Food and Chemical Toxicology 50 (2012) 1405–1412

Contents lists available at SciVerse ScienceDirect

Food and Chemical Toxicology

journal homepage: www.elsevier .com/locate / foodchemtox

Bowman-Birk inhibitors, proteasome peptidase activities and colorectalpre neoplasias induced by 1,2-dimethylhydrazine in Swiss mice

Alessandra de Paula Carli a,d, Paula Melo de Abreu Vieira a, Karina Taciana Santos Silva a,Renata Guerra de Sá Cota a,b, Cláudia Martins Carneiro a,c, William Castro-Borges a,b, Milton HérculesGuerra de Andrade a,b,⇑a Núcleo de Pesquisas em Ciências Biológicas, Universidade Federal de Ouro Preto, Ouro Preto, Brazilb Departamento de Ciências Biológicas, Universidade Federal de Ouro Preto, Ouro Preto, Brazilc Departamento de Análises Clínicas, Escola de Farmácia, Universidade Federal de Ouro Preto, Brazild Centro Universitário de Caratinga, Caratinga, Brazil

a r t i c l e i n f o a b s t r a c t

Article history:Received 30 May 2011Accepted 19 January 2012Available online 2 February 2012

Keywords:Colon cancerBowman-Birk inhibitorsProteasome activityDimethylhydrazineCD44

0278-6915/$ - see front matter � 2012 Elsevier Ltd. Adoi:10.1016/j.fct.2012.01.036

Abbreviations: BBIs, Bowman-Birk inhibitors; DMMCA, 7-amide-4-methylcoumarin.⇑ Corresponding author. Address: Instituto de C

Departamento de Ciências Biológicas, Núcleo de PesqUniversidade Federal de Ouro Preto, Campus MorroOuro Preto, MG, Brazil. Tel.: +55 31 3559 1705; fax: +

E-mail addresses: alessandrapcarli@hotmail.com (peb.ufop.br (P.M. de Abreu Vieira), karinasilva353@rguerrasa@gmail.com (R.G. de Sá Cota), carneirocmwilliamcborges@hotmail.com (W. Castro-Borges), mgde Andrade).

Bowman-Birk inhibitors (BBIs) are protein molecules containing two inhibitory domains for enzymessimilar to trypsin and chymotrypsin. Interest in these inhibitors arose from their properties against thecancer chemically induced by 1,2-dimethylhydrazine (DMH). In this study the effect of two BBI prepara-tions (from Glycine max and Macrotyloma axillare) were evaluated for the prevention of colorectal neopla-sia induced by intraperitoneal injections of DMH, given at a dose of 30 mg/kg, during 12 weeks. Micetreated with DMH presented histopathological alterations consistent with tumor development, aug-mented CD44 expression and increased proteasome peptidase activities. Lysosomal fractions, obtainedfrom the intestines, were chromatographed in a Sepharose-BBI column and increased activity for trypsinand chymotrypsin-like proteases recovered from DMH-treated animals. In parallel, mice treated for eightweeks with BBIs showed a decrease in the chymotrypsin and trypsin-like proteasome activities comparedto animals fed on normal diet. For the groups receiving simultaneous treatment with DMH and BBIs,dysplasic lesions were not observed and proteasome peptidase activities were similar to the controlgroup after the 24th week. These results suggest that the mechanism by which BBIs could prevent theappearance of pre neoplastic lesions is associated with inhibition of both the lysosomal and protea-some-dependent proteolytic pathways.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Advances in industrialization generate an increasing frequencyof exposure to new chemicals that are potentially harmful to cellsand hence capable of inducing cancer. Active substances known tobe important for the prevention of colorectal cancer are present ina variety of plants (Forman et al., 2004). In this regard, soybean hasbeen considered an important food source containing chemopre-

ll rights reserved.

H, N,N0-dimethylhydrazine;

iências Exatas e Biológicas,uisas em Ciências Biológicas,do Cruzeiro, s/n, 35400000

55 31 3559 1680.A. de Paula Carli), paula@nuyahoo.com.br (K.T.S. Silva),

@gmail.com (C.M. Carneiro),uerra@nupeb.ufop.br (M.H.G.

ventive compounds (Losso, 2008). Earlier studies indicated theBowman-Birk inhibitor (BBI/Mr � 8 kDa) which contains two dis-tinct inhibitory domains for enzymes similar to trypsin and chy-motrypsin as the main soy component endowed with anticanceractivity (Deshimaru et al., 2004; Odani and Ikenaka, 1973). Inparticular, the antichymotrypsin inhibitory activity of BBIs hasbeen correlated to their anticarcinogenic activity (Kennedy et al.,2008). Nevertheless, it has been demonstrated that other soyconstituents are known to contribute to this effect such as lunasinand a variety of small organic components (Gomes et al., 2005;Hernandez-Ledesma et al., 2009; Kris-Etherton et al., 2002).

The 20/26S proteasome complex is currently considered animportant intracellular target for antitumor drugs (Montagutet al., 2006). Although the proteasome is classified as a threonineprotease, its proteolytic core contains catalytic subunits exhibitingcaspase (b1), trypsin (b2) and chymotrypsin (b5)-like activities(Tanaka, 2009). Bortezomib is the first proteasome inhibitor ap-proved by the FDA (Food and Drug Administration), known to beeffective in some hematological malignancies for treatment of

1406 A. de Paula Carli et al. / Food and Chemical Toxicology 50 (2012) 1405–1412

multiple myeloma (Cavo, 2006; Chen et al., 2011). It was demon-strated that the use of proteasome inhibitors results in multiplecellular effects, including promotion of anti-inflammatory, anti-carcinogenic, anti-proliferative and apoptotic effects in cells exhib-iting abnormal development (Mitsiades et al., 2002; Myung et al.,2001). Other studies demonstrated that BBIs specifically inhibitedthe chymotrypsin-like proteasomal activity in MCF7 tumor cellscausing accumulation of ubiquitinated substrates and a decreasein the levels of regulatory cyclins (Chen et al., 2005).

The present study refers to an evaluation of the protective effectsof BBIs isolated from Glycine max and Macrotyloma axillare duringdevelopment of colorectal tumors caused by administration of thecarcinogen 1,2-dimethylhydrazine to mice. In addition to displaytrypsin and chymotrypsin inhibitory properties M. axillare BBIsshare relevant primary and 3D structural features with soybeanBBI, such as the conserved cysteine residues involved in disulfidebond formation (Kumar et al., 2002). However, specific differencesobserved in amino acid composition may account for some particu-lar pharmacokinetic properties of M. axillare BBIs, including a higherdistribution to stomach (Santana et al., 2011), most likely due to ahigher interaction followed by enhanced internalization in the gas-tric tissue. In the present study the protective effects of these twoBBI preparations were assessed through histopathological examina-tions, evaluation of the proteasome peptidase activities, CD44 pro-tein expression by Western blotting and evaluation of the enzymeactivities retained in a Sepharose-BBI affinity column.

2. Materials and methods

2.1. Ethics statement, experimental animals and reagents

Swiss male mice aged 120 days were obtained from the Center for AnimalScience, Universidade Federal de Ouro Preto. All procedures involving animals werecarried out in accordance with the national guidelines provided by the local ethicscommittee. The animals were maintained under a light/dark cycle of 12 h and givenfood and water ad libitum. N,N0-Dimethylhydrazine (DMH) 99% and fluorogenic sub-strates Boc-Val-Pro-Arg-MCA (amide-4-methylcoumarin) and Suc-Ala-Ala-Pro-Phe-MCA were from Sigma (Sigma–Aldrich, St. Louis, USA). Purified anti-mouse CD44was purchased from BD Biosciences Pharmingen™ (BD Biosciences PharMingen,New Jersey, USA). The fluorogenic substrates, Cbz-Gly-Gly-Arg-MCA and Suc-Leu-Leu-Val-Tyr-MCA were purchased from Biomol International (Biomol, Exeter, UK).

2.2. Experimental groups

The group of animals named BBIS received a diet supplemented with 0.1% w/wenriched soybean (G. max) BBI, produced according to methods previously de-scribed (Yavelow et al., 1985). BBIM group refers to animals treated with a prepa-ration of enriched BBIs from the leguminous M. axillare, produced according toour patented isolation protocol available at http://www.patentesonline.com.br/pro-cesso-de-preparacao-de-extrato-ativado-de-bowman-birk-de-macrotyloma-axil-lare-115536.html. In the last case, the dose added to the diet showed antitrypsinactivity equivalent to that contained in 0.1% w/w soybean BBI. The homogeneityof the two BBI preparations was assessed using 15% SDS–PAGE. For the inductionpre neoplastic lesions and neoplasia, mice were subjected to intraperitoneal injec-tions of DMH weekly, at a dose of 30 mg/kg in citrate buffer/EDTA (1 mM EDTA,10 mM sodium citrate, 0.9% NaCl, pH 8.0) during 12 weeks. Injection volume was150 lL. Animals were maintained for a latency period of 12 additional weeks.Overall, mice were divided into six treatment groups (n = 18 animals/group) as fol-lows – Control group (citrate/EDTA), DMH group (DMH), DMHS group (DMH anddiet supplemented with BBI from G. max) and DMHM group (DMH and diet supple-mented with BBI from M. axillare). Two additional control groups were given a dietsupplemented with either 0.1% G. max BBI (BBIS) or M. axillare BBIs (BBIM) for8 weeks.

2.3. Histopathological analyses

Representative animals of the various treatment groups had their colon andintestine removed during necropsy for histopathological analyses. The organs werefixed in buffered formalin and embedded in paraffin. Histological sections of 5 lmwere obtained using a microtome. These were fixed on slides followed by stainingwith a Hematoxylin Eosin preparation. Qualitatively evaluated parameters were:

presence/absence of atypical cells either present in the mucous, middle or serosalayer; mucosa inflammation; necrosis; atypical structures and hyperplasia of thelymphoid nodules.

2.4. In gel proteolytic digestion and peptidase activities of the fraction purified from aSepharose-BBI column

Protein samples from the small intestine and colon of each experimental groupwere subjected to affinity chromatography using a Sepharose column containingimmobilized M. axillare BBI. Coupling of purified M. axillare BBI to the Sepharosematrix was performed as previously described (Matsumoto et al., 1981). Homoge-nates obtained from 100 mg of colon and intestine tissues were prepared in PBSpH 7.4 containing 1 mM MgCl2 using a Branson 250 sonifier. After centrifugationat 1000g for 10 min, to eliminate cell/tissue debris, the supernatant was removedand recentrifuged at 10,000g for 20 min to obtain a soluble and a pellet fraction.Lysosomes were further enriched from the pellet fraction following the protocolas described elsewhere (Billings et al., 1988). The lysosome-enriched pellet fractionwas resuspended in 50 mM phosphate buffer pH 7.0 containing 1 mM MgCl2 and0.1% Triton X-100. Lysosomal proteins were extracted by five cycles of sonicationfor 30 s with 1 min intervals on ice. After centrifugation at 10,000g for 5 min, thesupernatant containing lysosomal proteins was obtained. The lysosomal and solu-ble fractions were separately loaded onto the Sepharose-BBI column. The columnwas extensively washed with 50 mM Tris–HCl pH 7.0, 1 mM MgCl2, 500 mM NaCland the bound fraction obtained by decreasing the pH to 1 using 100 mM HCl. Pro-teolytic activity present in this fraction was demonstrated by 12% PAGE with the gelmatrix containing 0.1% gelatin as described by (Billings et al., 1991). In parallel, pep-tidase activities related to trypsin and chymotrypsin were evaluated by hydrolysisof the fluorogenic substrates Boc-Val-Pro-Arg-MCA and Suc-Ala-Ala-Pro-Phe-MCA,respectively. The assays were conducted using an aliquot of 40 lL of each fractionand 1.6 mM of the substrates in a final volume of 240 lL. The hydrolysis reactionoccurred for 30 min at 37 �C and stopped by addition of 3 mL 50 mM sodium phos-phate, pH 8.0. Fluorometric readings were obtained at the wavelengths of 380 nm(excitation) and 460 nm (emission) using the RF-5301PC fluorometer (Shimadzu,Japan). Results were expressed as fluorescence arbitrary units/lg protein appliedto the column.

2.5. Assessment of proteasome peptidase activities

Approximately 100 mg of the small intestine and colon of animals from the dif-ferent experimental groups were homogenized in 3 mL of 20S buffer (25 mM Tris–HCl pH 7.5, 1 mM DTT, 10% glycerol). Protein extraction was first performed using atissue homogenizer (Potter Elvejhem) followed by sonication under three cycles of40 and 15 s intervals in an ice bath. Samples were centrifuged for 30 min at10,000g and the supernatant submitted to ultracentrifugation for 6 h at 100,000g.One mL fractions taken from the bottom of the tubes were used to assess the pro-teasome’s trypsin- and chymotrypsin-like peptidase activities, using the respectivefluorogenic substrates Cbz-Gly-Gly-Arg-7-amide-4-methylcoumarin and Suc-Leu-Leu-Val-Tyr-7-amide-4-methylcoumarin. The reactions were performed using15 lg total protein and 13 lM of fluorogenic substrates for a final volume of240 lL, in 50 mM Tris–HCl pH 8.0 containing 10 mM MgCl2. The chymotrypsin-likeassay was also performed in the presence of 20 lM MG132, a proteasome inhibitor,to control for proteasome-specific activity. Fluorometric readings were performedat wavelengths of 380 nm (excitation) and 460 nm (emission), and resultsexpressed as fluorescence arbitrary units/lg protein.

2.6. Analysis of CD44 expression by Western blotting

Approximately 100 mg of the colon of animals from the different experimentalgroups were used for preparing protein extracts in 1 mL of 50 mM Tris–HCl pH 7.5,1 mM DTT, 10% glycerol in the presence of 1� PIC (Protease Inhibitor Cocktail, Sig-ma) through sonication, as described in Section 2.5. The protein homogenate wasthen centrifuged at 10,000g for 20 min and the supernatant stored at �80 �C. Onemillilitre of a buffered solution (100 mM Tris–HCl, pH 7.5, 0.1% SDS, 1% Triton X-100) was used to resuspend the remaining pellet. The membrane suspension wasextracted by agitation through vortexing, the preparation boiled for 5 min and cen-trifuged at 10,000g for 10 min. The integrity of the proteins extracted from themembrane fractions was assessed using 12% SDS–PAGE and the protein contentestimated by densitometric analysis. Fifteen micrograms of each of the fractionswere then separated using 12% SDS–PAGE and immediately transferred to a PVDFmembrane essentially as described by (Towbin et al., 1979). After a blocking stepof the membrane in TBST (50 mM Tris–HCl pH 8.3, 150 mM NaCl, 0.05% Tween-20) containing 5% non-fat milk for 12 h, the membrane was incubated with therat anti-mouse CD44 monoclonal antibody at a dilution of 1:1000, in the same buf-fer. Incubation was performed for 3 h at room temperature. Visualization of theCD44 protein isoforms was possible after incubation of the membrane with alkalinephosphatase-labelled anti rat IgG antibody in TBST/milk for 1.5 h, followed byaddition of the enzyme substrates nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate, used according to the manufacturer’s instructions.

A. de Paula Carli et al. / Food and Chemical Toxicology 50 (2012) 1405–1412 1407

2.7. Statistical analysis

Statistical analysis was performed by means of the Prism software, using analy-sis of variance between groups (ANOVA), which guarantees normal distribution ofdata and the complementary Tukey’s test, considering a significance level ofp 6 0.05.

3. Results

3.1. BBIs from M. axillare and G. max protect against the developmentof pre neoplastic lesions induced by DMH

The percentage of animals that presented macroscopic andmicroscopic alterations in the colon and small intestine from thedifferent experimental groups is indicated in Table 1. It was ob-served that 45% of the mice that received DMH presented macro-scopic alterations in the colon and small intestine. With respect tomicroscopic alterations, dysplastic and neoplastic lesions were ob-served in 56% and 33% of the colon and small intestine, respectively,only in animals under DMH-treatment. Fig. 1A shows a macroscopiclesion corresponding to a polyp formation in the colon of DMH-trea-ted animals. These were found distributed throughout the intestine.Fig. 1B reveals the histologic aspects of the polipomatous lesion andits insertion in the mucosa, squared area. Fig. 1D demonstratesarchitectural details of the proliferative lesion from DMH-treatedanimals and its associated-inflammatory response. No tissue alter-ations were found for animals in the remaining groups (DMHS andDMHM), attesting for the protective effect of diet supplementationwith BBIs. At the employed dosage regimen BBI from the twosources were equally effective for the prevention of pre neoplasticlesions with no significant differences in the parameters used toevaluate the small intestine and colon. Nevertheless, specific inhib-itory activity assays comparing the properties of the main BBI iso-forms present in these two preparations and isolated according to(César et al., 2009) revealed BBI from M. axillare being 33% and69% more active against bovine trypsin and chymotrypsin, respec-tively, compared to G. max BBI – Supplementary Table 1.

It’s worth emphasizing that a severe inflammatory responsewas observed in the colon and small intestine of animals in theDMH-treated group. A representative result from the colon is pre-sented in Fig. 2 showing the inflammatory infiltrate in the musclelayer in animals of the DMH group (Fig. 2B) and the absence ofsuch response in animals from DMHS and DMHM groups (Figs.2C and D). Altogether these results revealed the protective effectsof our two BBI preparations for DMH-induced inflammation anddevelopment of pre neoplastic lesions.

3.2. DMH-treatment induces increase in trypsin- and chymotrypsin-like activities in the lysosomal fraction

Lysosomal fractions from intestine and colon of the various ani-mal groups were used to evaluate their ability to hydrolyze fluoro-genic substrates to trypsin and chymotrypsin-like enzymes, afterthe 24th week of experimentation. Prior to the enzymatic assays,the lysosomal fractions were submitted to affinity chromatographyusing a Sepharose-BBI column, to recover proteolytic activity

Table 1Percentage and number of animals that presented macroscopic and microscopic alteration

Groups Macroscopic alterations Dysplasia and p

Small intestine (%) Colon (%) Small intestine

Control 0(0/6) 0(0/6) 0(0/6)DMH 45(4/9) 45(4/9) 56(5/9)DMHS 0(0/6) 0(0/6) 0(0/6)DMHM 0(0/6) 0(0/6) 0(0/6)

related to trypsin and chymotrypsin enzymes. Fig. 3 reveals thatboth activities are increased in the intestine and colon of animalsfrom the DMH group. The low trypsin- and chymotrypsin-likeactivities in the lysosomal fractions derived from the control groupallowed us to establish a correlation between the increase in pro-teolytic activity and the presence of neoplastic lesions. Animalsfrom DMHS and DMHM groups exhibited proteolytic activitiessimilar to the untreated control group.

3.3. CD44 protein expression increases in neoplastic lesions fromDMH-treated animals

Total membrane fractions were obtained from the small intes-tines of control and DMH-treated groups for evaluation of CD44expression levels by Western blotting. Protein samples from threerepresentative animals in each group were separated by SDS–PAGEand the gel image used for quantification of total protein loading.Densitometric analysis revealed no significant difference betweenthe amounts of protein transferred to the PVDF membrane com-paring control and DMH-treated animals. When the membranewas immunoblotted with anti-CD44, it was clearly observed in-creased levels of CD44 expression in the fractions correspondingto membrane extracts from the intestinal polyps induced by treat-ment with DMH, Fig. 4A. The appearance of CD44 positive bandsunique to DMH-treated animals is also indicated by arrows.Fig. 4B corresponds to the densitometric analysis of the blottedmembrane shown in Fig. 4A. Diet supplementation with both BBIpreparations revealed no significant difference in the expressionpattern of CD44, compared to control groups (data not shown).This result provided molecular evidence for the development ofpre neoplastic lesions induced by DMH.

3.4. Treatment with BBIs inhibits the proteasome’s chymotrypsin andtrypsin-like activities

Treatment with either 0.1% w/w G. max BBI or M. axillare BBIs(the latter given at a dose equivalent to the inhibitory activity pres-ent on 0.1% G.max BBI) added to the animal diet for 8 weeks pro-moted decrease in the chymotrypsin- and trypsin-like activitiesof the proteasome extracted from the colon. The assays revealeda reduction of approximately 50% in both activities for animalstreated with both BBI preparations, Fig. 5. Apparently BBI from soy-bean seem to be more effective at inhibiting both the trypsin andchymotrypsin-like proteasome activities but the observed differ-ences were not statistically significant. In addition, aiming to eval-uate the proteasome activities from mice simultaneously treatedwith these inhibitors plus DMH, colon protein extracts from thevarious groups were subjected to proteasome’s activity assays.The results presented in Fig. 6 clearly show a remarkable increasein the trypsin- and chymotrypsin-like activities of the proteasomefor the DMH group. Proteasome peptidase activities were similar tothe control for DMHS and DMHM groups. Use of MG132, a classicalproteasome b5 subunit inhibitor, revealed approximately 80%reduction in the chymotrypsin-like activity; meaning this enzymeactivity was major attributed to proteasome function, Fig. 6.

s in the colon and small intestine from the different experimental groups.

re neoplastic lesions Inflammation

(%) Colon (%) Small intestine (%) Colon (%)

0(0/6) 0(0/6) 0(0/6)33(3/9) 45(4/9) 56(5/9)0(0/6) 0(0/6) 0(0/6)0(0/6) 0(0/6) 0(0/6)

Fig. 1. Colon photomicrographs of DMH-treated animals. (A) Macroscopic lesions inthe colon (polyp). (B) Polipomatous histological appearance of the lesion. Squaredarea represents the insertion of the polyp in the mucosa. (C) Higher magnification ofthe polyp formation. (D) Architectural details of the polyp and its associated-inflammatory response. Hematoxylin Eosin staining. Bar = 50 lm.

1408 A. de Paula Carli et al. / Food and Chemical Toxicology 50 (2012) 1405–1412

4. Discussion

Considering that pre neoplastic lesions precede malignanttransformation our study confirms previous results demonstratingthe ability of soybean BBIs to inhibit carcinogenesis induced byDMH, (Kennedy, 1998; Kennedy et al., 1996; Kennedy et al.,

2002). Experiments aimed at elucidating the mechanism of actionof BBIs, as cancer preventive agents, showed an increased expres-sion of proteolytic enzymes exhibiting trypsin- and chymotryp-sin-like activities. These could be recovered from the lysosomalfraction after affinity chromatography on a Sepharose-BBI column(Billings et al., 1990; St Clair et al., 1990). Therefore, the cell molec-ular targets of BBIs have been primarily centered on lysosomal pro-teases. Despite their unknown nature, zymogram analyses andpeptidase activities demonstrated that BBI protease ligands werein the mass range of 45–66 kDa, as determined by SDS–PAGE. Morerecently, it was demonstrated that purified soybean BBIs inhibitedthe 20S proteasome’s chymotrypsin-like in breast cancer cells(Chen et al., 2005). Reinforcing the proteasome’s inhibition by BBIs,it was reported reduced degradation of ubiquitinated connexin 43in tumor cells treated with BBIs, leading to enhanced cell adhesion(Saito et al., 2007). In addition, specific inhibitors of the protea-some’s chymotrypsin-like activity have been shown to induceapoptosis of cancer cells and suppress tumor growth in severalexperimental models (Chen et al., 2005; Kisselev et al., 2006).

Until now, few studies have demonstrated the use of BBIs aspotential inhibitors of proteasome activity. The present data revealsfor the first time the in vivo effects of two BBI preparations, given asa diet supplement to healthy animals, upon proteasomecapabilities. The eight week’s treatment with either G. max BBI orM. axillare BBIs incorporated in the diet, resulted in a decrease ofapproximately 50% in the trypsin- and chymotrypsin-like activitiesof the proteasome isolated from the colon of mice. This in vivo effectis consistent with both a considerable absorption of these inhibitorsgiven orally, reaching 50% bioavailability (Billings et al., 1991) andthe need for cellular internalization of BBI to gain access to their po-tential targets (Chen et al., 2005). No histopathologic alterationswere observed in the small intestine and colon from animals receiv-ing a diet supplemented with BBI from the two sources. Further-more, the protective effects of these two BBI preparations werecorrelated with maintenance of normal proteasome-dependentproteolytic activity, during simultaneous treatment with DMH.Altogether, our results provide evidence for the suppressive effectof BBIs being correlated with inhibition of at least two intracellularproteolytic systems: the lysosomal and the proteasome-dependentpathway.

Histopathological analyses attested for the protective effect ofBBIs based on the absence of tissue alterations indicative of tumordevelopment. At the molecular level, control levels of CD44 proteinexpression was observed in DMHS and DMHM groups, as opposedto the clearly increased levels of CD44 isoforms in the colon in-duced solely by DMH-treatment. It is important to emphasize thatincreased CD44 protein expression and particularly the appearanceof aberrant CD44 isoforms, due to alternative splicing, is a hallmarkof the cell proteome associated to the development of colontumors (Hynes et al., 2009). Moreover, considering that CD44 areproteoglycans their molecular diversity is also explained by differ-entially attached glycans, adding complexity to their electropho-retic pattern (Arcinas et al., 2009). Particularly, as our primaryantibody does not discriminate among CD44 isoforms we believe,judged by the electrophoretic pattern, we have detected the stan-dard CD44 (CD44s) at approx. 80 kDa and high molecular massCD44 molecules, which should account for other CD44 splice vari-ants (Brown et al., 2011; Godge and Poonja, 2011; Hanley et al.,2006). In addition to the appearance of high mass CD44 positivebands, unique to the DMH-treated group, we have also observeda marked increase in the amount of detected CD44 at the massrange 25–30 kDa. As this mass range is not compatible with the fulllength CD44, we hypothesized that it corresponds to CD44 cleav-age products previously shown to accumulate at the neoplastic tis-sue and known to contribute to tumor progression in variousanimal cell lines (Sugahara et al., 2008; Wang et al., 2009).

Fig. 2. Colon photomicrographs of control animals or DMH-treated associated or not with diet supplementation with G.max and M. axillare BBIs. (A) Normal histologicalaspect of the colon. (B) Inflammatory infiltrate (arrowed) in the muscle layer in animals from the DMH-treated group. Absence of inflammatory infiltrates from animals inDMHS (C) and DMHM (D) groups. Hematoxylin Eosin staining. Bar = 50 lm.

Fig. 3. Chymotrypsin and trypsin-like activities recovered from lysosomal fractions after Sepharose-BBI affinity chromatography. Enzyme assays were performed usingsamples from six animals in each treatment group and the results were expressed as means through fluorescence arbitrary units/lg protein. Bars represent the standarderrors of the means. Note the increased chymotrypsin- and trypsin-like activities found in the colon and intestine in the DMH-group and similar to control levels for bothactivities in animals receiving DMH and diet supplemented with either G.max BBI (DMHS) or M. axillare BBIs (DMHM). Values significantly different from the control wereasterisked considering p 6 0.05.

A. de Paula Carli et al. / Food and Chemical Toxicology 50 (2012) 1405–1412 1409

The use of two BBI preparations aimed the comparison of theirtumor-preventive activities. The dose to be incorporated in the dietwas standardized based on their inhibitory activity over bovinetrypsin. The results revealed that both preparations were equally

effective at preventing the appearance of tissue biomarkers associ-ated with tumor development. Nevertheless, a comparative analysisof the primary sequences of BBIs isolated from G. max and M. axil-lare, in particular at the regions corresponding to their inhibitory

Fig. 4. CD44 expression pattern in the colon from control and DMH-treated animals. (A) Extracted membrane proteins from the colon of three control and three DMH-treatedanimals were separated using 12% SDS–PAGE, transferred to a PVDF membrane and immunoblotted using an anti-CD44 antibody. Note the increased expression of CD44s inDMH-treated animals and the appearance of high mass CD44 isoforms unique to this group (arrowed). Detection of CD44 cleavage products at the mass range 25–30 kDa wasalso particularly prominent in the DMH group (arrowed). (B) Densitometric analysis of the blotted membrane shown in (A). Significantly different values were observed fromthe control using ANOVA/Tukey’s test at p 6 0.05.

Fig. 5. Animals fed a diet containing either G. max BBI (BBIS) or M. axillare BBIs(BBIM) during 8 weeks displayed decreased proteasome activity. Proteasome’strypsin (T) and chymotrypsin (Q)-like activities were assayed for control and BBIS/BBIM groups. Q-like activity was also performed in the presence of MG132 tocontrol for proteasome’s specific activity. Assays were performed using samplesfrom six animals in each treatment group and the results were expressed as meansthrough fluorescence arbitrary units/lg protein. Bars represent the standard errorsof the means. Values significantly different from the control were observed atp 6 0.05 using ANOVA/Tukey’ test.

Fig. 6. DMH-treatment induces increase in the colon proteasome’s trypsin- andchymotrypsin-like activities whereas DMH plus BBIs keep them equivalent tocontrol levels. Q-like activity was also performed in the presence of MG132 tocontrol for proteasome’s specific activity. Assays were performed using samplesfrom six animals in each treatment group and the results were expressed as meansthrough fluorescence arbitrary units/lg protein. Bars represent the standard errorsof the means. Values significantly different from the control were observed atp 6 0.05 using ANOVA/Tukey’ test.

1410 A. de Paula Carli et al. / Food and Chemical Toxicology 50 (2012) 1405–1412

heads, permitted observation of critical amino acid differences. Inthis regard, the active sequences of G. max BBI are CTKSNPPQCand CALSYPAQC for the inhibitory heads of trypsin and chymotryp-sin, respectively. M. axillare BBIs are composed of two types of mol-ecules named DE3 and DE4, displaying identical amino acidsequences for the trypsin-inhibitory head (CTKSIPPQC) and distinctsequences (CTFSIPAQC and CALSEPAQC) for the chymotrypsin-inhibitory head, respectively, (Joubert et al., 1979). Given that theamino acid found in the P2’ position of M. axillare native BBI inhib-itor is Ile and the best observed modification in P20 position of a cyc-lic synthetic peptide was found to be associated with the sameamino acid, resulting in a change of Ki from 2.5 to 0.009 lM (Garianiet al., 1999), we suggest that M. axillare BBIs and/or their byproductsgenerated once given orally, have the potential to act as potent pro-tease inhibitors. This argument is strengthened considering that

during germination of M. axillare there is proteolytic processing ofthe native inhibitor, giving rise to short molecules known to be 5times more potent than the full length inhibitor found in dormantseeds (Kumar et al., 2002). In addition, as for the chymotrypsininhibitory head, particularly between the P2 e P20 positions, signif-icant amino acid differences in DE3 and DE4 isoforms exist com-pared to the respective sequence found in soybean BBI.Altogether, these findings provided our rationale for proposing theutilization of an alternative BBI source (M. axillare BBI) to preventDMH-induced lesions.

Given the structural differences and the potential therapeuticrole of BBIs, detailed investigations are necessary for elucidatingthe structure/activity relationship of such inhibitors. This would al-low exploration of their activity upon intracellular protein turn-over, either mediated by the lysosome or the 20S/26S

A. de Paula Carli et al. / Food and Chemical Toxicology 50 (2012) 1405–1412 1411

proteasome, towards prevention of various diseases includingcancer.

5. Conclusions

In our experiments mice simultaneously treated with DMH andeither G.max or M. axillare seed BBIs, the induction of neoplastic le-sions could be prevented. Since their effects correlated with theirinhibitory potential over bovine trypsin and given the increasedactivity of M. axillare BBIs over this enzyme, we suggest that thelatter inhibitors could be used in doses much lower than that re-quired for soybean BBI. Histopathological analyses revealed thatboth BBI preparations protected against the onset of severe inflam-matory processes and ultimately the appearance of pre-malignantlesions, observed in DMH-treated animals. The two investigatedbiomarkers (expression of CD44 and lysosomal protease activitiesassociated with a BBI-affinity column) showed consistent resultsunder the different experimental conditions, demonstrating theirreliability and usefulness in this experimental model. Finally, thisstudy demonstrated a significant increase in the trypsin and chy-motrypsin-like activities of the proteasome in pre neoplastic le-sions of the colon from DMH-treated animals. In contrast, normallevels were observed for proteasome activity found in animalstreated simultaneously with BBIs and DMH.

Conflict of Interest

The authors declare that there are no conflicts of interest.

Acknowledgments

The authors would like to acknowledge the financial supportprovided by FAPEMIG (Fundação de Amparo à Pesquisa do Estadode Minas Gerais), CNPq (Conselho Nacional de DesenvolvimentoCientífico e Tecnológico), Centro Universitário de Caratinga andUniversidade Federal de Ouro Preto. The authors are also greatfulto José H.B. Fortes for his technical assistance.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.fct.2012.01.036.

References

Arcinas, A., Yen, T.Y., Kebebew, E., Macher, B.A., 2009. Cell surface and secretedprotein profiles of human thyroid cancer cell lines reveal distinct glycoproteinpatterns. J. Proteome Res. 8 (8), 3958–3968.

Billings, P.C., Brandon, D.L., Habres, J.M., 1991. Internalisation of the Bowman-Birkprotease inhibitor by intestinal epithelial cells. Eur. J. Cancer 27 (7), 903–908.

Billings, P.C., Newberne, P.M., Kennedy, A.R., 1990. Protease inhibitor suppression ofcolon and anal gland carcinogenesis induced by dimethylhydrazine.Carcinogenesis 11 (7), 1083–1086.

Billings, P.C., St Clair, W., Owen, A.J., Kennedy, A.R., 1988. Potential intracellulartarget proteins of the anticarcinogenic Bowman Birk protease inhibitoridentified by affinity chromatography. Cancer Res. 48 (7), 1798–1802.

Brown, R.L., Reinke, L.M., Damerow, M.S., Perez, D., Chodosh, L.A., Yang, J., Cheng, C.,2011. CD44 splice isoform switching in human and mouse epithelium isessential for epithelial-mesenchymal transition and breast cancer progression.J. Clin. Invest. 121 (3), 1064–1074.

Cavo, M., 2006. Proteasome inhibitor bortezomib for the treatment of multiplemyeloma. Leukemia 20 (8), 1341–1352.

César, J., de Santana, M., de Oliveira, M., Santos, A., de Miranda, A., Santoro, M., deAndrade, M., 2009. A New Extraction and Purification Methodology of Bowman-Birk Inhibitors from Seeds and Germinated Seeds of<i> Macrotylomaaxillare</i&gt. Chromatographia 69 (3), 357–360.

Chen, D., Frezza, M., Schmitt, S., Kanwar, J., 2011. Bortezomib as the first proteasomeinhibitor anticancer drug: current status and future perspectives. Curr. CancerDrug Targets 11 (3), 239–253.

Chen, Y.W., Huang, S.C., Lin-Shiau, S.Y., Lin, J.K., 2005. Bowman-Birk inhibitor abatesproteasome function and suppresses the proliferation of MCF7 breast cancercells through accumulation of MAP kinase phosphatase-1. Carcinogenesis 26(7), 1296–1306.

Deshimaru, M., Yoshimi, S., Shioi, S., Terada, S., 2004. Multigene family for Bowman-Birk type proteinase inhibitors of wild soja and soybean: the presence of twoBBI-A genes and pseudogenes. Biosci. Biotechnol. Biochem. 68 (6), 1279–1286.

Forman, M.R., Hursting, S.D., Umar, A., Barrett, J.C., 2004. Nutrition and cancerprevention: a multidisciplinary perspective on human trials. Annu. Rev. Nutr.24, 223–254.

Gariani, T., McBride, J.D., Leatherbarrow, R.J., 1999. The role of the P20 position ofBowman-Birk proteinase inhibitor in the inhibition of trypsin. Studies on P2’variation in cyclic peptides encompassing the reactive site loop. Biochim.Biophys. Acta 1431 (1), 232–237.

Godge, P.Y., Poonja, L.S., 2011. Quantitative assessment of expression of celladhesion molecule (CD44) splice variants: CD44 standard (CD44s) and v5, v6isoforms in oral leukoplakias: An immunohistochemical study. Indian J. Dent.Res. 22 (3), 493–494.

Gomes, C.E., Barbosa, A.E., Macedo, L.L., Pitanga, J.C., Moura, F.T., Oliveira, A.S.,Moura, R.M., Queiroz, A.F., Macedo, F.P., Andrade, L.B., Vidal, M.S., Sales, M.P.,2005. Effect of trypsin inhibitor from Crotalaria pallida seeds on Callosobruchusmaculatus (cowpea weevil) and Ceratitis capitata (fruit fly). Plant Physiol.Biochem. 43 (12), 1095–1102.

Hanley, W.D., Napier, S.L., Burdick, M.M., Schnaar, R.L., Sackstein, R.,Konstantopoulos, K., 2006. Variant isoforms of CD44 are P- and L-selectinligands on colon carcinoma cells. FASEB J. 20 (2), 337–339.

Hernandez-Ledesma, B., Hsieh, C.C., de Lumen, B.O., 2009. Lunasin and Bowman-Birk protease inhibitor (BBI) in US commercial soy foods. Food Chem. 115 (2),574–580.

Hynes, M.J., Huang, K.M., Huang, E.H., 2009. Review paper: Implications of the‘‘cancer stem cell’’ hypothesis on murine models of colon cancer and colitis-associated cancer. Vet. Pathol. 46 (5), 819–835.

Joubert, F.J., Kruger, H., Townshend, G.S., Botes, D.P., 1979. Purification, someproperties and the complete primary structures of two protease inhibitors(DE-3 and DE-4) from Macrotyloma axillare seed. Eur. J. Biochem. 97 (1),85–91.

Kennedy, A.R., 1998. Chemopreventive agents: protease inhibitors. Pharmacol. Ther.78 (3), 167–209.

Kennedy, A.R., Beazer-Barclay, Y., Kinzler, K.W., Newberne, P.M., 1996. Suppressionof carcinogenesis in the intestines of min mice by the soybean-derivedBowman-Birk inhibitor. Cancer Res. 56 (4), 679–682.

Kennedy, A.R., Billings, P.C., Wan, X.S., Newberne, P.M., 2002. Effects of Bowman-Birk inhibitor on rat colon carcinogenesis. Nutr. Cancer 43 (2), 174–186.

Kennedy, A.R., Davis, J.G., Carlton, W., Ware, J.H., 2008. Effects of dietary antioxidantsupplementation on the development of malignant lymphoma and otherneoplastic lesions in mice exposed to proton or iron-ion radiation. Radiat. Res.169 (6), 615–625.

Kisselev, A.F., Callard, A., Goldberg, A.L., 2006. Importance of the differentproteolytic sites of the proteasome and the efficacy of inhibitors varies withthe protein substrate. J. Biol. Chem. 281 (13), 8582–8590.

Kris-Etherton, P.M., Hecker, K.D., Bonanome, A., Coval, S.M., Binkoski, A.E., Hilpert,K.F., Griel, A.E., Etherton, T.D., 2002. Bioactive compounds in foods: their rolein the prevention of cardiovascular disease and cancer. Am. J. Med. 113 (9B),71S–88S.

Kumar, P., Sreerama, Y.N., Gowda, L.R., 2002. Formation of Bowman-Birk inhibitorsduring the germination of horsegram (Dolichos biflorus). Phytochemistry 60(6), 581–588.

Losso, J.N., 2008. The biochemical and functional food properties of the bowman-birk inhibitor. Crit. Rev. Food Sci. Nutr. 48 (1), 94–118.

Matsumoto, I., Kitagaki, H., Akai, Y., Ito, Y., Seno, N., 1981. Derivatization of epoxy-activated agarose with various carbohydrates for the preparation of stable andhigh-capacity affinity adsorbents – their use for affinity-chromatography ofcarbohydrate-binding proteins. Anal. Biochem. 116 (1), 103–110.

Mitsiades, N., Mitsiades, C.S., Poulaki, V., Chauhan, D., Fanourakis, G., Gu, X., Bailey,C., Joseph, M., Libermann, T.A., Treon, S.P., Munshi, N.C., Richardson, P.G.,Hideshima, T., Anderson, K.C., 2002. Molecular sequelae of proteasomeinhibition in human multiple myeloma cells. Proc. Natl. Acad. Sci. USA 99(22), 14374–14379.

Montagut, C., Rovira, A., Albanell, J., 2006. The proteasome: a novel target foranticancer therapy. Clin. Transl. Oncol. 8 (5), 313–317.

Myung, J., Kim, K.B., Crews, C.M., 2001. The ubiquitin-proteasome pathway andproteasome inhibitors. Med. Res. Rev. 21 (4), 245–273.

Odani, S., Ikenaka, T., 1973. Scission of soybean Bowman-Birk proteinase inhibitorinto two small fragments having either trypsin or chymotrypsin inhibitoryactivity. J. Biochem. 74 (4), 857–860.

Saito, T., Sato, H., Virgona, N., Hagiwara, H., Kashiwagi, K., Suzuki, K., Asano, R., Yano,T., 2007. Negative growth control of osteosarcoma cell by Bowman-Birkprotease inhibitor from soybean; involvement of connexin 43. Cancer Lett.253 (2), 249–257.

Santana, M.A., Castro-Borges, W., Amorin, L.L., Santos, A.G., Cristine Leal, S., Andrade,M.H.G., 2011. Perennial Horse Gram (Macrotyloma axillare) Seeds:biotechnology applications of its peptide and protein content – Bowman-Birkinhibitors and lectin. In: Preedy, V.R., Watson, R.R., Patel, V.B. (Eds.), Nuts &Seeds in Health and Disease Prevention, 1st ed. Burlington, San Diego, London,pp. 899–907.

1412 A. de Paula Carli et al. / Food and Chemical Toxicology 50 (2012) 1405–1412

St Clair, W.H., Billings, P.C., Carew, J.A., Keller-McGandy, C., Newberne, P., Kennedy,A.R., 1990. Suppression of dimethylhydrazine-induced carcinogenesis in miceby dietary addition of the Bowman-Birk protease inhibitor. Cancer Res. 50 (3),580–586.

Sugahara, K.N., Hirata, T., Tanaka, T., Ogino, S., Takeda, M., Terasawa, H., Shimada, I.,Tamura, J., ten Dam, G.B., van Kuppevelt, T.H., Miyasaka, M., 2008. Chondroitinsulfate E fragments enhance CD44 cleavage and CD44-dependent motility intumor cells. Cancer Res. 68 (17), 7191–7199.

Tanaka, K., 2009. The proteasome: overview of structure and functions. Proc. Jpn.Acad. Ser. B Phys. Biol. Sci. 85 (1), 12–36.

Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer of proteins frompolyacrylamide gels to nitrocellulose sheets: procedure and some applications.Proc. Natl. Acad. Sci. USA 76 (9), 4350–4354.

Wang, S.J., Wong, G., de Heer, A.M., Xia, W., Bourguignon, L.Y., 2009. CD44 variantisoforms in head and neck squamous cell carcinoma progression. Laryngoscope119 (8), 1518–1530.

Yavelow, J., Collins, M., Birk, Y., Troll, W., Kennedy, A.R., 1985. Nanomolarconcentrations of Bowman-Birk soybean protease inhibitor suppress X-ray-induced transformation in vitro. Proc. Natl. Acad. Sci. USA 82 (16), 5395–5399.