Dichapetalins from dichapetalum species and their cytotoxic properties

Article abstract of DOI:10.1016/j.phytochem.2013.03.023

Six dichapetalins named dichapetalins N-S were isolated from Dichapetalum mombuttense, Dichapetalum zenkeri and Dichapetalum leucosia. They were accompanied in the same plants by the known dichapetalins A, B, C, I, L and M. The structures of the compounds were elucidated by 1D and 2D NMR experiments and mass spectrometry. They all possessed the dammarane skeleton substituted at position C-3 by a C₆-C₂ unit forming a 2-phenylpyran moiety. All contained a lactone ring in the side chain except dichapetalins O, Qand R, in which this ring was replaced by a lactol. Dichapetalin Qand R were also the first dichapetalins bearing a tertiary methyl and a double bond instead of the cyclopropane of the dammaranes. All these compounds were assayed against cancer cell lines HCT116 and WM 266-4 and displayed cytotoxic and anti-proliferative activities in the 10^-6 to 10^-8 M range.

Full text of DOI:10.1016/j.phytochem.2013.03.023

Phytochemistry 94 (2013) 184–191
Contents lists available at SciVerse ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Dichapetalins from Dichapetalum species and their cytotoxic properties
Christophe Long a, Yannick Aussagues a, Nicolas Molinier a, Laurence Marcourt a, Laure Vendier b,
Arnaud Samson a, Valérie Poughon a, Patrick B. Chalo Mutiso c, Frédéric Ausseil a, François Sautel a,
Paola B. Arimondo a, Georges Massiot a,
⇑
a USR CNRS-Pierre Fabre No. 3388 ETaC, Centre de Recherche et Développement Pierre Fabre, 3 avenue Hubert Curien, 31035 Toulouse Cedex 01, France
b UPR CNRS No. 8241, Laboratoire de Chimie de Coordination, 205 route de Narbonne, 31077 Toulouse Cedex 04, France
c School of Biological Sciences, University of Nairobi, P.O. Box 30197, 00100 Nairobi, Kenya
a r t i c l e i n f o
a b s t r a c t
Article history:
Six dichapetalins named dichapetalins N–S were isolated from Dichapetalum mombuttense, Dichapetalum
zenkeri and Dichapetalum leucosia. They were accompanied in the same plants by the known dichapeta-
lins A, B, C, I, L and M. The structures of the compounds were elucidated by 1D and 2D NMR experiments
and mass spectrometry. They all possessed the dammarane skeleton substituted at position C-3 by a C6–
C2 unit forming a 2-phenylpyran moiety. All contained a lactone ring in the side chain except dichapet-
alins O, Q and R, in which this ring was replaced by a lactol. Dichapetalin Q and R were also the first dicha-
petalins bearing a tertiary methyl and a double bond instead of the cyclopropane of the dammaranes. All
these compounds were assayed against cancer cell lines HCT116 and WM 266-4 and displayed cytotoxic
and anti-proliferative activities in the 10–6 to 10–8 M range.
Received 29 November 2012
Received in revised form 27 March 2013
Available online 22 April 2013
Keywords:
Dichapetalum mombuttense
Dichapetalum leucosia
Dichapetalum ruhlandii
Dichapetalum zenkeri
Dichapetalum eickii
Dichapetalaceae
Ó 2013 Elsevier Ltd. All rights reserved.
Dammarane
Cytotoxicity
1. Introduction
cytotoxicity, measured in a brine shrimp bioassay, was evaluated
to an EC50 of 2.0 ꢀ 10ꢁ8 M (Osei-Safo et al., 2008). Finally, five
The dichapetalins constitute a small family of secondary metab-
olites of mixed biosynthesis, composed of a dammarane part and a
C6–C2 unit. Dichapetalin A was the first to be discovered and was
isolated from the roots of Dichapetalum madagascariense (Achen-
bach et al., 1995; Osei-Safo et al., 2012). It presented a very potent
cytotoxic effect against L1210 murine leukemia cells, with an EC90
below 1.7 ꢀ 10ꢁ10 M, but was four orders of magnitude less effi-
cient towards KB carcinoma cells and murine bone marrow cells
stimulated with GM-CSF (granulocyte-macrophage colony-stimu-
lating factor). In 1996, Achenbach’s group established the absolute
configuration of dichapetalin A and identified seven new deriva-
tives, dichapetalins B–H, presenting poor or no cytotoxic activities
(Weckert et al., 1996; Addae-Mensah et al., 1996). In 2006, four
new dichapetalin I–L were extracted from the stem bark of Dicha-
petalum gelonioides; these derivatives showed poor cytotoxic ef-
fects on a panel of nine (I, J) and five (K, L) cancer cell lines, with
EC50s superior to 1.7 ꢀ 10ꢁ6 M, except on the SW626 human ovar-
ian adenocarcinoma cell line where EC50s ranged from 3.3 ꢀ 10ꢁ7
to 8.3 ꢀ 10ꢁ7 M for dichapetalin I and J (Fang et al., 2006). Dicha-
petalin M was isolated in 2008 from D. madagascariense and its
new dichapetalin-derived triterpenoids, the acutissimatriterpenes,
were isolated from Phyllanthus acutissima. These molecules were
tested against a panel of six cancer cell lines and one of them (acu-
tissimatriterpene E), presented a strong and selective effect on the
P388 murine leukemia cell line (EC50 of 5.0 ꢀ 10ꢁ9 M) (Tuchinda
et al., 2008).
Aiming at finding new cytotoxic natural products, we investi-
gated a series of plants from the genus Dichapetalum, namely, Dich-
apetalum mombuttense Engl., Dichapetalum leucosia Engl.,
Dichapetalum ruhlandii Engl., Dichapetalum zenkeri Engl. and Dicha-
petalum eickii Ruhland. We focused on isolating dichapetalins,
which give blue TLC spots upon vanillin/H2SO4 spray. Herein, we
report the identification of six new compounds and their cytotoxic
properties. Table 1 reports a summary of the isolation of the differ-
ent dichapetalins in the genus Dichapetalum.
2. Results
2.1. Isolation and structural elucidation
All species but D. leucosia, originating from Madagascar,
contained dichapetalin A (1), which was also the sole ‘‘blue’’
compound isolated from D. ruhlandii. It was identified by its spec-
⇑
Corresponding author. Tel.: +33 5 34506517.
E-mail address: georges.massiot.externe@pierre-fabre.com (G. Massiot).
0031-9422/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.phytochem.2013.03.023

C. Long et al. / Phytochemistry 94 (2013) 184–191
185
Table 1
Occurrence of the dichapetalins in Dichapetalum species.
Dichapetalin
Dichapetalum
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
p
p
p
p
p
p
p
p
p
p
p
p
madagascariense
gelonoïdes
ruhlandii
mombuttense
leucosia
p
p
p
p
p
p
p
p
p
p
p
p
p
p
p
p
p
p
zenkiri
eickii
p
compound was missing the C-11/C-12 double bond and the lactone
carbonyl (no signal after 145 ppm in the 13C NMR spectrum). A
close examination of the 1H and 13C NMR spectra showed that most
signals came as pairs of almost equal intensity, suggesting that 3
was a mixture. The observation of carbons at d 102.1 and 98.2
was in favor of a hemiketal instead of the lactone. All attempts to
separate the isomers were unsuccessful and led to recovery of the
mixture. Signal overlap precluded configuration assignment using
interproton couplings, but the high similarities in 13C chemical
shifts showed that 1 and 3 shared the same backbone configuration.
Furthermore, when 1 was partially reduced to hemiketal 7, also a
mixture, the NMR spectra closely resembled those of 3. The reduc-
tion was performed with a slight excess of Dibal and tetrol 9 was
obtained as a by-product. Comparison of the NMR spectra of 3
and 7 indicated that the natural product 3 had same configuration
of dichapetalin A.
Dichapetalin P (4) had a C40H48O8 composition (C40H48O8Na
meas. 679.3232, calc. 679.3241). Its NMR spectra strongly resem-
bled those of dichapetalin M and particularly we observed signals
for a ketone at d 213.6 and for two isolated methylenes, as those
present in the side chain of M. Since the composition of the two
compounds differed only by an oxygen atom, it remained to deter-
mine which oxygen atom was missing in 4. Comparison of the 1H,
13C and HSQC experiments showed that C-6 was now a methylene
whose protons coupled with the C-7 ketone. Dichapetalin P was
thus 6-deoxy dichapetalin M.
tral properties, which matched those published in the literature
(Achenbach et al., 1995). Nevertheless, a severe discrepancy was
20
observed in the rotation values between our samples ([a]
D
20
ꢁ2.4 (c 0.9, CHCl3)) and the original report ([
a]
35 (c 1.5,
D
CHCl3)). To elucidate this point, a monocrystal suitable for X-ray
diffraction was grown in absolute EtOH and the structure solved
using direct methods.1 It was found in all respects identical to the
proposed one in the literature (Achenbach et al., 1995). The rotation
value was measured three times on different instruments and com-
pound 1 was always found levorotatory in contrast to what was pre-
viously described.
In addition to dichapetalin A, D. eickii from Kenya contained
dichapetalin M, previously isolated from D. madagascariensis
(Osei-Safo et al., 2008).
D. mombuttense collected in RDC (Democratic Republic of Con-
go) provided the known dichapetalins A and L (Fang et al., 2006),
the abietic acid derivative pyracrenic acid (Otsuka et al., 1981)
and a novel compound named dichapetalin N (2). The mass spec-
trum of 2 showed a molecular ion at m/z 582 corresponding to a
C
38H46O5 formula. The differences in the 1H and 13C NMR spectra
of compounds 1 and 2 were minor except for the signals belonging
to the side chain. The presence of signals for an aldehyde and the
absence of signals for CH2OH-26 led us to propose structure 2, in
which the terminal primary alcohol is replaced by an aldehyde. A
final structural proof was obtained by chemical transformation of
1 into 2 by MnO2 oxidation.
O
R
O
X
O
O
H
OCOCH3
H
22
12
19
H
H
11
H
OH
18
30
28
H
H
β
7
OH
O
H
H
α
R
29
O
O
1 X = O, R = CH2OH: dichapetalinA
4
dichapetalin P
R = H,
dichapetalin N
2 X = O, R = CHO:
R = OH dichapetalin M
11,12-dihydro: dichapetalin O
3 X = H, OH, R = CH2OH,
7 X = H, OH, R = CH2OH
D. zenkeri from Kenya yielded dichapetalin A and its 22-hydroxy
derivative dichapetalin B as major compounds. They were accompa-
nied by dichapetalin L and by four other compounds named here
dichapetalins O, P, Q and R (3–6); compounds 3, 5 and 6 appeared
as unseparable mixtures of isomers. Dichapetalin O (3) displayed
a molecular ion at m/z 611, which analyzed for C38H52O5Na (meas.
611.3717, calc. 611.3707), i.e., 4 amu more than the reference com-
pound 1. The methyl count was identical in 1 and 3 but this latter
OH
OH
HO
O
OH
OH
OH
O
OH
OH
1
The crystal structure has been deposited at the Cambridge Crystallographic Data
Fig. 1. HMBC correlations in the side chain of compound 6 (hydrogen atoms have
Centre and allocated the deposition number CCDC 910128.
been omitted for sake of clarity) and tentative stereochemical assignments.

186
C. Long et al. / Phytochemistry 94 (2013) 184–191
O
H
Dichapetalin Q (5) was an isomer of dichapetalin O, with a C38H52O5
HO
OH
analysis (meas. 611.3719, calc. 611.3707 for C38H52O5Na). Its NMR
spectra also displayed the typical signals for a 1:1 mixture of hemi-
ketals presenting all the molecule backbone signals except those
belonging to the C/D ring junction. Noteworthy was the absence
of the protons of the methylene bridge of the cyclopropane, which
appear at high field in all the dichapetalins, either directly observa-
ble in the 1H spectrum or indirectly in the HSQC experiment in case
of severe overlap. The missing methylene was replaced by a methyl
on a quaternary carbon atom and a trisubstituted double bond was
present between C-14 and C-15. In the dichapetalin series, cyclo-
propane ring opening was previously observed under acidic condi-
tions (Tuchinda et al., 2008) and this was used to secure
configurations in the side chain. Despite the fact that no acid was
used during the extraction or purification stages, it cannot be totally
ruled out that 5 derived from 3 and therefore had the same config-
urations, identical to the one determined for dichapetalins A and L
(Fang et al., 2006).
H
H
13
14
15
H
OH
H
O
5 dichapetalin Q
HO
HO
OH
H
H
12
7
H
11
Dichapetalin R (6) was the most oxygenated compound in the
series with a C38H54O7 composition (C38H54O7Na meas. 645.3753,
calc. 645.3762). It was also a mixture of hemiketals in an approx-
imate 80:20 ratio. The 13C NMR spectrum showed that it had the
same pentacyclic nucleus of compound 5 with a characteristic C-
14 at d 162. Signal assignment was done in a straightforward fash-
ion up to C-21, which appeared at d 97.3. A methyl (d 22.1), two
methylenes (d 31.2 and 67.6), two methines (d 76.1 and 77.8)
and a quaternary carbon (d 74.6) remained to be located. Fig. 1
reports some of the HMBC correlations, which allowed determina-
tion of the side chain. As far as relative configurations are
H
OH
H
O
9
Table 2
13C NMR of compounds 1–9 (CDCl3, except for 7 and 9 measured in CD3OD).
1
2
3
3a
4
5
5a
6
7
7a
8
9
C
1
40.0
40.1
40.2
40.3
40.1
39.5
39.2
39.5
41.4
41.4
40.0
41.4
2
3
4
5
6
7
8
9
117.8
140.0
38.3
43.7
24.1
72.3
36.4
45.7
36.2
124.1
128.9
30.0
35.1
24.9
22.7
40.9
17.4
18.1
42.1
178.3
31.3
75.1
122.0
141.7
67.2
14.2
23.8
72.5
14.9
81.8
40.7
142.6
125.8
128.4
127.5
117.7
140.2
38.3
43.8
24.2
72.3
36.3
45.8
36.4
124.4
128.7
30.1
35.4
25.0
23.0
41.1
17.5
18.2
41.7
177.2
30.5
74.4
147.7
140.7
193.7
9.8
118.3
139.7
38.5
43.9
24.0
74.2
38.7
42.4
36.8
16.8
25.8
28.5
37.3
26.1
26.5
44.9
19.7
16.8
50.2
98.1
37.6
75.4
127.7
138.2
68.0
14.0
24.1
72.8
13.9
82.0
40.9
142.8
126.0
128.6
127.7
118.3
139.7
38.6
43.9
24.0
74.1
38.8
42.6
36.8
16.9
25.8
28.9
37.3
26.2
27.6
48.5
19.8
16.9
51.6
102.0
34.8
74.0
124.8
139.5
70.2
14.3
24.1
72.8
14.1
82.0
40.9
142.8
126.0
128.6
127.7
117.8
139.9
39.2
53.4
35.9
213.6
48.4
52.5
36.4
120.6
131.5
31.5
36.2
27.1
23.7
40.6
17.2
17.9
47.4
174.4
72.1
111.5
45.9
85.3
78.7
22.2
23.9
72.2
14.5
82.1
40.6
118.4
139.7
38.4
44.6
23.4
72.0
44.3
40.3
37.2
16.3
32.8
47.2
162.4
119.4
34.9
58.0
27.3
16.2
48.5
102.4
39.2
75.3
127.8
139.8
68.0
14.0
24.1
72.9
19.7
82.0
40.9
142.8
126.0
128.6
127.7
118.4
139.7
38.4
44.6
23.4
72.0
44.3
40.3
37.2
16.3
33.2
46.9
162.1
119.9
35.3
52.9
27.3
16.2
44.6
97.4
35.4
73.9
124.6
138.4
68.0
14.3
24.1
72.9
20.2
82.0
40.9
142.8
126.0
128.6
127.7
118.4
139.8
38.4
44.7
23.4
72.0
44.3
40.3
119.5
141.3
39.7
45.2
26.3
73.5
39.2
47.1
37.5
123.3
132.6
32.7
37.7
25.8
28.2
46.0
18.6
18.8
52.0
99.1
36.1
75.3
126.2
140.1
68.1
13.9
24.4
73.8
16.7
83.5
42.0
144.1
127.1
129.5
128.8
119.5
141.3
39.7
45.2
26.4
73.6
39.2
47.1
37.5
123.7
133.4
33.5
37.8
26.1
29.9
43.0
18.6
18.8
52.8
103.3
38.3
76.8
129.3
138.5
68.3
14.2
24.4
73.8
17.3
83.5
42.0
144.1
127.1
129.5
128.8
117.8
140.0
38.3
43.7
24.1
72.3
36.2
45.7
36.4
124.0
128.9
30.0
35.1
24.9
22.8
40.8
17.4
18.1
41.2
177.7
35.0
114.2
51.3
78.2
80.2
24.7
23.8
72.5
14.9
81.8
40.7
142.6
125.8
128.4
127.5
119.5
141.3
39.6
45.2
26.3
73.5
37.5
47.1
37.5
123.8
132.8
31.5
37.6
26.1
26.6
45.3
18.4
18.8
41.9
65.8
37.6
68.3
129.6
138.4
68.4
14.4
24.4
73.8
17.9
83.5
42.0
144.1
127.1
129.5
128.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
a
16.4
33.2
46.8
162.1
119.9
35.2
53.0
27.6
16.4
45.2
97.4
77.8
31.2
76.2
74.6
67.6
22.1
24.0
72.9
20.2
82.0
40.9
142.8
126.0
128.6
127.7
23.8
72.5
15.1
81.9
40.7
142.6
125.8
128.4
127.5
b
10
20,60
30,50
40
142.4
126.0
128.5
127.9
170.7
22.0
OCOCH3
OCOCH3

C. Long et al. / Phytochemistry 94 (2013) 184–191
187
Table 3
Dichapetalin A chemosensitivity on 16 cancer cell lines.
O
Cell line
Tumor types
Dichapetalin A
EC50 (M)
O
H
O
HCT-116
NAMALWA
SKOV-3
HOP-62
A549
NCI-H460
T47D
BxPC3
KM-12
OVcAR-3
HL-60
Colo-205
TK10
MDAMB-231
DU145
Human colorectal carcinoma
Human Burkitt’s lymphoma
Human ovarian adenocarcinoma
Human lung cancer
Human lung adenocarcinoma
Human prostate carcinoma
Human breast ductal carcinoma
Human pancreatic adenocarcinoma
Human colon carcinoma
Human ovarian adenocarcinoma
Human acute promyelocytic leukemia
Human colorectal adenocarcinoma
Human renal adenocarcinoma
Human breast adenocarcinoma
Human prostate carcinoma
Human melanoma
2.5 ꢀ 10ꢁ7
3.0 ꢀ 10ꢁ7
8.6 ꢀ 10ꢁ7
9.1 ꢀ 10ꢁ7
1.2 ꢀ 10ꢁ6
1.4 ꢀ 10ꢁ6
2.3 ꢀ 10ꢁ6
3.0 ꢀ 10ꢁ6
8.7 ꢀ 10ꢁ6
1.0 ꢀ 10ꢁ5
1.1 ꢀ 10ꢁ5
1.2 ꢀ 10ꢁ5
1.2 ꢀ 10ꢁ5
1.3 ꢀ 10ꢁ5
1.3 ꢀ 10ꢁ5
1.7 ꢀ 10ꢁ5
OH
H
H
OH
H
O
8 dichapetalin S
WM 266-4
Table 4
D. leucosia was a source of dichapetalins C and I, previously isolated
from D. madagascariensis and D. genoloides, of the new dichapetalin
Q (5) and of a new one named here dichapetalin S (8). This com-
pound had a C38H46O6 composition according to HRMS (C38H46O6Na
meas. 679.3232, calc. 679.3241). The hexacyclic part of the mole-
cule was very similar to the one of dichapetalin A and the 1H
NMR spectrum showed the typical signals of the pyran ring, of
the two double bonds, of the cyclopropane and of hydroxylation
at C-7. The originality therefore was in the side chain, which
showed 13C signals for a lactone and for a quaternary carbon atom
at d 114.2. This latter signal was reminiscent of C-23 in the spirok-
etal moiety of compound 4 and of dichapetalin M (Osei-Safo et al.,
2008). It differed however by the absence of an alcohol on C-22,
and the tertiary acetate on C-25 was replaced by an alcohol. Paucity
of material did not allow full structural elucidation in the side chain.
Dichapetalins chemosensitivity on HCT116 and WM 266-4 cell lines reported as the
concentration at which 50% of cell proliferation is inhibited.
Dichapetalin derivatives
HCT116
EC50 (M)
WM 266-4
EC50 (M)
A (1)
B
C
I
L
M
N (2)
O (3)
P (4)
Q (5)
R (6)
S (8)
7
9
2.5 ꢀ 10ꢁ7
8.1 ꢀ 10ꢁ8
5.0 ꢀ 10ꢁ7
2.8 ꢀ 10ꢁ7
6.8 ꢀ 10ꢁ7
9.9 ꢀ 10ꢁ9
9.2 ꢀ 10ꢁ8
8.9 ꢀ 10ꢁ7
5.8 ꢀ 10ꢁ8
2.5 ꢀ 10ꢁ6
4.2 ꢀ 10ꢁ6
4.5 ꢀ 10ꢁ7
4.7 ꢀ 10ꢁ7
3.0 ꢀ 10ꢁ6
3.4 ꢀ 10ꢁ8
3.5 ꢀ 10ꢁ9
1.7 ꢀ 10ꢁ5
3.4 ꢀ 10ꢁ7
2.5 ꢀ 10ꢁ6
1.0 ꢀ 10ꢁ5
3.1 ꢀ 10ꢁ6
7.8 ꢀ 10ꢁ8
1.5 ꢀ 10ꢁ6
8.4 ꢀ 10ꢁ6
2.3 ꢀ 10ꢁ7
2.7 ꢀ 10ꢁ5
3.1 ꢀ 10ꢁ5
1.3 ꢀ 10ꢁ6
7.1 ꢀ 10ꢁ6
1.8 ꢀ 10ꢁ5
4.1 ꢀ 10ꢁ8
4.9 ꢀ 10ꢁ9
2.2. Biological activities of isolated compounds
Doxorubicin
Camptothecin
Before evaluating the biological activities of the new dichapet-
alin derivatives, we profiled dichapetalin A cytotoxicity on a set
of 16 cancer cell lines (Table 3). HCT116 (human colorectal carci-
noma) was found to be the most sensitive to dichapetalin A
(EC50 = 2.5 ꢀ 10ꢁ7 M) and WM 266-4 (human melanoma) the
most resistant (EC50 = 1.7 ꢀ 10ꢁ5 M): HCT116 chemosensitivity
being 68-fold higher than the WM 266-4 one. Because of this sig-
nificant difference, we used the HCT116 and WM 266-4 cell line
couple to analyze the cytotoxic and anti-proliferative properties
of dichapetalins B, C, I, L, M, O, P, Q, R and S as well as of two hemi-
synthetic derivatives (Table 4). All new derivatives were found ac-
concerned, interproton coupling constants measurement (or line-
width) indicated that H-21 and -24 were equatorial and H-22 axial.
Even if it was not possible to assign with certainty the configura-
tion of C-25, we assumed it to be 25R as in all dichapetalins studied
so far.
OH
25
O
tive on HCT116 cancer cells and new dichapetalin
P
24
HO
OH
(EC50 = 4.8 ꢀ 10ꢁ8 M) was 4-fold more active than dichapetalin A
(EC50 = 2.5 ꢀ 10ꢁ7 M). Dichapetalin P was also more potent than
dichapetalin A on WM 266-4 cells (EC50 = 2.3 ꢀ 10ꢁ7 M). Interest-
ingly, derivative P contains the intact lactone ring, a constraint
moiety in the side chain and a carbonyl group in position 7. Its
21
20
H
22
H
OH
17
13
14
6a-hydroxy derivative, dichapetalin M, is the most potent cyto-
toxic agent measuring an EC50 of 9.9 ꢀ 10ꢁ9 and 7.8 ꢀ 10ꢁ8 M in
HCT116 and WM 266-4 cell lines, respectively. The lactone is
clearly important for activity, as seen for compounds R, Q, 9, O
and L, as well as the tricyclic motif (R and Q being the least active
on both cell lines).
15
H
OH
H
O
3. Conclusions
6 dichapetalin R
Six dichapetalins are here described for the first time (named N,
O, P, Q, R and S): three of them possess a partially reduced form of

188
C. Long et al. / Phytochemistry 94 (2013) 184–191
the side chain lactone and two the genuine skeleton of the dam-
maranes without the cyclopropane ring.
HPLC was performed on a Merck–Hitachi apparatus equipped with
an L-7200 automated sample injector, a L-7100 pump, a L-7450
diode array detector, a D-7000 interface and EZChrom software.
A prepacked C18 reversed-phase column (Lichrospher 100 RP-18,
In the literature, the biological activity profile of the dichapeta-
lin derivatives is mainly based on cytotoxicity measurements in
various cancer cell lines or brine shrimps assays. Here, we first pro-
filed dichapetalin A on 16 cancer cell lines finding EC50 values in
agreement with previous reports. Achenbach et al. found an EC90
below 1.7 ꢀ 10ꢁ10 M on L1210 murine leukemia cells, whereas
KB carcinoma cells and murine bone marrow cells stimulated with
GM-CSF were found four orders of magnitude less sensitive
(Achenbach et al., 1995). We found an EC50 of 3 ꢀ 10ꢁ7 M in hu-
man Burkitt’s lymphoma (Namalwa) and of 1.1 ꢀ 10ꢁ5 M in human
acute promyelocytic leukemia HL60 cells. Similar micromolar cyto-
toxicity levels were found for dichapetalin A in KB, KB-V+ and KB-
Vꢁ cell lines (Weckert et al., 1996; Addae-Mensah et al., 1996) and
in seven other cancer cell lines, excepted in the SW626 human
ovarian adenocarcinoma cell line 10 times more sensitive with
an EC50 of 3.4 ꢀ 10ꢁ7 M (Fang et al., 2006). In agreement, human
ovarian adenocarcinoma SKOV-3 cells in our study gave an EC50
of 8.6 ꢀ 10ꢁ7 M for dichapetalin A, while human ovarian adenocar-
cinoma OVcAR-3 was 12-fold less sensitive (EC50 1.1 ꢀ 10ꢁ5 M).
We measured a subnanomolar cytotoxicity in human colorectal
carcinoma HCT116 cells (EC50 2.5 ꢀ 10ꢁ7 M), while the human
melanoma WM 266-4 cells was 68-fold more resistant. Therefore,
we used the HCT116 and WM 266-4 cell line couple to evaluate the
cytotoxic and anti-proliferative activities of the other dichapetalin
derivatives. New derivative dichapetalin P, constraint on the side
chain, resulted 4-fold more active than dichapetalin A on HCT116
and 74-fold more active on WM 266-4. Its hydroxy derivative,
dichapetalin M, was the most potent cytotoxic compound. Dicha-
petalin B, initially isolated by Addae-Mensah et al. but never de-
scribed in terms of biological activity, was found nearly as active
as dichapetalin P on the two cell lines. Newly described compound
N was also very potent, while new dichapetalin derivatives O, Q
and R, lacking the lactone, showed a decreased toxicity on the
two cell lines. The hemisynthetic compounds 7 and 9 were tested
o and 7 was found 10-fold more active than 9 on both cell lines,
showing how limited structural modifications can lead to an
important difference in biological activity.
4 ꢀ 125 mm, 5
lm) was used for analytical HPLC with a binary gra-
dient elution (solvent A: H2O and solvent B: MeCN) and a flow rate
of 1 mL minꢁ1. Semipreparative HPLC was performed on an appa-
ratus equipped with a VWR International LaPrep pump P110, a
VWR LaPrep P314 Dual k absorbance detector and EZChrom soft-
ware. A prepacked C18 reversed-phase column (Hibar-Lichrospher
100 RP-18, 25 ꢀ 250 mm, 5
lm) was used for semipreparative
HPLC with a binary gradient elution (solvent A: H2O and solvent
B: MeCN), a flow rate of 30 mL minꢁ1 and the chromatogram was
monitored at 210 and 320 nm.
4.2. Plant material
D. ruhlandii and D. zenkeri were collected near Kaya Muhaka
(Kenya), D. eickeii was collected in the Ngaongao Forest (all in
the Coast Province, Kenya) by one of us (P. B. C. M.) in March and
April, 2010 for the first two species and in May, 2003 for the third
one. They were authentified by comparison with authentic sam-
ples. Specimens have been deposited at the University of Nairobi
Herbarium under respective references 2010/059, 2010/101 and
2003/0152. D. mombuttense was collected near Isiro (RDC) in
November 1987 by René Bellé (Pierre Fabre Research Institute). It
was identified by R. Bellé and an herbarium specimen was depos-
ited under reference RBL-068 in the Pierre Fabre Botanical Conser-
vatory in Cambounet-sur-le-Sor (France). D. leucosia was collected
near Ampasy Naompoana (Madagascar) by Raymond Gérold in
November 2001; an herbarium specimen was deposited under ref-
erence SEAR1122 in the Pierre Fabre Botanical Conservatory.
4.3. Extraction and isolation of compounds
The dried roots of D. mombuttense (100 g) were powdered and
extracted with EtOAc at room temperature for 24 h. After filtration,
the organic solvent was concentrated under reduced pressure. The
crude extract (1.55 g) was further subjected to silica gel CC (70 g,
40 ꢀ 100 mm) using a CH2Cl2–MeOH gradient [1:0 to 8.5:1.5;
15 mL each], to give 60 fractions. All the fractions were analyzed
by TLC on silica gel using the solvent mixture CH2Cl2–MeOH
(95:5) and pooled according to TLC into seven fractions (frs.1–7).
The spots of fraction 2 on the TLC plate were blue, as detected by
heating after spraying with 3% H2SO4 + 1% vanillin. Fraction 2
(749 mg) was purified by repeated semi-preparative HPLC RP-18
chromatography, eluting with a linear gradient (70–100% B) to give
1 (146.7 mg) and an impure fraction (17.0 mg). This fraction was
further purified by extensive preparative TLC on silica gel using
the solvent mixture CH2Cl2–MeOH (98:2) to afford 1 (0.2 mg),
dichapetalin L (3.3 mg) and 2 (1.7 mg).
Finally, our study constitutes the more exhaustive analysis of
dichapetalin molecules based on both cytotoxic activity and
molecular diversity. Dichapetalins M and P remained the most po-
tent representatives of this class of dammarane-type triterpenoids.
The structure activity relationships determined here inspires the
road to simplify the compounds and keeping the pharmacological
activity.
4. Experimental section
4.1. General experimental procedures
Optical rotations were determined in CHCl3 solution on a Per-
kin-Elmer 341 automatic polarimeter. UV spectra were obtained
in MeOH using a UV MC2 Safas spectrophotometer. A FT-IR Bruker
Tensor 27 spectrophotometer was used for scanning IR spectros-
copy. The NMR spectra were recorded on a Bruker Avance II spec-
trometer equipped with a 13C cryoprobe at 500 MHz for 1H and
125 MHz for 13C; 2D experiments were performed using standard
Bruker programs. The ESIMS and MS/MS were performed using a
Bruker Esquire-LC ion trap mass spectrometer; the samples were
introduced by infusion in a solution of MeOH. HRESIMS were ob-
tained on a Bruker MicrOTOF. TLC was carried out on precoated sil-
ica gel 60F254 (Merck) with CH2Cl2–MeOH (95:5) and spots were
visualized by heating after spraying with 3% H2SO4 + 1% vanillin.
Column chromatography (CC) was carried out on prepacked
The dried roots of D. leucosia (100 g) were powdered and ex-
tracted with EtOAc at room temperature for 24 h. After filtration,
the organic solvent was concentrated under reduced pressure.
The crude extract (2.22 g) was further subjected to silica gel CC
(70 g, 40 ꢀ 100 mm) using
a CH2Cl2–MeOH gradient [1:0 to
8.5:1.5; 15 mL each], to give 60 fractions. All the fractions were
analyzed by TLC on silica gel using the solvent mixture CH2Cl2–
MeOH (95:5) and pooled according to TLC into height fractions
(frs.1–8). The spots of fraction 4 on the TLC plate were blue, as de-
tected by heating after spraying with 3% H2SO4 + 1% vanillin. Frac-
tion 4 (646 mg) was purified by repeated semi-preparative HPLC
RP-18 chromatography, eluting with a linear gradient (50–80% B)
to give dichapetalin I (6.4 mg), dichapetalin C (8.0 mg), dichapeta-
lin S 8 (12.0 mg) and an impure fraction (10.4 mg). This fraction
was further purified by preparative TLC on silica gel using the sol-
cartridge Kieselgel (40–60 lm) with CH2Cl2–MeOH. Analytical

C. Long et al. / Phytochemistry 94 (2013) 184–191
189
vent mixture CH2Cl2–MeOH (98:2) to afford dichapetalin Q 5
(2.4 mg).
The dried roots of D. ruhlandii (200 g) were powdered and ex-
tracted with EtOAc at room temperature for 24 h. After filtration,
the organic solvent was concentrated under reduced pressure.
The crude extract (2.14 g) was further subjected to silica gel CC
model, excepted for the two hydrogens of the water molecule, which
were located by Fourier differences. Although the studied Dichapet-
alin is an organic compound, it was possible to determine its abso-
lute configuration after recording the data under Cu-K
a instead of
Mo-K radiation. All non-hydrogens atoms were anisotropically re-
a
fined, and in the last cycles of refinement, a weighting scheme was
(70 g, 40 ꢀ 100 mm) using
a CH2Cl2–MeOH gradient [1:0 to
used, where weights were calculated from the following formula:
8.5:1.5; 15 ml each], to give 60 fractions. All the fractions were
analyzed by TLC on silica gel using the solvent mixture CH2Cl2–
MeOH (95:5) and pooled according to TLC into seven fractions
(frs.1–6). The spots of fraction 3 (1.04 g) on the TLC plate were
blue, as detected by heating after spraying with 3% H2SO4 + 1% van-
illin. Fraction 3 (100 mg) was purified by semi-preparative HPLC
RP-18 chromatography, eluting with a linear gradient (60–100%
B) to give dichapetalin A (2.4 mg).
w = 1/[r
2(Fo2)+(aP)2 + bP] where P = (Fo2 + 2Fc2)/3. Drawing of the
molecule was performed with the program ORTEP32 (Farrugia,
1997) with 30% probability displacement ellipsoids for non-hydro-
gen atoms.
4.5. Dichapetalin N (2)
20
Yellow gum, [
a]
D +12 (c 0.33, CHCl3); UV (MeOH) kmax (log
e)
The dried roots of D. eickii (110 g) were powdered and extracted
with EtOAc at room temperature for 24 h. After filtration, the organic
solvent was concentrated under reduced pressure. The crude extract
(360 mg) was further subjected to silica gel CC (30 g, 30 ꢀ 100 mm)
using a CH2Cl2–MeOH gradient [1:0 to 8.5:1.5; 15 mL each], to give
30 fractions. All the fractionswere analyzed by TLC on silica gel using
the solvent mixture CH2Cl2–MeOH (95:5) and pooled according to
TLC into seven fractions (frs.1–8). The spots of fraction 4 (138 mg)
on the TLC plate were blue, as detected by heating after spraying
with 3% H2SO4 + 1% vanillin. Fraction 4 was purified by semi-pre-
parative HPLC RP-18 chromatography, eluting with a linear gradient
(20–100% B) and by repeated preparative TLC on silica gel using the
solvent mixture CH2Cl2–MeOH (98:2) to afford dichapetalin A
(0.8 mg) and dichapetalin M (0.2 mg).
226 (4.0) nm; IR (film) mmax 1770, 1692 cmꢁ1
;
1H NMR (CDCl3,
500 MHz) d 9.48 (1H, s, H-26), 7.39–7.32 (4H, m, H-20, -30, -50, -
60), 6.46 (1H, dq, J = 7, 1 Hz, H-24), 6.16 (1H, dd, J = 10, 3 Hz, H-
12), 5.49 (1H, dd, J = 10, 2.5 Hz, H-11), 5.41 (1H, br d, J = 7 Hz, H-
2), 5.31 (1H, dt, J = 11, 6 Hz, H-23), 4.27 (1H, dd, J = 11.5, 2.5 Hz,
H-a), 3.95 (1H, br t, J = 2.3 Hz, H-7), 3.77 (1H, d, J = 10.5 Hz, H-
29), 3.61 (1H, d, J = 10.5 Hz, H-29), 3.15 (1H, ddd, J = 12.5, 8.5,
5 Hz, H-20), 2.65 (2H, m, H-b/H-17), 2.54 (1H, ddd, J = 14, 8, 6 Hz,
H-22), 2.21 (1H, dd, J = 13, 2.5 Hz, H-b), 1.84 (3H, br d, J = 1 Hz,
CH3-27), 1.33 (3H, s, CH3-28), 1.22 (1H, d, J = 5 Hz, H-30), 1.09
(3H, s, CH3-19), 0.93 (3H, s, CH3-18), 0.78 (1H, d, J = 5 Hz, H-30);
13C NMR, see Table 2; HRESIMS m/z 608.3228 (calcd for C38H46O5-
Na: 608.3237).
The dried roots of D. zenkeri (1 kg) were powdered and ex-
tracted with EtOAc at room temperature for 24 h. After filtration,
the organic solvent was concentrated under reduced pressure.
The crude extract (6.90 g) was further subjected to silica gel CC
(190 g, 40 ꢀ 250 mm) using a CH2Cl2–MeOH gradient [1:0 to
8.5:1.5; 15 mL each], to give 400 fractions. All the fractions were
analyzed by TLC on silica gel using the solvent mixture CH2Cl2–
MeOH (95:5) and pooled according to TLC into seven fractions
(frs.1–7). The spots of fraction 4 (3200 mg) and 5 (755 mg) on
the TLC plate were blue, as detected by heating after spraying with
3% H2SO4 + 1% vanillin. Fraction 4 (400 mg) was purified by semi-
preparative HPLC RP-18 chromatography, eluting with a linear gra-
dient (60–100% B) and by repeated preparative TLC on silica gel
4.6. Dichapetalin O (3)
20
Yellow gum, [
a]
+18 (c 0.1, CHCl3); UV (MeOH) kmax (log e)
D
206 (4.0) nm; IR (film) mmax 3375, 1563 cmꢁ1 1H NMR (CDCl3,
;
500 MHz) d 7.37–7.29 (2 ꢀ 4H, m, H-20, -30, -50, -60), 5.50 (1H, d,
J = 8 Hz, H-24), 5.42 (1H, d, J = 8 Hz, H-24), 5.37–5.33 (4H, m,
2 ꢀ H-21, 2 ꢀ H-12), 4.86 (1H, ddd, J = 13, 9, 5 Hz, H-23), 4.73
(1H, dt, J = 6.5, 9.5 Hz, H-23), 4.24 (2 ꢀ 1H, dd, J = 11.5, 2.5 Hz,
H-
a
), 4.0 (2 ꢀ 2H, d, J = 2 Hz, H-26), 3.79 (2 ꢀ 1H, m, H-7), 3.73
(2 ꢀ 1H, d, J = 10.8 Hz, H-29), 3.56 (2 ꢀ 1H, d, J = 10.8 Hz, H-29),
2.60 (2 ꢀ 1H, t, J = 12.5 Hz, H-b), 1.71 (3H, br s, CH3-27), 1.69
(3H, br s, CH3-27), 1.30 (3H, s, CH3-28), 1.04 (3H, s, CH3-18),
1.03 (3H, s, CH3-18), 1.00 (3H, s, CH3-19), 0.72 (1H, d, J = 5 Hz,
H-30), 0.65 (1H, d, J = 5 Hz, H-30), 0.47 (2 ꢀ 1H, m, H-30); 13C
NMR, see Table 2; HRESIMS m/z 611.3717 (calcd for C38H52O5Na:
611.3707).
using the solvent mixture CH2Cl2–MeOH (97:3) to give
1
(14.8 mg), dichapetalin B (1.4 mg), dichapetalin L (0.7 mg) and 4
(2.8 mg). Fraction 5 (300 mg) was purified by semi-preparative
HPLC RP-18 chromatography, eluting with a linear gradient (60–
100% B) and by repeated preparative TLC on silica gel using the sol-
vent mixture CH2Cl2–MeOH (97:3) to give dichapetalin B (4.8 mg),
3 (1.7 mg), 5 (12.3 mg) and 6 (2.6 mg).
4.7. Dichapetalin P (4)
20
4.4. Dichapetalin A (1) X-ray analysis
Yellow gum, [
a]
D ꢁ29 (c 0.11, CHCl3); UV (MeOH) kmax (log
e)
208 (4.47) nm; IR (film) mmax 3367, 1769, 1749, 1706 cmꢁ1
;
1H
Data were collected at low temperature (180 K) on an Agilent
Gemini diffractometer using a graphite-monochromated Cu-K
NMR (CDCl3, 500 MHz) d 7.37–7.31 (4H, m, H-20, -30, -50, -60), 6.26
(1H, dd, J = 10, 3 Hz, H-12), 5.42 (1H, br d, J = 7 Hz, H-2), 5.39 (1H,
dd, J = 10, 2.5 Hz, H-11), 4.30 (1H, d, J = 10 Hz, H-26), 4.27 (1H,
a
radiation (k = 1.54184 Å) and equipped with an Oxford Instrument
Cooler Device. The final unit cell parameters have been obtained by
means of a least-squares refinement. The structure has been solved
by direct methods using SIR92 (Altomare et al., 1993) and refined
by means of least-squares procedure on F2 with the aid of the pro-
gram SHELXL97 (Sheldrick, 2008) included in the software package
WinGX (Farrugia, 1999). The atomic scattering factors were taken
from the International Tables for X-ray Crystallography2. All hydro-
gens atoms were geometrically placed and refined by using a riding
dd, J = 9, 2.5 Hz, H-a), 4.16 (1H, t, J = 9.7 Hz, H-22), 4.07 (1H, d,
J = 10 Hz, H-26), 3.79 (1H, d, J = 10.5 Hz, H-29), 3.53 (1H, d,
J = 10.5 Hz, H-29), 2.95 (1H, dd, J = 10, 5.5 Hz, H-20), 2.82 (1H, d,
J = 15 Hz, H-24), 2.52 (1H, m, H-17), 2.48 (1H, d, J = 15 Hz, H-24),
2.41 (1H, d, J = 10 Hz, OH), 2.34 (1H, dd, J = 13, 2.5 Hz, H-7), 2.22
(1H, dd, J = 13.5, 2.5 Hz, H-b), 2.00 (3H, s, OAc), 1.76 (1H, dd,
J = 14.7, 2.6 Hz, H-5), 1.67 (3H, s, CH3-27), 1.63 (1H, br d,
J = 15 Hz, H-1), 1.29 (3H, s, CH3-28), 1.26 (3H, s, CH3-19), 1.22
(1H, d, J = 6 Hz, H-30), 1.16 (3H, s, CH3-18), 0.95 (1H, d, J = 6 Hz,
H-30); 13C NMR, see Table 2; HRESIMS m/z 679.3232 (calcd for C40-
H48O8Na: 679.3241).
2
International Tables for X-ray Crystallography, 1974, Vol IV, Kynoch Press,
Birmingham, United Kingdom.

190
C. Long et al. / Phytochemistry 94 (2013) 184–191
4.8. Dichapetalin Q (5)
an hour at room temperature; the organic phase was separated,
dried over Na2SO4 and evaporated. The crude oil (45 mg) was puri-
fied by HPLC to yield 7 (5 mg) and 9 (15 mg) as amorphous solids.
20
Yellow gum, [
a]
0 (c 0.12, CHCl3); UV (MeOH) kmax (log
e
)
D
206 (4.34) nm; IR (film) mmax 3374, 1601 cmꢁ1
;
1H NMR (CDCl3,
Compound 7 (1:1 mixture of isomers): UV (MeOH) kmax (log
e) 206
500 MHz) d 7.40–7.32 (2 ꢀ 4H, m, H-20, -30, -50, -60), 5.52 (1H, dq,
J = 9, 1.3 Hz, H-24), 5.49 (2 ꢀ 1H, m, H-15), 5.44 (1H, dq, J = 8.5,
1.3 Hz, H-24), 5.39 (2 ꢀ 1H, d, J = 7 Hz, H-2), 5.36 (1H, m, H-21),
5.33 (1H, t, J = 3 Hz, H-21), 4.90 (1H, ddd, J = 10.5, 9, 5 Hz, H-23),
4.79 (1H, dt, J = 6.5, 9.5 Hz, H-23), 4.27 (2 ꢀ 1H, dd, J = 11.8,
(4.2) nm; IR (film) mmax 3355, 1647 cmꢁ1 1H NMR (CD3OD,
;
500 MHz) d 7.38–7.23 (2 ꢀ 4H, m, H-20, -30, -50, -60), 6.34 (1H, dd,
J = 10.2, 3.3 Hz, H-12), 6.18 (1H, dd, J = 10.3, 3 Hz, H-13), 5.53
(1H, dq, J = 8,1.3 Hz, H-24), 5.45 (1H, br d, J = 7 Hz, H-2), 5.42 (1H,
dq, J = 8, 1.3 Hz, H-24), 5.39–5.30 (6H, m, 2 ꢀ H-21, 2 ꢀ H-11),
4.88 (1H, m, H-23), 4.76 (1H, ddd, J = 10, 9, 6 Hz, H-23), 4.27
2.5 Hz, H-
a
), 4.03 (2 ꢀ 2H, br s, H-26), 3.96 (2 ꢀ 1H, br s, H-7),
3.78 (2 ꢀ 1H, d, J = 10.5 Hz, H-29), 3.62 (2 ꢀ 1H, d, J = 10.5 Hz, H-
29), 2.63 (2 ꢀ 1H, br t, J = 12 Hz, H-b), 2.58 (1H, br s, OH), 2.40
(1H, m, H-20), 1.74 (3H, br d, J = 1 Hz, CH3-27), 1.73 (3H, br d,
J = 1 Hz, CH3-27), 1.33 (2 ꢀ 3H, s, CH3-28), 1.12 (3H, s, CH3-30),
1.09 (2 ꢀ 3H, s, CH3-18), 1.05 (2 ꢀ 3H, s, CH3-30), 1.04 (2 ꢀ 3H, s,
CH3-19); 13C NMR, see Table 2; HRESIMS m/z 611.3719 (calcd for
C38H52O5Na: 611.3707).
(2 ꢀ 1H, dd, J = 11.5, 2.5 Hz, H- ), 3.93 (2 ꢀ 2H, m, H-26), 3.93
a
(2 ꢀ 1H, m, H-7), 3.73 (2 ꢀ 1H, d, J = 10.8 Hz, H-29), 3.57 (2 ꢀ 1H,
d, J = 10.8 Hz, H-29), 2.62 (2 ꢀ 1H, t, J = 12.5 Hz, H-b), 1.71 (3H, d,
J = 1.2 Hz, CH3-27), 1.69 (3H, d, J = 1.2 Hz, CH3-27), 1.32 (2 ꢀ 3H,
s, CH3-28), 1.11 (2 ꢀ 3H, s, CH3-19), 0.92 (3H, s, CH3-18), 0.91
(3H, s, CH3-18), 0.85 (1H, d, J = 5 Hz, H-30), 0.84 (1H, d, J = 5 Hz,
H-30); 13C NMR, see Table 2; HRESIMS m/z 609.3540 (calcd for C38-
H50O5Na: 609.3550). Compound 9: UV (MeOH) kmax (log
e
) 206
4.9. Dichapetalin R (6)
(4.5) nm; IR (film) mmax 3343, 1646 cmꢁ1 1H NMR (CD3OD,
;
500 MHz) d 7.41–7.26 (4H, m, H-20, -30, -50, -60), 6.18 (1H, dd,
J = 10.2, 3 Hz, H-12), 5.49 (1H, br d, J = 7 Hz, H-2), 5.42–5.38 (2H,
m, H-24, H-11), 4.53 (1H, dt, J = 9, 7 Hz, H-23), 4.30 (1H, dd,
20
Yellow gum, [
a]
ꢁ4 (c 0.1, CHCl3); UV (MeOH) kmax (log
e)
D
208 (4.28) nm; IR (film) mmax 3353, 1575 cmꢁ1 1H NMR (CDCl3,
;
500 MHz) d 7.37–7.29 (4H, m, H-20, -30, -50, -60), 5.47 (1H, br t,
J = 2.2 Hz, H-21), 5.37 (1H, br d, J = 7 Hz, H-2), 5.31 (1H, d,
J = 3.8 Hz, H-21), 4.37 (1H, t, J = 8.7 Hz, H-22), 4.24 (1H, dd,
J = 11.6, 2.6 Hz, H-a), 3.98 (2H, s, H-26), 3.94 (1H, t, J = 2.5 Hz, H-
7), 3.84 (1H, d, J = 10.8 Hz, H-29),3.67 (1H, dd, J = 5.7, 10.7, H-21),
3.60 (2H, m, H-21, H-29), 2.65 (1H, br t, J = 12.5 Hz, H-b), 1.75
(3H, d, J = 1.2 Hz, CH3-27), 1.34 (3H, s, CH3-28), 1.14 (3H, s, CH3-
19), 0.93 (3H, s, CH3-18), 0.90 (1H, d, J = 5.5 Hz, H-30); 13C NMR,
J = 11.8, 2.2 Hz, H-a), 3.94 (1H, br s, H-7), 3.76 (1H, d, J = 11 Hz,
H-29), 3.67 (1H, d, J = 12 Hz, H-26), 3.59 (1H, d, J = 11 Hz, H-29),
3.51 (1H, d, J = 12 Hz, H-26), 3.38 (1H, br s, H-24), 2.60 (1H, br t,
J = 12 Hz, H-b), 1.30 (3H, s, CH3-28), 1.16 (3H, s, CH3-27), 1.73
(3H, br d, J = 1 Hz, CH3-27), 1.12 (3H, s, CH3-30), 1.09 (3H, s, CH3-
18), 1.02 (3H, s, CH3-30), 1.01 (3H, s, CH3-19); 13C NMR, see Table 2;
HRESIMS m/z 645.3753 (calcd for C38H54O7Na: 645.3762).
see Table 2; HRESIMS m/z 611.3705 (calcd for
C38H52O5Na:
611.3707).
4.13. Cell cytotoxicity assay
The cytotoxicity activities of all tested molecules were mea-
sured in vitro using the ATP quantification method as described
elsewhere (Long et al., 2012). Briefly, cells were seeded in 96-well
plates and incubated during 24 h for adhesion in RPMI1640 med-
ium, 10% FBS, 2 mM glutamine, 50 U/mL penicillin/streptomycin
4.10. Dichapetalin S (8)
20
Yellow gum, [
a
]
+50 (c 0.5, CHCl3); UV (MeOH) kmax (log e)
D
212 (4.26) nm; IR (film) mmax 1762 cmꢁ1
;
1H NMR (CDCl3,
500 MHz) d 7.40–7.32 (4H, m, H-20, -30, -50, -60), 6.16 (1H, dd,
J = 10, 3 Hz, H-13), 5.46 (1H, dd, J = 10, 2.5 Hz, H-11), 5.41 (1H, br
and 1.25 lg/mL fungizone. Different concentrations of dichapetalin
derivatives or vehicle were added and cells were incubated for 72 h
at 37 °C in humidified 5% CO2 atmosphere. Cell viability was eval-
uated by determining the level of ATP released by viable cells after
molecule contact. ATP was measured with the ATPlite assay (Per-
kin Elmer) according to the manufacturer conditions. EC50 values
were determined with curve fitting analysis method (non linear
regression model with a sigmoid dose response, variable Hill slope
coefficient) provided by the Prism Software (GraphPad). Results
were expressed as average EC50 values (concentration of tested
compound that inhibits 50% of the maximum effect for the consid-
ered compound).
d, J = 7 Hz, H-2), 4.27 (1H, dd, J = 11.6, 2.7 Hz, H-a), 3.94 (1H, br s,
H-7), 3.91 (1H, d, J = 9.5 Hz, H-26), 3.89 (1H, d, J = 9.5 Hz, H-26),
3.77 (1H, d, J = 11 Hz, H-29), 3.61 (1H, d, J = 11 Hz, H-29), 3.79
(1H, d, J = 10.5 Hz, H-28), 3.53 (1H, d, J = 10.5 Hz, H-28), 3.38 (1H,
ddd, J = 12, 8, 5 Hz, H-20), 2.48 (1H, dd, J = 12, 8 Hz, H-14), 2.47
(1H, d, J = 14 Hz, H-24), 2.34 (1H, d, J = 14 Hz, H-24), 1.50 (3H, s,
CH3-27), 1.33 (3H, s, CH3-28), 1.20 (1H, d, J = 5 Hz, H-30), 1.08
(3H, s, CH3-19), 0.91 (3H, s, CH3-18), 0.76 (1H, d, J = 5 Hz, H-30);
13C NMR, see Table 2; HRESIMS m/z 623.3343 (calcd for C38H48O6-
Na: 623.3343).
4.11. Oxidation of dichapetalin A (1 to 2)
Acknowledgment
Dichapetalin A (8 mg) was dissolved in 1 mL CH2Cl2 and 40 mg
(large excess) of MnO2 was added in one portion. After 24 h of stir-
ring at room temperature, the mixture was filtered on a Celite plug
in a Pasteur pipette and eluted with 10 mg of CH2Cl2. After evapo-
ration, the residue was purified by HPLC, which yielded 3.4 mg
(42%) of pure 2 as a pale waxy solid.
The authors are indebted to Dr Bruno David (IRPF) for the
botanical sample collections.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.phytochem.
2013.03.023. These data include MOL files and InChiKeys of the
most important compounds described in this article.
4.12. DIBAL reduction of dichapetalin A (1 to 7 and 9)
Dichapetalin A (37 mg) was dissolved in a mixture of 2 mL CH2-
Cl2 and 0.2 mL dry THF cooled at 0 °C. A 1 M solution of DIBAL in
hexane (800 lL) was added slowly and the reaction mixture was
References
stirred at 0 °C for three hours. A 1 M solution of Rochelle salt
(10 mL, excess) was then added and the mixture was stirred for
Addae-Mensah, I., Waibel, R., Asunka, S.A., Oppong, I.V., Achenbach, H., 1996. The
dichapetalins – a new class of triterpenoids. Phytochemistry 43, 649–656.

C. Long et al. / Phytochemistry 94 (2013) 184–191
191
Achenbach, H., Asunka, S.A., Waibel, R., Addae-Mensah, I., Oppong, I.V., 1995.
Dichapetalin A, a novel plant constituent from Dichapetalum madagascariensis.
Nat. Prod. Lett. 7, 93–100.
Osei-Safo, D., Chama, M.A., Addae-Mensah, I., Waibel, R., Asomaning, W.A., Oppong,
I.V., 2008. Dichapetalin M. from Dichapetalum madagascariensis. Phytochem.
Lett. 1, 147–150.
Altomare, A., Cascarano, G., Giacovazzo, C., Guagliardi, A., 1993. SIR92 – a program
for crystal structure solution. J. Appl. Crystallogr. 26, 343–350.
Fang, L.Q., Ito, A., Chai, H.-B., Mi, Q., Jones, W.P., Madulid, D.R., Oliveros, M.B., Gao, Q.,
Orjala, J., Farnsworth, N.R., Soejarto, D.D., Cordell, G.A., Swanson, S.M., Pezzuto,
J.M., Kinghorn, A.D., 2006. Cytotoxic constituents from the stem bark of
Dichapetalum gelonioides collected in the Philippines. J. Nat. Prod. 69, 332–337.
Farrugia, L., 1997. ORTEP3 for Windows. J. Appl. Crystallogr. 30, 565.
Farrugia, L., 1999. WINGX – 1.63 Integrated System of Windows Programs for the
Solution, refinement and analysis of single crystal X-ray diffraction data. J. Appl.
Crystallogr. 32, 837–838.
Osei-Safo, D., Chama, M.A., Addae-Mensah, I., Waibel, R., 2012. The dichapetalins –
unique cytotoxic constituents of the Dichapetalaceae in phytochemicals as
nutraceuticals. InTech, pp. 245–254.
Otsuka, H., Fujioka, S., Komiya, T., Goto, M., Hiramatsu, Y., Fujimura, H., 1981.
Studies on anti-inflammatory agents. 5. A new anti-inflammatory constituent of
Pyracantha crenulata Roem. Chem. Pharm. Bull. 29, 3099–3104.
Sheldrick, G.M., 2008. A short history of SHELX. Acta Cryst. A64, 112–122.
Tuchinda, P., Kornsakulkam, J., Pohmakotr, M., Kongsaeree, P., Prabpai, S., Yoosook,
C., Kassisit, J., Napaswad, C., Sophasan, S., Reutrakul, V., 2008. Dichapetalin-type
triterpenoids and lignans from the aerial parts of Phyllanthus acutissima. J. Nat.
Prod. 71, 655–663.
Long, C., Beck, J., Cantagrel, F., Marcourt, L., Vendier, L., David, B., Plisson, F.,
Derguini, F., Vandenberghe, I., Aussagues, Y., Ausseil, F., Lavaud, C., Sautel, F.,
Massiot, G., 2012. Proteasome inhibitors from Neoboutonia melleri. J. Nat. Prod.
75, 34–47.
Weckert, E., Mattern, G., Addae-Mensah, I., Waibel, R., Achenbach, H., 1996. The
absolute configuration of dichapetalin A. Phytochemistry 43, 657–660.