Effects of an Anticarcinogenic Bowman-Birk Protease Inhibitor on Purified 20S Proteasome and MCF-7 Breast Cancer Cells

Proteasome inhibitors have been described as an important target for cancer therapy due to their potential to regulate the ubiquitin-proteasome system in the degradation pathway of cellular proteins. Here, we reported the effects of a Bowman-Birk-type protease inhibitor, the Black-eyed pea Trypsin/Chymotrypsin Inhibitor (BTCI), on proteasome 20S in MCF-7 breast cancer cells and on catalytic activity of the purified 20S proteasome from horse erythrocytes, as well as the structural analysis of the BTCI-20S proteasome complex. In vitro experiments and confocal microscopy showed that BTCI readily crosses the membrane of the breast cancer cells and co-localizes with the proteasome in cytoplasm and mainly in nucleus. Indeed, as indicated by dynamic light scattering, BTCI and 20S proteasome form a stable complex at temperatures up to 55°C and at neutral and alkaline pHs. In complexed form, BTCI strongly inhibits the proteolytic chymotrypsin-, trypsin- and caspase-like activities of 20S proteasome, indicated by inhibition constants of 10−7 M magnitude order. Besides other mechanisms, this feature can be associated with previously reported cytostatic and cytotoxic effects of BTCI in MCF-7 breast cancer cells by means of apoptosis.


Introduction
Proteases are involved in many biological processes such as the hydrolysis of intracellular proteins, transcription, cell cycle, cell invasion and apoptosis [1]. The activity of these proteases can be regulated by proteolytic degradation and inhibitors that display variable degrees of affinity with the enzymes [2,3]. Natural protease inhibitors are classified into about 20 different families [4,5], among which the Bowman-Birk inhibitors (BBI) and Kunitz have been the most studied [6,7].
Bowman-Birk inhibitors are found in mono and dicotyledons, especially in leguminous seeds [8]. Diets rich in these legumes have been associated with low incidence of cancer in human populations, in which protease inhibitors are considered to be responsible for this protective action [9][10][11]. In addition, BBIs are the most characterized inhibitors for their role as carcinogenesis suppressors [12][13][14][15][16], and they have been studied in a human phase IIa clinical trial [17].
The Black-eyed pea Trypsin/Chymotrypsin Inhibitor (BTCI) is a natural plant protease inhibitor isolated from Vigna unguiculata (Cowpea) seeds, and it belongs to the BBI family. Members of this protease inhibitor family are proteins that inactivate the functions of serine proteases by providing a reactive site, present in the canonical loop connecting the b-hairpin motif, which acts competitively as a pseudo or analogue substrate for the cognate enzyme [2,18,19]. The remarkable complementarities of these inhibitors, in particular BTCI, determine their high affinity for cognate enzymes. The dissociation constants of 10 27 -10 29 M magnitude order for BBIs and BTCI are compatible with their low dissociation process from the S1 active site of the enzymes [3,20,21].
BTCI is a globular protein containing 83 amino acid residues presenting seven disulfide bonds and molecular weight of 9.1 kDa [22][23][24]. It has two different and independent reactive sites for trypsin (Lys26) and chymotrypsin (Phe53) [23][24][25][26]. Its binary and ternary complexes with these proteases were isolated and physicochemically characterized by analytical ultracentrifugation, viscometry and light scattering, which showed the hydrodynamic parameters and high stability of these complexes at pH 7.0 [25].
The binding constants were calculated by enzymatic assays resulting in values of 10 7 -10 9 M 21 magnitude for chymotrypsin and trypsin, respectively [27,28]. Additionally, thermodynamic parameters calculated for the formation of trypsin-BTCI and chymotrypsin-BTCI complexes characterized these associations as endothermic, spontaneous and entropy-driven processes [27][28]. In spite of the slow process of peptide bond cleavage in the P1 reactive sites of BTCI and the characteristic reversibility of the inhibition process, the presence of one disulfide bond flanking each loop containing the P1 residues prevents the displacement of the product from the S1 enzyme pocket [24].
The biochemical, biophysical and biotechnological properties of BTCI have been extensively characterized [14,23,24,[27][28][29][30][31][32][33][34][35]. BTCI is a thermally stable protein that retains 96% of its inhibitory activity after heating at 95uC for 60 min, as well as when it is exposed from pH 3 to 10 [30]. BTCI presented in vitro and in vivo effects on development of the boll weevil (Anthonomus grandis), a cotton pest. It caused larval growth delay and reduced the insect population, showing its potential for expression in transgenic plants engineered for pest resistance [31]. Doseresponse experiments showed that BTCI was capable of inducing changes in renal functions by enhancing guanylin-induced natriuresis, a strong but transient diuretic response. Additionally, BTCI promoted increases in urine flow, in fractional excretion of Na + and K + , in perfusion pressure, in glomerular filtration rate and osmolar clearance [32]. It was the first description of renal effects induced by a protease inhibitor belonging to the Bowman-Birk family. Both effects are due to the ability of BTCI to inhibit chymotrypsin-like protease, thus playing a fundamental role in the renal metabolism of guanylin in a natriuretic process [32] and trypsin-and chymotrypsin-like protease from the midguts of larvae and adult insects from the cotton boll weevil Antonomus grandis [31]. Furthermore, our previous results indicated that BTCI induced significant cytostatic and cytotoxic effects on MCF-7 breast cancer cells associated with severe cell morphological alteration, lysosome membrane permeabilization and apoptosis [14]. Although BTCI induced a significant reduction in MCF-7 cancer cell viability and proliferation (arrest in S and G2/M phase), it did not affect the viability of normal MCF-10A breast cells. BTCI induced severe morphological changes in MCF-7 cancer cells, such as plasma membrane fragmentation, cytoplasm disorganization, and presence of double-membrane vesicles, mitochondrial swelling, and an increase in the size of lysosomes. In addition, meaningful DNA fragmentation, annexin-V+ cell number increase, mitochondrial membrane potential reduction, and cytoplasm acidification were also detected [14].
In another study, it was shown that topical applications of BTCI on the skin of mice significantly reduced the incidence and volume of pre-malignant lesions during chemical induction of nonmelanoma cancer in this tissue [35]. Reduction was also observed both in the number of histopathological features as well as in production of inflammatory mediators involved in tumor progression. Regarding the preventive effects, BTCI treatments were able to retard the progression of non-melanoma skin cancer, probably inducing anti-inflammatory effects [14,35]. Non-melanoma skin and breast cancer are of high incidence, and breast cancer is the most frequent malignancy among women worldwide [36]; both are leading causes of cancer-related mortality.
The high stability of BTCI in terms of maintaining its inhibitory activity at temperatures as high as 95uC in a wide range of pHs (3)(4)(5)(6)(7)(8)(9)(10)(11), besides its lack of effect on normal breast cells MCF-10A, are of fundamental importance in evaluating the reduction of side effects using, in the future, BTCI in breast cancer treatment. Therefore, the aforementioned results are relevant in focusing on BTCI as a potential anticarcinogenic drug, so as to design new strategies for current breast cancer treatments, as well as skin cancer prevention.
As is known, cancer is related to dysfunction in the cell cycle in which the ubiquitin-proteasome system plays a fundamental role in the proteolysis of intracellular proteins [37,38]. The 20S proteasome is a catalytic complex mediating caspase-like, trypsinlike and chymotrypsin-like activities in the b1, b2, and b5 catalytic subunits, respectively [39], important in ubiquitinated protein breakdown. This process occurs after recognition of ubiquitinated proteins by 26S proteasome [40][41][42], which is formed by 20S proteasome binding to one or two 19S regulatory caps [43].
Cancer cells are generally associated with increased constitutive proteasome activity compared to non-malignant cells, due to cell proliferation, increased oxidative stress, and elevated cytokine levels, which can also induce the expression of immunoproteasomes [44]. The constitutive 20S proteasome is present in all eukaryotic cells, whereas the immunoproteasome is mainly expressed in immunocompetent tissues. The b1, b2, and b5 catalytic subunits present in the constitutive proteasome are replaced by b1i (low-molecular-weight protein-2, LMP2), b2i (multicatalytic endopeptidase complex-like-1, MECL1) and b5i (LMP7) in the immunoproteasome [45,46]. Despite both proteasome diversities, the expression and function of the immunoproteasome in different cancer types seem to be discrepant in the literature and still remain under investigation [47]. Regarding MCF-7 breast cancer cells, few descriptions of the expression and activity of the immunoproteasome subunits have been reported [48][49][50]. As shown, in MCF-7 cells both b5i/LMP7 and b1i/ LMP2 subunits presented minimum expression levels when compared to other cancer cells, even with MCF-10A cells, indicating that immunoproteasomes are almost absent in MCF-7 breast tumor cells [48].
Proteasome inhibition results in cellular homeostasis disruption [51] and in the induction of apoptosis [52,53]. In this context, proteasome inhibitors have been characterized as important compounds for cancer therapy [54]. Among them, Bowman-Birk inhibitors (BBI) appear as a very promising compound of anticarcinogenic drugs. As reported in 2005 [55], BBI from soybean inhibits the chymotrypsin activity of the proteasome in breast cancer cells, but the molecular mechanism involved in this process is not fully understood. BBI in MCF-7 cells decreases proteasome function and results in up-regulation of MAP kinase phosphatase-1, which suppresses ERK1/2 activity. These results led to BBI being indicated as a novel mechanism that contributes to prevention of cancer, as well as a potential chemopreventive agent [55].
Here, we report the effects of BTCI on the 20S proteasome of MCF-7 breast cancer cells and on the catalytic activity of the 20S proteasome purified from horse erythrocytes, as well as the molecular analysis of the BTCI-proteasome complex. Our findings also identified BTCI as a proteasome inhibitor, which could be related to ubiquitin-proteasome pathway inhibition and its previously reported anticarcinogenic effects [14].

Protein Purification
BTCI was purified as previously described [29] and its purity was confirmed by mass spectrometry analysis [30]. The 20S proteasome used in enzymatic assays, BTCI-proteasome association and structural stability studies was purified from horse erythrocytes. The purification was performed by sequential chromatography in DEAE-Sepharose, sucrose density gradient ultracentrifugation and chromatography on mono-Q column using FPLC facility (Pharmacia) [56]. This proteasome source was chosen considering its well established purification procedure resulting in a high amount of protein and also the conserved structural features of the proteasome in all organisms. For all experiments, BTCI and 20S proteasome were sterilized by filtration with a 0.22 mm membrane (Millipore, USA).

BTCI and 20S Proteasome Association Evaluated by Dynamic Light Scattering (DLS)
The interaction of BTCI with 20S proteasome was analyzed by DLS assays carried out through a laser wavelength of 800 nm, using a DynaPro -DLS model (Wyatt Technology Corporation, Santa Barbara, CA, USA) molecular-sizing instrument equipped with a Peltier system for temperature control. Solutions of proteins were centrifuged at 15,000 g for 20 min at 4uC, and the supernatant filtered through a 0. 22  The intensity of scattered light from each sample was normalized considering the buffer scattering contribution. Polydispersity (Pd), hydrodynamic diameter (D H ) and molecular weight were determined from the intensity correlation function using the cumulant method [57,58] and the Dynamics V.6 software. The experiments were performed with an average of 100 acquisitions.

Antibody Anti-BTCI Preparation
The antibody against BTCI was obtained from immunized Swiss mice lineage with 3 applications of purified BTCI (0.1 mg/ mL) in phosphate-buffered saline (PBS) (10 mM phosphate, 137 mM NaCl, 2.7 mM KCl, pH 7.4) and complete adjuvant, then with incomplete adjuvant and without adjuvant, sequentially. The blood was collected from the carotid vein after anesthetization with ketamine (75 ml) and xylazine (120 ml) in PBS (805 ml) and then clotted and centrifuged at 10,000 g for 7 minutes in order to obtain the serum. A polyacrylamide gel with 400 mg of BTCI was transferred to a nitrocellulose membrane and incubated for 2 hours with the serum anti-BTCI. The membrane was washed with PBS and the antibody was eluted with 0.2 mM glycine-HCl buffer at pH 2. The solution containing the antibody was adjusted to pH 7.0 and concentrated in amicon ultra-0.5 mL centrifugal filter (Merck Millipore). The functionality of the antibody was confirmed by Western blotting. This study was approved by the Ethics Committee (Proc. No 47028/2007) of the Institute of Biology, University of Brasilia.
Immunofluorescence Analysis MCF-7 cells were seeded on 12-well plates at a density of 5610 4 cells in culture medium and incubated with 200 mM BTCI for 2, 6, 12 and 24 h at 37uC. The negative control was done with cells incubated in the absence of BTCI. Cells were harvested, washed three times with phosphate-buffered saline (PBS) and fixed with 3.7% paraformaldehyde. Cells were permeabilized with 0.2% Triton-X in PBS, blocked with PBS -5% of skimmed milk (w/v) and incubated with primary antibodies: anti-BTCI at 1:100 dilution (serum obtained from immunized mice with purified BTCI), and rabbit anti-20S proteasome core (Affiniti -Biomol, USA) in PBS -1% skimmed milk (w/v), at room temperature for 2 h. Cells were washed with PBST (PBS, 0.2% Tween 20, v/v -Pharmacia Biotech, Buckinghamshire, UK). Goat anti rabbit Alexa Fluor 633-conjugated and goat anti mouse Alexa Fluor 488conjugated secondary antibodies (Invitrogen, USA), at 1:400 dilutions, were incubated in PBS -1% skimmed milk (w/v), for 2 h, in dark room, at room temperature. A blue nuclear counterstain 49,6-diamidino-2-phenylindole (DAPI) was used to visualize nuclear DNA in fixed cells. Cells were washed with PBST and MilliQ water, and then analyzed in Immunofluorescence Confocal Microscopy (Leica Microsystems, TCS SP5 German). The fluorescence intensity of the images was quantified by ImageJ program (Wayne Rasband; Research Services Branch, National Institute of Mental Health, Bethesda, Maryland, USA). All values were expressed as mean 6 SD of three independent experiments. The differences in the fluorescence intensity of each group were analyzed by t-test. Asterisk indicates that the values are significantly different from the group analyzed (*, p,0.001).

Enzymatic Assays
Purified 20S proteasome at concentration of 2.0 mg/mL was incubated with BTCI (2.0 to 30.0 mg/mL) and 62.5 mg/mL of the following fluorogenic substrates in 20.0 mM Tris-HCl pH 7.5, 1.0 mM EDTA, 1.0 mM NaN 3 , 1.0 mM DTT: t-butyloxycarbonyl-L-leucyl-L-arginyl-L-arginine-4-methylcoumaryl-7-amide (Boc-leu-argarg-AMC) for trypsin-like; benzyloxycarbonyl-L-leucyl-L-leucyl-L-glutamyl-naphthylamide (Z-leu-leu-glu-bna) for caspase-like and N-succinyl-Leu-Leu-Val-Tyr-AMC (Suc-leu-leu-val-tyr) for chymotrypsin-like activities. The control of proteasome inhibition assays was done using the proteasome inhibitor named MG132 (carbobenzoxylleucyl-leucyl-leucinaI-H, also called Z-LLL-CHO) at a similar concentration to that at which BTCI showed ,80-95% of proteases inhibition. The assays were carried out in triplicate, at room temperature for 60 min. The hydrolysis of fluorogenic substrates was monitored at 480 nm (wavelength of emission, l em ), after excitation at 380 nm (l exc ) for chymotrypsin-like and trypsinlike, and l exc of 410 and l em of 335 nm for caspase-like. Inhibition curves were obtained by plotting the relative activities of the proteases versus BTCI concentration. Inhibition constant of the enzyme-inhibitor complex, K i , was calculated from fitted inhibition curve considering the Morrison equation using the GRAFIT program version 3 (Erithacus Software Ltd., UK). All inhibition constants were calculated until the chi-square (x 2 ) reached a value less than 0.012, compatible with the best fit curves representing the experimental data. The control for this experiment was done using the proteasome inhibitor named MG132 at a similar concentration to that used when BTCI showed ,80-95% of protease inhibition.

BTCI and 20S Proteasome Association Evaluated by Dynamic Light Scattering
The hydrodynamic parameters of the 20S proteasome in complex with BTCI, as well as the self-association tendency of BTCI and the proteasome, in a temperature and pH dependent manner, were investigated by using dynamic light scattering. This technique has been commonly used mainly to analyze protein crystallization tendency [59]. Furthermore, in the last decade, DLS has also been a useful tool in a number of structural studies involving proteins and nanoparticles, such as association of proteins with surfactants [60], protein unfolding process [61], protein stability [62,63], dimension of nanoparticles [64][65][66], protein-protein interaction and protein self-association process [67][68][69][70], in addition to others.
In the present work, the parameters obtained from DLS measurements indicate that the 20S proteasome appears as a monomer at 21.4 nM and pH 7.5 with hydrodynamic diameter of 15 nm (Fig. 1a), in agreement with previously reported data [71,72]. BTCI forms a trimer at 15 mM (Fig. 1b) and pH 7.5, with hydrodynamic diameter of 4.5 nm, which is consistent with the tendency of this inhibitor to form oligomers [33,73]. In contrast, at pH 2.0 and 4.0 (Fig. 1c) BTCI and 20S proteasome assembled into oligomeric or aggregate structures. BTCI interacts with the 20S proteasome in the first 15 min and, after that, induces low conformational changes in the 20S proteasome, as indicated by the increase in the hydrodynamic diameter of the complex from 21.8 nm to 22.7 nm (Fig. 1d) and concomitant disappearance of scattering peak corresponding to BTCI.
The proteasome-BTCI complex presented tertiary conformational changes as indicated by differences in hydrodynamic diameters under these conditions (Fig. 2a). Indeed, the complex was characterized as thermally stable at pH 7.5 and up to 55uC, but dissociated and formed an aggregate above 60uC (Fig 2b). These structural changes were attributed mainly to the proteasome, since BTCI was previously reported as a highly thermostable protein within a wide pH range (3.0-10.0), preserving its tridimensional structure and inhibitory activity up to 95uC after incubation for 60 min [30].
The structure of the 20S proteasome is conserved through fungi to humans [74]. Its physicochemical features, such as sedimenta-tion and diffusion coefficients, secondary structure content and amino acid composition, are similar among eukaryotic organisms [75]. The 20S proteasome was here characterized as a stable protein complex from pH 6.0 to 12.0, with aggregation tendency at acid pH, and structural stability up to 60uC, where dissociation of proteasome ring structures probably occurs. Our finding corroborates previous data reported for the 20S proteasome from ostrich skeletal muscle, which presented pH and temperature optima ranged between 8.0 and 11.0, and 40 to 70uC, respectively, and stabilities for enzymatic activities between pH 5.0 and 12.0 up to 60uC [76].

Immunocolocalization of BTCI and Proteasome in MCF-7 Breast Cancer Cells
The immunocolocalization of BTCI and the proteasome in MCF-7 breast cancer cells was investigated by confocal microscopy. Cells were treated with 200 mM BTCI from 2 to 24 h. The negative control was performed in the absence of the inhibitor. Figure 3a shows that BTCI (green) and the proteasome (red) are distributed all over the cells after 2 h incubation. The two molecules co-localize in the cytoplasm and nucleus of cells, as indicated by the yellow color, corresponding to the overlapping Alexa Fluor 488 (green) and Alexa Fluor 633 (red) antibody conjugated fluorophores, respectively. These results indicate that MCF-7 cellular membranes are very permeable to BTCI, and that once BTCI is inside cells, it co-localizes in similar regions, such as the proteasome.
The anti-20S proteasome used in the present study was a general antibody that is able to detect multiple constitutive (b1, b2, b5) and immunoproteasome (b1i, b2i, b5i) subunits. The co-localization observed in the immunofluorescence assay (Figure 3a) using the anti-20S core antibody probably occurs between BTCI and the constitutive proteasome. It is supported by the reportedly very low expression of immunoprotesomes in MCF-7 cells [48][49][50]; hence, their representation in this assay is negligible. Therefore, the detection of i20S or BTCI targeting the i20S subunits in MCF-7 cells could be difficult to verify.
Fluorescence intensity quantification (Fig. 3b-c)    Moreover, the fluorescence intensity of the proteasome (Fig. 3c) decreased significantly in the presence of the inhibitor (p,0.001).
This result suggests that BTCI enters through the MCF-7 cell (high initial fluorescence intensity with 2 h incubation) compatible with continuous flux of the inhibitor through the cell, coincident with cellular membrane disruption and other morphological alterations in the cell, as previously reported by Joanitti et al. (2010) [14]. In contrast, the fluorescence intensity corresponding to the proteasome (Fig. 3c) decreased until 12 h, which may have occurred as a consequence of the steric hindrance of antibody binding due to conformational changes in the proteasome caused by BTCI.
These immunofluorescent assays showed that this inhibitor is taken up by the MCF-7 cells in a time-dependent manner and is present in the cells for 24 h. In addition, according to our previous results, the ultra-structural analysis of MCF-7 cell morphology indicated a pronounced effect of BTCI on plasma membrane fragmentation, cytoplasm disorganization, presence of doublemembrane vesicles, and lysosome size increase [14]. The recognition and internalization processes of BTCI by MCF-7 are unknown. However, according to those results, this process can be activated after structural and/or functional alterations of plasma membrane integrity occurred with exposure of phosphatidyl serine outside the inner membrane.

BTCI Inhibits the 20S Proteasome Catalytic Activities
Three protease activity sites are present in the b subunits of the 20S proteasome, including the caspase-like (b1), trypsin-like (b2) and chymotrypsin-like (b5) sites [77][78][79]. In the present work BTCI was characterized as a novel and potent Bowman-Birk inhibitor of the 20S proteasome through specific inhibition of those three protease activities. BTCI presented high affinity to the 20S proteasome, as indicated by inhibition or dissociation constants (K I or K d ) values of 1.0610 27 M, 7.0610 27 M and 14.0610 27 M for trypsin-like (Fig. 4a), chymotrypsin-like (Fig. 4b) and caspase-like sites (Fig. 4c), respectively. The calculated K I magnitude order of 10 27 to 10 28 M is similar to the previous estimate for most BBI inhibitors [3,25,80] and also BTCI [27,28]. In addition, BTCI was able to inhibit all proteases in a similar way to the known proteasome inhibitor MG132 (carbobenzoxyl-leucylleucyl-leucinaI-H), here used as a control of proteasome inhibition assays (Fig. 5). It can be observed that BTCI was a more potent inhibitor for trypsin than MG132 and presented a similar inhibition to MG132 against caspase-and chymotrypsin-like activities.
MG132 was among the first developed proteasome inhibitors and the most widely used in research. It is a peptide aldehydebased reversible proteasome inhibitor, which inhibits the proteasome primarily on the chymotrypsin-like site, but also inhibits trypsin-and caspase-like sites. Although it is a potent proteasome inhibitor, MG132 is rapidly oxidized into inactive carbonic acid in vivo and, for this reason, its therapeutic use is usually prevented [81][82][83].
BTCI was the first member of the Bowman-Birk family to be characterized as a potent inhibitor of all three trypsin-like, chymotrypsin-like and caspase-like proteasomal activities. As previously reported, BBI, a Bowman-Birk Inhibitor isolated from soybean, inhibits only chymotrypsin-like activity (inhibition of 70%) of the 26S proteasome in vitro at 40 mM [55]. In contrast, BTCI inhibited almost 100% of the three enzymatic activities of the 20S proteasome at 20 mM. This indicates that although BTCI and BBI present similar structures [24], low differences in primary and tertiary structure of BTCI, compared to BBI, are essential for its significant inhibition of three proteasome proteases. Surprisingly, the way that BTCI inhibits the caspase-like proteasomal activity has never been reported before in the case of any other Bowman-Birk type inhibitor. This ambiguous feature is sustained by the structural changes in the 20S proteasome which occurred when in complex with BTCI (Fig. 1c). The BTCI-20S proteasome complex probably favors the steric hindrance effects, where the caspase-like catalytic site may be buried within a large protein surface, when BTCI binds specifically to chymotrypsin-like or trypsin-like catalytic sites. These conformational changes reflect the differences in proteasome enzymatic activities in the presence of their specific substrates and BTCI. This is in accordance with previous work reporting the conformational changes in the proteasome when a particular active site is activated or inhibited [84].
Studies of proteasome inhibitors and yeast mutants have shown that chymotrypsin-like activity is a rate-limiting step responsible for the ubiquitinated protein breakdown [85][86][87][88][89]. In mammalian cells, the caspase-like and trypsin-like sites also play a significant role in protein degradation, but the relative importance of the three active sites depends on the protein being degraded. Indeed, previous studies show that simultaneous inhibition of the chymotrypsin-like and the caspase-or trypsin-like sites were needed to decrease protein degradation by the proteasome [90]. As discussed, the inactivation of one site is not sufficient to markedly block protein degradation by the proteasome. As previously reported, BTCI can inhibit the activity of chymotrypsin and trypsin enzymes independently and simultaneously through its two specific reactive sites (P1), Lys26 and Phe53, acting as a substrate analogous for cognate serine proteases by forming stable ternary or binary complexes with those enzymes [23][24][25]27,28,34]. Similarly, BTCI can inhibit the activity of all three enzymes of the b1, b2, and b5 subunits of the 20S proteasome.
As is known, BBIs from legumes are potent inhibitors of both trypsin and chymotrypsin, with Ki values within the nanomolar units [3,21]. These inhibitors, such as BTCI, are well-characterized as ''double-headed'' protease inhibitors, presenting two conserved reactive sites located in two opposite and independent domains. The structure of BTCI folds into two similar domains with three b-strands forming an antiparallel b-sheet stabilized by disulfide bonds. Each domain contains one reactive site in position P1, Lys26 and Phe53, located in a loop connecting two strands, capable of binding to the active site of the trypsin and chymotrypsin, respectively. The distance between the Ca atoms of the two reactive sites (32.5 Au) is very similar to that of the BBIs, which allows the inhibition of two protease molecules simultaneously and independently [24,25]. According to our results, the inhibition of the proteasome by BTCI probably occurs in an independent manner for all three proteases, since their active sites are positioned in the N-terminal region of independent b1, b2 and b5 subunits of the 20S proteasome. In this case, independent monomers of BTCI are oriented towards each specific active site of the different b subunits composing the 20S proteasome. Additionally, it is known that the three b1, b2 and b5 subunits work independently [91].
The magnitude of dissociation constants was similar to those obtained for free BBIs with cognate enzymes. These results indicate that reactive P1 sites of the BTCI bind to the specific S1 pocket of the catalytic region of the subunits b2 (trypsin-like), b5 (chymotrypsin-like) and block the active site of b1 (caspase-like) of the 20S proteasome [92]. This property strongly indicates that this protein can act as a potent inhibitor of the proteasome, one of the most important macromolecular complexes involved in intracellular protein breakdown. These results can partially corroborate recently reported cytostatic and cytotoxic effects of BTCI on MCF-7 breast cancer cells [14]. In this case, the induction of apoptosis cell death by BTCI is associated with severe morphological alterations in the cell and lysosome membrane permeabilization. These alterations are related to degradation of the cytoskeleton proteins, structural protein-phospholipid membrane disruption, chromatin condensation and DNA fragmentation, where cytoplasmatic caspase proteins play vital roles. These proteins belong to a group of enzymes known as cysteine proteases and exist within the cell, differently from the caspase-like protein of the proteasome, as inactive pro-forms or zymogens. These zymogens can be cleaved to form active enzymes following the induction of apoptosis [93].
Currently, there are numerous reports of proteasome inhibitors which act only on one or two protease-like sites of the proteasome, but few proteasome inhibitors acting simultaneously on three proteases-like sites [94]. One of the most investigated proteasome inhibitors is bortezomib, the first proteasome inhibitor approved for clinical use by the FDA in 2004, which is a more potent inhibitor for b5 than b1 subunit. Bortezomib is a reversible inhibitor of the proteasome in spite of its very slow dissociation rate [82,83]. This inhibitor has several molecular effects, such as cell cycle arrest in G2/S phase [95], stabilization of cell cycle regulatory proteins, inhibition of NF-kB activation, induction of apoptosis, among others [96,97], mainly involving the inhibition of subunit b5 (chymotrypsin-like activity) of the proteasome [98].
Like bortezomib, BTCI is a reversible inhibitor that inhibits the proteases of the b1 and b5 subunits. Moreover, like MG132 (Fig. 5), BTCI inhibits all three chymotrypsin-, trypsin-and caspase-like activities of the 20S proteasome. Although bortezomib and MG132 are considered to be potent proteasome inhibitors, exploratory investigation of others is needed. This is principally due to the occurrence of bortezomib resistance after clinical use in multiple myeloma and other diseases, such as malignant lymphoma [99], and the fact that the therapeutic use of MG132 is usually prevented. In these contexts, therefore, the fact that BTCI inhibits all three chymotrypsin-, trypsin-and caspase-like activities may be an advantage in its application in cancer therapy, compared to caspase-like. The dissociation or inhibition constants, calculated from the fitted curve by GraFit program (Erithacus software Ltd.), were 1.0610 27 M, 7.0610 27 M, and 14.0610 27 M, respectively. doi:10.1371/journal.pone.0086600.g004 Figure 5. Inhibitory activity of BTCI and MG132 (positive control), toward 20S proteasome caspase-like (Casp), chymotrypsin-like (Chym) and trypsin-like (Tryp). The concentration of both inhibitors is 10 mM. As shown, BTCI is a more potent inhibitor for trypsin than MG132 and presented similar inhibition to MG132 against caspase-like and chymotrypsin-like. doi:10.1371/journal.pone.0086600.g005 other reported proteasome inhibitors. These inhibitory properties of BTCI are probably associated with its previously reported cytostatic and cytotoxic effects on breast cancer cells [14], as discussed above. Furthermore, as is known, the inhibited ubiquitin-proteasome pathway can induce cell death. This is one of the molecular pathways of cell death that need to be investigated in order to verify the involvement of BTCI.
The mechanism by which BTCI induces apoptosis of the MCF-7 cells is not yet known. However, as indicated here, BTCI activity against the 20S proteasome may be related to apoptosis by inhibiting protein breakdown processes. This mechanism, along with others promoting cytostatic and cytotoxic effects, may be responsible for the anticarcinogenic effect of BTCI on MCF-7 cells by means of apoptosis.
Finally, this work is the first study analyzing the molecular association of BTCI with the 20S proteasome in breast cancer pathways. BTCI forms a stable complex with the 20S proteasome, independently inhibiting its three catalytic activities. Our data also indicate that due to its inhibitory action on the proteasome, BTCI may be associated with the anticarcinogenic effects previously reported in MCF-7 cancer cells [14]. Therefore, BTCI can become an even more promising anticancer agent in the treatment and prevention of breast cancer, but other studies should be undertaken to elucidate the mechanism of BTCI action in breast cancer cells.