Cleavage of ring A and formation of an unusual nor-triterpene skeleton via the Baeyer-Villiger reaction

Article abstract of DOI:10.1016/j.tetlet.2012.08.068

With the aim to generate a compound library for our biological screening studies, cycloastragenol was subjected to chemical transformation studies. The Baeyer-Villiger oxidation experiments provided an interesting 3,5-seco-4-nor-triterpene skeleton via ring opening followed by an unusual rearrangement. A new methodology is described for transforming triterpenoids into 3,5-seco-4-nor derivatives.

Full text of DOI:10.1016/j.tetlet.2012.08.068

Tetrahedron Letters 53 (2012) 5864–5867
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Cleavage of ring A and formation of an unusual nor-triterpene skeleton via the
Baeyer–Villiger reaction
Özgür Tag˘ a, Ali Çag˘ır b, Ikhlas A. Khan c, Erdal Bedir d,
⇑
a Department of Chemistry, Faculty of Science, Ege University, Bornova-Izmir, Turkey
b Department of Chemistry, Faculty of Science, Izmir Institute of Technology, 35430 Urla-Izmir, Turkey
c Department of Pharmacognosy, School of Pharmacy, The University of Mississippi, University, MS 38677, USA
d Department of Bioengineering, Faculty of Engineering, Ege University, 35100 Bornova-Izmir, Turkey
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 July 2012
Revised 1 August 2012
Accepted 16 August 2012
Available online 23 August 2012
With the aim to generate a compound library for our biological screening studies, cycloastragenol was
subjected to chemical transformation studies. The Baeyer–Villiger oxidation experiments provided an
interesting 3,5-seco-4-nor-triterpene skeleton via ring opening followed by an unusual rearrangement.
A new methodology is described for transforming triterpenoids into 3,5-seco-4-nor derivatives.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Cycloastragenol
Semi-synthesis
Baeyer–Villiger
3,5-Seco-4-nor-triterpenes
Saponins have been given attention due to their bioactivities,
and have become one of the primary targets for new pharmacolog-
ically active antitumor molecules.
intermediate in the biosynthesis of different phytosterols.9 For this
reason, cycloartenol and its weakly polar derivatives are wide-
spread in the plant kingdom. The plants of the Astragalus genera
are the richest source of this class of compounds. The main sapoge-
nol of many cycloartane-type glycosides found in this genus is
Cellular screening of various semi-synthetic triterpenoid com-
pounds formally derived from lupane, the aglycone of a well-known
anti-cancer agent betulinic acid (BA), has identified a number of ana-
logues as potential anti-cancer drug candidates.1,2 It is also worth
mentioning that BA itself, as well as some of its derivatives, have
been described as potent anti-HIV agents.3 A recent study revealed
that a semi-synthetic oleanane-type saponin, 2-cyano-3,12-dioxo-
olean-1,9-dien-28-oic acid (CDDO) is a multi-functional molecule
with promising clinical potential as a chemo-preventive agent and
as a therapeutic agent for the treatment of cancer.4,5 In vitro studies
have shown that CDDO induces cell differentiation, growth inhibi-
tion, and apoptosis in human leukemia,6,7 osteosarcoma,6 and breast
cancer8 cell lines. As a result, lupane derivatives and CDDO are cur-
rently undergoing clinical trials.
Semi-synthetic anti-cancer drug-discovery programs focusing
on saponins have mainly engaged with commercially available trit-
erpenoids such as oleanolic acid and ursolic acid, less common
miscellaneous aglycones such as cycloartanes, lanostanes, and ho-
panes have not been evaluated for their potentials.
The cycloartanes, which are unique triterpenoids with a charac-
teristic 9,19-cyclopropane ring, occupy a special position among
low molecular weight bioregulators since cycloartenol is a key
cycloastragenol
(CA),
20(R),24(S)-epoxy-3b,6a,16b,25-
tetrahydroxycycloartane.
However, studies on the preparation of CA analogues have so far
been limited.10–12 In continuation of our studies on the chemistry
and biological activities of Astragalus cycloartanes,13–16 we have di-
rected our efforts toward the exploration of various analogues of
CA to generate a library of structurally similar compounds for cyto-
toxicity screening studies.
Herein, we report an unusual modification of ring A of CA by
utilizing the Baeyer–Villiger (BV) oxidation.
Following on from our research into the transformations of
cycloastragenol (1), we first aimed to obtain astragenol (AG), which
is the main artifact of the acidic hydrolysis process of CA produc-
tion from the glycosidic precursors, and subsequently other deriv-
atives (Fig. 1).
Acid-catalyzed cyclopropane ring-opening of CA yielded astrag-
enol (AG) (2) as a white solid in 65% yield.17 The structure of the
semi-synthetic product was characterized by HRESIMS and NMR
analyses. The structure of hydrolyzed product AG was readily de-
duced based on the following observations: (i) the disappearance
of the characteristic 9,19-cyclopropane protons; (ii) the observa-
tion of the extra methyl group at d 1.14 (H3-19) and d 23.4 (C-
19); (iii) the appearance of a trisubstituted olefinic system due to
⇑
Corresponding author. Tel./fax: +90 232 3884955 32.
E-mail address: erdalbedir@gmail.com (E. Bedir).
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2012.08.068

Özgür Tag˘ et al. / Tetrahedron Letters 53 (2012) 5864–5867
5865
cated the presence of hydroxy (3406 cmꢀ1) and carbonyl (1730,
1710, 1769 cmꢀ1) functional groups. The HRESIMS of 5 provided
[M+H]+ at m/z 461.2549 [(+) mode] and [M+Cl]ꢀ at m/z 495.1696
[(ꢀ) mode] implying the molecular formula C26H36O7.
To our surprise, inspection of the 1H NMR spectrum of 5 showed
noteworthy discrepancies compared to starting molecule 4. Firstly,
instead of five tertiary methyls, a primary (d 1.16, t, 6.8 Hz) and
four tertiary methyl groups (d 1.04, 1.07, 1.11, 1.47, each s) were
observed together with two low-field signals at
J = 6.8 Hz) and 3.36 (d, J = 6.0 Hz). Second, the absence of
d
4.15 (q,
D
Figure 1. Structures of cycloastragenol (1) and astragenol (2).
9(11) ole-
finic protons was apparent from the 1H NMR spectrum. The 13C and
DEPT spectra showed that the resonances assigned to the product
comprised 5 methyl, 9 methylene, 3 methine, and 9 quaternary
carbons, four of which were carbonyl carbons as ester/lactone (d
174.5 and 177.3) or ketone carbons (d 209.6 and 215.9) (Table 1).
The characteristic signals of the 20,24-olide (ring-E) and ring-D
[C(13)? C(17)] in compound 5 were readily deduced from the pro-
ton and carbon chemical shifts (Table 1). Taking into account the
results of comprehensive 1D- and 2D NMR studies, and the previ-
ously reported spectroscopic data of related isolated metabolites, it
was inferred that 5 had a seco-structure with missing carbons.
The combined use of DQF–COSY and HMQC spectra, allowed the
assignment of five spin systems (Fig. 3) excluding the signals
ascribable to the 20,24-olide and the E-ring: ‘A’ (H2-1 ? H2-2), ‘B’
(ethanoyl) and ‘C’(H2-5), ‘D’ (H2-7 ? H-8) and ‘E’ (H-11 ? H2-12).
The d 1.52 and 1.91 signals, corresponding with a methylene
carbon at d 30.9 in the HMQC spectrum, coupled with another
methylene system at d 2.34 and 2.70 (dC 30.2) (‘A’). The spin sys-
tem B started with the primary methyl resonance (d 1.16, t,
J = 6.8 Hz), which showed a cross peak with a quartet signal at
d 4.15 (J = 6.8 Hz), suggesting an ethanoyl unit. These two spin
systems (A and B) were connected to each other based on the
long-range correlations from an ester carbonyl (d 174.5) to the
primary alcohol proton (d 4.15) and H2-1, in the HMBC spec-
trum. The low-field proton at d 3.36 (d, J = 6.0 Hz) showed a
cross peak with the d 2.12 signal (dd, J = 14.8 and 5.7 Hz), while
the latter coupled with a resonance at 2.22 ppm (d, J = 14.8) (‘E’).
the
D
9(11) double bond (dH 5.27; dC 146.6 and 115.6) in the 1H and
13C NMR spectra.18
After obtaining AG, the oxidation of the secondary alcohols into
carbonyl groups was undertaken. For this transformation, high-va-
lent metal oxides or their mineral salts are the reagents of choice,
amongst which the Jones reagent is the best known Cr(VI) reagent
for the oxidation of secondary alcohols to ketones.19
When we subjected compound 2 to Jones oxidation conditions,
compound 3 or 4 was isolated as the major product depending on
the reaction temperature (Fig. 2).20 At 0 °C, Jones oxidation re-
sulted in 3 in 64% yield as a white powder. When compared to 2,
the absence of low-field oxymethine signals due to H-3, H-6, and
H-16 in the 1H NMR, and the observation of three carbonyl carbons
(213.7, 209.6, 214.6) instead of the corresponding oxymethine car-
bons in the 13C NMR spectrum confirmed the formation of com-
pound 3. The accurate mass measurement also substantiated
these findings (HRESIMS: m/z 507.3115 [M+Na]+; Calcd:
484.3189 for C30H44O5).
On the other hand, when the oxidation was performed at room
temperature, an alternative major product (4, 72% yield) was ob-
served. The HRESIMS spectrum of 4 showed a major ion at m/z
458.2924 which was assigned to [M+NH4]+. Together with the 13C
NMR data, accurate mass data supported a molecular formula of
C27H36O5 (Calcd: 440.2563). The NMR spectra of 3 and 4 were iden-
tical except the 20,24-lactone ring. The C-25(O) resonance and its at-
tached methyl groups (Me-26 and Me-27) were missing in the 1H
and 13C NMR spectra of 4, whereas two low-field carbon signals at
d 86.9 and 176.3 were attributed to C-20 and C-24, ‘respectively’,
indicating a 20,24-olide ring system (c-lactone). The c-butyrolac-
Table 1
13C and 1H NMR data of
respectively)a
5 recorded in pyridine d5 (100 MHz and 400 MHz,
tones of acylated CA and its glycosides were previously prepared
by Isaev et al. via Jones oxidation to afford more potent cardiotonic
agents.11 A comparison of the spectral data of 4 and the reported
analogues confirmed the elucidated 25-nor structure.
The Baeyer–Villiger oxidation is generally a reliable process for
the conversion of cyclic and acyclic ketones into lactones and es-
ters, ‘respectively’.21 BV oxidations are also of great importance
for the field of natural products to obtain interesting products via
stereo- and regioselective reactions.22–24 Particularly on the basis
of its great success in synthetic steroid chemistry toward bioactive
agents, we utilized the BV oxidation on the 25-nor-AG (4) as a key
step to afford lactones and seco-cycloartane building blocks via
subsequent ring-opening.
C/H
dC (5) (ppm)
dH (5) (ppm), J (Hz)
1
2
3
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
30
OEt
30.9
30.2
174.5
50.8
209.6
43.0
1.52b, 1.91b
2.34b, 2.70b
2.30b, 2.59 (d, 13.3)
2.25b, 2.70b
2.68b
38.9
66.5
41.6
55.8
34.0
45.9
41.9
50.9
215.9
64.5
19.5
23.6
87.5
27.2
31.3
29.9
177.2
19.6
61.4, 15.2
3.36 (d, 5.6)
2.12 (dd, 14.8, 5.7), 2.22 (d, 14.4)
When compound 4 was subjected to BV oxidation using an ex-
cess of m-CPBA (5 equiv), a white crystalline solid (5) was obtained
as the major product after purification.25 The IR spectrum of 5 indi-
1.95 (d, 17.8), 2.22 (d, 18.0)
3.07 s
1.04 s
1.11 s
1.47 s
1.51b, 1.91b
2.34b, 2.66b
1.07 s
4.15 q (6.8), 1.16 t (7.2)
a
Assignments confirmed by COSY, HMQC and HMBC experiments.
Overlapping signals.
b
Figure 2. Structures of oxidation products 3 and 4.

5866
Özgür Tag˘ et al. / Tetrahedron Letters 53 (2012) 5864–5867
of C-10, C-13, and C-14 were readily assigned on the basis of
long-range connectivities to H3-19, H3-18, and H3-30,
‘respectively’.
The relative stereochemistry of the new chiral center at C-11
was assigned by a combination of coupling constant analysis and
molecular modeling studies performed on 5. The minimum energy
conformations of
a- and b-oriented epoxy rings were calculated
with a 3-D computer-generated model. Since the H-11 resonance
was observed as a doublet with a 6.0 Hz coupling constant (d
3.36), it must be equatorial, while the dihedral angle between H-
11 equiv and H-12ax is required to be about 90°. Only in the case
of an
out the b-orientation of the epoxy ring.
Consequently, 4,25,26,27,28,29-hexanor framework was
a-oriented epoxy ring is this dihedral angle possible, ruling
Figure 3. The spin systems (COSY) and key long-range correlations (HMBC) present
in compound 5.
a
established based on the spectral data, and a mechanistic sequence
for compound 5 is shown in Scheme 1.
The HMQC spectrum revealed two carbons (d 55.8 ? 3.36, H-11;
d 34.0 ? 2.12 and 2.22, H2-12) for the spin system E. The carbon
and proton chemical shifts of C(H)-11 was indicative of an epoxy
ring via oxidation of the D9(11) double bond. Observation of the
C-9 signal at d 66.5, and key long-range connectivities from C-
9 to H2-12 and H3-19, from C-11 to H3-19, and from C-12 to
H3-18 not only helped in locating the spin system E, but also
connected the fragments.
In the DQF–COSY spectrum, an isolated spin system (‘C’) due to
two methylene protons was observed at d 2.30 and 2.59 (d,
J = 13.3 Hz; dC 38.9 from HMQC), whereas the signal at d 2.68 (dC
38.9) displayed cross peaks with the d 2.25 and 2.70 methylene sig-
nals correlating with a carbon at d 43.0 (‘D’). The low-field shifts of
the methylene carbons and protons of the ‘C’ and ‘D’ spin systems
were significant for their adjacent position to a ketone carbon.
Accordingly, the ketone carbon at d 209.6 displayed long-range
correlations to the protons of the ‘C’ and ‘D’ spin systems, whereas
the d 50.9 signal showed connectivity with H3-19 allowing it to be
assigned unambiguously to C-5.
BV oxidation of 4 can yield esters 5 and 10 depending on the
added alcohol in the work-up steps. Although the formation of 6
was expected from BV oxidation of 4, the carbonyls can be further
protonated by excess m-chloroperbenzoic acid as shown in struc-
ture 6. After the addition of an alcohol, an equilibrium will be
formed between compounds 6 and 7. This equilibrium can be dis-
turbed toward the formation of an enol product 8 by a retro-aldol
reaction with elimination of acetone. Enol 8 rearranges into its keto
form 9 in alcoholic solution. So the addition of ethanol or methanol
in the work-up step yields the corresponding ethyl 5 or methyl 10
esters.
In conclusion, Jones reaction, followed by ring cleavage through
BV oxidation and an in situ retro-aldol reaction provided an unu-
sual 3,5-seco-4,25,26,27,28,29-hexanor-triterpenoid framework.
This skeleton was reported for the first time demonstrating the po-
tential of the BV oxidation for the transformation of the steroidal/
triterpenoid nucleus. The structure activity relationships of these
novel compounds including A-ring opened products will be evalu-
ated for their biological potential and the results will be reported in
due course.
Additionally, the C-8 (d 38.9) and H3-30 (d 1.07) correlation in
the HMBC spectrum revealed the location of the spin system ‘D’
in the B-ring. In a similar fashion, the quaternary carbon signals
O
OH
O
O
O
O
O
O
O
O
+
H
O
O
O
Cl
O
O
4
6
HO
R
O
O
H
O
O
O
~H+
RO
O
H
R
O
H
O
HO
R
O
OH
H
R
H
7
8
-H+
O
O
O
RO
RO
O
O
H
5 R=Et;
10 R=Me;
9
Scheme 1. A plausible mechanism for the transformation of 4 into compound 5 under BV reaction conditions.

Özgür Tag˘ et al. / Tetrahedron Letters 53 (2012) 5864–5867
5867
14. Gülcemal, D.; Çalısßkan, Ö. A.; Perrone, A.; Özgökçe, F.; Piacente, S.; Bedir, E.
Phytochemistry 2011, 72, 761–768.
15. Nalbantsoy, A.; Nesil, T.; Erden, S.; Çalısß, I.; Bedir, E. J. Ethnopharmacol. 2011,
Acknowledgements
_
This work is supported by TUBITAK (109S345) and COST Action
CM 0804. We are very grateful to Dr. Markus Ganzera and Dr. Bha-
rathi Avula for their help in mass measurements. We also acknowl-
edge partial support from the Global Research Network for
Medicinal Plants (GRNMP) and King Saud University.
134, 897–903.
_
_
16. Gür, C. S.; Onbasßılar, I.; Atilla, P.; Genç, R.; Çakar, N.; Gürhan, I. D.; Bedir, E. J.
Ethnopharmacol. 2011, 134, 844–850.
17. A solution of CA (1) (1.0 g, 2.28 mmol) in MeOH (50 mL) was treated with
concd H2SO4 (5 mL) and the mixture was heated under reflux for 6 h. The
reaction mixture was poured into H2O (50 mL) and extracted with EtOAc
(2 ꢁ 50 mL). The EtOAc extract was washed successively with aq satd NaHCO3
and H2O, dried over Na2SO4 and filtered. The filtrate was evaporated, and
chromatographic separation of the crude product on a silica gel column was
performed using hexanes:EtOAc (6:4) as eluent to give AG (2) as a white solid
(65% yield).
18. Kitagawa, I.; Wang, K. H.; Takagi, A.; Fuchida, M.; Miura, I.; Yoshikawa, M.
Chem. Pharm. Bull. 1983, 31, 689–697.
19. Cainelli, G.; Cardillo, G. Chromium Oxidations in Organic Chemistry; Springer-
Verlag: Berlin, 1984.
20. To a stirred mixture of 2 (500 mg, 1.14 mmol) in acetone (150 mL), Jones
reagent (0.4 mL, 3.3 mmol) was added slowly at 0 °C to furnish major
compound 3 as a white solid (64% yield). Freshly prepared Jones reagent
(CrO3, 3 g: distilled H2O, 9 mL and concd H2SO4, 1.5 mL) was added to a
solution of 2 (500 mg, 1.14 mmol) in acetone and stirred at room temperature
until completion of the reaction. After adding satd aq NaHCO3 (25 mL), the
aqueous mixture was extracted with CH2Cl2 (3 ꢁ 25 mL). The combined
organic layers were washed with brine solution (25 mL), dried over Na2SO4
and filtered. The filtrate was evaporated, and chromatographic separation of
Supplementary data
Supplementary data (HR-MS spectra and 1D-NMR spectra of 3
and 4, HR-MS spectra, 1D-and 2D-NMR spectra of 5, and carbon
NMR data of 3 and 4) associated with this article can be found,
in
the
online
version,
at
http://dx.doi.org/10.1016/
j.tetlet.2012.08.068.
References and notes
1. Sarek, J.; Urban, M.; Klinot, J.; Korinkova, G.; Hajduch, M. J. Nat. Prod. 2004, 67,
1100–1105.
2. Santos, R. C.; Salvador, J. A. R.; Marín, S.; Cascante, M. Bioorg. Med. Chem. Lett.
2009, 17, 6241–6250.
3. Mayaux, J. F.; Bousseau, A.; Pauwels, R.; Huet, T.; Henin, Y.; Dereu, N.; Evers, M.
Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 3564–3568.
4. Suh, N.; Yongping, W.; Tadashi, H.; Gordon, W. G.; Ethan, D.; William, F. H.;
Robert, A. M. Cancer Res. 1999, 59, 336–341.
5. Koohang, A.; Majewski, D. N.; Szotek, E. L.; Mar, A. A.; Eiznhamer, D. A.; Flavin,
M. T.; Xu, Z. Q. Bioorg. Med. Chem. Lett. 2009, 19, 2168–2171.
6. Ito, Y.; Pandey, P.; Sporn, M. B.; Datta, R.; Kharbanda, S.; Kufe, D. Mol.
Pharmacol. 2001, 59, 1094–1099.
7. Pedersen, I. M.; Shinichi, K.; Aaron, S.; Youngsoo, K.; Juan, M. Z.; Lula, C.; Laura,
R.; Andreeff, M.; Bennett, F.; Sporn, M. B.; Liotta, L. D.; Kipps, T. J.; Reed, J. C.
Blood 2002, 100, 2965–2972.
the crude product on
a silica gel column was performed by using
hexanes:EtOAc (8:2) as eluent to give 4 as a white solid (72% yield); mp
204–206 °C; ½a 2D3
ꢀ57.9 (c 0.005, MeOH).
ꢂ
21. Baeyer, A.; Villiger, V. Ber. 1900, 33, 858–864.
22. Harmata, M.; Rashatasakhon, P. Tetrahedron Lett. 2002, 43, 3641.
23. Rivera, D. G.; Pando, O.; Suardiaz, R.; Coll, F. Steroids 2007, 72, 466–473.
24. ten Brink, G. J.; Arends, I. W. C. E.; Sheldon, R. A. Chem. Rev. 2004, 104, 4105–
4123.
25. To a stirred solution of 4 (200 mg, 0.454 mmol) in CHCl3 (8 mL), 5 equiv of m-
CPBA (0.509 g, 2.27 mmol) were added portionwise at room temperature. The
mixture was refluxed with stirring for 8 h. The mixture was allowed to cool to
room temperature followed by evaporation in vacuo. The residue was
8. Lapillonne, H.; Konopleva, M.; Tsao, T.; Gold, D.; McQueen, T.; Madden, T.;
Andreeff, M. Exp. Hematol. 2003, 31, 66–69.
dissolved in
a minimum amount of EtOH, and then saturated NaHCO3
(25 mL) solution was added, followed by EtOAc extraction (3 ꢁ 25 mL). The
combined organic layers were washed with H2O (3 ꢁ 25 mL), brine
(3 ꢁ 25 mL), dried over anhyd Na2SO4 and filtered. The resulting filtrate was
evaporated in vacuo to give a solid. The solid residue was subjected to column
chromatography [hexanes–EtOAc (7:3)] to give 5 as a crystalline solid (128 mg,
9. Dewick, P. M. Medicinal Natural Products: A Biosynthetic Approach; Wiley, 2002.
10. Kuban, M.; Öngen, G.; Bedir, E. Org. Lett. 2010, 12, 4252–4255.
11. Isaev, I. M.; Iskenderov, D. A.; Isaev, M. I. Chem. Nat. Compd. 2009, 45, 381–384.
12. Yang, W.; Ye, M.; Huang, F.; He, W.; Guo, D. Adv. Synth. Catal. 2012, 354, 527–
539.
64% yield); mp 146–149 °C; ½a D23
ꢀ62.3 (c 0.005, MeOH).
ꢂ
_
13. Horro, I.; Bedir, E.; Perrone, A.; Özgökçe, F.; Piacente, S.; Çalısßkan, Ö. A.
Phytochemistry 2010, 71, 956–963.