Divergent synthesis of cytotoxic styryl lactones from D-xylose. The first total synthesis of (+)-crassalactone C

Article abstract of DOI:10.1021/ol701734s

A new divergent approach to (+)-goniofufurone (1) and 7-epi-(+)- goniofufurone (2), as well as the first total synthesis of crassalactone C (3), has been achieved starting from o-xylose. In a preliminary bioassay, all three natural products 1, 2, and 3 sh

Full text of DOI:10.1021/ol701734s

ORGANIC
LETTERS
2007
Vol. 9, No. 21
4235-4238
Divergent Synthesis of Cytotoxic Styryl
Lactones from -Xylose. The First Total
Synthesis of ( )-Crassalactone C
D
+
Velimir Popsavin,*,† Goran Benedekovic´,† Bojana Srec´o,† Mirjana Popsavin,†
Jovana Francuz,† Vesna Kojic´,‡ and Gordana Bogdanovic´‡
Department of Chemistry, Faculty of Sciences, UniVersity of NoVi Sad,
Trg D. ObradoVic´a 3, 21000 NoVi Sad, Serbia, and Institute of Oncology Sremska
Kamenica, Institutski put 4, 21204 Sremska Kamenica, Serbia
popsaVin@ih.ns.ac.yu
Received July 26, 2007
ABSTRACT
A new divergent approach to (
+
)-goniofufurone (1) and 7-epi-(+)-goniofufurone (2), as well as the first total synthesis of crassalactone C (3),
has been achieved starting from
D-xylose. In a preliminary bioassay, all three natural products 1, 2, and 3 showed remarkable in vitro
antiproliferative activities against K562, Raji, and HeLa neoplastic cell lines.
Asian trees of the genus Goniothalamus of the plant family
Annonaceae have long been recognized as a source of
biologically active styryl lactones.1 Many styryl lactones that
have been isolated from Goniothalamus species or synthe-
sized exhibited a notable cytotoxic activity against certain
human tumor cell lines.2 (+)-Goniofufurone (1) and 7-epi-
(+)-goniofufurone3 (2) are naturally occurring styryl lactones
that have attracted considerable attention since their isolation
from the stem bark of Goniothalamus giganteus (Annon-
aceae).4,5 Due to their unique structural features and promis-
ing antitumor activities, both natural products 1 and 2, along
with a number of their analogues, have been the targets of
many total syntheses.6,7 (+)-Crassalactone C (3) is a natural
7-O-cinnamoyl derivative of (+)-goniofufurone that was very
recently isolated from the leaves and twigs of Polyalthia
crassa.8 Its structure was determined on the basis of
spectroscopic methods. The absolute stereochemistry of 3
was established by treatment of the isolated (+)-1 with
cinnamoyl chloride. Apart from this nonselective and low-
yielding single step route,8 no total synthesis of 3 was hitherto
reported. As a part of our continuing interest in the synthesis
of natural products and analogues having γ-lactone rings,9
we have planned the synthesis of 1-3. We report herein their
† Department of Chemistry.
‡ Institute of Oncology.
(1) For a review on chemistry, biogenesis, and biological activities of
styryl lactones from Goniothalamus species, see: Blazquez, M. A.; Bermejo,
A.; Zafra-Polo, M. C.; Cortes, D. Phytochem. Anal. 1999, 10, 161.
(2) For recent reviews on cytotoxicity of styryl lactones and their
analogues, see: (a) de Fatima, A.; Modolo, L. V.; Conegero, L. S.; Pilli,
R. A.; Ferreira, C. V.; Kohn, L. K.; de Carvalho, J. E. Curr. Med. Chem.
2006, 13, 3371. (b) Mereyala, H. B.; Joe, M. Curr. Med. Chem. Anti-Cancer
Agents 2001, 1, 293. (c) See also ref. 1
(3) Due to differences in numbering systems, compound 2 is sometimes
named as 8-epi-goniofufurone (e.g., see ref 8).
(4) Fang, X. P.; Anderson, J. E.; Chang, C. J.; Fanwick, P. E.;
McLaughlin, J. L. J. Chem. Soc., Perkin Trans. 1 1990, 1655.
(5) Fang, X. P.; Anderson, J. E.; Chang, C. J.; McLaughlin, J. L.;
Fanwick, P. E. J. Nat. Prod. 1991, 54, 1034.
(6) For recent reviews on syntheses of styryl lactones, see: (a) Mondon,
M.; Gesson, J.-P. Curr. Org. Synth. 2006, 3, 41. (b) Zhao, G.; Wu, B.; Wu,
X. Y.; Zhang, Y. Z. Mini-ReV. Org. Chem. 2005, 2, 333.
(7) (a) Sartillo-Melendez, C.; Cruz-Gregorio, S.; Quintero, L.; Sartillo-
Piscil, F. Lett. Org. Chem. 2006, 3, 504. (b) Mihovilovic, M. D.; Bianchi,
D. A.; Rudroff, F. Chem. Commun. 2006, 3214. (c) Ferna´ndez de la Pradilla,
R.; Ferna´ndez, J.; Viso, A.; Ferna´ndez, J.; Go´mez, A. Heterocycles 2006,
68, 1579. (d) Prasad, K. R.; Gholap, S. L. Synlett 2005, 2260. (e) Ruiz, P.;
Murga, J.; Carda, M.; Marco, J. A. J. Org. Chem. 2005, 70, 713.
(8) Tuchinda, P.; Munyoo, B.; Pohmakotr, M.; Thinapong, P.; Sophasan,
S.; Santisuk, T.; Reutrakul, V. J. Nat. Prod. 2006, 69, 1728.
10.1021/ol701734s CCC: $37.00
© 2007 American Chemical Society
Published on Web 09/15/2007

divergent synthesis using D-xylose as a chiral precursor and
a preliminary bioassay against human K562, Raji, HeLa, and
MRC-5 cell lines. The data related to cytotoxic activities of
1-3 against the mentioned cell lines are herein for the first
time reported.
All three target compounds 1-3 contain five contiguous
stereocenters and display a clear structural similarity. We
thus wanted to design a divergent synthesis for all three
compounds from a common intermediate. Among other
methods, the required [3.3.0] bicyclic lactone core could be
formed through condensation of Meldrum’s acid with a
protected sugar lactol derivative.10 Accordingly, we envisaged
the retrosynthetic concept depicted in Scheme 1. For both 1
Disconnection of 3 shows that it can be derived from the
diol 14 by a number of successive transformations that
involve regioselective Mitsunobu reaction in the presence
of cinnamic acid and hydrolytic removal of the 1,2-O-
cyclohexylidene protective group, followed by γ-lactone
formation. Intermediate 14 in turn should be accessible from
5 by removal of benzyl protective group from C-3.
Scheme 2. Synthesis of (+)-Goniofufurone (1) and
7-epi-Goniofufurone (2)
Scheme 1. Retrosynthetic Analysis of (+)-Goniofufurone (1),
7-epi-Goniofufurone (2), and (+)-Crassalactone C (3)
and 2, lactone ring-removal would give rise via, respectively,
1a and 2a to the two epimeric alcohols 6 and 5. These can
be prepared from the same aldehyde 4 by means of
stereoselective addition of PhMgBr. Synthesis of 4 itself is
visualized from D-xylose by established chemical reactions.9b
The synthesis of 1 and 2 is presented in Scheme 2.
Addition of phenylmagnesium bromide in ether to 4 led to
two diastereomeric alcohols 5 and 6 in a 6:1 ratio and 82%
combined yield. The observed diastereoselectivity may be
explained as a result of the 1,2-chelation control.11 The major
L-ido isomer 5 has the same stereochemistry as target 2, while
the minor D-gluco stereoisomer 6 is of the same absolute
(9) (a) Popsavin, V.; Srec´o, B.; Krstic´, I.; Popsavin, M.; Kojic´, V.;
Bogdanovic´, G. Eur. J. Med. Chem. 2006, 41, 1217. (b) Popsavin, V.; Krstic´,
I.; Popsavin, M.; Srec´o, B.; Benedekovic´, G.; Kojic´, V.; Bogdanovic´, G.
Tetrahedron 2006, 62, 11044. (c) Popsavin, V.; Grabezˇ, S.; Popsavin, M.;
Krstic´, I.; Bogdanovic´, G.; Divjakovic´, V. Tetrahedron Lett. 2004, 45, 9409.
(d) Popsavin, V.; Krstic´, I.; Popsavin, M. Tetrahedron Lett. 2003, 44, 8897.
(10) Bruns, R.; Wernicke, A.; Koll, P. Tetrahedron 1999, 55, 9793.
(11) Gracza, T.; Szolcsa´nyi, P. Molecules 2000, 5, 1386.
Org. Lett., Vol. 9, No. 21, 2007
4236

configuration as both targets 1 and 3. Efforts to change the
isomeric ratio in favor of 6 were unsuccessful (e.g., use of
phenylcerium dichloride11 gave 5 and 6 in a 2:1 ratio, but in
only 20% yield). To prepare compound 6, an efficient two-
step route was developed that involved configurational
inversion at C-5 in 5 under the Mitsunobu conditions.12
Accordingly, treatment of 5 with diethyl azodicarboxylate,
triphenylphosphine, and benzoic acid gave the expected 5-O-
benzoyl derivative 7 which upon deprotection with metha-
nolic sodium methoxide formed 6 in 55% yield (from three
steps). Hydrolytic removal of the cyclohexylidene protective
group in 6 with aqueous acetic acid, gave the corresponding
lactol 8. Since the pyridine solution of 8 mutarotated to a
more positive equilibrium value {[R]20D ) +23.7 f +39.6
(77 h)}, the crystalline 8 was the â anomer (Hadson’s rule).
Attempted condensation of 8 with Meldrum’s acid in the
presence of tert-butylamine10 failed to give the expected
γ-lactone 9. However, when the last reaction was carried
out in the presence of triethylamine as a catalyst,13 the desired
γ-lactone 9 was obtained in 52% yield. Final cleavage of
benzyl protecting group in 9 gave (+)-goniofufurone (1).
The physical and spectral data of thus-obtained sample 1
were identical to those reported in the literature.7e,10
Scheme 3. Synthesis of (+)-Crassalactone C (3)
The stereoisomer 5 was converted to (+)-7-epi-goniofu-
furone (2) by using the same methodology as that already
applied for the conversion of 6 to 1. Treatment of 5 with
aqueous acetic acid gave 10 (64%) along with a minor
amount of 11 (9%).14 Reaction of 10 with Meldrum’s acid,
followed by hydrogenolytic 5-O-deprotection gave (+)-7-
epi-goniofufurone (2). All physical constants and spectral
data of thus prepared natural product 2 were in good
agreement with those reported in the literature.10,15
The synthesis of (+)-crassalactone C (3) is presented in
Scheme 3. Removal of the benzyl protective group in 5 gave
the corresponding diol 13 (70%) that was further allowed to
react with cinnamic acid under the standard Mitsunobu
conditions. The corresponding cinnamic ester 14 was thus
obtained (63%) accompanied by a minor amount of 3,5-
anhydro derivative 15 (16%). The side product 15 was
presumably formed by a competitive intramolecular nucleo-
philic displacement process.12 Indeed, when the last reaction
was carried out in the absence of cinnamic acid (Ph3P,
DEAD, refluxing toluene, 1.5 h), the oxetane 15 was isolated
as a main reaction product in 71% yield. The assignment of
stereochemistry at the C-5 in product 15 was confirmed by
an NOE interaction between H-1 and H-5, indicating that
these protons are in close proximity on the same side of the
ring. Compound 15 might be of use for a synthesis of a
hitherto unknown conformationally constrained analogue of
1 (5,7-anhydrogoniofufurone). Hydrolytic removal of the
cyclohexylidene protective group in 14 gave the expected
lactol 16 (59%), which upon treatment with Meldrum’s acid
in the presence of triethylamine gave (+)-crassalactone C
(3), with physical and spectral properties in reasonable
agreement with those reported in the literature.8
Compounds 1-3 were evaluated for their in vitro anti-
proliferative activity toward human myelogenous leukemia
(K562), Burkitt’s lymphoma (Raji cells), cervix carcinoma
(HeLa), and normal fetal lung (MRC-5) cell lines. Cytotoxic
activities were evaluated by using standard MTT assay16 after
exposure of cells to the tested compounds for 72 h. The
commercial cytotoxic agent doxorubicin (DOX) was used
as a positive control in this bioassay. The results are shown
in Table 1.
Table 1. In Vitro Cytotoxicity of 1-3 and DOX
IC50, µMa
compd
K562
Raji
HeLa
MRC-5
1
2
3
DOX
0.41
0.028
3.56
0.25
18.45
1.25
15.46
2.98
8.32
0.89
11.25
0.07
>100
>100
>100
0.10
a IC50 is the concentration of compound required to inhibit the cell growth
by 50% compared to an untreated control.
(12) Mitsunobu, O. Synthesis 1981, 1.
(13) Mata, F. Z.; Martinez, M. B.; Perez, J. A. G. Carbohydr. Res. 1990,
201, 223.
(14) The side product 11 was presumably formed as a result of the
competitive intramolecular dehydration of 10 (in furanose form), during
removal of acetic acid by codistillation with toluene.
(15) Yang, M.; Li, H. M.; Zhao, G.; Yu, Q.-S.; Ding, Y. Chin. J. Chem.
2000, 18, 225.
In general, all three natural products 1-3 exhibited
antiproliferative activities against all of the malignant cell
(16) Scudiero, D. A.; Shoemaker, R. H.; Paull, K. D.; Monks, A.; Tierney,
S.; Nofziger, T. H.; Currens, M. J.; Seniff, D.; Boyd, M. R. Cancer Res.
1988, 48, 4827.
Org. Lett., Vol. 9, No. 21, 2007
4237

lines (K562, Raji, and HeLa) but were devoid of any
cytotoxicity against the normal fetal lung fibroblasts (MRC-
5). 7-epi-(+)-Goniofufurone (2) was found to be the most
efficient cytotoxic agent against all neoplastic cells, with IC50
values ranging from 0.028 to 1.25 µM. The most pronounced
antiproliferative activity of compound 2 was recorded against
the K562 cells, being ca. 9-fold more potent than the
commercial cytotoxic agent doxorubicin. Against the Raji
cells, 7-epi-(+)-goniofufurone (2) was over 2-fold more
active with respect to the reference compound, doxorubicin.
This compound was also active against HeLa cells, but it
was almost 13-fold less potent than the reference compound
(DOX). Remarkably, compound 2 exhibited 1 order of
magnitude higher cytoxicity against all the malignant cell
lines when compared to (+)-goniofufurone (1) and (+)-
crassalactone C (3). The previous biological data2 in different
neoplastic cell lines suggested that (+)-goniofufurone (1) was
more active than 7-epi-(+)-goniofufurone (2). It appears that
K562, Raji, and HeLa represent the first neoplastic cell lines
in which stereoisomer 2 shows a stronger in vitro antipro-
liferative activity with respect to (+)-goniofufurone (1).
In addition to providing a divergent access to the natural
products 1-3, this approach is flexible and straightforward.
It uses nonexpensive reagents and a readily available starting
material.17 These advantages make the synthetic methodology
suitable for easy preparation of a variety of 7-O-substituted
(+)-goniofufurone analogues for biological evaluation. In a
preliminary MTT bioassay, all three natural products 1, 2,
and 3 showed remarkable in vitro antiproliferative activities
against K562, Raji, and HeLa neoplastic cell lines. 7-epi-
(+)-Goniofufurone (2) was found to be the most active
compound, being at least 1 order of magnitude more
cytotoxic than (+)-goniofufurone (1). To the best of our
knowledge, the K562, Raji, and HeLa cells represent the first
malignant cell lines, for the time being, in which 7-epi-(+)-
goniofufurone (2) exhibits more potent in vitro antiprolif-
erative activity with respect to (+)-goniofufurone (1). Finally,
based upon observed antitumor activities of 1-3, as well as
their inactivity against the normal MRC-5 cells, we believe
that these natural products may serve as important leads in
the synthesis of more potent and selective antitumor agents
derived from the parent compounds 1-3.
In conclusion, by utilizing a chiral pool strategy based on
D-xylose as the starting material, we have completed a new
divergent synthesis of the natural styryl lactones (+)-
goniofufurone (1) and (+)-7-epi-goniofufurone (2), as well
as the first total synthesis of (+)-crassalactone C (3), a novel
7-O-cinnamoyl (+)-goniofufurone derivative that has been
recently isolated from the tropical plant Polyalthia crassa.8
Acknowledgment. Financial support from the Ministry
of Science of the Republic of Serbia (Project No. 142005)
is gratefully acknowledged.
Supporting Information Available: Experimental pro-
cedures and full spectroscopic data for all compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(17) Bols, M. Carbohydrate Building Blocks; John Wiley & Sons Ltd.:
New York, 1996.
OL701734S
4238
Org. Lett., Vol. 9, No. 21, 2007