A concise synthesis of ginkgolide M, a minor component of a terpene trilactone fraction from Ginkgo biloba roots

Article abstract of DOI:10.1021/np050403i

Ginkgolide M (GM), which is found only in the roots of the Ginkgo biloba tree and is an inhibitor of ligand-operated ion channels in the central nervous system, has been prepared in three steps from 10-benzylginkgolide C, an intermediate generated during the isolation and separation of ginkgolides from Ginkgo biloba leaf extract. The described synthetic sequence can be applied to access GM derivatives for biological studies.

Full text of DOI:10.1021/np050403i

J. Nat. Prod. 2006, 69, 429-431
429
A Concise Synthesis of Ginkgolide M, a Minor Component of a Terpene Trilactone Fraction
from Ginkgo biloba Roots
Sergei Bolshakov, Sergei V. Dzyuba,* John Decatur, and Koji Nakanishi*
Department of Chemistry, Columbia UniVersity, 3000 Broadway, New York, New York 10027
ReceiVed October 14, 2005
Ginkgolide M (GM), which is found only in the roots of the Ginkgo biloba tree and is an inhibitor of ligand-operated
ion channels in the central nervous system, has been prepared in three steps from 10-benzylginkgolide C, an intermediate
generated during the isolation and separation of ginkgolides from Ginkgo biloba leaf extract. The described synthetic
sequence can be applied to access GM derivatives for biological studies.
Ginkgo biloba leaf extract, an over-the-counter dietary supple-
ment or prescription product with multimillion dollar worldwide
sales, has long been advocated for its ability to improve memory
in the elderly.1 The potential of Ginkgo biloba leaf extract for the
treatment of cognitive impairment has been demonstrated in an array
of studies.1,2 However, neurophysiological effects and long-term
neuroprotective properties of the extract still need to be clarified.
Activation of ligand-gated ion channels and G-protein coupled
receptors is responsible for many processes, such as basal excitatory
synaptic transmission and many forms of synaptic plasticity that
are thought to control memory and learning.3 Ligand-gated ion
channels, including glycine, GABAA, NMDA, and AMPA recep-
tors, are involved in mediation of excitatory synaptic transmission,
a process by which neurons communicate with each other.3 Small
molecules that can modulate activity of these receptors are valuable
as mechanistic probes for the processes in the central nervous
system as well as pharmacological leads for developing remedies
for neurological disorders.
Although the molecular mechanisms of G. biloba extract
components are largely unknown, they are often attributed to the
presence of structurally unique compounds named ginkgolides
(Figure 1).4 Ginkgolides are believed to be responsible for a variety
of neuromodulatory effects exhibited by G. biloba leaf extract,
including learning and memory functions.5 Several studies addressed
structure-activity relationships of ginkgolides toward platelet-
activating factor6 and glycine7 receptors and have indicated a very
fine balance between the number and position of the hydroxy groups
around the ginkgolide skeleton and biological activity.
It was recently demonstrated that ginkgolides could block and
modulate the responses of several ion channel receptors.7a Note-
worthy, GM, which is found only in the root of Ginkgo biloba L.
(Ginkgoaceae), unlike other ginkgolides that are found in the leaves
as well, was shown to be the most potent natural ginkgolide in
blocking the responses of several receptor-gated channels, whereas
other tested ginkgolides (GA, GB, GC, and GJ) showed antago-
nistic properties exclusively toward glycine receptor. In particular,
of all the ginkgolides, GM exhibited the highest inhibition of
GABAA receptor and efficiently displaced TBPS8 from the con-
vulsant binding site of GABAA.7a Ion channel blocker properties
make GM a lead for potential treatment of neurodegenerative
disorders, such as Alzheimer’s disease.9
Scheme 1. Attempted 3,14-Dehydration of GC
to address the effect of subtle structural changes on ginkgolide-
receptor interactions. Yet the biological scope and potential of this
ginkgolide is not broadly studied; apparently, the available quantities
of GM are relatively small as compared to other ginkgolides, thus
making structure-activity relationship studies quite challenging.
Therefore, a practical synthetic preparation of GM is desirable.10
Dehydration of OH-3 from unprotected GC would create a plausible
intermediate, such as 1, en route to efficient synthesis of GM
(Scheme 1). However, subjection of GC to known dehydration
procedures10 led to either decomposition of the starting material or
formation of the double-dehydrated product 2 (Scheme 1).11
During the course of studies directed toward regiocontrolled
synthesis of fluorinated ginkgolides, we found that treatment of
GA with (diethylamino)sulfur trifluoride (DAST)12 provided no
fluorodehydroxylation at the C-10 position, but instead led to a
high yield selective elimination of the tertiary hydroxy group, OH-
3, affording GL (Scheme 2).13 Upon hydrogenation of the unsatur-
ated trilactone moiety of GL in the presence of Crabtree’s catalyst,14
a clean formation of epi-derivative 3 took place (Scheme 2). The
cis orientation of H-3 and H-14 in 3-dehydroxy-14-epi-GA, 3, was
confirmed by NOE studies: an NOE was observed between the
14-Me and 12-H, which indicated that the 14-Me and 3-H are in a
trans orientation; the 14-Me extends back toward the backbone,
i.e., “â-oriented” (Scheme 2).15
Attempts to achieve dehydration of GC upon reaction with DAST
under the conditions outlined in Scheme 2 led to decomposition of
the starting material.16 No desired elimination product was obtained
when the reaction was conducted at different temperatures (-78
and 0 °C); instead, the starting material was recovered. Application
of bases, i.e., pyridine and 4-(dimethylamino)pyridine (DMAP),
to facilitate the dehydration process17 was also unsuccessful.
Since extra hydroxy groups of GC (as compared to GA and GB)
are likely to contribute to the inefficiency of the dehydration, we
turned our attention to monoprotected GC analogues. It is relevant
From a structural point of view, GM lacks the tertiary hydroxy
group at the C-3 position, which is present in other ginkgolides
from G. biloba extract, and, therefore, represents a unique analogue
Dedicated to Dr. Norman R. Farnsworth of the University of Illinois
at Chicago for his pioneering work on bioactive natural products.
* To whom correspondence should be addressed. (S.V.D.) Tel: +1 212
854 5356. Fax: +1 212 932 1289. E-mail: svd2102@columbia.edu. (K.N.)
Tel: +1 212 854 2169. Fax: +1 212 932 1289. E-mail: kn5@columbia.edu.
10.1021/np050403i CCC: $33.50
© 2006 American Chemical Society and American Society of Pharmacognosy
Published on Web 02/22/2006

430 Journal of Natural Products, 2006, Vol. 69, No. 3
Notes
Figure 1. Ginkgolide structures from Ginkgo biloba (leaves and roots).
Scheme 2. DAST-Mediated GA to GL Conversion and Stereochemical Assignment of Hydrogenated GL
Scheme 3. Synthesis of GM
achieved in a few steps starting from the commercial G. biloba
leaf extract.20
Experimental Section
General Experimental Procedures. All reagents and solvents were
purchased from Aldrich and used as received. Ginkgolides were
available from earlier studies or isolated from BioGinkgo 27/7 extract
according to a literature procedure.17,20 1H NMR spectra were recorded
on a Bruker DPX-300 (300 MHz) spectrometer and are reported in
ppm from CDCl3 internal standard (7.26 ppm). 2D NOESY spectra
were acquired on a Bruker DMX 500 spectrometer under the following
conditions: TPPI mode; SW ) 3500 Hz; TD2 ) TD1*4 ) 1200; D1
) 5 s; NS ) 40. For the NOESY volume ratio analysis, several different
reference distances and NOEs were used, and the choice did not affect
the calculated distances significantly.
Example of Dehydration Procedure. 10-Benzyl-GC 4 (15.0 mg,
0.029 mmol) was dissolved in 1 mL of THF and cooled to -78 °C.
Pyridine (100 µL, 1.23 mmol) and DAST (100 µL, 0.76 mmol) were
added dropwise at this same temperature. The reaction mixture was
stirred at -78 °C for 10 min, warmed to room temperature, and then
kept for 40 min before quenching by addition of 2 mL of water. The
aqueous layer was extracted with 3 × 2 mL of EtOAc. The combined
organic layers were washed with brine, dried over sodium sulfate, and
concentrated in vacuo. Purification by flash chromatography on silica
gel (eluted with 1:1 EtOAc/hexane) afforded the desired 3,14-dehydro-
10-benzyl-GC 5 (13.1 mg, 0.026 mmol, 90% yield) as a white solid:
1H NMR (300 MHz, CDCl3) δ 7.40 (5H, m), 6.00 (s, 1H), 5.85 (1H,
s), 5.42 (1H, d, J ) 10.2 Hz), 4.82 (1H, s), 4.63 (1H, d, J ) 10.3 Hz),
4.60 (1H, d, J ) 5.0 Hz), 4.39 (1H, m), 2.20 (3H, s), 2.03 (2H, m),
1.92 (1H, d, J ) 12.4 Hz), 1.23 (9H, s).
to note that the 10-benzyl derivatives of GB and GC are
intermediates prepared for the separation of individual ginkgolides
from G. biloba leaf extract.18 Therefore, 10-benzyl-GC is an
attractive starting point to explore the synthesis of ginkgolides and
their derivatives. It was found that the reaction of 10-benzyl-GC 4
with DAST in the presence of pyridine in THF led to a clean
elimination of the OH-3 group, giving unsaturated lactone 5 in good
yield (Scheme 3).19 Hydrogenation of the unsaturated lactone moiety
yielded the known 14-epi-GM, which in turn was converted into
GM under previously reported conditions.10
Synthesis of GM. A glass liner for a stainless steel 45 mL Parr
high-pressure reactor equipped with a stir bar was charged with a
solution of 3,14-dehydro-10-benzyl-GC 5 (12.0 mg, 0.024 mmol) in a
mixture of EtOAc (1 mL) and MeOH (3 mL), then Pd/C (10% w/w, 2
mg) was added. The liner was inserted into the Parr reactor, and the
In conclusion, a short synthesis of GM has been achieved, which
features selective removal of the tertiary hydroxy group in the
presence of two unprotected secondary alcohol moieties and does
not require the use of isolated GC. Thus the whole process can be

Notes
Journal of Natural Products, 2006, Vol. 69, No. 3 431
(7) (a) Chatterjee, S. S.; Kondratskaya, E. L.; Krishtal, O. A. Pharma-
copsychiatry 2003, 36 (Suppl. 1), S68-S77. (b) Ivic, L.; Sands, T.
T. J.; Fishkin, N.; Nakanishi, K.; Kriegstein, A. R.; Stromgaard, K.
J. Biol. Chem. 2003, 278, 49279-49285. (c) Jaracz, S.; Nakanishi,
K.; Jensen, A. A.; Stromgaard, K. Chem. Eur. J. 2004, 10, 1507-
1518. (d) Kondratskaya, E. L.; Betz, H.; Krishtal, O. A.; Laube, B.
Neuropharmacology 2005, 49, 945-951.
pressure gauge and gas assembly were attached. The reactor was sealed,
charged and vented with 3 × 400 psi with H2, and recharged to 600
psi H2. The reaction mixture was stirred at room temperature for 18 h,
and then the reactor was vented. The solution was filtered through a
pad of Celite and concentrated in vacuo. The residue was dissolved in
1 mL of MeCN, and DMAP (12.0 mg, 0.096 mmol) and 0.1 mL of
water were added. The reaction mixture was stirred 80 °C in sealed
tube for 4 days and then quenched by addition of 2 mL of 1 M HCl.
The aqueous layer was extracted with 3 × 2 mL of EtOAc, and the
combined organic layers were washed with brine, dried over sodium
sulfate, and concentrated in vacuo. Purification by flash chromatography
on silica gel (eluted with 2:1 EtOAc/hexane) afforded GM (7.9 mg,
0.019 mmol, 77% yield over two steps), whose spectral data matched
those of the natural compound.4a
(8) TBPS: 35S-tert-butylbicyclophosphorothionate. Edgar, P. P.; Schwartz,
R. D. J. Neurosci. 1990, 10, 603-612.
(9) Schmitt, H. P. Med. Hypothesis 2005, 65, 259-265.
(10) Synthetic studies leading to formation of GM were reported
previously: (a) Weinges, K.; Rummler, M.; Schick, H. Liebigs Ann.
Chem. 1993, 1023-1027. (b) Maruyama, M.; Terahara, A.; Naka-
daira, Y.; Woods, M. C.; Takagi, Y.; Nakanishi, K. Tetrahedron Lett.
1967, 8, 315-319.
(11) Maruyama, M.; Terahara, A.; Nakanishi, K. Sci. Rep. Tohoku UniV.
1967, 50, 92-99.
(12) Middleton, W. J. J. Org. Chem. 1975, 40, 574-578.
(13) GA to GL conversion using POCl3/pyridine was previously re-
ported: Weinges, K.; Ru¨mmler, M.; H. Schick, H.; Schilling, G.
Leibigs Ann. Chem. 1993, 287-291.
Acknowledgment. We are grateful to Pharmanex for a generous
gift of Ginkgo biloba leaf extract. This work was financially supported
by NIH (GM-MH068817).
References and Notes
(14) Attempts to achieve the desired hydrogenation using Pd/C failed.
Upon screening several hydrogenation catalysts, Crabtree’s catalyst
was identified as the most efficient. For a review of Crabtree’s
catalysts see: Grabtree, R. H. Acc. Chem. Res. 1990, 23, 95-101.
(15) Further proof of stereochemistry is as follows: the 2D NOESY cross-
peak volumes were analyzed using a NOE ratio method upon which
the volume of the Me-14/H-12 cross-peak yielded a distance of 0.35
nm, consistent with the cis orientation of H-3 and H-14. The cis
orientation of H-2 and H-3 was supported by NOESY volume
analysis. The cross-peak volume between H-2 and H-3 was similar
in magnitude to the peak volume for several other cis-oriented
protons, such as between the H-6 and H-7R protons and between
the H-2 and H-1R protons. For the NOE ratio method see: Newhaus,
D.; Williamson, M. The Nuclear OVerhauser Effect in Structural and
Conformational Analysis, 2nd ed.; Wiley-VCH: New York, 2000.
(16) GB, however, was successfully converted into GK, using a DAST-
mediated protocol, Scheme 2.
(17) Somekh, L.; Shanzer, A. J. Org. Chem. 1982, 48, 907-908.
(18) Jaracz, S.; Malik, S.; Nakanishi, K. Phytochemistry 2004, 65, 2897-
2902.
(19) The DAST-mediated procedure appears to be quite general for the
dehydration of ginkgolides and ginkgolide derivatives, as 10-benzyl-
GB and 10-methyl-GC underwent a clean elimination of the OH-3
group, yielding the corresponding 3,14-unsaturated products in 90
and 85% yields, respectively.
(1) Stromgaard, K.; Vogensen, S. B.; Nakanishi, K. In Encyclopedia of
Dietary Supplements; Coates, P. M., Ed.; Marcel Dekker Inc.: New
York, 2005; pp 249-257.
(2) (a) van Dongen M. C. J. M.; van Rossum, E.; Knipschild, P. In
Ginkgo biloba; van Beek, T. A., Ed.; Harwood Academic Publish-
ers: Australia, 2000; Chapter 22, pp 385-442. (b) Crews, W. D.,
Jr.; Harrison, D. W.; Griffin, M. L.; Falwell, K. D.; Crist, T.; Longest,
L.; Hehemann, L.; Rey, S. T. HerbalGram 2005, 67, 43-62.
(3) (a) Guderman, T.; Schoneberg, T.; Schultz, G. Annu. ReV. Neurosci.
1997, 20, 399-427. (b) Ito, M, Physiol. ReV. 2001, 81, 1143-1195.
(c) El Far, O.; Betz, H. Biochem. J. 2002, 365, 329-336. (d)
Debanne, D.; Daoudal, G.; Sourdet, V.; Russier, M. J. Physiol. 2004,
97, 403-414. (e) Malbon, C. C. Nature ReV. Mol. Cell Biol. 2005,
6, 689-701. (f) Lujan, R.; Shigemoto, R.; Lopez-Bendito, G.
Neuroscience 2005, 130, 567-580.
(4) (a) van Beek, T. A. Bioorg. Med. Chem. 2005, 13, 5001-5012. (b)
GK and GL were identified as native ginkgolides present in trace
amounts in the terpene trilactone fraction of Ginkgo biloba leaf
extract: Wang, Y.; Sheng, L. S.; Lou, F. C. Acta Pharmcol. Sin.
2001, 31, 606-608.
(5) (a) Nakanishi, K. Bioorg. Med. Chem. 2005, 13, 4983-5000. (b)
Stromgaard, K.; Nakanishi, K. Angew. Chem., Int. Ed. 2004, 43,
1640-1658.
(6) (a) Hosford, D.; Mencia-Huerta, J. M.; Page, C.; Braquet, P.
Phytotherapy Res. 1988, 2, 1-24. (b) Hu, L.; Chen, Z.; Xie, Y.;
Jiang, Y.; Zhen, H. Bioorg. Med. Chem. 2000, 8, 1515-1521. (c)
Vogensen, S. B.; Stromgaard, K.; Shindou, H.; Jaracz, S.; Suehiro,
M.; Ishii, S.; Shimizu, T.; Nakanishi, K. J. Med. Chem. 2003, 46,
601-608. (d) Zhu, W.; Chen, G.; Hu, L.; Luo, X.; Gui, C.; Luo, C.;
Puah, C. M.; Chen, K.; Jiang, H. Bioorg. Med. Chem. 2005, 13, 313-
322.
(20) All ginkgolides can be selectively isolated from the Ginkgo biloba
leaf extract in one step: Nakanishi, K.; Jaracz, S.; Malik, S.; Ishii,
H.; Dzyuba, S. V. PCT Int. Appl. WO 2005046829, 2005, 61 pp.
NP050403I