Divergent synthesis of cytotoxic styryl lactones related to goniobutenolides A and B, and to crassalactone D

Article abstract of DOI:10.1021/ol302860z

Goniobutenolides A (1) and B (2), crassalactone D (3), 4-epi-crassalactone D (4), and the corresponding 7-epimers have been synthesized starting from d-glucose. The key step in the synthesis of 1 and 2 is a new one-pot sequence comprised of a Z-selective Wittig olefination/lactonization/β-elimination. Preparation of 3 and 4 included the final 5-endo-trig spirocyclization of 1 and 2. The synthesized products were evaluated for their in vitro antiproliferative activity against selected tumor cell lines.

Full text of DOI:10.1021/ol302860z

ORGANIC
LETTERS
2012
Vol. 14, No. 23
5956–5959
Divergent Synthesis of Cytotoxic Styryl
Lactones Related to Goniobutenolides
A and B, and to Crassalactone D
,†
Velimir Popsavin,* Ivana Kovacevic, Goran Benedekovic, Mirjana Popsavin,
†
ꢀ
ꢁ †
Vesna Kojic, and Gordana Bogdanovic
ꢁ †
ꢁ ‡
ꢁ‡
Department of Chemistry, Biochemistry and Environmental Protection, Faculty of
ꢁ
Sciences, University of Novi Sad, Trg D. Obradovica 3, 21000 Novi Sad, Serbia, and
Oncology Institute of Vojvodina, Institutski put 4, 21204 Sremska Kamenica, Serbia
velimir.popsavin@dh.uns.ac.rs
Received October 17, 2012
ABSTRACT
Goniobutenolides A (1) and B (2), crassalactone D (3), 4-epi-crassalactone D (4), and the corresponding 7-epimers have been synthesized starting
from D-glucose. The key step in the synthesis of 1 and 2 is a new one-pot sequence comprised of a Z-selective Wittig olefination/lactonization/
β-elimination. Preparation of 3 and 4 included the final 5-endo-trig spirocyclization of 1 and 2. The synthesized products were evaluated for their in
vitro antiproliferative activity against selected tumor cell lines.
Goniobutenolides A and B (1 and 2, Figure 1) were
isolated, together with other related styryl-lactones, from
the ethanolic extracts of the stem bark of Goniothalamus
giganteus Hook. f. and Thomas (Annonaceae).1 These
compounds were found to be cytotoxic against certain
human tumor cell lines and have been the targets of
synthetic efforts by several groups.2 The structures of 1
and 2 were originally determined by spectral methods; in
particular, a threo relationship was proposed for the diol
total syntheses of goniobutenolides A and B as well as their
7-epimers reassigned the relative configuration of the diol
moiety as erythro and established their absolute stereo-
chemistries as 1 and 2, respectively.3 A number of styryl
lactones were recently isolated from the tropical plant
Polyalthia crassa.4 Separation of the tree’s leaves and twigs
ethyl acetate-soluble extract led to the isolation of several
new cytotoxic compounds including the (þ)-crassalactone
D (3). The relative configuration of 3 was established by
single-crystal X-ray diffraction analysis, and its absolute
stereochemistry was determined by NMR studies on (R)-
and (S)-MTPA esters of 3. Recently, two asymmetric total
syntheses of (þ)-3 have been reported both employing an
oxidative spirocyclization of a furan as the key step.5
1
moiety, based on the H NMR data.1 However, the first
† University of Novi Sad.
‡ Oncology Institute of Vojvodina.
(1) Fang, X.-P.; Anderson, J. E.; Chang, C.-J.; McLaughlin, J. L.
Tetrahedron 1991, 47, 9751.
(2) (a) Shing, T. K. M.; Tsui, H.-C.; Zhou, Z.-H. J. Org. Chem. 1995,
60, 3121. (b) Ko, S. Y.; Lerpiniere, J. Tetrahedron Lett. 1995, 36, 2101. (c)
Mukai, C.; Hirai, S.; Kim, I. J.; Kido, M.; Hanaoka, M. Tetrahedron
(3) Shing, T. K. M.; Tai, V. W.-F.; Tsui, H.-C. J. Chem. Soc., Chem.
Commun. 1994, 1293.
(4) Tuchinda, P.; Munyoo, B.; Pohmakotr, M.; Thinapong, P.;
ꢂ
1996, 52, 6547. (d) Surivet, J.-P.; Vatele, J.-M. Tetrahedron 1999, 55,
ꢁ
13011. (e) Solladie, G.; Hanquet, G.; Rolland, C. TetrahedronLett. 1999,
40, 177.
Sophasan, S.; Santisuk, T.; Reutrakul, V. J. Nat. Prod. 2006, 69, 1728.
r
10.1021/ol302860z
Published on Web 11/16/2012
2012 American Chemical Society

Scheme 1. Retrosynthetic Analysis of (þ)-Crassalactone D (1)
Figure 1. Chemical structures of goniobutenolides A (1) and B
(2), and of (þ)-crassalactone D (3).
As part of our continuing efforts,6 in the preparation
of naturally occurring styryl-lactones and their analogs
as potential antitumor agents from monosaccharides, we
now disclose novel total syntheses of goniobutenolides
A (1) and B (2), (þ)-crassalactone D (3), and the related
C-7 epimers, along with their effects on the proliferation of
some human malignant cell lines.
The retrosynthetic analysis of (þ)-crassalactone D (3)
is shown in Scheme 1. We envisaged that the spiroketal
center in 3 could be established by spirocyclization of 1
through a 5-endo-trig ring closure process, under the
conditions similar to those reported in the literature.7 It
was further envisioned that goniobutenolides A (1) and
B (2) could be synthesized from 9, via 3b, by a one-pot
cascade comprised of Z-selective olefination, followed by
lactonization and E2 elimination. It is therefore essential
that the key intermediate 3b and its precursor 9 contain a
good leaving group at C-3 and a convenient protective
group at O-5. Cyclic carbonate was chosen because it can
act as a protecting group for the diol functionality,8 but at
the same time, it can undergo ring-opening reactions, due
to its good leaving group properties.9
Scheme 2. Preparation of Intermediates 9 and 10
Lactol of type 9 is visualized from D-glucose, via 5, by
well established chemical reactions.6e,10
At the outset, we focused on the synthesis of the key
building blocks 9 and 10 starting from commercially
available diacetone-D-glucose (4, Scheme 2).
Treatment of 5 with 1,10-carbonyldiimidazole in boiling
toluene provided the cyclic carbonate 7 in 95% yield.
Under similar reaction conditions stereoisomer 6 gave 8
(76%). Hydrolytic removal of the isopropylidene protec-
tive groups in both 7 and 8 afforded the corresponding
lactols 9 and 10 in almost quantitative yields. Both pro-
ducts, and particularly stereoisomer 9, are rather hygro-
scopic. They should be therefore used in the next synthetic
step immediately after its brief isolation.
It was expected that lactol 9 could directly give natural
product 1 (and/or its E-isomer 2) under Z-selective Wittig
olefination conditions,11 presuming a subsequent two-step
cascade comprised of γ-lactonization of the intermediate
conjugated ester followed by concomitant E2 elimination
of cyclic carbonate functionality (Scheme 1). In our first
experiment, lactol 9 was submitted to the reaction with a
stabilized ylide (reagent A, Table 1) in dry methanol, to
afford a 2:1 mixture of 1 and 2 in 30% combined yield.
Next, the olefination step was examined in the presence of
(5) (a) Pavlakos, E.; Georgiou, T.; Tofi, M.; Montagnon, T.;
Vassilikogiannakis, G. Org. Lett. 2009, 11, 4556. (b) Yang, Z.; Tang,
P.; Gauuan, J. F.; Molino, B. F. J. Org. Chem. 2009, 74, 9546.
ꢁ
ꢁ
(6) (a) Popsavin, V.; Benedekovic, G.; Sreco, B.; Popsavin, M.;
ꢁ
ꢁ
Francuz, J.; Kojic, V.; Bogdanovic, G. Org. Lett. 2007, 9, 4235. (b)
ꢁ
ꢁ
Popsavin, V.; Benedekovic, G.; Sreco, B.; Popsavin, M.; Francuz, J.;
ꢁ
ꢁ
Kojic, V.; Bogdanovic, G. Bioorg. Med. Chem. Lett. 2008, 18, 5178. (c)
ꢁ
ꢁ
Popsavin, V.; Benedekovic, G.; Sreco, B.; Francuz, J.; Popsavin, M.;
ꢁ
ꢁ
ꢁ
Kojic, V.; Bogdanovic, G.; Divjakovic, V. Tetrahedron 2009, 65, 10596.
ꢁ
ꢁ
(d) Popsavin, V.; Sreco, B.; Benedekovic, G.; Francuz, J.; Popsavin, M.;
ꢁ
ꢁ
Kojic, V.; Bogdanovic, G. Eur. J. Med. Chem. 2010, 45, 2876. (e)
ꢁ
ꢁ
ꢁ
Popsavin, V.; Benedekovic, G.; Popsavin, M.; Kojic, V.; Bogdanovic,
ꢁ
G. Tetrahedron Lett. 2010, 51, 3426. (f) Francuz, J.; Sreco, B.; Popsavin,
ꢁ
ꢁ
ꢁ
ꢁ
M.; Benedekovic, G.; Divjakovic, V.; Kojic, V.; Bogdanovic, G.; Kapor,
A.; Popsavin, V. Tetrahedron Lett. 2012, 53, 1819.
(7) (a) Perron, F.; Albizati, K. F. Chem. Rev. 1989, 89, 1617. (b)
Knight, D. W.; Shaw, D. E.; Staples, E. R. Eur. J. Org. Chem. 2004,
1973.
(8) Parrish, J. P.; Salvatore, R. N.; Jung, K. W. Tetrahedron 2000, 56,
8207.
(9) Verhelst, S. H. L.; Wennekes, T.; vander Marel, G. A.; Overkleeft,
H. S.; van Boeckel, C. A. A.; van Boom, J. H. Tetrahedron 2004, 60,
2813.
(11) For examples of Z-selective Wittig olefinations, see: (a) Shing,
T. K. M.; Tsui, H.-Ch; Zhou, Z.-H. J. Org. Chem. 1995, 60, 3121. (b)
Ramirez, E.; Sanchez, Meza-Leon, M. R. L.; Quintero, L.; Sartillo-Piscil,
ꢁ
ꢁ
ꢁ
(10) Gracza, T.; Szolcsanyi, P. Molecules 2000, 5, 1386.
F. Tetrahedron Lett. 2010, 51, 2178.
Org. Lett., Vol. 14, No. 23, 2012
5957

Table 1. One-Pot Conversion of Lactols 9 and 10 into the Unsaturated Lactones 1 (or 11) and 2 (or 12)
a Isolated yields. b ir ꢀ Z/E isomeric ratio.
Still’s reagent12 (reagent B, entry 2) derived from unstable
ylide using the Z-selective HornerꢀWadsworthꢀEmmons
(HWE) reaction protocol.13 A mixture of isomeric natural
products 1 and 2 (Z/E ca. 2:1) was obtained in 45% yield.
Very careful chromatographic separation afforded pure 1
and 2. Our optical rotation values were different from the
values reported for synthetic2 goniobutenolides A (1) and
B (2) but were in reasonable agreement with the reported
data of isolated 1 and 2.1 However, the NMR data of
synthesized samples 1 and 2 were in full agreement with
those reported in the literature.1ꢀ3,14
In an attempt to generalize this methodology by using a
similar substrate, but with opposite stereochemistry at the
C-5 position, we treated lactol 10 under the same Wittig
olefination conditions. Gratifyingly, within 140 min, TLC
indicated full conversion of the starting material into un-
natural products 11 and 12, which were isolated in 74%
combined yield. Unfortunately, a drop in selectivity was
observed, giving thus an almost equimolar mixture of 11 and
12 (entry 3). Treatment of 10 with Still’s reagent (reagent B)
in the presence of NaH gave even a lower yield of 11 and 12,
and the lack of stereoselectivity was again observed (entry 4).
Apart from the main products 1 and 2, minor amounts
of (þ)-crassalacone D (3) and the corresponding 4-epimer
13 were isolated from both Wittig and HWE reactions of 9.
These side products were presumably formed by a 5-endo-
trig spirocyclization of the unsaturated lactones 1 and 2
after nucleophilic attack of the hydroxyl group from C-7 at
the C-4 position. Accordingly it was assumed that prolong-
ing the reaction time of the Wittig reaction would afford the
required spiro-lactones as the major reaction product.
Indeed, when the lactol 9was treated with 2-(triphenylphos-
phoranylidene)-acetic acid methyl ester (dry MeOH, 0 °C
for 0.5 h, then rt for 47 h), a 1:2 mixture of spiro-lactones
3 and 13 was obtained in 51% yield (Scheme 3). A small
amount of 1 and 2 was isolated in 10% combined yield.
Under the same reaction conditions stereoisomer 10
gave a mixture of spiro-lactones 14 and 15 in 67% com-
bined yield, along with their synthetic precursors 11 and
12, which were isolated in 19% combined yield. The
products 3 and 13, as well as 14 and 15, were separated
with difficulties. Typically, flash column chromatography
was required, followed by preparative TLC, in order to
obtain pure products.14
Interestingly, it was observed that the less stable isomer
13 could be converted to the more stable 3, after treatment
of their mixture (2:1 in favor of 13) with a solution of
trifluoroacetic acid in dichloromethane, to give a 2:1
(12) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405.
(13) For examples of Z-selective HWE olefinations, see: (a) Pihko,
P. M.; Salo, T. M. Tetrahedron Lett. 2003, 44, 4361. (b) Touchard, F. P.
Tetrahedron Lett. 2004, 45, 5519.
1
mixture of 3 and 13 by H NMR analysis. In contrast,
(14) See the Supporting Information for details.
5958
Org. Lett., Vol. 14, No. 23, 2012

Table 2. In Vitro Cytotoxicity of the Natural Products (1, 2, and 3)
and the Corresponding Analogs (11ꢀ15)
Scheme 3. One-Pot Conversion of Lactol 9 to the Spiro-lactones
3 and 13a,b and the Conversion of 10 to 14 and 15a
IC50, μMa
compd
K562
HL-60
Jurkat
Raji
HeLa
MRC-5
1
0.97
46.17
2.01
30.27
0.97
1.02
5.25
1.06
0.36
8.79
2.11
6.15
54.31
32.04
36.65
8.64
7.68
9.99
6.87
4.09
9.48
>100
>100
>100
>100
>100
>100
>100
>100
0.12
2
51.69
41.21
32.54
2.64
29.64
81.02
25.45
0.97
1.67
1.23
2.03
0.39
29.03
1.24
11
12
3
8.85
1.69
13
14
15
DOX
5.97
11.32
48.40
10.87
1.17
10.51
3.54
4.62
a IC50 is the concentration of compound required to inhibit cell
growth by 50% compared to an untreated control. The values are means
of three independent experiments done in quadruplicates. Coefficients of
variation were <10%.
a Isolated yields: 51% (isomeric ratio 3/13 ∼1:2), 67% (isomeric ratio 14/15
∼1:1). b Apart from 3 and 13, a mixture of 1 and 2 was isolated in 10% yield.
c Apart from 14 and 15, a mixture of 11 and 12 was isolated in 19% yield.
cells, in contrast to the commercial antitumor agent doxo-
rubicin (DOX) that exhibited potent cytotoxic activity
against this cell line.
In conclusion, we have developed a new divergent
synthesis of goniobutenolide A (1), goniobutenolide B
(2), crassalactone D (3), 4-epi-crassalactone D (4), and
the corresponding 7-epimers starting from D-glucose. The
key step in the synthesis of butenolides 1 and 2 involved a
new one-pot sequence comprised of an initial Z-selective
Wittig olefination, followed by successive lactonization
and β-elimination. Preparation of spiro-lactones 3 and 4
required an additional step, which included final 5-endo-
trig spirocyclization of 1 or 2. This work provided clues
regarding the possible biosynthetic pathway of natural
product 3, which is probably formed from the unsaturated
lactone 1 or 2 by a similar 5-endo-trig spirocyclization. The
synthesized natural products and analogs were evaluated
for their in vitro antiproliferative activity against a panel of
human tumor cell lines. The obtained biological data revealed
that natural products 1and 3as well as synthetic analogues 13
and 15 exhibited potent cytotoxic activities against all the
tumor cells under evaluation but were devoid of any signifi-
cant cytotoxicity against the normal MRC-5 cells. However,
the styryl-lactones 2, 11, 12, and 14 exhibited diverse anti-
proliferative activities. Based upon these results, we believe
that these molecules may serve as convenient leads in the
synthesis of more potent and selective antitumor agents.
when pure 15 was treated under similar reaction conditions
a 1:1 mixture of 14 and 15 was obtained indicating a
similarity in stability of both steroisomers.
A side-by-side comparison of the antiproliferative activ-
ity of the synthesized styryl lactones (1ꢀ3 and 11ꢀ15)
against a range of human tumor cell lines is presented in
Table 2. The use of human fetal lung fibroblasts (MRC-5)
serves to demonstrate the toxicity of selected lactones
toward normal cells. Cytotoxic activity was evaluated by
using the standard MTT assay after exposure of cells to the
tested compounds for 48 h. The commercial antitumor
agent doxorubicin (DOX) was used as a reference com-
pound in this assay.
As shown in Table 2, natural products 1 and 3 as well as
synthetic analogues 13 and 15 exhibited potent cytotoxic
activities against all the tumor cells under evaluation with
IC50 values in the range of 0.97 to 8.64 μM. The most
potent antiproliferative activity of lactones 3 and 15 were
recordedin the HL-60 cells, being notablymoreactivethan
the commercial cytostatic doxorubicin (DOX). Natural
product 2 demonstrated moderate to weak antiprolifera-
tive effects against all the malignant cells under evalua-
tion with IC50 values in the range of 29.03 to 54.31 μM.
A similar cytotoxicity profile was observed for synthetic
analogue 12 (IC50 values in the range of 8.85 to 36.55 μM).
Unnatural lactone 11 exhibited the most potent growth-
inhibitory activity against HeLa (1.24 μM) and K562
(2.01 μM) cell lines but showed a moderate to weak anti-
proliferative activity toward HL-60, Jurkat, and Raji
malignant cells. Compounds 3 and 11 inhibited the growth
of HeLa cells, being as active as DOX in the same cell
line. The most potent antiproliferative activity of the
unnatural lactone 14 was recorded in the Jurkat cell line
(IC50 1.23 μM), although this molecule exhibited 4-fold
lower activity than DOX. Moreover, none of the synthe-
sized styryl lactones exhibited toxicity toward MRC-5
Acknowledgment. Financial support of the Ministry of
Education, Science and Technological Development of the
Republic of Serbia is acknowledged (Grant No. 172006).
Supporting Information Available. Experimental pro-
cedures, characterization data for all compounds, and
1
copies of H, 13C NMR spectra of final products. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 23, 2012
5959