Breast Cancer Research and Treatment 69: 153–164, 2001.© 2001 Kluwer Academic Publishers. Printed in the Netherlands.

Report

Chemoprevention of DMBA-induced mammary cancer in rats by dietarysoy

Daniela Gallo1, Sabrina Giacomelli1, Fabio Cantelmo1, Gian Franco Zannoni2, GabriellaFerrandina1, Erika Fruscella1, Antonella Riva3, Paolo Morazzoni3, Ezio Bombardelli3,Salvatore Mancuso1, and Giovanni Scambia1

1Department of Obstetrics and Gynecology, 2Department of Histopathology, Catholic University of the SacredHeart, Rome; 3Indena S.p.A., Milan, Italy

Key words: chemoprevention, mammary cancer, rat, soy, tumor differentiation

Summary

This study was designed to assess the potential chemopreventive effect of the administration of a standardized soyextract, SOYSELECTTM, on 7,12-Dimethylbenz[a]anthracene (DMBA)-induced mammary tumors in rats. Threegroups, 24 females each, were used. Animals were fed either a phytoestrogen-free diet alone (control) or the samediet supplemented with 0.35% or 0.7% of soy extract. Treatment started at weaning and continued to the end ofthe study (24 weeks after DMBA administration). At day 50 of age all animals received via oral gavage 80 mg/kgDMBA. Only tumors subsequently classified as adenocarcinomas were considered for data evaluation. In rats onthe soy diet, mammary tumors took a longer period of time to develop as compared to control rats. However, atthe end of the study, no relevant difference in tumor incidence and multiplicity was observed among the groups.The most significant changes were seen between control and soy-treated groups when tumor dimension and resultsfrom histopathologic examination were considered. The latter, in fact, showed a dose-dependent reduction in thepercentage of poorly differentiated tumors in treated animals. This change was statistically significant in animalsreceiving 0.7% soy. In addition, assessment of estrogen and progesterone receptor (ERα, PR) levels, revealed asignificant reduction in the percentage of ERα and PR positive tumors in animals receiving 0.7% dietary soy, whencompared to controls. Interestingly, genistein and daidzein plasma levels determined at the end of the study werewithin the range of those detected in people consuming large amounts of soyfoods.

Introduction

It has been well established that cancer rates show astrikingly different geographical distribution in vari-ous populations of the world: while cancer of thebreast, colon and prostate are major public healthproblems in the western world, rates of these can-cers are significantly lower in Asian populations witheastern lifestyles [1]. However, when Asians mi-grate to the United States, their offspring develop thecommon western cancers at about the same rates asAmericans [2–4], suggesting that the differences ob-served are largely attributable to environmental factorsrather than genetics. In particular, there is strongepidemiological evidence that diet has a role in the

development of breast cancer. Asian diet is high insoy and several researchers have suggested that itmay exert a protective action on the developmentof breast cancer, even though other diet features,such as a much lower content of saturated fats anda higher amount of fibre may have a role in redu-cing cancer risk. Soybeans and most soy productscontain the isoflavones genistein and daidzein whichhave been suggested to possibly prevent hormone-dependent breast cancer by virtue of their potentialestrogen-modulator activity. Overall, however, res-ults from in vitro, in vivo and analytical epidemi-ological studies aimed at evaluating the associationbetween soy intake and reduction of breast cancer,are conflicting. In a case-control study, Ingram et al.

154 D Gallo et al.

[5] reported a significant reduction in breast cancerrisk in premenopausal as well as postmenopausal wo-men consuming phytoestrogens. In contrast, in a dif-ferent case-control study of women in Singapore, thiseffect was apparent in premenopausal but not in post-menopausal women [6], while other authors do notsupport this property of soy consumption at all [7].

When considering animal studies, results obtainedare difficult to evaluate, as reviewed by Messina etal. [8], because of poor experimental design and/ormarked differences in quali-quantitative compositionof the soy products used. This latter point is partic-ularly crucial, since isoflavone bioavailability can bestrongly influenced by the food matrix [9] and onlyfew researchers have determined their plasma concen-trations after administration to animals. Some morerecent studies have shown that genistein (which is con-sidered the most important bioactive soy compound)by itself significantly protects rats against DMBA-induced mammary cancer when it is administered toneonatal or prepubertal rats [10, 11] or, again, duringperinatal life [12]. In contrast, other authors showedthat short in utero [13] or perinatal [14] exposure togenistein results in an increased susceptibility againstchemically-induced mammary tumors.

Finally, with regard to the in vitro studies, despitethe identification of several mechanisms that couldaccount for the hypothesized anticancer effects ofphytoestrogens, it has been noted that, in many cases,the concentration required to obtain an effect washigher than phytoestrogen plasma levels in peopleconsuming large amounts of soyfoods.

However, despite the lack of definitive evidence fora chemopreventive effect of soy, scientific interest onits possible application as a protective agent againstcancer is growing for several reasons: first of all, oneshould consider that even a small reduction in tumorincidence could have a great impact when a largepopulation is exposed; secondary, diet is a relativelymodifiable entity and thus a potential target for changeand importantly, if consumed in amounts comparablewith those on eastern diets, soy is considered to beessentially safe.

The aim of this study was to determine whether adietary soy supplement could affect DMBA-inducedmammary tumors in female Sprague-Dawley rats.For this purpose we used a standardized soy extract,SOYSELECTTM, a product commercially available,which has been already tested in our laboratory andproved to be a safe and efficacious therapy for relief ofvasomotor symptoms in postmenopausal women [15].

Dosages of the soy extract in the diet gave genisteinand daidzein plasma levels comparable to those foundin Japanese population [16] and the treatment sched-ule was intended to mimic a potential modification ofdietary habit in women starting from peripubertal age.Particular attention was given to the histopathologicalclassification of tumors to delineate separate subtypesof adenocarcinomas. Moreover, estrogen and proges-terone receptor levels in mammary tumors was alsoinvestigated.

Materials and methods

Animals

A total of 90 weaningly female Sprague-Dawley ratswas obtained from Harlan Nossan S.r.l. (Correzzana,Milan, Italy). Animals were assigned to the threetreatment groups to give approximately equal groupmean body weight. Each group was comprised of 24females. Animals were housed in a purpose-built fa-cility with a controlled environment. Environmentalcontrols were set at the following levels: temperat-ure 22 ± 2◦C, humidity 50%, 12 h light and 12 h dark.Procedure and facilities followed the requirementsof Commission Directive 86/609/EEC concerning theprotection of animals used for experimental and otherscientific purposes. Italian legislation is defined in theDecreto Legislativo No. 116 of 27 January, 1992. Inaddition the UKCCCR guidelines for the welfare ofanimals in experimental neoplasia were followed.

Diet

Control or treated diets were presented to the anim-als from weaning (median weaning date as nominalage 21 days) until the end of the study (24 weeksafter DMBA administration). The diet used in thisstudy was based on the purified AIN-76A diet andwas prepared by A. Rieper (Vandoies, Bolzano, Italy).The diet contained the following ingredients: casein20%, cornstarch 15%, sucrose 46.5%, cellulose 5%,corn oil 5%, mineral mixture 5.2%, vitamin mix-ture 2.0%, trace and ultratrace elements 0.8%, DL-methionine 0.3%. Animals were treated with eitherthe phytoestrogen-free diet alone (control group) orthis diet supplemented with 0.35% or 0.7% (w/w) ofa standardized extract of soy, SOYSELECTTM (thiscontaining 12% of isoflavones and 35% of saponines)(Indena S.p.A, Milan, Italy). Three different batches ofthe diet were used during the study. Samples of each

Soy activity in chemically-induced mammary cancer 155

batch were taken during the study to verify the stabilityof the diet over the utilization period, in terms of con-centration of genistin/genistein and daidzin/daidzein.Results obtained by HPLC analysis indicated that iso-flavones were stable in the diet (data not shown). Foodand water were supplied ad libitum during the study.

Tumor induction and experimental design

At day 50 of age all animals received via oral gavage80 mg DMBA/kg body weight. DMBA was dissolvedin sesame oil (8 mg/ml) and was administered to eachanimal at a dose volume of 10 ml/kg. Sesame oil andDMBA were obtained from Sigma-Aldrich (Milan,Italy). Animals were palpated twice a week in order torecord the presence, location, size and date of detec-tion of all tumors. Tumor weights (mg) were estimatedfrom two-dimensional measurements (mm): tumorweight = (length × width2)/2 [17]. Body weight andfood consumption were recorded weekly during thestudy. Animals were killed 24 weeks after DMBA ad-ministration. Those animals bearing excessive tumorburden were sacrificed before termination of the studyfor ethical reasons. At the time of necropsy all tumorswere removed, fixed in 10% buffered formol salineand subsequently dehydrated and blocked in paraffin.The paraffin block was cut into 5 µm sections, fixedon slides and processed for light microscopy (stainedwith hematoxylin and eosin) or immunohistochem-istry (see below). Histopathological evaluation wascarried out on coded slides following the IARC (In-ternational Agency for Research on Cancer) classific-ation of mammary rodent tumors [18]. Uterus, ovariesand pituitary were also rapidly removed and weighed.

Immunohistochemical analysis

Cells expressing progesterone and estrogen receptorswere identified after 1 h incubation at room temper-ature by using the rabbit polyclonal antibody PR (C-19 : sc-538, Santa Cruz Biotechnology, Rome, Italy) at1 : 50 dilution and the rabbit polyclonal antibody ERα

(M-20 : sc-542, Santa Cruz Biotechnology) at 1 : 80 di-lution, respectively. Tissue slices were then incubatedfor 30 min with the anti rabbit mouse/human adsorbedBiotin conjugated antibody (IgG-B sc-2040, SantaCruz Biotechnology) at 1 : 400 dilution, treated withthe avidin-peroxidase complex (LSAB Kit peroxidase,DAKO, Copenhagen, Denmark) and the product ofthe reaction was revealed by incubation with 3-amino-9-ethylcarbazole (AEC Substrate System, DAKO,Copenhagen, Denmark).

Positive cells from control and treated sampleswere identified in a count from five randomly selec-ted fields each containing at least 300 histologicallyidentified neoplastic cells. Immunoreactivity was de-termined by two independent observers according toa simplified scoring system: cases were rated negativeif none of the cells within the lesions were stained.Discrepant cases were reviewed. Immunohistochem-ical scoring was determined without any knowledge towhich group the rats belonged. Evaluation was per-formed on 38 tumors from both control and 0.7%soy-treated animals.

Phytoestrogen analysis

At necropsy, venous blood samples were collec-ted from the caudal vena cava of animals. Bloodsamples were collected into heparinized tubes andcentrifuged as soon as possible after collection at2500 × g, at +4◦C for 10 min. Daidzein, genisteinand equol were determined in all plasma sampleseither as such or after enzymatic hydrolysis with aβ-glucuronidase/-arylsulfatase (type H-1 from Helixpomatia; Sigma-Aldrich, Milan, Italy) mixture toevaluate the compounds both as free and as total iso-flavones. The analytical method, previously describedby Supko and Phillips [19] has been modified byLimonta et al. (manuscript in preparation). Briefly,plasma samples for the determination of total daidzein,genistein and equol were prepared by adding 4-hydroxybenzophenone (internal standard), methanol,acetate buffer (500 mM, pH 5) and β-glucoronidase/-arylsulfatase solution to 250 µl of plasma aliquots;samples were incubated overnight at 37◦C. For ana-lysis of free isoflavones, enzymatic hydrolysis wasomitted. For the extraction procedure methyl ter-butylether (TBME) was added to each sample; the sampleswere subsequently centrifuged at 2800 × g for 10 minand the organic layer was then evaporated under astream of nitrogen. The residue was reconstituted witha solution consisting of 75% 50 mM ammonium form-ate buffer pH 4, 25% acetonitrile v/v, centrifugedand the supernatant was used for the determinationby HPLC. Extracted samples were loaded into a C18chromatographic column (4 µm, 3.9 × 15 cm), andeluted using a linear gradient of 20–50% acetonitrile in50 mM ammonium formate buffer pH 4, over a periodof 12 min at a flow rate of 1 ml/ min. The eluate wasmonitored by UV absorbance at 260 nm, to determ-ine genistein and daidzein and at 280 nm to determineequol. Isoflavone concentrations were quantified by

156 D Gallo et al.

comparison of peak areas with standard curves. In-strumentation consisted of a model LC-10 AS binarypump (Shimadzu), a model SPD-10 A UV-vis detector(Shimadzu), a model SIL- 10 A Autosampler injectionsystem (Shimadzu) and a Nova-Pak C18 analyticalcolumn (Waters).

Statistics

Data pertaining to body weight, food consumptionand organ weight were analyzed for homogeneity ofvariance using Bartlett’s test. If the group variance ap-peared homogenous a parametric ANOVA was used,followed by Dunnett’s multiple comparison test. If thevariances were heterogeneous, log or reciprocal trans-formations were made in an attempt to stabilize thevariances. If the variances remained heterogeneous, anon parametric test such as the Kruskall–Wallis test,followed by the Dunn’s test for multiple comparisons,was used. Only mammary tumors classified as adeno-carcinomas were included in the statistical evaluationof the data. Moreover, we omitted from the analysistumors found at necropsy.

The log-rank test was used to analyze tumorlatency. Incidence was evaluated with the M-L chi-square test and Fisher’s exact test. Tumor multi-plicity and tumor weight were analyzed by usingthe Kruskall–Wallis test followed by the Dunn’s testfor multiple comparison. Results from immunohisto-chemical analysis were evaluated by the Fisher’s exacttest for proportion.

Results

Tumor induction

In rats on the soy diet mammary tumors took a longerperiod of time to develop as compared to control rats:the mean appearance time for the first tumor was11.7 ± 5.4, 12.9 ± 5.2 and 14.2 ± 5.5 (mean ± SD)weeks for control, 0.35 and 0.7% soy fed groups,respectively, these differences approaching statisticalsignificance (p = 0.07). No relevant difference in thepercentage of tumor-bearing animals was observed atthe end of the study when comparing treated groupsto control (Figure 1(A)). At this time, the percent ofanimals affected was 89%, 89 and 86% for control,0.35 and 0.7% soy fed groups, respectively. However,evaluation of the data obtained during the early ob-servation period (i.e., up to week 10) showed a lowerincidence of affected animals in soy treated groups.

Throughout the study, the mean number of adeno-carcinomas per tumor-bearing rat remained lower intreated groups with respect to control (Figure 1(B))although this difference did not reach any statisticalsignificance. At the end of the study, the averagenumber of tumors/tumor-bearing rat was 3.9 ± 2.5,2.5 ± 1.4 and 3.2 ± 1.7 (mean ± SD) for control, 0.35and 0.7% soy groups, respectively.

A statistically significant difference in total tu-mor burden/rat was observed between controls andanimals receiving 0.7% soy on weeks 12, 13 and14 of the study (Figure 1(C)). Median total tumorburden/tumor-bearing rat, calculated at the end ofstudy, was approximately 2800, 1200 and 800 mgfor control, 0.35 and 0.7% soy groups, respectively,even though, due to interindividual data variability, nostatistical significance was achieved.

Histopathological analysis revealed that approx-imately 90% of the palpable masses from controland treated rats were adenocarcinomas and 6% werefibroadenomas or adenomas. The remaining 4% weremainly represented by lymph nodes or epidermalcysts. The histological classification of mammary ad-enocarcinomas, did not reveal any differences amongthe groups on the distribution of various tumorsubtypes (data not shown). On the contrary, whenconsidering the degree of tumor differentiation, adose-dependent reduction in the percent of poorly dif-ferentiated tumors, with a concomitant increase of thewell-differentiated ones, was observed in treated an-imals, this difference reaching statistical significancefor animals fed 0.7% soy (Table 1). Representativemorphological patterns are shown in Figure 2.

Results from immunohistochemical analysis re-vealed that 25/38 (65.8%) of control masses and 14/38(36.8%) of 0.7% soy treated tumors were ERα pos-itive, while 13/38 (34.2%) control and 1/38 (2.6%)soy treated tumors were PR positive (Table 2). Thesedifferences were statistically significant (p = 0.02 andp = 0.0006 by Fisher’s exact test, respectively). Thepattern of immunohistochemical staining for ERα isdemonstrated in Figure 3. When considering all tu-mor masses, the ERα and PR percentage of positivecells was lower in soy-treated tumors than in controls.However, when negative tumors were excluded fromcalculation no similar differences were found (data notshown).

Body weight, food consumption and organ weight

A reduction in body weight gain was observed inanimals given 0.7% dietary soy when compared to

Soy activity in chemically-induced mammary cancer 157

Figure 1. (A) Effects of dietary soy on percent incidence of mammary tumors, (B) number of mammary tumors/tumor-bearing rat and (C)median total tumor burden/tumor-bearing rat, in Sprague Dawley females given a single oral dose of DMBA. The rat treatment groups were-O-, DMBA and control diet; -∗-, DMBA and 0.35% dietary soy; -�-, DMBA and 0.7% dietary soy. Animals were fed control or soy addeddiet from weaning up to the end of the study (24 weeks after DMBA administration).∗ p < 0.05 vs. control, Dunn’s test.

158 D Gallo et al.

Table 1. Degree of differentiation of DMBA-induced mammarytumours in female Sprague-Dawley rats

Total no. of Degree of tumor differentiation

tumor masses (% of masses)

Well Moderately Poorly

Control 87 32.2 25.3 42.5

Soy 0.35% 59 47.5 20.3 32.2

Soy 0.7% 65 61.5∗∗ 18.5 20∗

∗ =p < 0.01.∗∗ = p < 0.001 vs. control, Fisher’s exact test.

controls (data not shown). This difference never ex-ceeded 7%. A tendency towards decreased food in-take was also seen in the same animals (data notshown). Organ weight data (uterus, ovaries and pitu-itary weights) did not show any relevant differences incontrol and treated groups (data not shown).

Isoflavone plasma levels

Results obtained from HPLC analysis showed that nodetectable levels of isoflavones were present in plasmasamples of control animals fed the phytoestrogen-freediet (Figures 4(A) and (B)).

Analysis of plasma samples from treated anim-als revealed a high interindividual variability. Freedaidzein levels were seen to be 0.03 ± 0.01 µM (mean± SE) in both 0.35 and 0.7% soy groups. Free gen-istein levels were 0.01 ± 0.005 and 0.02 ± 0.004µM(mean ± SE) for 0.35 and 0.7% soy group, respect-ively (Figure 4(A)). Under our experimental condi-tions, equol, a daidzein metabolite, was not detectableas free in any treatment group, while it was detectedin all study samples after hydrolysis (Figure 4(B)).Equol plasma levels were seen to be dose-relatedsince they grew from 0.94 ± 0.26 µM (mean ± SE) inanimals fed 0.35% soy to 1.98 ± 0.54 µM in anim-als fed 0.7% dietary soy, this difference approachingstatistical significance (p = 0.07).

When considering total daidzein and genisteinplasma concentration, no clear dose-response effectwas observed at the end of the study in treated an-imals for the two doses used. Daidzein levels were0.44 ± 0.09 and 0.61 ± 0.14 µM and genistein levelswere 0.77 ± 0.16 and 0.5 ± 0.12 µM (mean ± SE)for animals fed 0.35 and 0.7% dietary soy, respect-ively. As shown, most of phytoestrogens were asglucuronide and/or sulfate conjugate (93–98%), thepercentage of conjugated fraction being similar in thetreatment groups.

Discussion

The present study was aimed at investigating whethera dietary supplement of the standardized soy extract(SOYSELECTTM), administered since peripubertalage, can influence chemically-induced mammary car-cinogenesis in female rats. We could not see anystatistically significant difference in tumor incidenceand tumor multiplicity at the end of the study, althoughsoy-treated rats exhibited lower total tumor burden/ratwith respect to controls. This difference can not bemerely related to the reduction in body weight gainobserved in animals fed soy diet since the depressionseen never exceed 7%, a value which is generallyaccepted not to be able to bias the interpretation ofresults obtained in chemoprevention studies [20].

Importantly, we showed, for the first time, thatsoy-treated rats exhibited a decrease in the percentof poorly differentiated tumors and a concomitant in-crease in the percent of the well-differentiated ones.These data suggest that dietary soy supplement canpromote the maintenance of a less aggressive state andpossibly justify the slower rate of tumor enlargement(i.e., the lower total tumor burden in soy-treated rats).These observations emphasize the need to consideradditional qualitative parameters in the evaluation ofresults from chemopreventive studies, since not onlythe crude number of tumors but, possibly more im-portantly, the morphological and biological character-istics of the occurring tumors, should be taken intoaccount.

We also showed that the percentage of ERα pos-itive tumors was significantly lower in animals re-ceiving 0.7% dietary soy with respect to controls, inkeeping with other reports which have shown, both invivo and in vitro, that prolonged exposure to phytoes-trogen results in a decrease in estrogen receptor levelsin normal tissues [21] as well as in breast cancercells [22–25]. Therefore, it is conceivable that pro-longed exposure to soy extract, by reducing the ERexpression, leads to decreased responsiveness to endo-genous estrogens, as also confirmed by the reductionin the expression of PR, which is a classical estrogen-dependent protein. It has been reported that genisteinin vitro can up-regulate PR levels in breast cancer cells[25], however, the different experimental conditions,including the different method of receptor determina-tion, make any comparison with our in vivo data quitedifficult.

Our data, suggesting that in rats soy feeding res-ulted in more differentiated ER negative tumors, are

Soyactivity

inchem

ically-inducedm

amm

arycancer

159

Figure 2. Representative morphological pattern of (A) an undifferentiated mammary tumor (solid pattern) from a control animal and (B) a well-differentiated one (tubular pattern) from asoy-fed animal (HE). In the insert the same fields are shown at lower magnification.

160 D Gallo et al.

Table 2. Estrogen/progesterone receptor levels of DMBA-induced mammary tumours in female Sprague-Dawley rats

No. of tumor No. (%) Median Range p value

masses examined + ve masses % + ve cells % + ve cells (Fisher)

ERα

Control 38 25 (65.8) 10 0–60

0.02

Soy 0.7% 38 14 (36.8) 0 0–60

PR

Control 38 13 (34.2) 0 0–20

0.0006

Soy 0.7% 38 1 (2.6) 0 0–10

Evaluation was performed on 38 tumours from both control and 0.7% soy-treated animals.

barely comparable to those reported in literature show-ing the association between a more differentiatedgrade and ER positive status in human breast can-cer [26]. Unfortunately, no data are available thatcould help explaining this discrepancy. Moreover, thenatural history of spontaneously occurring tumors inhumans can be hardly comparable to tumors exposedfor prolonged periods to a pharmacological modu-lation. In this context, it remains to be clarifiedwhether the morphological features of a better dif-ferentiated status associated with soy exposure couldhave a biological basis in terms of expression ofspecific biochemical parameters. To this end, the sim-ultaneous evaluation of a panel of biological markersof proliferation (PCNA), cellular apoptosis (caspasecleaved products), angiogenesis (CD31) or hormoneindependence (neu) is ongoing in our laboratory.

Besides the direct effect of genistein on the de-crease of ER biosynthesis, other possible mechan-isms modulating the estrogenic pathways have to betaken into account. The final resulting in vivo activ-ity of phytoestrogens contained in soy extract may beheavily conditioned by estrogen hormonal changes oc-curring during animal maturation. In immature rats,which have very low levels of endogenous estrogens,phytoestrogens would behave as estrogen agonists, assuggested by Lamartiniere et al. [10], who reportedthat the administration of genistein during neonatalperiod was associated to alterations of the endocrinesystem similar to those seen in rats exposed neonatallyto estrogens [27]. Accordingly, the growth of the ERpositive MCF-7 human breast cancer cells, cultured ina medium containing dextran-coated stripped serum,was shown to be promoted by genistein concentrationsup to 10 µM [23].

Conversely, in postpubertal animals with adultlevels of endogenous estrogens, phytoestrogens, atphysiological achievable concentrations, are morelikely to behave as estrogen antagonists. Barnes et al.[22] showed that dietary soy, administered since 25days of age, inhibited DMBA-induced mammarycancer in rats, an effect supposed to be the resultof an antiestrogenic action. Also in vitro data byMiodini et al. [25] showed that genistein can sig-nificantly counteract the estrogen-promoted growthstimulation of MCF-7 cells, suggesting that this effectcan be mediated by a direct interaction with the ERbinding site, even though alternative mechanisms werenot ruled out.

Finally, different behavior can be exhibited byphytoestrogens according to the achievable concentra-tions: at very high concentrations (>10 µM) genisteininhibits the proliferation of both MCF-7 as well as theER-negative MDA-MB-231 cell line, thus indicatingthat, at pharmacological doses, the antiproliferativeactivity occurs independently of ER [23]. This is notsurprising considering that, at high doses, phytoes-trogens can affect a very wide range of signal trans-duction pathways and enzymatic activities involved inthe energy metabolism and other biologically relevantcellular processes [28–33].

This picture seems to be even more complex invivo: we previously reported that dietary adminis-tration of soy extract at doses which give plasmalevels 10-fold higher than those obtained in the presentstudy, can produce estrogen-like effects in estrogen-dependent tissue and parameters [34, unpublisheddata]. Evidences have also been reported that indifferent animal species prolonged exposure to diet-ary phytoestrogens may negatively affect reproduc-

Soyactivity

inchem

ically-inducedm

amm

arycancer

161

Figure 3. Immunohistochemical staining for ERα showing (A) numerous positively stained (brown) cells in a tumor from a control animal and (B) absence of stained cells in a tumor from asoy-fed animal.

162 D Gallo et al.

Figure 4. Free (A) and total (B) isoflavone plasma levels (µM) detected at the end of the study. Animals were fed control or soy added dietfrom weaning up to the end of the study (24 weeks after DMBA administration). Equol was not detectable as free in any treatment group.Daidzein; � Genistein. � Equol.

tive physiology [35] as result of their estradiol-likeaction.

Taken together, all these findings support theconcept that the potential importance of phytoestro-gens in human health and disease is certain to deriveboth from their mimicry of steroidal estrogens andtheir failure to closely mimic the action of steroidalestrogens. The final effect observed in vivo will de-pend not only on their concentration but, also, on thebiological effect considered.

It is conceivable that this complex picture will takeadvantage of the recent acquisitions on the existenceof two different ER (ERα and ERβ) [36] and on thehigher affinity of genistein for ERβ with respect toERα [37]; the simultaneous assessment of the statusof both receptor subtypes at the target tissues, togetherwith a more complete understanding of the preciserole of ERβ will help to gain insight into mechanismsunderlying the biological effects of phytoestrogens.

In the present study the extract at 0.35 and 0.7%in the diet gave final levels of approximately 420and 840 mg isoflavones/kg diet, respectively, beingthe genistein/daidzein ratio in the diet of 1.2. As-suming a 200 g rat and an average food intake of12 g/rat/day, the estimated isoflavones intake was ap-proximately 25 and 50 mg/kg/day for animals fed0.35% and 0.7% diet, respectively. Japanese con-sumption of isoflavones from soy products has beenestimated at up to about 3 mg/kg/day. Despite the rel-atively higher phytoestrogen intake of rat compared tohumans, interestingly, in the present study, genisteinand daidzein concentrations in rat serum were foundto be within the range of those detected in people con-suming large amounts of soyfoods [16]. On the otherhand, treated animals had equol plasma levels sub-stantially higher than those reported in humans [16].In animals (as well as in humans), equol is formedin the gastrointestinal tract as the result of bacterial

Soy activity in chemically-induced mammary cancer 163

degradation of the ingested daidzein and our resultsshow that it is the major phenolic compound foundin blood of rats maintained on a soy diet. Relativeto physiological estrogen, the biological potencies ofgenistein and equol are similar, whereas daidzein hasbeen shown to be somewhat less potent than genistein.In view of this, synergistic or at least additive effectsbetween genistein and daidzein/equol, should be takeninto account.

In conclusion, our research has shown that di-etary administration from weaning of the commer-cially available, clinically tested [15], soy extractSOYSELECTTM to rats has no major effects onDMBA-induced tumors, in terms of incidence, mul-tiplicity or tumor burden, but it can probably modifythe histomorphological characteristic of the occurringtumors which express a higher grade of differenti-ation. The treatment schedule of soy administrationwe studied could provide useful information from aclinical point of view as it was chosen to mimic apotential modification of dietary habit in women fromperipubertal age onward.

References

1. Tominaga S: Cancer incidence in Japanese in Japan, Hawaiiand western United States. NCI Monograph 69: 83–92, 1985

2. Haenszel W, Kurihara M: Studies of Japanese immigrants. I.Mortality from cancer and other diseases among Japanese inthe US. J Natl Cancer Inst 40: 43–69, 1968

3. Buell P: Changing incidence of breast cancer in Japanese-American women. J Natl Cancer Inst 51: 1479–1483, 1973

4. Ziegler RG, Hoover RN, Hildeshein RN, Nomura MY, PikeMC, West DW, Wu-Williams A, Kolonel LN, Horn-Ross PL,Rosenthal JF, Hyer MB: Migration patterns and breast cancerrisk in Asian-American women. J Natl Cancer Inst 85: 1819–1827, 1993

5. Ingram D, Sanders K, Kolybaba M, Lopez D: Case-controlstudy of phyto-oestrogen and breast cancer. Lancet, 350: 990–994, 1997

6. Lee HP, Gourley L, Duffy SW, Esteve J, Lee J, Day NE: Di-etary effects on breast cancer risk in Singapore. Lancet 337:1197–1200, 1991

7. Yuan JM, Wang QS, Ross RK, Henderson BE, YU MC: Dietand breast cancer in Shanghai and Tianjin, China. Br J Cancer71: 1353–1358, 1995

8. Messina MJ, Persky V, Setchell KDR, Barnes S: Soy intakeand cancer risk: a review of the in vitro and in vivo data. NutrCancer 21: 113–131, 1994

9. Setchell KDR: Absorption and metabolism of soy isoflavones-from food to dietary supplements and adults to infants. J Nutr130: 654S–655S, 2000

10. Lamartiniere CA, Moore JB, Brown NM, Thompson R,Hardin MJ, Barnes S: Genistein suppresses mammary cancerin rats. Carcinogenesis 16: 2833–2840, 1995

11. Murrill WB, Brown NM, Zhang JX, Manzolillo PA, Barnes S,Lamartiniere CA: Prepubertal genistein exposure suppresses

mammary cancer and enhances gland differentiation in rats.Carcinogenesis 17: 1451–1457, 1996

12. Fritz WA, Coward L, Wang J, Lamartiniere CA: Dietary gen-istein: perinatal mammary cancer prevention, bioavailabilityand toxicity testing in the rat. Carcinogenesis 19: 2151–21581998

13. Hilakivi-Clarke L, Cho E, Onojafe I, Raygada M, Clarke R:Maternal exposure to genistein during pregnancy increasescarcinogen-induced mammary tumorigenesis in female ratoffspring. Oncol Rep 6: 1089–1095, 1999

14. Yang J, Nakagawa H, Tsuta K, Tsubura A: Influence of peri-natal genistein exposure on the development of MNU-inducedmammary carcinoma in female Sprague-Dawley rats. CancerLett 149: 171–179, 2000

15. Scambia G, Mango D, Signorile PG, Anselmi RA, Palena C,Gallo D, Bombardelli E, Morazzoni P, Riva A, Mancuso S:Clinical effects of a standardized soy extract in postmeno-pausal women: a pilot study. Menopause 7: 105–111, 2000

16. Adlercreutz H, Markkanen H, Watanabe S: Plasma concen-trations of phyto-oestrogens in Japanese men. Lancet 342,1209–1210, 1993

17. Corbett T, Valeriote F, LoRusso P, Polin L, Panchapor C,Pugh S, White K, Knight J, Demchik L, Jones J, Jones L,Lisow L: In vivo methods for screening and preclinical testing.In: Teicher BA (ed) Anticancer Drug Development Guide.Humana Press, Totowa, New Jersey, 1997, pp 75–99

18. International classification of rodent tumors. Part I: The Rat.5. Integumentary System. Mohr U (ed) IARC Scientific Pub-lication, No.122, Lyon, France, 1993

19. Supko JG, Phillips LR: High-performance liquid chroma-tographic assay for genistein in biological fluids. J ChromBiomed Appl 666: 157–167, 1995

20. Mc Gregor D: Tumour incidence reductions in carcinogenicitybioassays. In: Stewart BW, Mc Gregor D, Kleihues P (eds)Principles of Chemoprevention. IARC Scientific Publication,No. 139, Lyon, 1996, pp 277–290

21. Cotroneo MS, Lamartiniere CA: Estrogen receptor modula-tion by genistein in the uterus. In: Beck WT, Colvin OM,Matrisian LM, Simone JV (eds) Proceedings of the Amer-ican Association for Cancer Research. 91st Annual Meeting,Vol 41, April 1–5, 2000, San Francisco, CA, 844 (Abstr no.5362).

22. Barnes S, Grubbs C, Setchell KDR, Carlson J: Soybeans in-hibit tumors in models of breast cancer. Mutagens CarcinoDiet 239–253, 1990

23. Wang TTY, Sathyamoorthy N, Phang JM: Molecular effectsof genistein on estrogen receptor mediated pathways. Carci-nogenesis, 17: 271–275, 1996

24. Cao S, Hudnall SD, Lu LJW: Downregulation of estrogenreceptor by phytoestrogens daidzein and genistein in MCF-7breast cancer cells. In: Beck WT, Colvin OM, Matrisian LM,Simone JV (eds) Proceedings of the American Association forCancer Research. 91st Annual Meeting, Vol 41, April 1–5,2000, San Francisco, CA, p 310 (Abstr no. 1970)

25. Miodini P, Fioravanti L, Di Fronzo G, Cappelletti V: Thetwo phyto-oestrogens genistein and quercetin exert differenteffects on oestrogen receptor function. Br J Cancer 80: 1150–1155, 1999

26. Osborne CK: Receptors. In: Harris JR, Hellman S, HendersonIC, Kinne DW (eds) Breast Diseases, 2nd edn, JB LippincottCompany, Philadelphia, Pennsylvania, 1991, p 301

27. Lamartiniere CA: Neonatal estrogen treatment alters sexualdifferentiation of hepatic histidase. Endocrinology 105: 1031–1035, 1979

164 D Gallo et al.

28. Akiyama T, Ishida J, Nakagawa S, Ogawara H, Watanabe S,Itoh NM, Shibuya M, Fukami Y: Genistein, a specific inhibitorof tyrosine-specific protein kinases. J Biol Chem 262: 5592–5595, 1987

29. Markovits J, Linassier C, Fosse P, Couprie J, Pierre J,Jacquemin-Sablon A, Saucier JM, Le Pecq JB, Larson AK:Inhibitory effects of the tyrosine kinase inhibitor genistein onmammalian DNA topoisomerase II. Cancer Res 49, 5111–5117, 1989

30. Matsukawa Y, Marui N, Sakai T, Satomi Y, Yoshida M,Matsumoto K, Nishino H, Aoike A: Genistein arrests cellcycle progression at G2-M. Cancer Res 53: 1328–1331, 1993

31. Fotsis T, Pepper M, Adlercreutz H, Fleischmann G, Hase T,Montesano R, Schweigerer L: Genistein, a dietary-derived in-hibitor of in vitro angiogenesis. Proc Natl Acad Sci USA 90,2690–2694, 1993

32. Wei H, Bowen R, Cai Q, Barnes S, Wang Y: Antioxidant andantipromotional effects of the soybean isoflavone genistein.Proc Soc Exp Biol Med 208: 124–130, 1995

33. Kim H, Peterson TG, Barnes S: Mechanisms of action of thesoy isoflavone genistein: emerging role for its effects via trans-forming growth factor β signaling pathways. Am J Clin Nutr68(S): 1418–1425S, 1998

34. Gallo D, Cantelmo F, Distefano M, Ferlini C, Zannoni G, RivaA, Morazzoni P, Bombardelli E, Mancuso S, Scambia G: Re-productive effects of dietary soy in female Wistar rats. FoodChem Toxicol 37: 493–502, 1999

35. Hughes CL Jr: Phytochemical mimicry of reproductive hor-mones and modulation of herbivore fertility by phyotestro-gens. Environ Health Perspect 78: 171–175, 1988

36. Kuiper GGJM, Enmark E, Pelto-Huikko M, Nilsson S, Gust-afsson JA: Cloning of a novel estrogen receptor expressed inrat prostate and ovary. Proc Natl Acad Sci 93: 5925–5930,1996

37. Kuiper GGJM, Carlsson B, Grandien K, Enmark E,Haggblad J, Nilsson S, Gustafsson JA: Comparison ofthe ligand binding specificity and transcript tissue distributionof estrogen receptors α and β. Endocrinology 138: 863–870,1997

Address for offprints and correspondence: Prof. Giovanni Scam-bia, Department of Obstetrics and Gynecology, Catholic Univer-sity of the Sacred Heart, Lgo A. Gemelli, 8 - 00168, Rome,Italy; Tel.: 06/35508736; Fax: 06/35508736, 06/3051160; E-mail:giovanni.scambia@libero.it