Food and Chemical Toxicology 39 (2001) 12111224

www.elsevier.com/locate/foodchemtox

Research Section

Identication, quantitation and biological activity of phytoestrogens

in a dietary supplement for breast enhancement
N.G. Coldham*, M.J. Sauer
Department of Risk Research, Veterinary Laboratories Agency, Addlestone, Surrey KT15 3NB, UK
Accepted 19 July 2001

Abstract
A hop-based dietary supplement, marketed for natural breast enhancement, was analysed to determine the identity and biological
activity of active constituents and potential biological eects in man. Extracts of the dietary supplement were analysed by LCMSn
and phytoestrogens identied and quantitated by reference to appropriate standards. Only hop-associated phytoestrogens were
found in the dietary supplement at signicant concentrations as follows (mean
1 S.D.); 8-prenylnaringenin 10.9
0.3, 6-prenylnaringenin 27.4
1.2, 6,8-diprenylnaringenin 0.9
0.1, xanthohumol 321
17 and isoxanthohumol 81.1
1.6 mg/g of dietary
supplement. The oestrogenic activity of extracts in an ERa reporter gene assay was equivalent to 48
6.3 ng 17b-oestradiol/g supplement and consistent with the 8-prenylnaringenin content. The dietary supplement extract also inhibited reductive 17b-hydroxysteroid oxidoreductase activity, but to a greater extent than a concentration matched reference mixture of hop phytoestrogens.
However, the supplement was only weakly active in mouse uterotrophic assays following administration in feed or after subcutaneous injection of extract at doses of 8-PN up to 250 times higher than that recommended for women. These preliminary
ndings suggest that the dietary supplement is unlikely to produce oestrogenic eects in vivo at the level of the uterus; supporting
evidence is still required to demonstrate ecacy. Crown Copyright # 2001 Published by Elsevier Science Ltd. All rights reserved.
Keywords: Hops; Phytoestrogens; Breast enhancement; Oestrogen receptor; 17b-Hydroxysteroid oxidoreductase; Uterotrophic

1. Introduction
Hormonal breast augmentation has been reported in
a range of clinical situations such as treatment with 17boestradiol (E2) (Hartmann et al., 1998), anti-depressants
(Amsterdam et al., 1997) and during oral contraceptive

use (Jernstrom and Olsson, 1997). A variety of nonmedicinal dietary supplements (DS), marketed on the
basis of claimed breast enhancing action, are widely
available and advertised as low cost natural alternatives
to surgical augmentation. No peer-reviewed scientic
data are available in support of claimed ecacy rates of

Abbreviations (text): APCI, atmospheric pressure chemical ionisation; CID, collision induced dissociation; DS, dietary supplement; E1, oestrone; E2, 17b-oestradiol; EE2, 17a-ethynyl oestradiol; EC50, eective concentration for 50% response; ER, oestrogen receptor; ESI, electrospray
ionisation; FSPI, full scan product-ion; HS-OR, 17b-hydroxysteroid oxidoreductase; IGF-1, insulin like growth factor-1; isoxan, isoxanthohumol;
LCMS, liquid chromatographymass spectrometry; MS1, parent-ion mass spectrometry; MS2, product-ion mass spectrometry; m/z, mass/charge;
N/D, none detected; N/E, not evaluated; ovx, ovariectomised; 6-PN, 6-prenylnaringenin; 8-PN, 8-prenylnaringenin; 6,8-PN, 6,8-diprenylnaringenin;
QA, quality assurance; RCBA, recombinant cell bioassay; SIR, selected ion recording; SD, standard deviation; xan, xanthohumol; [M-H], deprotonated molecular ion.
Abbreviations (chemicals): anhydrosecoisolariciresinol, 3,4-bis[(3-methoxy-4-hydroxy-phenyl)methyl] tetrahydrofuran; coumestrol, 7,12-dihydroxycoumestan; daidzein, 7,40 -dihydroxyisoavone; daidzin, 7-O-glucosyl-40 -hydroxyisoavone; 6,8-diprenylnaringenin, 5,7,40 -trihydroxy-6,8diprenylavanone; 17b-oestradiol, 1,3,5-estratriene-3,17b-diol; 17a-ethynyloestradiol, 17a-ethynyl 1,3,5-estratriene-3,17b-diol; genistein, 5,7,40 -trihydroxyisoavone; genistin, 7-O-glucosyl-5,40 -dihydroxyisoavone; glycitein, 7,40 -dihydroxy-6-methoxyisoavone; isoxanthohumol, 7,40 -dihydroxy5-methoxy-prenylavanone; matairesinol, dihydro-3,4-bis[(4-hydroxy-3-methoxyphenyl)-methyl]-2-furanone; naringenin, 5,7,40 -trihydroxyavanone; naringin, naringenin-7-rhamnoglucoside; 8-prenylnaringenin, 5,7,40 -trihydroxy-8-prenylavanone; 6-prenylnaringenin, 5,7,40 -trihydroxy6-prenylavanone; secoisolariciresinol, 2,3-bis[4-hydroxy-3-methoxyphenyl)methyl]-1,4-butanediol; xanthohumol, 20 ,40 , 4-trihydroxy-60 -methoxy-30 prenylchalcone.
* Corresponding author. Tel.: +44-1932-357827; fax: +44-1932-57445.
E-mail address: n.g.coldham@vla.ma.gsi.gov.uk (N.G. Coldham).
0278-6915/01/$ - see front matter. Crown Copyright # 2001 Published by Elsevier Science Ltd. All rights reserved.
PII: S0278-6915(01)00081-3

1212

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

over 70% for some of these products; to date, trials of

ecacy have instead been conducted, and outcomes
widely reported, by the media (Anon., 1999; Law, 1999).
Some suppliers suggest that such products act by selective stimulation of glandular breast tissue by phytoestrogens acting through the oestrogen receptor. Many
available DSs contain a comparable range of natural
products which may include undisclosed varieties of
hops, kava kava root, saw palmetto berry, don quai and
cereal grains. The biologically active chemical entities in
these supplements and their potential biological activity
have not been identied.
Signicant changes in breast size occur naturally during pregnancy, at puberty, the menopause, and to a
much lesser extent through the menstrual cycle (Jernstrom and Olsson, 1997). The endogenous hormonal
signals regulating the opposing processes of breast cell
mitosis and apoptosis include the female gonadal steroids (Anderson et al., 1982), prolactin and growth factors such as IGF-1 (Hartmann et al., 1998), although
the precise mechanisms by which they act through different classes and forms of steroid receptors are not
completely understood (McCarty, 1989). Perturbation
of the mitotic/apoptotic equilibrium, by hormonally
active chemicals such as oestrogens of endogenous or
exogenous origin, in favour of breast cell accumulation,
provides a plausible mechanism for increased gland
size. However, risk factors for breast cancer, including
oral contraceptive use and hormone replacement
therapy, have dictated a precautionary approach to use
of oestradiol for breast augmentation (Hartmann et al.,
1998).
Commonly known phytoestrogens arise from a
variety of food sources. Soya is a rich source of the
isoavones daidzein, genistein and glycitein which act
as weak oestrogens (Garrett et al., 1999). Estimates of
their oestrogenic potency depend on the characteristics
of the test system (Kupfer, 1988). The most potent isoavone is genistein, and relative potency values between
0.1 and 0.001% relative to E2 (100%) have been reported (Messina et al., 1994). Diets rich in soya have also
been shown to evoke weak oestrogenic eects in
humans, promoting vaginal cell maturation in postmenopausal women (Baird et al., 1995) and slightly
extending the follicular phase of the menstrual cycle
(Cassidy et al., 1994). Conversely, genistein may antagonise the action of E2 on target tissues and thus act as
an anti-oestrogen (Messina et al., 1994). Many other
foods, such as the female owers of hops (Humulus
lupulus L), are also reported to be oestrogenic (Verdeal
and Ryan, 1979). Estimates of the oestrogenic activity
of hops range between 0 (inactive) and 300 mg E2
equivalents/g (Verdeal and Ryan, 1979). A range of
phytoestrogens have been identied in hops (Milligan et
al., 1999) including 8-prenylnaringenin (8-PN), 6-prenylnaringenin (6-PN), xanthohumol (xan) and iso-

xanthohumol (isoxan). The potency of these hop

phytoestrogens has been measured with a variety of in
vitro oestrogen receptor based test systems including
competitive binding, reporter gene and breast cancer
cell proliferation assays. Literature values for the
potency of 8-PN, the most potent, range between 20 and
0.2% (Kitaoka et al., 1998; Milligan et al., 1999)
compared with E2 (100%). This evidence indicates that
8-PN is substantially more oestrogenic than other dietary phytoestrogens including the isoavones (Breinholt
and Larsen, 1998). Normal dietary intake of hop phytoestrogens is relatively low; for instance the concentration of 8-PN in beer is typically less than 18 ng/ml
(Stevens et al., 1999; Tekel et al., 1999). The lignans
secoisolariciresinol and matairesinol occur in certain
foods of plant origin and are the putative respective
precursors of the weakly oestrogenic mammalian lignans enterodiol and enterolactone (Welshons et al.,
1987).
In the present study, the potential oestrogenic activity
of the DS was investigated with a recombinant cell bioassay (RCBA) expressing the human oestrogen receptor
(a) linked to a reporter gene. This approach oers a
robust and highly sensitive means of measuring receptor
mediated oestrogenic activity which has been validated
by comparison with other in vivo and in vitro screening
methods (Coldham et al., 1997) and by external assessment (Fang et al., 2000). Rodent uterotrophic bioassays
also take account of pharmacokinetic factors which
may have a signicant impact on biological activity;
these bioassays provide a means of placing in perspective oestrogenic activity found using in vitro screening
methods (Galey et al., 1993; Coldham et al., 1997). Key
enzymes in oestrogen metabolism provide a further
target for endocrine modulation. 17b-Hydroxysteroid
oxidoreductase (HS-OR) is a pivotal enzyme catalysing
the inter-conversion of oestrone to the more potent E2.
Reductive type-1 HS-OR activity is sensitive to inhibition by phytoestrogens, may be extracted from human
cells and used to investigate the potential anti-oestrogenic eect of other chemical entities and food extracts
by inhibition of this enzyme (Makela et al., 1995).
Complex mixtures of potentially active chemicals
derived from natural products may be readily analysed
by LCMSn, which enables their separation and identication of specic components by acquisition of product-ion mass spectra (Coldham et al., 1999).
Identication and quantitation of the active chemical
entities may be conrmed since, when corrected for
analyte concentration, the biological activity of a test
sample should be consistent with that of appropriate
concentration matched reference calibration standards.
The objective of the present study was to determine the
identity, concentration and potential biological signicance to the consumer of phytoestrogens in a DS
which is promoted for cosmetic breast enhancement.

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

2. Materials and methods

2.1. Chemicals
Genistein, naringenin, naringin, genistin, daidzein, E2
and 17a-ethynyl oestradiol (EE2) were supplied by
Sigma (Poole, Dorset, UK); coumestrol, daidzin and
glycitein were obtained from Apin Chemicals (Abingdon, Oxon, UK); 6-PN, 8-PN, 6,8-diPN, xan, isoxan,
matairesinol, secoisolariciresinol and anhydrosecoisolariciresinol were from Plantech Chemicals (Reading
University, Berks, UK). Stock standard solutions of
phytoestrogens (1 mg/ml) were prepared in DMSO, or
methanol (coumestrol only), and stored frozen at 4 C.
[4-14C]Genistein (2.07 GBq/mmol) was obtained from
Amersham International (Coldham et al., 1999) (Little
Chalfont, Bucks, UK) to a purity of 98%. Ammonium
acetate, hydrochloric acid, acetonitrile and methanol
were purchased from BDH, (Poole, Dorset, UK). T-47D cells were obtained from the European Collection of
Cell Cultures (Salisbury, Wiltshire, UK).
2.2. Food samples
Two batches of a readily available brand of DS for
natural breast enhancement were obtained directly from
the supplier in the United Kingdom (batch 1) and via
the Medicines Control Agency, UK (batch 2). White
bread was obtained from a local supermarket and a
quality assurance (QA) sample prepared from soya our
was provided by the Central Science Laboratory (Sand
Hutton, York, UK). Food samples were stored at
80 C prior to analysis.

1213

methanol (9 ml) and the combined ltrates taken to dryness under reduced pressure at 55 C in a centrifugal evaporator. Residues were dissolved in either water (1 ml) or,
for bread and soya our extracts only (which contained
glycoside conjugates), subjected to acid treatment by
incubation with 1 m HCl (1 ml) at 100 C for 2 h for
hydrolysis of glycoside conjugates, prior to extraction of
phytoestrogens into ethylacetate (25 ml). The phases
were separated by centrifugation (2000 g for 5 min) and the
pooled ethyl acetate extracts taken to dryness under nitrogen at 55 C. The residues were dissolved in acetonitrile (1
ml) and further diluted in either HPLC mobile phase
(80:20 ammonium acetate:acetonitrile), minimal yeast
medium or phosphate buer, for analysis by HPLCMSn,
RCBA or the HS-OR assays (see below), respectively.
2.5. HPLC
Conditions were established empirically to enable
separation of a wide range of potential analytes including the isoavone glycosides daidzin and genistin, the
aglycones daidzein, genistein and glycitein, coumestrol,
naringin, naringenin, the lignans secoisolariciresinol,
anhydrosecoisolariciresinol and matairesinol, and the hop
phytoestrogens 6-PN, 8-PN, 6,8-diPN, xanthohumol
and isoxanthohumol. Samples of the nal extracts (20
ml) were chromatographed at a ow rate of 0.3 ml/min on
a Hypersil Elite C18 column (1502.1 mm; 5 mm particle
size) and eluted with a linear binary gradient of ammonium acetate (10 mm) (A) and acetonitrile (B) (t=0 min,
A 80%; t=26.6 min, A 0%; t=30 min, A 80%; t=40
min, A 80% and next injection). The HPLC retention
times of phytoestrogen standards (1 mg/ml) under these
chromatography conditions are provided in Table 1.

2.3. Animals
2.6. LCMS optimisation
The uterotrophic activity of the DS was evaluated in
prepubertal (day 18) and adult ovariectomised (ovx)
mouse models. The mice were obtained from Charles
River, U.K. Ltd (Margate, Kent, UK) and maintained
on RM1 (soya free) feed (Special Diet Services,
Witham, Essex, UK). Care and treatment of the mice
was in accordance with the Animals (Scientic Procedures) Act 1986 of the United Kingdom and under the
supervision of a veterinary surgeon.
2.4. Phytoestrogen extraction
DS tablets (n=5; average weight 0.996
0.023 g) and
bread (5 g) were mechanically ground to a ne powder or
crumbs. Representative weighed portions of DS powder,
bread crumbs (both 1 g) and soya our (40 mg) were
transferred to volumetric asks and made up with 80%
methanol (v/v; 10 ml) and phytoestrogens extracted by
probe sonication (230 s). Extracts (1 ml) were passed
through 1 mm PTFE lters, the lters were washed with

Optimal mass spectrometry conditions were investigated for protonated and deprotonated molecular ions
of 8-PN, as appropriate, by continuous ow infusion of
a 1 mg/ml standard into the instrument (LCQ, ThermoQuest, Hemel Hempstead, Herts, UK) at 10 ml/min via
the electrospray (ESI) or atmospheric pressure chemical
ionisation (APCI) probes in positive and negative ionisation modes. Negative ion APCI provided the greatest
response for 8-PN and for all other analytes except
anhydrosecoisolariciresinol, for which ESI (negative
ion) was most favourable. A collision induced dissociation (CID) energy of 40% proved optimal for acquisition of product ion (MS2) mass spectra from the
deprotonated phytoestrogen molecular ions [M-H].
2.7. Phytoestrogen identication
Phytoestrogens were tentatively identied in DS
extracts by acquisition of full scan (MS1) mass spectra

1214

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

(m/z 240650) and extraction of appropriate [M-H]

chromatograms and comparison with those produced
from reference standards. Identity was conrmed by
comparison of chromatographic retention time and [MH] product ion (MS2) mass spectra with reference
standards prior to quantitation.
2.8. Phytoestrogen quantitation
The hop phytoestrogens 8-PN, 6-PN, 6,8-diPN, xanthohumol and isoxanthohumol were quantitated on the
base ion extracted from full scan product ion mass
spectra of each analyte. This data acquisition mode
oered the exibility for simultaneous quantitation and
conrmation of identity. The isoavones genistein,
daidzein and glycitein were quantitated by selected ion
recording (SIR) of their [M-H]. Chromatographic
peaks were automatically detected and the mass of
phytoestrogen calculated after interpolation from
appropriate calibration curves over the range 11000 ng/
ml for 8-PN, 6-PN, 6,8-diPN, xan and isoxan and 10
10,000 ng/ml for the isoavones using the LCQuanTM
(ThermoQuest) software package. Data acquisition
modes, parent and target product ions used for the
quantitation of these analytes are provided in Table 1.
2.9. Assay validation
Analytical recovery through the extraction procedure
was determined by fortication of DS powder (1 g) with

5 mg 8-PN or soya our (40 mg) with [14C]genistein (18

kBq) immediately prior to extraction. The extraction
procedure was further validated by investigating the
relationship between the quantity of DS (0.11 g) or
soya our (780 mg) extracted and the mass of hop and
isoavone phytoestrogens quantitated. The within and
between assay coecients of variation were determined
by analysis of replicate samples (n=5) in a single assay
and those in separate assays (n=36), respectively. The
extent of acid hydrolysis of isoavone conjugates was
evaluated by LCMS analysis (scanning m/z 250600)
of HCl and untreated soya our extracts and comparison of selected ion chromatograms corresponding to
the [M-H] of the glycosides and aglycones.
2.10. Recombinant cell bioassay
The potency of 8-PN relative to E2 and oestrogenic
activity of DS extracts were determined on two separate
occasions using the recombinant yeast cell bioassay
(RCBA) described elsewhere (Coldham et al., 1997).
The oestrogenic potency of 8-PN in the RCBA was calculated by estimation of the concentration required to
give a half maximal response (EC50) from calibration
curves for E2 (1100,000 pm) and 8-PN (11000 nm) and
application of the formula: potency=(EC50 E2/EC50 8PN)  100. The oestrogenic activity of dilutions of the
DS extract (0.0250.000025, v/v) were adjusted for 8PN content (g 8-PN/g DS as determined by LCMS; see
below) to enable comparison of the response with that

Table 1
Phytoestrogen analytical criteria
Analyte

Mass

HPLC rt (min)

Acquisition mode

m/z target ions for quantitation [M-H]

Daidzin
Genistin
Naringin
Secoisolariciresinol
Daidzein
Glycitein
Coumestrol
Genistein
Matairesinol
Naringenin
Anhydrosecoisolariciresinol
Isoxan
17b-Oestradiol
17a-Ethynyl oestradiol
8-PN
6-PN
Xan
6,8-diPN

416.1
432.1
580.2
362.1
254.1
284.1
268.1
270.1
358.1
272.1
344.1
354.1
272.1
296.1
340.1
340.1
340.1
408.1

2.8
5.0
6.0
6.8
8.3
8.6
10.5
10.6
10.6
10.9
11.7
12.7
13.1
14.3
15.5
17.9
18.9
21.7

253
283

269

[353] 233b

[339] 219b
[339] 219b
[353] 233b
[407] 287b

n/d
n/d
SIR
SIR
n/d
SIR
n/d
n/d
n/d
FSPI
n/d
n/d
FSPI
FSPI
FSPI
FSPI

Summary of potential phytoestrogen target analytes, HPLC retention times, quantitative mass spectrometer data acquisition modes and m/z ratio of
selected target ions. The isoavones genistein, daidzein and glycitein were quantitated with the mass spectrometer set in the selected ion recording
(SIR) mode. The hop phytoestrogen were analysed by the more specic data acquisition mode of full scan product ion (FSPI) mass spectrometry of
the parent molecular ion [M-H] followed by extraction and quantitation on the base product ion.
a
Indicates that these analytes were quantitated following acid hydrolysis to the parent aglycone, n/d indicates not detected in any food matrix analysed.
b
Denotes base product ion extracted from the full scan product ion spectrum for quantitation of hop phytoestrogens.

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

of calibration curve of the 8-PN standard. The oestrogenic potency of 8-PN and the other hop phytoestrogens has been extensively evaluated and reported
elsewhere (Milligan et al., 1999). The oestrogenic activity of the DS relative to E2 (ng E2 equivalents/g DS)
was determined by application of the formula: relative
activity=EC50 E2/EC50 DS.
2.11. HS-OR assay
Type-1 HS-OR was extracted from the T-47D human
breast cancer cells (Miettinen et al., 1996). T-47D cells
were grown to 50% conuence in 450 cm2 asks and
harvested by treatment with trypsin (0.05%, w/v)
EDTA (0.02%, w/v) and collected by centrifugation (5
min; 100 g). Type-1 HS-OR was released from the cells
by probe sonication, precipitated with 50% (v/v)
ammonium sulphate, dialysed against phosphate buer
(50 mm; pH 7.4) for 48 h and stored frozen at 80 C.
HS-OR activity was measured by incubation of enzyme
preparation (50 ml; 13.6 mg protein) with 500 nm oestrone (E1) containing 18.3 kBq [3H] E1, a cofactor generating system (1 mm NADP+, 10 mm glucose-6phosphate and 2.5 units of glucose-6-phosphate oxidoreductase) and phytoestrogen standards or food extracts
(5 ml; in phosphate buer) in a total volume of 250 ml at
37 C for 2 h. Substrate and product were extracted into
diethyl ether (5 ml), taken to dryness under nitrogen at
40 C and redissolved in HPLC mobile phase. Extracts
were chromatographed by HPLC on a Hypersil Elite
C18 column (4.6100 mm) and eluted at a ow rate of 1
ml/min with an isocratic mobile phase system consisting
of 60:40 H2O:acetonitrile and detected by on line ow
scintillation counting (A525, Packard, Pangbourne,
Berks, UK). Inhibition of HS-OR activity was expressed
as a percentage of activity of control incubations (without phytoestrogen standard or food extract). DS
extracts were diluted over the range (0.10.0001, v/v)
and normalised for total hop phytoestrogen content
(mg/g) to enable comparison with a calibration curve
prepared from hop phytoestrogens at the ratio found in
the DS. Performance of the HS-OR assay was assessed
by comparison of the inhibitory eect of genistein with
literature values and by estimation of between and
within batch coecients of variation using genistein (10
mm) in separate (n=3) assays and replicate (n=5)
extracts of the DS in a single assay.
2.12. Prepubertal mouse uterotrophic bioassay
DS (12 g; batch 1) was extracted using the procedure
described above, the residue was dissolved in dimethyl
sulphoxide (0.5 ml) and oestrogenic activity evaluated
using the prepubertal mouse uterotrophic bioassay
using E2 as calibrant as described elsewhere (Coldham
et al., 1997). The uterotrophic activity of EE2 was also

1215

determined as a control to enable comparison between

the prepubertal and adult ovx mouse bioassays (see
below). The limited commercial availability of 8-PN
prevented its use as a calibration standard. Day 18 prepubertal mice (n=7 per treatment) were injected subcutaneously with 0.1 ml of test compound or DS extract
dissolved in corn oil (Sigma, Poole, Dorset, UK) on 3
consecutive days at the doses shown in Table 3. Mice
were sacriced on day 4 and the weights of the animals
and uteri recorded. The mean fractions of blotted uterus
to body weight 100 were calculated.
2.13. ovx mouse uterotrophic bioassay
DS was administered to adult ovx mice diluted with
feed using EE2 as a calibrant. EE2 was selected as a
calibrant in this bioassay due to the poor oral bioavailibility of E2. Feed containing DS or EE2 was prepared
by mixing powdered RM1 feed and DS in the proportions (DS:RM1) 1:1, 1:9 and 1:99 (w/w). Feed (200 g)
was mixed with water (150 ml), fortied with EE2 (10,
100, 1000, 10000 ng EE2/g RM1) where appropriate, to
produce a thick paste which was extruded with a piston
from an open syringe barrel to produce 1570 mm pellets. These pellets were dried to their original weight
(200 g) by incubation at 60 C for 24 h. The concentration of phytoestrogens in these feeds was not determined since inactivation during the drying process at
the relatively low temperature of 60 C was considered
unlikely. ovx mice were provided with feed containing
EE2 or DS (n=7 mice per treatment) for 5 days and the
quantity of feed consumed per day recorded. Mice were
sacriced on day 6 and the weights of the animals and
uteri recorded. The mean fraction of blotted uterus to
body weight 100 were calculated. The daily dose (ng
EE2 or 8-PN/g mouse) was calculated from the mass of
EE2 or DS added to the feed, the concentration of 8-PN
in the DS, the average daily consumption of the feeds
and the average mouse weight.

3. Results
3.1. Phytoestrogen identication
Selected mass (MS1) chromatograms of DS extracts
contained ions at m/z 339, 353 and 407 corresponding to
[M-H] of 8-PN and 6-PN (both 339), xanthohumol
and isoxanthohumol (both 353) and 6,8-diPN (407).
These were positively identied from their respective
chomatographic retention times (Fig. 1) and production mass spectra (Fig. 2), as isoxanthohumol (12.7 min;
spectrum A), 8-PN (15.5 min; spectrum B), 6-PN (17.9
min, spectrum identical to B), xanthohumol (18.9 min;
spectrum identical to A), and 6,8-diPN (21.7 min; spectrum C). The unidentied peak in the product ion (MS2)

1216

Table 2
Summary of phytoestrogen concentrations in DS, soya our and white bread
Analytes
8-PN (mg/g)

6-PN (mg/g)

Xan (mg/g)

Isoxan (mg/g)

6,8-diPN (mg/g)

genistein (mg/g)

daidzein (mg/g)

glycitein (mg/g)

D.S. (batch 1)
D.S. (batch 2)
Soya our (QA)
White bread
Within batch CV% (n=5)
Between batch CV% (n)
Formula (y=Ax+B) and correlation coecient
(r) between mass phytoestrogen (ng for hop
phytoestrogens) and sample size.
Calculated daily intake (mg) of phytoestrogen
from the DS for breast enhancement.

9.7
0.6
10.929
0.3
N/D
N/D
6
2 (3)
A=9.8
B=169 (0.983)

27.2
1.9
27.396
1.2
N/D
N/D
4
N/E
A=25
B=419 (0.994)

323.0
41.0
321.0
17.0
N/D
N/D
5
N/E
A=303
B=7632 (0.992)

77.3
4.9
81.062
1.6
N/D
N/D
2
N/E
A=80
B=350 (0.999)

1.0
0.2
0.9
0.1

8
N/E
A=0.73
B=260 (0.967)

N/D
N/D
1509
174
5.6
0.4
3.6
11 (6)
A=1.25
B=0.64 (0.992)

N/D
N/D
1052
150
4.2
0.1
1.8
14 (6)
A=1.07
B=0.33 (0.999)

1.8
0.5
2.4
0.1
155
79
0.6
0
14.8
51 (6)
A=0.10
B=0.16 (0.985)

103155

273409

32204830

7921188

1930

Phytoestrogen concentrations in two batches (1 & 2) of a DS for breast enhancement (D.S.), soya our and white bread, assay performance criteria and daily intake of hop phytoestrogen from the
DS. The average concentration
1 SD are provided for each phytoestrogen where appropriate. The quoted concentrations are not corrected for analytical recovery. The soya our material was a
quality assurance (QA) sample. N/D indicates none detected; N/E not evaluated. The daily intake of hop phytoestrogens is calculated for a dose of 1015 tablets as recommended by the suppliers.
The correlation (r) and best t formula where y=Ax+B are shown for the masses of food (DS or soya our) analysed and phytoestrogen (ng for hop phytoestrogens; mg for isoavones) detected.

Table 3
Utrotrophic activity of extract of DS in prepubertal mice
Treatment (dose/0.1 ml/day)

Dose (ng/g/day)

Body weight (g; mean
S.D.)

Uterine weight (mg; mean
S.D.)

Uterine wt/body wt fraction100 (mean
S.D.)

Control (vehicle)
E2 (5 ng)
E2 (10 ng)
E2 (25 ng)
E2 (50 ng)
E2 (100 ng)
EE2 (5 ng)
EE2 (25 ng)
EE2 (100 ng)
DS (0.024 mg 8-PN)
DS (0.084 mg 8-PN)
DS (0.24 mg 8-PN)
DS (0.84 mg 8-PN)
DS (2.4 mg 8-PN)

0
0.33
0.66
1.65
3.21
7.48
0.33
1.94
6.99
1.7
5.8
17.2
63.8
186.7

13.5
2.7
15.4
2.5
15.2
1.9
15.1
1.3
15.6
1.7
13.4
1.4
14.9
1.2
12.9
1.8
14.3
3.4
14.0
3.4
13.7
1.9
13.9
1.9
12.5
1.1
12.9
2.0

16.4
5.2
32.1
11.5
47.9
9.8
64.0
16.0
94.9
25.1
79.5
20.5
39.7
8.7
83.4
6.0
96.4
24.2
21.1
10.2
18.0
5.1
23.1
8.9
18.3
1.7
17.6
5.2

0.122
0.04
0.209
0.08**
0.315
0.07***
0.423
0.11***
0.609
0.16***
0.595
0.15***
0.266
0.06***
0.647
0.05***
0.674
0.17***
0.151
0.07
0.131
0.04
0.166
0.06*
0.146
0.01*
0.137
0.04

Female 18-day-old prepubertal mice were administered E2, EE2 or DS extract dissolved in 0.1 ml of corn oil by subcutaneous injection for 3 consecutive days. The daily dose rate provided (ng E2,
EE2 or 8-PN/g mouse) was calculated from the mass of E2, EE2 or 8-PN present in the DS extract, or dilutions thereof, divided by the average body weight of the mice in that group. The statistical
signicance of treated relative to the untreated animals was evaluated by the Students t-test (*P <0.05; **P<0.01; ***P< 0.001).

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

Food

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

chromatograms of m/z 219 at 16.9 min (Fig. 1) yielded a

spectrum identical to B (Fig. 2) indicating the likelihood
that this represented a isomer of PN. Genistein, daidzein, coumestrol, matairesinol, secoisolarisiresinol,
anhydrosecoisolarisiresinol, naringin, naringenin and
glycoside conjugates of the hop phytoestrogens were not
detected in extracts of the DS. The isoavonoid phytoestrogens genistein, daidzein and glycitein were positively identied in parallel acid hydrolysed extracts of
soya our and white bread. The lignans and coumestrol
were not detected in either soya our or white bread.
3.2. Validation of quantitative phytoestrogen assays
The recoveries of 8-PN and [14C]genistein following
extraction were 97
7% and 91
4%, respectively. Linear relationships were obtained between the quantity of
DS and soya our analysed and the masses of phytoestrogen determined (Table 2). The within and between
assay coecients of variation for the phytoestrogens
quantitated are provided in Table 2. Hydrolysis of the
isoavone glycosides (glucose, acetyl glucose and malonyl glucose) was complete after treatment of soya our
for 2 h with 1 m HCl.
3.3. Quantitation of phytoestrogens
The concentration of hop phytoestrogens in two batches of a DS (measured in the same assay) and iso-

1217

avones in the soya our QA sample and white bread

are provided in Table 2. The concentration of genistein
and daidzein in the soya our QA sample was within 1
SD of target values (mean
1 S.D.) of 1376
246 and
1100
104 mg/g, respectively. The intake of individual
hop phytoestrogens by a consumer taking the recommended dose of 1015 tablets was calculated from their
concentration in the DS (Table 2).
3.4. Oestrogenic activity
The eect of E2, 8-PN, DS extract (batch 1) and DS
extract adjusted for 8-PN content (g 8-PN/g DS) on bgalactosidase expression in the RCBA is shown in Fig. 3.
The relative oestrogenic potency (17b-oestradiol
=100%) of 8-PN was 0.4
0.1%. The oestrogenic
activity of the DS extract (mean+1 S.D.) indicated a
content of 48
6.3 ng of E2 equivalents/g of DS.
3.5. HS-OR assay
The within and between batch coecients of variation
in the HS-OR assay were 4 and 14%, respectively. The
inhibitory eect of hop phytoestrogens and genistein on
HS-OR activity are shown in Fig. 4. Genistein (10 mm)
produced a 32% inhibition of HS-OR activity, a value
similar to that reported in literature (Makela et al., 1995).
A reference mixture of hop phytoestrogens in the ratio
1:024:0.09:0.03:0.007
xan:isoxan:6-PN:8-PN:6,8-diPN

Fig. 1. Representative product ion chromatograms derived from analysis of an extract of a DS for breast enhancement. Parent ions ([M-H]) were
fragmented with a collision energy of 40% to yield product ions. (A) Total product ion chromatogram; (B) product ion chromatogram for m/z 219
derived from [M-H] m/z 339; (C) product ion chromatogram for m/z 233 derived from [M-H] m/z 353; (D) product ion m/z 287 derived from
[M-H] m/z 407. Each chromatogram was normalised to the most abundant peak. The identity of each peak is indicated where appropriate. The
NL value (top right) indicates the intensity (counts per second) of the most abundant peak in that chromatogram.

1218

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

similar to that found in the DS (Table 2) was selected

for analysis since, unlike oestrogenic potency (Garrett et
al., 1999), their inhibitory eects on HS-OR activity
were of a similar order (Fig. 4). The inhibitory eect of
a DS extract (batch 1; adjusted for hop phytoestrogen
content) on HS-OR activity was greater than that of the
reference calibration curve containing the mixture of
hop phytoestrogens (Fig. 5).

3.6. Prepubertal mouse uterotrophic bioassay

The eect of E2, EE2 and DS extract administered
subcutaneously to prepubertal mice is shown in Table 3.
The uterus/body weight fraction was signicantly
(P < 0.01) increased by treatment E2 and EE2 compared
with the control group. The EC50 of E2 and EE2 were
0.82 and 0.51 ng, respectively, which gave a relative

Fig. 2. Representative product-ion mass spectra of hop phytoestrogens. Xan and isoxan yielded product-ion spectrum A from an [M-H] of m/z
353, 8-PN and 6-PN yielded spectrum B from an [M-H] of m/z 339 and 6,8-di-PN yielded spectrum C from an [M-H] of m/z 407.

1219

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

potency for EE2 of 161% (E2=100%). The uterus/

body weight fraction was increased by the DS extract
(63.8 and 17.2 ng 8-PN/g doses only); however, this
eect was of marginal statistical signicance (P < 0.05)
and not dose dependent.
3.7. ovx mouse uterotrophic bioassay
The eect on the uterine/body weight fraction of EE2
and DS administered orally to ovx mice in feed is shown
in Table 4. The uterine/body weight fraction was signicantly (P < 0.01) increased by EE2 at all doses tested
and slightly reduced (P < 0.05) by the DS (1:99) compared with the control group.
Fig. 3. Oestrogenic activity of 17b-oestradiol (square symbols), 8-PN
(circles), DS (triangles) and DS corrected for 8-PN content (stars)
determined with the RCBA. Log b-galactosidase activity is plotted
against log mass (g)/well of reference standard (17b-oestradiol or 8PN), DS or DS corrected for 8-PN content (g 8-PN/g DS) assayed per
well. Error bars indicate
1 S.D.

Fig. 4. Inhibitory eect (0.310 mm) of 8-PN (squares; 0.13.4 mg/ml),

6-PN (diamonds), xan (triangles), isoxan (circles) and genistein (10 mm;
bar) on reductive HS-OR activity. Error bars indicate
1 S.D.

Fig. 5. Inhibitory eect of a mixture of hop phytoestrogens (circles)

and DS extract (squares) (10 mm; bar) on reductive HS-OR activity.
The hop phytoestrogen mixture was prepared from stock reference
standards to the fraction 1:0.24:0.09:0.03:0.007 xan:isoxan:6-PN:8PN:6,8-diPN which was consistent with their fraction in the DS
determined by LC-MS (Table 2). Percentage inhibition of HS-OR
activity is plotted against the total (xan+isoxan+8-PN+6-PN+6,8diPN) hop phytoestrogen concentration in the standard mixture and
DS. Error bars indicate
1 S.D.

Table 4
Uterotrophic activity of DS in ovariectomised mice
Treatment
(concentration in feed)

Average daily food

consumption (g)/mouse/day

Dose
(ng/g/day)

Body weight
(g; mean
S.D.)

Uterine weight
(mg; mean
S.D.)

Control (vehicle)
EE2 (2 ng/g feed)
EE2 (20 ng/g feed)
EE2 (200 ng/g feed)
EE2 (2000 ng/g feed)
DS:SFF (1:99, w/w)
DS:SFF (1:9, w/w)
DS:SFF (1:1, w/w)

4.9
4.8
3.2
3.4
1.6
4.4
3.1
3.4

0
0.3
1.9
23.4
111.9
13.6
95.0
542.6

32.8
5.6
32.4
2.0
33.2
2.2
29.2
3.7
29.1
5.2
32.5
5.9
32.2
2.7
31.0
2.0

18.0
2.6
22.9
3.3
33.6
5.8
118.7
20.0
139.1
40.4
14.4
2.3
19.0.
3.5
17.4
3.3

Uterine wt/body wt
fraction100 (mean
S.D.)
0.056
0.008
0.071
0.010**
0.101
0.017***
0.417
0.110***
0.503
0.188***
0.045
0.009*
0.060
0.013
0.056
0.009

Ovx mice (n=7/group) were treated with either EE2 or DS diluted with soya free feed (SFF) at the concentrations shown in parentheses. The daily
dose rate (ng EE2 or 8-PN/g mouse) was calculated from the mass of EE2 or DS added to the feed, the concentration of 8-PN in the DS, the average
daily consumption of such feeds and the average body weight of mice in that group. The statistical signicance of treated relative to the untreated
animals was evaluated by the Students t-test (*P<0.05; **P<0.01; ***P <0.001).

1220

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

Fig. 6. Chemical structures of the hop phytoestrogens, 8-PN, 6-PN, isoxan and xan found in the DS and the isoavone genistein found in soya our
and the natural steroidal oestrogen 17b-oestradiol. The broken arrows indicates the potential pathway of isoxan biotransformation by O-demethylation to 8-PN.

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

4. Discussion
Extracts of the DS contained a range of candidate
phytoestrogens, which were positively identied on the
basis of HPLC retention time and product-ion mass
spectra as isoxanthohumol, xanthohumol, 6-PN, 8-PN,
6,8-diPN (Fig. 6). Trace amounts of glycitein were also
found in these extracts. An isomer of PN was detected
on the basis of identical product-ion spectra but dierent chromatographic properties to the reference standards. This unidentied entity is unlikely to represent a
40 isomers since prenylation of the avanone C ring
would yield a dierent product-ion spectrum to that
observed. Isomerism in the alkyl side-chain (Seo et al.,
1997) has been reported for other avanones and provides the basis for a candidate structure for this unidentied compound since further sites for prenylation
are not available in the avanone A ring. Xanthohumol
and isoxanthohumol produced identical product-ion
spectra suggesting similar routes of collision induced
dissociation.
Mass spectrometry also provided a highly specic
means for quantitation following extraction of the baseion from the product-ion spectra of each analyte. The
concentrations of hop phytoestrogens in the DS were in
the order xan > isoxan > 6-PN > 8-PN > 6,8-diPN. The
8-PN content of the DS was approximately an order of
magnitude less than that reported for hop cones (Milligan et al., 1999). This may reect dilution with other
constituents of the DS and the reported variation in
phytoestrogen concentration between hop varieties
(Milligan et al., 1999); the concentration of the other
phytoestrogens were negligible. The concentration of
the hop phytoestrogens in the DS were substantially less
than those of daidzein and genistein present in soya
our. Hydrolysed extracts of the soya our QA sample
contained isoavones at concentrations within target
ranges and similar to values reported in the literature
(Garrett et al., 1999). Isoavones were also detected
in white bread, which reected the inclusion of soya
our in this particular brand. The selected lignans
and coumestrol were not detected in any of the foods
analysed for phytoestrogen content. Identication and
quantitation of hop phytoestrogens, especially 8-PN, in
the DS warranted further investigation of its biological
activity.
In the present study, the oestrogenic potency of 8-PN
determined with the RCBA was 0.4% which was within
the range (200.2%) of literature values (Kitaoka et al.,
1998; Milligan et al., 1999) and compared favourably to
a value of 0.6% obtained with another recombinant
yeast cell bioassay (Milligan et al., 1999). The presence
and position of the prenyl moiety contributed to the
oestrogenic activity of 8-PN and since naringenin (Ruh
et al., 1995) and 6-PN (Milligan et al., 1999) are signicantly less potent. The DS extract was also oestro-

1221

genic in the RCBA and such activity was consistent with

8-PN content as determined by LC-MSn. The relatively
high content of other weakly oestrogenic hop phytoestrogens including 6-PN and isoxan also identied in the
DS may be responsible for the observed divergence of
the DS curve (adjusted for 8-PN content) from the
reference calibration curve for 8-PN. However, the correspondence between these curves provides an element
of conrmation that 8-PN is the major oestrogen in the
DS extract. The oestrogenic activity of the dietary supplement determined with RCBA (48 ng of E2 equivalents/g) was considerably lower than the highest value
reported in the literature for hops (Verdeal and Ryan,
1979). Similar oestrogen receptor based assays have
been employed for the analysis of the phytoestrogen
content of soya based foods and provided statistically
signicant correlations with isoavone content (Garrett
et al., 1999). Thus, at least in vitro, both 8-PN and the
DS were biologically active through the oestrogen
receptor in the RCBA.
In addition to receptor binding, prenylation of naringenin is likely to have a signicant impact on other
determinants of oestrogenic activity. In keeping with
expectation, comparison of HPLC retention times indicated that prenylation of naringenin increased hydrophobicity substantially, and to a greater extent than that
of ethynylation of 17b-oestradiol; indeed 6,8-diPN had
a longer retention time than either 8- or 6-PN. Ethynylation of oestrogens and progestins at the 17a position
greatly increases their bioavailibility and thus potency
as oral contraceptives. Thus it is tempting to speculate
at this point that this could substantially augment the
oestrogenic activity of 8-PN in vivo. Prenylation is a
common post-translational C-terminal modication
used to anchor proteins in cell membranes by increasing
hydrophobicity (Khwaja et al., 2000). Thus prenylation
is likely to have a signicant inuence on the pharmacokinetics of 8-PN compared with many other phytoestrogens, perhaps encouraging disposition into
hydrophobic compartments of the body. Preclinical
studies in laboratory species using radiotracer techniques would help resolve these pharmacokinetic issues
and establish the concentration of 8-PN in target tissues
such as the breast.
In common with the isoavones and coumestrol
(Makela et al., 1995), the hop phytoestrogens also
inhibited reductive HS-OR activity demonstrated by
reduced conversion of oestrone to the more potent oestrogen 17b-oestradiol. In contrast to oestrogenic
potency (Milligan et al., 1999), all the hop phytoestrogens inhibited HS-OR activity by a similar order and
over the same concentration range. Thus xan, the most
abundant hop phytoestrogen in the DS, is likely to be
the most biologically signicant inhibitor of HS-OR
activity. However, the dilution curve of the DS extract
adjusted for total hop phytoestrogen content was

1222

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

signicantly more active than the calibration curve of

hop phytoestrogens suggesting the presence of other, as
yet unidentied, inhibitors of HS-OR activity. The
relative importance of receptor and enzyme mediated
mechanisms of any phytoestrogen action in vivo will
depend on the concentration and potency of the active
entities. Indirect mechanisms, such as inhibition of HSOR activity, are likely to be subordinate where phytoestrogen acts directly at the oestrogen receptor in target organs. Activity of 8-PN in the RCBA was found at
concentrations three orders of magnitude lower than
those required for inhibition of HS-OR and further
argue for the relative importance of the oestrogen
receptor route of action.
Demonstration of true oestrogenic eects requires
corroborative evidence from in vivo studies. In contrast
with expectation from in vitro ndings, DS administered subcutaneously as an extract or orally in feed, had
little uterotrophic activity in the prepubertal and adult
ovx mouse bioassays respectively. Laboratory prepared
and commercial extracts of hops have been intensively
evaluated elsewhere (Fenselau and Talalay, 1973) for
uterotrophic activity following subcutaneous administration to prepubertal mice and were found to be inactive. In the present study, oral and subcutaneous routes
of administration were compared since their pharmacokinetic properties are likely to dier and impact on
observed biological activity. Historically, subcutaneous
administration has most commonly been used for uterotrophic bioassays (Fenselau and Talalay, 1973; Coldham et al., 1997) and has the potential for superior
sensitivity compared with the oral route since test compounds are not subjected to rst-pass metabolism and
inactivation in the gut during absorption and liver.
However, oral administration is the most meaningful
route of administration in the context of food intake by
man.
In contrast to the ndings here and with hop extracts
(Fenselau and Talalay, 1973), 8-PN has been shown to
have uterotrophic activity following subcutaneous
administration to ovx rats (Miyamoto et al., 1998).
However, the dose administered (30 mg/g/day) was 160
times higher than the highest achieved in the present
study following subcutaneous dosing with the DS
extract (0.187 mg/g/day) and represents a 14,000 fold
excess over the likely intake of 8-PN (2.1 ng/g) from the
DS following consumption of the daily recommended
dose of 15 tablets. Administration of DS in feed to ovx
mice gave reasonable representation of the route of
human exposure and eliminated potential problems
associated with achieving quantitative solvent extraction of all active components. This route also enabled
administration of a large quantity of DS in feed; the top
dose of 8-PN (542 ng/g) fed to the ovx mice represented
a 258-fold excess over the likely daily intake (2.1 ng/g)
for natural breast enhancement in women. The oral

route also considered the bioavailability of active entities across the gut and through the liver including the
potential for biotransformation to active or inactive
metabolites. Although the DS had a small inhibitory
eect on uterine weight at the lowest dose (13.6 ng 8PN/g) provided in feed, this eect was of low statistical
signicance (P < 0.05) and not dose dependent. All
models commonly employed for the evaluation of oestrogenic activity have inherent limitations (Kupfer,
1988; Coldham et al., 1997); in vitro test systems, such
as recombinant yeast and breast cancer cells, take little
account of pharmacokinetic factors which may have a
profound eect on biological activity. Moreover, such
pharmacokinetic factors are likely to dier between test
compounds, model test systems and routes of administration and provide a plausible explanation for the
detection of oestrogenic activity in vitro using the
RCBA but not with the uterotrophic bioassays in prepubertal and adult ovx mice at the doses indicated.
Further studies are required to provide an estimate of
hop phytoestrogens bioavailability following oral
adminstration.
Although biotransformation is generally regarded as
a process of metabolic inactivation prior to excretion of
polar metabolites there are many well known examples
of metabolic activation. Demethylation by cytochrome
P450 enzymes in the liver or by gut microora represents a common metabolic pathway for activation of
proestrogens. The oestrogenic potency of the isoavone
phytoestrogen formononetin and synthetic steroid mestranol is increased following demethylation to daidzein
(Batterham et al., 1971; Lundh et al., 1988) and EE2
(Schmider et al., 1997), respectively. Similarly, in the
event that isoxan was O-demethylated to 8-PN in vivo,
this would signicantly increase the potential oestrogenic activity of DS (Fig. 6). Relatively modest levels of
such O-demethylation would greatly augment the
intrinsic oestrogenic activity of the DS since the concentration of isoxan exceeded that of 8-PN by over
sevenfold. Clearly, the ndings from the mouse bioassays indicate that other mechanisms operate which
result in the net reduction of activity at the uterus rather
than an increase.
The DS is marketed for breast enhancement, and
supporting product literature for some of these products
indicate that this eect is achieved through the oestrogen receptor. The ndings of the present study raise
many issues including the potential for biological activity through mechanisms other than the oestrogen
receptor and limitations of rodent models at the level of
the uterus for the estimation of potential oestrogenic
activity in the breast tissue of women. Progesterone is
also thought to play an important role in breast epithelial cell proliferation since the peak of mitotic activity is
found during the luteal phase of the menstrual cycle
(Anderson et al., 1989) when the plasma concentrations

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

of this gonadal steroid are maximal. Furthermore, oral

contraceptives containing progestin only are more
potent stimulants of breast cell proliferation than combined preparations (Anderson et al., 1989). Recent
studies (Kuiper et al., 1998) indicate that the relative
binding anity of a range of isoavanoid phytoestrogens is signicantly greater to the b than a form of the
oestrogen receptor. The relative binding anity of 8-PN
and other hop phytoestrogens to the dierent forms of
ER and the progesterone receptor has not been determined nor their distribution in epithelial breast tissue or
relative roles in regulating proliferative processes.
Consumption of foods rich in isoavones and lignans
has been associated with a range of benecial health
eects (Messina et al., 1994). Reports of serious adverse
health eects are limited to animals consuming relatively
large quantities of isoavones (Setchell et al., 1998).
There are no reports of breast development in infants or
augmentation in adults following dietary exposure to
isoavones (Klein, 1998). However, the weakly oestrogenic isoavone formononetin has been shown to
enhance mammary gland proliferation and raise the
plasma concentration of prolactin in ovariectomised mice
following subcutaneous doses of 40 mg/kg (Wang et al.,
1995). In contrast to the isoavones there is relatively little scientically-based information concerning the health
implications following dietary exposure to hop phytoestrogens. Breast enhancement aside, hop preparations
may represent a valuable food source of phytoestrogens.
This said, signicant dierences in relation to both in vivo
potency and concentration in food exist between isoavonoid and hop phytoestrogens which are likely to
modulate biological activity and thus health eects and
limit the utility of such direct comparisons. Further
information concerning the full range of biological activities, pharmacokinetics and modifying inuence of metabolism on the biological activity of prenylnaringenins in
man should provide a better understanding of the eects
of hop phytoestrogens and thus potential risks or benets
that they may present to the consumer.

Acknowledgements
This study was commissioned by the Joint Food
Safety and Standards Group of the Department of
Health and Ministry of Agriculture, Fisheries and
Food, UK. The authors express their thanks to Dr. Don
Clarke of the Central Science Laboratory, Sand Hutton,
York for provision of the soya our QA sample.

References
Amsterdam, J.D., Garcia-Espana, F., Goodman, D., Hooper, M.,
Hornig-Rohan, M., 1997. Breast enlargement during chronic antidepressant therapy. Journal of Aective Disorders 46, 151.

1223

Anderson, T.J., Battersby, S., King, R.J.B., McPherson, K., Going,

J.J., 1989. Oral contraceptive use inuences resting breast proliferation. Human Pathology 20, 11391144.
Anderson, T.J., Ferguson, D.J.P., Raab, G.M., 1982. Cell turnover in
the resting human breast: inuence of parity, contraceptive pill,
age and laterality. British Journal of Cancer 44, 177181.
Anon., British Broadcasting Corporation On-line, Watchdog Face
Value, Bigger Breasts, 17 June, 1999 (www.bbc.co.uk/watchdog/
stories/fvbigbre).
Baird, D.D., Umbach, D.M., Lansdell, L., Hughes, C.L., Setchell,
K.D.R., Weinberg, C.R., Haney, A.F., Wilcox, A.J., McLachlan,
J.A., 1995. Dietary intervention study to assess estrogenicity of
dietary soy among postmenopausal women. Journal of Clinical
Endocrinology and Metabolism 80, 16851690.
Batterham, T.J., Shutt, D.A., Hart, N.K., Braden, A.W.H., Tweeddale, H.J., 1971. Metabolism of intraruminally administered
[4-14C]formononetin and [414C]biochanin A in sheep. Australian
Journal of Agricultural Research 22, 131138.
Breinholt, V., Larsen, J.C., 1998. Detection of weak estrogenic avonoids using a recombinant yeast strain and a modied MCF-7 cell
proliferation assay. Chemical Research in Toxicology 11, 622629.
Cassidy, A., Bingham, S., Setchell, K.D.R., 1994. Biological eects of
a diet of soy protein rich in isoavones on the menstrual cycle of
premenopausal women. American Journal of Clinical Nutrition 60,
333340.
Coldham, N.G., Howells, L.C., Santi, A., Montesissa, C., Langlais,
C., King, L., Sauer, M.J., 1999. Biotransformation of genistein in
the rat: Elucidation of metabolite structure by product ion mass
fragmentologyn. Journal of Steroid Biochemistry and Molecular
Biology 70, 169184.
Coldham, N.G., Dave, M., Sivapathasundaram, S., McDonald, D.,
Connelly, C., Sauer, M.J., 1997. Evaluation of a recombinant yeast
cell estrogen screening assay. Environmental Health Perspectives
105, 734742.
Fang, H., Tong, W., Perkins, R., Soto, A., Prechtl, N.V., Sheehan,
D.M., 2000. Quantitative comparisons of in vitro assays for estrogenic activities. Environmental Health Perspectives 108, 723729.
Fenselau, C., Talalay, P., 1973. Is oestrogenic activity present in hops?
Food and Cosmetics Toxicology 11, 597603.
Galey, F.D., Mendez, L.E., Whitehead, W.E., Holstege, D.M.,
Plumle, K.H., Johnson, B., 1993. Estrogenic activity of forages:
diagnostic use of the classical mouse uterine bioassay. Journal of
Veterinary Diagnosis and Investigation 5, 603608.
Garrett, S.D., Lee, H.A., Friar, P.M.K., Morgan, M.R.A., 1999.
Validation of a novel estrogen receptor-based microtitration plate
assay for the determination of phytoestrogens in soy based foods.
Journal of Agricultural and Food Chemistry 47, 41064111.
Hartmann, B.W., Kirchengast, S., Albrecht, A.E., Huber, J.C., 1998.
Hormonal breast augmentation: prognostic relevance of insulin-like
growth factor-I. Gynecological Endocrinology 12, 123127.
Jernstrom, H., Olsson, H., 1997. Breast size in relation to endogenous
hormone levels, body constitution, and oral contraceptive use in
healthy nulligravid women aged 19-25 years. American Journal of
Epidemiology 145, 571580.
Khwaja, A., Connolly, J.O., Hendry, B.M., 2000. Prenylation inhibitors in renal disease. Lancet 355, 741744.
Kitaoka, M., Kadokawa, H., Sugano, M., Ichikawa, K., Taki, M.,
Takaishi, S., Iijima, Y., Tsutsumi, S., Boriboon, M., Akiyama, T.,
1998. Prenylavonoids: a new class of non-steroidal phytoestrogen
(Part 1) (Isolation of 8-isopentenylnaringenin and an initial study on
its structure-activity relationship). Planta Medica 64, 511515.
Klein, K.O., 1998. Isoavones, soy-based infant formulas, and relevance to endocrine funtion. Nutrition Reviews 56, 193204.
Kuiper, G.G.J.M., Lemmen, J.G., Carlsson, B.O., Corton, J.C., Safe,
S.H., Van Der Saag, P.T., VanDerBurg, B., Gustafsson, J.-A., 1998.
Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor b. Endocrinology 139, 42524263.

1224

N.G. Coldham, M.J. Sauer / Food and Chemical Toxicology 39 (2001) 12111224

Kupfer, D., 1988. Critical evaluation of methods for detection and

assessment of estrogenic compounds in mammals: strengths and
limitations for application to risk assessment. Reproductive Toxicology 1, 147153.
Law C. Sunday Times, Lifestyle magazine, 5 September 1999, pp. 3334
(www.sunday-times.co.uk/news/pages/sti/99/09/05/stihealif02001).
Lundh, T.J.O., Pettersen, H., Kiessling, K., 1988. Demethylation and
conjugation of formononetin and daidzein in sheep and cow liver
microsomes. Journal of Agricultural and Food Chemistry 36, 2225.
Makela, S., Poutanen, M., Lehtimaki, J., Kostian, M.-L., Santti, R.,
Vihko, R., 1995. Estrogen-specic 17b-hydroxysteroid oxidoreductase type 1 (E.C. 1.1.1.62) as a specic target for the action of
phytoestrogen. Proceedings of the Society for Experimental Biology
and Medicine 208, 5159.
McCarty, K.S., 1989. Proliferative stimuli in the normal breast: estrogens or progestins. Human Pathology 20, 11391144.
Miettinen, M.M., Mustonen, M.V.J., Poutanen, M.H., Isomaa, V.V.,
Vikho, R.K., 1996. Human 17b-hydroxysteroid dehydrogenase type
1 and type 2 isoenzymes have opposite activities in cultured cells and
characteristic cell- and tissue-specic expression. Biochemical Journal 314, 839845.
Messina, M.J., Persky, V., Setchell, K.D.R., Barnes, S., 1994. Soy
intake and cancer risk: a review of the in vitro and in vivo data.
Nutrition and Cancer 21, 113131.
Milligan, S.R., Kalita, J.C., Heyerick, A., Rong, H., De Cooman, L.,
DeKeukeleire, D., 1999. Identication of a potent phytoestrogen in
hops (Humulus lupulus L.) and beer. Journal of Clinical Endocrinology and Metabolism 83, 22492252.
Miyamoto, M., Matsushita, Y., Kiyokawa, A., Fukuda, C., Iiijima,
Y., Sugano, M., Akiyama, T., 1998. Estrogenic eects of 8-isopentenylnaringenin on bone metabolism. Planta Medica 64, 516
519.

Ruh, M., Zacharewski, T., Connor, K., Howell, J., Chen, I., Safe, S.,
1995. Naringenin: A weakly estrogenic bioavonoid that exhibits
antiestrogenic activity. Biochemical Pharmacology 50, 14851493.
Schmider, J., Greenblatt, D.J., Von Moltke, L.L., Karsov, D., Vena,
R., Friedman, H.L., Shader, R.I., 1997. Biotransformation of mestranol to ethinyl estradiol in vitro: the role of cytochrome P-450 2C9
and metabolic inhibitors. Journal of Clinical Pharmacology 37, 193
200.
Seo, E.-K., Silva, G.L., Chai, H.-B., Chagwedera, T.E., Farnsworth,
N.R., Cordell, G.A., Pezzuto, J.M., Kinghorn, A.D., 1997. Cytotoxic prenylated avanones from Monotes engleri. Phytochemistry
45, 509515.
Setchell, K.D.R., Zimmer-Nechemias, L., Cai, J., Heubi, J.E., 1998.
Isoavone content of infant formulas and the metabolic fate of
phytoestrogens in early life. American Journal of Clinical Nutrition
68, 1453s1461s.
Stevens, J.F., Taylor, A.W., Clawson, J.E., Deinzer, M.L., 1999. Fate
of xanthohumol and related prenylavonoids from hops to beer.
Journal of Agricultural and Food Chemistry 47, 24212428.
Tekel, J., De Keukeleire, D., Rong, H., Daeseleire, E., Van Peteghem,
C., 1999. Determination of the hop-derived phytoestrogen, 8-prenylnaringenin, in beer by gas chromatography/mass spectrometry.
Journal of Agricultural and Food Chemistry 47, 50595063.
Wang, W., Tanaka, Y., Han, Z., Higuchi, C.M., 1995. Proliferative
response of mammary glandular tissue to formononetin. Nutrition
and Cancer 23, 131140.
Welshons, W.V., Murphy, C.S., Koch, R., Calaf, G., Jordan, V.C.,
1987. Stimulation of breast cancer cells in vitro by the environmental estrogen enterolactone and the phytoestrogen equol. Breast
Cancer Research and Treatment 10, 169175.
Verdeal, K., Ryan, D.S., 1979. Naturally-occurring estrogens in plant
foodstusa review. Journal of Food Protection 42, 577583.