Tylosema esculentum extractives and their bioactivity

Article abstract of DOI:10.1016/j.bmc.2011.07.006

The investigation of Tylosema esculentum (Morama) husks, cotyledons, and tuber yielded griffonilide 2, compound 1, griffonin 3, gallic acid 4, protocatechuic acid 5, β-sitosterol 6, behenic acid 7, oleic acid 8, sucrose 9, 2-O-ethyl-α-d-glucopyranoside 10, kaempferol 11 and kaempferol-3-O-β-d-glucopyranoside 12. The structures of the isolates were determined by NMR, HR-TOF EIMS, IR and UV-vis spectroscopy, and by comparison with literature data. The husk EtOAc and n-butanol extracts demonstrated >90% DPPH radical scavenging activity at concentrations of 25, 50 and 250 μg/mL. Furthermore the husk extracts showed higher total phenolic content (233 mg GAE/g). The extractives exhibited minimum inhibitory quantities of 50-100 μg or no activity against Staphylococcus aureus, Escherichia coli, Bacillus subtilis, Pseudomonas aeruginosa and Candida albicans. The tuber extracts were inactive against Caco-2 and Hela cell lines, while the husk extracts showed low activity against Caco-2 and Vero cell line with IC₅₀ values >400 μg/mL. The GC-MS analysis showed the beans and tuber non-polar (n-hexane) extracts major constituents as fatty acids.

Full text of DOI:10.1016/j.bmc.2011.07.006

Bioorganic & Medicinal Chemistry 19 (2011) 5225–5230
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Tylosema esculentum extractives and their bioactivity
Ofentse Mazimba a, Runner R. T. Majinda a, , Ceciliah Modibedi , Ishmael B. Masesane , Avrelija Cencic ,
ˇ b
⇑
a
a
Walter Chingwaru b
a Department of Chemistry, University of Botswana, Private Bag 00704, Gaborone, Botswana
b
ˇ
Department of Microbiology, Faculty of Agriculture and Life Sciences, University of Maribor, Pivola 10, 2311 Hoce, Slovenia
a r t i c l e i n f o
a b s t r a c t
Article history:
The investigation of Tylosema esculentum (Morama) husks, cotyledons, and tuber yielded griffonilide 2,
compound 1, griffonin 3, gallic acid 4, protocatechuic acid 5, b-sitosterol 6, behenic acid 7, oleic acid 8,
sucrose 9, 2-O-ethyl-a-D-glucopyranoside 10, kaempferol 11 and kaempferol-3-O-b-D-glucopyranoside
12. The structures of the isolates were determined by NMR, HR-TOF EIMS, IR and UV–vis spectroscopy,
and by comparison with literature data. The husk EtOAc and n-butanol extracts demonstrated >90% DPPH
Received 4 April 2011
Revised 29 June 2011
Accepted 4 July 2011
Available online 14 July 2011
radical scavenging activity at concentrations of 25, 50 and 250
showed higher total phenolic content (233 mg GAE/g). The extractives exhibited minimum inhibitory
quantities of 50–100 g or no activity against Staphylococcus aureus, Escherichia coli, Bacillus subtilis, Pseu-
domonas aeruginosa and Candida albicans. The tuber extracts were inactive against Caco-2 and Hela cell
lines, while the husk extracts showed low activity against Caco-2 and Vero cell line with IC50 values
lg/mL. Furthermore the husk extracts
Keywords:
Tylosema esculentum
Antioxidant
Antimicrobial
Cytotoxicity
Fatty acids
l
>400
lg/mL. The GC–MS analysis showed the beans and tuber non-polar (n-hexane) extracts major con-
stituents as fatty acids.
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
compositional levels and domestication.3,10,11 T. esculentum seed
coat (husk) ethanolic extracts (0.01–0.001 mg/mL), cotyledon
ethanolic extracts (P0.1 mg/mL) and tuber water extracts (0.1–
0.01 mg/mL) are reported to exhibit high in vitro inhibition of rota-
viruses which are a major cause of diarrhoea among infants and
immunocompromised people.12 The seed coat is reported to have
a higher concentration of total phenolic acids and total flavonoids
than the cotyledon, where levels were determined using reversed-
phase high performance liquid chromatography.13 Some of the de-
tected phenolics in the cotyledon and seed coat were myricetic,
kaempeferol, rutin, gallic acid, protocatechuic acid and sinapic
acid.13 The Literature12,13 has recommended the isolation phyto-
chemicals of T. esculentum extracts and identification of the
responsible bioactive compounds, as isolations have not been re-
ported.13 The objective of the study was to investigate the phyto-
chemical constituents, antioxidant potential of Morama extracts
and phenolic compounds as well as to establish the composition
of the non-polar extracts using GC–MS analysis. The cell growth
inhibition and antimicrobial activity were also purposed to be
carried out to determine the antimicrobials.
Morama [Tylosema esculentum (Burch.) Schreiber; Fabaceae:
Caesalpiniaceae] bean and tuber are gathered from the wild and
eaten by the indigenous people of the Kalahari Desert of Botswana,
spreading to Namibia and South Africa. The bean oil is reported to
be rich in mono- and di-unsaturated fatty acids and contains no
cholesterol.1,2 It is also a good source of calcium, iron, zinc, phos-
phate, magnesium and B-vitamins. Morama bean is an excellent
source of good quality proteins. The protein content is 32–45%,
while oil content is 30–42%.1–3
The much publicized healthy living concept has embraced the
consumption of fresh fruit, tea, herbs and vegetables, alongside
physical exercises. The increased intake of dietary antioxidant
helps to maintain a balance between antioxidants and oxidants
in living organisms, to alleviate oxidative stress. The role of reac-
tive oxygen species/and or oxidants in many diseases including
malaria, acquired immunodeficiency syndrome, heart disease, dia-
betes, cancer,4 asthma, inflammatory joint disease and immune
system decline are well documented.5,6 The resurgence of interest
in the natural control of phytopathogens and increasing consumer
demand for effective, safe and natural products means that quan-
titative data on plant extracts are required.7–9 Prior to this work,
documented research on Morama focused mostly on nutritional
2. Results and discussion
2.1. Phytochemistry
The phytochemical work-up of T. esculentum (Morama or
Marama) husks, cotyledons and tuber yielded twelve secondary
⇑
Corresponding author. Tel.: +2673552505; fax: +2673553826.
E-mail address: majindar@mopipi.ub.bw (R.R.T. Majinda).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.07.006

5226
O. Mazimba et al. / Bioorg. Med. Chem. 19 (2011) 5225–5230
Table 1
metabolites. The chemical structures (Fig. 1) of the compounds
were identified based on spectral data analyses (NMR, MS, IR, UV
1H and 13C NMR data for compounds 1 and 2
a
a
and [a]) D and comparison with the cited Literature data. The cya-
H/
C
dH (J)
dC
dH (J)
dC
nogenic glucoside, griffonin (3),14 and the phenolic compounds
gallic acid (4), protocatechuic acid (5),15 kaempferol (11) and
1
2
2
À
À
kaempferol-3-O-b-D
-glucopyranoside (12),16 were isolated from
175.8
178.2
3
5.90, d (1.8)
111.6
2.37 dd (8.7, 17.4) 2.75 dd (9.3, 17.7)
35.4
the non edible part, which are the husks (very hard seed coat).
The Morama edible parts are the dehulled beans (cotyledons)
and tubers. The dehulled beans afforded b-sitosterol (6),17 oleic
3a
4
À
3.54 dd (2.4, 8.1)
37.5
164.8
acid (8),18 sucrose (9)19 and 2-O-ethyl-
a-D-glucopyranoside
6.27 dd (2.1, 9.9)
120.6
5.72 m
131.8
(10).20 Griffonilide (1),14 compound (2) and behenic acid (7)21 were
obtained from the tuber. It was of significance importance that the
cytotoxic compound 3 was not detected from the edible parts of
Morama bean.
5
6.62 dd (2.4, 9.9)
144.2
5.72 m
126.6
6
4.33 dt (2.4,2.4, 7.8)
73.6
4.06 dt (2.4, 8.1)
71.2
7
3.54 dd (2.4, 8.1)
80.0
4.90 dd (2.1, 10.8)
85.2
4.33 dd (2.4, 7.8)
74.7
4.54 dd (1.2, 8.1)
83.3
Compound 2 was isolated as a mixture (1:3) with 1 and was
recognized from NMR data as the minor peaks in 1H and 13C
NMR spectrum of the mixture (Table 1). The 13C NMR spectra of
1 exhibited signals for three oxygenated methine carbons (C-6, 7,
7a), three methine carbons (C-3, 4, 5), a quaternary carbon (C-3a)
and a carbonyl carbon (C-2). The 1H NMR spectra of compound 1
exhibited six signals, each integrating for one proton and analysis
of the HSQC spectra revealed that the protons were all attached
to methine carbons. Three protons resonated at dH 3.54 (1H, dd,
J = 2.4, 8.1 Hz, H-7), 4.33 (1H, dt, J = 2.4, 2.4, 8.4 Hz, H-6) and 4.90
(1H, dd, J = 2.1, 10.8 Hz, H-7a). Two cis-vicinally coupled (3J) pro-
tons resonated at dH 6.27 (1H, dd, J = 2.1, 9.9 Hz, H-4) and 6.62
(1H, dd, J = 2.4, 9.9 Hz, H-5) and an olefinic proton signal was ob-
served at dH 5.89 (1H, br s, H-3). The IR absorption bands for the
7a
a
MeOD, (x) proton multiplicity in Hz.
H-3a) and 2.75 (1H, dd, J = 9.3, 17.7 Hz, H-3b). The signals were as-
signed to C-3 (dC 35.4) using HMQC spectrum correlations.
Compound 1 molecular formula C8H8O4 was established by HR-
TOF EIMS [M]+ = 168.0420. The HR-TOF EIMS spectra molecular ion
at m/z 170.0576 was used to establish the molecular formula of 2
as C8H10O4. The four degrees of unsaturation in the molecule were
accounted for by a bicyclic structure with one carbonyl group and
one double bond. Based on the NMR, MS data and comparison with
data of 1 in Table 1 and published data,14 compound 2 was identi-
fied as 3,3a,7,7a-tetrahydro-6,7-dihydroxy-2(6H)-benzofura-none.
It is also important to mention that compound 1 was also obtained
from the hydrolysis of compound 3 using trifluroacetic acid.
The relative stereochemistry of 1 was established by the NOE
correlations observed in a NOESY experiment. In the NOESY spec-
trum, H-7a showed NOE interactions with H-6 and none with H-7,
suggesting the same orientation for H-6 and H-7a protons. The
NOE correlations and the small coupling constants (J = 2.1–
butenolide moiety were observed at 1726 and 1649 cmÀ1 22
IR and NMR spectral data of 1 indicated the presence of
.
The
,b,
a
c
,
d-unsaturated carbonyl system. The HR-TOF EIMS spectra molecu-
lar ion at m/z 170.0576 was used to establish the molecular for-
mula as C8H10O4. The five degrees of unsaturation in the
molecule were accounted for by a bicyclic structure with one car-
bonyl group and two double bonds.
The spectral data of 1 was similar to literature data.14 The 13C
NMR spectral pattern of 2 was almost identical to that of 1 showing
eight signals. The difference was in the presence of two downfield
13C resonances at d 35.4 (3-CH2) and 37.5 (3a-CH) in the spectra of
2. The absence of the methine signal at dC 111.6 (dH 5.90, d,
J = 1.8 Hz, H-3) and the quaternary carbon signal at dC 164.8
(C-3a) found in the spectra of 1 were key observations. These
observations indicated that 2 had the same basic structure as 1,
but have a reduced C-3/C-3a double bond. The 1H NMR spectra
of 2 exhibited germinally (J = 9.3, 17.4 Hz) coupled methylene pro-
ton signals resonating up-field at dH 2.37 (1H, dd, J = 8.7, 17.4 Hz,
2.4 Hz) between H-6 and H-7a corroborated an
a-orientation for
H-6 and H-7a in 1. The small coupling constants (J = 2.1–2.4 Hz)
between H-6 and H-7a occurs presumably due to the ‘W-coupling’,
where the H–C–C–C–H system lies in the same plane.23 The same
coupling constants and the negative sign of optical rotation similar
to those of compound 1, were used to assign the
H-6 and H-7a and b-orientation for H-7 in 2.
a-orientation for
2.2. Total phenolic content and antioxidant activity
The total phenolic content (TPC) of Morama extracts were esti-
mated using gallic acid as a standard and was expressed as milli-
grammes of gallic acid equivalents (GAE) per gramme dry extract
(mg GAE/g) (Table 2). The EtOAc and n-butanol extracts (Table 2)
exhibited the highest total phenolic content (>233 mg GAE/g),
CN
3a
7a
H
O
OGluc
OH
O
2
O
6
HO
O
HO
H
H
OH
OH
OH
Table 2
1
2
3
Total phenolic content (TPC) of T. esculentum extracts
OH
OH
COOH
Extract
TPC (mg GAE/g)
O
HO
O
HO
Fresh milk
Dried milk
EtOAc T
70% MeOH T
CHCl3
0.41 0.07
1.54 0.02
91.94 0.12
92.36 0.23
5.81 0.05
HO
OH
HO
R
O
OR
OH
OH
O
EtOAc H
n-butanol H
70% EtOH de-hulled bean
233.42 2.50
235.11 3.00
39.99 0.08
10
R = OH 4
R = H 5
11 R = H
12 R = β-D-glucopyranose
Values are average of three replications. T-tuber, H-husk.
Figure 1. Chemical structure of compounds isolated from T. Esculentum.

O. Mazimba et al. / Bioorg. Med. Chem. 19 (2011) 5225–5230
5227
while morama milk showed low detection of phenolics (0.41 mg
GAE/g).
Morama extracts and compounds 1, 3, 4, 5, 11 and 12 were as-
sayed for DPPH radical scavenging activity. The Morama husk
EtOAc extract (EC50 = 4.93 g/mg) and n-butanol extract
(EC50 = 3.47 g/mL) exhibited higher DPPH free radical scavenging
activities than the Morama tuber extracts (EC50 = 91.76 g/mL),
morama milk (EC50 >1000 g/mL) and cotyledons EtOH
(EC50 = 282.81 g/mL) extracts. Morama milk poor DPPH radical
l
l
l
l
l
scavenging activity was attributed to its high content of oils, 30–
42%2,3 and low phenolic content. Compounds 4, 5, 11, 12 showed
activities comparable to the standard (ascorbic acid), with
compound 4, after 30 min of reaction time showing EC50 of
Figure 3. Scavenging of hydrogen peroxide by T. esculentum extracts.
1.85
l
g/mL compared to that of the reference compound, ascorbic
minimum inhibitory quantities of 50–100 lg or no activity com-
pared to the reference compound, chloramphenicol (0.001 lg).
acid (EC50 = 41.08
ing activity was also shown as % of DPPH radical scavenged (after
l
g/mL). The quantitative DPPH radical scaveng-
3 h) in Figure 2. The husk extracts scavenged 94% of the DPPH
radical at concentration of 25 lg/mL, this was attributed to the
2.4. Cell growth inhibition
presence of phenolic compounds 4–5 and 11–12 in the extracts.
Morama husk extracts could, thus, have an antioxidant potential.-
after 3 h of reaction time. Values are means of triplicate determina-
tions, SD 0.02–0.05. H-Husks, T-Tuber.
A positive linear correlation between antioxidant activity and
total phenolic content for extracts at 5, 25, 50 and 250 lg/mL (cor-
relation coefficient r = 0.9395 and 0.9543, 0.9328 and 0.8804
respectively) was observed. These correlations suggested that the
phenolic compounds contributed significantly to the antioxidant
capacity of the investigated plant extracts. These results were con-
sistent with the trends in literature reporting such positive corre-
lation between total phenolic content and antioxidant activity.24,25
The effect of Morama extracts and gallic acid on scavenging of
hydrogen peroxide is shown in Figure. 3.
The results show that Morama extracts had an effective H2O2
scavenging activity in a concentration dependant manner. All the
tested extracts and controls (gallic acid and ascorbic acid) demon-
strated greater than 50% scavenging activity at all concentrations
The Caco-2 (human colon adenocarcinoma cells), HeLa (cervical
cancer cells) cells and normal cells derived from African green
monkey kidney cells (Vero) were used for the cytotoxicity tests
in this study. The cytotoxicity of the tuber and husk extracts
against the cell lines are summarized in Table 3. The result show
that tuber extracts had no cytotoxicity against Caco-2 and Hela cell
lines, but these extracts showed lower activity against Vero cell
lines with values of IC50 >400
the husk showed stimulation of cell proliferation in Hela cells with
ED50 values of 430 g/mL. Epidemiological evidence suggest that
lg/mL. The n-butanol extract of
l
polyphenolic phytochemicals posses cancer chemopreventive
properties,26 but in this study the husk extracts which have shown
good antioxidant properties show high IC50 values in cell growth
inhibition. There was a negative correlation between the DPPH
and H2O2 scavenging properties and cell growth inhibition test.
2.5. GC–MS analysis
tested [12.5, 25, 50, 125 and 250 lg/mL]. The husk EtOAc extract
The IR spectrum of the non-polar extracts of both the tuber and
dehulled bean (cotyledons) displayed the strong bands at 2960 and
2873 cmÀ1 (C–H stretching of methylene groups), 1736 cmÀ1 (car-
bonyl groups), 1471 (C–H bending) and 1184 cmÀ1 (C–O stretch-
ing). These bands were indicative of the fatty acid nature of the
non-polar extracts.27
demonstrated the highest H2O2 inhibition values (64–83%) while
the tuber methanol extract exhibit the lowest H2O2 inhibition val-
ues at 50–71%.
2.3. Antimicrobial studies
The non-polar extracts from the edible parts (tuber and cotyle-
dons) were analyzed using GC–MS. A total of thirteen compounds
were identified in the sample non-polar extracts and theseaccounted
for 99.36% of cotyledons and 85.48% of tuber extract content. The
main components are listed in Table 4. The cotyledons extract main
constituents were oleic acid (50.7%) and palmitic acid (23.9%), while
the tuber extract main constituents were palmitic acid (55.1%) and
oleic acid (26.8%). Saturated fatty acids accounted for 27.9% in the
The preliminary antimicrobial activities of the extractives were
evaluated using the TLC auto-bioautography method. The results
indicated very weak antimicrobial activities against Staphylococcus
aureus, Escherichia coli, Bacillus subtilis, Pseudomonas aeruginosa and
Candida albicans. The extracts and compounds 1, 3–10 exhibited
Table 3
Cytotoxic activities of Morama extracts
Extract
Cell types
HeLa
Caco-2
Vero
MeOH T
EtOAc T
EtOAc H
n-butanol H
Inactivea
Inactivea
Activeb
Inactivea
Inactivea
Activec
Activeb
Activeb
Activeb
Activeb
Activeb
stimulated
a
b
c
IC50 >1000 lg/mL.
400
l
g/mL < IC50 < 1000
g/mL
g/mL. Data shown represent the mean SD of three independent
lg/mL.
IC50 = 120
l
d
ED50 = 430
l
Figure 2. T. esculentum extractives DPPH radical scavenging activity (%) after 3 h of
reaction time.
determinations. T-tuber, H-husk.

5228
O. Mazimba et al. / Bioorg. Med. Chem. 19 (2011) 5225–5230
Table 4
reagent from Rochelle Chemicals (South Africa). All reagents were
of analytical grade or better.
GC–MS analysis of T. esculentum non-polar extracts from the bean and tuber
Components
RIa
(%) Compositionb
3.3. Plant Material
Cotyledons
Palmitic acid
Oleic acid (9E)
Linoleic
Oleic acid (9Z)
Stearic acid
1960
2102
2161
2175
2184
2463
23.9
40.3
18.1
10.4
4.0
The Morama bean and tuber samples were supplied in April
2007 by Thusano Lefatsheng agency, in Mmankgodi village, in
the South District of Botswana. A voucher specimen (voucher #
MB4kg1990) has been deposited at the University of Botswana
Herbarium.
Icosanoic acid
Tuber
1.08
4-Methyldodecane
Pentadecanoic acid
Palmitic acid
14-Methylpentadecanoic
Oleic acid (9E)
Oleic acid (9Z)
Stearic acid
1200
1870
1960
2099
2102
2175
2184
2.06
0.22
3.4. Extraction and Isolation
55.07
1.08
15.76
11.04
0.23
3.4.1. Husk
Powdered husk (4.5 kg) was first defatted with n-hexane and
then extracted with 70% aqueous ethanol. The solvent was evapo-
rated using rotary evaporator (Buchi R210, Switzerland) and the
residual water was removed by freeze drying to give 300 g of dry
extract. Part of the dry extract (178.8 g) was reconstituted in dis-
tilled water (200 mL) and extracted successively with ethyl acetate
and n-butanol to yield the EtOAc (20.8 g) and n-butanol (14.7 g)
extracts respectively. The n-butanol extract (14.7 g) was fraction-
ated using silica gel column (500 g silica gel) eluting with n-hex-
ane/chloroform (4:1) and increasing polarity to 100% chloroform
and to 30% methanol in chloroform. Fractions (500 ml each) were
collected and pooled according to similarity in TLC profiles to give
fractions H1 to H7. White crystals identified as 6 came out of H1 to
H3. H4 (340 mg) was eluted through a Sephadex LH-20 (Sigma) col-
umn (MeOH/CHCl3 1:1) to give subfractions H1’ to H3’. H2’ gave 4 as
white crystals (70 mg), H1’ gave 11 (10 mg), H3’ afforded 5 (10 mg),
H5 yielded more 4 (1.2 g) after further purification using Sephadex
LH-20 column. H6 afforded 3 (430 mg) and H7 yielded 12 (15 mg)
on further purification on a Sephadex LH-20 column and repeated
preparative TLC (CHCl3/MeOH 4:1).
a
Retention indices (RI) relative to C9-C24 n-alkanes on the HP 5MS column.
% composition based on peak areas calculated in GC on HP 5MS column.
b
cotyledons and 56.6% in the tuber non-polar extract, while monoun-
saturated fatty acids composition was 51.8% and 26.8% in the cotyle-
dons and tuber extracts respectively. The polyunsaturated fatty acids
were detected only in the cotyledons extract at 18.1%. These results
are in agreement with the published data,3,13 which reports 27.6%
saturated fatty acids, 50.2% monounsaturated fatty acids and 21.2%
poly unsaturated fatty acids in the dehulled beans.
In conclusion, the phytochemical work-up of Morama husks,
cotyledons and tuber yielded twelve compounds. The results
established that the non-polar extracts of the edible parts (tuber
and dehulled bean) of Morama were constituted by fatty acids
while the polar extracts of the non-edible husks showed good anti-
oxidant properties but weak antimicrobial activity and cytotoxici-
ties. The known high content of amino acids, fatty acid in Morama
and its long storage period13 and now the isolation of the second-
ary metabolites from the different parts of T. esculentum, which
showed the absence of cynanogenic glucoside from the cotyledons,
add to the widely held view that Morama cotyledons has potential
as a food source, while the husk extracts antioxidant potential
need to be investigated further.
3.4.2. Cotyledons
The powdered cotyledons (952.7 g) were extracted with 70% eth-
anol which on evaporation and freeze drying gave 50.2 g extract,
which was defatted with chloroform to give 4.8 g extract. In the fatty
fraction 8 (9 mg) crystallized out. The mother liquor of this fraction
was prepared for GC–MS analysis. The residual extract (40 g) was
fractionated over silica gel (200 g) VLC column and eluted in increas-
ing polarity from n-hexane/CHCl3 (1:1) to pure chloroform, to which
was added methanol increments until 30% methanol in chloroform.
Seven fraction were collected and combined to give four main frac-
tions B1 (20.2 g), B2 (8 g), B3 (4 g) and B4 (12.2 g). Part of B1 (5 g) was
fractionated over silica gel column using EtOAc/MeOH/H2O (EMW)
7:2:1 (B1’–B4’), ethyl acetate/methanol/water (EMW) 6:3:1
(B5’–B6’) and 50% methanol (B7’–B8’). B5’–B6’ gave 9 (2 g), B7’–B8’
afforded 10 (3 g) while B4’, B2 and B3 yielded more 9 (800 mg). B4
was not interesting and was not worked on.
3. Experimental
3.1. General
Merck (Darmstadt, Germany) silica gel 60 (size 0.040–0.063 mm)
was used for column chromatography. TLC was carried out on
0.25 mm layer of Merck silica gel 60 F254 pre-coated on aluminium
sheets. Merck silica gel 60 HF254 + 366 nm coated on 20 x 20 cm glass
plates (0.5 mm thickness) were used for preparative TLC (PTLC). The
UV light (kmax 254 and 366 nm) and 1% vanillin-sulphuric acid spray
were used for visualization. The 1H NMR, 13C NMR, DEPT-135, COSY,
HMQC, HSQC, HMBC and TOCSY were acquired on Bruker Avance
DPX 300, and DRX 600 MHz. HR-MS were obtained on a GCT Premier
Mass Spectrometer (Waters) in EI Mode. UV-Vis spectra were
recorded on Shimadzu (UV-2101 PC) UV-Vis spectrometer (Kyoto,
Japan). IR spectra were measured on a Shimadzu Hyper FT-IR 8700
(Kyoto, Japan). Melting point was recorded (uncorrected) on Stuart
Scientific melting point apparatus SMP1 (UK). Specific rotation
3.4.3. Tubers
The powdered tubers (1.35 kg) were extracted sequentially with
EtOAc for three days and 70% MeOH for a further three days at 25 °C.
Solvent evaporation yielded 8.21 g EtOAc and 19.4 g 70% MeOH ex-
tracts. Part of theEtOAc extract (7 g) was applied to VLC silica gel giv-
ing seven fractions which were combined to give main fractions T1
(1.9 g), T2 (3.0 g), T3 (50 mg), T4 (90 mg) and T5 (0.9 g). T1 was ana-
lyzed using GC–MS. T2 gave white crystals (2.0 g) identified as behe-
nic acid 7. T3 yielded crystals (30 mg) which are yet to be identified.
T4 gave a mixture (11.2 mg) of 1 and 2. Repeated preparative TLC
(EMW 7:2:1) yielded only 1 (7 mg).
[a]D was determined through Autopol IV (Rudolph Research Analyt-
ical) Automatic Polarimeter at k = 589 nm.
3.2. Chemicals
The structures of all known compounds were identified based
Hydrogen peroxide, 1,1-diphenyl-2-picrylhydrazyl (DPPH) were
obtained from Aldrich (Munich, Germany) and Folin-Ciocalteu
on spectral data analysis (NMR, MS, IR, UV and [
with literature data.
a]) and comparison

O. Mazimba et al. / Bioorg. Med. Chem. 19 (2011) 5225–5230
5229
3.5. Griffonilide
were tested. Then, the cells were incubated at 37 °C in the humid-
ified incubator for 48 h. The plates were stained with Crystal Violet
in ethanol, rinsed with water, and destained with 10% (v/v) acetic
acid. The arbsorbance was measured at 590 nm, and the results
were expressed, for each dilution, by the mean ratios (%, SD) of
absorbances in treated wells to that in control wells.
The cell viability was determined by the following formula: %
Cell viability = (Mean absorbance in test wells/Mean absorbance
in control wells) Â 100.
Compound 1 was white solid; mp 180–182 °C; [
a]
D = À11.0 (c
0.98, MeOH); Rf: 0.51 (EMW, 7:2:1); UV/Vis (MeOH) kmax nm (
e)
254 (4.5); IR (KBr) mmax cmÀ1: 3371, 3076, 1726, 1649, 1375,
1166, 1082, 1010 cmÀ1 1H NMR (300 MHz, CD3OD) and 13C NMR
.
(75.5 MHz, CD3OD), 1H and 13C data in Table 1; HR-TOF EIMS m/z
168.0420 (calcd for C8H8O4 168.0423).
3.6. Compound 2
Dose–response curves between percentages of cell viability and
concentrations of the extracts were constructed. The results are ex-
pressed to show the concentration necessary to show 50% inhibi-
tion. The 50% inhibitory concentration (IC50), in the case of cell
inhibition, or effective concentration (ED50), in the case of cell
stimulation, was determined from the plotted curve. Correlations
were established using linear regression analysis, employing
Microsoft ExcelTM 2007 software.
Compound 2 was mixed with 1 (minor peaks). Mixture [a]
D
À14.0 (c 0.98, MeOH); Rf: 0.50 (EMW, 7:2:1). IR (KBr) mmax cmÀ1
:
3371, 2914, 1736, 1649, 1375, 1166, 1082, 1010 cmÀ1 1H and
;
13C data in Table 1; HR-TOF EIMS m/z 170.0576 (calcd for
C8H10O4 170.0579).
3.7. Bioassays
3.7.3. DPPH free radical scavenging assay
3.7.1. TLC bioautography antimicrobial testing
Methanolic samples (2 mL) of various concentrations (0.005,
0.05, 0.1, 0.2 and 0.5) mg/mL were mixed with 2 mL of 2% DPPH.
Morama milk was used as raw milk and 1 mL of the milk was di-
luted to 10 ml and treated as such. The absorbance of the mixture
was measured at 517 nm using a spectrophotometer (UV 2101,
Japan) at time intervals of 0.5, 1, 3, 6 and 24 h after incubation
in the dark at room temperature. The% of the DPPH radical
scavenged was calculated using the formula:% of DPPH scav-
enged = (AblankÀAsample/Ablank) Â 100. Where, Ablank is the absor-
bance of DPPH solution with zero sample concentration and
Asample is the absorbance of DPPH and known sample concentra-
tion. Ascorbic acid was used as a reference compound. The
experiment was performed in triplicate.
The microorganisms used were: S. aureus (ATCC 9144), E. coli
(ATCC 11229), B. subtilis (ATCC 6633), P. aeruginosa (NCTC 10332)
and C. albicans (ATCC 10231). The microorganisms were obtained
from the Department of Biological Sciences, University of Botswa-
na. Chloramphenicol and miconazole (Aldrich, Germany) were
used as standards for bacteria and fungi respectively.
The antimicrobial activities of extracts and compounds were
evaluated using agar-overlay bioautography method. Stock solu-
tions (10 mg/mL) of the test sample were prepared and serially-
diluted to obtain concentrations of 5.0, 1.0, 0.05, 0.001 mg/mL.
Aliquots of 10 lL of each concentration was spotted on Merck
pre-coated Silica gel 60 HF254 TLC plates (0.25 mm thickness;
10 cm  10 cm), corresponding to loading quantities of 100, 50,
10, 0.5 and 0.01
l
g. The spots were of the same size and the solvent
3.7.4. Total phenolic content
was allowed to evaporate in the fume hood. The seed layers were
prepared by inoculating 10 mL aliquot of culture into 100 mL agar
(Oxoid, UK: Sabouraud Dextrose) solution. 24-h old (37 °C) cul-
tures were used. Using a sterile Pasteur pipette the TLC plates were
overlaid with the agar. Plates were run in duplicates. After the
medium congealed the TLC plates were incubated at 37 °C (S. aur-
eus, E. coli) and 25 °C (B. subtilis, P. aeruginosa, C. albicans) for 24 h.
The bioautograms were sprayed with an aqueous solution of thia-
zoyl blue (methylthiazolyltetrazolium bromide (Sigma); 200 mg in
100 mL distilled water) and further incubated for 4 h after which
results were scored. White spots against purple background indi-
cated inhibition zones. The minimum inhibitory quantity (MIQ in
The procedure followed is in Yeboah & Majinda, 2009.29 Briefly,
1 mg/mL solution in methanol was prepared for each extract. The
fresh milk concentration was 1 mL in 10 mL of 90% methanol/
water. Five different concentrations of gallic acid in 90% aqueous
methanol ranging from 0.01 to 0.05 mg/mL were prepared.
Aliquots of 0.5 mL test sample (standard or extract) and 0.5 mL
of Folin-Cioucalteau reagent were added to a 10 mL screw cap
test-tube. After 3 min 1 mL of 20% sodium carbonate solution
was added. The mixture was shaken vigorously for 5 min and
allowed to stand for 2 h. Finally, 5 mL of 80% aqueous methanol
was added and the absorbance of the supernatant solution was
determined at 725 nm using a Shimadzu UV-2101 spectrophotom-
eter. The total phenolic content of the sample was determined by
comparing the optical density of the sample with those of different
concentrations of gallic acid, a standard phenolic compound.
Analysis for each sample was performed in triplicate, and values
for total phenolic content are expressed in milligrams of gallic acid
equivalent (GAE) per gram of samples.
lg) was taken as the lowest loading quantity to exhibit an inhibi-
tion zone.
3.7.2. Growth inhibition assay
The Caco-2, HeLa cells and normal cells (Vero) were used. All
cell lines were grown in advanced Dulbecco’s Modified Eagle’s
Medium (DMEM) (Sigma-Aldrich, Grand Island, USA), supple-
mented with 5% foetal calf serum (Lonza, Basel, Switzerland),
L-glutamine (2 mmol/L, Sigma), penicillin (100 U/mL, Sigma) and
streptomycin (1 mg/mL, Fluka, Buchs, Switzerland). Cell lines were
routinely grown in 25 cm2 culture flasks (Corning, New York, USA)
at 37 °C in a humidified atmosphere of 5% CO2 and 95% air, until
confluent monolayers were obtained. Culture medium was rou-
tinely changed. Before introduction of samples, the monolayers
3.7.5. Hydrogen peroxide scavenging activity
Hydrogen peroxide scavenging activity of the plant extract was
determined using methods described in Gulcin et al., 200430 and
Ogunlana and Ogunlana., 2008.31 Briefly, 32.7 mM solution of
hydrogen peroxide was prepared in phosphate-buffer (PBS): pH
7.4. Hydrogen peroxide concentration was determined spectropho-
tometrically (UV-2101) from absorbance (A0) at 230 nm using mo-
were washed twice with 100
supplements.
l
l DMEM without phenol red and
lar absorptivity of 81 MÀ1 cmÀ1. 12.5–250
lg plant extract and
standard (gallic acid and ascorbic acid) corresponding to 0.05,
0.1, 0.2, 0.5, 1 ml of 1 mg/mL plant extract stock solution were
added to 0.6 mL hydrogen peroxide-PBS solution. The total volume
of solution was 4 mL (topped-up using distilled water). Absorbance
of hydrogen peroxide at 230 nm was determined after 30 min (A1).
The inhibition study was performed when the cells grew up to
80–90% confluency and were seeded into 96-well plates at a den-
sity of 6 x 106 cells per well using method described in Agelis
et al. 2007.28 In assay 3-fold serial dilutions of each compound

5230
O. Mazimba et al. / Bioorg. Med. Chem. 19 (2011) 5225–5230
A blank solution containing phosphate buffer without hydrogen
peroxide was prepared for background correction. The percentage
of scavenged hydrogen peroxide of sample and standard com-
pounds was calculated using the following equation:% Scavenged
[H2O2] = [(A0ÀA1)/A0] Â 100. Where A0 was the absorbance of con-
trol and A1 was the absorbance in the presence of the sample. The
DPPH, TPC and H2O2 assays data were processed using Microsoft
ExcelTM 2007 software.
tests. This project was funded by the European Union under the
Morama II project (EU FP-6 grant (FP6-2004-INCO-DEV-3-MARA-
MA II – 032059).
A. Supplementary data
Supplementary data (1H and 13C NMR spectras for compounds 1
and 2) associated with this article can be found, in the online ver-
sion, at doi:10.1016/j.bmc.2011.07.006.
3.8. GC–MS analysis
References and notes
3.8.1. Non-polar extract preparation
1. Keegan, A. B.; Van Staden, J. S. Afr. J. Sci. 1981, 77, 387–390.
2. Francis, C. M.; Campbell, M. C. New high quality oil seed crops for temperate and
tropical Australia, RIRDC Publication No 03/045, 2003, pp 9-11.
3. Bower, N.; Hertel, K.; Oh, J.; Storey, R. Econ. Bot. 1988, 42, 533–540.
4. Liao, K. L.; Yin, M. C. J. Agri. Food Chem. 2000, 48, 2266–2270.
5. Han, X.; Shen, T.; Lou, H. Int. J. Mol. Sci. 2007, 8, 950–988.
6. Young, I. S.; Woodside, J. V. J. Clin. Pathol. 2001, 54, 176–186.
7. Antolovich, M.; Prenzler, P. D.; Patsalides, E.; McDonald, S.; Robards, K. The
Analyst 2002, 127, 183–198.
8. Tawaha, K.; Alali, F. Q.; Gharaibeh, M.; Mohammad, M.; El-Elimat, T. Food Chem.
2007, 104, 1372–1378.
9. Li, J.; Liu, J.; Lan, H.; Zheng, M.; Rong, T. Front. Agric. China 2009, 3, 40–42.
10. Hartley, L. M. PhD Thesis, Falmer, Sussex University, 1997, pp 27-31.
11. Yeaboah, S. O.; Ketshajwang, K. K.; Holmback, J. J. Am. Oil Chem. Soc. 1998, 75,
741–743.
12. Chingwaru, W.; Majinda, R. R. T.; Yeboah, S. O.; Jackson, J. C.; Kapewangolo, P.
T.; Kandawa-Schulz, M.; Cencic, A. Evid. Based Compl. Altern. Med. 2011.
doi:10.1155/2011/284795.
13. Jackson, J. C.; Duodu, K. G.; Holse, M.; de Faria, M. D.; Jordaan, D.; Chingwaru,
W.; Hansen, A.; Cencic, A.; Kandawa-Schultz, M.; Mpotokwane, S. M.;
Chimwamurombe, P.; de Kock, H. L.; Minnaar, A. Adv. Food Nutr. Res. 2010,
61, 187–246.
14. Dwuma-Badu, D.; Watson, W. H.; Gopalakrishna, E. M.; Okarter, T. U.; Knapp, J.
E., ; Schiff, P. L., Jr.; Slatkin, D. J. J. Nat. Prod. 1976, 39, 385–390.
15. Peter, R. A Master of Science dissertation. University of Botswana, Botswana,
2004, pp 49-50.
16. Jian-Hong, Y.; Kondratyuk, T. P.; Marler, L. E.; Qiu, X.; Choi, Y.; Cao, H.; Yu, R.;
Sturdy, M.; Pegan, S.; Liu, Y.; Wang, L. Q.; Mesecar, A. D.; Van Breemen, R. B.;
Pezzuto, J. M.; Fong, H. H. S.; Chen, Y. C.; Zhang, H. J. Phytochemistry 2010, 71,
641–647.
The chloroform cotyledons extract (4.6 g) and tuber EtOAc ex-
tract (Fraction T1, 1.9 g) provided the non-polar extracts after being
further extracted in a Soxhlet apparatus with n-hexane (2 h). The
solvent was allowed to evaporate at room temperature. The coty-
ledons and tuber yielded 1.6 g (35%) and 0.25 g (13%) viscous
non polar extracts respectively. The yield was expressed as weight
per weight (w/w) based on the dry weight of extract. The non-polar
extracts were dried over Na2SO4 and stored at 4 °C prior to qualita-
tive analysis by GC–MS in the conditions described.
3.8.2. FT-IR of non polar extracts
The viscous non-polar extracts were subjected to FT-IR (Shima-
dzu Hyper FT-IR 8700 spectrometer, KBr disc, 0.6 mg) analysis for
determination of their fatty acid nature.
3.8.3. GC–MS analysis
GC–MS analysis were performed using an HP-5 MS capillary
column (25 m  250
lm i.d., 0.25 lm film thickness, Agilent) in
an Agilent 6890 gas chromatograph coupled to a Waters GCT Pre-
mier mass spectrometer. The carrier gas was helium with a con-
stant flow rate of 1 mL/min. The oven temperature was initially
kept at 50 °C for 6 min then ramped at 4 °C/min to 230 °C then
gradually increased to 300 °C and held isothermally for 30 min.
Solutions of the samples (100 ppm in chloroform) were injected
17. Rowshanul Habib, M.; Nikkon, F.; Rahman, M.; Ekramul Haque, M.; Rezaul
Karim, M. Pakistan J. Biol. Sci. 2007, 10, 4174–4176.
18. Farooqi, J. A.; Ahmed, I.; Ahmed, M. Studies J. Oil Technol. Assoc. India 1983, 15,
25–26.
manually at 250 °C. Injection volume was 1.0 lL in the split-less
mode. Mass spectra were obtained by EI at electron energy of
70 eV.
19. Mazimba, O.
A Bachelor of Education Science Degree report; University of
Botswana: Botswana, 2001. pp 28–29.
20. Kapustina, I. I.; Makaréva, T. N.; Kalinovskii, A. I.; Stonik, V. A. Chem. Nat. Prod.
2005, 41, 109–110.
21. Hilditch, T. P.; Meara, M. L.; Patel, C. B. J. Sci. Food Agric. 2006, 2, 142–148.
22. Chao, C. H.; Hsieh, C. H.; Chen, S. P.; Lu, C. K.; Daid, C. F.; Sheu, J. H. Tetrahedron
Lett. 2006, 47, 5889–5891.
23. Fouraste, I.; Moulis, C. J. Nat. Prod. 2002, 65, 1180–1182.
24. Hodzic, Z.; Pasalic, H.; Memisevic, A.; Srabovic, M.; Saletovic, M.; Poljakovic, M.
Eur. J. Sci. Res. 2009, 28, 471–477.
25. De Oliveira, A. C.; Valentim, I. B.; Silva, C. A.; Bechara, E. J. H.; Paes de Barros,
M.; Mano, C. M.; Goulart, M. O. F. Food Chem. 2009, 115, 469–475.
26. Hemaisvarya, S.; Doble, M. Phytother. Res. 2006, 20, 239–249.
27. Badoni, R.; Semwal, D. K.; Rawat, U. J. Sci. Res. 2010, 2, 397–402.
28. Agelis, G.; Tzioumaki, N.; Botic, T.; Cencic, A.; Komiotis, D. Bioorg. Med. Chem.
2007, 15, 5448–5456.
3.8.4. Identification and quantification of constituents
The relative percent composition of the non-polar extract con-
stituents was determined by computerized peak area measure-
ments using internal normalization method. Identification of
components was based on GC retention time on the HP-5MS cap-
illary column, retention indices and by computer matching of the
acquired mass spectra with those stored in the spectrometer data
base using the NIST 05L Mass Spectral Library.9,32 The identity of
the spectra above 95% was considered for identification of
constituents.
29. Yeboah, E. M. O.; Majinda, R. R. T. Nat. Prod. Commun. 2009, 4, 89–94.
30. Gülçin, I.; Küfrevioglu, O. I.; Oktay, M.; Büyükokuroglu, M. E. J. Ethnopharmacol.
2004, 90, 205–215.
Acknowledgments
31. Ogunlana, O. E.; Ogunlana, O. O. Res. J. Agric. Biol. Sci. 2008, 4, 666–671.
32. Guido, F.; Pier, L. C.; Ivano, M.; Ammar, B. Food Chem. 2007, 100, 732–735.
Mr. I. Morobe of Department of Biological Sciences, University
of Botswana is thanked for his assistance during antimicrobial