Anti-inflammatory isoflavonoids from the stems of Derris scandens

Article abstract of DOI:10.1055/s-2004-827147

Fractionation of the aqueous extract of Derris scandens stems extract using tests for eicosanoid inhibition resulted in the isolation of three isoflavonoids, genistein, its 7-O-α-rhamno(1→6)-β-glucosyl glycoside, a new compound, and two known isoprenyl derivatives 3′-γ,γ-dimethylallylweighteone and scandenin. The isoprenylated compounds showed a high inhibitory effect on eicosanoid production in vitro but HPLC analysis showed that the genistein accounted for most of the activity of the total extract. Antioxidant studies showed that genistein and the isoprenylated compounds showed activity comparable to standard antioxidants. Genistein and its glycoside demonstrated no cytotoxicity in the MTT test but the prenylated compounds showed some toxicity and also increased LDH release from polymorphonucleocytes, at concentrations much greater than would be encountered in an aqueous extract of D. scandens.

Full text of DOI:10.1055/s-2004-827147

Pisamai Laupattarakasem1,2
Peter J. Houghton1
J. Robin S. Hoult3,4
Anti-Inflammatory Isoflavonoids from the Stems of
Derris scandens
Abstract
compounds showed activity comparable to standard antioxi-
dants. Genistein and its glycoside demonstrated no cytotoxicity
Fractionation of the aqueous extract of Derris scandens stems ex- in the MTT test but the prenylated compounds showed some
tract using tests for eicosanoid inhibition resulted in the isola- toxicity and also increased LDH release from polymorphonucleo-
tion of three isoflavonoids, genistein, its 7-O-a-rhamnoꢀ1®6)-b- cytes, at concentrations much greater than would be encounter-
glucosyl glycoside, a new compound, and two known isoprenyl ed in an aqueous extract of D. scandens.
derivatives 3¢-g,g-dimethylallylweighteone and scandenin. The
isoprenylated compounds showed a high inhibitory effect on ei- Key words
cosanoid production in vitro but HPLC analysis showed that the Derris scandens ´ Leguminosae ´ isoflavonoids ´ eicosanoid inhibi-
genistein accounted for most of the activity of the total extract. tion ´ antioxidant ´ anti-inflammatory
Antioxidant studies showed that genistein and the isoprenylated
4
96
Introduction
vitro assays related to inflammation. These assays explored var-
ious functions of neutrophils, the metabolic products of arachi-
Aqueous and ethanolic extracts of the stems of Derris scandens donic acid and the role played by reactive oxygen species ꢀROS).
Benth. ꢀLeguminosae) were found to be the most active of those
from four species used to treat arthritis in traditional Thai medi- The release of elastase myeloperoxidase ꢀMPO) and lactase dehy-
cine when tested in a variety of in vitro and in vivo test systems drogenase ꢀLDH) measures the ability of test substances to alter
for activity associated with reduction of inflammation [1]. The neutrophil secretory activity and cellular integrity respectively
present paper describes the identification of active compounds and utilises rat peritoneal leukocytes challenged with the cal-
from the aqueous extract and describes the activity, cytotoxicity cium ionophore A23187 [2].
and concentration of the isolated compounds using a range of in
Affiliation
1
Department of Pharmacy, King's College London, London, U.K.
Current address: Pharmacology Department, Faculty of Medicine, Khon Kaen University, Khon Kaen,
2
Thailand
3
Messengers and Signalling Research Group, GKT School of Biomedical Sciences, King's College London,
London, U.K.
Deceased 24 April 2001
4
th
Correspondence
Prof. Peter J. Houghton ´ Department of Pharmacy ´ King's College London ´ Franklin-Wilkins Building ´
150 Stamford Street ´ London, SE1 9NN ´ U.K. ´ Fax: +44-20-7848-4800 ´ E-mail: peter.houghton@kcl.ac.uk
Funding
Our grateful thanks to Faculty of Medicine, Khon Khaen University, Thailand for a PhD scholarship for
Dr. Laupattarakasem.
Received November 11, 2003 ´ Accepted February 15, 2004
Bibliography
Planta Med 2004; 70: 496±501 ´ ꢁ Georg Thieme Verlag KG Stuttgart ´ New York
DOI 10.1055/s-2004-827147
ISSN 0032-0943

Reductions in the amounts of prostaglandins and leukotrienes tion gave IC50 values of > 500 mg/mL, 255 mg/mL, 120 mg/mL and
can also be used as an indicator of inflammatory activity [3], [4]. < 62.5 mg/mL for A to D, respectively, so fraction D was chosen for
A reduction in production of ROS and tissue-damaging enzymes further fractionation and was subjected to preparative TLC ꢀsilica
such as MPO from neutrophils is also an indication of anti-in- gel PF254, EtOAc:MeOH:water, 77:13:10) using UV 254 nm for
flammatory activity. Inflammation may be reduced by anti-oxi- detection. Two bands D1 ꢀhR : 34±38) and D2 ꢀhR : 48±65) were
f
f
dants since tissue damage is caused by excess ROS [5].
removed and eluted with MeOH to yield D1 ꢀ1334 mg) and D2
367 mg). D1 was purified using CC ꢀsilica gel, CHCl :MeOH:
ꢀ
3
water gradient) followed by preparative TLC ꢀsilica gel PF254
/
Materials and Methods
EtOAc:MeOH:water, 77:13:10) to yield 1 ꢀ375mg) hRf: 35. D2
was chromatographed by CC ꢀsilica gel, hexane: CHCl :MeOH,
3
Plant material and preparation of extract
2:1:1) monitored by TLC ꢀsilica gel GF254, n-BuOH:HOAc:water,
D. scandens stem was collected in Songkla province in southern 4:1:1, using UV 254 nm and 1% methanolic diphenylboric acid
Thailand and identified by Dr. Arunporn Itharat. A voucher speci- b-ethylamino ester followed by 5% ethanolic polyethylene gly-
men ꢀSKP A 1410101) is deposited in the Department of Pharma- col-4000 and viewing under UV 365 nm for detection) and like
ceutical Botany and Pharmacognosy, Faculty of Pharmaceutical fractions were combined to give fractions D2a ꢀtR of D2a be-
Science, Prince of Songkla University, Hat Yai, Thailand.
tween 50 and 65 mL) and D2b ꢀtR of D2b between 90and 115
mL) each showing only one major zone. Both D2a and D2b were
The materials were dried in the shade and the dried material purified by passage through Sephadex LH-20ꢂ to yield two pure
100 g) was macerated in boiling water for 30 min, centrifuged compounds, recrystallised from chloroform, 3 ꢀ8.5 mg) and 4 ꢀ5.6
at 800 g for 10 min and the supernatant was freeze-dried. The mg) with hR 74 and 72, respectively, on silica gel, CHCl :MeOH,
ꢀ
f
3
process was repeated twice to yield 8.96 g of freeze-dried extract. 9:1. The identity of the three compounds was determined by
spectroscopic means and, in the case of 1 ꢀ250 mg), by hydrolysis
Chemicals
using 2 M HCl at 1008C for 60 min and extraction into EtOAc, to
3
HBSS was obtained from Gibco Ltd. H-TXB was obtained from yield 2 which was purified using preparative TLC ꢀsilica gel,
2
Amersham International, LTB4 from Cascade Biochemicals Ld EtOAc:MeOH:HOAc:water, 65:15:20:15) to afford pure 2; yield:
3
and H-LTB from Du Pont ꢀUK) Ltd. Polyclonal rabbit anti-TXB
102 mg.
4
2
and anti-LTB antibodies were a kind gift from Dr. R. Forder, Xe-
4
neca Pharmaceuticals and Liquiscint scintillation fluid was ob- Compound 1 was isolated as a white amorphous powder, m.p.
tained from National Diagnostics. All other chemicals used were 216 ± 219 8C; UV ꢀMeOH): lmax = 262, 328 sh; + NaOMe: 270; +
obtained from Sigma.
AlCl :271; + AlCl /HCl: 271; + NaOAc: 262; + NaOAc/H BO :
3
3
3
3
+
262. Electrospray-MS: m/z ꢀrel. int. %) = 579 [M±H] ꢀ20);
Animals
C H O ꢀmeasured: 579.1922 calcd.: 579.1713), 433 ꢀ100), 271
27
31 14
1
Leukocytes were obtained from male or female Wistar rats under ꢀ10); H-NMR ꢀDMSO-d /TMS, signals due to most sugar protons
6
497
a protocol approved by the Home Office [1]. Rats were kept un- not given): d = 12.88 ꢀ1H, broad s, 5-OH), 9.65 ꢀ1H, broad s, 4¢-
der conditions approved by the Home Office licence and fed wa- OH), 8.40 ꢀ1H, s, H-2), 7.40 ꢀ2H, d, J = 8.5 Hz, H-2¢, H-6¢), 6.83
ter and food ad libitum before being sacrificed.
ꢀ2H, d, J = 8.5 Hz, H-3¢, H-5¢), 6.73 ꢀ1H, d, J = 2.1 Hz, H-8), 6.44
1H, d, J = 2.1 Hz, H-6), 5.18 ꢀ1H, d, J = 6.9 Hz, H-1¢¢), 4.62 ꢀ1H,
s, H-1¢¢¢), 3.91 ꢀ1H, dd, J = 14.2, 3.2 Hz, H-6¢¢axial), 3.47 ꢀ1H, dd,
ꢀ
Spectroscopic methods
UV spectra were acquired on an AVIV 17DS spectrometer using J = 14.2, 3.2, H-6¢¢equatorial), 2.28 ꢀ3H, d, J = 4.2 Hz, CH3±6¢¢¢);
MeOH as solvent. 1H- and C-NMR spectra were recorded at
00/100 MHz on a Bruker AMX 400 NMR spectrometer ꢀppm, J 5),157.9 ꢀC-8a), 157.6 ꢀC-4¢),155.11 ꢀC-2),130.5 ꢀC-6¢),122.8 ꢀC-3),
in Hz, using TMS as internal standard). DEPT experiments were 121.4 ꢀC-1¢), 115.5 ꢀC-3¢), 106.5 ꢀC-4a), 101.0 ꢀC-1¢¢), 100.3 ꢀC-6),
13
13
C-NMR ꢀDMSO-d /TMS): d = 180.9 ꢀC-4), 163.2 ꢀC-7), 161.9 ꢀC-
6
4
o
o
carried out with the polarization pulse t = 458, 90 and 135 . 100.1 ꢀC-1¢¢¢), 95.0 ꢀC-8), 79.6 ꢀC-3¢¢), 79.3 ꢀC-3¢¢¢), 76.9 ꢀC-2¢¢),
Standard programs from the library 3JCH = 10 Hz was used for 76.0 ꢀC-5¢¢), 73.4 ꢀC-4¢¢¢), 72.5 ꢀC-2¢¢¢), 71.1 ꢀC-3¢¢¢), 70.6 ꢀC-4¢¢),
the COLOC experiment. FAB-MS and EI-MS data were recorded 70.3 ꢀC-5¢¢¢), 68.7, 66.8 ꢀC-6¢¢), 18.23 ꢀC-6¢¢¢).
on high-resolution spectrometer ꢀKRATOS MS890MS and JEOL
JMS-AX505W spectrometers).
Enzyme release from rat peritoneal leukocytes
A suspension of leukocytes containing approximately 85% poly-
morphonuclear leukocytes ꢀPMNs) and 15% mononuclear cells
Isolation and identification of compounds
8
.5 g of the freeze-dried extract were fractionated using VLC ꢀLi- was prepared as detailed previously [6].
Chroprep RP18, Merck; 2”10 cm) and eluting with a water:me-
thanol gradient of 10:0; 9:1; 6;1; 3:1; 1:1 and 0:10 in 100 mL The assay of b-glucuronidase activity followed the method of Fri-
aliquots. Fractions were monitored by TLC [silica gel GF254/n- dovich [7] which measures by fluorimetry the release of 4-me-
BuOH:HOAc:water ꢀ4:1:1) using UV 254 nm and Natural Prod- thylumbelliferone from its b- -glucuronide as a model system.
D
uct Reagent for detection]. Fractions were combined and taken to Total cellular b-glucuronidase was measured using cells disrup-
dryness under reduced pressure at 408C followed by freeze-dry- ted by treatment with 20% Triton X-100 so that the amounts of
ing to yield fractions A ꢀ5370 mg ) t of A between 50 and 200 mL, enzyme released could be expressed as a percentage of the total.
R
B ꢀ511 mg) t of B between 250 and 375 mL, C ꢀ276 mg) t of C
R
R
between 500 and 650 mL, and D ꢀ1701 mg) t of D between 800 Lysozyme release was assessed by taking 50 mL of leukocyte su-
R
and 1000 mL. The results from testing for cycloxygenase inhibi- pernatants and mixing with 200 mL of vigorously shaken 0.3 mg/
Laupattarakasem P et al. Anti-inflammatory isoflavonoids from¼ Planta Med 2004; 70: 496±501

mL Micrococcus lysodeikticus suspension. The mixture was incu- The test for LDH release also used rat peritoneal PMNs and fol-
bated at 378C for 30 min and the reaction was terminated by ad- lowed an established method [2].
dition of 125 mL of ethanol. 200 mL were transferred to a micro-
plate and the OD450nm values were read using a microplate read- Determination of compounds in plant extract by HPLC
er. The total lysosyme content was measured by performing the The method was adopted from one used and validated previous-
same test on supernatants from neutrophils subjected to two cy- ly to determine similar compounds [11]. The aqueous extract was
cles of freeze-thawing. The results for activity of released lyso- dissolved in 50% MeOH to give two solutions at 50 and 250 mg/
zyme were expressed as a percentage of the total lysozyme.
mL. The three compounds 1, 2 and 4 were dissolved separately
in MeOH ꢀ50 mg/mL) and diluted to give solutions of 4.17, 8.33
The determination of MPO was carried out in 96-well micro- and 16.67 mg/mL. The solutions were examined by HPLC: col-
plates, each determination being carried out in triplicate. To umn: Lichrosorb RP-18 ꢀMerck) ꢀ250”4.6 mm I.D., 5 mm); mobile
each well were added 50 mL of supernatant from the peritoneal phase: linear gradient MeOH:0.5% HOAc, 1:4 to 4:1, over 20
leukocytes, 50 mL aqueous o-dianisidine ꢀ0.67 mg/mL) and 50 mL min; flow rate: 0.8 mL/min; detection: 266 nm ꢀWaters 440 de-
0
.5% hexadecyltrimethylammonium bromide in 50 mM phos- tector); injection: 20 mL. Calibration curves were drawn for the
phate buffer pH 6.0. 50 mL 0.003% hydrogen peroxide solution three compounds so that their concentrations in the extract
were added to commence the reaction and the absorbance at could be determined.
4
50 nm was read using a Dynatech Plate Reader at 1 min and at
timed intervals thereafter at room temperature with intermit- Statistical analysis
tent shaking. The value for total MPO release was obtained by All values are expressed as mean  standard deviation ꢀSD). Data
using cells disrupted by two cycles of freeze-thawing.
were analysed using one-way ANOVA and P < 0.05 was regarded
as significant. Data obtained were further analysed by the Dun-
nett's t-test for significance.
In vitro test for eicosanoid synthesis inhibition
Several methods of measuring levels of these compounds are
used but the use of a radioimmunoassay has been accepted by
many workers and the method used was identical to that used
for assessing the activity of the crude extracts of the plant [1],
[8]. Test extracts and compounds were used to give final concen-
trations in the range 10±500 mg/mL for the extract and fractions
and 31.5 to 250 mg/mL for 1 and 2 and 1 to 50 mg/mL for 3 and 4.
All determinations were carried out in triplicate and indometha-
cin 5 mM was used as positive control for COX inhibition and the
methoxyalkylthiazole ZM-211965 5 mM for 5-lipoxygenase inhi-
bition.
4
98
In vitro anti-oxidant effects of compounds
Two systems were used in this study, one based on the effect of
superoxide radicals on ferric cytochrome c [9] and the other as-
sessing the amount of malondialdehyde ꢀMDA) formed from oxi-
dation of lipids [8].
The effect of extracts on the hypoxanthine/xanthine oxidase sys-
tem was carried out in microwells using ferricytochrome c as a
reducing agent following an established method [8]. The thiobar-
bituric acid ꢀTBA) test for amount of MDA formed also followed
established procedures [8]. Propyl gallate ꢀPG) and the coumarin
fraxetin ꢀCou) were used as positive controls whilst negative
controls were tubes omitting the iron chloride-ascorbate mix, re-
agent blanks and samples without test substances.
Tests for cytotoxicity
Since it is important to assess the toxicity as well as the activity
of compounds present in traditional remedies, the substances
isolated were investigated for their effect on rat peritoneal leu-
kocyte viability using the MTT assay [10] and on the integrity of
the cell walls of the leukocytes by measuring lactate dehydro-
genase ꢀLDH) released [2].
The MTT test was carried out as previously described [8]. The ex-
periment was done in triplicate and Triton-X and melittin were
used as positive controls.
Laupattarakasem P et al. Anti-inflammatory isoflavonoids from¼ Planta Med 2004; 70: 496±501

Results and Discussion
100 g dried plant material which were found to be 1440 mg,
5 mg and 7 mg for 1, 2 and 4, respectively.
2
The structures of the compounds isolated from A4 were deter-
mined as follows by examination of their spectra and compari- The isolated compounds inhibited the synthesis of both throm-
son of spectral and chromatographic data with published values. boxane and leukotriene in a dose-dependent manner ꢀFig.1).
When 1 was hydrolysed the product 2 was identified as genistein The IC50 values were calculated from the data and were found to
since it gave spectroscopic and chromatographic characteristics be 1500 mM, 100 mM, 3 mM and 8 mM for COX inhibition and
1
13
identical to a commercial sample of genistein. The H- and C- 2500 mM, 80 mM, 6 mM and 1.6 mM for 5-LOX inhibition for 1, 2, 3
NMR spectra of 1 showed signals very similar to those given for and 4, respectively. The IC50 values of the positive controls were
genistein, but showed additional signals in the areas where sig- 0.2 mM and 62 mM for indomethacin and 1.8 mM and 0.03 mM for
nals for sugar molecules would be expected, 1 was therefore con- ZM211965 against COX and 5-LOX, respectively. It can be seen
sidered to be a glycoside of genistein. The lack of shift with that the two prenylated compounds 3 and 4 are much more ac-
NaOAc for 1 indicates that the sugars are probably attached at tive than 2 or its glycoside 1. However, when the concentrations
C-7, a fact confirmed by the downfield shift of H-6 and H-8 in of these compounds present in the extract were measured, it was
1
the H-NMR spectrum and the signal at 5.18 ppm for the anome- found that 1 was much higher ꢀ30.2 mg/mL) than 2 ꢀ0.51 mg/mL)
ric proton H-1¢¢ of the glucose [12]. TLC ꢀsilica gel, n-BuOH:- or 4 ꢀ0.15 mg/mL) thus showing that the greatest contribution to
HOAc:water, 4:1:5, upper layer; detection with aniline phthalate the activity comes from 1 and 2. Genistein has been shown to in-
0
.5% in water followed by heating at 1058C for 10 min) of the hibit prostaglandin E so it accounts for some of the activity of D.
2
aqueous layer of the acid hydrolysis mixture showed that glu- scandens [17]. The concentrations of the compounds in an extract
cose and rhamnose were present. The mass spectrum also sup- with a concentration equivalent to the IC50 are much less than
ported the presence of glucose and rhamnose in 1 by fragments would be predicted from their individual IC50 values. Compound
at m/z = 433, characteristic of a loss of rhamnose, and at m/ 2, the aglycone of 1, is more active than 1, and hydrolysis of 1 may
z = 271 which would indicate the subsequent loss of glucose, take place in vivo to release 2 and so increase the potency of the
thus indicating that the rhamnose was the terminal sugar. The aqueous extract.
signals given for the sugar carbons in the 13C-NMR spectrum
were typical of a rutinose [a-rhamnosylꢀ1®6)-b-glucose] moiety Data in Table 1 demonstrate that 1 and 2 showed some inhibition
[13] so 1 was therefore considered to be 7-O-a-rhamnoꢀ1®6)-b- of granular release, with 2 exerting a greater activity, a finding
glucosylgenistein. This appears to have not been previously re- which agrees with a previous report [18]. Compound 4 also
ported. Compound 3 was identified as 5,7,4¢-trihydroxy-6,5¢-di- showed appreciable inhibition of MPO release. To clarify whether
prenylisoflavone [14], which had previously been reported from the compounds exert this effect by interfering with the release of
D. scandens[15]. Compound 4 was identified as scandenin, also enzymes or by inhibiting the enzymes directly, tests on MPO pre-
previously isolated from D. scandens [16].
pared from rat PMNs were carried out with a range of concentra-
tions of 2 and 4. A dose-related response was obtained and IC50
499
Compounds 2 and 4 were present in detectable amounts only at values were calculated to be 0.22 mM and 0.14 mM, respectively,
the higher concentration of extract injected, although 3 could be thus demonstrating inhibitory effects.
detected at lower concentrations. The calibration curves obtain-
ed for the three compounds 1 ꢀRt = 9.3 min), 2 ꢀRt = 11.47 min) Fig. 2 shows that no anti-oxidant activity was expressed by 1 and
and 4 ꢀRt = 15.13 min) all gave a good correlation value 2 while 3 and 4 did have some effect, although at doses much
ꢀ
r = 0.999) and were used to determine the amounts present in higher than those which would be found in an extract of the
Fig. 1 Effect of compounds from D. scandens on eicosanoid
production.
Laupattarakasem P et al. Anti-inflammatory isoflavonoids from¼ Planta Med 2004; 70: 496±501

Table 1 Effect of compounds from D. scandens aqueous extract on en- Table 2 Effect on brain liposomal peroxidation by compounds from
zyme degranulation from rat PMNs D. scandens
Substance
Concentration % Total enzyme ꢀmean  s.e.m.) released
Compound
Concentration
mg/mL
Optical density % of DMSO control
ꢀ s.e.m.)
b-Glucuroni- Lysozymes
Myeloperoxi-
dase ꢀMPO)
dase
Liposomes alone
DMSO
9.6**
100
Blank control
TritonX-100
Cells alone
A23187
3.9  0.1
100.5  0.3
7.6  0.2
14.2  2.2
98.7  1.7
13.9  1.6
64.8  4.1
69.2  8.0
62.9  2.7
51.9  5.4
4.8  0.04
100.0  0.5
11.7  1.7
70.7  2.7
79.2  7.6
66.2  7.6
84.4  1.8
Propyl gallate
Fraxetin
7.5 mM
20.6  0.2**
17.7  0.1**
100.0  0.2
100.2  0.1
100.1  0.2
103.3  0.4
101.9  0.2
86.6  0.4
77.6  0.5
52.8  0.3**
32.3  0.2**
26.7  0.5**
98.3  0.1
83.8  0.2
38.6  0.1**
19.3  0.3**
84.6  0.1
61.4  0.2**
18.1  0.1**
16.9  0.1**
20%
10 mM
1
15.6
1 mM
5mM
28.1  0.7
29.2  1.2
3
6
1.25
2.5
Indomethacin
1
31.25 mg/mL 32.6  2.4
1
2
25
50
6
2.5 mg/mL
31.0  0.3
29.1  1.7
32.8  4.9
1
25 mg/mL
50 mg/mL
47.2  2.9*73.5  5.6
34.7  11.6*58.7  7.3
2
15.6
2
3
6
1.25
2.5
2
31.25 mg/mL 25.2  2.5
31.0  5.8**
33.8  8.1**
6
2.5 mg/mL
20.6  2.5* 10.3  1.9**
19.3  1.6** 8.9  1.9**
19.6  1.7** 4.1  0.8**
20.4  3.1**
14.1  1.1**
6.1  0.1**
86.5  5.3
1
2
25
50
1
1
25 mg/mL
50 mg/mL
2
3
4
3
4
15.62 mg/mL 34.2  0.8
94.7  1.6
77.9  2.2
84.4  2.1
86.6  1.1
5
1
25 mg/mL
50 mg/mL
35.8  1.0
35.7  0.8
62.0  8.0
2
5
0
2
65.6  4.8
5
15.62 mg/mL 35.2  1.4
40.7  6.0**
1
5
1
25 mg/mL
50 mg/mL
34.3  1.1
32.1  1.5
42.2  12.2*24.7  3* .5*
11.2  4.6** 6.2  0.6**
2
2
5
n = 3; *p < 0.05, ** p < 0.001 difference between release from cells treated with A23187
and those treated with compound.
50
Significance with Dunnett's t-test between release from cells treated with A23187 and those
treated with compound.
n = 3; ** P < 0.001 in comparison with cells treated with DMSO alone.
Significance with Dunnett's t-test in comparison with cells treated with DMSO alone.
Fig. 2 Effect of compounds on superoxide generation by
rat PMN treated with 1mM phorbol myristate acetate
ꢀPMA). ** p < 0.001 significant inhibition compared with
5
00
value given by PMA; + = significance with Dunnett's t-test.
plant. A similar pattern, with 3 and 4 having the strongest pro- Only two concentrations of compounds ꢀ50 and 125 mg/mL) were
tective effect ꢀIC50 0.06 mM and 0.026 mM, respectively) was seen tested in the MTT assay. Compounds 1 and 2 actually seemed to
with the TBA test ꢀTable 2). Compound 2 showed some activity stimulate cell growth and division at the higher concentrations
with an IC50 of 0.3 mM, an activity previously ascribed to the fact ꢀ128% of control) whilst 3 and 4, gave a small reduction in appar-
that the tyrosine kinase inhibitory activity of genistein would de- ent cell viability ꢀ72% and 68% of control, respectively) at the
crease superoxide production [19]. Although 3 and 4 are more higher dose, which was not considered significant. The apparent
active, the much greater concentrations of 1, and therefore its ac- increase given by 1 and 2 may be due to a direct reduction effect
tive hydrolysis product 2, are likely to mean that it contributes by the compounds on MTT, a phenomenon which has been noted
most to any anti-oxidant activity of the extract.
recently for some isoflavones [20]. An increase in LDH release is
another indicator of cytotoxicity and Fig. 3 shows that 3 and 4 did
Laupattarakasem P et al. Anti-inflammatory isoflavonoids from¼ Planta Med 2004; 70: 496±501

Fig. 3 Effect of compounds on release of lactate dehydro-
genase ꢀLDH) from rat PMN.
4
cause leakage at higher concentrations, but at levels unlikely to
occur in extracts.
Bauer R. Cycloxygenase and 5-lipoxygenase as targets for medicinal
plant research. in: Bohlin L, Bruhn JG ꢀEds.). Bioassay Methods in Nat-
ural Products Research and Drug Development Kluwer Academic, Dor-
drecht, Netherlands: 1999: pp 119±42
Cuzzocrea S, Riley DP, Caputi AP, Salvemini D. Antioxidant therapy: A
new pharmacological approach in shock, inflammation and ischemia/
reperfusion injury. Pharmacology Reviews 2001; 53: 135±9
Hoult JRS, Pang L-H, Bland-Ward P, Forder R, Williams C, Harborne JB.
Inhibition of leukocyte 5-lipoxygenase and cyclooxygenase but not
constitutive nitric oxide synthase by tanetin, a novel flavonoid derived
from feverfew, Tanacetum parthenium. Pharmaceutical Sciences 1995;
The anti-inflammatory effect of D. scandens stem aqueous extract,
noted previously by in vivo tests [1], appears to be due, at least in
part, to the isoflavone content of the extract. The overall effect ap-
pears to be due to a mixture of effects with inhibition of eicosa-
noid synthesis predominating over anti-oxidant activity and inhi-
bition of release of granular enzymes such as MPO. The most ac-
tive compounds were the two prenylated isoflavones and their
ability to inhibit the two major enzymes involved in eicosanoid
synthesis is of interest and could make them candidates for lead
molecules for novel anti-inflammatory compounds. Although 3
and 4 are much more potent than 1 and 2, it is likely that these lat-
ter two make the major contribution to the observed effects since
they are present in much greater concentrations in the extract. In
addition, the glycoside may act as a reservoir of the more potent
aglycone as it is likely that hydrolysis may occur in vivo. Although
the results reported here indicate that none of the compounds has
appreciable toxicity at the concentrations in which they occur in
the extract, it would be unwise to deduce that no toxicity would
occur when the extract is ingested.
5
6
1
: 71±4
7
8
9
Fridovitch I. Superoxide dismutase: an adaptation to a paramagnetic
gas. Journal of Biological Chemistry 1989; 264: 7761±4
Aruoma OI. Assessment of potential prooxidant and antioxidant ac-
tions. Journal of the American Oil Chemists Society 1996; 73: 1617±25
Paya M, Halliwell B, Hoult JRS. Interaction of a series of coumarins
with reactive oxygen species. Biochemical Pharmacology 1992; 44:
2
05±14
10
Mosmann T. Rapid colourimetric assay for cellular growth and survi-
val: application to proliferation and cytotoxicity assays. Journal of Im-
munological Methods 1983; 65: 55±63
van de Casteele K, Geiger H, van Sumere C. Separation of flavonoids by
high-performance liquid chromatography. Journal of Chromatography
501
11
1982; 240: 81±94
12
Mabry T, Markham K, Thomas M. The systematic identification of fla-
vonoids. New York: Springer, 1970: p 55
Slimestad R, Andersen éM, Francis GW, Marston A, Hostettmann K.
Syringetin 3-O-ꢀ6¢¢-acetyl)-b-glucopyranoside and other flavonols
from needles of Norway spruce, Picea abies. Phytochemistry 1995;
40: 1537±42
Singhal A, Sharma R, Thyagarajan G, Herz W, Govindan S. New preny-
lated isoflavones and a prenylated dihydroflavonol from Millettia
pachycarpa. Phytochemistry 1980; 19: 929±34
Rao M, Krupadanam G, Srimannarayana G. Four isoflavones and two
3-arylcoumarins from stems of Derris scandens. Phytochemistry
13
Acknowledgements
14
Jane Hawkes and Andy Cakebread of King's College London are
thanked for their help in running NMR and mass spectra, respec-
tively. Dr Arunporn Itharat of Pharmacy Faculty, Prince of Songkla
University, Hat Yai, Thailand is thanked for supply of plant material.
15
1994; 37: 267±9
16
Johnson A, Pelter A, Stainton P. The extractives from Derris scandens.
Part I. The structures of scandenin and lonchocarpic acid. Journal of
the Chemical Society C, 1969: 192±202
Kniss D, Zimmerman P, Su H, Fertal R. Genistein suppresses EGF-in-
duced prostaglandin synthesis by a mechanism independent of EGF
receptor tyrosine kinase inhibition. Prostaglandins 1996; 57: 87±105
Ozaki Y, Yatomi Y, Jinnai Y, Kume S. Effects of genistein, a tyrosine ki-
nase inhibitor, on platelet functions. Biochemical Pharmacology 1993;
46: 395±403
Kusunoki T, Higashi H, Hosoi S, Hata D, Sugie K, Mayumi M, Mikawa H.
Tyrosine phosphorylation and its possible role in superoxide produc-
tion by human neutrophils stimulated with fMLP and IgG. Biochem-
ical and Biophysics Research Communications 1992; 183: 789±96
Bruggiser R, van Daeniken K, Jundt G, Schaffner W, Tullberg-Reinert H.
Interference of plant extracts, phytoestrogens and antioxidants with
the MTT tetrazolium assay. Planta Medica 2002; 68: 445±8
17
References
1
Laupattarakasem P, Houghton PJ, Hoult JRS, Itharat A. An evaluation of
18
the activity related to inflammation of four plants used in Thailand to
treat arthritis. Journal of Ethnopharmacology 2003; 85: 207±15
De las Heras B, Hoult JRS. Non-cytotoxic inhibition of macrophage ei-
2
19
cosanoid biosynthesis and effects on leukocyte functions and reactive
oxygen species of two novel anti-inflammatory plant diterpenoids.
Planta Medica 1994; 60: 501±6
3
Scott KF, Bryant KJ, Bidgood MJ. Functional coupling and differential
20
regulation of the phospholipase A -cycloxygenase pathways in in-
2
flammation. Journal of Leukocyte Biology 1999; 66: 535±41
Laupattarakasem P et al. Anti-inflammatory isoflavonoids from¼ Planta Med 2004; 70: 496±501