Expedient synthetic transformation of ptychantins into Forskolin

Article abstract of DOI:10.1055/s-2008-1042930

Forskolin has been synthesized in 11 steps with a 17% overall yield from ptychantins A and B, which have been isolated from the liverwort Ptychanthus striatus in good yield. The 1α-hydroxy group was furnished by stereoselective reduction of the corresponding carbonyl group by sodium cyanoborohydride. The 9α-hydroxy group was introduced stereoselectively by epoxidation of Δ^9,11-enol ether. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2008-1042930

LETTER
929
Expedient Synthetic Transformation of Ptychantins into Forskolin
Synthetic
T
ransfor
i
m
ation
s
of Ptycha
a
ntins into F
h
orskolin iro Hagiwara,*a Masashi Tsukagoshi,a Takashi Hoshi,b Toshio Suzuki,b Toshihiro Hashimoto,c
Yoshinori Asakawac
a
Graduate School of Science and Technology, Niigata University, 8050, 2-Nocho, Ikarashi, Nishi-ku, Niigata 950-2181, Japan
Fax +81(25)2627368; E-mail: hagiwara@gs.niigata-u.ac.jp
Faculty of Engineering, Niigata University, 8050, 2-Nocho, Ikarashi, Nishi-ku, Niigata 950-2181, Japan
Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima, 770-8514, Japan
b
c
Received 22 December 2007
11
O
O
Abstract: Forskolin has been synthesized in 11 steps with a 17%
overall yield from ptychantins A and B, which have been isolated
from the liverwort Ptychanthus striatus in good yield. The 1a-hy-
droxy group was furnished by stereoselective reduction of the cor-
responding carbonyl group by sodium cyanoborohydride. The 9a-
hydroxy group was introduced stereoselectively by epoxidation of
HO
9
O
O
1
OH
OH
H
OAc
OAc
H
H
OH
D
9,11-enol ether.
2
1
Key words: terpenoids, total synthesis, forskolin, labdane diterpe-
noid
AcO
AcO
O
H
The highly oxygenated labdane diterpenoid forskolin (1,
Figure 1) was isolated from the Indian herb Coleus for-
skolii by a research group from Hoechst India in 1977.1
Forskolin (1) and congeners initially attracted much atten-
tion due to characteristic physiological activities of blood
pressure lowering and cardioprotective properties,2 and
subsequently due to the property of activating adenylate
cyclase.3
OR
H
OAc
3 R = H
4 R = Ac
Figure 1
O
O
HO
Owing to interesting physiological activities, numerous
synthetic studies appeared in the 1980’s. Ziegler et al.4 re-
ported a formal total synthesis, and the Hashimoto5 and
Corey6 groups successively completed total syntheses us-
ing an intramolecular Diels–Alder strategy. Later, Lett et
al.7 described an alternative total synthesis starting from
Corey’s intermediate in 1996.7c However, all synthetic
routes toward forskolin (1) provided a racemic material,
even though Corey6b and Lett7c demonstrated the possibil-
ity to a nonracemic synthesis by employing an optically
active intermediate.
O
O
O
3
2
H
O
H
OH
H
H
OH
6
5
O
O
HO
HO
O
O
O
5
1
OH
O
OH
OH
OH
H
H
7
8
Recently, we have demonstrated successful synthetic
transformations of ptychantin A (3) to forskolin (1) in 12
steps with a 12% overall yield and to 1,9-dideoxyforsko-
lin (2) in 8 steps with a 37% overall yield in the naturally
enantiomeric form (Scheme 1).8 Ptychantins A (3) and B
(4) were isolated in sufficient quantity (6.7 g of 3 from 1
kg of dry rhizome) from the common liverwort Ptychan-
thus striatus in western Japan.9
Scheme 1 Previous synthetic transformations of ptychantin A (3) to
forskolin (1) and 1,9-dideoxyforskolin (2)
prevention of easy epimerization at C-9 in C-11 keto com-
pounds. Although these issues have been solved success-
fully in our previous work, one major issue still remained,
that is, difficulty in deprotection of acetonide moieties
leading to compounds 6 and 8. Both acetonides at C-6 and
C-7 were very stable and deprotections were very slug-
gish, resulting in moderate yield as already mentioned by
Corey6a and Lett.7b
Major issues in our previous transformations were intro-
duction of two a-hydroxyl groups at C-1 and C-9 by selec-
tive manipulations of four hydroxyl groups considering
Herein, we present our studies on an alternative and more
expedient synthesis of forskolin (1) from ptychantins (3)
and (4) to address the long-standing issue on protection at
C-6 and C-7.
SYNLETT 2008, No. 6, pp 0929–0931
Advanced online publication: 11.03.2008
DOI: 10.1055/s-2008-1042930; Art ID: U12807ST
© Georg Thieme Verlag Stuttgart · New York
x
x.
x
x.
2
0
0
8

930
H. Hagiwara et al.
LETTER
Our initial efforts to protect as methylidene, cyclopentyl-
idene, and cyclohexylidene acetals, or bis-MOM or bis-
TES ether at C-6 and C-7 gave no satisfactory results in
either protection or deprotection. Some protecting groups
were fragile for further transformations and others resist-
ed deprotection. Strongly acidic reaction conditions had to
be avoided due to side reactions such as opening of the tet-
rahydropyran ring.
AcO
AcO
HO
HO
O
O
i
ii
H
H
OR
OH
H
H
OAc
OH
3 R = H
4 R = Ac
9
O
O
To resolve this issue, we chose carbonate group among
other possible protecting groups of 1,2-diols, which was
expected to be stable under aprotic basic reaction condi-
tions and could be hydrolyzed easily under protic basic re-
action conditions at the end of the synthetic sequence.
Towards this end, we had to reinvestigate the whole syn-
thetic sequence.
O
O
O
O
iv
iii
H
H
O
OH
O
O
H
H
OH
11
10
HO
HO
O
Ptychantins A (3) and B (4) were reduced with lithium
aluminum hydride to give tetraol 9 in high yield. Sterical-
ly less demanding two hydroxyl groups at C-1 and C-11
of tetraol 9 were protected with 2,2-dimethoxypropane re-
gioselectively as an acetonide to afford acetonide 9
(Scheme 2). Treatment of acetonide 10 with triphosgene
provided carbonate 11 in quantitative yield, while protec-
tion with carbonyldiimidazole gave no satisfactory re-
sults. Deprotection of acetonide 11 proceeded smoothly to
afford diol 12, which was oxidized by chromium pyridine
complex generated in situ (Sarret reagent) to give dike-
tone 13, although oxidation with other chromium reagents
or Swern oxidation was unsatisfactory. In previous syn-
thetic transformations,8 reduction of the carbonyl group at
C-1 was achieved with sodium in tert-butyl alcohol, giv-
ing thermodynamically more stable a-alcohol. Since the
protecting group at C-6 and C-7 is a carbonate, reduction
under nonbasic reaction condition was required to prevent
hydrolysis of the carbonate protecting group. In actuality,
reduction with sodium borohydride in pyridine according
to the known procedure11 provided 1a-alcohol 14 in only
8% yield along with a large amount of recovered starting
material 13 and its deprotected product. Among various
reducing agents tested, reduction was carried out with so-
dium cyanoborohydride in acetic acid and methanol to
give stereoselectively the desired 1a-alcohol 14 accompa-
nied by diol 15, which was oxidized by Sarett reagent
back to diketone 13 in 86% yield. The carbonate protect-
ing group was intact under these reaction conditions. Re-
gioselectivity might be ascribed to steric interaction by
three axial methyl groups at C-8, C-10, and C-13, which
prevented approach of the reducing agent at C-11. Higher
stereoselectivity might be explained by attack of hydride
from the less-hindered b-face of the carbonyl group at C-
1 avoiding steric interference by three 1,3-a-diaxial C–H
O
O
O
v
vi
H
H
O
O
O
O
H
H
O
O
12
O
13
HO
HO
HO
O
O
vii
H
H
O
+
O
O
O
O
H
H
O
14
15
MeO
HO
MeO
HO
O
O
viii
ix
OH
O
O
O
O
O
H
H
O
16
17
O
O
HO
HO
O
O
x
xi
1
OH
OH
O
OH
O
O
H
H
OH
18
19
Scheme 2 Reagents and conditions: i, LiAlH4, Et2O, r.t., 6 h, 97%;
ii, 2,2-dimethoxypropane, PPTS, r.t., 17 h, 85%; iii, triphosgene, pyr-
idine, CH2Cl2, r.t., 18 h, quant.; iv, 10% HClO4, THF, r.t., 67 h, 86%;
v, CrO3, pyridine, CH2Cl2, r.t., 54 h, 81%; vi, NaBH3CN, MeOH,
AcOH, r.t., 31 h, 63%, diol 15, 30%; vii, KH, Me2SO4, THF, 0 °C, 2
h, 82%; viii, MCPBA, Cs2CO3, DCE, r.t., 13 h, 85%; ix, 10% CSA,
dioxane–H2O, r.t., 14 d, 78%; x, 10% K2CO3 aq, MeOH, r.t., 1.5 h,
bonds at C-3, C-5, and C-9. In the absence of acetic acid, 86%; xi, Ac2O, pyridine, 0 °C, 18 h, r.t., quant.
only epimerization at C-9 occurred. Treatment of keto al-
cohol 14 with potassium hydride and dimethyl sulfate pro-
tively by MCPBA epoxidation of the D9,11-enol ether 16
ceeded in a thermodynamically controlled manner to
afford D9,11-enol ether 16.10 According to MOPAC AM-1
calculations, D9,11-enol ether 16 is 4 kcal/mol more stable
than the corresponding D11,12-enol ether. Introduction of
the C-9 hydroxyl group was accomplished stereoselec-
probably via interaction with the a-hydroxyl group at C-1
followed by concomitant in situ epoxide opening to give
hydroxyl enol ether 17. Among other organic and inor-
ganic bases tested, cesium carbonate gave the best result.
Hydrolysis of enol ether 17 was carried out in the presence
Synlett 2008, No. 6, 929–931 © Thieme Stuttgart · New York

LETTER
Synthetic Transformation of Ptychantins into Forskolin
931
(2) Bhat, S. V.; Dohadwalla, A. N.; Bajwa, B. S.; Dadkar, N. K.;
Dornauer, H.; de Souza, N. J. J. Med. Chem. 1983, 26, 486.
(3) (a) Abbadi, M.; Morin, C. Bioorg. Med. Chem. Lett. 1999, 9,
1779. (b) Laurenza, A.; Sutkowski, M. E.; Seamon, K. B.
Trends Pharm. Sci. 1989, 10, 442.
of camphorsulfonic acid (CSA) to provide ketone 18 with
2% of 19. Since hydrolysis of enol ether 17 often resulted
in ring opening of the hydropyran ring, mild reaction con-
ditions were employed. Deprotection of the carbonate
protecting group proceeded easily as anticipated with
methanolic potassium carbonate to give 7-deacetylforsko-
lin (19). Final acetylation completed the synthetic trans-
formation to forskolin (1), whose spectral data were
identical to those of natural 1.12
(4) Ziegler, F. E.; Jaynes, B. H.; Salndane, M. T. J. Am. Chem.
Soc. 1987, 109, 8115.
(5) Hashimoto, S.; Sakata, S.; Sonegawa, M.; Ikegami, S. J. Am.
Chem. Soc. 1988, 110, 3670.
(6) (a) Corey, E. J.; Da Silva Jardine, P.; Rohloff, J. C. J. Am.
Chem. Soc. 1988, 110, 3672. (b) Corey, E. J.;
In summary, we have achieved alternative and more expe-
dient synthetic transformations to forskolin (1) from pty-
chantins A (3) and B (4) in 11 steps with a 17% overall
yield.
Da Silva Jardine, P. Tetrahedron Lett. 1989, 30, 7297.
(7) (a) Delpech, B.; Calvo, D.; Lett, R. Tetrahedron Lett. 1996,
37, 1015. (b) Delpech, B.; Calvo, D.; Lett, R. Tetrahedron
Lett. 1996, 37, 1019. (c) Calvo, D.; Port, M.; Delpech, B.;
Lett, R. Tetrahedron Lett. 1996, 37, 1023.
(8) Hagiwara, H.; Takeuchi, F.; Kudou, M.; Hoshi, T.; Suzuki,
T.; Hashimoto, T.; Asakawa, Y. J. Org. Chem. 2006, 71,
4619.
Acknowledgment
This work was partially supported by a Grant-in-Aid for Scientific
Research on Priority Areas (17035031 for H.H. and T.S.) from The
Ministry of Education, Culture, Sports, Science and Technology
(MEXT), Japan.
(9) Hashimoto, T.; Horie, M.; Toyota, M.; Taira, Z.; Takeda, R.;
Tori, M.; Asakawa, Y. Tetrahedron Lett. 1994, 35, 5457.
(10) Hrib, N. J. Tetrahedron Lett. 1987, 28, 19.
(11) Khandelwal, Y.; Morares, G.; de Souza, N. J.; Fehlhaber, H.
W.; Paulus, E. F. Tetrahedron Lett. 1986, 27, 6249.
(12) Compounds 9–18 have satisfactory 1H NMR, 13C NMR, IR,
and high-resolution mass spectral data along with optical
rotation values.
References and Notes
(1) Bhat, S. V.; Bajwa, B. S.; Dornauer, H.; de Souza, N. J.;
Fehlhaber, H. W. Tetrahedron Lett. 1977, 18, 1669.
Synlett 2008, No. 6, 929–931 © Thieme Stuttgart · New York