Enantioselective synthesis of (-)-paeonilide

Article abstract of DOI:10.1039/c2cc18172j

The first enantioselective synthesis of (-)-paeonilide is reported. Starting from inexpensive furan-3-carboxylic acid the targeted monoterpene was obtained in 12 steps via an asymmetric cyclopropanation-lactonization cascade and a stereoselective side chain insertion at an acetal-like position.

Full text of DOI:10.1039/c2cc18172j

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 3457–3459
www.rsc.org/chemcomm
COMMUNICATION
Enantioselective synthesis of (À)-paeonilidew
Klaus Harrar and Oliver Reiser*
Received 30th December 2011, Accepted 9th February 2012
DOI: 10.1039/c2cc18172j
The first enantioselective synthesis of (À)-paeonilide is reported.
Starting from inexpensive furan-3-carboxylic acid the targeted
monoterpene was obtained in 12 steps via an asymmetric
cyclopropanation-lactonization cascade and a stereoselective
side chain insertion at an acetal-like position.
Natural products of herbal origin had an indisputable impact
on modern medicinal chemistry and still break ground for new
drugs. Paeonia root bark containing prescriptions like the
multiherbal formula GuiZhiFuLing-Wan are employed in
China, Japan and Korea to alleviate the syndromes of blood
stasis and stiffness of abdominal muscles.1–3 (+)-Paeonilide
(+)-(1), a monoterpenoid isolated from Paeonia delavayi,4 is a
substructure of the privileged class of ginkgolides. In bioassays it
showed selective inhibition of the platelet aggregation induced by
PAF (platelet activating factor) with an IC50 value of 8 mg mLÀ1
,
Scheme 1 Retrosynthetic analysis of (À)-paeonilide (À)-(1).
Table 1 Asymmetric cyclopropanation of 6
but importantly, no effect on the platelet aggregation induced by
adenoside diphosphate (ADP) or arachidonic acid (AA).4
Two racemic5,6 and one ex-chiral pool7 synthesis of paeonilide
have been reported to date, the latter achieved the synthesis of
(+)-paeonilide (+)-(1) starting from chiral (À)-carvone. We report
here a strategy that allows facile access to either enantiomer of
paeonilide. Since only the (+)-enantiomer was tested in bioassays
until now, we opted to demonstrate our strategy with the synthesis
of its antipode to gain further insight into the biological activity of 1.
The three key steps of our strategy are the synthesis of 4 by an
enantioselective cyclopropanation of 6 derived from commercially
available furan-3-carboxylic acid (5) (Scheme 1 and Table 1),
its transformation to the core structure 3 by donor–acceptor
substituted cyclopropane ring-opening/intramolecular lactonization
cascade of 4, and synthesis of 2 by introducing an allyl side
chain at the acetal-like position in 3.
Entry
Diazoester R
Ligand
Yield [%]
% ee
1
2
3
4
5
6
Et
tBu
Et
Et
tBu
tBu
7a
7
27
38
22
31
38b
34
74
65
74
83
83
19
8aa
8b
8b
8c
a
b
Ref. 8b. 53% based on recovered starting material.
We had previously developed the cyclopropanation of
furan-2-carboxylic acid esters with ethyl diazoacetate8 as a facile
entry into various g-butyrolactone containing natural products,9
which proceeds in the presence of copper(I)-bis(oxazoline)
CuOTfÁ8a in 94% ee. The asymmetric cyclopropanation of the
corresponding 3-methylester 6 turned out to be more challenging,
Institut fur Organische Chemie, Universitat Regensburg,
¨
¨
Universitatsstr. 31, 93053 Regensburg, Germany.
¨
E-mail: oliver.reiser@chemie.uni-regensburg.de;
Fax: +49 941 943 4121; Tel: +49 941 943 4631
w Electronic supplementary information (ESI) available: Preparation
and analytical data for all new compounds. CCDC 861132À861134.
For ESI and crystallographic data in CIF or other electronic format.
See DOI: 10.1039/c2cc18172j
most likely due to the fact that the ester group in the 3-position
is too far away to effectively interact with the catalyst, as depicted
in the transition state 9 that rationalizes the asymmetric pathway
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 3457–3459 3457

View Online
Scheme 2 Synthesis of (À)-paeonilide (À)-(1).
of the process. Nevertheless, 4 could be obtained in 83% ee
using CuOTfÁ8b (Table 1, entries 4 and 5), being surprisingly
more selective than the sterically more bulky catalysts CuOTfÁ
8a (entry 3) or CuOTfÁ8c (entry 6). In order to avoid the
competing twofold cyclopropanation, it turned out to be
advantageous to run the reaction only to about 50% conversion,
nevertheless, 4b could be readily produced on a 20 g scale.
Unfortunately, all efforts to enhance the enantiomeric excess
via recrystallization at this stage failed due to the excellent
solubility of 4b in most solvents.
In order to transform the hydroxycyclopropane ester unit
into the desired lactone, it was necessary to first reduce the
double bond in 4b to avoid rearomatization to a furan during the
ring opening of the donor–acceptor-substituted cyclopropane.
However, direct hydrogenation of 4b under a variety of conditions
only proceeded in low yields.
In contrast, the carboxylic acid 10 being obtained by
selective hydrolysis of 4b allowed the quantitative transformation
to 11, in which the carboxylic acid group is placed selectively
on the concave face of the bicycle. Acid induced cyclopropane
ring-opening/lactonization10 proceeded smoothly to 12, establishing
the core structure of paeonilide in which the relative stereochemistry
of the acid substituent with respect to the ring junctions needs to be
corrected. Treatment of 12 with pyridine gave rise to 3, exploiting
the higher stability of bicyclo[3.3.0] frameworks that orient larger
groups on its convex face. Notably, the stereocenters on the ring
junctions undergo epimerization as proved by X-ray-structure
analysis of 12 and the ammonium salt obtained from 3 and
(R)-1-(4-chlorophenyl)ethylamine (Scheme 3), being also in
agreement with epimerization studies of similar furolactones.11
The final challenge towards paeonilide (1) consisted of the
introduction of the side chain at the acetal-like position of 3,
thus requiring the creation of a quaternary stereocenter to
arrive at 15 (Scheme 4).
Scheme 3 Base induced isomerization of 12 to 3.
the addition of nucleophiles to 16 followed by relactonization
to give the desired 15.
In practice (Scheme 2), lactone 3 was treated with Jones-
reagent, simultaneously opening the lactone and oxidizing the
hemi-acetale to the a,b-dicarboxy-g-lactone 13 in very good
yield. Addition of allylmagnesiumbromide followed by acidic
workup gave rise to 2 in which an allyl group was installed
stereoselectively into the bicyclic framework.
All attempts to realize this goal by a direct C–H insertion of
a carbene following leads known in the literature12 for achieving
such transformations at the 2-position of THF or at the acetal
carbon of 1,3-dioxolanes with diazoesters in the presence of copper
or rhodium catalysts failed. As an alternative, lactone opening
followed by oxidation to 13 was envisioned, which should allow
Reduction of the carboxylic acid at this stage to create the
required alcohol functionality afforded only very low yields.
Oxidizing the allyl substituent with Jones-reagent combined
with catalytic amounts of mercury(II)-acetate led to 14, while
the oxidation under Wacker conditions afforded the desired
c
3458 Chem. Commun., 2012, 48, 3457–3459
This journal is The Royal Society of Chemistry 2012

View Online
In conclusion we reported the first enantioselective synthesis
of the natural product (À)-paeonilide (À)-(1), which was
synthesized in 12 steps and an overall yield of 4.4% (7.7%
brsm) starting from 3-furoic acid (5). Biological tests implicate
the inactivity of (À)-1 against the PAF receptor, giving further
credit to the unique and selective activity of naturally occur-
ring (+)-paeonilide (+)-(1).
We thank Dr Egon Koch, Schwalbe GmbH&CoKG, for
carrying out biological testing on (À)-1 as well as Dr Manfred
Zabel, Sabine Stempfhuber and Dr Michael Bodensteiner,
Department of Crystallography, University of Regensburg,
for X-ray structure analyses.
Notes and references
Scheme 4 Synthetic strategy for the introduction of side chains R at
1 M. Kimura, I. Kimura, H. Nojima, K. Takahashi, T. Hayashi,
M. Shimizu and N. Morita, Jpn. J. Pharmacol., 1984, 35, 61–66.
2 L. Chen, D. Wang, J. Wu, B. Yu and D. Zhu, J. Pharm. Biomed.
Anal., 2009, 49, 267–275.
the acetal-like position of furolactones.
3 S. Hatakeyama, M. Kawamura, E. Shimanuki and S. Takano,
Tetrahedron Lett., 1992, 33, 333–336.
4 J. K. Liu, Y. B. Ma, D. G. Wu, Y. Lu, Z. Q. Shen, Q. T. Zheng and
Z. H. Chen, Biosci., Biotechnol., Biochem., 2000, 64, 1511–1514.
5 Y. Du, J. Liu and R. J. Linhardt, J. Org. Chem., 2007, 72,
3952–3954.
6 C. Wang, J. Liu, Y. Ji, J. Zhao, L. Li and H. Zhang, Org. Lett.,
2006, 8, 2479–2481.
7 C. Wang, H. Zhang, J. Liu, Y. Ji, Z. Shao and L. Li, Synlett, 2006,
1051–1054.
8 (a) C. Bohm, M. Schinnerl, C. Bubert, M. Zabel, T. Labahn,
E. Parisini and O. Reiser, Eur. J. Org. Chem., 2000, 2955;
(b) O. Reiser, M. Schinnerl, C. Bohm and M. Seitz,
Tetrahedron: Asymmetry, 2003, 14, 765–771; (c) C. Bohm and
O. Reiser, Org. Lett., 2001, 3, 1315–1318; (d) E. Jezek, A. Schall,
P. Kreitmeier and O. Reiser, Synlett, 2005, 915–918.
9 (a) A. P. G. Macabeo, A. Kreuzer and O. Reiser, Org. Biomol.
Chem., 2011, 9, 3146–3150; (b) M. Schanderl, W. B. Jeong,
M. Schwarz and O. Reiser, Org. Biomol. Chem., 2011, 9,
2543–2547; (c) S. Kalidindi, W. B. Jeong, A. Schall,
R. Bandichhor, B. Nosse and O. Reiser, Angew. Chem., Int. Ed.,
2007, 46, 6361; (d) B. Nosse, R. B. Chhor, W. B. Jeong, C. Bohm
and O. Reiser, Org. Lett., 2003, 5, 941; (e) R. B. Chhor, B. Nosse,
S. Sorgel, C. Bohm, M. Seitz and O. Reiser, Chem.–Eur. J., 2003, 9,
260–270; (f) M. Seitz and O. Reiser, Curr. Opin. Chem. Biol., 2005,
9, 285–292.
10 For related methodology see (a) T. P. Brady, S. H. Kim, K. Wen
and E. A. Theodorakis, Angew. Chem., Int. Ed., 2004, 43, 739–742;
(b) S. D. Haveli, P. R. Sridhar, P. Suguna and S. Chandrasekaran,
Org. Lett., 2007, 9, 1331–1334; For a review on the utilization of
donor–acceptor substituted cyclopropanes see H.-U. Reissig and
R. Zimmer, Chem. Rev., 2003, 103, 1151–1196.
11 R. Weisser, W. Yue and O. Reiser, Org. Lett., 2005, 7, 5353–5356.
12 (a) J. M. Fraile, J. I. Garcia, J. A. Mayoral and M. Roldan, Org.
Lett., 2007, 9, 731–733; (b) A. Caballero, M. M. Diaz-Requejo,
T. R. Belderrain, M. C. Nicasio, S. Trofimenko and P. J. Perez,
Organometallics, 2003, 22, 4145–4150; (c) H. M. L. Davies and
Ø. Loe, Synthesis, 2004, 2595–2608; (d) M. P. Doyle, Top. Organomet.
Chem., 2004, 13, 203–222.
Fig. 1 X-Ray structure of (À)-paeonilide (À)-(1).
product as well but was accompanied by several by-products.
Finally, a reduction–benzoylation–oxidation sequence of 14 via
the corresponding diol selectively allowed functionalization of
the primary alcohol. Final oxidation of the secondary alcohol
with Dess–Martin periodinane (DMP) yielded (À)-paeonilide
(À)-(1) in 83% ee corresponding to the enantioselectivity
achieved in the cyclopropanation of 6. All spectroscopic data
were in accordance to literature data, furthermore, the structure
of (À)-1 was confirmed by X-ray crystallography (Fig. 1).
The antagonistic activity of (À)-1 against the PAF receptor
was found to be negligible compared to the naturally occurring
enantiomer (+)-1: the thrombocyte aggregation was inhibited
only by 20% at a concentration of 30 mg mLÀ1 (94 mM) of
(À)-1 (cf. (+)-1 (IC50 = 8 mg mLÀ1, 25 mM)). Presumably, the
antagonism observed is caused by the impurity of the eutomeric
(+)-1 in the measured sample.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 3457–3459 3459