Revision of the absolute configuration of the tricyclic sesquiterpene (+)-kelsoene by chemical correlation and enantiospecific total synthesis of its enantiomer

Article abstract of DOI:10.1039/b101579f

The absolute configuration of the recently identified sesquiterpene (+)-kelsoene was revised by chemical correlation with (R)-(+)-pulegone; the correct structure is (1R,2S,5R,6R,7R,8S)-2,8-dimethyl-6-(1-methylethenyl) tricyclo[5.3.0.0^2,5]decane

Full text of DOI:10.1039/b101579f

View Article Online / Journal Homepage / Table of Contents for this issue
Revision of the absolute configuration of the tricyclic sesquiterpene
+)-kelsoene by chemical correlation and enantiospecific total synthesis
(
of its enantiomer
a
a
a
b
Sonja Fietz-Razavian, Stefan Schulz,* Ina Dix and Peter G. Jones
a
Institut für Organische Chemie, Technische Universität Braunschweig, Hagenring 30, D-38108,
Braunschweig, Germany. E-mail: stefan.schulz@tu-bs.de; Fax: +49 531 391 5272;
Tel: +49 531 391 7353
b
Institut f u¨ r Anorganische Chemie, Technische Universit a¨ t Braunschweig, Hagenring 30, D-38108
Braunschweig, Germany
Received (in Cambridge, UK) 8th March 2001, Accepted 4th September 2001
First published as an Advance Article on the web 2nd October 2001
The absolute configuration of the recently identified sesqui-
terpene (+)-kelsoene was revised by chemical correlation
with (R)-(+)-pulegone; the correct structure is
lactonization of pure cis-5 to 6, followed by elimination with a
10,11
bulky base, as has been described for anti-5.
The double
bond of the resulting acid 7 was now internally acylated via the
acid chloride to form the ketone 8, which was photochemically
transformed by a [2 + 2]-cycloaddition with ethylene into the
(
1R,2S,5R,6R,7R,8S)-2,8-dimethyl-6-(1-methylethenyl)tri-
2,5
cyclo[5.3.0.0 ]decane.
desired
ketone
(1S,2R,5R,7R,8R)-2,8-dimethyltricyclo-
[5.3.0.02 ]decan-6-one (ent-3). The ethylene attack occurs from
the less hindered side. In an identical manner, rac-3 was
synthesized starting from rac-4. To exclude the unlikely
possibility of inversion of the stereogenic centers in the course
,5
Recently the sesquiterpene hydrocarbon (+)-kelsoene (1, also
called tritomarene) was isolated from the marine sponge
1
Cymbastela hooperi as well as from the liverworts Ptychanthus
2
3
straitus, Calypogeia muelleriana, and Tritomaria quinque-
4
dentata. The relative configuration was elucidated during these
4
of our synthesis, ent-3 was reduced with LiAlH . Preferential
attack of the hydride from the less hindered side and reaction of
the alcohol with tosyl chloride resulted in formation of the
studies; 1 contains a rare cis,anti,cis 4-5-5 carbotricyclic ring
5
system, previously only found in the tetraterpene poduran and
6
the sesquiterpenoid sulcatine G. The analysis of NMR data
obtained from two dihydroisoxazoles formed after addition
of
(aS)-2A-methoxy-1,1A-binaphthalene-2-carbohydroximoyl
chloride (MBCC) to the double bond of 1 according to the
7
method of Fukui et al. led to the conclusion that (+)-1
possesses
a (1S,2R,5S,6S,7S,8R)-2,8-dimethyl-6-(1-methyl-
ethenyl)tricyclo[5.3.0.0 ]decane structure.8 In the present
study we will show by chemical correlation with the ketone 3
and further with (R)-(+)-pulegone (4) that this assignment is
incorrect. We will also present the first synthesis of enantiomer-
ically pure 1, which has previously only been synthesized in
racemic form.9
2,5
Natural 1 can be isomerized by a short treatment with
toluene-p-sulfonic acid into the more stable isomer 2, which in
turn is cleaved by ozonolysis to the tricyclic ketone 3. Under the
reaction conditions no excessive degradation of 3 or 1 or
formation of byproducts occurs. The ketone 3 was the target of
a short enantiospecific synthesis, because we reasoned that its
racemate may be well suited to separation by GC on chiral
phases due to its rigidity. It would also be formed by
degradation of other terpenes exhibiting the same ring system,
5
such as e.g. the tetraterpene poduran recently identified by us,
thus allowing its use for the determination of the absolute
configuration of several natural products.
The ketone 3 was synthesized starting from (R)-(+)-pulegone
(
4), which was transformed by bromination and Favorskii-
rearrangement into a mixture of cis- and trans-pulegonic acids
10
(
5), separable by column chromatography. Isomerization of
the double bond into the terminal position was achieved by
2 2
Scheme 2 Reagents and conditions: (a) Br , Et O, 0 °C; (b) 25% aq. KOH,
D, column separation; (c) aq. HCl, MeOH, D; (d) t-BuOK, DMF, 140 °C;
(
(
e) (COCl)
h) LiAlH
O; (m) MeLi, THF, 278 °C; (n) PDC, CH
MeOH, D; (p) Cp TiMe , THF, D.
2
, CH
2
Cl
2
; (f) AlCl
3
, CH
2
Cl
2
; (g) CH
2
NCH
SOI, DMSO; (l) H
Cl ; (o) NaOMe,
2 2 2
, hn, CH Cl , 0 °C;
9
,13
4
, Et O; (i) p-TsCl, py; (j) see ; (k) Me
2
3
2
,
Scheme 1 Reagents and conditions: (a) p-TSA, 15 min, CH
CH Cl , 278 °C, then Me S.
2
Cl
2
, D; (b) O
3
,
Pd/C, Et
2
2
2
2
2
2
2
2
2154
Chem. Commun., 2001, 2154–2155
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b101579f

View Article Online
tosylate 9. An X-ray analysis of 9† allowed the determination of
Notes and references
its absolute configuration.12 This tosylate could be thus
†
Crystal data for 9: orthorhombic, space group P2
1
3
2
1
2
1
, a = 9.1578(8), b
identified
as
(1S,2R,5R,6S,7R,8R)-2,8-dimethyltricyclo-
=
11.7038(10), c = 16.1527(14) Å, U = 1731.3 Å , Z = 4, T = 2130 °C.
Data collection: a crystal ca. 0.4 3 0.16 3 0.09 mm was used to record
6944 intensities on a Bruker SMART 1000 CCD diffractometer (Mo-Ka
radiation, 2qmax 57°). Structure refinement: the structure was refined
anisotropically on F (program SHELXL-97, G. M. Sheldrick, University of
,5
[
5.3.0.02 ]dec-6-yl tosylate, exhibiting the expected absolute
configuration.
2
Being confident on the absolute configuration of ent-3, we
were now able to perform GC analyses. The absolute configura-
tion of natural (+)-1 can be deduced by analysis of its
degradation product 3, because its relative configuration has
2
G o¨ ttingen) to wR2 0.083, R1 0.032 for 211 parameters and 4388 unique
reflections. The hydrogens were refined using a riding model or rigid
methyl groups. The Flack parameter refined to 20.07(5). CCDC 171200.
See http://www.rsc.org/suppdata/cc/b1/b101579f/ for crystallographic files
in .cif or other electronic format.
1
been elucidated during its isolation. Separation of rac-3 was
achieved on a chiral Hydrodex-6-TBDMS stationary phase.‡
The (2)-enantiomer of 3 eluted first, as did the degradation
product of synthetic (2)-kelsoene. The ketone 3 derived from
natural (+)-kelsoene eluted as second peak. In addition, the
synthetic (2)-1 prepared by us showed the expected sign of
optical rotation of [a] 2D 0 = 278.3. These data unequivocally
prove that natural (+)-kelsoene has the opposite configuration as
previously reported and possesses a (1R,2S,5R,6R,7R,8S)-
‡
Separations were performed on a 15 m Hydrodex-6-TBDMS capillary
column (Macherey & Nagel), id 0.25 mm, programmed from 60 to 180 with
5
§
min after a 2 min isothermal period with H
2
as the carrier gas (45 psi).
1
13
25
The H NMR, C NMR, and MS data are identical to those reported in the
1
literature.
[a] = +78.1 (c = 1.98, CHCl ).
3
[a]D = 278.3 (c = 0.31, CHCl ). Literature data for (+)-1:
3
2
5
1
D
8
¶ We also tried to find out why the NMR study by Nabeta et al. furnished
the opposite result. In their work Nabeta et al. presented a preferred
conformation for one of the two formed kelsoene-MBCC adducts.
Unfortunately, they did not explain how they obtained this conformation.
These adducts contain a single bond between the heterocyclic ring and the
,8-dimethyl-6-(1-methylethenyl)tricyclo[5.3.0.02 ]decane
structure.
We proceeded with the synthesis of enantiomerically pure
2)-1 to further underline the consistency of our data set. In
their synthesis of rac-1, Mehta and Srinivas showed that ketone
,5
2
(
7
cyclic core of 1, while in the original paper introducing this method all
examples contain a dihydroisoxazole ring directly linked to the parent cyclic
hydrocarbon, thus reducing the inherent flexibility. Finally, only one of two
formed diastereomers was investigated by NMR and no comparison with
the respective (aR)-MBCC was performed, as should be done in Mosher-
like methods. In essence, their determination of the absolute configuration
relies on several NOE signals observed, while in the present study a sound
chemical correlation is performed. It should be noted that recently a very
informative critique of the Mosher method has been published, which also
applies to related methods.15
3
is strongly hindered and not readily attacked by nucleophiles.
We therefore chose their strategy to reduce steric hindrance:
The bicyclic ketone 8 was transformed into the cyclobutene
derivative 10 as described.
9,13
The side chain was then
elaborated in a new way. Even after reduction of the steric
hindrance, only small nucleophiles can be used to attack the
carbonyl carbon atom. Nevertheless, epoxidation with trime-
thylsulfoxonium iodide was feasible with satisfactory yield.
The epoxide 11 was then hydrogenated and rearranged in one
step to form the aldehyde 12. Addition of methyllithium
followed by oxidation with PDC afforded the ketone 13 in a
1 G. M. König and A. D. Wright, J. Org. Chem., 1997, 62, 3837.
2 K. Nabeta, K. Yamamoto, M. Hashimoto, H. Koshino, K. Funatsuki and
K. Katoh, Chem. Commun., 1998, 1485.
3
U. Warmers, K. Wihstutz, N. Bülow, C. Fricke and W. A. König,
Phytochemistry, 1998, 48, 1723.
8
0+20 dr in favor of the unwanted diastereomer. Obviously the
less hindered hydrogen moves more readily in the epoxide–
aldehyde rearrangement of 12. Nevertheless, simple treatment
with base epimerized this center to the thermodynamically more
favored arrangement, which is also found in the natural product.
Kelsoene was then formed by final methenylation with
4
5
U. Warmers and W. A. König, Phytochemistry, 1999, 52, 1519.
S. Schulz, C. Messer and K. Dettner, Tetrahedron Lett., 1997, 53,
2
077.
A. Arnone, G. Nasini and O. V. de Pava, J. Chem. Soc., Perkin Trans.
, 1993, 2723.
6
1
14
Me
2
TiCp2, because Wittig olefination, despite reported by
7 H. Fukui, Y. Fukushi and S. Tahara, Tetrahedron Lett., 1999, 40,
325.
13
Mehta and Srinivas, did not proceed with satisfactory yields.
Thus enantiomerically pure (2)-1 has been synthesized,
confirming our previous assignments.§,¶
8 K. Nabeta, M. Yamamoto, H. Koshino, H. Fukui, Y. Fukushi and S.
Tahara, Biosci., Biotechnol., Biochem., 1999, 63, 1772.
9
G. Mehta and K. Srinivas, Tetrahedron Lett., 1999, 40, 4877.
We thank Professor Nabeta and Professor G. König for
kindly providing us with samples of (+)-kelsoene. We also
thank the Deutsche Forschungsgemeinschaft and the Fonds der
chemischen Industrie for financial support.
Note added in proof. Similar results were obtained recently
by G. Mehta and K. Srinivas, Tetrahedron Lett., 2001, 42, 2855,
using a different synthetic strategy.
1
1
1
1
1
0 J. Wolinsky, H. Wolf and T. Gibson, J. Org. Chem., 1963, 28, 274.
1 J. Wolinsky and E. J. Eustache, J. Org. Chem., 1972, 37, 3376.
2 P. Jones, I. Dix, S. Fietz-Razavian and S. Schulz, unpublished results.
3 G. Mehta and K. Srinivas, Synlett, 1999, 5, 555.
4 N. A. Petasis and E. I. Bzowej, J. Am. Chem. Soc., 1990, 112, 6392.
15 J. M. Seco, E. Quiñoá and R. Riguera, Tetrahedron: Asymmetry, 2000,
11, 2781.
Chem. Commun., 2001, 2154–2155
2155