Steroidal saponins from the leaves of Yucca de-smetiana and their in vitro antitumor activity: Structure activity relationships through a molecular modeling approach

Article abstract of DOI:10.1007/s00044-013-0497-4

Four steroidal saponins were isolated from the leaves of Yucca de-smetiana Baker. Their structures were established using one- and two- dimensional NMR spectroscopy and mass spectrometry. The structure of the new steroidal saponin was identified as: (25R)-3β-hydroxy-5α-spirostan-3-O-β-d- xylopyranosyl-(1 → 2)-β-d-galactopyranosyl-(1 → 2)-O-[β-d-glucopyranosyl-(1 → 3)]-β-d-glucopyranosyl-(1 → 4)-β-d-galactopyranoside (desmettianoside C) along with three known spirostanol and furostanol saponins. The isolated saponins were evaluated for their antitumor activity against HCT116, MCF7, HepG2, and A549 cell lines. Saponins 3 and 4 showed potent activity against HCT116, MCF7, and HepG2 cell lines in comparison with the positive control doxorubicin. A molecular modeling approach was performed to establish conformational criteria that could affect the biological activity of the isolated saponins.

Full text of DOI:10.1007/s00044-013-0497-4

MEDICINAL
CHEMISTRY
RESEARCH
Med Chem Res (2013) 22:4877–4885
DOI 10.1007/s00044-013-0497-4
ORIGINAL RESEARCH
Steroidal saponins from the leaves of Yucca de-smetiana
and their in vitro antitumor activity: structure activity
relationships through a molecular modeling approach
Jacqueline Eskander
Dominique Harakat
•
Ola K. Sakka
Catherine Lavaud
•
•
Received: 2 October 2012 / Accepted: 15 January 2013 / Published online: 30 January 2013
Ó Springer Science+Business Media New York 2013
Abstract Four steroidal saponins were isolated from the
leaves of Yucca de-smetiana Baker. Their structures were
established using one- and two- dimensional NMR spec-
troscopy and mass spectrometry. The structure of the new
steroidal saponin was identified as: (25R)-3b-hydroxy-5
a-spirostan-3-O-b-D-xylopyranosyl-(1 ? 2)-b-D-galactopyr-
anosyl-(1 ? 2)-O-[b-D-glucopyranosyl-(1 ? 3)]-b-D-gluco-
pyranosyl-(1 ? 4)-b-D-galactopyranoside (desmettianoside C)
along with three known spirostanol and furostanol saponins.
The isolated saponins were evaluated for their antitumor activity
against HCT116, MCF7, HepG2, and A549 cell lines. Saponins
Keywords Yucca de-smetiana ꢀ Steroidal saponins ꢀ
Desmettianoside C ꢀ Antitumor activity ꢀ
Molecular modeling
Introduction
Yucca genus (Agavaceae) is widely distributed in central
and northern America; many of them are cultivated in the
tropical gardens of the world as ornamental plants. The
main application of Yucca products is in animal nutrition as
a food additive to reduce ammonia and fecal odors in
animal excreta (Cheeke, 2000) and improve the growth and
productivity of livestock and poultry (Anthony et al., 1994;
Balog et al., 1994). Another important application for
Yucca extracts is as a foaming agent in soft drink, phar-
maceutical, cosmetic, and food industries, which is attrib-
uted to the very high steroidal saponins content in Yucca
species (Piacente et al., 2005; Oleszek et al., 2001;
Hostettmann and Marston, 1995; Mahato et al., 1982).
Studies of steroidal saponins from Yucca genus have
mainly focused on Y. schidigera (Oleszek et al., 2001),
Y. gloriosa (Skhirtladze et al., 2006), Y. filamentosa
3
and 4 showed potent activity against HCT116, MCF7, and
HepG2 cell lines in comparison with the positive control
doxorubicin. A molecular modeling approach was performed to
establish conformational criteria that could affect the biological
activity of the isolated saponins.
J. Eskander (&) ꢀ C. Lavaud
Laboratoire de Pharmacognosie, Institut de Chimie Mol e´ culaire
de Reims (ICMR), UMR-CNRS 7312, B aˆ timent 18,
BP 1039, 51687 Reims Cedex 2, France
e-mail: Jacqueline.eskander@hotmail.com
(Dragalin and Kintia, 1975), Y. elephantipes (Zhang
et al., 2008), Y. aloifolia (Bahuguna and Sati, 1990), and
Y. smalliana (Jin et al., 2007). Some of these species
showed antifungal (Jin et al., 2007), antibacterial, anti-
protozoal, hypocholesterolemic (Piacente et al., 2005), and
antitumor (Balestrieri et al., 2006; Ali et al., 1978) activ-
ities. As part of our chemical investigations on steroidal
saponins of medicinal values (Eskander et al., 2010, 2011),
we studied the leaves of Y. de-smetiana Baker cultivated in
Egypt. Recently, two furostanol saponins named desmet-
tianosides A and B were isolated from the aqueous meth-
anolic extract of the same source which was incorrectly
O. K. Sakka
Department of Chemistry, American University in Cairo,
P.O. Box 74, New Cairo 11835, Egypt
D. Harakat
Service d’analyse, Institut de Chimie Mol e´ culaire de Reims
(
5
ICMR), UMR-CNRS 7312, B aˆ timent 18, BP 1039,
1687 Reims Cedex 2, France
Present Address:
J. Eskander
Pharmacognosy Department, Faculty of Pharmacy, Helwan
University, POB 11795, Ain-Helwan, Cairo, Egypt
123

4
878
Med Chem Res (2013) 22:4877–4885
worded Y. desmettiana; these compounds exhibited high
molluscicidal activity (Diab et al., 2012). In addition, this
paper reports the isolation and structural elucidation of four
steroidal saponins from
HR-ESI–MS, corresponding to the empirical molecular
1
formula C H O Na. The H NMR spectrum displayed
5
6 92 27
signals for two tertiary methyl groups at d 0.61 (s), 0.80
(s) and two secondary methyl groups at d 0.67 (d, J =
5.6 Hz), 1.12 (d, J = 7.0 Hz). On the basis of the HSQC
and HMBC correlations, the aglycone moiety was identi-
fied as (25R)-5a- spirostan-3b-ol known as tigogenin
(Agrawal et al., 1985) previously isolated from Yucca
species (Mahato et al., 1982). The C-25 configuration was
deduced to be (R) based on the difference in chemical shifts
of the geminal protons CH -26 (Dd = 0.09 ppm). It has
Y. de-smetiana, among which, one is a new compound
named desmettianoside C. The antitumor activities of
saponins 1–4 were evaluated against HCT116 (colon),
MCF7 (breast), HepG2 (hepatoma), and A549 (lung) cell
lines in an attempt to discover new antitumor compounds.
Secondly, a comprehensive structural activity relationship
(
SAR) was performed using conformational analysis,
2
H
construction of hydrophobic and hydrophilic bonding map-
pings, rigid and flexible alignment to infer a relationship
between the cytotoxicity of our isolated saponins and their
structures.
been established that DdH is \0.1 ppm in 25R- spiros-
tane steroids, whereas DdH 0.63–0.79 ppm confirmed the
25S- configuration (Agrawal, 2003). The ROE correlation
observed between H-26ax at d 3.58 and Me-27 at d 0.67
confirmed the 25R- configuration. The ROE correlations
between H-1a (ax) and H-3, H-3 and H-5 provided evi-
dence of the b-configuration of C-3 oxygen atom and the
a-configuration of H-5.
Results and discussion
The analysis of the saccharide part of 4 was deduced by
the observation of five anomeric proton signals at d 4.86 (d,
J = 7.8 Hz), 5.09 (d, J = 7.4 Hz), 5.13 (d, J = 7.9 Hz),
The leaves of Y. de-smetiana were extracted with MeOH to
give a crude extract which was precipitated by acetone.
The saponin mixture was obtained after removal of sac-
charides by passing the precipitate through a porous gel
polymer column (Diaion HP-20). The fractions containing
saponins were purified using combination of reversed-
phase C18 column, silica gel column chromatography, and
preparative TLC to afford four saponins (1–4) (Fig. 1).
Three known compounds (1–3) were identified the struc-
tural assignments of which were made by analysis of ESI–
MS and 2D- NMR spectral data including COSY, TOCSY,
HSQC, HMBC, and ROESY. Characterization was carried
out by comparison of their data with those reported in
literature.
1
5.20 (d, J = 7.8 Hz), and 5.55 (d, J = 6.8 Hz) in the H
NMR spectrum associated with five anomeric carbons
resonating at dC 102.2, 105.9, 104.6, 104.3, and 103.8,
respectively, through the HSQC experiment. The nature of
monosaccharides was identified as D-galactose, D-glucose,
and D-xylose from acid hydrolysis of saponin mixture.
Complete assignment of the glycosidic protons was achieved
by analysis of the COSY and TOCSY experiments, while
those of the corresponding glycosidic carbons were deter-
mined through the direct H–C correlations in the HSQC
spectrum. The anomeric signals at d 4.86 (d, J = 7.8 Hz)
and 5.55 (d, J = 6.8 Hz) were identified as two b-D-
galactopyranosyl units from their characteristic equatorial
H-4 d 4.58 (dd, J = 4.0, 2.6 Hz) and 4.04 (d, J = 2.5 Hz).
The signals at d 5.13 (d, J = 7.9 Hz) and 5.20 (d,
J = 7.8 Hz) belonged to the typical spins systems of b-D-
?
?
Saponin 1 with a high-resolution ESI–MS [M ? Na]
atm/z943.4885(calc. 943.4879, C H O Na)wasidentified
4
5 76 19
as (25R)-26-O-b-D-glucopyranosyl-5b-furostan-3b,22a,26-
triol 3-O- [b-D-glucopyranosyl-(1 ? 2)-O-b-D-galactopy-
ranoside], previously isolated from Y. gloriosa (Skhirtladze
et al., 2006) and Y. elephantipes (Zhang et al., 2008).
glucopyranosyl moieties with their H-1 to H-5 in axial
3
positions ( J
[ 7.0 Hz). The last anomeric proton sig-
H–H
?
?
Saponin 2 with a high-resolution ESI–MS [M ? Na]
at m/z 957.5025 (calc. 957.5035, C H O Na) was
nal at d 5.09 (d, J = 7.4 Hz) was identified as b-D-xylo-
3
pyranosyl unit. The large J coupling constant for
H-1, H-2
4
6
78 19
established as (25R)-26-O-b-D-glucopyranosyl-22a-meth-
oxy-5b-furostan-3b, 26-diol 3-O-[b-D-glucopyranosyl-
those five sugar units indicated the b- anomeric configu-
ration. Glycosidation shifts were observed for C-4 gal
(
1 ? 2)-O-b-D-galactopyranoside], previously isolated
from Y. gloriosa (Skhirtladze et al., 2006).
(d 79.7), C-2 glc
0
(d 80.6), C-3 glc
0
(d 88.3), and C-2 gal
0
(d 86.6). The linkage of the sugar units and their sequenc-
ing were confirmed using the HMBC spectrum, which
showed key correlation peaks between the proton signal at
?
?
Saponin 3 with a high-resolution ESI–MS [M ? Na]
at m/z 763.4250 (calc. 763.4245, C H O Na) was assigned
3
9 64 13
to smilagenin 3-O-[b-D-glucopyranosyl-(1 ? 2)-O-b-D-galac-
topyranoside], previously isolated from Y. gloriosa (Nakano
et al., 1989) and Y. elephantipes (Zhang et al., 2008).
d 4.86 (H-1 ) and (d 77.3) C-3 of the aglycone, 5.13 (H-1
0
)
gal
glc
and (d 79.7) C-4 , 5.20 (H-1 00) and (d 88.3) C-3glc
0
, 5.55
, 5.09 (H-1 ) and (d 86.6) C-2 .
0
gal
gal
glc
(H-1gal
0
) and (d 80.6) C-2glc
0
xyl
Saponin 4, isolated as an amorphous solid, showed an
?
accurate [M ? Na] ion at m/z 1219.5735 in the positive
ROEs observed across the glycosidic bonds confirmed the
previous assignments of the HMBC spectrum. Thus, saponin
1
23

Med Chem Res (2013) 22:4877–4885
4879
OH
OH
HO
O
O
O
R
OH
O
O
OH
OH
OH
OH
O
O
O
O
H
HO
HO
HO
O
H
HO
O
R
O
O
OH
HO
HO
1-OH
HO
HO
3
OH
2-OCH3
O
O
OH Glc'
OH Glc''
O
O
HO
OH
O
Gal
HO
HO
O
O
O
OH
HO
O
O
H
HO
HO
OH
O
HO
Gal'
OH
O
4
Xyl
HO
OH
Fig. 1 Isolated saponins (1–4) from Yucca de-smetiana leaves
4
was deduced to be (25R)-3b-hydroxy-5a-spirostan-3-O-b-
D-xylopyranosyl-(1 ? 2)-b-D-galactopyranosyl-(1 ? 2)-O-
b-D-glucopyranosyl-(1 ? 3)]-b-D-glucopyranosyl-(1 ? 4)-
Table 1 The in vitro cytotoxicity (IC50, lM) of isolated saponins
from Y. de-smetiana
[
Cell line
Doxorubicin
3
4
b-D-galactopyranoside, and named desmettianoside C.
HCT116
MCF7
A549
6.86
5.46
0.84
7.36
4.4 ± 0.47
4.0 ± 0.85
2.4 ± 0.57
2.6 ± 0.49
In vitro anti-tumor evaluation
16.5 ± 1.45
3.5 ± 0.41
10.2 ± 0.97
1.1 ± 0.56
HepG2
The cytotoxic activity of saponins 1–4 was investigated
against HCT116 (colon), MCF7 (breast), HepG2 (hepa-
toma), and A549 (lung) cell lines. The results are presented
as the inhibitory concentration (IC ) that inhibits the
Compounds 1–2 having IC50 values larger than 100 lM are not
included in the table
5
0
saponin 3 against all cell lines. Upon examination of
structures of saponins 3 and 4, we can conclude that an
increase in the number of sugar moieties leads to an
increase in cytotoxicity. Saponin 4 showed the highest
inhibitory effect against HCT116 cells (IC50 2.4 lM) as
compared to 3 (IC50 4.4 lM). Saponin 4 also showed
higher potency against MCF7 cells (IC50 2.6 lM) than
saponin 3 (IC50 4.0 lM) and was almost two folds more
active than the positive control doxorubicin. Saponins 3
and 4 exhibited a relatively weak in vitro cytotoxic activity
against A549 tumor cell line with IC50 values of 16.5 lM
and 10.2 lM, respectively, compared with the positive
growth of cells by 50 % when compared with untreated
cells after using MTT assay (Table 1). The anticancer drug
Doxorubicin was also included in the study as a compari-
son. Furostanol saponins 1 and 2 showed no cytotoxicity
against all cell lines while spirostanol saponins 3 and 4
were able to inhibit tumor cells in a dose-dependent man-
ner (Fig. 2). Since inactive saponins 1 and 2 shared the
same common b-D-glucopyranosyl-(1 ? 2)-O-b-D-galac-
topyranoside sugar chain as active saponin 3, the difference
in cytotoxicity could be attributed to the presence of the
spirostan aglycone. Saponin 4 showed higher potency than
123

4
880
Med Chem Res (2013) 22:4877–4885
Fig. 2 a The effects of
saponins 3 and 4 on cell death in
HCT116, MCF7 cells. b The
effects of saponins 3 and 4 on
cell death in A549 and HepG2
cells
control doxorubicin (IC50 0.84 lM). Desmettianoside C (4)
showed the highest inhibitory effect against HepG2 cells
aglycones and their cytotoxicity is available. In this study,
we report a comprehensive SAR analysis on both the spi-
rostan and furostan aglycones existing in our isolated
saponins 1–4 using molecular modeling to pursue infor-
mation regarding their different cytotoxic activities.
(
IC50 1.1 lM) in comparison to 3 (IC50 3.5 lM) while 3
was almost two folds more active than positive control
doxorubicin against HepG2 tumor cell line.
The fact that both furostan saponins showed no significant
cytotoxic activity against all cell lines, although disappoint-
ing, provides new information regarding important structural
elements that seem to be required for cytotoxicity. Since
isolated spirostanol saponin 3 shared the same common
b-D-glucopyranosyl-(1 ? 2)-O-b-D-galactopyranoside sugar
chain, the difference in cytotoxicity could be attributed to the
type of aglycone and to the possible existence of the sugar
chain in different conformations with varying contributions
to the partition function. The lowest energy-minimized
structures for saponins 1–4 are shown in Fig. 3. To allow for a
bettervisualizationofthe conformationaldifferences between
Conformational analysis
The correlation between the presence of a spirostanol
structure and anticancer activity of steroidal saponins was
previously reported (Yuen-Nei Cheung et al., 2005; Gonzalez
et al., 2003, Liu et al., 2004), and remarkable reports
(
Labrada et al., 2011; Mimaki et al., 2001; Trouillas et al.,
005) studied conformational and electronic differences
2
between natural and synthetic saponins possessing the
spirostan aglycone. To date, limited information on the
structure activity relationship (SAR) of different steroidal
1
23

Med Chem Res (2013) 22:4877–4885
4881
Fig. 3 Lowest energy
conformers of saponins 1–4 (left
panel). Hydrophobic (blue) and
hydrophilic (red) mappings of
the lowest energy conformers
1–4 (right panel) (Color figure
online)
the four molecules, we rigidly aligned the local minima best
steroidal plane remained intact. Such dissimilarity in orien-
tation of the sugar chain between the pair of saponins 1, 2 and
that of 3, 4 could be partly attributed to the substitution of the
furostan aglycone in 1 and 2 with the spirostan aglycone in 3
and 4. With regard to this result, it can be concluded that the
three dimensional orientation of the galactose moiety that is
directly attached to the spirostan aglycone of cytotoxic sap-
onins 3 and 4 contributes to their activity and is of importance
to their mode of action. Interestingly, a correlation between
spatial geometry and biological activity was previously
reported (Labrada et al., 2011; Mimaki et al., 2001). There-
fore, it can be concluded that variations in biological activity
of saponins 1–4 are influenced by their spatial parameters.
To further assess the importance of the spirostanol system
for cytotoxic activity, flexible alignment was employed in
addition to rigid alignment. Flexible alignment is an appli-
cation for flexibly aligning small molecules. It is a stochastic
search procedure that accepts as input a collection of small
representing each saponin 1–4 based on the steroid ring
(
Fig. 4). Theresultsshowedthatthelowestenergy-minimized
structures of the inactive saponins 1 and 2 possessing the
common sugar chain b-D-glucopyranosyl-(1 ? 2)-O-b-D-
galactopyranoside, exhibited a common orientation of their
sugar moieties with respect to the steroidal plane. In this ori-
entation, the b-D-galactose unit (green) attached to C-3 of the
aglycone was roughly in the same plane as the aglycone. A
dissimilar orientation of the sugar chain with respect to the
steroidal aglycone was noticed upon close inspection of the
lowest-energy minimized structure of cytotoxic saponins 3
and 4. Therein, the b-D-galactose moiety (blue) attached to the
aglycone adopted a conformation that is vertical to the ste-
roidal spirostan plane. Even though the sugar moieties of local
minima of saponins 3 and 4 failed to align together due to the
presence of a different sugar chain at C-3 of aglycone, the
vertical orientation of the galactose moiety with respect to the
123

4
882
Med Chem Res (2013) 22:4877–4885
alignments is an almost complete superposition of the spi-
˚
rostan aglycone of the most active saponins 3 and 4 (0.35 A)
˚
(Fig. 5a); however, adeviation (0.87 A) is observed at C-3 of
the spirostan aglycone, which may indicate the reason
behind the slight difference in the cytotoxicity of saponins
3
and 4.
By the same methodology, saponin 1 was taken as a
representative example of the inactive saponin and was
subjected to flexible alignment along with 3 and 4. Saponin
1
(Fig. 5b) was flexibly aligned in a different manner when
compared to the congruous alignment of active saponins 3
and 4, explaining why such a compound was void of
antitumor activity. This further proves that the steroid
spirostan conjugate of 3 and 4 could be an important factor
behind their significant cytotoxicity.
Fig. 4 Rigid alignment of the molecular geometries of the most
stable conformers of saponins 1–4. b-D-galactose unit attached to C-3
of the aglycone is shown in green for saponins 1, 2 and in blue for
saponins 3, 4 (Color figure online)
Hydrophobic interactions are the primary driving force
in protein–ligand binding and protein folding (Horton and
Lewis, 1992; Sarkar and Kellogg, 2010). Recent studies
demonstrated that saponin-induced autophagy is governed
by protein–ligand interactions through their association
with suppression of Akt and ERK1/2 proteins (Dou et al.,
2001; Ellington et al., 2006; Tong et al., 2011). Therefore,
the construction of hydrophobic mappings for compounds
1–4 (Fig. 3) allows us to utilize the spatial conformations
of molecules in examining the similarity and dissimilarity
in the electronic binding characteristics of their surfaces to
possible biological receptors. It can be noted that the
presence of the 26-O- b-D glucose moiety in 1 and 2
resulted in a different distributions of hydrophilic regions.
Comparison of the hydrophobic mappings of 1–4 indicated
that hydrophilic regions are distributed on both sides (C-3
and C-26 glycoside chains) of saponins 1 and 2, with a
hydrophobic region occupying the agylcone. On the other
hand, similarly cytotoxic saponins 3 and 4 share a common
one-sided hydrophilic surface distribution located on C-3
of the aglycone owing to the presence of the spirostan ring.
This insightful information is important in elucidating the
molecules with 3D coordinates and computes a collection of
alignments by generating conformations through random
rotation of bonds (including ring bonds) instead of only
Cartesian coordinate perturbation. Each alignment is given a
score that quantifies the quality of the alignment in terms of
both internal strain and overlap of molecular features. Often,
atomic-level details of the structures of pharmaceutically
relevant receptors are not available. In such cases, 3D
alignment (or superposition) of putative ligands can be used
to deduce structural requirements for biological activity.
Methodologies based upon 3D alignment for finding bio-
logically active ligands generally make use of the qualitative
assumption that if two ligands align well, they will possess
similar biological activity (Kearsley, 1990).
After assigning MMFF94 charges to all molecules, flex-
ible alignment was employed to scan and rank overlays of
saponins 1–4 based on stearic, electrostatic field, hydro-
phobic areas overlap, hydrogen bond donors, and acceptors
overlap. A common feature of the MOE-generated
Fig. 5 a Flexible alignments of the most active saponins: 3 (in red), 4 (in green). b Flexible alignments of the inactive saponin 1 (in blue) and
active saponins 3 (in red), 4 (in green) (Color figure online)
1
23

Med Chem Res (2013) 22:4877–4885
4883
preferred binding mode of saponins 3 and 4 to comple-
mentary amino acid residues of biological receptors.
extract (170 g). The methanol extract was suspended in
MeOH (350 ml) and precipitated by addition of a large
excess of Me CO (2 l). The resulting precipitate was filtered
2
and dried to give (75 g) of crude saponin mixture. This
mixture was passed through a porous polymer gel column
Conclusion
(Mitsubishi Diaion HP-20), eluted with H O then MeOH:
2
Analysis of geometric characteristics by molecular modeling
demonstrated their direct influence on biological activity.
Our study demonstrated a dissimilar spatial conformation
adopted by the 3-O- b-D galactosyl moiety of both spirostan
and furostan aglycones. Hydrophobic mappings and flexible
alignment of the lowest energy conformations of saponins
H O (50: 50 and 75: 25) and finally 100 % MeOH. Saponins
2
of fractions eluted with MeOH: H O (50: 50) (5 g) were
2
purified by silica gel CC using a gradient of CHCl : MeOH
3
(80: 20 to 70: 30) to give 69 fractions. Fractions 31–38
(146 mg) were chromatographed successively on silica gel
column eluted with CHCl : MeOH (80: 20 to 75: 25) then by
3
1
–4 indicated that biological activity of saponins 3 and 4
preparative TLC eluted with CHCl –MeOH–H O (70:30:5)
2
3
could also be attributed to the presence of the spirostan
aglycone. Thus, this report represents the first study ratio-
nalizing the cytotoxicity of the spirostan aglycone both
experimentally and theoretically and is thus important for
future SAR studies that might bring out information con-
cerning the apoptotic mechanism of action of saponins.
to yield pure saponins 1 (12 mg) and 2 (10 mg). Saponins of
fractions eluted with 100 % MeOH (1 g) were subjected to
silica gel column eluted with CHCl : MeOH (90: 10 to 78:
3
22) to afford pure saponins 3 (23 mg) and 4 (20 mg).
Saponin 4
2
D
1
1
Experimental
White amorphous powder; [a] –32.4 (c 0.1, MeOH); H
NMR of the aglycone moiety (C D N):
5
5
General
d 4.53 (1H, q, J = 6.7 Hz, H-16), 3.89 (1H, m, H-3),
.58 (1H, m, H-26a), 3.49 (1H, t, J = 10.9 Hz, H-26b),
3
Optical rotations were determined in MeOH with a Perkin-
1 13
Elmer 241 polarimeter. H and C NMR spectra were
0.47 (1H, td, J = 12.0, 3.7 Hz, H-5), 0.61 (3H, s, Me-19),
0.67 (3H, d, J = 5.6 Hz, Me-27), 0.80 (3H, s, Me-18), 1.12
1
(3H, d, J = 7.0 Hz, Me-21); C NMR of the aglycone
3
recorded in pyridine-d on a Bruker DRX-500 at 500 and
5
1
25 MHz, respectively, and 2D-NMR experiments were
moiety (C D N):
5
5
performed using standard Bruker microprograms. ESI–MS
and high resolution MS were recorded on Micromass
Q-TOF micro instrument (Manchester, UK) with an elec-
trospray source. The samples were introduced by infusion
in a solution of MeOH (5 ll/min). TLC was carried out on
pre-coated silica gel 60 F254 (Merck), and spots were
visualized by spraying with 50 % H SO . Kieselgel 60
d 36.9 (C-1), 29.7 (C-2), 77.3 (C-3), 34.6 (C-4), 54.1 (C-
5), 28.7 (C-6), 32.2 (C-7), 35.0 (C-8), 44.4 (C-9), 35.6 (C-
10), 21.0 (C-11), 39.9 (C-12), 40.5 (C-13), 56.2 (C-14),
31.9 (C-15), 81.2 (C-16), 62.8 (C-17), 16.4 (C-18), 12.1 (C-
19), 41.7 (C-20), 14.8 (C-21), 109.0 (C-22), 31.6 (C-23),
1
29.0 (C-24), 30.4 (C-25), 66.8 (C-26), 17.1 (C-27). H
NMR of the sugar moiety (C D N): d 4.86 (d, J = 7.8 Hz,
2
4
5 5
(63–200 lm, Merck) and Lichroprep RP-18 (40–63 lm,
Merck) were used for CC.
H-1 ), 4.38 (dd, J = 9.2, 7.6 Hz, H-2 ), 4.08 (m, H-3 ),
gal gal gal
4.58 (dd, J = 4.0, 2.6 Hz, H-4 ), 4.07 (m, H-5 ), 4.18
gal gal
(
m, H-6 ), 4.67 (m, H-6 ), 5.13 (d, J = 7.9 Hz, H-1
gal
0
),
gal
glc
Plant material
4.34 (t, J = 8.4 Hz, H-2 glc
H-3glc ), 4.12 (t, J = 9.0 Hz, H-4 glc
4.01 (dd, J = 11.0, 5.0 Hz, H-6glc
J = 10.8 Hz, H-6glc
), 5.20 (d, J = 7.8 Hz, H-1glc00), 4.02
0
), 4.14 (dd, J = 8.8, 8.4 Hz,
), 3.82 (m, H-5glc ),
), 4.47 (brd,
0
0
0
The leaves of Y. de-smetiana Baker were collected in May
0
2
010 from El-Orman Public Botanical Garden, Giza,
0
Egypt. The plant was identified by Dr. Th e´ r e` se Labib,
senior specialist of plant identification at El-Orman Public
Botanical Garden. Avoucher Herbarium specimen(H.M.G. 38,
(dd, J = 8.9, 8.0 Hz, H-2glc00), 4.13 (m, H-3glc00), 4.08 (t,
J = 9.2 Hz, H-4 glc00), 3.82 (m, H-5glc00), 4.22 (dd,
J = 10.8, 5.2 Hz, H-6glc00), 4.53 (dd, J = 11.0, 7.1 Hz,
2
010) was deposited in the Herbal Medicinal Garden of
H-6glc00), 5.55 (d, J = 6.8 Hz, H-1gal
7.0 Hz, H-2gal ), 3.82 (m, H-3gal ), 4.04 (d, J = 2.5 Hz,
), 4.00 (m, H-5gal ), 4.28 (dd, J = 11.0, 7.5 Hz,
), 4.47 (brd, J = 10.8 Hz, H-6gal ), 5.09 (d,
0
), 4.07 (dd, J = 8.0,
Helwan University, Egypt.
0
0
H-4gal
H-6gal
0
0
0
Extraction and isolation
0
J = 7.4 Hz, H-1xyl), 3.92 (t, J = 8.6 Hz, H-2 xyl), 4.05 (t,
J = 8.5 Hz, H-3 xyl), 4.08 (m, H-4 xyl), 3.58 (m, H-5 xyl),
4.18 (dd, J = 10.2, 5.0 Hz, H-5 xyl).
The air dried powdered leaves of Y. de-smetiana (1 kg) were
extracted three times with MeOH to obtain a concentrated
123

4
884
Med Chem Res (2013) 22:4877–4885
13
C NMR of the sugar moiety (C D N): d 102.2 (C-1gal),
37 °C under 5 % CO . To stop the reaction and to dissolve
5
5
2
7
6
7
7
6
7
7
2.9 (C-2gal), 75.3 (C-3gal), 79.7 (C-4gal), 75.3 (C-5gal),
0.4 (C-6gal), 104.6 (C-1glc ), 80.6 (C-2glc ), 88.3 (C-3glc ),
1.4 (C-4glc ), 77.9 (C-5glc ), 62.8 (C-6glc
), 104.3 (C-1glc00),
5.2 (C-2glc00), 78.2 (C-3glc00), 70.5 (C-4glc00), 77.2 (C-5glc00),
2.1 (C-6glc00), 103.8 (C-1gal
), 86.6 (C-2gal ), 70.6 (C-3gal ),
4.8 (C-4gal ), 78.4 (C-5gal ), 61.8 (C-6gal ), 105.9 (C-1xyl),
5.2 (C-2xyl), 77.5 (C-3 ), 68.8 (C-4xyl), 66.6 (C-5xyl).
the formed crystals, 200 ll of 10 % Sodium dodecyl sul-
fate (SDS) in deionized water was added to each well and
incubated overnight at 37 °C. A positive control composed
of 100 lg/ml was used as a known cytotoxic natural agent,
which gives 100 % lethality under the same conditions
(Thabrew et al., 1997).The absorbance was then measured
using a microplate multi-well reader (Bio-Rad Laboratories
Inc., model 3350, Hercules, California, USA) at 595 nm
and a reference wavelength of 620 nm. A statistical sig-
nificance was tested between samples and negative control
(cells with vehicle) using independent t test by SPSS 11
program. DMSO is the vehicle used for dissolution of plant
extracts, and its final concentration on the cells was less
than 0.2 %. The percentage of change in viability was
calculated according to the formula: (Reading of extract/
Reading of negative control) -1 9 100. A probit analysis
was carried for IC50 and IC90 determination using SPSS
11 program.
0
0
0
0
0
0
0
0
0
0
0
0
xyl
?
?
High-resolution ESI–MS [M ? Na] : m/z 1219.5735
?
calc. 1219.5724, C H O Na); ESI–MS : 1088.5
(
[
[
56 92 27
?
M ? Na-xylosyl] , 925.5 [M ? Na-galactosyl] , 763.4
?
?
?
M ? Na-glucosyl] , 601.4 [M ? Na-glucosyl] .
Acid hydrolysis
The crude saponin mixture (100 mg) was refluxed with
1
0 ml of 2 N HCl at 100 °C for 5 h. After cooling, the
reaction was extracted with CHCl to remove aglycones.
3
The acid aqueous layer was neutralized with 1 N KOH and
evaporated. Three sugars were identified and compared
with authentic samples by TLC using solvent MeCOEt–
iso-PrOH–Me CO–H O (20:10:7:6) as glucose, galactose,
Molecular modeling calculations
2
2
and xylose. The purification of these sugars was achieved
by prep. TLC using Kieselgel 60 plates which were eluted
three times with solv., CHCl –MeOH–H O (70:30:1) to
Energy minimization
This was performed by use of the Merck forcefield
(MMFF94) as implemented in MOE 2008.10. program
(Chemical Computing Group Inc., Montreal, Canada). This
energy minimization methodology is capable of calculating
constrained geometries through the use of chiral, distance,
angle, and dihedral restraints (Halgren, 1996). Hydrogen
atoms and lone pairs were added to each molecule. Energy
minimization was terminated when the root mean square
gradient fell below 0.05. Force field partial charges were
calculated prior to energy minimization.
3
1
2
2
afford D-glucose (R 0.26, [a] ?53; H O), D-galactose (R
f
D
2
f
2
1
21
0
.27, [a] ?56; H O), D-xylose (R 0.48, [a] ?15; H O).
D
2
f
D
2
Cytotoxicity assays
Cell viability was assessed by the mitochondrial dependent
reduction of yellow MTT (3-(4,5-dimethylthiazol-2-yl)-
2
,5-diphenyl tetrazolium bromide) to purple formazan
(
Mosmann, 1983). All the following procedures were done
in a sterile area using a Laminar flow cabinet biosafety
class II level (Baker, SG403INT, Sanford, ME, USA).
Cells were suspended in RPMI 1640 medium [(for HePG2-
MCF7 and HCT116–DMEM for A549)], 1 % antibiotic–
antimycotic mixture (10,000 lg/ml Potassium Penicillin,
Flexible alignment
Using MOE/MMFF94, 200 conformers of each compound
were generated and minimized with a distance-dependent
dielectric model. Each resulting conformation was checked
to determine if it had already been generated by comparing
all atom positions using a predefined root-mean-square
1
0,000 lg/ml Streptomycin Sulfate, and 25 lg/ml Ampho-
tericin B), and 1 % L-glutamine at 37 °C under 5 % CO2.
Cells were batch cultured for 10 days, then seeded at a
3
concentration of 10 9 10 cells/well in fresh complete
˚
(RMS) tolerance (0.5 A). The procedure was terminated
growth medium in 96-well micro titer plastic plates at
when the number of failures to find new conformations
exceeded 20 in a row, a stringent criterion. Conformations
with energies over 10 kcal/mol above the respective min-
ima were rejected. Chirality was preserved using signed
volume restraints on all chiral centers. All open chain
bonds and flexible ring bonds were set to random dihedral
angles. The alignment mode was chosen as flexible. In the
calculations, a bias of *30° was selected, which means
that dihedral angles were rotated by random angle with a
sum-of-gaussians distribution with peaks at multiples of
3
7 °C for 24 h under 5 % CO using a water jacketed
2
carbon dioxide incubator (Sheldon, TC2323, Cornelius,
OR, USA). Media was aspirated, fresh medium (without
serum) was added, and cells were incubated either alone
(
negative control) or with different concentrations of
sample to give a final concentration of (100–50–25–12.5–
.25–3.125–0.78 and 1.56 lM). After 48 h of incubation,
6
medium was aspirated, 40 ll MTT salt (2.5 lg/ml) was
added to each well, and incubated for further four hours at
1
23

Med Chem Res (2013) 22:4877–4885
4885
3
0°. From the top scoring superpositions obtained, only one
Horton N, Lewis M (1992) Calculation of the free energy of
association for protein complexes. Prot Sci 1:169–181
Hostettmann KA, Marston A (1995) Saponins chemistry and phar-
macology of natural products. Cambridge University Press,
Cambridge
was consistent with the structure–activity relationships of
the aligned saponins, and this was subjected to a more
refined searching using MOE/flexible alignment module.
Jin Y, Kuk J, Oh K, Kim Y, Piao X, Park R (2007) A new steroidal
saponin, yuccalan, from the leaves of Yucca smalliana. Arch
Pharm Res 30:543–546
Kearsley S (1990) An alternative method for the alignment of
molecular structures: maximizing electrostatic and steric over-
lap. Tetrahedron Comput Methodol 3:615–633
Acknowledgments The authors are grateful to Dr. Khaled Mohamed,
Bioassay-Cell Culture Laboratory, National Research Center- Cairo,
Egypt, for carrying the antitumor activity.
Labrada KP, Brouard I, Morera C, Estevez F, Bermejo J, Rivera DG
(2011) ‘Click’ synthesis of triazole- based spirostan saponin
analogs. Tetrahedron 67:7713–7727
References
Agrawal PK (2003) Spectral assignments and reference data. Magn
Reson Chem 41:965–968
Agrawal PK, Jain DC, Gupta RK, Thakur RS (1985) Carbon-13 NMR
spectroscopy of steroidal sapogenins and steroidal saponins.
Phytochemistry 24:2479–2496
Ali MS, Sharma GC, Asplund RO, Nevins MP, Garb S (1978)
Isolation of antitumor polysaccharide fractions from Yucca
glauca Nutt. (Lilliaceae). Growth 42:213–223
Anthony NB, Balog JM, Staudinger FB, Walker RD, Huff WE (1994)
Environmental variability and incidence of ascites. Poult Sci
Liu MJ, Wang Z, Ju Y, Zhou JB, Wang Y, Wong RNS (2004) The
mitotic-arresting and apoptosis-inducing effects of diosgenyl
saponins on human leukemia cell lines. Biol Pharm Bull
2
7:1059–1065
Mahato SB, Ganguly AN, Sahu NP (1982) Steroid saponins. Phyto-
chemistry 21:959–978
Mimaki Y, Yokosuka A, Kuroda M, Sashida Y (2001) Cytotoxic
activities and structure-cytotoxic relationships of steroidal sap-
onins. Biol Pharm Bull 24:1286–1289
Mosmann T (1983) Rapid colorimetric assays for cellular growth and
survival: application to proliferation and cytotoxicity assays.
J Immunol Methods 65:55–63
Nakano K, Yamasaki T, Imamura Y, Murakami K, Takaishi Y,
Tomimatsu T (1989) The steroidal glycosides from the stem of
Yucca gloriosa. Phytochemistry 28:1215–1217
7
3:801–809
Bahuguna S, Sati OP (1990) Spirostanol saponins from Yucca
aloifolia rhizomes. Phytochemistry 29:342–343
Balestrieri C, Piacente S, Pizza C, Montoro P, Oleszek W, Visciano
V, Balestrieri ML (2006) Relative effects of phenolic constit-
uents from Yucca schidigera Roezl. bark on kaposi’s sarcoma
cell proliferation, migration, and PAF synthesis. Biochemphar-
macol 71:1479–1487
Oleszek W, Sitek M, Stochmal A, Piacente S, Pizza C, Cheeke P
(2001) Steroidal saponins of Yucca schidigera Roezl. J Agric
Food Chem 49:4392–4396
Balog JM, Anthony NB, Wall CV, Walker RD, Rath NC, Huff WE
(1994) Blood variables, ascites scores and body and organ
weights. Poult Sci 73:810–816
Piacente S, Pizza C, Oleszek W (2005) Saponins and phenolics of
Yucca schidigera Roezl: chemistry and bioactivity. Phytochem
Rev 4:177–190
Cheeke PR (2000) Actual and potential applications of Yucca
schidigera and Quillaia saponaria saponins in human and
animal nutrition. J Anim Sci 77:1–10
Diab Y, Ioannou E, Emam A, Vagias C, Roussis V (2012) Desmettiano-
sides A and B, bidesmosidic furostanol saponins with molluscicidal
activity from Yucca desmettiana. Steroids 77:686–690
Dou DQ, Zhang YW, Zhang L, Chen YJ, Yao XS (2001) The
inhibitory effects of ginsenosides on protein tyrosine kinase
activated by hypoxia/reoxygenation in cultured human umbilical
vein endothelial cells. Planta Med 67:19–23
Dragalin IP, Kintia PK (1975) Steroidal saponins of Yucca filamen-
tosa: yuccoside C and protoyuccoside C. Phytochemistry 14:
Sarkar A, Kellogg GE (2010) Hydrophobicity-shake flasks, protein
folding and drug discovery. Curr Top Med Chem 10:67–83
Skhirtladze A, Plaza A, Montoro P, Benidze M, Kemetelidze E, Pizza
C, Piacente S (2006) Furostanol saponins from Yucca gloriosa L.
rhizomes. Biochem Syst Ecol 34:809–814
Thabrew MI, Hughes RD, McFarlane IG (1997) Screening of
hepatoprotective plant components using a HepG2 cell cytotox-
icity assay. J Pharm Pharmacol 49:1132–1135
Tong QY, Qing Y, Shu D, He Y, Zhao YL, Li Y, Wang ZL, Zhang
SY, Xing ZH, Xu C, Wei YQ, Huang W, Wu XH (2011)
Deltonin, a steroidal saponin, inhibits colon cancer cell growth
in vitro and tumor growth in vivo via induction of apoptosis and
antiangiogenesis. Cell Physiol Biochem 27:233–242
1
817–1820
Ellington AA, Berhowl MA, Singletary KW (2006) Inhibition of Akt
signaling and enhanced ERK1/2 activity are involved in induction
of macroautophagy by triterpenoid B-group soyasaponins in colon
cancer cells. Carcinogenesis 27:298–306
Eskander J, Lavaud C, Harakat D (2010) Steroidal saponins from the
leaves of Agave macroacantha. Fitoterapia 81:371–374
Eskander J, Lavaud C, Harakat D (2011) Steroidal saponins from the
leaves of Beaucarnea recurvata. Phytochemistry 72:946–951
Gonzalez AG, Hernandez JC, Leon F, Padron JI, Estevez F, Quintana J,
Bermejo J (2003) Steroidal saponins from the bark of Dracaena
draco and their cytotoxic activities. J Nat Prod 66:793–798
Halgren TA (1996) Merck molecular force field. I. Basis, form, scope,
parameterization, and performance of MMFF94. J Comp Chem
Trouillas P, Corbiere C, Liagre B, Duroux JL, Beneytout JL (2005)
Structure-function relationship for saponin effects on cell cycle
arrest and apoptosis in the human 1547 osteosarcoma cells: a
molecular modelling approach of natural molecules structurally
close to diosgenin. Bioorg Med Chem 13:1141–1149
Yuen-Nei Cheung J, Chik-Ying Ong R, Suen YK, Ooi V, Nai-Ching
Wong H, Chung-Wai Mak T, Fung KP, Yu B, Kong SK (2005)
Polyphyllin D is a potent apoptosis inducer in drug-resistant
HepG2 cells. Cancer Lett 217:203–211
Zhang Y, Zhang Y-J, Jacob MR, Li X, Yang C (2008) Steroidal
saponins from the stem of Yucca elephantipes. Phytochemistry
6
9:264–270
1
7:490–519
123