Concise Synthesis of the Core Structures of Saundersiosides

Article abstract of DOI:10.1021/acs.orglett.5b00821

A divergent synthesis of three core pentacyclic lactones of nine rearranged cholestane sapogenins, saundersiosides A-H (1-8) and candicanoside A (9), is reported. Key features include a one-flask CBS reduction/Brown hydroboration-oxidation, a SmI₂

Full text of DOI:10.1021/acs.orglett.5b00821

Letter
pubs.acs.org/OrgLett
Concise Synthesis of the Core Structures of Saundersiosides
Shou-Ling Cheng, Xiao-Ling Jiang, Yong Shi,* and Wei-Sheng Tian*
The Key Laboratory of Synthetic Chemistry of Natural Substances, Shanghai Institute of Organic Chemistry, Chinese Academy of
Sciences, 345 Lingling Road, Shanghai 200032, China
S
* Supporting Information
ABSTRACT: A divergent synthesis of three core pentacyclic
lactones of nine rearranged cholestane sapogenins, saundersio-
sides A−H (1−8) and candicanoside A (9), is reported. Key
features include a one-flask CBS reduction/Brown hydro-
boration−oxidation, a SmI2-mediated intramolecular Refor-
matskii reaction, and an intramolecular transesterification. This
synthesis provides a general strategy and key precursors for the
collective synthesis of natural and designed saundersiosides.
An efficient formal synthesis of candicanoside A is also achieved.
he disclosure of the exceptionally potent antitumor
activities of OSW sapogenins has triggered a continuous
(IC50 from nM to μM against HL-60 human promyelocytic
leukemia cells). The first and only synthesis of candicanoside A
by the Yu group4 is elegant but not flexible to the synthesis of
saundersiosides. Our goal is to provide a unified solution to
these natural products and their analogues. Herein we report a
divergent synthesis of the core structures of their sapogenins
(compounds 10−12, Scheme 1).
T
phytochemical investigation of the genus Ornithogalum, which
resulted in the isolation of more than 100 steroidal glycosides.1
Among many structurally novel glycosides, a number of them
stood out for their rearranged cholestane sapogenins, as
depicted in Figure 1. Saundersiosides A−H (1−8, isolated
from Ornithogalum Saundersiae bulbs by Mimaki between
1993−1999)2 and candicanoside A (9, isolated from Galtonia
candicans by Mimaki in 2000)3 are closely related 24(23 → 22)
abeo-cholestane glycosides and show potent antitumor activities
Scheme 1. Retrosynthesis Analysis of the Core Structures
The steroidal sapogenins of 1−9 differed with each other
mainly in the oxidative states and assembly patterns of C18,
C20, and C23. Since the Yu group has successfully introduced
the C24−C27 side chain at the late stage of the candicanoside
A synthesis, we likewise identified lactones 10−12 as the core
structures and thus planned our synthesis toward them. We
envisaged that the six-membered lactones of 10−12 could be
Figure 1. Structures of saundersiosides A−H (1−8), candicanoside A
(9), and their analogues.
Received: March 20, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b00821
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Synthesis of the Core Structures of I−III
obtained through an intramolecular transesterification of the
seven-membered lactone of 13 (O18 → O16) and the oxidative
states of C18/C20 could be easily modified via simple
transformations. In turn, 13 could be prepared by an
intramolecular Reformatskii reaction from the triol derivative
14 which was designed to be prepared from dihydropreg-
nenolone 15 through reduction (C20-OH), hydroboration−
oxidation (C16-OH), and remote functionalization (C18-OH).
The remote functionalization of the angular C18 methyl
group often requires a C20-OH or a C11β-OH to perform the
transformation,5 and the C16-substituent, if there was one, at
the α face of the D-ring to minimize side reactions.6 Therefore,
diol 18, with the C20(S)-7 and C16α-hydroxyl groups and the
3,5-cyclo-6-methoxy-protected AB ring, was prepared from 15,
as shown in Scheme 2. Through a slightly modified two-step
procedure,8,9 the C5−C6 double bond of 15 was protected,
affording enone 16 in 85% yield on the 60 g scale. Then the
C20 ketone of 16 was stereoselectively transformed into the
C20(S)-OH of 17 through Corey−Bakshi−Shibata (CBS)
reduction,4b,10 and the C16α-OH was introduced through a
substrate-controlled Brown hydroboration−oxidation to pro-
vide diol 18. Since both CBS reduction and Brown hydro-
boration−oxidation use borane as a stoichiometric reagent, we
envisioned that incorporation of them in one flask would
simplify the operation. As expected, performing the reduction
at −15 °C for 6 h, and then keeping the reaction at 25 °C for
12 h before NaOH/H2O2 was added, effectively delivered diol
18 (89% yield, 50 g scale). The structure of 18 was secured by
an X-ray analysis (Figure 2). In this manner, three contiguous
stereocenters (C16, C17, and C20) were established in one
flask.
Figure 2. Ortep structures of 18 (left) and 22 (right).
yield on the 20 g scale. It was noted that the transformation
required 3 equiv of PhI(OAc)2 and 2 equiv of iodine.
Furthermore, despite being widely used in steroid synthesis,
the free radical remote functionalization of C18-Me was never
followed by a direct reduction step to provide 18,20-diol;
therefore, our method provided an important complement to
the existing protocol especially for the acid-labile substrates.
We then entered the next stage of the synthesis. A selective
esterification of the C18-OH in 14 with bromoacetic acid
(EDCI, DMAP, CH2Cl2, rt, 1 h) and a Dess−Martin oxidation
of the C20-OH gave 20 in good yield.14 Compound 20 quite
smoothly underwent the key transformation, an intramolecular
15
Reformatskii reaction mediated by SmI2 at ambient temper-
ature, to form the seven-membered lactone 13 in 89% yield as a
single isomer. In contrast, at −78 °C, this reaction only gave the
debromination product of 20, as did the reaction at ambient
temperature using additives or cosolvents such as HMPA, LiCl,
MeOH, and t-BuOH. We reasoned that the reductive
debromination of 20 occurred swiftly to form the SmIII enolate,
but it failed to attack the C20 carbonyl group at low
temperature because the reaction sites were seven atoms
away from each other. This distance was shortened through the
chelation of the SmIII ion with the reacting functional groups,
an effect which was fortunately achieved by running the
reaction at ambient temperature but was broken by using
additives or cosolvents. The high stereoselectivity was also
viewed as a consequence of chelation (as in structure A).16
With 13 in hand, we began to explore the synthesis of 10 and
11. Replacement of the C20-OH with hydrogen through a
Before remote functionalization of the C18-Me was
performed on 18, the C16-OH was selectively protected as
the TBDPS ether in 80% yield (TBDPSOTf, iPr2NEt, DCM, 0
°C to rt, 10 h). Employing Meystre’s hypoidodite method11
(Pb(OAc)4/I2, hv) we obtained the desired diol 14 in 40% yield
after LiAlH4 reduction,12 along with the cyclic ether in 42%
yield. Owing to the inefficiency in converting the byproduct to
14 through RuCl3−NaIO4 oxidation/LiAlH4 reduction, we
investigated other C18-oxidation methods. Finally, the
combination of a Suar
I2, hv) and a LiAlH4 reduction was employed to give 14 in 78%
́
ez iodine(III) oxidation13 (PhI(OAc)2/
B
DOI: 10.1021/acs.orglett.5b00821
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
dehydration−hydrogenation process (SOCl2, pyridine, 0 °C, 30
min; 10% Pd/C, 20 atm, EtOH-EtOAc, rt, 24 h17) and
deprotection of the TBDPS ether on the resulting product with
TBAF provided compound 21. Then, inverting the config-
uration of the C16α-OH was achieved through a Dess−Martin
oxidation/NaBH4 reduction process, which was accompanied
by a spontaneous intramolecular transesterification of the C22
carboxyl group from the C18-OH to the newly generated
C16β-OH, thus providing the desired lactone 11 (five steps,
64% overall yield from 13). The exposed C18-OH was oxidized
with Collins reagent to give aldehyde 10 in 69% yield.18
Yu and Tang have used the TBS-protected 11 (23) as an
intermediate in their synthesis, so we decided to obtain a formal
synthesis of candicanoside A. Interestingly, we found that 23,
prepared from 11 (TBSOTf, 2,6-lutidine, DCM, 0 °C, 69%), is
acid-labile and partly desilylated in neutralized CDCl3,
presumably owing to the steric crowding of this position, and
that the TMS-protected product 24 (TMSOTf, Et3N, DCM, 0
°C, 90%) is much more stable. Moreover, compound 11 was
protected as the MOM ether (MOMCl, iPr2NEt, Bu4NI, DCM,
83%) and the resulting 22 generated a crystal suitable for X-ray
analysis, thereby securing the stereochemistry. Until then we
have achieved a formal synthesis of candicanoside A in 16%
yield over 15 steps from dihydropregnenolone 15, which is
apparently superior to the previous route in efficiency (4% yield
over 18 steps from dehydroisoandrosterone).
esterification, provided another diol 26 with C18-OH exposed
for further transformation.
Oxidation of the C18-OH of 26 with PCC reagent directly
gave hemiacetal 27 as an inseparable mixture of epimers (1.6/1
1
at C18 by H NMR) in moderate yield. The 3,5-cyclo-6-
methoxy protection of 27 was then removed with aqueous HF
in MeCN because it is acid-sensitive and apt to react with
MeOH in an acidic medium. The resulting crude was treated
with p-toluenesulfonic acid in MeOH to deliver the desired
acetal 28 in 79% yield, also as an inseparable mixture of
epimers. To render the chromatographic separation of the
epimers, the C3-OH of 28 was protected as the TBS ether,
giving acetal 29 in 54% yield and 18-epi-29 in 42% yield. Acetal
29 is a more advanced intermediate than 12; therefore, we
accomplished the synthesis of the core structure of IV in 5%
(9% as mixture) yield over 15 steps.
In summary, we have achieved a divergent and effective route
for the core structures of saundersiosides and candicanoside A,
with lactone 13 as a common intermediate. The key
transformations featured herein include a one-flask CBS
reduction/Brown hydroboration−oxidation to install three
stereocenters from the starting enone, a SmI2-mediated
intramolecular Reformatskii reaction to connect the C20−
C22 bond in a highly stereoselective manner, and an
intramolecular transesterification to furnish the desired six-
membered lactone and to expose the C18-OH for further
manipulations. Our synthesis also exhibits an efficient formal
synthesis of candicanoside A. Installation of the C24−C27 side
chain and progress toward the natural products and the
designed analogues are ongoing in this laboratory and will be
reported in due course.
Finally, we moved on toward another target 12. As depicted
in Scheme 3, deprotection of the TBDPS group on 13 with
TBAF in THF gave diol 25 whose crystal was suitable for X-ray
analysis. The stereochemistry of C20 was secured as the R
configuration. Again, the Dess−Martin oxidation/NaBH4
reduction process, accompanied by an intramolecular trans-
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental details, spectral data, 1H and 13C NMR spectra of
all the new compounds, the X-ray crystallographic data and cif
files for 18, 22, and 25. The Supporting Information is available
free of charge on the ACS Publications website at DOI:
10.1021/acs.orglett.5b00821.
Scheme 3. Synthesis of the Core Structure of IV
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: shiong81@sioc.ac.cn.
*E-mail: wstian@sioc.ac.cn.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Financial support is provided by the National Natural Science
Foundation of China (21272258).
■
REFERENCES
■
(1) (a) Kubo, S.; Mimaki, Y.; Terao, M.; Sashida, Y.; Nikaido, T.;
Ohmoto, T. Phytochemistry 1992, 31, 3969−3973. (b) Mimaki, Y.;
Kuroda, M.; Kameyama, A.; Sashida, Y.; Hirano, T.; Oka, K.;
Maekawa, R.; Wada, T.; Sugita, K.; Beutler, J. A. Bioorg. Med. Chem.
Lett. 1997, 7, 633−636. (c) Morzycki, J.; Wojtkielewicz, A. Phytochem.
Rev. 2005, 4, 259−277. (d) Mimaki, Y. Nat. Prod. Commun. 2006, 1,
247−253. (e) Tang, Y.; Li, N.; Duan, J.-A.; Tao, W. Chem. Rev. 2013,
113, 5480−5514.
(2) (a) Kuroda, M.; Mimaki, Y.; Sashida, Y.; Nikaido, T.; Ohmoto, T.
Tetrahedron Lett. 1993, 34, 6073−6076. (b) Mimaki, Y.; Kuroda, M.;
C
DOI: 10.1021/acs.orglett.5b00821
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Kameyama, A.; Sashida, Y.; Hirano, T.; Oka, K.; Dobashi, A.; Koike,
K.; Nikaido, T. Bioorg. Med. Chem. Lett. 1996, 6, 2635−2638.
(c) Mimaki, Y.; Kuroda, M.; Sashida, Y.; Hirano, T.; Oka, K.; Dobashi,
A.; Koshino, H.; Uzawa, J. Tetrahedron Lett. 1996, 37, 1245−1248.
(d) Kuroda, M.; Mimaki, Y.; Sashida, Y.; Hirano, T.; Oka, K.; Dobashi,
A.; Li, H.-y.; Harada, N. Tetrahedron 1997, 53, 11549−11562.
(e) Kuroda, M.; Mimaki, Y.; Sashida, Y. Phytochemistry 1999, 52,
435−443.
(3) Mimaki, Y.; Kuroda, M.; Sashida, Y.; Yamori, T.; Tsuruo, T. Helv.
Chim. Acta 2000, 83, 2698−2704.
(4) (a) Tang, P.; Yu, B. Angew. Chem., Int. Ed. 2007, 46, 2527−2530.
(b) Tang, P.; Yu, B. Eur. J. Org. Chem. 2009, 2009, 259−269.
(5) (a) Majetich, G.; Wheless, K. Tetrahedron 1995, 51, 7095−7129.
(b) Reese, P. B. Steroids 2001, 66, 481−497. (c) Cekovic, Z.
Tetrahedron 2003, 59, 8073−8090.
(6) (a) Shi, Y.; Jia, L.-Q.; Xiao, Q.; Lan, Q.; Tang, X.-H.; Wang, D.-
H.; Li, M.; Ji, Y.; Zhou, T.; Tian, W.-S. Chem.Asian J. 2011, 6, 786−
790. (b) Gui, J.; Wang, D.; Tian, W. Angew. Chem., Int. Ed. 2011, 50,
7093−7096.
(7) Our experience suggests that C20(S)-OH is more suitable for
C18-Me remote functionalization than its C20(R)-OH counterpart,
owing to the faster reaction rate and higher yield (see ref 6).
(8) Julian, P. L.; Meyer, E. W.; Ryden, I. J. Am. Chem. Soc. 1950, 72,
367−370.
(9) By employing pyridine as the base instead of KOAc in the 3,5-
cyclo-6-methoxy-forming step, our procedure eliminates the gener-
ation of the 3,5-cyclo-6-acetoxy compound as a side product, which is
difficult to remove from 16.
(10) (a) Corey, E. J.; Bakshi, R. K.; Shibata, S. J. Am. Chem. Soc.
1987, 109, 5551−5553. (b) Corey, E. J.; Helal, C. J. Angew. Chem., Int.
Ed. 1998, 37, 1986−2012.
(11) Heusler, K.; Wieland, P.; Meystre, C. Org. Synth. 1965, 45, 57−
63.
(12) Jones oxidation was frequently performed after the remote
funtionalization of the C18-Me group to give a lactone in good yield.
As in our case, 3,5-cyclo-6-methoxy protection of the AB ring is acid-
labile and cannot survive Jones oxidation, so direct reduction of the
intermediate 19 was investigated.
(13) Betancor, C.; Freire, R.; Per
Tetrahedron 2005, 61, 2803−2814.
́ ́ ́
ez-Martín, I.; Prange, T.; Suarez, E.
(14) The outcome of the selective esterification of the C18-OH
group in 14 was unstable. We found that Dess-Martin oxidation of the
unwanted C20-OH protected byproduct also provided the desired
ketone 20 in 50% yield which might be caused by the shift of the
bromoacetyl group from C20-OH to C18-OH. So direct oxidation of
the crude product in esterification step gave an reasonable yield of
ketone 20.
(15) (a) Molander, G. A.; Harris, C. R. Chem. Rev. 1996, 96, 307−
338. (b) Edmonds, D. J.; Johnston, D.; Procter, D. J. Chem. Rev. 2004,
104, 3371−3404.
(16) (a) Molander, G. A.; Kenny, C. J. Org. Chem. 1988, 53, 2132−
2134. (b) Carroll, G. L.; Little, R. D. Org. Lett. 2000, 2, 2873−2876.
(c) Hamon, S.; Birlirakis, N.; Toupet, L.; Arseniyadis, S. Eur. J. Org.
Chem. 2005, 2005, 4082−4092. (d) Ogawa, Y.; Kuroda, K.; Matsuo, J.-
I.; Mukaiyama, T. Bull. Chem. Soc. Jpn. 2005, 78, 677−697.
(e) Nicolaou, K. C.; Ellery, S. P.; Chen, J. S. Angew. Chem., Int. Ed.
2009, 48, 7140−7165.
(17) Hydrogenation with NaBH4/NiCl2 gave a 1/1 mixture of 20(S)-
and 20(R)-epimers. The high stereoselectivity of the Pd/C catalyzed
hydrogenation was realized in a substatrate-controlled manner. The
catalyst approached the double bond from the less hindered convex
face.
(18) Dess−Martin, Swern, and Parikh−Doering oxidations only
provided aldehyde 10 in less than 20% yield.
D
DOI: 10.1021/acs.orglett.5b00821
Org. Lett. XXXX, XXX, XXX−XXX