Flavonoids from Lupinus texensis and their free radical scavenging activity

Article abstract of DOI:10.1080/14786419.2010.523423

Seventeen flavonoids including one new compound were isolated from Texas bluebonnet (Lupinus texensis), the state flower of Texas. Their structures were determined by extensive nuclear magnetic resonance and high-resolution electrospray ionization mass sp

Full text of DOI:10.1080/14786419.2010.523423

This article was downloaded by: [The Aga Khan University]
On: 25 December 2014, At: 04:09
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Natural Product Research: Formerly
Natural Product Letters
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/gnpl20
Flavonoids from Lupinus texensis and
their free radical scavenging activity
a
a
a
b
Zhizhen Zhang , Wei Yuan , Ping Wang , Greg Grant & Shiyou
a
Li
a
National Center for Pharmaceutical Crops, Arthur Temple
College of Forestry and Agriculture, Stephen F. Austin State
University , Nacogdoches, Texas 75962-6109, USA
b
Pineywoods Native Plant Center, Arthur Temple College of
Forestry and Agriculture, Stephen F. Austin State University ,
Nacogdoches, Texas 75962-6109, USA
Published online: 27 Jun 2011.
To cite this article: Zhizhen Zhang , Wei Yuan , Ping Wang , Greg Grant & Shiyou Li (2011)
Flavonoids from Lupinus texensis and their free radical scavenging activity, Natural Product
Research: Formerly Natural Product Letters, 25:17, 1641-1649, DOI: 10.1080/14786419.2010.523423
To link to this article: http://dx.doi.org/10.1080/14786419.2010.523423
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“
Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions

Natural Product Research
Vol. 25, No. 17, October 2011, 1641–1649
Flavonoids from Lupinus texensis and their free radical
scavenging activity
a
a
a
b
a
Zhizhen Zhang , Wei Yuan , Ping Wang , Greg Grant and Shiyou Li *
a
National Center for Pharmaceutical Crops, Arthur Temple College of Forestry and
Agriculture, Stephen F. Austin State University, Nacogdoches, Texas 75962-6109, USA;
b
Pineywoods Native Plant Center, Arthur Temple College of Forestry and Agriculture,
Stephen F. Austin State University, Nacogdoches, Texas 75962-6109, USA
(
Received 19 February 2010; final version received 28 July 2010)
Seventeen flavonoids including one new compound were isolated from
Texas bluebonnet (Lupinus texensis), the state flower of Texas. Their
structures were determined by extensive nuclear magnetic resonance and
high-resolution electrospray ionization mass spectrometry analyses. High-
performance liquid chromatography analytic method for simultaneous
determination of the 17 compounds was established and validated. Eleven
isolated flavonoids were first evaluated for their free radical scavenging
activity using ꢀ,ꢀ-diphenyl-ꢁ-picrylhydrazyl scavenging assay and they
ꢀ1
showed activity with EC50 48.6–172.5 mg mL
.
Keywords: Lupinus texensis; Fabaceae; flavonoids; HPLC analysis;
free radical scavenging activity
1
. Introduction
The genus Lupinus L. (Fabaceae) has over 400 species worldwide. It is well known
that alkaloids (Reinhard, Rupp, Sager, Streule, & Zollern, 2006), flavonoids (Garcıa-
Lopez, Kachlicki, Zamora-Natera, Ruiz-Moreno, & Stobiecki, 2006; Tahara,
Ingham, Nakahara, Mizutani, & Harboren, 1984) and triterpenoid saponins
Woldemichael, Montenegro, & Timmermann, 2003; Woldemichael & Wink, 2002)
´
´
(
are the main secondary metabolites of this genus. To date, more than 150 alkaloids
of the quinolizidine, piperidine and indole groups have been found in lupins
(
Reinhard et al., 2006). Lupin alkaloids such as lupanine, sparteine, multiflorine and
N-methylcytisine have been reported to possess hypoglycemic activity (Kubo et al.,
000; Mohamed, Kamel, El-Moghazy, & Murakoshi, 1993), while the triterpenoid
saponins from lupins have antifungal activity against Candida albicans
Woldemichael & Wink, 2002) and hepatoprotective property (Kinjo, Udayama,
2
(
Okawa, & Nohara, 1999).
Lupins have long been used as a source of protein and oil in the Andean region of
South America (Woldemichael et al., 2003). Because lupins and soya have
comparable nutritive characteristics, they are possible substitutes for genetically
*Corresponding author. Email: lis@sfasu.edu
ISSN 1478–6419 print/ISSN 1478–6427 online
ß 2011 Taylor & Francis
http://dx.doi.org/10.1080/14786419.2010.523423
http://www.tandfonline.com

1
642
Z. Zhang et al.
modified soya in human foodstuffs (Reinhard et al., 2006; van Barneveld, 1999).
Four lupine species are of agronomic interest: Lupinus albus (white lupin), Lupinus
angustifolius (blue lupin), Lupinus luteus (yellow lupin) and Lupinus mutabilis
(Andean lupin) (Reinhard et al., 2006).
Lupinus texensis Hook., known as Texas bluebonnet, Texas lupine, buffalo
clover, wolf flower and (in Mexican) el conejo, is the state flower of Texas. It is an
annual plant endemic to Texas. This species has larger more sharply pointed leaves
and more numerous flower heads than similar lupins and is often planted by the
Texas Department of Transportation and garden clubs. To the best of our
knowledge, there has been no report on the chemical constituents, nor a recorded
traditional use or agronomic interest of this plant. Our interest in the discovery of
novel bioactive natural products (Zhang & Li, 2007; Zhang, Wang, Yuan, & Li,
2
008) from native plants in Texas prompted us to investigate the chemical
constituents of L. texensis. This study described the isolation and structure
determination of 17 flavonoids including one new compound from L. texensis and
reported the free radical scavenging activities of 11 flavonoids for the first time. In
addition, a validated HPLC method for simultaneous analysis of the 17 compounds
was also presented.
2
. Results and discussion
Repeated column chromatography following by HPLC purification of an ethanol
extract of the whole plant of L. texensis yielded one new flavonoid (2), together with
1
(
6 known ones (1, 3–17) (Figure 1). On the basis of the nuclear magnetic resonance
NMR) analysis and the comparison of the reference data, the known compounds
0
were identified as genistein 7,4 -di-O-ꢁ-D-glucopyranoside (1) (Watanabe, Kinjo, &
Nohara, 1993), pratensein 7-O-ꢁ-D-glucopyranoside (3) (Anhut et al., 1984), orientin
0
(
(
4) (Burns, Ellis, & March, 2007), vitexin (5) (Burns et al., 2007), 2 -hydroxygenistein
6) (Tahara et al., 1984), diosmetin 7-O-ꢁ-D-glucopyranoside (7) (Wang, Yang, &
Guo, 2008), genistein (8) (Tahara et al., 1984), acacetin (9) (Rahman & Moon, 2007),
diosmetin (10) (Domınguez, Marın, Esquivel, & Cespedes, 2007), crotarin (11)
Chaturvedi, Pant, Garg, & Bhakuni, 1987), 3-[2,3-dihydro-4-hydroxy-2-(1-hydroxy-
´
´
´
(
1
-methylethyl)-7-benzofuranyl]-5,7-dihydroxy-4H-1-benzopyran-4-one (12) (Tahara,
Nakahara, Mizutani, & Ingham, 1985), erythrinin C (13) (Deshpande, A. Pendse, &
R. Pendse, 1977), sophoraisoflavone A (14) (Yoshiaki, Ichiro, Mami, Tsuyoshi, &
Manki, 1988), erythrinin B (15) (Deshpande et al., 1977), licoisoflavone A (16)
(
1
Kinoshita, Saitoh, & Shibata, 1978) and lupinisoflavone A (17) (Tahara et al.,
984).
Compound 2, a yellow powder, had a molecular formula of C H O as
2
2
22 11
þ
1
deduced from its HRESIMS at m/z 463.1229 [M þ H] and NMR data. The H-
0
NMR of 2 showed one methoxy signal at ꢂ 3.89 (3H, s, H-4 ) and six aromatic proton
signals, of which one appeared as a singlet at ꢂ 8.09 and was assigned to H-2, two
were doublets at ꢂ 6.53 and 6.73 (J ¼ 2.1 Hz, each) and attributed to H-6 and H-8,
0
0
and three at ꢂ 6.40 (d, J ¼ 2.3 Hz, H-3 ), 6.37 (1H, dd, J ¼ 8.3/2.3 Hz, H-5 ), and 7.05
0
0
0
0
(
d, J ¼ 8.3 Hz, H-6 ) corresponded to an 1 ,2 ,4 -oxygenation-aromatic ring. These
0
0
data indicated that the aglycone of 2 was 5,7,2 -trihydroxy-4 -methoxyisoflavone
dehydroferreirin) (Aparecida dos Santos, Geraldo de Carvalho, & Braz-Filho,
(

Natural Product Research
1643
OH
HO
O
O
O
O
O
HO
OH
OH
OH
OH
OR3
OH
HO
O
OH
OH
R1
R2
1
2
3
R1 = H, R2 = H, R3 = b-D-glucopyranosyl
R1 = OH, R2 = H, R3 = CH3
R1 = H, R2 = OH, R3 = CH3
R
OH
O
4
5
R = OH
R = H
HO
O
OH
R1O
O
OCH3
OH
O
R
R2
6
8
R = OH
R = H
OH
O
HO
O
7
9
R1 = b-D-glucopyranosyl, R2 = OH
R1 = H, R2 = H
OH
10 R1 = H, R2 = OH
OH
O
HO
O
O
O
11
OH
O
O
OH
O
HO
OH
OH
1
2
OH
OH
O
O
HO
O
13
HO
OH
OH
O
O
OH
O
1
4
15
HO
O
O
O
OH
OH
OH
O
HO
OH
O
HO
16
17
Figure 1. Chemical structures of flavonoids 1–17.
1
3
1
995). The C-NMR spectra of 2 exhibited 22 carbon signals, 16 of these signals
were identical to those reported for dehydroferreirin (Aparecida dos Santos et al.,
995). The remaining six carbon signals were indicative of the presence of one hexose.
1
Acidic hydrolysis of 2 produced dehydroferreirin identified by NMR spectroscopic
analysis with the reference data (Aparecida dos Santos et al., 1995), and D-glucose
confirmed by co-HPLC analysis with authentic sugar and by measurement of optical
activity after separation by HPLC. The monosaccharide was assigned to be in
1
3
the pyranose form from analysis of its C-NMR data. The ꢁ-anomeric configuration
(161 Hz). A key HMBC
CH
3
of glucose was defined from its J
1
(7.3 Hz) and J
H1,H2
correlation between ꢂ 5.06 (glc-H-1) and ꢂ 164.1 (C-7) established that the glucose
was attached to the assigned position. The full assignments of the protons and
1 13
carbons were achieved by a combination of H, C, HMQC and HMBC

1
644
Z. Zhang et al.
0
experiments. The structure of 2, a new isoflavone glycoside, was determined as 5,7,2 -
0
trihydroxy-4 -methoxyisoflavone 7-O-ꢁ-D-glucopyranoside (dehydroferreirin 7-O-ꢁ-
D-glucopyranoside).
The chemical profiles of L. texensis were characterised by defining and verifying
the chemical ingredients and determining the concentration of 17 individual flavo-
noids using HPLC method. The simultaneous separation of these compounds was
achieved on a reverse-phase HP Hypersil C18 column (250 ꢁ 4.6 mm, 5 mm) with grade
mobile system consisting of methanol and 0.5% HOAc in water in 90 min. Fourteen
compounds isolated and identified from the whole plant of Texas bluebonnet were
used as the marker compounds. The content of compounds 9, 13 and 15 were calcu-
lated from the linear regression equation of compounds 10, 12 and 16, respectively.
Table 1 presented the concentration of individual and total flavonoid compounds.
The results showed that Texas bluebonnet contained a total of over 7.79% flavonoids
with vitexin (5) as the major one (3.70%), following by genistein (8, 0.84%).
This HPLC method was validated for linearity and reproducibility. The linearity
of the standard curves was studied for compounds 1–8, 10–12, 14, 16, 17. The
calibration curve of each compound was investigated between the peak area (y) and
the quantity of each reference compound (x, mg). The linearity is expressed in terms
of the correlation coefficient (r). The correlation coefficient was found to be better
ꢀ
1
than 0.999 for most of the reference compounds in the range of 25–300 mg mL
.
The reproducibility of the method was assessed by within and between run
validations. RSD% and RE% were found to be ꢂ2.05% and 1.93%, respectively.
Accelerated solvent extraction (ASE) was used for the preparation of sample
solution for HPLC analysis. The recovery of the 14 reference compounds extracted
using ASE with 95% ethanol as solvent was investigated. The results showed that the
recovery for each analysed compound was over 98.1%, indicating that using ASE
with 95% ethanol as extract solvent is a validated method to prepare the sample
solution for HPLC analysis.
All tested flavonoids (1–3, 6, 11–17) have free radical scavenging activities with
ꢀ1
EC50 48.6–172.5 mg mL (Table 1). To the best of our knowledge, the free radical
scavenging activities of these flavonoids were reported for the first time.
Table 1.
(
Contents of flavonoids and their free radical scavenging activities
mean ꢃ SD, n ¼ 3).
ꢀ1
ꢀ1
)
Compound Content (%) EC50 (mg mL
)
Compound Content (%) EC50 (mg mL
1
2
3
4
5
6
7
8
9
0.427 ꢃ 0.005
0.320 ꢃ 0.022
0.203 ꢃ 0.009
0.670 ꢃ 0.022
3.707 ꢃ 0.009
0.207 ꢃ 0.017
0.163 ꢃ 0.012
0.843 ꢃ 0.009
0.227 ꢃ 0.005
172.5 ꢃ 2.3
10
11
12
13
14
15
16
17
0.290 ꢃ 0.008
0.113 ꢃ 0.005
0.053 ꢃ 0.005
0.163 ꢃ 0.005
0.053 ꢃ 0.005
0.083 ꢃ 0.005
0.143 ꢃ 0.005
0.123 ꢃ 0.005
7.790 ꢃ 0.043
–
93.1 ꢃ 1.9
71.8 ꢃ 1.8
82.8 ꢃ 2.3
98.1 ꢃ 1.9
78.5 ꢃ 1.6
73.8 ꢃ 1.7
56.3 ꢃ 1.5
48.6 ꢃ 2.0
74.2 ꢃ 1.2
a
a
–
–
66.9 ꢃ 1.8
–
–
–
a
Total
a
Notes: The content of compounds 9, 13 and 15 was calculated from the linear regression
equation of compounds 10, 12 and 16, respectively.

Natural Product Research
1645
3
3
. Experimental
.1. General experimental procedures
NMR experiments were performed on a Bruker 600 MHz NMR instrument. NMR
data were reported as ꢂ (ppm) values and referenced to the solvent used. HRESIMS
were acquired on an electrospray instrument (MDS Sciex Pulsar Qstar). The UV
spectra were recorded in MeOH with a mQuant spectrophotometer (Bio-Tek
Instruments, Inc.). Optical rotation values were measured on a Jasco P-1010
polarimeter. Octadecyl-functionalised silica gel (ODS, Aldrich) and silica gel (60–200
˚
mesh, 60 A, Aldrich) were used for low-pressure chromatography. HPLC analysis
was performed on an Agilent 1100 HPLC system with an Agilent 1100 diode array
detector or an Agilent 1100 refractive index detector using a Hypersil ODS column
(
column A, 250 ꢁ 4.6 mm, 5 mm, Supelco) or a SupelcoGel CA column for sugar
(
column B, 300 ꢁ 7.8 mm, Supelco). Preparative HPLC was performed with an
Acuflow Series III pump connected with an Acutect 500 UV/VIS detector using an
Econosil ODS column (250 ꢁ 22 mm, 10 mm, Alltech; Detection: 256 nm; flow rate:
ꢀ1
3.5 mL min ). ASE (Dionex model ASE 200) was used to prepare the sample
solution for HPLC analysis.
3
.2. Plant material
The whole plants of L. texensis were collected in Nacogdoches, Texas, USA in April
2
2
007 and were identified by Greg Grant. A voucher specimen (TX-NAC-SFATC-
0070402-#0002-LT) and a sample of the experimental plant material (NCPC-
LT20070402) were deposited at the National Center for Pharmaceutical Crops at
Stephen F. Austin State University, USA.
3
.3. Extraction and isolation
The air-dried whole plant materials (500 g) were ground to a coarse powder and
percolated with 95% ethanol thrice (each 4000 mL). The combined percolated
solution was concentrated under vacuum to give an ethanol extract (40.5 g). This
extract was suspended in MeOH–H O (1 : 1, v/v), and then partitioned successively
2
with hexanes, chloroform and n-butanol to yield hexane (5.0 g), chloroform (10.1 g),
n-butanol (15.1 g) and aqueous parts (10.1 g). The n-butanol part was applied to a
column (60 ꢁ 5 cm) of ODS (300 g) eluting with water (1000 mL) and then 30%
methanol (2000 mL). The fractions 8–13 (each 100 mL) collected from the 30%
methanol eluate were combined and then dried to afford fraction A (1.6 g). Part of
fraction A (300 mg) was separated by preparative HPLC using methanol/0.5% HOAc
in water (35:65) as flow phase to yield compounds 1 (18.9 mg, t 30 min), 2 (15.8 mg, t
R
R
4
3 min), 3 (9.6 mg, t 48 min), 4 (22.6 mg, t 56 min) and 5 (123.8 mg, t 70 min). The
R R R
chloroform part (5.0 g) was chromatographed on a column (60 ꢁ 5 cm) of silica gel
500 g) eluting with a mixture of chloroform and methanol (19 : 1, 3000 mL) to give
fractions B1 (200 mg, 300–350 mL), B2 (100 mg, 450–550 mL), B3 (60 mg,
(
6
00–750 mL), B (600 mg, 800–900 mL), B (126 mg, 1150–1200 mL), B (208 mg,
4 5 6
1450–1500 mL), B7 (300 mg, 1600–1700 mL) and B8 (150 mg, 1900–2000 mL).
Compounds 8 (13.6 mg), 6 (12.2 mg), 17 (8.6 mg) and 7 (9.5 mg) were crystallised
from fractions B , B , B and B , respectively. Fraction B was purified by preparative
7
1
2
8
3

1
646
Z. Zhang et al.
HPLC (MeOH/0.5% HOAc in water, 70:30) to yield 9 (6.1 mg, tR 36.0 min).
Compounds 10 (13.5 mg, tR 49.7 min), 11 (7.9 mg, tR 58.6 min) and 12 (7.3 mg, tR
6
from fraction B . Similarly, fraction B was separated by preparative HPLC (MeOH/
6.6 min) were separated by preparative HPLC (MeOH/0.5% HOAc in water, 65:35)
4
5
0
6
.5% HOAc in water, 70 : 30) to give 13 (11.7 mg, tR 53.6 min) and 14 (7.8 mg, tR
1.7 min); and fraction B was separated by preparative HPLC (MeOH/0.5% HOAc
6
in water, 70:30) to give 15 (8.3 mg, tR 63.2 min) and 16 (9.1 mg, tR 68.5 min).
All isolated compounds have purity of over 95% determined by HPLC analysis.
0 0
.3.1. 5,7,2 -Trihydroxy-4 -methoxyisoflavone 7-O-ꢁ-D-glucopyranoside (2)
3
2
D
5
ꢅ
1
Yellow powder; ½ꢀꢄ þ 39.6 (c 0.1, MeOH); H-NMR (DMSO-d , 600 MHz): ꢂ 8.09
6
0
(
1H, s, H-2), 7.05 (1H, d, J ¼ 8.3 Hz, H-6 ), 6.73 (1H, d, J ¼ 2.1 Hz, H-8), 6.53 (1H, d,
0
0
J ¼ 2.1 Hz, H-6), 6.40 (1H, d, J ¼ 2.3 Hz, H-3 ), 6.37 (1H, dd, J ¼ 8.3/2.3 Hz, H-5 ),
5
.06 (1H, d, J ¼ 7.3 Hz, glc-H-1), 3.91 (1H, dd, J ¼ 12.2/2.0 Hz, glc-H-6a), 3.89 (3H,
0
s, OCH -4 ), 3.70 (1H, dd, J ¼ 12.2/5.6 Hz, glc-H-6b), 3.53 (1H, m, glc-H-5), 3.50
3
1
1H, m, glc-H-3), 3.47 (1H, m, glc-H-2) and 3.40 (1H, m, glc-H-4)); C-NMR
3
(
(DMSO-d , 150 MHz): ꢂ 182.2 (C, C-4), 164.1 (C, C-7), 163.3 (C, C-5), 159.7 (C,
C-2 ), 158.9 (C, C-9), 156.8 (C, C-4 ), 156.5 (CH, C-2), 132.6 (CH, C-6 ), 122.1 (C,
6
0
0
0
0
0
0
C-3), 109.8 (C, C-1 ), 107.4 (CH, C-5 ), 106.8 (C, C-10), 103.1 (CH, C-3 ), 100.3 (CH,
1
JCH ¼ 161 Hz, glc-C-1), 100.1 (CH, C-6), 94.7 (CH, C-8), 78.0 (CH, glc-C-5),
7
5
7.6 (CH, glc-C-3), 74.9 (CH, glc-C-2), 70.6 (CH, glc-C-4), 62.1 (CH , glc-C-6),
2
6.0 (CH3, OCH -4 ); HRESIMS: m/z 463.1229 [M þ H]þ {Calcd for
0
3
[
M(C H O ) þ H], 463.1240}.
2
2
22 11
3
.4. Acidic hydrolysis of compound 2
Compound 2 (7.0 mg) was refluxed with 2 mL of 1 M HCl (dioxane–H O, 1 : 1) at
2
ꢅ
8
0 C for 2 h. Following dioxane removal, the solution was extracted with EtOAc
(
5 mL ꢁ 3). The extraction was washed with water and then concentrated.
The residue was purified by preparative HPLC using methanol/0.5% HOAc in
0
0
water (55 : 45) to give 5,7,2 -trihydroxy-4 -methoxyisoflavone (dehydroferreirin,
.6 mg). The monosaccharide portion was neutralised by passing through an ion-
3
exchange resin (Amberlite MB-3) column, and then concentrated for analysis.
The monosaccharide of 2 was determined to be glucose (tR 11.6 min) based on
co-HPLC analysis with reference sugar (Column B; Detector: refractive index;
ꢀ
1
ꢅ
Mobile phase: deionised water; Flow rate: 0.5 mL min ; Temperature: 78 C). The
monosaccharide portion of 2 was then separated by HPLC using the above HPLC
ꢅ
conditions and the optical activity of the monosaccharide (D-glucose, þ46.6 , c. 0.18,
H O) was measured.
2
3
.5. HPLC analysis
HPLC analysis was performed on an Agilent 1100 HPLC system with an Agilent
100 diode array detector using a Hypersil ODS column (250 ꢁ 4.6 mm, 5 mm,
1
Supelco). The water with 0.05% acetic acid was employed as mobile phase A, and
methanol as mobile phase B. The gradient procedure was 0–25 min with 10–30% B,

Natural Product Research
1647
2
6
8
5–39 min with 30–40% B, 39–40 min with 40% B, 40–60 min with 40–65% B,
0–80 min with 65–90% B, 80–85 min with 90% B, 85–86 min with 90–10% B and
6–90 min with 10% B. The flow rate was 1.0 mL min . The column temperature
ꢅ
ꢀ1
was set at 36 C, and detection was carried out at 256 nm. Data collection and
integration were performed using HP ChemStation software.
3
.5.1. Preparation of sample solution for HPLC analysis
ASE (Dionex model ASE 200) was applied to prepare the sample solution for HPLC
analysis; 95% ethanol was used as the extract solvent. To prepare a single extraction
cartridge, about 20 g of 30–40-mesh sand SX0075-3 (EM Sciences, Gibbstown, NJ)
was mixed with 2.0 g of plant materials. Extractions were conducted at temperature
ꢅ
85 C with a pressure of 800 psi and 120 s nitrogen purges for a total of 30 min with
three 10 min cycles. The solution of each extraction was transferred into a 50 mL
ꢀ
1
volumetric flask and diluted with DMSO to make a final solution of 40 mg mL
.
The solution of each sample was filtered through a 0.2 mm nylon cartridge before
HPLC analysis. Extraction for each sample was performed in triplicate.
3
.5.2. Reference compounds
A total of 14 flavonoids (1–8, 10–12, 14, 16 and 17; purity: 95–98%) isolated and
purified from L. texensis were used as reference compounds. Each reference
compound (3.0 mg) was accurately weighed into a 1.0 mL volumetric flask and
ꢀ1
dissolved in methanol to make a stock solution (3.0 mg mL ) of the 14 components.
The final solution for HPLC analysis was prepared from the stock solution.
3
.5.3. Linear regression equation and calibration curve
The linear regression equation and the calibration curve were investigated between
peak area (y) and the quantity (x, mg) of each component. Seven injections were
ꢀ1
performed to obtain the absorption plots ranging from 25–300 mg mL in seven
increments.
3
.5.4. Reproducibility of the HPLC method
The precision and accuracy of the HPLC method were assessed by within and
between run validations. The variation was evaluated by three injections of a mixture
of 14 reference compounds each day on three consecutive days after the preparation
of reference compound solution. By substituting the peak area into the linear
regression equation from the same run, the measured concentrations were obtained.
The coefficient of variance RSD% was calculated by comparing the measured
concentrations. The relative error RE% was obtained by comparing calculated and
theoretical concentrations.
3
.5.5. Recovery of the 14 reference compounds
A certain amount of each reference compound was added into the plant material
2.0 g) of L. texensis, and the mixtures were extracted by ASE using 95% ethanol as
solvent as described previously. The extract solution was transferred into a 50 mL
(
ꢀ
1
volumetric flask, diluted with DMSO to make a final solution of 40 mg mL and

1
648
Z. Zhang et al.
then filtered through a 0.2 mm nylon cartridge prior to HPLC analysis. For
comparison, a blank sample without spiking with marker compounds was prepared
and similarly analysed.
3
.6. Determination of the free-radical-scavenging activity
The ꢀ,ꢀ-diphenyl-ꢁ-picrylhydrazyl (DPPH) scavenging assay was carried out
according to the procedure described previously (Tadic et al., 2008). Briefly,
different concentrations of the samples (100 mL) were mixed with 900 mL of
´
ꢀ1
0
temperature for 20 min and then the UV absorbance at 517 nm was measured.
.04 mg mL
methanolic solution of DPPH. The mixture was kept at room
The inhibition percentage was calculated using the following equation: I ¼
[
(A ꢀ A )/A ] ꢁ 100, where I was the inhibition percentage, A the absorbance of
c
s
c
c
the negative control (containing 100 mL of methanol instead of the tested samples)
and A the absorbance of the samples. The inhibition percentage was plotted against
the concentration of the samples, and EC50 values (mean ꢃ SD) were determined by
s
linear regression analysis of three determinations.
Acknowledgements
This study was funded by the US CDC grant R01 CI000315-01 and USDA grant 2008-38928-
19308. The authors thank Keck/IMD NMR Center, which is funded by the W.M. Keck
Foundation and the University of Houston; Dr Youlin Xia for NMR analysis assistance; and
Dr Shane Tichy of Texas A&M University for HRESI-MS analysis.
References
Anhut, S., Zinsmeister, H.D., Mues, R., Barz, W., Mackenbrock, K., Ko
K.R. (1984). The first identification of isoflavones from a bryophyte. Phytochemistry,
3, 1073–1075.
Aparecida dos Santos, S., Geraldo de Carvalho, M., & Braz-Filho, R. (1995). Unambiguous
¨
ster, J., & Markham,
2
1
13
H- and C-NMR assignments of isoflavones from Virola caducifolia. Journal of the
Brazilian Chemical Society, 6, 349–352.
1
3
Burns, D.C., Ellis, D.A., & March, R.E. (2007). A predictive tool for assessing C-NMR
chemical shifts. Magnetic Resonance in Chemistry, 45, 835–845.
Chaturvedi, R., Pant, N., Garg, H.S., & Bhakuni, D.S. (1987). Isoflavanoids of Crotalaria
madurensis. Journal of Natural Products, 50, 266–269.
Deshpande, V.H., Pendse, A.D., & Pendse, R. (1977). Erythrinins A, B and C, three new
isoflavones from the bark of Erythrina variegata. Indian Journal of Chemistry B, 15,
205–207.
nguez, M., Marın, J.C., Esquivel, B., & Cespedes, C.L. (2007). Pensteminoside, an
Domı
´
´
´
unusual catalpol-type iridoid from Penstemon gentianoides HBK (Plantaginaceae).
Phytochemistry, 68, 1762–1766.
Garcı
´
a-Lo
2006). Profiling isoflavone conjugates in different organs of Lupinus exaltatus Zucc.
Phytochemical Analysis, 17, 184–191.
´
pez, P.M., Kachlicki, P., Zamora-Natera, F., Ruiz-Moreno, J., & Stobiecki, M.
(
Kinjo, J., Udayama, M., Okawa, M., & Nohara, T. (1999). Study of structure –
hepatoprotective relationships of oleanene-type triterpenoidal glucuronides obtained
from several fabaceous plants on rat primary hepatocyte cultures. Biological and
Pharmaceutical Bulletin, 22, 203–206.

Natural Product Research
1649
Kinoshita, T., Saitoh, T., & Shibata, S. (1978). Chemical studies on the oriental plant drugs.
XLIII. A new isoflavone from licorice root. Chemical and Pharmaceutical Bulletin, 26,
141–143.
Kubo, H., Kobayashi, J., Higashiyama, K., Kamei, J., Fuji, Y., & Ohmiya, S. (2000). The
hypoglycemic effect of (7R,9aS)-7-phenyloctahydroquinolizin-2-one in mice. Biological
and Pharmaceutical Bulletin, 23, 1114–1117.
Mohamed, M.H., Kamel, M.S., El-Moghazy, S.A., & Murakoshi, I. (1993). The hypoglycemic
effect of some lupin alkaloids. Bulletin of Faculty of Pharmacy (Cairo University),
31, 107–111.
Rahman, A.A., & Moon, S.S. (2007). Antimicrobial phenolic derivatives from Dendranthema
zawadskii var. latilobum Kitamura (Asteraceae). Archives of Pharmacal Research,
30, 1374–1379.
Reinhard, H., Rupp, H., Sager, F., Streule, M., & Zollern, O. (2006). Quinolizidine alkaloids
and phomopsins in lupin seeds and lupin containing food. Journal of Chromatography
A, 1112, 353–360.
Tadic
´
, V.M., Dobric, S., Markovic, G.M., Dordevic, S.M., Arsic, I.A., Menkovic, N.R., &
Stevic, T. (2008). Anti-inflammatory, gastroprotective, free-radical-scavenging, and
antimicrobial activities of hawthorn berries ethanol extract. Journal of Agricultural and
Food Chemistry, 56, 7700–7709.
Tahara, S., Ingham, J.L., Nakahara, S., Mizutani, J., & Harboren, J.B. (1984). Fungitoxic
dihydrofuranoisoflavones and related compounds in white lupin, Lupinus albus.
Phytochemistry, 23, 1889–1890.
Tahara, S., Nakahara, S., Mizutani, J., & Ingham, J.L. (1985). Fungal metabolism of the preny-
lated isoflavone licoisoflavone A. Agricultural and Biological Chemistry, 49, 2605–2612.
van Barneveld, R.J. (1999). Understanding the nutritional chemistry of lupin (Lupinus spp.)
seed to improve livestock production efficiency. Nutrition Research Reviews, 2, 203–230.
Wang, Y.J., Yang, X.W., & Guo, Q.S. (2008). Studies on chemical constituents in Huangjuhua
(
flowers of Chrysanthemum morifolium). Zhongguo Zhongyao Zazhi, 33, 526–530.
Watanabe, K., Kinjo, J., & Nohara, T. (1993). Leguminous plants. XXXIX. Three new
isoflavonoid glycosides from Lupinus luteus and L. polyphyllus ꢁ arboreus. Chemical &
Pharmaceutical Bulletin, 41, 394–396.
Woldemichael, G.M., Montenegro, G., & Timmermann, B.N. (2003). Triterpenoidal lupin
saponins from the Chilean legume Lupinus oreophilus Phil. Phytochemistry, 63, 853–857.
Woldemichael, G.M., & Wink, M. (2002). Triterpene glycosides of Lupinus angustifolius.
Phytochemistry, 60, 323–327.
Yoshiaki, S., Ichiro, Y., Mami, N., Tsuyoshi, T., & Manki, K. (1988). Studies on the
constituents of Sophora species. XXII. Constituents of the root of Sophora
moorcroftiana Benth. ex Baker. (1). Chemical & Pharmaceutical Bulletin, 36, 2220–2225.
Zhang, Z.Z., & Li, S.Y. (2007). Cytotoxic triterpenoid saponins from the fruits of Aesculus
pavia L. Phytochemistry, 68, 2075–2086.
Zhang, Z.Z., Wang, P., Yuan, W., & Li, S.Y. (2008). Steroids, alkaloids, and coumarins from
Gelsemium sempervirens. Planta Medica, 74, 1818–1822.