3): 3510, 3026, 2934, 2862, 1778,
1713, 1456 cm21; 1H NMR -CDCl3): 3.77 -1H, s, OH), 2.89
-1H, ddd, Jˆ11.7, 6.9, 5.4 Hz), 2.80 -1H, dq, Jˆ6.9,
7.1 Hz), 2.27 -1H, ddd, Jˆ13.0, 3.2, 3.2 Hz), 1.92 -1H,
1.74 -3H, s), 1.12-3H, d,
Jˆ7.3 Hz), 0.92-3H, d,
Jˆ6.4 Hz)Ðsee Table 4 for full H and 13C NMR assign-
ments. Compound 26 -Rt 17.2min, 0.6 mg, 0.0013 mmol;
3%): Oil. 1H NMR -d CDCl3) ppm: 7.52 -2H, m), 7.43 -3H,
m), 3.94 -1H, s), 3.54 -3H, s), 3.06 -1H, dq, Jˆ6.9, 7.1 Hz),
2.61 -1H, ddd, Jˆ10.7, 6.9, 5.1 Hz), 2.47 -1H, d, Jˆ12.4 Hz),
2.34 -1H, br s, OH),1.91 -1H, ddd, Jˆ12.9, 12.9, 3.7 Hz), 1.75
-3H, s), 1.12-3H, d, Jˆ7.1 Hz), 0.92-3H, d, Jˆ6.2Hz)Ðsee
Table 4 for full 1H and 13C NMR assignments.
1
3.8. Synthesis of labelled compounds 1a, 15a, 16a and 17a
13C-Labelled methyl iodide -13CH3I) -Aldrich 27,718-5, 99

L.-K. Sy et al. / Tetrahedron 57 )2001) 8481±8493
8493
atom %) was used in place of isotopically normal CH3I in
the Grignard reaction with 6 which forms 3a. Subsequent
epoxidation of 3a to 14a and acid hydrolysis of 14a yielding
1a, 15a, 16a and 17a was performed as described for the
isotopically-normal compounds in the preceding sections.
Compound 3a: physical properties as for 3, with the follow-
Chemistry Department of the University of Hong Kong
for providing a postdoctoral fellowship to Dr Jaclyn Sy.
Miss Crystal Cheung and Miss Jo Yip helped to isolate
arteannuin O from A. annua and Mr Gary Ho performed
some preliminary synthetic work. This project was funded
by a grant from the CRCG.
1
ing differences: H NMR -CDCl3): 5.64 -1H, dd, Jˆ5.6
-3JCH), 1.4 Hz), 1.69 -3H, d, Jˆ126 Hz -1JCH)); 13C NMR
-CDCl3): 142.2 -d, Jˆ43 Hz -1JCC)), 121.7 -d, Jˆ2Hz
-2JCC)), 83.3 -d, Jˆ4 Hz -3JCC)), 30.8 -d, Jˆ3 Hz -2JCC)),
23.4 -ca. 90 times normal intensity); HREIMS m/z -rel. int.)
235.1643 [M1, C1413C1H22O2 requires 235.1653] -10), 191
-95), 162-100). Compound 14a: physical properties as for
References
1. Liu, J.-M.; Ni, M.-Y.; Fan, J.-F.; Tu, Y.-Y.; Wu, Z.-H.; Wu,
Y.-L.; Chou, W.-S. Acta Chim. Sin. 1979, 37, 129±143.
2. Sy, L.-K.; Brown, G. D.; Haynes, R. Tetrahedron 1998, 54,
4345±4356.
1
14, with the following differences: H NMR -CDCl3): 1.32
-3H, d, Jˆ127 Hz -1JCH)); 13C NMR -CDCl3): 63.1 -d,
Jˆ45 Hz -1JCC)), 27.3 -d, Jˆ4 Hz -2JCC)), 24.4 -ca. 95
times normal intensity); HREIMS m/z -rel. int.) 251.1602
[M1, C1413C1H22O3 requires 251.1602] -1), 223 -10), 179
-100), 167 -35). Compound 1a: physical properties as for
3. Sy, L.-K.; Brown, G. D. Phytochemistry 1998, 48, 1207±
1211.
4. Brown, G. D. Phytochemistry 1994, 36, 637±641.
5. Brown, G. D. J. Nat. Prod. 1992, 55, 1756±1760.
6. Imakura, Y.; Hachiya, K.; Ikemoto, T.; Yamashita, S.; Kihara,
M.; Kobayashi, S.; Shingu, T.; Milhous, W. K.; Lee, K.-H.
Heterocycles 1990, 31, 1011±1016.
1
1, with the following differences: H NMR -CDCl3): 3.53
-1H, d, Jˆ4 Hz -3JCH)), 3.37 -1H, d, Jˆ6.8 Hz, 4-OH
-3JCH)), 1.22 -3H, d, Jˆ126 Hz -1JCH)); 13C NMR
-CDCl3): 72.2 -d, Jˆ42Hz - 1JCC)), 26.7 -ca. 90 times
normal intensity); 20.2 -d, Jˆ3 Hz -2JCC)). Compound
15a: physical properties as for 15, with the following
7. Li, Y.; Yu, P.-L.; Chen, Y.-X.; Zhang, J.-L.; Wu, Y.-L. Kexue
Tongbao 1986, 31, 1038±1041.
8. Hui, S.-M.; Ngo, K.-S.; Brown, G. D. J. Chem. Soc., Perkin
Trans. 1 1997, 3435±3442.
1
differences: H NMR -CDCl3): 5.67 -1H, dq, Jˆ5.7 -3JCH),
9. Mi, J.-F.; Shen, C.-Y. Yaoxue Xuebao 1984, 19, 184±189.
10. Li, Y.; Chen, Y.-X.; Yu, P.-L.; Ji, R.-Y. Acta Pharm. Sin.
1986, 21, 899±905.
1.7 Hz), 1.78 -3H, dd, Jˆ126 -1JCH), 1.7 Hz); 13C NMR
-CDCl3): 131.2-d, Jˆ45 Hz -1JCC)), 126.4 -d, Jˆ2Hz
-2JCC)), 70.1 -d, Jˆ3 Hz -2JCC)), 27.2 -d, Jˆ4 Hz -3JCC)),
21.2 -ca. 75 times normal intensity); HREIMS m/z -rel. int.)
251.1606 [M1, C1413C1H22O3 requires 251.1602] -2), 167
-100). Compound 16a: physical properties as for 16, with
11. Ngo, K.-S.; Brown, G. D. Tetrahedron 1999, 55, 15099±
15108.
12. Xu, X.-X.; Zhu, J.; Huang, D.-Z.; Zhou, W.-S. Tetrahedron
1986, 42, 819±828.
1
13. Li, L-Q.; Zheng, Y.-P.; Gai, Y.-Z. Kexue Tongbao 1984, 29,
396±398.
the following differences: H NMR -CDCl3): 4.93 -1H, dd,
Jˆ157 -1JCH), 1.8 Hz), 4.90 -1H, dd, Jˆ156 -1JCH), 1.5 Hz);
13C NMR -CDCl3): 114.6 -ca. 90 times normal intensity).
Compound 17a: physical properties as for 17, with the
following differences: 1H NMR -CDCl3): 1.62-3H, d,
Jˆ128 Hz -1JCH)); 13C NMR -CDCl3): 31.4 -ca. 95 times
normal intensity).
14. Zhou, W.-S. Pure Appl. Chem. 1986, 58, 817±824.
15. Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34,
2543±2549.
16. Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512±
519.
17. Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am.
Chem. Soc. 1991, 113, 4092±4096.
18. Minato, M.; Yamamoto, K.; Tsuji, J. J. Org. Chem. 1990, 55,
766±768.
Acknowledgements
We thank the Generic Drug Research Programme of the