Flavonoid Constituents of Chorizanthe diffusa with Potential Cancer Chemopreventive Activity

Article abstract of DOI:10.1021/jf980784o

An ethyl acetate-soluble extract of Chorizanthe diffusa was found to exhibit significant antioxidant activity, as judged by scavenging stable 1,1-diphenyl-2-picrylhydrazyl (DPPH) free radicals and inhibition of 12-O-tetradecanoylphorbol 13-acetate (TPA)-i

Full text of DOI:10.1021/jf980784o

36
J. Agric. Food Chem. 1999, 47, 36−41
F la von oid Con stitu en ts of Ch or iza n th e d iffu sa w ith P oten tia l
Ca n cer Ch em op r even tive Activity
Ha Sook Chung,†,‡ Leng Chee Chang,† Sang Kook Lee,† Lisa A. Shamon,†,§
Richard B. van Breemen,† Rajendra G. Mehta, Norman R. Farnsworth,† J ohn M. Pezzuto,†, and
A. Douglas Kinghorn*,†
Program for Collaborative Research in the Pharmaceutical Sciences and Department of Medicinal
Chemistry and Pharmacognosy, College of Pharmacy, and Department of Surgical Oncology, College of
Medicine, University of Illinois at Chicago, Chicago, Illinois 60612
An ethyl acetate-soluble extract of Chorizanthe diffusa was found to exhibit significant antioxidant
activity, as judged by scavenging stable 1,1-diphenyl-2-picrylhydrazyl (DPPH) free radicals and
inhibition of 12-O-tetradecanoylphorbol 13-acetate (TPA)-induced free radical formation with cultured
HL-60 cells. Bioassay-directed fractionation of this extract using the DPPH antioxidant assay as a
monitor led to the isolation of five structurally related flavonoids (1-5), including the novel compound
5,8,3′,4′,5′-pentahydroxy-3,7-dimethoxyflavone (1). Isolates 1-5 demonstrated varying degrees of
antioxidant or antimutagenic activity. Two of the compounds, 5,7,3′,4′-tetrahydroxy-3-methoxyflavone
(2) and quercetin (4), were subsequently found to inhibit carcinogen-induced preneoplastic lesions
in a mouse mammary organ culture model. Inhibitory activity of this type is known to correlate
with cancer chemopreventive effects in full-term models of tumorigenesis.
Keyw or d s: Chorizanthe diffusa; Polygonaceae; flavonoids; 5,8,3′,4′,5′-pentahydroxy-3,7-dimethoxy-
flavone; antioxidant activity; antimutagenic activity; mouse mammary organ culture; cancer
chemoprevention
INTRODUCTION
with the activated (electrophilic) forms of carcinogens
and oxygen free radicals. Free radicals can be detri-
mental by reacting with, and sometimes by destroying,
critical cellular components including the polyunsatu-
rated fatty acids (PUFA) that comprise lipoprotein
particles and plasma membranes. Damage by free
radicals can result in loss of membrane fluidity, receptor
misalignment, cell lysis, damage to sulfur-containing
enzymes and other proteins (producing inactivation,
cross-linking, and denaturation), and damage to carbo-
hydrates (leading to altered cellular receptor functions,
including those associated with cytokine activities and
prostaglandin formation). Free radicals can also induce
brain disorders, atherosclerosis, and colon cancer (In-
serra et al., 1997). Several cancer chemopreventive
agents exhibit antioxidant activity through their ability
to scavenge oxygen radicals, inclusive of singlet oxygen,
peroxy radicals, superoxide anion, and hydroxyl radicals
(Pryor, 1988). Antioxidant defenses normally protect
against DNA damage caused by reactive oxygen species
from endogenous and exogenous sources, and reduced
levels of antioxidants are associated with increased
cancer risk (Carmia, 1997).
Chemoprevention of cancer involves the use of dietary
or pharmacological agents to enhance intrinsic physio-
logical mechanisms that protect the organism against
the development and progression of transformed cells
(Sporn et al., 1976; Wattenberg, 1985). Chemopreven-
tive agents can block initiation of the carcinogenic
process, or arrest or reverse further progression of
premalignant cells before they become invasive or
metastatic, during the extensive latency period that is
characteristic of most human cancers (Sporn, 1993).
Compounds of this type are classified as either inhibi-
tors of carcinogen formation, blocking agents, or sup-
pressing agents. Most blocking agents are inhibitors of
tumor initiation and can be assigned to one or more of
five major categories: inhibitors of cytochrome P450
enzymes; inducers of cytochrome P450; inducers of
phase II enzymes such as glutathione S-transferase
(GST), uridine diphosphate (UDP)-glucuronyltrans-
ferase, and glutathione peroxidase; scavengers of elec-
trophiles and free radicals; and inducers of DNA repair
(Hong and Sporn, 1997; Stoner et al., 1997). Scavenging
or trapping agents are compounds that physically react
There are 50 species in the genus of Chorizanthe
which occur in arid regions in the western part of North
America (Mabberley, 1987). Previous work on a 50%
ethanol extract of Chorizanthe glabrescens Benth. showed
weak antimicrobial activity against Sarcina lutea and
Staphylococcus aureus and cytotoxic activity with cul-
tured KB cells (ED50 < 20.0 µg/mL) (Azarowicz et al.,
1952; Bhakuni et al., 1974, 1976). However, there has
been no detailed phytochemical work to date on any
species in this genus, inclusive of C. glabrescens.
* Author to whom correspondence should be addressed
[telephone, (312) 996-0914; fax, (312) 996-7107; e-mail,
kinghorn@uic.edu].
† College of Pharmacy.
‡ Present address: Department of Food and Nutrition,
College of Natural Science, Duksung Women’s University,
Seoul 132-174, Korea.
§ Present address: Geraldine Brush Cancer Research In-
stitute, California Pacific Medical Center, 2330 Clay St., San
Francisco, CA 94115.
College of Medicine.
In the current investigation, Chorizanthe diffusa
10.1021/jf980784o CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/16/1998

Bioactive Flavonoids of C. diffusa
J. Agric. Food Chem., Vol. 47, No. 1, 1999 37
are maintained. The dried plant material was stored at
ambient temperature and milled just prior to the present
investigation.
Benth. (Polygonaceae) was selected for study, since the
ethyl acetate-soluble extract of the entire plant was
found to exhibit significant antioxidant activity based
on scavenging activity of stable 1,1-diphenyl-2-picryl-
hydrazyl (DPPH) free radicals (IC50: 12.8 µg/mL) and
inhibition of TPA-induced free radical formation with
cultured HL-60 cells (IC50: 22.0 µg/mL). In addition, a
secondary biological analysis demonstrated the extract
inhibited 7,12-dimethylbenz[a]anthracene (DMBA)-
induced preneoplastic lesion formation with a mouse
mammary organ culture model (Mehta and Moon, 1991)
(57% inhibition). Bioassay-guided fractionation of the
active extract using the DPPH antioxidant assay led to
the isolation of one novel (1) and four known (2-5)
flavonoids, which were then evaluated for their indi-
vidual biological activities. The initial screening biologi-
cal data leading to the selection of C. diffusa for activity-
monitored isolation have been published in a preliminary
communication (Lee et al., 1998). At the onset of the
present study, C. diffusa was considered a native species
to the United States and of potential importance for its
possible cancer chemopreventive effects.
An tioxid a n t P oten tia l of Test Com p ou n d s. 1. Assay for
DPPH Free Radical Scavenging Potential. This assay is based
on the scavenging activity of stable 1,1-diphenyl-2-picrylhy-
drazyl (DPPH) free radicals (Fujita et al., 1988; Lee et al.,
1998). Reaction mixtures containing test samples (dissolved
in DMSO) and 300 µM DPPH ethanolic solution in 96-well
microtiter plates were incubated at 37 °C for 30 min. Absor-
bances were then measured at 515 nm, and percent inhibition
by sample treatment was calculated. IC50 values denote the
concentration of sample required to scavenge 50% of DPPH
free radicals.
2. Assay for Inhibition of TPA-Induced Free Radical Forma-
tion in Cultured HL-60 Cells. This assay method is based on
inhibiting TPA-induced free radical formation in HL-60 cells
by measuring cytochrome c reduction (Sharma et al., 1994).
HL-60 cells were grown in RPMI 1640 containing 5% heat-
inactivated fetal calf serum and penicillin/streptomycin. Cells
were differentiated with culture medium containing 1.3%
DMSO for 7 days and then harvested and washed twice with
and resuspended in HBSS (Hank’s balanced salt solution)
(without phenol red; Cellgro, Mediatech, DMSO-HL60). Cells
were suspended in HBSS and added to 96-well plates (1 × 106
cells/well). Free radical formation was induced by addition of
TPA (8 µM). This was followed by the addition of 5 half-log
concentrations of test agents together with cytochrome c (160
µM). After an incubation period of 1 h at 37 °C, cytochrome c
reduction was measured at 515 nm using an ELISA reader.
Relative to incubations performed with TPA alone, the percent
inhibition of radical formation in TPA plus agent-treated
samples was determined. Blank reaction mixtures were pre-
pared in which cell suspensions were omitted, and these values
were subtracted from the treatment groups. Percent inhibition
by sample treatment was determined by comparison with a
solvent-treated control group. IC50 values were determined
from the concentration of sample required to reduce cyto-
chrome c oxidation by 50%.
An tim u ta gen icity Assa y w ith Sa lm on ella typh im u -
r iu m Str a in TM677. Test compounds were evaluated for
potential to inhibit forward mutation to 8-azaguanine resis-
tance with S. typhimurium strain TM677 (Gru¨ter et al., 1990;
Shamon et al., 1994; Ito et al., 1998). Reaction mixtures were
prepared consisting of 0.77 mL of bacteria in minimal essential
medium, 0.1 mL of S9 liver homogenate derived from Aroclor
1254-pretreated rats, and 0.11 mL of a NADPH-generating
system. To monitor antimutagenic activity, test substances
(100 µM final concentration in 10 µL of DMSO) were added to
duplicate incubation mixtures 1 min prior to the addition of
DMBA (80 µM final concentration in 10 µL of DMSO).
Following incubation in a rotating dry-air incubator (2 h, 37
°C), the reaction mixtures were quenched by the addition of 4
mL of phosphate-buffered saline. The bacteria were recovered
by centrifugation, resuspended in 5 mL of phosphate-buffered
saline, diluted, and plated on minimal agar in the presence
and absence of 8-azaguanine. The plates were then incubated
(48 h, 37 °C) and scored with an automatic colony counter
(Fisher model 600). Results were expressed as a mutant
fraction, and percent inhibition was calculated relative to
controls.
Mou se Ma m m a r y Gla n d Or ga n Cu ltu r e Assa y. BALB/c
female mice (4 weeks old; Charles River, Wilmington, MA)
were pretreated for 9 days with 1 µg of estradiol and 1 mg of
progesterone. On the tenth day, the mice were sacrificed and
the second thoracic mammary glands were dissected on silk
and transferred to 60-mm culture dishes containing 5 mL of
Waymouth’s 752/1 MB medium supplemented with 100 units
of streptomycin and penicillin and 35 µg/mL glutamine. The
glands were incubated for 10 days (37 °C, 95% O2 + 5% CO2)
in the presence of growth-promoting hormones (5 µg of insulin,
5 µg of prolactin, 1 µg of aldosterone, and 1 µg of hydrocortisone
per mL of medium). Glands were exposed to 2 µg/mL DMBA
between 72 and 96 h. After the exposure, glands were rinsed
and transferred to new dishes with fresh medium. The fully
MATERIALS AND METHODS
1H NMR, 1H-1H COSY, and 13C NMR spectra were mea-
sured on a Varian XL-300 instrument operating at 300 and
75.4 MHz, respectively. Compounds were analyzed in DMSO-
d6, with tetramethylsilane (TMS) as internal standard. 1H-
1
13C HETCOR and H-13C FLOCK NMR spectra were run with
standard pulse sequences. A General Electric Omega 500 NMR
spectrometer, operating at 499.9 MHz, was used to perform
1H-13C HMQC and 1H-13C HMBC NMR experiments. Posi-
tive-ion electron-impact (EIMS), high-resolution EIMS (HR-
EIMS), chemical-ionization (CIMS), and fast-atom-bombard-
ment (FABMS) mass spectra were measured on a Finnigan
MAT-90 (Bremen, Germany) double-focusing mass spectrom-
eter. The CI reagent gas was methane, and glycerol was used
as the FAB matrix. High-performance liquid chromatography-
electrospray mass spectrometry (HPLC-ESMS) was carried
out using a Hewlett-Packard (Palo Alto, CA) 5989B mass
spectrometer equipped with a 1050 HPLC system. HPLC
separation was carried out using a 60-min linear gradient from
12 to 90% methanol in water containing 0.1% formic acid on
a Vydac (Separations Group, Hesperia, CA) C18 reversed-phase
column, 5-µm packing, 150 × 2 mm, at a flow rate of 200 µL/
min. Elution of compounds was monitored using negative-ion
electrospray ionization and by scanning the mass range m/z
150-800 in 2 s. Negative-ion electrospray tandem mass
spectra (MS-MS) were obtained using a Micromass (Manches-
ter, U.K.) Quattro II mass spectrometer with argon collision
gas and a collison energy of 50 V. IR spectra were taken on a
Midac Collegian FT-IR spectrometer and UV spectra were
measured on a Beckman DU-7 spectrometer. Melting points
were determined using a Fisher-J ohns melting point apparatus
and are uncorrected, and optical rotations were obtained on a
Perkin-Elmer model 241 polarimeter.
Thin-layer chromatographic (TLC) analysis was performed
on Kieselgel 60 F254 (Merck, Darmstadt, Germany) plates
(silica gel, 0.25-mm layer thickness), with compounds visual-
ized by spraying with 10% (v/v) H2SO4 followed by heating at
110 °C on a hot plate. Silica gel (Merck 60 A, 230-400 mesh
ASTM), Sephadex LH-20 (25-100 µm; Pharmacia Fine Chemi-
cals, Piscataway, NJ ), and Sorbisil C18 reversed-phase silica
gel (Phase Separations, Ltd., Deeside, Clywd, U.K.) were used
for column chromatography. All solvents used for chromato-
graphic separations were purchased from Fisher Scientific
(Fair Lawn, NJ ) and distilled before use.
P la n t Ma ter ia l. Whole plants of Chorizanthe diffusa
Benth. were collected in California, in August 1978, under the
auspices of Dr. Robert E. Purdue, J r., Medicinal Plants
Laboratory, USDA, Beltsville, MD, where voucher specimens

38 J. Agric. Food Chem., Vol. 47, No. 1, 1999
Chung et al.
differentiated glands were then permitted to regress by
withdrawing all hormones except insulin for 14 additional
days. Test compounds were present in the medium during days
1-10 of culture (10 µg/mL); mammary glands were scored for
incidence of lesions (Mehta and Moon, 1991).
°C. After being diluted with water the reaction mixture was
extracted with CHCl3. The CHCl3 layer was washed with
water, dried with Na2SO4, evaporated, and applied to prepara-
tive TLC to obtain the pure methylated compound 1a (3 mg).
5-Hyd r oxy-3,7,8,3′,4′,5′-h exa m eth ylfla von e (1a ): yellow
needles; mp 167 °C; UV (MeOH) λmax (log ꢀ) 247 (4.22), 352
(3.31) nm; unchanged by shift reagents; 1H NMR 3.96 (6H, s,
OMe), 3.95 (6H, s, OMe), 3.93 (3H, s, OMe), 3.90 (3H, s, OMe),
6.94 (1H, s, H-6), 7.52 (2H, s, H-2′ and -6′), 12.38 (1H, s, OH-
5).
Extr a ction a n d Isola tion P r oced u r e. The milled, dried
whole plants of C. diffusa (5 kg) were extracted with MeOH
and then concentrated and partitioned sequentially with
petroleum ether, EtOAc, and n-BuOH, producing petroleum
ether (16.7 g), EtOAc (15.6 g), n-BuOH (23.5 g), and aqueous
methanolic extracts. The EtOAc and n-BuOH extracts exhib-
ited significant activity in the DPPH antioxidant assay (IC50
values of 33.4 and 39.9 µg/mL, respectively). The EtOAc-
soluble (15.5 g) extract was adsorbed onto silica gel and
separated over additional silica gel (750 g) by open column
silica gel chromatography, using a gradient of 3-12% MeOH
in CHCl3, and eluates containing constituents with similar
TLC profiles were combined to provide 12 pooled fractions
(fractions 4-15). Fractions 13 and 14 (eluted with CHCl3-
MeOH, 93:7 and 90:10, respectively) were active in the DPPH
antioxidant assay (IC50 values of 20.4 and 24.6 µg/mL, respec-
tively) and yielded a pure flavone (1, 18.0 mg), which was
recrystallized from MeOH. A pure flavonol (4, 14.7 mg) was
obtained by further separation over a small column containing
Sephadex LH-20, eluted with MeOH. Column chromatography
of the n-BuOH-soluble extract (23.4 g) on silica gel with a
gradient of 5-20% MeOH in CHCl3 yielded 17 subfractions
(fractions 26-42), with fractions 34 and 35 being moderately
active in the DPPH antioxidant assay (IC50 values of 75.4 and
81.8 µg/mL, respectively). Further chromatography of fraction
34 using Sephadex LH-20 size-exclusion chromatography
(MeOH) yielded two further pure flavones (2, 8.2 mg, and 3,
17 mg), eluted from silica gel with 95% MeOH in H2O. A
flavonol glycoside (5, 12 mg) was obtained from fraction 35
using C18 reversed-phase silica gel, eluted with 85-90% MeOH
in H2O.
RESULTS AND DISCUSSION
The dried and ground whole plant of C. diffusa was
extracted with methanol and partitioned between pe-
troleum ether and 90% methanol in water, with the
more polar layer then partitioned with EtOAc and
n-BuOH. The dried EtOAc and n-BuOH extracts were
subjected individually to a series of activity-guided
chromatographic fractionation steps to afford, in turn,
compounds 1 and 4 and compounds 2, 3, and 5. All of
these pure isolates gave characteristic flavonoid color
reactions (purplish brown with FeCl3, yellow with
NaOH, yellowish orange with Mg-HCl, pale pink with
1
Zn-HCl) (Markham, 1975). The UV, H NMR, and 13C
NMR spectral data of the aromatic parts of compounds
1-3 and 5 were characterized by major bands that
resembled those of 3-O-substituted flavones and fla-
vonols (Agrawal et al., 1989; Mabry et al., 1970). The
previously known compounds 2-5 were established as
5,7,3′,4′-tetrahydroxy-3-methoxyflavone, 5,8,3′,4′-tet-
rahydroxy-3,7-dimethoxyflavone, quercetin, and 3′′-O-
acetylquercitrin, respectively, by comparison with lit-
erature data (Roitman and J ames, 1985; Sakakibara
and Mabry, 1975; Wenkert and Gottlieb, 1977; Ulubelen
and Timmermann, 1980; Tanaka et al., 1978; Markham
et al., 1978).
5,8,3′,4′,5′-P en ta h yd r oxy-3,7-d im eth oxyfla von e (1): yel-
low-orange needles; mp 282 °C; positive reactions with the
FeCl3, Mg-HCl, and Zn-HCl tests; UV (MeOH) λmax (log ꢀ) 278
(4.53), 305 (3.07), 346 (3.35) nm; (+NaOMe) 278 (3.85), 372
(2.09) nm; (NaOAc) 275 (4.28), 303 (2.76), 348 (2.96) nm;
(NaOAc + H3BO3) 263 (4.47), 307 (2.12), 387 (3.57) nm; (AlCl3)
281 (4.84), 315 (2.05 sh), 464 (3.04) nm; (AlCl3 + HCl) 285
(4.10), 323 (2.84), 370 (3.19) nm; IR (MeOH) νmax 3270, 1657,
Compound 1 was assigned a molecular formula of
C17H14O9 from its HR-EIMS (m/z 362.0624), which was
in accord with a flavone containing five hydroxyl and
1
two methoxyl groups. Comparison of its UV, H NMR,
and 13C NMR data with literature values indicated that
this compound was a flavone (Agrawal et al., 1989). In
addition, it was apparent that 1 contained five OH
groups (δC 126.2, 137.1, 145.9 (double intensity), 152.8)
and two OMe groups (δC 56.1, 59.1; δH 3.81, 3.93). On
methylation of 1 using dimethyl sulfate under standard
conditions, the hexamethoxylated product 1a was pro-
duced. The relative positions of the functional groups
in 1 were established from the following observations.
1
1607, 1507, 1454 cm-1; H NMR (DMSO, 300 MHz) δ 12.27
(1H, s, OH-5), 7.27 (2H, s, H-2′ and -6′), 6.57 (1H, s, H-6), 3.93
(3H, s, OMe-7), 3.81 (3H, s, OMe-3); 13C NMR (DMSO, 75.6
Hz) δ 178.2 (s, C-4), 156.1 (s, C-2), 153.8 (s, C-7), 152.8 (s, C-5),
145.9 (s, C-3′ and -5′), 143.7 (s, C-9), 137.6 (s, C-3), 137.1 (s,
C-4′), 126.2 (s, C-8), 119.9 (s, C-1′), 107.9 (d, C-2′ and -6′), 104.5
(s, C-10), 95.3 (d, C-6), 59.1 (q, OMe-3), 56.1 (q, OMe-7); EIMS
m/z (rel int) [M]+ 362 (100), 319 (36), 320 (6), 181 (4); HR-
EIMS m/z found 362.0624, calcd for C17H14O9 362.0633.
5,7,3′,4′-Tetr a h yd r oxy-3-m eth oxyfla von e (2): pale-yel-
low powder; UV, 1H NMR, and 13C NMR data consistent with
published values (Roitman and J ames, 1985); EIMS m/z (rel
int) [M]+ 316 (2), 302 (100), 273 (9), 161 (10), 137 (10), 108 (7).
5,8,3′,4′-Tetr a h yd r oxy-3,7-d im eth oxyfla von e (3): yellow-
orange powder; UV, 1H NMR, and EIMS data consistent with
published values (Sakakibara and Mabry, 1975); EIMS m/z
(rel int) [M]+ 346 (42), 345 (16), 315 (45), 303 (70), 287 (11).
Qu er cetin (4): yellow powder; mp 302 °C [lit. mp 313-314
°C (Wenkert and Gottlieb, 1977)]; UV, 1H NMR, and 13C NMR
data consistent with published values (Wenkert and Gottlieb,
1977; Ulubelen and Timmermann, 1980); CIMS m/z (rel int)
[M + H]+ 303 (13), 286 (100), 229 (3), 133 (22), 154 (10).
3′′-O-Acetylqu er citr in (5): yellow powder; mp 172 °C [lit.
mp 167-175 °C (Nobutoshi et al., 1978)]; UV, 1H NMR
(Tanaka et al., 1978), and 13C NMR data consistent with
published values (Markham et al., 1978); FABMS m/z (rel int)
[M + H]+ 491 (52), 303 (100), 245 (33), 189 (41), 129 (31), 165
(15).
1
The H NMR spectrum of 1 suggested that the A ring
was trisubstituted from the appearance of a singlet at
δH 6.57, and an aromatic proton singlet integrating for
two protons at δH 7.27 was assigned for H-2′ and H-6′
in ring B, respectively. After sodium acetate was added
to compound 1, the resultant UV spectrum did not
exhibit any bathochromic shift, thus indicating the
absence of any free hydroxyl group at the C-7 position
(Mabry et al., 1970). However, a large bathochromic
shift was observed upon addition of sodium methoxide,
indicating the presence of a hydroxyl group at C-4′ in
compound 1 (Markham and Mabry, 1975). The presence
of a chelated hydroxyl group at C-5 was indicated in
the UV spectrum by the observed bathochromic shift
with AlCl3, and the bathochromic shift observed with
AlCl3-HCl indicated the presence of o-dihydroxyl groups
in ring B (Mabry et al., 1970). The UV shifts observed
for compound 1 were consistent with data obtained for
the model compound 5,8,3′,4′-tetrahydroxy-3,7-dimeth-
oxyflavone (Voirin, 1983).
Meth yla tion of Com p ou n d 1. To a solution of compound
1 (3 mg) in acetone (1 mL) were added (CH3)2SO4 (0.6 mL)
and K2CO3 (4 mg), and the mixture was stirred for 5 h at 40

Bioactive Flavonoids of C. diffusa
J. Agric. Food Chem., Vol. 47, No. 1, 1999 39
F igu r e 3. Negative-ion electrospray mass spectrum (A) and
tandem mass spectrum (B) of compound 1 from C. diffusa
eluting at 32.5 min in Figure 2.
F igu r e 1. Structures of antioxidant flavonoids from C. diffusa
(rha, R-L-rhamnopyranosyl).
spectrum of this compound is shown in Figure 3B.
During MS-MS, the deprotonated molecule at m/z 361
fragmented to eliminate a methyl group at m/z 346 (also
observed in the standard mass spectrum), or a molecule
of formaldehyde, [M - H - 28]+, to give a product ion
of m/z 331. Of lower abundance are fragment ions at
m/z 303 and 166, which correspond to elimination of two
formaldehyde molecules or cleavage of the central ring,
respectively (see structure and proposed fragmentation
patterns in Figure 3B).
Diagnostic mass spectral retro-Diels-Alder (RDA)
fragmentation is almost entirely absent in the spectra
of flavones with four or more oxygen substituents (OH
or OCH3), and the spectra are dominated by ions such
as [M - 15]+, [M - 28]+, and [M - 43]+, along with the
molecular ion (Kingston, 1971). Thus, the presence of a
methoxyl group at C-3 was supported by the observation
of a [M - COCH3]+ peak in the EIMS of compound 1,
which is characteristic of 3-methoxyflavones (Kingston,
1971). In the 1H-13C HETCOR NMR spectrum of 1, the
signal at δH 7.27 (H-2′ and H-6′) correlated to δC 107.9,
and δH 6.57 (H-6) also showed a cross-peak with the
resonance at δC 95.3. In addition, the methoxy peak at
δH 3.93 correlated to δC 56.1, whereas that at δH 3.81
correlated to δC 59.1. The OCH3 group (δH 3.93) could
be assigned at C-7 based on 1D-NOE experiment, in
which irradiation at δH 6.57 (H-6) resulted in a NOE
for OCH3-7 (δH 3.93). Analysis of the HMBC NMR
F igu r e 2. LC-MS analysis of the ethyl acetate extract of C.
diffusa using C18 HPLC separation and negative-ion electro-
spray ionization. Ions of m/z 361, 315, and 345 correspond to
the deprotonated molecules of compounds 1, 2, and 3.
LC-MS and LC-MS-MS with electrospray ioniza-
tion were used to analyze the ethyl acetate-soluble
extract of C. diffusa. The negative-ion electrospray LC-
MS analysis in Figure 2 shows the total ion chromato-
gram and computer-reconstructed mass chromatograms
for the deprotonated molecules at m/z 361, 315, and 345,
which correspond to compounds 1, 2, and 3. Compound
1 eluted at 32.5 min and formed a deprotonated mol-
ecule of m/z 361 (Figure 2, second trace). The peak for
this compound was 20 times more abundant than the
peaks of compounds 2 or 3. The mass spectrum corre-
sponding to 1, which eluted at 32.5 min, is shown in
Figure 3A and indicates a high degree of purity (the ion
of m/z 346 is probably a fragment ion). The MS-MS
1
spectrum allowed the complete assignments of the H
and 13C NMR spectra of 1. Thus, the δH 6.57 resonance
also showed cross-peaks with signals at δC 104.5 (C-
10), 152.8 (C-5), and 126.2 (C-8), which strongly sup-
ported this being the H-6 signal, rather than at H-7 or
H-8. Furthermore, the δH 7.27 (H-2′ and H-6′) signal
exhibited connectivities with signals at δC 145.9 (C-3′
and C-5′), 137.1 (C-4′), 107.9 (C-2′ and C-6′), and 156.1
(C-2). The HMBC data supported the location of an OH
group (δH 12.27) at C-5, because correlations were
observed with the δC 152.8 (C-5) and 104.5 (C-10)
signals. A diagnostic feature of 3-methoxyflavones is the
appearance of the OMe-3 group at δC 59.5 ( 0.5,

40 J. Agric. Food Chem., Vol. 47, No. 1, 1999
Chung et al.
Ta ble 1. An tioxid a n t a n d An tim u ta gen ic Activities of
F la von oid s fr om C. d iffu sa a n d In h ibition of
DMBA-In d u ced P r en eop la stic Lesion s in Mou se
Ma m m a r y Or ga n Cu ltu r e
As shown in Table 1, compounds 2 and 4 mediated
significant inhibitory activity and, as such, seem worthy
of evaluation in experimental carcinogenesis models.
Quercetin (4) is one of the most common and abundant
flavonoids in the human diet, found in tea, wine, and
onions (de Vries et al., 1997; Hertog and Katan, 1998),
and has already been reported to exhibit pronounced
antioxidant activity that enables it to scavenge active
oxygen and electrophiles (Huang et al., 1992; Saija et
al., 1995). It is believed able to exert beneficial effects
on human health and has been shown to inhibit the
incidence of papillomas and tumors induced by DMBA
(Balasubramanian and Govindasamy, 1996), as well as
mammary adenocarcinomas in a DMBA-induced rat
mammary carcinogenesis model (Verma et al., 1988).
A study of quercetin (4) as a chemopreventive agent in
sarcoma-180-induced ascites tumors model has been
also reported (Ravichandran et al., 1997). Compounds
1, 3, and 5 were not active in the MMOC model,
irrespective of their potential to facilitate antimutagenic
(compound 1) or antioxidant responses. We have previ-
ously established a general correlation between anti-
mutagenic (desmutagenic) activity in TM677 and in-
hibitory activity in the MMOC system (Shamon et al.,
1994; Ito et al., 1998), so it is possible that this
mechanism applies in the case of compound 2.
antioxidant activity
anti-
(IC50, µg/mL)
mutagenicity
compound
DPPHa
HL-60b (% inhibition)c MMOCd
1
2
3
4
5
10.4
39.2
29.8
15.0
12.0
22.0
21.0
21.6
25.7
>50.0
9.3
18.4
25.5
50.0
90.0
92.0
nde
17
67
45
67
0
39.0
nde
ascorbic acid
2(3)-tert-butyl-
hydroxyanisole
caffeic acid
gallic acid
nordihydro-
guaiaretic acid
12.0
5.0
12.0
6.7
1.5
1.2
a
DPPH, 1,1-diphenyl-2-picrylhydrazyl free radical scavenging
b
activity. HL-60, inhibition of TPA-induced free radical formation
in cultured HL-60 cells. c Inhibition of DMBA-induced forward
d
mutation with S. typhimurium strain TM677. MMOC, inhibition
percent of DMBA-induced preneoplastic lesion formation in mam-
mary gland organ culture at the test concentration of 10 µg/mL.
Based on historical controls, inhibition of g60% is classified as
active at this test concentration. e Not determined.
whereas aromatic methoxyl groups resonate at δC 61.5
( 1.5 (with di-ortho-substitution) and 55.5 ( 1.0 (with
mono- or no ortho-substituents) (Agrawal et al., 1989).
This inference was consistent with the observation of
cross-peaks between δC 59.6 and 137.6 (C-3) and be-
tween δC 56.1 and 153.8 (C-7) in the HMBC experiment.
ACKNOWLEDGMENT
We are grateful to Mr. R. B. Dvorak, Mr. D. Nikolic,
and Mr. N.-C. Kim, Department of Medicinal Chemistry
and Pharmacognosy, College of Pharmacy, University
of Illinois at Chicago, for recording MS data, for obtain-
ing LC-MS and MS-MS data, and for assistance with
500 NMR experiments, respectively, and to Mr. M.
Hawthorne for technical assistance with the mouse
mammary organ culture experiments. The Nuclear
Magnetic Resonance Laboratory of the Research Re-
sources Center, University of Illinois at Chicago, is
thanked for assistance and for the provision of spectro-
scopic equipment used in this study.
1
The ring B H and 13C NMR chemical shifts of 1 were
comparable with published values for those of a model
compound with a symmetrical 4′-hydroxy-3′,5′-dimethoxy-
substituted B ring, pumilaisoflavone D (Yenesew et al.,
1989). From an analysis of all the data described above,
the structure of 1 was assigned as the novel compound
5,8,3′,4′,5′-pentahydroxy-3,7-dimethoxyflavone.
It is known that many flavonoids are effective anti-
oxidants (Bors and Saran, 1987; Robak and Glyglewski,
1988; Bors et al., 1990) and that radical scavenging
activity is potentiated with 3,4-disubstitution or 2,3,4-
trisubstitution in the B ring (Cotelle et al., 1996; van
Aker et al., 1996). Flavonoids 1, 4, and 5 were signifi-
cantly active as free radical scavengers in the DPPH
assay, and their activity compared favorably with the
activity observed with a number of standards in this
same assay (Table 1). Compounds 2 and 3 were only
moderately active in this assay, as well as in the free
radical quenching assay utilizing HL-60 cells, possibly
because of the lesser degree of hydroxylation compared
with 1. However, compounds 1 and 5 were not highly
active in the HL-60 cell-based assay, but compound 4
mediated a significant response. Three of the isolates
were also evaluated for antimutagenic potential in a
DMBA-induced forward mutation assay with S. typhi-
murium strain TM677 (Shamon et al., 1994), and
compounds 1 and 2 were found to be highly active.
These two compounds are likely to either inhibit the
metabolic activation of DMBA or scavenge the reactive
metabolites of DMBA.
LITERATURE CITED
Agrawal, P. K.; Thakur, R. S.; Bansal, M. C. Flavonoids. In
Carbon-13 NMR of Flavonoids; Agrawal, P. K., Ed.; Else-
vier: Amsterdam, 1989; pp 95-168.
Azarowicz, E. N.; Hughes, J . E.; Perkins, C. L. Antibiotics in
plants of Southern California active against Mycobacterium
tuberculosis 607 and Aspergillus niger. Antibiot. Chemother.
1952, 2, 532-536.
Balasubramanian, S.; Govindasamy, S. Inhibitory effect of
dietary flavonol quercetin on 7,12-dimethyl-benz[a]an-
thracene-induced hamster buccal pouch carcinogenesis.
Carcinogenesis 1996, 17, 877-879.
Bhakuni, D. S.; Bittner, M.; Marticorena, C.; Silva, M.; Weldt,
E.; Hoeneisen, M. Screening of Chilean plants for antimi-
crobial activity. Lloydia 1974, 37, 621-632.
Bhakuni, D. S.; Bittner, M.; Marticorena, C.; Silva, M.; Weldt,
E.; Hoeneisen, M. Screening of Chilean plants for anticancer
activity. I. Lloydia 1976, 39, 225-243.
Bors, W.; Saran, M. Radical scavenging by flavonoid antioxi-
dants. Free Radical Res. Commun. 1987, 2, 289-294.
Bors, W.; Heller, W.; Michel, C.; Saran, M. Flavonoids as
antioxidants: Determination of antiradical efficiencies.
Methods Enzymol. 1990, 186, 343-355.
Finally, compounds 1-5 were evaluated for their
potential to inhibit DMBA-induced preneoplastic lesions
with mouse mammary glands in organ culture (MMOC).
As noted previously (Mehta and Moon, 1991), com-
pounds active in this model system are considered good
candidates for full-term cancer chemopreventive studies.
Carmia, B. Antioxidants and cancer. Sci. Med. 1997, 4, 52-
61.
Cotelle, N.; Bernier, J . L.; Catteau, J . P.; Pommery, J .; Wallet,
J . C.; Gaydou, E. M. Antioxidant properties of hydroxy-
flavones. Free Radical Biol. Med. 1996, 20, 35-43.

Bioactive Flavonoids of C. diffusa
J. Agric. Food Chem., Vol. 47, No. 1, 1999 41
de Vries, J . H. M.; J anssen, P. L. T. M. K.; Hollman, P. C. H.;
van Staveren, W. A.; Katan, M. B. Consumption of quercetin
and kaempferol in free-living subjects eating a variety of
diets. Cancer Lett. 1997, 114, 141-144.
Fujita, Y.; Uehara, I.; Morimoto, Y.; Nakashima, M.; Hatano,
T.; Okuda, T. Studies on inhibition mechanism of autoxi-
dation by tannins and flavonoids. II. Inhibition mechanism
of caffeetannins isolated from leaves of Artemisia species
on lipoxygenase dependent lipid peroxidation. Yakugaku
Zasshi 1988, 108, 129-135.
Gru¨ter, A.; Friederich, U.; Wu¨rgler, F. E. Antimutagenic effects
of mushrooms. Mutat. Res. 1990, 231, 243-249.
Hertog, M. G. L.; Katan, M. B. Quercetin in foods, cardiovas-
cular disease, and cancer. Antioxid. Health Dis. 1998, 7,
447-467.
Hong, W. K.; Sporn, M. B. Recent advances in chemopreven-
tion of cancer. Science 1997, 278, 1073-1077.
Huang, M.-T.; Ho, C.-T.; Wang, Z.-Y.; Ferraro, T.; Finnegan-
Olive, T.; Lou, Y.-R.; Mitchell, J . M.; Laskin, J . D.; Yang,
C. S.; Conney, A. H. Inhibitory effects of topical application
of green tea polyphenol fraction on tumor initiation and
promotion in mouse skin. Carcinogenesis 1992, 13, 947-
954.
Inserra, P. F.; Ardestani, S. K.; Watson, R. R. Antioxidants
and immune function. In Antioxidants and Disease Preven-
tion; Garewal, H. S., Ed.; CRC Press: New York, 1997; pp
19-29.
Ito, A.; Shamon, L. A.; Yu, B.; Mata-Greenwood, E.; Lee, S.
K.; van Breemen, R. B.; Mehta, R. G.; Farnsworth, N. R.;
Fong, H. H. S.; Pezzuto, J . M.; Kinghorn, A. D. Antimu-
tagenic constituents of Casimiroa edulis with potential
cancer chemoprevention activity. J . Agric. Food Chem. 1998,
46, 3509-3516.
Kingston, D. G. I. Mass spectrometry of organic compounds
VI. Electron-impact spectra of flavonoid compounds. Tetra-
hedron 1971, 27, 2691-2700.
Robak, J .; Glyglewski, R. J . Flavonoids are scavengers of
superoxide anion. Biochem. Pharmacol. 1988, 37, 837-841.
Roitman, J . N.; J ames, L. F. Chemistry of toxic range plants.
Highly oxygenated flavonol methyl ethers from Gutierrezia
microcephala. Phytochemistry 1985, 24, 835-848.
Saija, A.; Scalese, M.; Lanza, M.; Marzullo, D.; Bonina, F.;
Castelli, F. Flavonoids as antioxidant agents: Importance
of their interaction with biomembranes. Free Radical Biol.
Med. 1995, 19, 481-486.
Sakakibara, M.; Mabry, T. J . A new 8-hydroxyflavonol from
Larrea tridentata. Phytochemistry 1975, 14, 2097-2098.
Shamon, L. A.; Chen, C.; Mehta, R. G.; Steele, V.; Moon, R.
C.; Pezzuto, J . M. A correlation approach for the identifica-
tion of antimutagens that demonstrate chemopreventive
activity. Anticancer Res. 1994, 14, 1775-1778.
Sharma, S.; Stutzman, J . D.; Kelloff, G. J .; Steele, V. E.
Screening of potential chemopreventive agents using bio-
chemical markers of carcinogenesis. Cancer Res. 1994, 54,
5848-5855.
Sporn, M. B. Chemoprevention of cancer. Lancet 1993, 342,
1211-1213.
Sporn, M. B.; Dunlop, N. M.; Newton, D. L.; Smith, J . M.
Prevention of chemical carcinogenesis by vitamin A and its
synthetic analogues (retinoids). Fed. Proc. 1976, 35, 1332-
1338.
Stoner, G. D.; Morse, M. A.; Kelloff, G. J . Perspectives in cancer
chemoprevention. Environ. Health Perspect. 1997, 105, 945-
951.
Tanaka, N.; Murakami, T.; Saiki, Y.; Chen, C.-M.; Gomez, P.
Chemical and chemotaxonomical studies of the Pteris family
and related families (Pteridaceae). Chem. Pharm. Bull.
1978, 26, 3580-3582.
Ulubelen, A.; Timmermann, B. N. Flavonoids from Brickellia
chlorolepis and B. dentata. Phytochemistry 1980, 19, 905-
908.
van Aker, S. A. B.; van den Berg, D. J .; Tromp, M. N. J . L.;
Griffioen, D. H.; van Bennekom, W. J . F.; van der Vijgh, W.
J . F.; Bast, A. Structural aspects of antioxidant activity of
flavonoids. Free Radical Biol. Med. 1996, 20, 331-342.
Verma, A. K.; J ohnson, J . A.; Gould, M. N.; Tanner, M. A.
Inhibition of 7,12-dimethylbenz(a)anthracene- and N-ni-
trosomethylurea-induced rat mammary cancer by dietary
flavonol quercetin. Cancer Res. 1988, 48, 5754-5758.
Voirin, B. UV spectral differentiation of 5-hydroxy-and 5-hy-
droxy-3-methoxyflavones with mono-(4′), di-(3′,4′) or tri-
(3′,4′,5′)-substituted B rings. Phytochemistry 1983, 22,
2107-2145.
Wattenberg, L. W. Chemoprevention of cancer. Cancer Res.
1985, 45, 1-8.
Wenkert, E.; Gottlieb, H. E. Carbon-13 nuclear magnetic
resonance spectroscopy of flavonoid and isoflavonoid com-
pounds. Phytochemistry 1977, 16, 1811-1816.
Lee, S. K.; Mbwambo, Z. H.; Chung, H.-S.; Luyengi, L.; Games,
E. J . C.; Mehta, R. G.; Kinghorn, A. D.; Pezzuto, J . M.
Evaluation of the antioxidant potential of natural products.
Combin. Chem. High Throughput Screen. 1998, 1, 35-46.
Mabberley, D. J . The Plant-Book; University Press: Cam-
bridge, U.K., 1987; p 123.
Mabry, T. J .; Markham, K. R.; Thomas, M. B. Reagents and
procedures for the ultraviolet spectral analysis of flavonoids.
The Systematic Identification of Flavonoids; Springer-Ver-
lag: New York, 1970; pp 35-250.
Markham, K. R. Isolation techniques for flavonoids. In The
Flavonoids; Harborne, J . B., Mabry, T. J ., Mabry, H., Eds.;
Chapman and Hall: London, 1975; pp 21-25.
Markham, K. R.; Mabry, T. J . Ultraviolet-visible and proton
magnetic resonance spectroscopy of flavonoids. In The
Flavonoids; Harborne, J . B., Mabry, T. J ., Mabry, H., Eds.;
Chapman and Hall: London, 1975; pp 45-126.
Markham, K. R.; Ternai, B.; Stanley, R.; Geiger, H.; Mabry,
T. J . Carbon-13 NMR studies of flavonoids III. Naturally
occurring flavonoid glycosides and their acylated derivatives.
Tetrahedron 1978, 34, 1389-1397.
Mehta, R. G.; Moon, R. C. Characterization of effective
chemopreventive agents in mammary gland in vitro using
an initiation promotion protocol. Anticancer Res. 1991, 11,
593-596.
Pryor, W. A. Why is the hydroxyl radical the only radical that
commonly adds to DNA? Hypothesis: It has a rare combi-
nation of high electrophilicity, high thermochemical reactiv-
ity, and a mode of production that can occur near DNA. Free
Radical Biol. Med. 1988, 4, 219-223.
Ravichandran, P.; Elangovan, V.; Govindasamy, S. Chemo-
preventive effect of quercetin in sarcoma-180-bearing mice.
J . Clin. Biochem. Nutr. 1997, 22, 149-154.
Yenesew, A.; Dagne, E.; Waterman, P. G. Flavonoids from the
seed pods of Tephrosia pumila. Phytochemistry 1989, 28,
1291-1292.
Received for review J uly 20, 1998. Revised manuscript received
October 26, 1998. Accepted October 27, 1998. This research
was supported by Program Project P01 CA48112, funded by
the National Cancer Institute, NIH, Bethesda, MD. LC-MS
instrumentation was provided by the Hewlett-Packard Co.,
and the MS-MS equipment was supported by NIH Grant
RR10485 (R.B.v.B.).
J F980784O