Rearrangement of germacranolides. Synthesis and absolute configuration of elemane and heliangolane derivatives from cnicin

Article abstract of DOI:10.1002/ejoc.200300161

A study of the Cope rearrangement of 15-oxo-germacranolides to 15-oxo-elemanolides has been carried out. The synthesis of two natural elemanolides, isolated from Centaurea paui, and an efficent isomerization of the C-4/C-5 double bond of 15-oxo-germacranolides to form heliangolides are reported. The absolute configuration of all the compounds has been ascertained. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejoc.200300161

FULL PAPER
Rearrangement of Germacranolides. Synthesis and Absolute Configuration of
Elemane and Heliangolane Derivatives from Cnicin
Sergio Rosselli,[a] Antonella Maggio,[a] Rosa Angela Raccuglia,[a] and Maurizio Bruno*[a]
Keywords: Synthetic methods / Sesquiterpenes / Terpenoids / Natural products / Rearrangement
A study of the Cope rearrangement of 15-oxo-germacranol- ported. The absolute configuration of all the compounds has
ides to 15-oxo-elemanolides has been carried out. The syn- been ascertained.
thesis of two natural elemanolides, isolated from Centaurea
paui, and an efficent isomerization of the C-4/C-5 double
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
bond of 15-oxo-germacranolides to form heliangolides are re- Germany, 2003)
Introduction
A remarkable reaction of the germacrane sesquiterpenes
is the [3,3] sigmatropic Cope rearrangement in which cyclo-
decadienes are thermally converted into 1,2-divinylcyclo-
hexane. The existence of an equilibrium between (E,E)-ger-
macradiene and trans-elemene sesquiterpenes has been
widely reported and its position is influenced by several fac-
tors such as substitution pattern, ring strain, conformation
and conjugation effects,[1] although it generally lies toward
the elemane isomer. The conditions for thermal conversion
of germacranes into elemanes are generally quite severe
since sealed tubes and temperatures between 150Ϫ210 °C
are required.[2] The use of catalytic amounts of PdII has
been reported[3] to enhance reaction rates.
Much lower temperatures and quantitative conversions,
probably due to a better conjugation of the aldehydic
groups, have been observed for the Cope reaction of 15-
oxo-trans,trans-germacra-1(10),4-dienolides with respect to
the corresponding alcohols or C-15 unfunctionalised de-
rivatives, although experimental conditions vary consider-
ably. Compound 1 (Figure 1) rearranges to the correspond-
ing elemane at 60 °C, whereas 2 rearranges to 3 at 100 °C[4]
and 4 gives 5 only at 125 °C.[5] Finally elemane 6 has been
obtained directly by oxidation of germacrane 7.[4] The dif-
ferent behaviour of these substrates has been ascribed to
several unpredictable stereoelectronic effects of the substitu-
ents in the C-6/C-9 moiety.[4]
Figure 1. Structure of compounds 1Ϫ7
monly found in this genus, whose antibacterial, cytotoxic
and antifungal properties have been investigated.[6Ϫ8] This
provided us with enough starting material to plan a study
of the germacrane-elemane conversion with the aim of syn-
thesising natural compounds with potential antitumour ac-
tivity.
In order to avoid unwanted oxidation of the two hydroxyl
groups of the side chain of cnicin, we decided to protect
them by the formation of an acetonide 9 (Scheme 1). Com-
pound 9, dissolved in CHCl3, was oxidized by treatment
with 10 equivalents of MnO2 at Ϫ78 °C to give, after 3
hours, the germacrane aldehyde 10 as the sole product. The
same reaction performed at 20 °C for 24 h gave an anal-
ogous result.
Results and Discussion
We have been able to isolate from some species of Centau-
rea large amounts of cnicin (8), a germacranolide com-
A solution of compound 10 in CDCl3 was allowed to
stand at 20 °C. A slow transformation took place and the
formation of a slightly less polar product was observed.
After 20 days no more starting material was detected. The
1H NMR spectrum of the new derivative showed signals of
an aldehydic proton at δ ϭ 9.48 ppm, a monosubstituted
[a]
`
Dipartimento di Chimica Organica, Universita di Palermo,
Viale delle Scienze, 90128 Palermo,Italy
Fax: (internat.) ϩ39Ϫ091596825
E-mail: bruno@dicpm.unipa.it
2690
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200300161
Eur. J. Org. Chem. 2003, 2690Ϫ2694

Rearrangement of Germacranolides
FULL PAPER
Table 1. Percentage of elemanolide aldehydes 11 and 17 formed at
20 and 60 °C
Time
(h)
11
17
20 °C
60 °C
20 °C
60 °C
0.5
1
2
3
5
10
24
72
96
168
240
312
408
480
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
16%
35%
44%
70%
85%
89%
98%
100%
30%
49%
62%
71%
85%
100%
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
20%
49%
85%
98%
100%
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
40%
67%
76%
91%
97%
100%
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
protons of C-13 appear as two doublets at δ ϭ 6.40 and
5.76 ppm with very small coupling constants (1.5 Hz and
1.4 Hz, respectively), not consistent with a germacrane or
elemane skeleton. Selective decoupling allowed us to ident-
ify H-7 at δ ϭ 3.08 ppm as a broad ddd. Irradiation of H-
7 collapsed the signal of H-8 (ddd at δ ϭ 5.11 ppm) into a
dd and sharpened the broad doublet at δ ϭ 4.94 ppm that
was so attributed to H-6, as subsequently confirmed by ir-
radiation of H-5 (d at δ ϭ 6.24 ppm).
Scheme 1. Reagents and conditions: (i) 2-methoxypropene, PTSA,
THF, room temp., 2 h, 96%; (ii) MnO2, CHCl3, room temp.,
30 min, 100%; (iii) CHCl3, HCl/CHCl3(sat.), room temp., 2 h,
100%; (iv) heating to 60 °C, 12 h, 100%; (v) PTSA, THF, room
temp., 24 h; (vi) Na2CO3 (0.1 ), dioxane/H20 (2:1), 16 h, 72%
The small value of the long-range coupling constants H-
7/H-13a and H-7/H-13b and of the coupling constant H-6/
H-7 (Ͻ 0.5 Hz) allowed us to propose an isomerization of
double bond and, besides clear signals of the γ-lactone and the C-5/C-6 double bond. Therefore, we assigned to this
the ester moiety, another olefinic methylene group. Conse- compound the structure 13, with a heliangolide skeleton,
quently we assigned to this compound the structure 11, confirmed also by the 13C NMR spectrum (Table 2) and the
with an elemane skeleton. Removal of the protecting group literature data.[10] The acetonide group of compound 13
afforded 12, whose physical and spectroscopic data were was then removed to give 14.
identical to those reported for a natural elemane isolated
To the best of our knowledge, compound 14 has not been
described previously, although similar heliangolides have
from Centaurea paui.[9]
The rate of the Cope rearrangement was enhanced by the been isolated from Centaurea tweediei[10] and Centaurea
use of silica gel. In fact, when SiO2 was added to a chloro- paui.[11] The former has a different ester moiety, and the
form solution of the germacrane 10 at 20 °C, the conversion latter has a methyleneacetoxy group at C-4 instead of the al-
into elemanolide 11 was complete in three days. A similar dehyde.
effect was obtained by increasing the temperature. Table 1
reports the percentage of rearranged products at different for 72 hours we detected the presence of a 1:1 mixture of
temperatures. elemanolide and heliangolide, showing that the ratio of the
When the same oxidative reaction was allowed to stand
In order to study the reactivity of cnicin (8) with a differ- latter had not increased. On the other hand, when we
ent oxidizing agent, three equivalents of PDC were added stopped the reaction after the disappearance of the starting
to a solution of acetonide 9 in CH2Cl2. After 18 hours, material (2 hours) we observed just the germacranolide 10
quite surprisingly we detected the presence of three com- and the heliangolide 13, in a 9:1 ratio. This shows that com-
pounds in the ratio 1:4:5. The less polar (10%) and the se- pound 13 is formed from aldehyde 10. In order to investi-
cond one (40%) were the previously described elemane 11 gate a probable mechanism, two different solutions of the
and germacrane 10, respectively. The third one (50%) was germacranolide 10 were mixed with pyridinium dichromate
an isomer of the former as shown by its mass spectrum.
(PDC) and CrCl3 respectively. In both cases no isomeriz-
1
The H NMR spectrum of this compound revealed the ation was observed. Consequently, we hypothesise that such
presence of an aldehydic proton at δ ϭ 9.45 ppm, similar an isomerization could arise due to the formation of
to that of the elemane (11), showing a better conjugation H2CrO3 during the oxidation, as reported in the litera-
of the α,β-unsaturated system with respect to the aldehydic ture.[12] The influence of an acid catalysis was confirmed by
group of the germacrane (δ ϭ 10.01 ppm). The exocyclic adding one drop of CHCl3 saturated with HCl to a solution
Eur. J. Org. Chem. 2003, 2690Ϫ2694
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2691

S. Rosselli, A. Maggio, R. A. Raccuglia, M. Bruno
FULL PAPER
Table 2. 13C NMR spectroscopic data for compounds 9Ϫ11, 13,
16, 18 in CDCl3
the literature. The second and less polar product was not
the expected germacrane 16 but the elemane 17 formed dur-
ing the chromatographic purification. In this case the heli-
angolide 18 was also synthesised in very good yield by acid
treatment of the germacranolide 16.
Finally, since the absolute stereochemistry of cnicin (8)
has been determined by X-ray analysis,[13] we were able to
ascertain the absolute configuration of the natural com-
pounds 12 and 17 as well as the heliangolides 14 and 18.
The very easy transformation of 15-oxo-germacranolides
into 15-oxo-elemanolides and 15-oxo-heliangolides, points
out an interesting problem: it is possible to argue that such
metabolites isolated as natural products (i.e. Centaurea
paui,[9,11] Centaurea tweediei[10]) could be artifacts generated
from unstable 15-oxo-germacranolides during extraction
and/or purification procedures. At this point there is no
clear answer to this and more investigation is required.
C
9
9[a]
10
11
13
16
18
1
2
3
4
5
6
7
8
9
129.8 d 131.1 d 130.1 d 145.0 d 128.7 d 129.6 d 128.8 d
26.3 t 27.5 t 26.3 t 113.1 t 24.6 t 26.5 t
34.6 t 35.6 t 30.3 t 137.7 t 22.7 t 30.6 t
24.4 t
22.4 t
143.9 s 145.8 s 141.5 s 144.6 s 142.4 s 141.7 s 142.1 s
128.5 d 129.7 d 144.6 d 46.1 d 146.8 d 144.1 d 147.0 d
76.5 d 78.0 d 74.1 d 77.3 d 75.6 d 73.4 d 75.8 d
53.0 d 54.2 d 52.4 d 52.3 d 49.9 d 54.3 d 51.6 d
73.1 d 75.5 d 72.5 d 69.5 d 69.6 d 70.9 d 67.1 d
48.6 t 49.7 t 48.2 t 44.7 t 46.3 t 52.3 t
49.0 t
10 132.2 s 133.5 s 133.9 s 42.0 s 131.8 s 133.2 s 132.6 s
11 135.4 s 138.2 s 133.9 s 136.4 s 133.5 s 135.4 s 133.8 s
12 169.6 s 170.7 s 168.7 s 168.9 s 168.5 s 168.7 s 169.1 s
13 124.9 t 124.5 t 126.3 t 120.2 t 128.2 t 127.9 t 127.7 t
14 16.8 q 17.6 q 17.4 q 17.9 q 17.0 q 17.5 q 17.0 q
15 61.4 t 61.6 t 188.7 d 193.5 d 193.8 d 189.1 d 193.8 d
1Ј 164.5 s 166.0 s 164.5 s 164.7 s 164.4 s
2Ј 139.4 s 141.7 s 139.3 s 138.9 s 139.3 s
3Ј 125.1 t 125.6 t 125.2 t 125.7 t 124.5 t
2ЈЈ 109.5 s 110.7 s 109.6 s 109.7 s 109.7 s
4ЈЈ 74.0 d 74.8 d 73.9 d 73.9 d 74.0 d
Experimental Section
5ЈЈ 70.2 t 71.5 t 70.2 t 70.1 t 70.2 t
6ЈЈ 26.3 q 27.3 q 26.3 q 26.3 q 26.3 q
7ЈЈ 25.4 q 26.4 q 25.4 q 25.5 q 25.4 q
General: IR spectra were obtained on a Shimadzu FTIR 8300 spec-
trometer. H NMR spectra were recorded in CDCl3 solution on a
1
Bruker AC 250E instrument at 250 MHz, and chemical shifts are
reported with respect to residual CHCl3 solvent signal (δ ϭ
7.27 ppm). 13C NMR spectra were recorded in CDCl3 solution on
the same apparatus at 62.7 MHz, and chemical shifts are reported
with respect to the solvent signals (δC ϭ 77.00 ppm). 13C NMR
assignments were determined from the DEPT spectra. Optical ro-
tations were measured on a PerkinϪElmer 141 polarimeter. MS
were recorded on a Finnigan TSQ70 instrument (PCI, isobutane,
[a]
In [D6]acetone.
of compound 10 in CHCl3. After three hours a complete
isomerization of the C-4/C-5 double bond was observed
and the heliangolide 13 was recovered with a 100% yield.
In order to investigate the influence of the C-8 side chain
on the germacrane-elemane equilibrium we decided to per-
form these reactions on salonitenolide (15), a germacranol-
ide often occurring in species of Centaurea genus.
direct inlet). Elemental analysis was carried out with
a
PerkinϪElmer 240 apparatus. Merck Si gel (70Ϫ230 mesh), deacti-
vated with 15% H2O (w/w), was used for column chromatography.
A careful saponification of the side chain of cnicin (8)
yielded salonitenolide (15), which was oxidized with 10
equivalents of MnO2 at 20 °C to give the germacrane alde-
hyde 16 as the only product.
A Cope rearrangement of 16, performed in CDCl3 solu-
tion at 20 °C, afforded the elemane 17. The different con-
ditions of this conversion are reported in Table 1. The spec-
troscopic and physical data of compound 17 are in perfect
agreement with those reported for a natural elemane iso-
lated from Centaurea paui.[11]
Isolation of (6R,7R,8S,1ЈЈR)-8-[(1ЈЈ,2ЈЈ-Dihydroxyethyl)acryloyl]-
15-hydroxygermacra-1(10),4,11(13)-trien-6,12-olide (cnicin) (8):
Cnicin (8) was isolated from Centaurea sphaerocephala and Centau-
rea napifolia both collected in June 2001 at Lascari, 60 km east of
Palermo, Italy following the purification procedures reported pre-
viously.[14,15]
(6R,7R,8S,4ЈЈR)-8-[2ЈЈ,2ЈЈ-Dimethyl-1ЈЈ,3ЈЈ-dioxacyclopentyl)-
acryloyl]-15-hydroxygermacra-1(10),4,11(13)-trien-6,12-olide (9): 2-
Methoxypropene (0.60 mL, 5.24 mmol) and p-toluenesulfonic acid
(40 mg, 0.20 mmol) were added at room temp. to a solution of com-
pound 8 (500 mg, 1.31 mmol) in 30 mL of THF. The mixture was
stirred for 2 h, 60 mL of CH2Cl2 was added, and the mixture
washed in turn with saturated aqueous NaHCO3 and H2O. The
organic layer was separated, dried (Na2SO4), concentrated and
In this case oxidation of salonitenolide (15) with three
equivalents of PDC in CH2Cl2 also gave a 1:4:5 mixture of
1
three compounds. The H NMR spectrum of the mixture
allowed us to identify the first two products as the elemane
17 and the germacrane 16, respectively. Silica gel column chromatographed on a column (silica gel not deactivated; petro-
leum ether/ethyl acetate, 3:1) to provide compound 9 (528 mg, 96%)
as an amorphous solid. [α]2D5 ϭ ϩ59.3 (c ϭ 0.76, CHCl3). IR (film):
ν˜max ϭ 3423, 2986, 2933, 2866, 1763, 1709, 1655, 1601, 1510, 1448,
chromatography of the mixture gave just two clean com-
pounds. The more polar one, also present in the original
mixture, showed an 1H NMR spectrum quite similar to that
of heliangolide 13, the only remarkable differences being
the upfield shift of the signals of H-8 and H-9b (ddd at δ ϭ
3.74 ppm and dd at δ ϭ 2.19 ppm, respectively) and the
downfield shift of the signals of H-13a and H-13b (br. s at
δ ϭ 6.56 ppm and br. s at δ ϭ 5.97 ppm, respectively), due
to the absence of the ester chain. Consequently we assigned
1381, 1261, 1149, 1064, 955 cmϪ1 1H NMR (CDCl3, 250 MHz):
.
δ ϭ 5.02 (br. dd, J ϭ 11.4, 4.6 Hz, 1 H, 1-H), 2.27 (m, 1 H, 2-HA),
2.20 (m, 1 H, 2-HB), 2.60 (ddd, J ϭ 11.6, 5.0, 2.6 Hz, 1 H, 3-HA),
2.00 (ddd, J ϭ 11.6, 11.6, 5.4 Hz, 1 H, 3-HB), 4.84 (d, J ϭ 9.8 Hz,
1 H, 5-H), 5.11 (dd, J ϭ 9.8, 8.2 Hz, 1 H, 6-H), 3.09 (m, 1 H, 7-
H), 5.12 (m, 1 H, 8-H), 2.61 (dd, J ϭ 12.2, 1.8 Hz, 1 H, 9-HA),
2.47 (dd, J ϭ 12.2, 11.8 Hz, 1 H, 9-HB), 6.29 (d, J ϭ 3.4 Hz, 1 H,
to this product the structure 18, not previously reported in 13-HA), 5.75 (d, J ϭ 2.8 Hz, 1 H, 13-HB), 1.51 (s, 3 H, 14-H3), 4.31
2692  2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org
Eur. J. Org. Chem. 2003, 2690Ϫ2694

Rearrangement of Germacranolides
FULL PAPER
(d, J ϭ 14.0 Hz, 1 H, 15-HA), 4.10 (d, J ϭ 14.0 Hz, 1 H, 15-HB), NMR (CDCl3, 62.7 MHz), see Table 2. PCI MS: m/z (%) ϭ 417
6.31 (br. s, 1 H, 3Ј-HA), 6.17 (br. s, 1 H, 3Ј-HB), 4.82 (br. t, J ϭ
(100) [M ϩ H]ϩ, 245 (27). C23H28O7 (416.45): calcd. C 66.33, H
6.9 Hz, 1 H, 4ЈЈ-H), 4.39 (dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-HA), 3.64 6.78; found C 66.27, H 6.81.
(dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-HB), 1.48 (s, 3 H, 6ЈЈ-H3), 1.44 (s, 3
(5R,6R,7R,8S,10S,1ЈЈR)-8-[(1ЈЈ,2ЈЈ-Dihydroxyethyl)acryloyl]-15-
1
H, 7ЈЈ-H3) ppm. H NMR ([D6]acetone, 250 MHz): δ ϭ 5.11 (m,
oxoelema-1,3,11(13)-trien-6,12-olide (12): Compound 11 (50 mg,
0.12 mmol) was dissolved in THF (5 mL), and p-toluenesulfonic
acid (34 mg, 0.18 mmol) was added at room temp., whilst stirring.
After 24 h, the mixture was concentrated and subjected to column
chromatography (silica gel; petroleum ether/ethyl acetate, 7:3) to
give 40 mg (88%) of compound 12. The physical and spectroscopic
data were identical to those reported in literature.[9]
1 H, 1-H), 2.33 (dddd, J ϭ 12.5, 12.5, 11.5, 5.1 Hz, 1 H, 2-HA),
2.20 (m, 1 H, 2-HB), 2.61 (ddd, J ϭ 11.6, 5.0, 2.6 Hz, 1 H, 3-HA),
1.98 (ddd, J ϭ 11.6, 11.6, 5.4 Hz, 1 H, 3-HB), 4.93 (d, J ϭ 9.8 Hz,
1 H, 5-H), 5.25 (dd, J ϭ 9.8, 8.2 Hz, 1 H, 6-H), 3.29 (m, 1 H, 7-
H), 5.12 (m, 1 H, 8-H), 2.60 (m, 2 H, 9-HA, 9-HB), 6.11 (d, J ϭ
3.4 Hz, 1 H, 13-HA), 5.72 (d, J ϭ 2.8 Hz, 1 H, 13-HB), 1.55 (s, 3
H, 14-H3), 4.30 (d, J ϭ 14.0 Hz, 1 H, 15-HA), 4.10 (d, J ϭ 14.0 Hz,
1 H, 15-HB), 6.32 (br. s, 1 H, 3Ј-HA), 6.11 (br. s, 1 H, 3Ј-HB), 4.77
(br. t, J ϭ 6.9 Hz, 1 H, 4ЈЈ-H), 4.33 (dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-
HA), 3.60 (dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-HB), 1.41 (s, 3 H, 6ЈЈ-H3),
1.37 (s, 3 H, 7ЈЈ-H3) ppm. 13C NMR (CDCl3, 62.7 MHz), see
Table 2. 13C NMR ([D6]acetone, 62.7 MHz), see Table 2. PCI MS:
m/z (%) ϭ 419 (100) [M ϩ H]ϩ, 247 (20). C23H30O7 (418.47): calcd.
C 66.01, H 7.23; found C 66.13, H 7.19.
(6R,7R,8S,4ЈЈR)-8-[(2ЈЈ,2ЈЈ-Dimethyl-1ЈЈ,3ЈЈ-dioxacyclopentyl)-
acryloyl]-15-oxohelianga-1(10),4,11(13)-trien-6,12-olide (13): Com-
pound 10 (50 mg, 0.12 mmol) was dissolved in CHCl3 (5 mL), and
0.1 mL of CHCl3 saturated with gaseous HCl was added at room
temp., whilst stirring. After 2 h, the solvents were evaporated in
vacuo to give the pure compound 13 (50 mg, 100%). Amorphous
solid. [α]2D5 ϭ Ϫ94.4 (c ϭ 0.12, CHCl3). IR (film): ν˜max ϭ 3371,
2986, 2935, 2870, 1772, 1711, 1689, 1637, 1450, 1371, 1296, 1269,
1157, 1065, 982, 959 cmϪ1. 1H NMR (CDCl3, 250 MHz): δ ϭ 5.28
(br. t, J ϭ 8.0 Hz, 1 H, 1-H), 2.50 (m, 1 H, 2-HA), 1.98 (m, 1 H,
2-HB), 2.77 (ddd, J ϭ 12.2, 3.4, 3.4 Hz, 1 H, 3-HA), 2.17 (ddd, J ϭ
12.2, 12.2, 4.4 Hz, 1 H, 3-HB), 6.24 (br. d, J ϭ 10.1 Hz, 1 H, 5-H),
4.94 (br. d, J ϭ 10.1 Hz, 1 H, 6-H), 3.08 (br. ddd, J ϭ 10.0, 1.5,
1.4 Hz, 1 H, 7-H), 5.11 (ddd, J ϭ 10.7, 10.0, 3.7 Hz, 1 H, 8-H),
2.64 (dd, J ϭ 12.0, 3.7 Hz, 1 H, 9-HA), 2.36 (dd, J ϭ 12.0, 10.7 Hz,
1 H, 9-HB), 6.40 (d, J ϭ 1.5 Hz, 1 H, 13-HA), 5.76 (d, J ϭ 1.4 Hz,
1 H, 13-HB), 1.87 (br. s, 3 H, 14-H3), 9.45 (br. s, 1 H, H-15), 6.29
(br. s, 1 H, 3Ј-HA), 6.12 (br. s, 1 H, 3Ј-HB), 4.79 (br. t, J ϭ 6.9 Hz,
1 H, 4ЈЈ-H), 4.37 (dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-HA), 3.61 (dd, J ϭ
8.3, 6.9 Hz, 1 H, 5ЈЈ-HB), 1.47 (s, 3 H, 6ЈЈ-H3), 1.45 (s, 3 H, 7ЈЈ-
H3) ppm. 13C NMR (CDCl3, 62.7 MHz), see Table 2. PCI MS: m/
z (%) ϭ 417 (100) [M ϩ H]ϩ, 245 (45), 199 (28), 171 (10). C23H28O7
(416.45): calcd. C 66.33, H 6.78; found C 66.40, H 6.84.
(6R,7R,8S,4ЈЈR)-8-[(2ЈЈ,2ЈЈ-Dimethyl-1ЈЈ,3ЈЈ-dioxacyclopentyl)-
acryloyl]-15-oxogermacra-1(10),4,11(13)-trien-6,12-olide (10): Acti-
vated manganese oxide (MnO2, 104 mg, 1.2 mmol) was added with
stirring to a solution of compound 9 (50 mg, 0.12 mmol) in 5 mL
of CHCl3. The mixture was stirred for 30 min at room temp. and,
after filtration through a Celite pad, concentrated to give an
amorphous solid (50 mg, 100%). [α]2D5 ϭ ϩ41.4 (c ϭ 0.84, CHCl3).
IR (film): ν˜max ϭ 2982, 2935, 2860, 1760, 1711, 1675, 1650, 1600,
1524, 1439, 1381, 1261, 1152, 1080, 955, 931 cmϪ1 1H NMR
.
(CDCl3, 250 MHz): δ ϭ 5.10 (br. dd, J ϭ 11.4, 5.0 Hz, 1 H, 1-H),
2.22 (m, 1 H, 2-HA), 2.18 (m, 1 H, 2-HB), 3.04 (ddd, J ϭ 11.7, 2.9,
1.8 Hz, 1 H, 3-HA), 1.82 (m, 1 H, 3-HB), 5.89 (d, J ϭ 10.3 Hz, 1
H, 5-H), 5.36 (dd, J ϭ 10.3, 8.1 Hz, 1 H, 6-H), 3.15 (m, 1 H, 7-H),
5.15 (m, 1 H, 8-H), 2.62 (br. d, J ϭ 12.3 Hz, 1 H, 9-HA), 2.48 (dd,
J ϭ 12.3, 10.3 Hz, 1 H, 9-HB), 6.36 (d, J ϭ 3.4 Hz, 1 H, 13-HA),
5.82 (d, J ϭ 3.0 Hz, 1 H, 13-HB), 1.31 (s, 3 H, 14-H3), 10.01 (br. s,
1 H, 15-H), 6.32 (br. s, 1 H, 3Ј-HA), 6.19 (br. s, 1 H, 3Ј-HB), 4.81
(br. t, J ϭ 6.9 Hz, 1 H, 4ЈЈ-H), 4.39 (dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-
HA), 3.63 (dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-HB), 1.48 (s, 3 H, 6ЈЈ-H3),
1.44 (s, 3 H, 7ЈЈ-H3) ppm. 13C NMR (CDCl3, 62.7 MHz), see
Table 2. PCI MS: m/z (%) ϭ 417 (100) [M ϩ H]ϩ, 245 (18).
C23H28O7 (416.45): calcd. C 66.33, H 6.78; found C 66.42, 6.68.
Oxidation of Compound 9 with PDC: Pyridinium dichromate
(270 mg, 0.72 mmol) was added to a solution of compound 9
(100 mg, 0.24 mmol) in 10 mL of CH2Cl2. The mixture was stirred
for 18 h and, after filtration through Florisil, subjected to column
chromatography (silica gel; dichloromethane/methanol, 49:1) to
give 10 mg (10%), 40 mg (40%) and 50 mg (50%) of compounds 11,
10, 13 respectively.
(5R,6R,7R,8S,10S,4ЈЈR)-8-[(2ЈЈ,2ЈЈ-Dimethyl-1ЈЈ,3ЈЈ-dioxacyclo-
pentyl)acryloyl]-15-oxoelema-1,3,11(13)-trien-6,12-olide (11): Com-
pound 10 was allowed to stand in a solution of CHCl3 at 20 °C for
20 days. After this period a complete conversion of 10 into com-
pound 11 was observed. Similar results (Table 1) were obtained by
(6R,7R,8S,1ЈЈR)-8-[(1ЈЈ,2ЈЈ-Dihydroxyethyl)acryloyl]-15-oxo-
helianga-1(10),4,11(13)-trien-6,12-olide (14): Compound 13 (50 mg,
0.12 mmol) was dissolved in THF (5 mL), and p-toluenesulfonic
acid (34 mg, 0.18 mmol) was added at room temp., whilst stirring.
leaving a solution of compound 10 (50 mg in 2 mL of CHCl3) to After 24 h, the mixture was concentrated and subjected to column
stand at 60 °C. Compound 11 was isolated as an amorphous solid.
[α]2D5 ϭ ϩ24.2 (c ϭ 0.26, CHCl3). IR (film): ν˜max ϭ 3018, 2986,
2926, 2854, 1774, 1715, 1693, 1637, 1456, 1410, 1371, 1292, 1263,
chromatography (silica gel; petroleum ether/ethyl acetate, 7:3) to
give 43 mg (95%) of compound 14. Amorphous solid. [α]2D5
Ϫ57.4 (c ϭ 0.12, CHCl3). IR (film): ν˜max ϭ 3385, 2930, 2872, 1763,
1H
ϭ
1161, 1065, 974 cmϪ1. 1H NMR (CDCl3, 250 MHz): δ ϭ 5.67 (dd, 1718, 1707, 1686, 1269, 1445, 1119, 1072, 1016, 980 cmϪ1
.
J ϭ 17.4, 10.7 Hz, 1 H, 1-H), 4.96 (d, J ϭ 10.7 Hz, 1 H, 2-HA), NMR (CDCl3, 250 MHz): δ ϭ 5.28 (br. t, J ϭ 8.0 Hz, 1 H, 1-H),
4.87 (d, J ϭ 17.4 Hz, 1 H, 2-HB), 6.35 (br. s, 1 H, 3-HA), 6.29 (br.
s, 1 H, 3-HB), 3.20 (d, J ϭ 12.1 Hz, 1 H, 5-H), 4.43 (dd, J ϭ 12.1,
2.49 (m, 1 H, 2-HA), 2.00 (m, 1 H, 2-HB), 2.75 (ddd, J ϭ 12.2, 3.4,
3.4 Hz, 1 H, 3-HA), 2.17 (ddd, J ϭ 12.2, 12.2, 4.4 Hz, 1 H, 3-HB),
10.8 Hz, 1 H, 6-H), 3.00 (dddd, J ϭ 10.8, 10.8, 2.8, 2.8 Hz, 1 H, 7- 6.24 (br. d, J ϭ 10.2 Hz, 1 H, 5-H), 4.94 (br. d, J ϭ 10.2 Hz, 1 H,
H), 5.31 (ddd, J ϭ 10.8, 10.6, 4.2 Hz, 1 H, 8-H), 2.07 (dd, J ϭ 6-H), 3.09 (br. ddd, J ϭ 10.0, 1.5, 1.4 Hz, 1 H, 7-H), 5.11 (ddd,
13.1, 4.2 Hz, 1 H, 9-HA), 1.73 (dd, J ϭ 13.1, 10.6 Hz, 1 H, 9-HB),
6.17 (d, J ϭ 2.8 Hz, 1 H, 13-HA), 5.54 (d, J ϭ 2.8 Hz, 1 H, 13-
J ϭ 10.7, 10.0, 3.7 Hz, 1 H, 8-H), 2.64 (dd, J ϭ 12.0, 3.7 Hz, 1 H,
9-HA), 2.35 (dd, J ϭ 12.0, 10.7 Hz, 1 H, 9-HB), 6.40 (d, J ϭ 1.5 Hz,
HB), 1.07 (s, 3 H, 14-H3), 9.48 (s, 1 H, 15-H), 6.31 (br. s, 1 H, 3Ј- 1 H, 13-HA), 5.77 (d, J ϭ 1.4 Hz, 1 H, 13-HB), 1.87 (br. s, 3 H, 14-
HA), 6.17 (br. s, 1 H, 3Ј-HB), 4.88 (br. t, J ϭ 6.9 Hz, 1 H, 4ЈЈ-H), H3), 9.45 (br. s, 1 H, 15-H), 6.33 (br. s, 1 H, 3Ј-HA), 6.02 (br. s, 1
4.38 (dd, J ϭ 8.3, 6.9 Hz, 1 H, 5ЈЈ-HA), 3.63 (dd, J ϭ 8.3, 6.9 Hz, H, 3Ј-HB), 4.55 (dd, J ϭ 6.9, 3.7 Hz, 1 H, 1ЈЈ-H), 3.80 (dd, J ϭ
1 H, 5ЈЈ-HB), 1.48 (s, 3 H, 6ЈЈ-H3), 1.45 (s, 3 H, 7ЈЈ-H3) ppm. 13C 11.0, 3.7 Hz, 1 H, 2ЈЈ-HA), 3.58 (dd, J ϭ 11.0, 6.9 Hz, 1 H, 2ЈЈ-HB)
Eur. J. Org. Chem. 2003, 2690Ϫ2694
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2693

S. Rosselli, A. Maggio, R. A. Raccuglia, M. Bruno
ppm. PCI MS: m/z (%) ϭ 377 (100) [M ϩ H]ϩ, 245 (46), 199 (40), J ϭ 10.8, 10.0, 3.3 Hz, 1 H, 8-H), 2.70 (dd, J ϭ 12.0, 3.3 Hz, 1 H,
FULL PAPER
171 (29). C20H24O7 (376.39): calcd. C 63.82, H 6.43; found C 63.67,
H 6.51.
9-HA), 2.19 (dd, J ϭ 12.0, 10.8 Hz, 1 H, 9-HB), 6.56 (br. s, 1 H,
13-HA), 5.97 (br. s, 1 H, 13-HB), 1.83 (br. s, 3 H, 14-H3), 9.44 (br.
s, 1 H, 15-H) ppm. 13C NMR (CDCl3, 62.7 MHz), see Table 2. PCI
MS: m/z (%) ϭ 263 (100) [M ϩ H]ϩ, 245 (30), 199 (23), 171 (18).
C15H18O4 (262.29): calcd. C 68.68, H 6.92; found C 68.52, H 6.99.
(6R,7R,8S)-8,15-Dihydroxygermacra-1(10),4,11(13)-trien-6,12-olide
(salonitenolide) (15): Cnicin (8; 700 mg, 1.85 mmol) was dissolved
in 15 mL of a mixture of dioxane and H2O (2:1) and an aqueous
solution of Na2CO3 (0.1 , 35 mL) was added dropwise at room
temp. over a period of 16 hours. The reaction mixture was diluted
with brine (30 mL) and extracted with CH2Cl2. The combined or-
ganic layers were dried (Na2SO4), concentrated and chromato-
graphed by column chromatography (silica gel; petroleum ether/
ethyl acetate, 1:1) to give salonitenolide (15) (350 mg, 72%). The
physical and spectroscopic data were identical to those reported in
the literature.[16]
Oxidation of Compound 15 with PDC: Pyridinium dichromate
(428 mg, 1.14 mmol) was added to a solution of compound 15
(100 mg, 0.38 mmol) in 10 mL of CH2Cl2. The mixture was stirred
1
for 6 h and then filtered through Florisil. The H NMR spectrum
of the mixture showed the signals of compounds 17, 16, 18 in a
1:4:5 ratio. Column chromatography (silica gel; dichloromethane/
methanol, 49:1) of the mixture gave 50 mg (50%), and 50 mg (50%)
of compounds 17 and 18, respectively.
(6R,7R,8S)-8-Hydroxy-15-oxogermacra-1(10),4,11(13)-trien-6,12-
olide (16): Activated manganese oxide (165 mg, 1.9 mmol) was ad-
Acknowledgments
This work was supported by Italian Government project MIUR
PRIN 2001 ‘‘Sostanze Naturali ed Analoghi Sintetici con Attivita
Antitumorale’’.
ded whilst stirring to
a solution of compound 15 (50 mg,
`
0.19 mmol) in 5 mL of CHCl3. The mixture was stirred for 24 hours
at room temp. and, after filtration through a Celite pad, concen-
trated to give compound 16 (50 mg, 100%). Amorphous solid. [α]
[1]
25 ϭ ϩ166.1 (c ϭ 0.90, CHCl3). IR (film): ν˜max ϭ 3421, 2930, 2866,
R. K. Hill, in Comprehensive Organic Synthesis, Vol 5 (Eds.: B.
D
1759, 1670, 1655, 1261, 1140, 1068, 1020, 970 cmϪ1 1H NMR
.
M. Trost, I. Fleming, L. A. Paquette), Pergamon Press, New
York, 1991, 785.
T. C. Jain, C. M. Banks, J. E. McCloskey, Tetrahedron Lett.
(CDCl3, 250 MHz): δ ϭ 5.06 (dd, J ϭ 11.4, 5.2 Hz, 1 H, 1-H), 2.22
(m, 2 H, 2-HA, 2-HB), 3.07 (ddd, J ϭ 11.7, 2.9, 1.8 Hz, 1 H, 3-HA),
1.95 (m, 1 H, 3-HB), 5.81 (d, J ϭ 10.1 Hz, 1 H, 5-H), 5.09 (dd, J ϭ
10.1, 8.4 Hz, 1 H, 6-H), 2.88 (dddd, J ϭ 10.0, 8.4, 3.4, 3.2 Hz, 1
H, 7-H), 4.12 (br. dd, J ϭ 10.9, 10.0 Hz, 1 H, 8-H), 2.71 (br. d,
J ϭ 12.7 Hz, 1 H, 9-HA), 2.41 (dd, J ϭ 12.7, 10.9 Hz, 1 H, 9-HB),
6.46 (br. d, J ϭ 3.4 Hz, 1 H, 13-HA), 6.44 (br. d, J ϭ 3.2 Hz, 1 H,
13-HB), 1.17 (br. s, 3 H, 14-H3), 9.94 (br. s, 1 H, 15-H) ppm. 13C
NMR (CDCl3, 62.7 MHz), see Table 2. PCI MS: m/z (%) ϭ 263
(100) [M ϩ H]ϩ, 245 (27). C15H18O4 (262.29): calcd. C 68.68, H
6.92; found C 68.59, H 6.83.
[2]
1970, 11, 841Ϫ844.
[3]
A. F. Barrero, J. E. Oltra, M. Alvarez, Tetrahedron Lett. 1998,
39, 1401Ϫ1404.
G. Appendino, H. C. Ozen, Gazz. Chim. Ital. 1993, 123, 93Ϫ94
and references cited therein.
[4]
[5]
A. Rustaiyan, L. Nazarians, F. Bohlmann, Phytochemistry
1979, 18, 879Ϫ880.
[6]
`
R. Vanhaelen-Fastre, M. Vanhaelen, Planta Med. 1976, 29,
179Ϫ189.
[7]
[8]
A. F. Barrero, J. E. Oltra, I. Rodriguez, A. Barragan, D. G.
Gravelos, P. Ruiz, Fitoterapia 1995, 66, 227Ϫ230.
A. G. Gonzalez, V. Darias, G. Alonso, J. N. Boada, M. Feria,
Planta Med. 1978, 33, 356Ϫ359.
L. Cardona, M. Casilo, F. I. Navarro, J. R. Pedro, Nat. Prod.
Lett. 1994, 5, 47Ϫ54.
(5R,6R,7R,8S,10S,)-8-Hydroxy-15-oxoelema-1,3,11(13)-trien-6,12-
olide (17): Compound 16 was allowed to stand in solution in CHCl3
at 20 °C for 13 days. After this period a complete conversion of 16
into compound 17 was observed. The physical and spectroscopic
data were identical to those reported in literature.[11]
[9]
[10]
[11]
[12]
[13]
A. M. Fortuna, E. C. de Riscala, C. A. N. Catalan, T. E.
Gedris, W. Herz, Biochem. Syst. Ecol. 2001, 29, 967Ϫ71.
L. Cardona, B. Garcia, M. C. Munoz, F. I. Navarro, J. R. Pe-
dro, Liebigs Ann./Recueil 1997, 527Ϫ32.
(6R,7R,8S)-8-Hydroxy-15-oxohelianga-1(10),4,11(13)-trien-6,12-
olide (18): Compound 16 (50 mg, 0.19 mmol) was dissolved in
CHCl3 (5 mL), and 0.1 mL of CHCl3 saturated with gaseous HCl
was added at room temp., whilst stirring. After 2 h, the solvents
were evaporated in vacuo to give the pure compound 18. Amorph-
ous solid. [α]2D0 ϭ Ϫ19.9 (c ϭ 0.80, CHCl3). IR (film): ν˜max ϭ 3450,
3018, 2933, 2868, 1763, 1686, 1655, 1448, 1317, 1120, 1094, 980
G. Piancatelli, A. Scettri, M. D’Auria, Synthesis 1982,
245Ϫ258.
S. M. Adekenov, V. Yu. Gatilov, I. Yu. Bagryanskaya, V. A.
Raldugin, Sib. Khim. Zh. 1993, 76Ϫ79 (Chem. Abstr. 1993,
119, 181139p).
M. Bruno, C. Fazio, S. Passannanti, M. P. Paternostro, J. G.
Diaz, W. Herz, Phytochemistry 1994, 35, 1371Ϫ1372.
M. Bruno, C. Fazio, M. P. Paternostro, J. G. Diaz, W. Herz,
Planta Med. 1995, 61, 374Ϫ375.
[14]
[15]
[16]
1
cmϪ1. H NMR (CDCl3, 250 MHz): δ ϭ 5.21 (br. t, J ϭ 8.0 Hz, 1
H, 1-H), 2.48 (m, 1 H, 2-HA), 1.98 (m, 1 H, 2-HB), 2.75 (ddd, J ϭ
12.2, 3.4, 3.4 Hz, 1 H, 3-HA), 2.12 (ddd, J ϭ 12.2, 12.2, 4.2 Hz, 1
H, 3-HB), 6.25 (br. d, J ϭ 10.2 Hz, 1 H, 5-H), 4.86 (br. d, J ϭ
10.2 Hz, 1 H, 6-H), 2.85 (br. d, J ϭ 10.0 Hz, 1 H, 7-H), 3.74 (ddd,
A. F. Barrero, J. F. Sanchez, I. Rodriguez, C. Soria Sanz, An.
Quim. 1988, 84, 344Ϫ347.
Received March 10, 2003
2694
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2690Ϫ2694