Bisdesmosidic saponins from securidaca longepedunculata roots: Evaluation of deterrency and toxicity to coleopteran storage pests

Article abstract of DOI:10.1021/jf901599j

Powdered dry root bark of Securidaca longepedunculata was mixed with maize and cowpea and effectively reduced the numbers of Sitophilus zeamais and Callosobruchus maculatus emerging from these commodities, respectively, more than 9 months after treatment.

Full text of DOI:10.1021/jf901599j

8860 J. Agric. Food Chem. 2009, 57, 8860–8867
DOI:10.1021/jf901599j
Bisdesmosidic Saponins from Securidaca longepedunculata
Roots: Evaluation of Deterrency and Toxicity to Coleopteran
Storage Pests
†
P
HILIP C. STEVENSON,*,†,‡
T
HAMARA K. DAYARATHNA
,
†,§ STEVEN R. BELMAIN
,
AND
‡
NIGEL C. VEITCH
†Natural Resources Institute, University of Greenwich, Central Avenue, Chatham Maritime, Kent ME4
4TB, United Kingdom, and ‡Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey TW9
3AB, United Kingdom. § Current address: Department of Biochemistry and the Siebens-Drake
Research Institute, Schulich School of Medicine and Dentistry, University of Western Ontario,
London, Ontario N6A 5C1, Canada.
Powdered dry root bark of Securidaca longepedunculata was mixed with maize and cowpea and
effectively reduced the numbers of Sitophilus zeamais and Callosobruchus maculatus emerging from
these commodities, respectively, more than 9 months after treatment. This effect was reciprocated in
grain treated with a methanol extract of the root bark, indicating that compounds were present that were
oviposition deterrents or directly toxic to the adults or larvae. Two new bisdesmosidic saponins, 3-O-β-
glucopyranosyl-28-O-(R- -arabinopyranosyl-(1 f 3)-β- -xylopyranosyl-(1 f 4)[β- -apiofuranosyl-(1 f 3)]-
R- -rhamnopyranosyl-(1 f 2)-[4-O-(4-methoxycinnamoyl-β- -fucopyranosyl)])-medicagenic acid (securi-
dacaside A) and 3-O-β- -glucopyranosyl-28-O-(R- -arabinopyranosyl-(1 f 3)-β- -xylopyranosyl-(1 f 4)-
[β- -apiofuranosyl-(1f3)]-R- -rhamnopyranosyl-(1 f 2)-[4-O-(3,4,5-trimethoxy-(E)-cinnamoyl-β- -fucopy-
D-
L
D
D
L
D
D
L
D
D
L
D
ranosyl)])-medicagenic acid (securidacaside B), were isolated from the methanol extract of the roots of
S. longepedunculata and characterized by spectroscopic methods. Securidacaside A, which occurred as
(E)- and (Z)-regioisomers, showed deterrency and toxicity toward C. maculatus and S. zeamais and
could contribute to the biological activity of the methanol extract. The potential to optimize the use of this
plant for stored product protection using water extracts, which would be appropriate technology for target
farmers, is discussed.
KEYWORDS: Securidaca; saponins; oviposition deterrent; Sitophilus; Callosobruchus; bruchid
INTRODUCTION
tropical African savannah, especially of Miombo and Caesalpi-
nioid woodland. This species has a wide variety of indigenous uses
including the protection of stored grain from weevil damage (6,7).
The activity is reportedly associated with nonpolar compounds in
the roots (4, 8). Compounds identified from Securidaca long-
epedunculata previously include methyl salicylate (9), tannins (10),
sucrose derivatives (11), phenolics (12) saponins (13, 14),
xanthones (15-17), and alkaloids (18,19). The aim of the present
study was to evaluate the effect of the powdered root bark and
methanol extracts of the root bark of this species against the
Coleopteran stored product pests, Sitophilus zeamais Motchulsky
and Callosobruchus maculatus F., up to 9 months after treatment
of the stored commodity, and identify and elucidate structures of
compounds responsible for the effects.
Despite the commercial difficulties associated with the regi-
stration of plant compounds as agrochemicals, interest in pesti-
cidal plants continues to grow (1,2). In the developed world, this
is linked to increasing demand for organic produce, for which
plant derived products are acceptable in pest control, despite
examples such as rotenone having well-known mammalian toxi-
city (3). Effective alternatives to synthetic pesticides, however, are
often a necessity rather than a choice for small-scale farmers in
sub-Saharan Africa. This is because synthetic pesticides can be
expensive, are often adulterated, are increasingly ineffective
owing to pest resistance, and may be difficult to accessreliably (4).
At best, pesticidal plants provide low-cost, safer, and environ-
mentally benign alternatives to synthetic pesticides.
Understanding why pesticidal plants are effective may facili-
tate the optimization of their use and therefore increase agricul-
turalproductivity, particularly among some ofthe world’s poorest
farmers (5). In this respect, we have investigated Securidaca
longepedunculata Fresen. (Polygalaceae), a widespread tree of
MATERIALS AND METHODS
Reagents. Methanol (HPLC grade) and acetic acid (HPLC grade)
were obtained from Merck (U.K.). All other chemicals were of analytical
grade. Deionized water was obtained from an in-house Milli-Q Plus
System (Millipore, Inc., Billerica, MA) at 18.2 MΩ.
Plant Material Extraction. S. longepedunculata roots were collected
from Tamale in North Ghana (Royal Botanic Gardens, Kew ref
*Corresponding author. Tel: þ44-208-332-5377. Fax: þ44-20-8332-
5310. E-mail: p.stevenson@kew.org.
pubs.acs.org/JAFC
Published on Web 09/21/2009
© 2009 American Chemical Society

Article
J. Agric. Food Chem., Vol. 57, No. 19, 2009 8861
BI-14863). Roots were ground to a fine powder (50 g) (KIKA mill, Janke
& Kunkel GmbH & Co. KG, Germany) and extracted in methanol
(150 mL) for 24 h at room temperature (22 °C), filtered, and evaporated to
dryness under reduced pressure.
δ 128.8 (C-1), 133.5 (C-2/6), 114.4 (C-3/5), 162.2 (C-4), 117.2 (C-R),
145.1 (C-β), 168.0 (CO), 55.8 (4-OMe).
Saponin 2 (Securidacaside B). Amorphous powder (2.0 mg); UV
(MeOH-H2O) λmax 321 nm. HRESIMS m/z: 1571.6904 [M - H]- (calcd.
for [C75H111O35]-, 1571.6911). 1H NMR data for the aglycone moiety: δ
2.06, 1.24 (2 ꢀ m, H-1a,b), 4.27 (br dd, J=7.2, 3.8 Hz, H-2), 4.08 (m, H-
3), 1.63 (m, H-5), 1.62, 1.20 (2 ꢀ m, H-6a,b), 1.52, 1.37 (2 ꢀ m, H-7a,b),
1.58 (m, H-9), 2.00, 1.91 (2 ꢀ m, H-11a,b), 5.30 (t, J=3.8 Hz, H-12), 1.59,
1.25 (2 ꢀ m, H-15a,b), 2.07, 1.63 (2 ꢀ m, H-16a,b), 2.84 (dd, J=14.0, 4.7
Hz, H-18), 1.75, 1.16 (2 ꢀ m, H-19a,b), 1.41, 1.25 (2 ꢀ m, H-21a,b), 1.77,
1.61 (2 ꢀ m, H-22a,b), 1.36 (s, 24-CH3), 1.24 (s, 25-CH3), 0.81 (s, 26-
CH3), 1.18 (s, 27-CH3), 0.92 (s, 29-CH3), 0.94 (s, 30-CH3). 13C NMR
data for the aglycone moiety: δ 45.0 (C-1), 71.3 (C-2), 86.3 (C-3), 53.9
(C-4), 53.3 (C-5), 21.8 (C-6), 34.3 (C-7), 41.3 (C-8), 49.7 (C-9), 37.4 (C-
10), 24.8 (C-11), 123.8 (C-12), 144.9 (C-13), 43.5 (C-14), 29.2 (C-15), 24.0
(C-16), 48.3 (C-17), 43.2 (C-18), 47.4 (C-19), 31.7 (C-20), 35.0 (C-21),
33.2 (C-22), 184.7 (C-23), 14.3 (C-24), 17.6 (C-25), 18.0 (C-26), 26.3 (C-
27), 178.1 (C-28), 33.6 (C-29), 24.2 (C-30). 1H and 13C NMR data for the
sugar moieties, see Table 1.
HPLC Analysis and Compound Isolation. The dried MeOH extract
of S. longepedunculata (5 g) was redissolved in MeOH and analyzed by
HPLC (Waters 600E pump and 996 PDA detector). The column used was
a 250 mm ꢀ 4.0 mm i.d., 5 μm, LiChrospher 100 RP-18 (Merck), with a
gradient elution program based on A=MeOH and B=2% AcOH; A=
65% at t=0 min, A=70% at t=26 min, and A=100% at t=40 min;
column temperature 30 °C and flow rate of 1 mL/min. Chromatograms
were extracted at 310 nm. Two compounds of interest, eluting at tR=28.3
(1) and 29.3 (2) min were collected by repetitive isolation yielding 4.5 mg
and 2.0 mg, respectively.
Sugar Analysis. Acid hydrolysis of 1 and 2 was carried out by
dissolving approximately 1.0 mg of each in 2 mL of dioxan/2M HCl
(1:1) and heating at 100 °C for 1 h. The hydrolysate was transferred to a
7 mL vial and dried under a stream of N2 on a heating block at 40 °C. The
absolute configurations of the constituent monosaccharides of 1 and 2
released by acid hydrolysis were determined by GC-MS analysis of their
trimethylsilylated thiazolidine derivatives, which were prepared using the
method of Ito et al. (20). Conditions for GC were as follows: capillary
column, DB5-MS (30 m ꢀ 0.25 mm ꢀ 0.25 μm), oven temperature
program, 180-300 at 6 °C/min; injection temperature, 350 °C; carrier
Insect Culturing and Handling. Strains of S. zeamais and C.
maculatus from West Africa were used. Insect cultures were maintained
(S. zeamais on maize and C. maculatus on cowpea) in 2.5 L glass jars in a
controlled temperature and humidity (CTH) room at 27 ( 5 °C, RH 60 (
5%, and 12:12 light/dark. C. maculatus is a highly mobile and fragile insect
compared to S. zeamais; therefore, counting and transfer of C. maculatus
to the test commodities was carried out with the aid of a low vacuum
suction pump (Charles Austen model DA7C, United Kingdom) and a
glass aspirator. A rubber tube was attached to the glass aspirator for
flexibility and to allow easy collection of the bruchids in the 2.5 L culturing
jars. S. zeamais were handled by sieving the cultures using 710 μm sieves
(Philip Harris Scientific, London) and with feather-light forceps.
Preparation of Known Age Insects. To prepare insects of known age
(7-14 day old S. zeamais and 1-7 day old C. maculatus), subcultures of
both insect species were prepared in their respective commodities. To
prevent contamination of the cultures, adults were removed and discarded
in the case of S. zeamais after three weeks of the initial setup, whereas C.
maculatus adults were removed after 10 days due to their shorter adult life
span of 7-10 days. After four (S. zeamais) and three (C. maculatus) weeks,
respectively, the cultures were checked daily for the F1 emergence. To
collect adults, the cultures were either sieved (S. zeamais) or adults
collected using an aspirator (C. maculatus) at the appropriate time.
Test Commodities. Whole organic cowpea (Canterbury Wholefoods,
Canterbury, Kent, U.K.) and whole organic maize (Gillet and Cook Ltd.,
Faversham, Kent, U.K.) were frozen at -20 °C for one week to kill any
existing infestation and then stored at 4 °C to prevent further infestation.
Three weeks prior to use, commodities were equilibrated to experimental
conditions in a controlled temperature and humidity room at 27 ( 5 °C
and 60 ( 5% RH 12:12 light/dark.
Effect of S. longepedunculata Root Bark Powder and Methanol
Extract of the Root Bark on the Emergence of S. zeamais and C.
maculatus in the F1 Generation. Pre-equilibrated commodity (100 g)
was placed in glass jars (250 mL), and after treatment and infestation,
the jars were sealed with black filter paper lids and molten wax, and
maintained at 27 ( 5 °C, 60 ( 5% RH, 12:12 light/dark. Powdered
S. longepedunculata root (0.5 g) was admixed into 10 jars for each test
insect at intervals of three months for a period of nine months (40
treatments plus 40 untreated controls). The concentration of crude
powdered material was informed by African farmer practice as well as a
series of experiments where a range of concentrations was evaluated
previously (8). A second, similar experiment was set up in which each
commodity was treated at the same time intervals but spraying 100 g
grain with 5 mL of a 2.5% (w/v) methanol extract of S. longepedunculata
roots that was the concentration equivalent of 0.5% root powder based
on HPLC analysis. The grain was allowed to dry in a fume hood for 3 h
at 22 °C. This was replicated 10 times with 100 g grain in each 250 mL
jar. Commodity treated with methanol only was used as the control.
After 9 months, 40 known-age adult insects of each species were
introduced into jars containing their respective commodity. The number
of live adult insects of the two species was recorded after 5 weeks
(C. maculatus on cowpea) or 8 weeks (S. zeamais on maize). The number
gas, He at 1 mL/min. Both 1 and 2 gave
D-apiose, D-xylose, L-arabinose,
L
-rhamnose, -fucose, and -glucose at tR 9.27, 9.48, 9.52, 10.19, 10.40,
D
D
and 12.20 min, respectively (identical to authentic standards).
Spectroscopic Analysis. HRESIMS was carried out using a Thermo
LTQ-Orbitrap XL instrument. Calibration was performed using factory
solutions containing sodium dodecyl sulfate, sodium taurocholate,
MRFA
(L-methionyl-arginylphenylalanyl-alanine acetate H2O), and
3
Ultramark1621 interfaced to an Accela autosampling LC system. For
chromatographic separation, a 150 mm ꢀ 3.0 mm i.d., 3 μm, Phenomenex
Luna C18(2) column was used with a linear mobile phase gradient, A=
H2O; B=MeOH; C=1% HCO2H in MeCN with A=90% and C=10% at
t=0 min; B=90% and C=10% at t=20 to 25 min at 400 μL/min flow rate
and 30 °C. Injection volumes were 2 μL, and data analysis was performed
using Xcalibur 2.0.7 software.
NMR spectra were acquired in CD3OD at 30 °C on a Bruker Avance
IIþ 700 MHz instrument equipped with a TCI-cryoprobe or on Varian
600 MHz or Bruker Avance 400 MHz instruments. Standard pulse
1
sequences and parameters were used to obtain 1D H, 1D 13C, and 1D
site selective ROE (rotating Overhauser enhancement), COSY
(correlation spectroscopy), TOCSY (total correlation spectroscopy),
HSQC (heteronuclear single quantum coherence), and HMBC
(heteronuclear multiple bond correlation) spectra. Chemical shift referen-
cing was carried out with respect to internal TMS at 0.00 ppm.
Saponin 1 (Securidacaside A; 5:3 Mixture of (E)- and (Z)-
Isomers). Amorphous powder (4.5 mg). UV (MeOH-H2O) λmax 315 nm.
HRESIMS m/z: 1511.6691 [M
-
H]- (calcd. for [C73H107O33]-,
1511.6700). 1H NMR data for the aglycone moiety: δ 2.08, 1.24 (2 ꢀ m,
H-1a,b), 4.26 (m, H-2), 4.08 (m, H-3), 1.64 (m, H-5), 1.62, 1.21 (2 ꢀ m, H-
6a,b), 1.53, 1.38 (2 ꢀ m, H-7a,b), 1.59 (m, H-9), 2.00, 1.92 (2 ꢀ m, H-11a,
b), 5.31 (m, H-12), 1.59, 1.23 (2 ꢀ m, H-15a,b), 2.07, 1.61 (2 ꢀ m, H-16a,b),
2.84 (br dd, J=13.4, 3.8 Hz, H-18), 1.75, 1.16 (2 ꢀ m, H-19a,b), 1.40, 1.25
(2 ꢀ m, H-21a,b), 1.76, 1.61 (2 ꢀ m, H-22a,b), 1.36 (s, 24-CH3), 1.26 (s,
25-CH3), 0.82 (s, 26-CH3), 1.18 (s, 27-CH3), 0.92 (s, 29-CH3), 0.94 (s,
30-CH3). 13C NMR data for the aglycone moiety: δ 45.0 (C-1), 71.4 (C-2),
86.4 (C-3), 54.1 (C-4), 53.3 (C-5), 21.9 (C-6), 34.3 (C-7), 41.3 (C-8), 49.8
(C-9), 37.4 (C-10), 24.8 (C-11), 123.6 (C-12), 144.9 (C-13), 43.6 (C-14), 29.2
(C-15), 23.9 (C-16), 48.3 (C-17), 43.2 (C-18), 47.5 (C-19), 31.7 (C-20), 35.0
(C-21), 33.3 (C-22), 186.5 (C-23), 14.4 (C-24), 17.5 (C-25), 18.0 (C-26), 26.4
(C-27), 178.1 (C-28), 33.6 (C-29), 24.2 (C-30). 1H and 13C NMR data for
the sugar moieties, see Table 1. 1H NMR data for the (Z)-isomer (where
different to that for the (E)-isomer): 4Fuc-O-(4-methoxy-(Z)-cinnamoyl)
moiety, δ 7.74 (d, J=8.9 Hz, H-2/6), 6.89 (d, J=9.0 Hz, H-3/5), 5.96 (d, J=
12.9 Hz, H-R), 6.95 (d, J=13.2 Hz, H-β), 3.82 (s, 4-OMe). 28-O-β-Fucp
moiety: δ 5.37 (d, J=8.2 Hz, H-1), 3.84 (m, H-2), 3.98 (m, H-3), 5.17 (dd,
J=3.4, 1.1 Hz, H-4). 13C NMR data for the (Z)-isomer (where different to
that for the (E)-isomer): 4Fuc-O-(4-methoxy-(Z)-cinnamoyl) moiety,

8862 J. Agric. Food Chem., Vol. 57, No. 19, 2009
Stevenson et al.
Table 1. 1H and 13C NMR Spectroscopic Data for the Glycosidic Moieties of Saponins 1 and 2 (CD3OD, 30 °C)
1
2
atom
δ1H (J in Hz)
δ13C
δ1H (J in Hz)
δ13C
At C-3
3-O-β-Glc
1
2
3
4
5
6
4.39 d (7.6)
104.4
75.5
77.7
71.3
77.8
62.5
4.37 d (7.7)
104.6
75.4
77.9
71.3
77.8
62.5
3.21 dd (9.0, 7.6)
3.36 m
3.33 m
3.21 dd (8.9, 7.6)
3.34 m
3.34 m
3.27 m
3.81 dd (12.2, 2.1)
3.67 dd (12.1, 5.0)
3.26 m
3.81 m
3.68 dd (11.9, 5.1)
At C-28
28-O-β-Fuc
1
2
3
4
5
6
5.40 d (8.1)
3.921 m
95.1
74.1
75.2
75.4
71.5
16.6
5.43 d (8.2)
3.91 m
95.2
74.8
75.0
75.5
71.5
16.7
3.99 dd (9.9, 3.5)
5.20 dd (3.5, 1.1)
3.915 m
3.99 dd (9.5, 3.7)
5.21 d (3.6)
3.92 m
1.10 d (6.4)
1.10 d (6.5)
2Fuc-O-R-Rha
3Rha-O-β-Api
4Rha-O-β-Xyl
3Xyl-O-R-Ara
1
2
3
4
5
6
5.46 br d (1.5)
4.06 dd (3.3, 1.5)
3.75 m
101.4
72.2
81.5
79.3
68.7
18.4
5.42 d (1.7)
4.07 m
3.78 m
101.6
72.2
81.4
79.2
68.9
18.6
3.64 m
3.90 m
1.28 d (6.3)
3.65 ’t’ (9.5)
3.88 m
1.27 d (6.2)
1
2
3
4
5.25 d (3.8)
4.02 d (3.9)
111.9
78.3
80.2
74.6
5.25 d (3.8)
4.02 d (3.7)
111.9
78.2
80.2
74.8
4.07 d (9.6)
3.75 d (9.6)
3.58 m
4.07 d (9.6)
3.75 d (9.6)
3.58 m
5
65.2
65.1
1
2
3
4
5
4.60 d (7.6)
3.31 m
105.0
75.3
89.9
69.9
66.6
4.61 d (7.7)
3.31 m
105.1
75.4
88.9
69.8
66.7
3.38 ’t’ (8.8)
3.57 m
3.89 m
3.41 ’t’ (8.9)
3.57 m
3.89 m
3.19 dd (11.7, 10.2)
3.19 dd (11.6, 10.1)
1
2
3
4
5
4.47 d (7.0)
3.75 m
3.72 m
106.7
73.5
74.5
69.9
67.8
4.49 d (7.4)
3.72 m
3.68 m
106.3
73.3
74.5
69.9
67.7
3.85 m
3.93 m, 3.61 m
3.84 m
3.92 m, 3.61 m
4Fuc-O-4-OMe-(E)-cinnamoyl
1
128.6
131.1
115.4
163.3
116.0
146.6
169.1
55.8
2/6
3/5
4
7.59 d (8.9)
6.96 d (8.9)
R
β
CO
4-OMe
6.53 d (15.9)
7.70 d (15.9)
3.84 s
4Fuc-O-3,4,5-triOMe-(E)-cinnamoyl
1
2/6
131.8
107.0
155.0
141.5
118.2
147.1
168.9
56.8
6.98 s
3/5
4
R
β
CO
6.64 d (15.9)
7.68 d (15.9)
3/5-OMe
4-OMe
3.89 s
3.80 s
61.3
of live adults in each commodity was analyzed using ANOVA. The
differences between the number of live insects in the treatments and
controls were compared at R=0.05 using the LSD test in the SPSS 10
statistical package.
Effect of the Methanol Extract of S. longepedunculata Root on
Adult Mortality and F1 Emergence of S. zeamais and C. maculatus
on Maize and Cowpea, respectively. The methanol extract of S.
longepedunculata root (5 mL; 2.5% w/v) was admixed with 100 g of either

Article
J. Agric. Food Chem., Vol. 57, No. 19, 2009 8863
TOCSY, HSQC, and HMBC data (Table 1 and Materials and
Methods). Among the characteristic resonances in the 1H NMR
spectrum were those for six quaternary methyl groups at δH 0.82,
0.92, 0.94, 1.18, 1.26, and 1.36 (all 3H, s), correlated in the HSQC
spectrum with δC 18.0, 33.6, 24.2, 26.4, 17.5, and 14.4, respec-
tively. Also of note was the characteristic resonance at δH 5.31
(1H, m, δC 123.6 by HSQC) corresponding to H-12 of a olean-12-
ene triterpenoid skeleton. Six anomeric proton resonances were
observed at δH 4.39, 4.47, 4.60, 5.25, 5.40, and 5.46, correlated in
the HSQC spectrum with δC 104.4, 106.7, 105.0, 111.9, 95.1, and
101.4, respectively. These preliminary data suggested that 1 was a
triterpenoid saponin with six sugars. The structure of the triter-
penoid moiety was confirmed to be that of medicagenic acid
(2β,3β-dihydroxyolean-12-ene-23,28-dioic acid), on the basis of
correlations observed in the HMBC spectrum and good agree-
ment between the 13C NMR assignments (Materials and Meth-
ods) and literature values acquired in the same solvent (21, 22).
Acid hydrolysis of 1 followed by determination of the absolute
configurations of the constituent monosaccharides confirmed the
presence of D-Api, D-Ara, D-Fuc, D-Glc, L-Rha, and D-Xyl.
1H and 13C NMR assignments for the six sugar residues were
obtained using a combination of COSY, TOCSY (τm=60 ms),
HSQC, and HMBC data, using the anomeric proton resonances
as starting points (Table 1). The sets of assignments were
Figure 1. Triterpenoid saponins from Securidaca longepedunculata root
extracts.
maize or cowpea equilibrated as described above and dried for 3 h in a
fume hood at 22 °C, and placed into 250 mL test jars. Methanol only
treated grain was used as a control. Forty known-age unsexed adult insects
of each species were introduced into jars containing their respective
commodity with C. maculatus on cowpea and S. zeamais on maize. The
experiment was maintained at 27 ( 5 °C, 60 ( 5% RH, and 12:12 light/
dark, and the commodity in each jar was sieved (aspirated in the case of
C. maculatus) every week for a period of eight weeks, and the number of
live adults in each replicate in each week was recorded. The live adults in
each replicate were replaced in their relevant jar in each week, and the dead
ones were collected separately. The total number of dead adults in each
commodity was also recorded weekly for eight weeks. The data were
analyzed using ANOVA, and the means were subjected to the LSD test at
R=0.05 (SPSS 10 statistical package).
Effect of Securidacaside A (1) on the Feeding Behavior and
Survival of S. zeamais. Compound 1 (2.5 mg) was dissolved in methanol
(2.5 mL) and mixed with maize (50 g) in a glass jar (250 mL) and dried on
aluminum foil for 3 h in a fume hood at room temperature (22 °C). The
treated maize was separated into glass vials (10 g in each=approximately
20 grains) to give five replicates. Methanol only treated maize was used as a
control. Ten unsexed S. zeamais adults aged 7-14 days were introduced
into each vial, and the number of damaged and undamaged maize seeds,
and the number of dead and live adults were recorded. Data were analyzed
using paired samples (Mann-Whitney U test), and differences in the
number of damaged and undamaged maize seeds, and in the number of
dead and live insects between the treatment and the control were tested for
significance at R=0.05.
Effect of Securidacaside A (1) on the Oviposition, Hatching, and
F1 Emergence of C. maculatus. Compound 1 (2.5 mg as a 1 mg/mL
solution) was mixed with cowpea (50 g) in a glass jar (250 mL) and dried
for 3 h in a fume hood at room temperature (22 °C). Treated cowpea was
separated into glass vials (10 g in each=approximately 50 peas) to give five
replicates. Methanol only treated cowpea was used as a control.
C. maculatus adults (1-7 days old) were released into each jar, and the
number of eggs, hatched eggs, and F1 adults in each vial was recorded.
Differences in the number of eggs, hatched eggs, and F1 C. maculatus
adults between 1 and the control were tested for significance at R=0.05
using paired samples (Mann-Whitney U test) in SPSS 10.
identified with specific sugars (D-Api, D-Ara, D-Fuc, D-Glc,
L
-Rha, and -Xyl) on the basis of coupling constant data for
D
1H NMR resonances, where resolved, and characteristic 1H and
13C NMR chemical shift values (21-23). The two deoxyhexose
residues were readily distinguished, as the full set of J values
obtained for one of them (H-1 at δH 5.40, d, J1,2=8.1, J2,3=9.9,
J3,4=3.5, J4,5=1.1, J5,6=6.4 Hz) was consistent with those of
β-Fucp (21, 23). The second deoxyhexose (H-1 at δH 5.46, br d,
J1,2 = 1.5, J5,6 = 6.3 Hz) was thus R-Rhap. The single hexose
residue was β-Glcp (H-1 at δH 4.39, d, J1,2=7.6 Hz), for which the
13C NMR assignments were comparable to literature values (22).
For the three pentose residues, a combination of J values and
chemicalshift data (Table 1) differentiatedβ-Apif (H-1 at δH 5.25,
d, J1,2=3.8 Hz, δC-1 111.9; J4a,4b=9.6 Hz) from R-Arap (H-1 at
δH 4.47, d, J1,2=7.0 Hz) and β-Xylp (H-1 at δH 4.60, d, J1,2=7.6
Hz) (21, 22).
Analysis of HMBC connectivities between sugar residues and
the medicagenic acid aglycone indicated that 1 was a bisdesmo-
side. At C-3, correlations were observed between H-3 of the
aglycone (δH 4.08) and C-1 ofGlc (δC 104.4), and from H-1 of Glc
(δH 4.39) to C-3 (δC 86.4). In a site selective ROE experiment, H-1
of Glc (δH 4.39) correlated to both H-2 (δH 4.26) and H-3 (δH
4.08) of the aglycone. Thus, a β-Glcp moiety was O-linked at C-3.
The remaining sugars constituted a branched pentasaccharide
O-linked at C-28. A correlation in the HMBC spectrum between
the anomeric proton at δH 5.40 and C-28 (δC 178.1) defined the
primary sugar as β-Fucp. The interglycosidic linkages of the
pentasaccharide chain followed from correlations in the HMBC
spectrum, in particular, H-1 of Rha to C-2 of Fuc (likewise, H-2
of Fuc to C-1 of Rha), H-1 of Api to C-3 of Rha (likewise H-3 of
Rha to C-1 of Api), H-1 of Xyl to C-4 of Rha (likewise, H-4 of
Rha to C-1 of Xyl), and H-1 of Ara to C-3 of Xyl (likewise, H-3 of
Xyl to C-1 of Ara). The pentasaccharide O-linked at C-28 was,
therefore, R-Arap-(1 f 3)-β-Xylp-(1 f 4)[β-Apif-(1 f 3)]-R-
Rhap-(1 f 2)-β-Fucp. In addition, the resonance of H-4 of Fuc
was downfield shifted (δH 5.20) and correlated in the HMBC
spectrum with the carbonyl group (δC 169.1) of a 4-methoxy-(E)-
cinnamoyl moiety (Table 1). A second, minor set of resonances
RESULTS AND DISCUSSION
Identification of Securidaca Saponins 1 and 2. Structures for two
saponins isolated from the MeOH extract of S. longepedunculata
roots were determined by MS, NMR, and sugar analysis
(Figure 1). Complete assignment of all proton and carbon
resonances in the NMR spectra of 1 was achieved using COSY,
corresponding to
also noted in the H NMR spectrum. The ratio of (E)- to (Z)-
a 4-methoxy-(Z)-cinnamoyl moiety was
1
forms was 5:3 by integration of proton intensities. Resonances

8864 J. Agric. Food Chem., Vol. 57, No. 19, 2009
Stevenson et al.
Figure 2. Effect of S. longepedunculata root bark powder and methanol extract treated commodities over that of untreated samples on S. zeamais and
C. maculatus F1 generation: mean number ((SEM, n = 10) of live adult insects in commodity mixed with S. longepedunculata root (0.5% w/w) or its methanol
extract (0.125% w/w) and stored for 0, 3, 6, and 9 months.
corresponding to a minor form could also be distinguished for
H-1 to H-4 of the primary Fuc residue (Materials and Methods),
most obviously at δH 5.37 (d, J=8.2 Hz, H-1) and δH 5.17 (dd, J=
3.4, 1.1 Hz, H-4), both slightly upfield (-0.03 ppm) of those for
the major (E)-form (Table 1). The resonance of Fuc H-4 (minor)
correlated in the HMBC spectrum with the carbonyl carbon
(δC 168.0) of the 4-methoxy-(Z)-cinnamoyl moiety. Thus,
saponin 1 was determined to be a 5:3 mixture of the (E)- and
triterpenoid aglycone (medicagenic acid) and glycosylation pro-
file (Table 1 and Materials and Methods). Acid hydrolysis of 2
followed by determination of the absolute configurations of the
constituent monosaccharides confirmed the presence of
-Ara, -Fuc, -Glc, -Rha, and -Xyl. Thus, saponin 2 was
3-O-β- -glucopyranosyl-28-O-(R- -arabinopyranosyl-(1 f 3)-β-
-xylopyranosyl-(1 f 4)[β- -apiofuranosyl-(1 f 3)]-R- -rhamno-
pyranosyl-(1 f 2)-[4-O-(3,4,5-trimethoxy-(E)-cinnamoyl-β-
D-Api,
D
D
D
L
D
D
L
D
D
L
D
-
(Z)-regioisomers of 3-O-β-
pyranosyl-(1 f 3)-β- -xylopyranosyl-(1 f 4)[β-
(1 f 3)]-R- -rhamnopyranosyl-(1 f 2)-[4-O-(4-methoxycinna-
moyl-β- -fucopyranosyl)])-medicagenic acid (securidacaside A).
Confirmation of the molecular formula of C73H108O33 for 1 was
by HR-ESIMS.
D
-glucopyranosyl-28-O-(R-
L
-arabino-
fucopyranosyl)])-medicagenic acid (securidacaside B). Con-
firmation of the molecular formula of C75H112O35 for 2 was by
HR-ESIMS.
D
D
-apiofuranosyl-
L
D
Effect of S. longepedunculata Root Bark Powder and Methanol
Extract of the Root Bark on the Emergence of S. zeamais and
C. maculatus in the F1 Generation over Nine Months. The number
of live adults emerging in the F1 generation was significantly
lower from maize and cowpea that had been mixed with either
S. longepedunculata root bark powder or sprayed with the
methanol extract of the root bark than emerged from the
methanol treated control (LSD, p < 0.05) across all treatment
times (Figure 2). This amounted approximately to an 80%
reduction in S. zeamais and 50% reduction in C. maculatus
emergence and confirmed that the dry powdered root bark
material, reportedly used by farmers in North Ghana to control
The 1H NMR spectrum of 2 was similar to that of 1, except that
resonances corresponding to three OMe groups were observed as
two singlets at δH 3.80 (3H) and 3.89 (2 ꢀ 3H). The aromatic
region comprised only a 2H singlet at δH 6.98 and two spin-
coupled doublets at δH 7.68 and 6.64 (both J=15.9 Hz). These
observations were consistent with the presence of a 3,4,5-
trimethoxy-(E )-cinnamoyl moiety. Full analysis of a set of one-
dimensional (1D) and two-dimensional (2D) NMR data sets
comparable to those acquired for 1 indicated that 2 had the same

Article
J. Agric. Food Chem., Vol. 57, No. 19, 2009 8865
Figure 3. Effect of S. longepedunculata methanol extract treated commodities on live F1 emergence of S. zeamais and C. maculatus: mean ((SEM, n = 10)
number of live S. zeamais (a) and C. maculatus (b) adult insects recorded in commodity treated with the methanol extract of S. longepedunculata (0.125% w/w)
over 8 weeks.
storage pests (4), had a quantifiable preventative effect against
stored product insect pest infestation of grain and that the effect
was chemical rather than physical as reported for other materi-
als (24). While a similar activity had been reported previously
(8, 25), the present study demonstrates that active components
were extractable in methanol and that the activity was effective
for up to 9 months with treatment applied every 3 months, a
treatment regime recommended for commercial alternatives such
as Actellic Super dust. A volatile component of the root extract,
methyl salicylate, is a known deterrent and also toxic to storage
pests, although there is little trace of it after 3 months (8). The
present work suggests that a different class of compound,
saponins, can contribute to the protective effect of S. long-
epedunculata root bark.
Effect of Methanol Extract of S. longepedunculata Root on Adult
Mortality and F1 Emergence of S. zeamais and C. maculatus. The
number of S. zeamais and C. maculatus adults emerging in the F1
generation from their respective commodity treated with the
methanol extract of S. longepedunculata roots was significantly
lower than the number emerging from solvent controls during an
8 week period (Figure 3). LSD analysis of the results showed that
the numbers of live S. zeamais increased significantly after the
fifth week (LSD, p < 0.05, Figure 3) as would be expected for this
species, while similarly, the number of C. maculatus adults
increased significantly after the third week (LSD, p < 0.05,
Figure 3). More importantly, significantly more adults emerged
in the F1 generation on control grain, treated with methanol only,
indicating that the methanol extract of S. longepedunculata roots
contained compounds that reduced the F1 emergence ofadults on
the grain. The results did not show a delay in the development
period of the F1 generation for either of the insect species,
suggesting that the effect is either an oviposition deterrent or
toxic to eggs or larvae, but not a development inhibitor.
Effect of Maize and Cowpea Treated with Securidacaside A (1)
on the Feeding Behavior of S. zeamais and on the Oviposition,
Hatching, and F1 Emergence of C. maculatus. The number of
damaged seeds in the maize treated with 1 was significantly lower
than that of the control maize (t test, p < 0.05, Figure 4), and
more dead adults were recorded on maize treated with 1 than on
the control maize, which was equivalent to >95% damage,
whereas in the treated seeds the damage was <35%. This
indicated that securidacaside A (1) was either deterrent or toxic
to S. zeamais adults and could explain, at least inpart, the effect of
the crude extracts against this insect and also the effect of the
whole plant material when mixed into the grain. It was not
possible, however, to determine whether the increased mortality
in the treated grain was caused by starvation owing to the
deterrent effect of 1 or whether this compound was toxic by
consumption.
The number of oviposited and hatched eggs on the surface of
cowpea seeds treated with 1 was significantly lower than that on
the control seeds (Figure 5). The number of adults emerging in the
F1 generation was also significantly lower in the treatment when
compared to the control (t test, p < 0.05) suggesting that 1 acts as
an oviposition deterrent as well as being toxic to C. maculatus
larvae as shown by the methanol extracts of the roots (Figure 5).

8866 J. Agric. Food Chem., Vol. 57, No. 19, 2009
Stevenson et al.
Figure 4. EffectofsecuridacasideA (1) treated commodities on feedingbehaviorandadultmortalityofS. zeamais:mean ((SEM, n = 5) numberofdamaged
seeds by S. zeamais recorded incontrol maize or maizetreated with1 (5 ꢀ 10-3% w/w) on the number of damaged seeds and dead S. zeamais adults after 2
weeks.
and efficient application procedure than hand mixing coarsely
powdered root bark. Although pounded roots of S. longepedun-
culata are effective when mixed into grain, ensuring that the roots
are sufficiently well powdered to mix efficiently throughout the
grain, is labor intensive and often wasteful. Furthermore,
maize grains have a glassy surface, and the powdered plant
material tends not to adhere to the surface, leaving many grains
exposed.
Some farmers in Ghana have reported effective treatment of
stored grain with water extracts from S. longepedunculata
roots (4). Jayasekara et al. (25) showed that methyl salicylate, a
volatile component of the methanol root extract, was toxic to
S. zeamais and the grain borers Rhyzopertha dominica and
Prostephanus truncatus. Methyl salicylate is poorly water-soluble,
however, and unlikely to entirely account for the protective effect
Figure 5. Effect of securidacaside A (1) treated commodities on oviposi-
tion and hatching of C. maculatus: mean ((SEM, n = 5) number of eggs
laidbyC. maculatus on cowpeatreated with1 or controlcowpeaina choice
of water extracts of the roots. The reported activity of water
extracts is also consistent with the presence of active saponin
constituents, which are water-soluble. Thus, the use of water
bioassay, mean((SEM, n=5) numberofhatchedeggsoncowpeaseeds,
extraction of S. longepedunculata roots by farmers could optimize
and ((SEM, n = 5) number of insects hatched in the F1 generation from
the same experiment.
the use of this plant and thus reduce the amount required
for treatment while facilitating the development of a more
sustainable use of this increasingly scarce plant material.
As with S. zeamais, this could explain, at least in part, the effect of
S. longepedunculata is not widely used in storage protection;
the crude extracts against bruchids and also the protective effect
therefore, there is scope for developing applications for this
of the whole plant material when used by farmers.
plant material in other parts of Africa, although optimization
Previously, Taylor et al. (26) demonstrated that a saponin,
of its use and information about its safety is required before it is
dehydrosoyasaponin I, isolated from the seed of the pea, Pisum
promoted widely. Belmain et al. (27) investigated the potential
sativum, showed antifeedant activity against S. zeamais, an effect
mammalian toxicity of several pesticidal plant species including
enhanced by the presence of three lysolecithin-type seed phos-
S. longepedunculata. They reported that rats, which had been fed
pholipids. These lysolecithins were inactive when presented to the
the root bark of S. longepedunculata, showed mild symptoms of
insects alone. Thus, there is evidence that saponins are deterrent
toxicity but only at the highest concentrations (5% w/w) tested
compounds to Sitophilus spp., although this property was miti-
and concluded that since farmers intentionally remove both
gated by lysolecithins (26).
synthetic pesticidal dusts or pesticidal plant materials from stored
The results presented here corroborate anecdotal reports on
grain by winnowing and washing with water they are highly
the practice by farmers in Ghana and Zambia in which the roots
unlikely to be exposed during consumption to toxic levels of
of S. longepedunculata are pounded and mixed into stored
material. In addition, the roots are consumed for traditional
grain (4). The longevity of activity recorded in the present work
medicinal purposes, suggesting further that the consumption of
is suitable for grain storage in the tropics by subsistence farmers
small quantities is of very low risk (28).
(at least 9 months) since most depend upon their own crops for
food, and thus, personal stores must last until the next crop is
harvested. This is especially relevant in the resource poor agri-
ACKNOWLEDGMENT
We thank Dr. G.C. Kite, Royal Botanic Gardens, Kew, U.K.
and EPSRC National Mass Spectrometry Service Centre, Uni-
versity of Wales for mass spectrometric data, and the Medical
Research Council Biomedical NMR Centre, National Institute
cultural environments of sub-Saharan Africa. The fact that the
extract applied to the stored product was as effective as the dried
root powder indicated that both contain active components.
However, the use of the extract may offer a more sustainable

Article
J. Agric. Food Chem., Vol. 57, No. 19, 2009 8867
for Medical Research, Mill Hill, London, U.K., for access to
higher-field NMR facilities. We are also grateful to Fuseini
Haruna Andan (Prince), Ministry of Forestry and Agriculture,
PO Box 950, Tamale, Ghana for providing the plant material.
(17) Galeffi, C.; Federici, E.; Msonthi, J. D.; Marini-Bettolo, G. B.;
Nicoletti, M. New xanthones from Extiadiopis oblongifolia and
Securidaca longepedunculata. Fitoterapia 1990, 61, 79–81.
(18) Scandola, M.; Games, D. E.; Costa, C.; Allegri, G.; Bertazzo, A.;
Curcuruto, O.; Traldi, P. Structural study of alkaloids from Secur-
idaca longipedunculata roots. 2. Isolation and characterization by
supercritical-fluid chromatography mass-spectrometry. J. Hetero-
cyclic Chem. 1994, 31, 219–224.
(19) Costa, C.; Bertazzo, A.; Allegri, G.; Curcuruto, O.; Traldi, P. Indole
alkaloids from the roots of an African plant Securidaca longipedun-
culata. Isolation by column chromatography and preliminary struc-
tural characterization by mass-spectrometry. J. Heterocyclic Chem.
1992, 29, 1641–1647.
LITERATURE CITED
(1) Isman, M. B. Botanical insecticides, deterrents, and repellents in
modern agriculture and an increasingly regulated world. Annu. Rev.
Entomol. 2006, 51, 45–66.
(2) Isman, M. B. Botanical insecticides: for richer, for poorer. Pest
Manag. Sci. 2008, 64, 8–11.
(3) PAN-UK (2008). Rotenone. http://www.pan-uk.org/pestnews/Actives/
rotenone.htm.
(4) Belmain, S. R.; Stevenson, P. C. Ethnobotanicals in Ghana: reviving
and modernising an age-old practice. Pesticide Outlook 2001, 6, 233–
238.
(20) Ito, A.; Chai, H.-B.; Kardono, L. B. S.; Setowati, F. M.; Afriastini,
J. J.; Riswan, S.; Farnsworth, N. R.; Cordell, G. A.; Pezzuto, J. M.;
Swanson, S. M.; Kinghorn, A. D. Saponins from the bark of
Nephelium maingayi. J. Nat. Prod. 2004, 67, 201–205.
(5) Belmain, S. R. Botanicals. In Crop Post-Harvest Science and
Technology. Volume 1: Principles and Practice; Golob, P., Farrell G.,
Orchard J. E., Eds.; Blackwell Science: Oxford, U.K., 2002; pp 280-283.
(6) Burkill, H. M. The Useful Plants of Tropical West Africa; Royal
Botanic Gardens: Kew, U.K.; 1997; Vol. 4, pp 454-458.
(7) Heywood, V. H. Flowering Plants of the World; Oxford University
Press: Oxford, U.K., 1985; pp 215-216.
(8) Jayasekara, T. K.; Belmain, S. R.; Stevenson, P. C.; Hall, D. R.
Securidaca longepedunculata As a Control for Stored Product Insect
Pests. In Advances in Stored Product Protection; Credland, P. F.,
Armitage, D. M., Bell, C. H., Cogan, P. M, Highley, E., Eds.; CAB
International: Wallingford, U.K., 2003; pp 596-599.
(9) Jayasekara, T. K.; Belmain, S. R.; Stevenson, P. C.; Hall, D. R.;
Farman, D. I. Identification of methylsalicylate as the principal
volatile component in the methanol extract of the root of Securidaca
longepedunculata Fers. J. Mass Spectrom. 2002, 37, 577–580.
(10) Makkar, H. P. S.; Becker, K. Do tannins in leaves of trees and shrubs
from African and Himalayan regions differ in level and activity?
Agroforest. Systems 1998, 40, 59–68.
(11) de Tommasi, N.; Piacente, S.; de Simone, F.; Pizza, C. New sucrose
derivatives from the bark of Securidaca longepedunculata. J. Nat.
Prod. 1993, 56, 134–137.
(21) Lavaud, C.; Voutquenne, L.; Massiot, G.; Le Men-Olivier, L.; Das,
B. P.; Laprevote, O.; Serani, L.; Delaude, C.; Becchi, M. Saponins
from the stem bark of Filicium decipiens. Phytochemistry 1998, 47,
441–449.
(22) Kapusta, I.; Janda, B.; Stochmal, A.; Oleszek, W. Determination of
saponins in aerial parts of barrel medic (Medicago truncatula) by
liquid chromatography-electrospray ionization/mass spectrometry.
J. Agric. Food Chem. 2005, 53, 7654–7660.
(23) Duus, J.Ø.; Gotfredsen, C. H.; Bock, K. Carbohydrate structural
determination by NMR spectroscopy: Modern methods and limita-
tions. Chem. Rev. 2000, 100, 4589–4614.
(24) Stathers, T. E.; Riwa, W.; Mvumi, B. M.; Mosha, R.; Kitandu, L.;
Mngara, K.; Kaoneka, B.; Morris, M. Do diatomaceous earths have
potential as grain protectants for small-holder farmers in sub-
Saharan Africa? The case of Tanzania. Crop Protection 2008, 27,
44–70.
(25) Jayasekara, T. K.; Stevenson, P. C.; Hall, D. R.; Belmain, S. R.
Effect of volatile constituents from Securidaca longepedunculata on
stored grain insect pests. J. Chem. Ecol. 2005, 31, 303–313.
(26) Taylor, W. G.; Fields, P. G.; Sutherland, D. H. Insecticidal compo-
nents from field pea extracts: Soyasaponins and lysolecithins.
J. Agric. Food Chem. 2004, 52, 7484–7490.
(12) Lognay, G.; Marlier, M.; Seck, D.; Haubruge, E. The occurrence of
2-hydroxy-6-methoxybenzoic acid methyl ester in Securidaca long-
epedunculata (Fresen) root bark. Biotechnol. Agron. Soc. Environ.
2000, 4, 107–110.
(27) Belmain, S. R.; Neal, G. E.; Ray, D. E.; Golob, P. Insecticidal and
vertebrate toxicity associated with ethnobotanicals used as post-
harvest protectants in Ghana. Food Chem. Toxicol. 2001, 39, 287–
291.
(13) Sofowora, A. The present status of knowledge of the plants used in
traditional medicine in Western Africa: A medical approach and a
chemical evaluation. J. Ethnopharmacol. 1980, 2, 109–118.
(14) Hultin, E. Securidaca-saponin. Acta Chem. Scand. 1967, 21, 1714–
1720.
(15) Lannang, A. M.; Lontsi, D.; Ngounou, F. N.; Sondengam, B. L.;
Nkengfack, A. E.; van Heerden, F. R.; Assob, J. C. N. Securida-
caxanthone A, a heptaoxygenated xanthone from Securidaca longe-
pedunculata. Fitoterapia 2006, 77, 199–202.
(28) Olajide, O. A.; Awe, S. O.; Makinde, J. M. Pharmacological screen-
ing of the root extract of Securidaca longepedunculata. Fitoterapia
1998, 69, 245–248.
Received May 15, 2009. Revised manuscript received August 27, 2009.
Accepted September 02, 2009. Financial support for this work was
provided by the Higher Education Funding Council of England
(HEFCE) and Implementation and Coordination of Agricultural
Research and Training: Competitive Regional Agricultural Research
Fund (ICART-CRARF) Programme in the SADC Region (9th
European Development Fund: 9 ACP SAD 1-11.).
(16) Meyer, J. J. M.; Rakuambo, N. C.; Hussein, A. A. Novel xanthones
from Securidaca longepedunculata with activity against erectile
dysfunction. J. Ethnopharmacol. 2008, 119, 599–603.