Steroidal Saponins from Furcraea hexapetala Leaves and Their Phytotoxic Activity

Article abstract of DOI:10.1021/acs.jnatprod.6b00702

Four new steroidal saponins (1-4) along with 13 known saponins were isolated from the leaves of Furcraea hexapetala. The new compounds were identified as (20R,22R,25R)-3β-hydroxy-5α-spirostan-12-one 3-O-{α-l-rhamnopyranosyl-(1→4)-O-β-d-glucopyranosyl-(1→3)-O-[β-d-glucopyranosyl-(1→3)-O-β-d-glucopyranosyl-(1→2)]-O-β-d-glucopyranosyl-(1→4)-O-β-d-galactopyranoside} (1), (25R)-3β-hydroxy-5α-spirost-20(21)-en-12-one 3-O-{α-l-rhamnopyranosyl-(1→4)-O-β-d-glucopyranosyl-(1→3)-O-[β-d-glucopyranosyl-(1→3)-O-β-d-glucopyranosyl-(1→2)]-O-β-d-glucopyranosyl-(1→4)-O-β-d-galactopyranoside} (2), (25R)-5α-spirostan-3β-ol 3-O-{β-d-glucopyranosyl-(1→2)-O-β-d-glucopyranosyl-(1→2)-O-β-d-glucopyranosyl-(1→4)-O-β-d-galactopyranoside} (3), and (25R)-5β-spirostan-3β-ol 3-O-{β-d-glucopyranosyl-(1→6)-O-β-d-galactopyranoside} (4) by spectroscopic analysis, including one- and two-dimensional NMR techniques, mass spectrometry, and chemical methods. The phytotoxicity of the isolated compounds against the standard target species Lactuca sativa was evaluated. Structure-activity relationships for these compounds with respect to phytotoxic effects are discussed.

Full text of DOI:10.1021/acs.jnatprod.6b00702

Article
pubs.acs.org/jnp
Steroidal Saponins from Furcraea hexapetala Leaves and Their
Phytotoxic Activity
Juan M. Calle,† Andy J. Perez,‡ Ana M. Simonet,† Jose O. Guerra,§ and Francisco A. Macías*,†
́
́
†Grupo de Alelopatía, Departamento de Química Organ
Cadiz, C/ Republica Saharaui, s/n, 11510-Puerto Real (Cad
́
ica, Instituto de Biomolec
iz), Spain
́
ico (UDT), Universidad de Concepcion, Bío-Bío, Chile
́
ulas (INBIO), Facultad de Ciencias, Universidad de
́
́
́
‡
́
Area Productos Químicos, Unidad de Desarrollo Tecnolog
́
§Departamento de Licenciatura en Química, Facultad de Química y Farmacia, Universidad Central “Marta Abreu” de Las Villas,
Carretera a Camajuaní km 5.5, 54830 Santa Clara, Cuba
S
* Supporting Information
ABSTRACT: Four new steroidal saponins (1−4) along with
13 known saponins were isolated from the leaves of Furcraea
hexapetala. The new compounds were identified as
(20R,22R,25R)-3β-hydroxy-5α-spirostan-12-one 3-O-{α-L-
rhamnopyranosyl-(1→4)-O-β-D-glucopyranosyl-(1→3)-O-[β-
D-glucopyranosyl-(1→3)-O-β-D-glucopyranosyl-(1→2)]-O-β-
D-glucopyranosyl-(1→4)-O-β-D-galactopyranoside} (1),
(25R)-3β-hydroxy-5α-spirost-20(21)-en-12-one 3-O-{α-L-
rhamnopyranosyl-(1→4)-O-β-D-glucopyranosyl-(1→3)-O-[β-
D-glucopyranosyl-(1→3)-O-β-D-glucopyranosyl-(1→2)]-O-β-
D-glucopyranosyl-(1→4)-O-β-D-galactopyranoside} (2),
(25R)-5α-spirostan-3β-ol 3-O-{β-D-glucopyranosyl-(1→2)-O-
β-D-glucopyranosyl-(1→2)-O-β-D-glucopyranosyl-(1→4)-O-β-
D-galactopyranoside} (3), and (25R)-5β-spirostan-3β-ol 3-O-{β-D-glucopyranosyl-(1→6)-O-β-D-galactopyranoside} (4) by
spectroscopic analysis, including one- and two-dimensional NMR techniques, mass spectrometry, and chemical methods. The
phytotoxicity of the isolated compounds against the standard target species Lactuca sativa was evaluated. Structure−activity
relationships for these compounds with respect to phytotoxic effects are discussed.
he genus Furcraea belongs to the Agavaceae family, which
Tconsists of around 20 species endemic to tropical America
and is cultivated mainly in warm climates as ornamental plants.
One of these species is Furcraea hexapetala (Jacq.) Urb., an
endemic plant from the western region of Cuba, the leaves of
which can reach up to two meters in length and are used as a
source of fibers.1 Steroidal sapogenins were identified from
leaves of F. hexapetala, with hecogenin and tigogenin being the
most prominent.2 However, in-depth phytochemical studies on
F. hexapetala have not been reported to date. The extracted
juice of the leaves of this plant is used by farmers of Cienfuegos
Province (Cuba), who spray a diluted solution over plants as a
natural pesticide, and this has led to prior studies being carried
out. The insecticidal and acaricidal activities of extracts
obtained from F. hexapetala have been reported recently, with
these activities attributed to the presence of saponins.3,4
As part of current research on the phytochemical study of
species from the Agavaceae family,5−8 along with the search for
phytotoxic natural products, a bioassay-guided isolation of
compounds from the leaves of F. hexapetala was performed.
The isolation and characterization of four new steroidal
saponins, named furcrosides A−D (1−4), together with
another 13 known steroidal saponins are described. The main
isolated compounds were evaluated in a phytotoxicity bioassay.9
RESULTS AND DISCUSSION
■
Dried and ground leaves of F. hexapetala were extracted with
EtOH−H2O (7:3). The extract was partitioned in n-BuOH−
H2O, and the organic phase was subjected to VLC on RP-18 to
give nine fractions. In order to evaluate the initial bioactivity of
the fractions from the butanol extract, these fractions were
assayed on etiolated wheat coleoptiles at 800, 400, and 200
ppm (Figure S41, Supporting Information). This bioassay has
been confirmed as an appropriate approach for bioguided
isolation, showing a direct correlation with the phytotoxicity of
saponins.7,8
Multiple separation procedures were carried out on the active
fractions (F4 to F9) to provide 17 pure steroidal saponins. Four
new steroidal saponins, named furcrosides A−D (1−4), were
isolated along with 13 known steroidal saponins identified as
yuccaloesides C (5) and B (6),10 atroposide E (7),11
petunioside F (8),12 chlorogenin-3-O-{α-L-rhamnopyranosyl-
(1→4)-O-β-D-glucopyranosyl-(1→3)-O-[β-D-glucopyranosyl-
(1→3)-O-β-D-glucopyranosyl-(1→2)]-O-β-D-glucopyranosyl-
(1→4)-O-β-D-galactopyranoside} (9),13 furcreastatin (10),14
Received: July 29, 2016
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Chart 1
hecogenin-3-O-{α-L-rhamnopyranosyl-(1→4)-O-β-D-glucopyr-
anosyl-(1→3)-O-[β-D-glucopyranosyl-(1→2)]-O-β-D-glucopyr-
anosyl-(1→4)-O-β-D-galactopyranoside} (11),15 rockogenin-3-
O-{α-L-rhamnopyranosyl-(1→4)-O-β-D-glucopyranosyl-(1→
3)-O-[β-D-glucopyranosyl-(1→3)-O-β-D-glucopyranosyl-(1→
2)]-O-β-D-glucopyranosyl-(1→4)-O-β-D-galactopyranoside}
(12),15 smilagenin-3-O-[β-D-glucopyranosyl-(1→3)-O-β-D-gal-
actopyranoside],16 agameroside-1 (13),17 chlorogenin-3-O-[β-
D-glucopyranosyl]-6-O-[β-D-glucopyranoside],18 cantalasapo-
nin-1,19 and chrysogenin 3-O-[β-D-fucopyranosyl]-6-O-[β-D-
glucopyranoside] (14).20
The structures of the new compounds were elucidated on the
basis of spectroscopic data obtained by 1D and 2D NMR
experiments (Tables 1 and 2), HRTOFESIMS spectra, and acid
hydrolysis. The absolute configurations of the sugar compo-
nents were determined by a slight modification of the method
reported by Tanaka et al.21 Briefly, sugars were converted into
the thiazolidine derivatives and then into the arylthiocarbamate
using L-cysteine methyl ester and o-tolylisothiocyanate. The
reaction mixture was then directly analyzed by reversed-phase
HPLC-UV, and the retention times (tR) were compared with
values obtained for derivatives of authentic sugar samples (D-
glucose, D-galactose, and L-rhamnose) with D- and L-cysteine
methyl ester. In this way, D-galactose, D-glucose, and L-
rhamnose were identified.
Compound 1 was isolated as an amorphous powder and
exhibited a deprotonated molecular ion at m/z 1385.6241 ([M
− H]−, calcd 1385.6225) in the negative-ion mode
HRTOFESIMS, and this corresponded to a molecular formula
of C63H102O33. This molecular formula is identical to that of the
main compound isolated from F. hexapetala leaves, furcreastatin
(10), indicating that these two compounds are isomers. The 1H
NMR spectrum of 1 contained six anomeric proton signals at δ
4.82 (d, J = 7.5 Hz), 5.08 (d, J = 7.7 Hz), 5.10 (d, J = 8.0 Hz),
5.16 (d, J = 7.8 Hz), 5.52 (d, J = 7.6 Hz), and 5.75 (br s), and
these showed correlations in the HSQC spectrum with carbons
resonating at δ 102.4, 105.6, 104.7, 104.2, 104.0, and 102.6,
respectively. Individual sugar units were identified by a
combination of one- and two-dimensional NMR experiments,
and their absolute configurations were determined under the
conditions described above. The sequence of the sugars in the
chain was established by means of interglycosidic HMBC/
ROESY correlations as a glycosidic chain formed by six sugar
units identical to that in compound 10.
1
The H NMR spectrum of 1 showed a typical signal pattern
for a spirostane aglycone, with two singlet signals for tertiary
methyl groups at δ 1.19 and 0.65 and two doublet signals for
secondary methyl groups at 1.23 (d, J = 7.9 Hz) and 0.67 (d, J
1
= 6.3 Hz). The H and 13C NMR signal assignments for the
aglycone were made through an analysis of the correlations
B
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
c
1
Table 1. 13C and H NMR Spectroscopic Data (J in Hz) for the Aglycone Moieties of Compounds 1−4 (Pyridine-d5)
a
b
b
b
furcroside A (1)
furcroside B (2)
furcroside C (3)
furcroside D (4)
position
δC
δH
δC
δH
δC
δH
δC
δH
d
1ax
1eq
2ax
2eq
3
36.7
0.69 m
1.29 m
1.54 (o)
1.97 m
3.84 m
1.32 m
36.6
0.69 (o)
37.1
0.77 m
1.49 (o)
1.62 m
2.02 (o)
3.91 m
1.34 m
1.77 (o)
31.0
1.66 (o)
1.47 (o)
1.53 (o)
1.98 m
1.27 m
1.54 (o)
1.96 m
3.84 m
1.31 (o)
29.6
29.6
30.0
27.0
77.1
34.6
77.1
34.6
77.3
34.8
74.4
30.7
4.38 (o)
1.78 br d (12.4)
1.66 (o)
1.94 (o)
4ax
4eq
5
1.77 br d (12.0)
0.82 (o)
1.77 br d (12.3)
0.83 (o)
44.4
28.6
44.4
28.5
44.6
28.9
0.88 dddd (12.2, 12.2, 2.9,
2.9)
36.8
27.0
6ax
6eq
7ax
1.10 (2H) (o)
0.73 m
1.09 (2H) (o)
0.73 m
1.04 m
1.08 m
0.76 m
1.69 m
1.03 (o)
31.6
31.6
32.4
26.7
0.92 dddd (13.4, 13.4, 13.4,
4.4)
7eq
8
1.50 (o)
1.69 (o)
0.84 (o)
1.52 (o)
1.69 (o)
0.86 (o)
1.49 (o)
1.24 (o)
1.46 (o)
1.25 (o)
34.2
55.8
36.2
37.8
34.7
55.5
36.2
37.7
35.2
54.4
35.8
21.2
1.38 (o)
35.5
40.2
35.2
21.1
9
0.47 ddd (12.3, 11.6, 4.1)
10
11ax
2.36 dd (13.5, 13.4)
2.15 dd (13.4, 4.4)
2.36 dd (14.6, 13.8)
2.20 dd (14.6, 4.5)
1.16 dddd (13.2, 12.8, 12.3,
3.8)
1.13 (o)
11eq
12ax
12eq
13
1.36 (o)
1.01 (o)
1.63 (o)
1.32 m
213.0
212.5
40.1
40.3
1.02 (o)
1.64 (o)
56.2
57.4
31.4
56.9
55.9
32.0
40.7
56.4
32.1
40.9
56.4
32.1
14
1.22 (o)
1.59 m
1.33 (o)
0.99 (o)
1.38 (o)
2.00 (o)
1.04 (o)
15ax
15eq
1.63 (o)
1.39 ddd (13.5, 12.2, 6.3)
2.02 ddd (12.4, 5.8, 7.7)
2.06 ddd (13.4, 7.2,
7.2)
2.13 ddd (12.5, 7.3, 7.3)
16
79.7
53.6
16.1
11.7
46.4
12.2
4.51 (o)
79.9
52.5
15.0
11.7
151.6
4.65 (o)
81.1
63.0
16.6
12.3
41.9
15.0
4.53 (o)
81.2
63.1
16.5
23.8
42.0
15.0
4.60 ddd (7.9, 7.9, 6.2)
1.84 dd (6.8, 8.1)
0.79 s
17
2.87 dd (9.2, 6.5)
1.19 s
3.64 ddd (7.6, 2.0,2.0)
0.99 s
1.77 dd (6.8, 8.4)
0.80 s
18
19
0.65 s
0.64 s
0.62 s
0.79 s
20
2.81 quin. (8.1)
1.23 d (7.9)
1.93 dq (7.0, 6.8)
1.11 d (7.0)
1.94 dq (6.9, 6.8)
1.14 d (6.9)
21a
21b
22
110.3 5.39 br s
5.54 br s
108.4
31.0
107.1
109.2
31.8
109.2
31.8
23ax
1.69 (2H) (o)
31.7
1.88 ddd (13.3, 13.2,
4.9)
1.63 m
1.64 (o)
23eq
24
1.67 (o)
1.68 m
1.68 (o)
28.9
30.8
68.1
1.57 (2H) (o)
1.58 (o)
29.0
31.5
67.6
1.58 (2H) (o)
1.66 (o)
29.2
30.6
66.8
1.54 (2H) (o)
1.55 (o)
29.2
30.6
66.8
1.55 (2H) (o)
1.56 (o)
25
26ax
26eq
27
3.56 (2H) (o)
3.52 dd (11.1, 11.1)
3.57 dd (11.1, 4.3)
0.68 d (6.5)
3.48 dd (10.6, 10.6)
3.56 dd (10.6, 3.7)
0.67 d (5.7)
3.50 dd (10.6, 10.5)
3.58 dd (10.6, 3.6)
0.67 d (5.4)
17.3
0.67 d (6.3)
17.3
17.3
17.3
a
b
Data were measured at 150 MHz (13C NMR) and 600 MHz (1H NMR). Data were measured at 125 MHz (13C NMR) and 600 MHz (1H NMR).
c
d
Assignments were confirmed by 1H−1H-COSY, 2D-TOCSY, HSQC, HSQC-TOCSY, and HMBC experiments. o: overlapped with other signals.
determined in HSQC, HMBC, 1H−1H COSY, and 2D-TOCSY
experiments.
As in furcreastatin (10), the 25R configuration for compound
1 was deduced according to Agrawal’s rule, which establishes a
25R configuration when the difference between the chemical
shifts of geminal protons in the methylene groups H2-23, H2-
24, and H2-26 (Δab = δa − δb) is less than 0.20.22 Moreover, the
typical A/B, B/C, and C/D “trans” and D/E “cis” junctions for
spirostanes were determined by an exhaustive study of
correlations in the 2D-ROESY experiment. Major differences
in the chemical shifts compared with compound 10 were
observed for the E/F rings, especially the notable deshielding of
H-20 (δ 2.81) in compound 1. A 1D-ROESY spectrum
acquired by the selective excitation of the proton at δ 4.51 (H-
16) showed correlations with H-17/H-15α, which is consistent
with an “α” orientation of H-16. Likewise, the effects observed
The presence of a carbonyl group was suggested by the signal
at δ 213.0 in the 13C NMR spectrum. This showed HMBC
correlations with protons at δ 2.36 and 2.15 (2H-11), 2.87 (H-
17), and 1.19 (H-18) of compound 1, consistent with its
location at C-12. Likewise, the sugar chain attached at C-3 of
the aglycone moiety was established by correlation of the
carbon signal at δ 77.1 (C-3) in the HMBC spectrum. Although
the functionalization of the aglycone was shown to be the same
as that in compound 10, significant differences were observed
in the spectroscopic data, thus indicating that the two
compounds are diastereoisomers.
C
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
c
1
,
Table 2. 13C and H NMR Spectroscopic Data (J in Hz) of the Sugar Portions of Compounds 1−4 (Pyridine-d5)
a
b
b
b
furcroside A (1)
furcroside B (2)
furcroside C (3)
furcroside D (4)
position
δC
δH
δC
δH
δC
δH
δC
δH
β-D-Gal
β-D-Gal
β-D-Gal
β-D-Gal
1
2
3
4
5
6
102.4
73.1
75.6
79.9
75.5
60.6
4.82 d (7.5)
102.4
73.1
75.6
79.7
75.5
60.6
4.84 d (7.6)
4.38 dd (9.5, 7.6)
4.08 (o)
102.3
73.2
75.6
80.8
75.1
60.5
4.88 d (7.7)
103.8
72.6
75.1
70.0
74.6
69.5
4.77 d (7.7)
d
4.39 (o)
4.44 ddd (10.0, 7.7, 4.3)
4.09 (o)
4.43 dd (7.7, 9.6)
4.08 dd (9.6, 3.5)
4.46 br d (3.5)
4.14 ddd(6.2, 6.2, 1.2)
4.43 m
4.09 (o)
4.56 br s
3.97 (o)
4.17 (o)
4.55 br d (3.0)
3.98 (o)
4.56 br d (3.3)
4.01 (o)
4.20 (o)
4.23 (o)
4.64 br d (8.2)
β-D-Glc
4.64 (o)
4.75 ddd (10.2, 10.2, 7.5)
β-D-Glc
4.66 dd (10.7, 5.7)
β-D-Glc
β-D-Glc
1
2
3
4
5
6
104.7
80.8
88.1
70.6
77.4
62.9
5.10 d (8.0)
4.30 dd (8.0, 8.5)
4.10 (o)
104.7
80.8
88.1
70.6
77.4
62.9
5.11 d (7.9)
4.30 dd (7.9, 8.8)
4.09 dd (8.8, 8.7)
3.72 dd (8.7, 9.4)
3.78 (o)
105.0
85.8
78.3
71.8
78.1
63.2
5.11 d (7.7)
4.12 dd (7.7, 8.6)
4.21 (o)
105.1
75.2
78.5
71.7
78.4
62.7
5.07 d (7.8)
4.06 dd (7.8, 8.6)
4.24 (o)
3.71 dd (8.5, 8.5)
3.78 (o)
4.20 (o)
4.24 (o)
3.94 m
3.92 ddd (8.5, 5.4, 2.5)
4.34 (o)
3.96 (o)
3.96 (o)
4.09 (o)
4.41 (o)
4.41 (o)
4.60 dd (10.2; 6.2)
4.49 br d (11.4)
β-D-Glc′
β-D-Glc′
β-D-Glc′
1
2
3
4
5
6
104.2
75.6
76.4
78.0
77.2
61.0
5.16 d (7.8)
3.95 dd (7.8, 8.2)
4.06 dd (9.8, 9.0)
4.33 (o)
104.2
75.6
76.4
78.0
77.2
61.1
5.16 d (7.8)
3.95 dd (7.8, 8.7)
4.05 dd (9.0, 8.7)
4.32 dd (9.0, 8.0)
3.77 (o)
106.2
87.2
68.4
75.4
78.4
61.3
5.14 d (6.9)
4.04 dd (6.9, 7.7)
4.08 (o)
4.11 (o)
3.77 (o)
3.72 ddd (9.6, 2.9, 2.7)
4.24 (o)
4.03 (o)
4.03 (o)
4.20 (o)
4.20 (o)
4.48 m
β-D-Glu″
β-D-Glu″
β-D-Glu″
1
2
3
4
5
6
104.0
74.7
88.0
69.3
77.9
62.0
5.52 d (7.6)
104.0
74.7
88.1
69.3
77.9
62.0
5.52 d (7.7)
105.4
78.6
78.2
71.6
75.5
62.5
5.22 d (7.7)
4.00 dd (7.7, 8.6)
4.21 dd (8.6, 9.6)
4.15 dd (9.6, 9.1)
4.01 (o)
4.04 (o)
4.05 (o)
4.02 dd (9.2, 8.6)
4.06 (o)
4.02 dd (9.0, 7.8)
4.06 (o)
3.76 (o)
3.76 (o)
4.23 (o)
4.24 (o)
4.28 br d (12.0)
4.51 m
4.40 (o)
4.41 (o)
β-D-Glu‴
β-D-Glu‴
1
2
3
4
5
6
105.6
75.3
78.0
71.5
78.5
62.4
5.08 d (7.7)
105.6
75.3
78.1
71.5
78.5
62.5
5.07 d (7.8)
3.98 dd (8.5, 7.8)
4.15 (o)
3.98 (o)
4.15 dd (8.5, 9.4)
4.12 (o)
4.13 (o)
3.88 ddd (9.0, 6.2, 2.1)
4.22 (o)
3.87 m
4.23 (o)
4.45 (o)
4.46 (o)
α-L-Rha
α-L-Rha
1
2
3
4
5
6
102.6
72.5
72.7
73.9
70.3
18.5
5.75 br s
102.6
72.5
72.7
73.9
70.4
18.5
5.76 br s
4.59 br s
4.61 br s
4.51 br d (7.2)
4.31 dd (7.2, 8.6)
4.90 dq (6.2, 8.6)
1.67 d (6.2)
4.50 br d (9.0)
4.30 (o)
4.91 (o)
1.67 d (6.2)
a
b
Data were measured at 150 MHz (13C NMR) and 600 MHz (1H NMR). Data were measured at 125 MHz (13C NMR) and 600 MHz (1H NMR).
c
d
Assignments were confirmed by 1H−1H-COSY, 2D-TOCSY, HSQC, HSQC-TOCSY, and HMBC experiments. o: overlapped with other signals.
in 1D-ROESY between H-16 and H-26ax suggested an “R”
H-18/H-23 confirmed a “β” disposition for this methyl group
and an “R” configuration for C-20 (Figure 1). Spirostane
saponins with different configurations at C-20 and C-22 were
synthesized by Tobari et al.,23 and chemical shifts greater than
2.47 ppm were observed for the H-20 signal, which means that
H-20 and the oxygen atom of the F ring were in a relative cis
disposition. This situation was consistent with the experimental
data obtained for compound 1.
configuration for the spiroketalic carbon C-22, as in
furcreastatin (10). An additional correlation was observed
between H-16 and H-20, which might suggest the epimeriza-
tion of C-20. This inference was confirmed by the 1D-ROESY
spectra acquired after the selective excitations of signals at δ
2.81 (H-20) and 1.23 (Me-21). In this way, correlations
between H-20 and H-16/H-17 indicated an “α” orientation for
H-20, while correlations between the methyl group at H-21 and
D
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
established as (25R)-3β-hydroxy-5α-spirost-20(21)-en-12-one
3-O-{α-L-rhamnopyranosyl-(1→4)-O-β-D-glucopyranosyl-(1→
3)-O-[β-D-glucopyranosyl-(1→3)-O-β-D-glucopyranosyl-(1→
2)]-O-β-D-glucopyranosyl-(1→4)-O-β-D-galactopyranoside}.
This compound has not been described previously, and its
trivial name proposed is furcroside B.
Compounds 1 and 2 were minor derivatives of furcreastatin
(10), where furcroside B (2) would be an intermediate
metabolite for the epimerization of furcreastatin (10) to
furcroside A (1), thus suggesting that an oxido-reduction is
involved in the metabolism of these compounds.
Compound 3 was isolated as a white, amorphous powder and
gave the molecular formula C51H84O23 (HRTOFESIMS, m/z
1063.5310 [M − H]−, calcd 1063.5325). The 1H NMR
spectrum of 3 was consistent with that of a spirostane saponin.
A detailed study of the 1D and 2D NMR spectra led the
aglycone moiety of compound 3 to be identified as tigogenin,
the same as that in compounds 5−7. The 1H NMR spectrum of
3 showed four anomeric signals at δ 4.88 (d, J = 7.7 Hz), 5.11
(d, J = 7.7 Hz), 5.14 (d, J = 6.9 Hz), and 5.22 (d, J = 7.7 Hz),
and these were correlated in the HSQC spectrum with carbons
at δ 102.3, 105.0, 106.2, and 105.4, respectively. Sugar units
were identified by 1D-TOCSY and 1D-ROESY experiments
involving selective excitation of each anomeric proton. Selective
1D-TOCSY experiments on the signals at δ 5.11, 5.14, and 5.22
showed a typical spin system for a β-glucopyranosyl moiety.
The anomeric signal at δ 4.88 showed a different TOCSY
pattern, in which the coupling constants were consistent with
trans-diaxial arrangements for H-1/H-2 and H-2/H-3. The
relatively small coupling constant between H-3 and H-4
indicated the equatorial position of H-4. The NOE correlations
with H-3 and H-5 observed in the 1D-ROESY spectra indicated
the presence of a β-galactopyranosyl moiety. Finally, HSQC
and HSQC-TOCSY experiments showed unambiguously the
complete correlations for proton and carbon signals of the
tetrasaccharide portion. The absolute configurations of sugars
were determined under the conditions described above. A study
of the HMBC and ROESY correlations between H-1Glc″ (δ
5.22) and C-2Glc′ (δ 87.2)/H-2Glc′ (δ 4.04), H-1Glc′ (δ 5.14) and
C-2Glc (δ 85.8)/H-2Glc (δ 4.12), H-1Glc (δ 5.11) and C-4Gal (δ
80.8)/H-4Gal (δ 4.56), and H-1Gal (δ 4.88) and C-3 (δ 77.3)/H-
3 (δ 3.91) of tigogenin allowed the sequence of the sugar to be
established. Thus, structure 3 was determined as tigogenin-3-O-
{β-D-glucopyranosyl-(1→2)-O-β-D-glucopyranosyl-(1→2)-O-β-
D-glucopyranosyl-(1→4)-O-β-D-galactopyranoside}. This sapo-
nin has not been described previously and has been named
furcroside C.
Figure 1. Key NOE correlations that show the β-orientation of
methyl-21 for compound 1. Correlations observed from the 1D-
ROESY spectrum performed by selective excitation of proton H-16 (δ
4.51, black arrows), H-20 (δ 2.81, blue arrows), and H-21 (δ 1.23,
green arrows).
Consequently, the aglycone moiety of compound 1 was
elucidated as 20-isohecogenin.24 The (20R) isomers of
spirostane sapogenins have usually been considered as
“unnatural”, and they are often obtained by synthetic methods.
However, in the literature, the (20R, 22S) configuration is
noted for brodioside B, a saponin isolated from Brodiaea
californica tubers,25 and for two steroids from Dracaena
cambodiana.26 To the best of our knowledge, compound 1 is
the first (20R,22R)-spirostanol glycoside to have been isolated
from a natural source. Thus, the structure of 1 was
characterized as (20R,22R,25R)-3β-hydroxy-5α-spirostan-12-
one 3-O-{α-L-rhamnopyranosyl-(1→4)-O-β-D-glucopyranosyl-
(1→3)-O-[β-D-glucopyranosyl-(1→3)-O-β-D-glucopyranosyl-
(1→2)]-O-β-D-glucopyranosyl-(1→4)-O-β-D-galactopyrano-
side}. This compound has not been reported previously, and it
has been named furcroside A.
Compound 2 was obtained as a white, amorphous powder.
The molecular formula was established as C63H100O33 by
HRTOFESIMS at m/z 1383.6047 ([M − H]−, calcd
1383.6069). The NMR features were typical of a spirostane
saponin, with signals from the sugar portion almost super-
imposable on those of compound 1. Scrutiny of the 1D-
TOCSY, 1D-ROESY, and HMBC spectra confirmed a sugar
chain identical to that of furcroside A (1) and furcreastatin
(10). Relative to the aglycone moiety, most of the differences in
chemical shifts between compounds 1 and 2 were observed for
1
rings D and E. The H NMR spectrum of 2 contained two
singlet signals for tertiary methyl groups at δ 0.99 and 0.64 and
only one doublet signal for a secondary methyl group at δ 0.68
(d, J = 6.5 Hz), which suggested the lack of a methyl group.
Two broad singlet signals were observed at δ 5.54 and 5.39,
typical of geminal olefinic protons, and these showed
correlations in the HSQC spectrum with the carbon at δ
110.3 (CH2). This observation, together with a carbon signal at
δ 151.6 corresponding to a quaternary carbon, suggested the
presence of an exomethylene group. The HMBC correlations
of the two olefinic protons with C-17 (δ 52.5) and C-22 (δ
107.1), along with an NOE correlation with H-17 (δ 3.64) and
H-23ax (δ 1.88), supported a double bond between C-20 and C-
21.
Thus, the aglycone of 2 was elucidated as 20-dehydroheco-
genin. The acetate derivative was synthesized by Tanabe and
Peters27 and is reported here for the first time from a natural
source. Only one 20(21)-dehydrospirosteroid has been
reported previously, and this was part of a saponin isolated
from Cordyline stricta.28 Accordingly, the structure of 2 was
Compound 4 was isolated as a white, amorphous powder,
with a molecular formula of C39H64O13 (HRTOFESIMS, m/z
763.4236 [M + Na]+, calcd 763.4245). The 1H NMR spectrum
contained signals corresponding to a steroidal glycoside with
two anomeric protons at δ 4.77 (d, J = 7.7 Hz) and 5.07 (d, J =
7.8 Hz) and four methyl groups at δ 1.14 (d, J = 6.9 Hz), 0.67
(d, J = 5.4 Hz), 0.79 (s), and 0.79 (s). From a complete
1
assignment of the H and 13C NMR signals of the aglycone
moiety of 4, an oxygenation at C-3 was determined as the only
functional group. A “cis” junction between the A and B rings of
compound 4 was indicated by 2D-ROESY correlations
observed from H-5 (δ 1.94) to H-4ec (δ 1.66) and H-19 (δ
0.79), as well as from H-4ax (δ 1.78) to H-7ax (δ 0.92) and H-6ax
(δ 1.69). Therefore, the aglycone moiety of 4 was identified as
smilagenin. Analysis of the 1D-TOCSY, 1D-ROESY, and
1H−1H COSY spectra allowed the sugar units to be identified
E
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 2. Effects of compounds 1 and 4−14 on root growth of Lactuca sativa L. Values are expressed as percentage difference from the control and
are not significantly different with p > 0.05 for the Welch’s test. (a) Values significantly different with p < 0.01. (b) Values significantly different with
0.01 < p < 0.05.
as β-galactopyranoside and β-glucopyranoside. The connec-
tions of this diglycoside were established by HMBC/ROESY
correlations between H-1Glc (δ 5.07) and C-6Gal (δ 69.5)/H2-
6Gal (δ 4.43, 4.66) and between H-1Gal (δ 4.77) and C-3 (δ
74.4)/H-3 (δ 4.38) of smilagenin. Consequently, the structure
of 4 (furcroside D) was formulated as smilagenin-3-O-{β-D-
glucopyranosyl-(1→6)-O-β-D-galactopyranoside}, which has
not been described before.
The isolated pure compounds 1 and 4−14 were evaluated for
their phytotoxic activity against Lactuca sativa L. as a standard
target species9 at 333, 100, 33, 10, 3.3, and 1 μM. The isolated
compounds chlorogenin-3-O-[β-D-glucopyranosyl]-6-O-[β-D-
glucopyranoside]18 and cantalasaponin-119 were excluded
from this test because they had been assayed previously
under identical conditions,7,8 and compounds 2, 3, and
smilagenin-3-O-[β-D-glucopyranosyl-(1→3)-O-β-D-galactopyra-
Figure 3. Dendogram of compounds tested in the phytotoxicity
noside]16 could not be assayed due to the small amounts
bioassay.
available.
The results of the bioassay are shown in Figure 2, where data
are presented as percentage differences from the control.
Positive values indicate stimulation and negative values
Table 3. Phytotoxicity of Compounds 1 and 4−14 on the
Roots of Lactuca sativa
compound
IC50 (μM)
r2
represent inhibition of the studied parameters. The effects on
root development were very significant, and the discussion is
therefore focused on the root growth effects. Despite the lack of
a high number of considered variables, a multivariate approach
was performed, in order to facilitate the discussion and organize
somehow the obtained data. The dendrogram resulting from
hierarchical cluster analysis of the saponins tested is shown in
Figure 3, in which two main clusters can be seen (S1 and S2).
Cluster S1 comprises compounds 1 and 10−12, which
exhibited the best inhibition values. Comparison of the
phytotoxic activities of 1 (IC50 37.8 μM), 10 (IC50 50.7 μM),
and 11 (IC50 51.4 μM), all of which have a carbonyl group at
C-12, and 12 (IC50 80.3 μM), with a hydroxy group at C-12,
suggests that the presence of an oxygenated functional group at
C-12 plays a key role in the activity displayed, with higher
inhibition values obtained for compounds with a carbonyl
group. In fact, compounds 5 (IC50 355.6 μM) and 6 (IC50
638.7 μM), which have identical structures to 10 and 11 but
lack oxygenated positions in the spirostanol, showed lower
levels of inhibition.
1
37.8
0.9915
n.d.
a
4
n.d.
b
5
355.6
0.9820
0.9772
n.d.
b
6
638.7
7
n.d.
b
8
354.9
0.9386
0.9677
0.9084
0.9590
0.9203
n.d.
9
230.8
50.7
51.4
80.3
n.d.
10
11
12
13
14
Logran
n.d.
n.d.
b
439.7
0.9949
a
b
n.d: not determined. The data were not adjusted to the dose−
response curve.
Compound 5 only has C-3 oxygenation, while compound 9 has
another hydroxy group at position C-6 of the spirostanol.
Compound 8 (IC50 354.9 μM) has a three-sugar glycosidic
chain and a hydroxy group at position C-2. On the other hand,
comparison of the data for compounds 7 (no significant
inhibition) and 8 (IC50 354.9 μM), which differ only in the
hydroxy group at C-2, suggested a moderate influence of the
hydroxy group when located in positions other than C-12. A
The second cluster (S2) could be divided into two
subclusters, S2a and S2b. The first included compounds 5, 8,
and 9 and also exhibited good inhibition values (Table 3).
Compounds 5 (IC50 355.6 μM) and 9 (IC50 230.8 μM) share
the same chain of six sugars but have a different aglycone.
F
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 4. Comparison of the structures for compounds 10 and 5. The upper part of the figure depicts the lowest energy conformers. Below these is a
representation of the spatial distribution of hydrophobic (blue) and hydrophilic (red) regions on the molecular surface.
detrimental effect on the activity was observed on decreasing
the size of the oligosaccharide portion for compounds 5−7,
which have the same aglycone moiety (tigogenin). The
aforementioned data suggest, therefore, that the length of the
glycoside chain and the presence of hydroxy groups in the
aglycone could also be key factors for activity. Finally,
subcluster S2b contained compounds 4, 6, 7, 13, and 14,
which did not exhibit significant activity. This lack of activity is
probably due to the lack of a long chain of sugars and/or the
absence of an oxygenated functional group on the aglycone
moiety.
crucial for activity. Thus, the results outlined above may
support a further investigation on the phytotoxic activity and
mechanism of action of these saponins against other species of
interest, especially weed species, before they can be considered
as potential natural herbicides.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured on a PerkinElmer model 241 polarimeter (589 nm, 20
°C). 1D and 2D NMR spectra were recorded on an Agilent 600 DD2
spectrometer and an Agilent 500 DD2 spectrometer equipped with a 5
1
mm H {15N−31P} PFG high-field inverse detection z-gradient probe.
Among the steroidal saponins tested, greater activity was
observed for compounds with an oxygenated functional group
at C-12. Molecular modeling was carried out with the aim of
studying the influence of this feature in the molecular structure.
A comparison of the minimized energy structures for
compounds 10 (IC50 50.7 μM) and 5 (IC50 355.6 μM),
which differ only in a carbonyl group at C-12, can be seen in
Figure 4. It can be seen from the lowest energy conformers of
the two compounds that the presence of the carbonyl group
does not significantly alter the structure of the aglycone or
sugar chain. Furthermore, it can be observed that the adopted
conformation of the sugar moieties precludes interactions
between the carbonyl group and the sugar units. This
conformational study showed that the carbonyl group is
accessible in space. Likewise, this functionalization results in a
different distribution of hydrophobicity for ring C of the
aglycone (Figure 4). Therefore, this change could affect the
interaction with plant material and provides a possible
explanation for the difference in activity observed.
1H (599.772 and 499.719 MHz) and 13C (150.826 and 125.666 MHz)
NMR spectra were recorded in pyridine-d5 at 25 °C, and chemical
shifts are given on the δ scale referenced to residual pyridine (δH 8.70,
7.55, 7.18 and δC 149.84, 135.50, 123.48). Adiabatic pulse sequences
using gradients were applied, and all 2D spectra, except for HMBC,
were recorded in the phase-sensitive mode. Exact masses were
measured on a UPLC-QTOF ESI (Waters Synapt G2, Manchester,
UK) high-resolution mass spectrometer (HRTOFESIMS). Mass
spectra were recorded in the negative- or positive-ion mode in the
range m/z 100−2000, with a mass resolution of 20 000 and an
acceleration voltage of 0.7 kV. HPLC in the isocratic mode was
performed on a Merck Hitachi apparatus equipped with a LaChrom
(L-2490) refractive index detector and a LaChrom (L-2400)
ultraviolet detector, using an analytical Phenomenex Gemini C18
column (4.6 × 250 mm, i.d.).
Plant Material. Leaves of Furcraea hexapetala were collected in
March 2010 and were authenticated by botanist D. Alfredo Noa, in the
neighborhood of Ranchuelo, in the province of Villa Clara, Cuba. A
voucher specimen was deposited in the Herbarium Dr. Alberto Alonso
Triana of the Universidad Central “Marta Abreu” de Las Villas, Cuba
(number 10407 ULV).
In conclusion, on the basis of the results obtained and in
agreement with our previous studies,7,8 the data obtained
indicate that steroidal saponins with four or more sugar units in
the glycosidic chain attached at C-3 can exhibit phytotoxic
activity on L. sativa. In general, this activity increases with the
length of the glycosidic chain and the presence of oxygenated
functional groups in the aglycone moiety, especially when such
a group is located at carbon 12, with 12-ketosaponins being the
most active. The combination of these two features seems to be
Extraction and Isolation. Dried and powdered leaves (1 kg) were
extracted three times with ethanol−H2O (7:3) for 48 h by maceration
at room temperature. The solvent was removed under reduced
pressure, and the syrupy residue (16%) was suspended in distilled
water, defatted with n-hexane, and then extracted with H2O-saturated
n-BuOH. After removing the solvent, 15 g of the n-BuOH extract
(11% of an ethanolic extract) was purified by VLC on LiChrospher
RP-18 and eluted with mixtures of Me2CO−H2O to give nine fractions
(F1: 1.35 g, F2: 4.69 g, F3: 1.97 g, F4: 1.27 g, F5: 2.45 g, F6: 1.14 g,
F7: 0.36 g, F8: 0.25 g, and F9: 1.11 g).
G
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Fraction F4 was subjected to MPLC on a Buchi 861 apparatus with
phytotoxicity bioassays with the dicot L. sativa L. (lettuce) as standard
target species.9 Both procedures were conducted under the conditions
reported previously by us.7 Control samples (buffered aqueous
solutions without any test compound) and the commercial herbicide
Logran, a combination of N-(1,1-dimethylethyl)-N′-ethyl-6-(methyl-
thio)-1,3,5-triazine-2,4-diamine (Terbutryn, 59.4%) and 2-(2-chlor-
oethoxy)-N-{[(4-methoxy-6-methyl-1,3,5-triazin-2-yl)amino]-
carbonyl}benzene-sulfonamide (Triasulfuron, 0.6%), were used as
internal references9 and were tested under the same conditions as the
samples.
The evaluated parameters in the phytotoxicity assay (germination
rate, root length, and shoot length) were recorded using a Fitomed
system,29 which allowed automatic data acquisition and statistical
analysis using its associated software. Data were analyzed statistically
using Welch’s test, with significance fixed at 0.01 and 0.05. Results are
presented as percentage differences from the control. Zero represents
the control, positive values represent stimulation, and negative values
represent inhibition. The concentration that resulted in a 50%
inhibition (IC50 values) was calculated from the dose−response curve.
The dendogram was obtained by using Statistica 7.0 software.30
Numerical clustering of bioassay data was carried out based on
percentage differences from control, using the root length parameter.
An average linkage was applied as an amalgamation criterion, and the
distance measurement (dissimilarity coefficient) was based on
Euclidean distances.
Method for Molecular Modeling Calculations. Calculations of
minimum energy conformers were performed using PCModel 9.2
software.31 The conformers created from molecular mechanics
GMMX calculations were refined, and those with energies higher
than 3.5 kcal/mol with respect to minima were not considered. The
resultant conformers were used as the starting point for semiempirical
calculations. These conformers were subsequently minimized using
PM3 calculations by HyperChem 8.0.3 software.32 3D molecular
models were constructed from lowest energy conformers using the
same software.
̈
a column filled with 40−63 μm of LiChrospher RP-18, using Me2CO−
H2O (1:1) as mobile phase. Six milliliter fractions were collected and
were checked by TLC on RP-18 F254S, developed with MeOH−H2O
(8:2), then sprayed with oleum reagent and heated at 150 °C.
Fractions with similar profiles were combined to give seven
subfractions. Subfraction F4-3 was separated by HPLC on an
analytical C18 column (1 mL/min) and MeOH−H2O (7:3) or
CH3CN−H2O (3.5:6.5) as mobile phase to yield compounds 1 (4.9
mg), 2 (2.5 mg), 12 (9.5 mg), 13 (8.6 mg), cantalasaponin-1 (8.1 mg),
and 14 (1.5 mg). Compound 10 (1.6 g) was obtained as a precipitate
from the Me2CO solution of F5. Fraction F6 was diluted in MeOH
and filtered to remove the precipitate. The solution was then subjected
to column chromatography (CC) on silica gel and eluted with a
stepwise gradient of CHCl3−MeOH−H2O (12:6:1 to 7:4:1) to yield
12 fractions. Subfractions containing the major saponins were
chromatographed by HPLC under the same conditions as described
above, using Me2CO−H2O (6:4) as the eluent, to afford compounds 5
(3.8 mg), 6 (3.6 mg), 8 (3.5 mg), 11 (6.3 mg), smilagenin-3-O-[β-D-
glucopyranosyl-(1→3)-O-β-D-galactopyranoside] (1 mg), and chlor-
ogenin-3-O-[β-D-glucopyranosyl]-6-O-[β-D-glucopyranoside] (6.0
mg). Finally, compounds 3 (2.5 mg), 4 (3.5 mg), and 7 (2.6 mg)
were purified using CH3CN−H2O (4.5:5.5) as mobile phase. Fraction
F7 was subjected to CC on silica gel with CHCl3−MeOH−H2O
(12:6:1 to 7:4:1) and purified by HPLC on an analytical C-18 column
with Me2CO−H2O (5.5:4.5) to furnish 9 (4.1 mg).
Furcroside A (1): white, amorphous solid; [α]20 −47.0 (MeOH, c
D
1
0.1); for H and 13C NMR data, see Tables 1 and 2; HRTOFESIMS,
m/z 1385.6241 [M − H]− (calcd for C63H101O33, 1385.6225).
Furcroside B (2): white, amorphous solid; [α]20 −27.0 (MeOH, c
D
0.1); for H and 13C NMR data, see Tables 1 and 2; HRTOFESIMS,
1
m/z 1383.6047 [M − H]− (calcd for C63H99O33, 1383.6069).
Furcroside C (3): white, amorphous solid; [α]20 −35.5 (MeOH, c
D
0.1); for H and 13C NMR data, see Tables 1 and 2; HRTOFESIMS,
1
m/z 1063.5310 [M − H]− (calcd for C51H83O23, 1063.5325).
Furcroside D (4): white, amorphous solid; [α]20 −39.6 (MeOH, c
D
0.1); for H and 13C NMR data, see Tables 1 and 2; HRTOFESIMS,
ASSOCIATED CONTENT
* Supporting Information
1
■
m/z 763.4236 [M + Na]+ (calcd for C39H64O13Na, 763.4245).
Acid Hydrolysis of Compounds 1−4 and Determination of
Sugar Absolute Configuration. Compounds 1−4 (2 mg each)
were treated with 2 M HCl in 1,4-dioxane−H2O (1:1, v/v, 2 mL) at 95
°C for 4 h. After cooling, the solvent was removed under reduced
pressure. The dried residue was suspended in water, and aglycones
were extracted with ethyl acetate (4 × 2 mL). The aqueous layer
containing sugars was neutralized with Amberlite IR-45 (OH− form),
dried under reduced pressure, and stored prior to analysis. The
absolute configurations of the monosaccharide constituents of
compounds 1−4 were determined according to the method reported
by Tanaka et al.21 with slight modifications. Sugars from each sample
were dissolved in pyridine (0.5 mL) containing L-cysteine methyl ester
hydrochloride (1 mg) and heated at 60 °C for 1 h; o-tolyl
isothiocyanate (2 μL) was then added, and the mixture was heated
at 60 °C for 1 h. Each reaction mixture was analyzed by reversed-phase
HPLC using a Merck Hitachi apparatus equipped with a LaChrom (L-
2400) UV detector and analytical Phenomenex Gemini C18 column
(4.6 × 250 mm, i.d): mobile phase: CH3CN−H2O (2.5:7.5)
containing 50 mM H3PO4; flow rate: 1 mL/min; detection: UV
(250 nm). The derivatives of monosaccharides of D-galactose, D-
glucose, and L-rhamnose, obtained from sugar hydrolysates, were
identified by comparison of their retention times (tR) with those of
authentic samples (Sigma-Aldrich, Steinheim, Germany) treated in the
same way as described above. The tR of L-galactose, L-glucose, and D-
rhamnose was obtained by reaction of D-galactose, D-glucose, and L-
rhamnose with D-cysteine methyl ester.21 Retention times of the
derivatives were as follows: D-galactose 15.08 min, L-galactose 16.00
min, D-glucose 17.30 min, L-glucose 15.80 min, L-rhamnose 28.36 min,
and D-rhamnose 20.23 min.
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.6b00702.
HRESIMS and 1D and 2D NMR spectra for compounds
1
1−4; complete H and 13C NMR data for compounds
5−8, 11−14, and smilagenin-3-O-[β-D-glucopyranosyl-
(1→3)-O-β-D-galactopyranoside] that have not been
reported to date completely (PDF)
AUTHOR INFORMATION
Corresponding Author
■
*Tel: +34 956 012770. Fax: +34 956 016193. E-mail:
famacias@uca.es.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This research was supported by the Ministerio de Economia y
■
́
Competitividad (MINECO) (Project AGL2013-42238-R) and
́
́
Consejeria de Economia Innovacion
́
y Ciencia, Junta de
́
Andalucia (P10 AGR-5822).
REFERENCES
■
́
(1) Alvarez de Zayas, A. Anales Inst. Biol. Univ. Nac. Auton
́
. Mex
́
ico,
Phytotoxicity Bioassay. Fractions were assayed at concentrations
of 800, 400, and 200 ppm in etiolated wheat coleoptile assays. Pure
compounds were bioassayed at 333, 100, 33, 10, 3.3, and 1 μM by
Ser. Bot. 1996, 67, 329−346.
(2) Blunden, G.; Carabot, A.; Jewers, K. Phytochemistry 1980, 19,
2489−2490.
H
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(3) Castellanos, L.; Fernan
Rev. Proteccion Veg. 2011, 26, 122−124.
(4) Castellanos, L.; Fernandez, A.; Ortega, I.; Guerra, J. O. J. Anim.
́
dez, A.; Ortega, I.; Soto, R.; Martin, C.
́
́
Plant Sci. 2011, 10, 1300−1305.
(5) Macías, F. A.; Guerra, J. O.; Simonet, A. M.; Nogueiras, C. M.
Magn. Reson. Chem. 2007, 45, 615−620.
́
(6) Macías, F. A.; Guerra, J. O.; Simonet, A. M.; Perez, A. J.;
Nogueiras, C. M. Magn. Reson. Chem. 2010, 48, 350−355.
(7) Perez, A. J.; Calle, J. M.; Simonet, A. M.; Guerra, J. O.; Stochmal,
A.; Macías, F. A. Phytochemistry 2013, 95, 298−307.
(8) Perez, A. J.; Simonet, A. M.; Calle, J. M.; Pecio, L.; Guerra, J. O.;
́
́
Stochmal, A.; Macías, F. A. Phytochemistry 2014, 105, 92−100.
(9) Macías, F. A.; Castellano, D.; Molinillo, J. M. G. J. Agric. Food
Chem. 2000, 48, 2512−2521.
(10) (a) Benidze, M. M.; Dzhikiya, O. D.; Pkheidze, T. A.;
Kemertelidze, E. P.; Shashkov, A. S. Khim. Prir. Soedin. 1987, 4, 537−
542. (b) Ahmad, V. U.; Basha, A., Eds. Spectroscopic Data of Steroid
Glycosides: Spirostanes, Bufanolides, Cardenolides; Springer: New York,
2006; Vol. 3, pp 1844−1845, 1899−1900.
(11) Shvets, S. A.; Latsterdis, N. V.; Kintia, P. K. Advances in
Experimental Medicine and Biology. In Saponins, Used in Traditional
and Modern Medicine; Plenum Press: New York, 1996; Vol. 404, pp
475−483.
(12) (a) Shvets, S. A.; Naibi, A. M.; Kintya, P. K.; Shashkov, A. S.
Khim. Prir. Soedin. 1995, 3, 391−395. (b) Ahmad, V. U.; Basha, A.,
Eds. Spectroscopic Data of Steroid Glycosides: Spirostanes, Bufanolides,
Cardenolides; Springer: New York, 2006; Vol. 3, pp 1588−1589.
(13) Ohtsuki, T.; Koyano, T.; Kowithayakorn, T.; Sakai, S.;
Kawahara, N.; Goda, Y.; Yamaguchi, N.; Ishibashi, M. Bioorg. Med.
Chem. 2004, 12, 3841−3845.
(14) Itabashi, M.; Segawa, K.; Ikeda, Y.; Kondo, S.; Naganawa, H.;
Koyano, T.; Umezawa, K. Carbohydr. Res. 2000, 323, 57−62.
(15) Kuroda, M.; Mimaki, Y.; Hasegawa, F.; Yokosuka, A.; Sashida,
Y.; Sakagami, H. Chem. Pharm. Bull. 2001, 49, 726−731.
(16) Yokosuka, A.; Jitsuno, M.; Yui, S.; Yamazaki, M.; Mimaki, Y. J.
Nat. Prod. 2009, 72, 1399−1404.
(17) Teng, R. W.; Jin, J. M.; Wang, D. Z.; Yang, C. R. Bopuxue Zazhi
2001, 18, 291−297.
(18) Sharma, S. C.; Sati, O. P. Phytochemistry 1982, 21, 1820−1821.
(19) Sati, O. P.; Pant, G. J. Nat. Prod. 1985, 48, 395−399.
(20) Yokosuka, A.; Sano, T.; Hashimoto, K.; Sakagami, H.; Mimaki,
Y. Chem. Pharm. Bull. 2009, 57, 1161−1166.
(21) Tanaka, T.; Nakashima, T.; Ueda, T.; Tomii, K.; Kouno, I.
Chem. Pharm. Bull. 2007, 55, 899−901.
(22) Agrawal, P. K. Magn. Reson. Chem. 2003, 41, 965−968.
(23) Tobari, A.; Teshima, M.; Koyanagi, J.; Kawase, M.; Miyamae,
H.; Yoza, K.; Yakasaki, A.; Nagamura, Y.; Saito, S. Eur. J. Med. Chem.
2000, 35, 511−527.
(24) Wall, M. E.; Serota, S.; Eddy, C. R. J. Am. Chem. Soc. 1955, 77,
1230−1237.
(25) Mimaki, Y.; Nagamura, O.; Sashida, Y.; Koike, K.; Nikaido, T.;
Ohmoto, A.; Satomi, Y.; Nishino, H. Chem. Pharm. Bull. 1995, 43,
971−976.
(26) Minh, C. V.; Dat, N. T.; Dang, N. H.; Nam, N. H.; Ban, N. K.;
Tuyen, N. V.; Huong, L. M.; Huong, T. T.; Kiem, P. V. Nat. Prod.
Commun. 2009, 4, 1197−1200.
(27) Tanabe, M.; Peters, R. H. J. Org. Chem. 1970, 35, 1238−1239.
(28) Mimaki, Y.; Takaashi, Y.; Kuroda, M.; Sashida, Y. Phytochemistry
1997, 45, 1229−1234.
(29) Castellano, D.; Macías, F. A.; Castellano, M.; Cambronero, R.
Span. Pat. No P9901565, 2001.
(30) STATISTICA, version 7.0; StatSoft Inc.: Tulsa, OK, USA, 2004.
(31) PCMODEL, version 9.2; Serena Software: Bloomington, IN,
USA, 2006.
(32) HyperChem, version 8.0.3; HyperCube Inc.: Gainsville, FL, USA,
2007.
I
DOI: 10.1021/acs.jnatprod.6b00702
J. Nat. Prod. XXXX, XXX, XXX−XXX