The impaired viability of prostate cancer cell lines by the recombinant plant kallikrein inhibitor

Background: a pivotal in establishing cancer. Results: In contrast to the classical Kunitz plant inhibitor SbTI, the recombinant kallikrein inhibitor (rBbKIm) lead to prostate cancer cell death, while fibroblast viability was not affected. Conclusion:

2 a relevant role in this disease; prostate cancer can be diagnosed through the quantification of PSA levels in the serum of sick individuals (7,8). The increase of the availability of insulinlike growth factors (IGFs) by tissue kallikreins proteolytic cleavage of insulin-like growth factor binding proteins (IGFBPs) can interfere on survival and cell differentiation (9).
The ability of cells to adhere to other cells or extracellular matrix (ECM) is important for growth, development, apoptosis, inflammation, migration, tumorigenesis and immune responses (10,11). Integrins play a central role in these processes (12)(13)(14).
The progression of a normal cell to a cancer cell involves the ability of the cell to stimulate angiogenesis. When a solid tumor reaches 1-2 mm, the growth of the tumor is due to their angiogenic capacity (21). Proangiogenic factors include vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), platelet-derived growth factor (PDGF), placenta growth factor (PlGF), transforming growth factor β (TGF -β), among others (22,23). Compounds that have the RGD sequence recognize αvβ3 integrin and act as antagonists that inhibit the formation of blood vessels (16,24). These compounds can be used as antiangiogenic for the treatment of cancer.
Protease inhibitors are capable of regulating apoptosis and the cell cycle (25). Bauhinia seeds have serine and cysteine proteases inhibitors (26)(27)(28). The Bauhinia bauhinioides Kalikrein Inhibitor (BbKI), which is obtained from seeds of Bauhinia bauhinioides, is an 18-kDa protein that has a primary structure similar to that of other plant Kunitz-type inhibitors but is devoid of disulfide bridges (29). BbKI inhibits plasma kallikrein, plasmin, bovine trypsin, bovine chymotrypsin, porcine pancreatic kallikrein and murine plasma kallikrein (29). BbKI recombinant protein was obtained by heterologous expression and production in Escherichia coli (30). Nakahata et al. (20) isolated an inhibitor of Bauhinia rufa named BrTI (Bauhinia rufa trypsin inhibitor) that inhibits human plasma kallikrein (K iapp = 14 nM) in addition to trypsin (K iapp = 2.9 nM). The distinctive features of the inhibitor structure are the RGD and RGE sequences responsible for the inhibition of cellular adhesion at positions 25-27 and 128-130 (26).
To evaluate the hypothesis that the presence of the RGD/RGE sequences in plant proteins is related to their defense against predators, Sumikawa et al. (31) expressed a recombinant inhibitor based on the primary sequence of BbKI. The amino acid residues around P21-28 in BbKI were replaced by those present in BrTI (V/E21, S/A24, H/R25, H/D27 and A/G28) and region P127-130 (E/D127 and Q/E130). These replacements resulted in the production of a modified inhibitor named rBbKIm (Bauhinia bauhinioides kallikrein inhibitor modified). rBbKIm inhibits trypsin (K iapp = 1.6 nM), chymotrypsin (K iapp = 7.4 nM) and human plasma kallikrein (K iapp = 3.6 nM) on the same order of magnitude as the native protein.
Due to the characteristics presented above, the use of this atypical inhibitor in a prostate cancer model is an important tool for studying the complex network involved in the proteolytic establishment of this malignancy. Therefore, this work aims to describe the action of the recombinant Bauhinia bauhinioides kallikrein inhibitor modified (rBbKIm) in DU145 and PC3 cell lines and its capacity as an antiangiogenic drug. 3 by Sumikawa et al. (31). Briefly, the recombinant pET-29a(+) plasmid was transformed into BL21(DE3) (Novagen, Madison, WI, USA) cells harboring pET29BbKIm and grown in LB-medium (Invitrogen, CA, USA) supplemented with 30 μg/mL kanamycin (Gibco, CA, USA) at 37°C.When the absorbance of the culture at 600 nm reached a value of 0.4, isopropyl β-Dthiogalactopyranoside (IPTG) (Invitrogen, CA, USA) was added at a final concentration of 0.2 or 0.5 mM, and the culture was grown for additional 3 h. Subsequently, the cells were harvested by centrifugation (4000 x g, 20 min, 4°C) (Hitachi himac CR21-GIII, Japan). The pellets were then resuspended in 10 mL of lysis buffer (0.05 M Tris/HCl, pH 8.0 and 0.15 M NaCl). After the bacteria were disrupted by sonication (Unique, Campinas, Brazil) and lysed in 0.5% Triton X-114 (Sigma Aldrich, St. Louis, USA) (eight cycles of ultrasound, 30 s, 40 W), the soluble protein fraction was isolated by centrifugation (4000 x g, 20 min, 4°C). The protein was precipitated from the supernatant by adding acetone (80%, v/v) at 4°C. The sediment was separated by centrifugation, dried at ambient temperature, dissolved in 10 mM Tris/HCl buffer, pH 8.5 and dialyzed against 50 mM Tris/HCl, pH 8.5. Subsequently, the soluble fraction was chromatographed on a DEAE-Sepharose Fast Flow (GE Healthcare, NJ, USA) and equilibrated with 0.1 M Tris/HCl, pH 8.5. The inhibitor was eluted with 0.05 M NaCl in the same buffer. The fractions that inhibited trypsin activity were pooled and chromatographed on a Superose 12 (GE Healthcare, NJ, USA) Äkta Purifier system for the removal of possible contaminants.
Protease Inhibition Assays -The inhibition of trypsin was used to evaluate the activity of the expressed protein as a model for enzymes with the same mechanism of action, called trypsin-type enzymes, such as the kallikrein family (32)(33)(34). The assays were performed in a 96-well plate in a final volume of 250 μL as described by Oliva et al. (35). Briefly, the inhibitor activity was determined by pre-incubation for 10 min at 37°C in 0.05 M Tris/HCl, pH 8.0, containing 0.02% CaCl 2 and trypsin (40 nM). Then, 1.0 mM of the substrate α-benzoyl-D-L-arginine-ρnitroanilide was added to the reaction and incubated for 30 min at 37ºC. K iapp values were determined by adjusting the experimental points to the equation for tight binding by nonlinear regression analysis using the Grafit software (36).
Evaluation of the Degradation of the Substrate HD-Pro-Phe-Arg-ρNa by Proteins of the Conditioned Medium of DU145 and PC3 -DU145 and PC3 cells (1x10 6 ) were seeded into plates 100 mm in diameter (TPP, Trasadingen, Switzerland) for 24 h in RPMI medium supplemented with 10% FBS. Subsequently, the medium was changed to RPMI medium without FBS for 12 h, and 50 µM rBbKIm was added. After 48 h, the supernatant was removed, centrifuged (563 x g for 10 min) and concentrated 4 times using an ultrafiltration membrane with an exclusion pore size of 14 kDa (Amicon, Millipore, Brazil), which resulted in the conditioned medium (CM). Total protein concentration of the CM was quantified by the micro BCA kit, according to manufacturer's instructions (Pierce, IL, USA).

4
To evaluate the degradation of the substrate HD-Pro-Phe-Arg-ρNa by proteins of the conditioned medium of DU145 and PC3, an enzymatic reaction composed of 50 mM Tris/HCl buffer, pH 8.0, 0.5 M NaCl, 50 µg of total protein present in the CM and 0.5 mM HD-Pro-Phe-Arg-ρNa substrate was used. After 24 h at 37ºC, the absorbance was measured at 405 nm in a Packard spectrophotometer (Packard, SpectraCount model).
MTT Cell Viability Assay -Cell viability was determined by the modified Cell Adhesion Assays -The cell adhesion assays were performed in triplicate according to Nakahata et al. (20). Briefly, 24well culture plates were coated with fibronectin (Millipore, Brazil), laminin (Sigma Aldrich, St. Louis, USA) and collagens I and IV (Sigma Aldrich, St. Louis, USA) (4 µg/100 µL /well) and incubated overnight at 4°C. The wells were blocked with 1% BSA in PBS (100 μL/well) for 1 h at 37°C and washed three times with PBS. DU145 or PC3 (5 x 10 4 cells/50 μL/well) cells were pre-incubated with rBbKIm in different concentrations for 30 min.
Subsequently, the cells and inhibitor were added to the wells and incubated for 4 h at 37ºC. Non-adherent cells were washed with PBS buffer, pH 7.4, three times. The remaining adherent cells were fixed with 100% methanol for 5 min, washed three times with PBS buffer, pH 7.4, and stained with 1% (v/v) toluidine blue in 1% sodium tetraborate for 5 min. The wells were exhaustively washed with PBS, and the absorbed stain was dissolved in 1% (v/v) SDS for 30 min at 37 °C. The absorbance was measured at 540 nm using a spectrophotometer (Packard, SpectraCount model). Each experimental condition was performed in triplicate.
Wound Healing -DU145 and PC3 cells and HUVECs (2 x 10 5 cells/well) were seeded in 24-well culture plates (TPP, Trasadingen, Switzerland) until complete confluence. With the support of a pipet tip, a slit of cells were scraped from the plate, the medium was aspirated and the cells were treated with rBbKIm (50 and 100 µM rBbKIm in DU145 and PC3 cells, and 50 µM rBbKIm in HUVECs) and SbTI (50 and 100 µM in DU145 and PC3 cells) in 10% FBS. The wound healing was observed at zero, 15 and 23 h (DU145 and PC3) or zero and 18 h (HUVECs). For each time of analysis, three measurements were performed from one ledge to another slit. The average distance between the edges at 15 and 23 h (DU145 and PC3) and 18 h (HUVEC) were subtracted from the average at time zero, and we then compared the mean migration of the treated cells with the control (untreated cells) (37).
Transwell migration chamber -To evaluate the migration of HUVEC cell in vitro, 4 x 10 4 cells in 250 µL serum-free RPMI 1640 medium with or without rBbKIm were placed in the top of cell culture inserts of 8 µm pore size (Millicell, Millipore, Brazil). The lower chambers were filled with 300 μL of RPMI supplemented with 2% heat-inactivated fetal calf serum (FCS; Sigma Aldrich, St. Louis, USA) to serve as a chemoattractant for cell migration. The cells were treated with 50 µM or 100 µM rBbKIm and the control cells with a medium containing 7 mM HEPES, pH 7.4 5 (vehicle). Plates were incubated for 24 h at 37 °C under 5.0% CO 2 atmosphere. Transwell units were removed from 24-well plates and fixed with methanol for 30 min, cells in the upper chamber were removed and the cells of the lower compartment, stained with 1% toluidine blue (Sigma Aldrich, St. Louis, USA) for 15 min. Then, these cells were counted under a light microscope by two independent observers.
Flow Cytometry Analysis of Cell Apoptosis -DU145 and PC3 cells (1 x 10 5 cells) were seeded on 6-well plates (TPP, Trasadingen, Switzerland) for 24 h for complete adhesion. After this time, the wells were washed three times with medium RPMI without FBS at incubation periods of 15 min at 37C and 5% CO 2 . The cells were incubated for 24 h in RPMI without FBS to synchronize the cell cycle. Thereafter, the cells were treated with rBbKIm and SbTI (50 and 100 µM) for 24 and 48 h in RPMI without FBS at 37C and 5% CO 2 . At the end of each incubation, the cells were removed from the plate using trypsin-EDTA solution (Cultilab, Campinas, Brazil), transferred to cytometry tubes and washed with 500 µL of binding buffer (BD PharMingen kit), centrifuged and resuspended in 50 µL of the same buffer, 3 µL of Annexin V-FITC and 5 µL of propidium iodide (PI). The reaction mixture was incubated for 30 min in the dark, and then 300 µL of the same binding buffer was added to each tube and analyzed by flow cytometry (FACSCalibur, BD, San Diego, USA) [38,39]. The cells were fixed with 2% (v/v) paraformaldehyde for 30 min and centrifuged. One hundred microliters of PBS containing 0.01% (v/v) saponin and 0.4 mg/mL RNAse A (Sigma Aldrich, St. Louis, USA) was added to each tube, and the tubes were incubated for 30 min at 37°C. At the end of the incubation period, 5 µL of 1 mg/mL propidium iodide and 200 µL of PBS were added to each tube and analyzed by flow cytometry (FACSCalibur, BD, San Diego, USA) (40). The results were analyzed by ModFit LT 3.2.
Cytochrome c release assay -DU145 and PC3 cells were seeded (5x10 4 cells/well) on 13-mm glass coverslips, placed in 24-well plates and incubated for 24 h at 37C and 5% CO 2 for full compliance. After this period, cells were incubated in RPMI medium without FBS for 24 h. Subsequently, the cells were treated with 50 and 100 µM rBbKIm in RPMI medium without FBS and incubated at 37C and 5% CO 2  Determination of Caspase-3 Activation by Flow Cytometry -DU145 and PC3 cells (1x10 5 cells) were seeded in each well of a 6well plate (TPP, Trasadingen, Switzerland) containing RPMI medium supplemented with 10% fetal bovine serum and 100 IU/mL penicillin/streptomycin solution for adhesion. The cells were then washed three times with RPMI medium without FBS and incubated for 24 h in RPMI without FBS. Twenty-four hours later, the medium was replaced with two different concentrations (50 µM and 100 µM) of rBbKIm and SbTI for 48 h at 37ºC and 5% 6 CO 2 . At the end of the treatment, the cells were removed from the plate using a trypsin-EDTA solution (Cultilab, Campinas, Brazil), transferred to cytometry tubes and fixed with 100 µL of 2% paraformaldehyde (v/v) for 30 min at room temperature. The cells were resuspended in 200 µL of 0.01% glycine in PBS and incubated for 15 min at room temperature. The cells were then resuspended in 200 µL of 0.01% saponin in PBS and incubated for 15 min at room temperature. Finally, the cells were incubated with 10 µL of anti-cleaved caspase 3 conjugated with Alexa Fluor 488 (BD, San Diego, USA) for 40 min. The results were analyzed by flow cytometry (FACSCalibur -BD, San Diego, USA).
Determination of Caspase-9 Activation -DU145 and PC3 cells (1x10 5 cells) were seeded into plates 100 mm in diameter (TPP, Trasadingen, Switzerland) containing RPMI medium supplemented with 10% fetal bovine serum and 100 IU/mL penicillin/streptomycin solution for 24 h. The cells were then washed three times with RPMI medium without FBS and incubated for 12 h in the same medium. After 12 h, the medium was replaced with RPMI medium without FBS, and the cells were treated with 50 µM rBbKIm for 48 h. After 48 h, the cells were washed twice with cold PBS, scraped and centrifuged. The pellet was frozen at -80ºC. For cell lysis, a lysis buffer containing 25 mM HEPES buffer, pH 7.4 plus 0.1% Chaps (Sigma Aldrich, St. Louis, USA), 1 mM EDTA, 2 mM MgCl2, 2 mM DTT and cocktail protease inhibitors (1:10) (Roche, Switzerland) was used. The pellet was resuspended in 50 µL of lysis buffer and incubated on ice for 20 min. The cell pellets were frozen in dry ice, thawed at 37ºC and vortexed; this cycle was repeated ten times. At the end, the cells were centrifuged at 15,771 x g for 10 min at 4ºC. The supernatant with the cytoplasmic proteins was quantified by the micro BCA method (Pierce, IL, USA). Thirty micrograms of total protein was incubated in buffer (25 mM Hepes buffer, pH 7.4 containing 10% sucrose, 0.1% Chaps and 1 mM DTT) with 20 mM Ac-Leu-Glu-His-Asp-AFC substrate (Calbiochem, Germany) at 37ºC, and readings were performed every 15 min in a fluorometer (Spectra Maz, Geminen MS) using λ exc = 400 nm and λ em = 505 nm.

Measurement of Ca 2+ Mobilization in Cells
-Ca 2+ was measured spectrophotometrically using the fluorescent probe fura-2/AM (Molecular Probe, OR, USA). DU145 (3 x 10 4 cell/well) and PC3 (3.5 x 10 4 cell/well) cells were seeded in 6-well plates (TPP, Trasadingen, Switzerland) for 24 h for complete adhesion. After this time, the cells were suspended in Hank's balanced salt solution (HBSS) and treated with rBbKIm (100 and 200 µM) for 1 h at 37ºC and 5% CO 2 . Subsequently, the cells were loaded with the probe (2 µM) for 40 min and transferred to a fluorometer (SPEX FluoroLog-2, AR-CM System, Perkin Elmer) using λ exc = 340 e 380 nm and λ em = 505 nm. The cells were either stimulated with thrombin (0.5 UI/mL), an agonist for PARs receptors, or not stimulated.
In Vitro Angiogenesis Assay on Matrigel -To evaluate angiogenic structure formation HUVECs were used. This assay was performed as described by Paschoalin et al. (41). The Matrigel TM (BD, San Diego, USA) was added to cold 96-well plates (TPP, Trasadingen, Switzerland) and incubated at 37ºC for 1 h for their polymerization. HUVECs supplemented with 0.2% FBS were seeded on the Matrigel TM at a density of 5x10 3 cells/well in the presence and absence of 50 and 100 µM of rBbKIm at 37ºC and 5% CO 2 . After 24 h, the tubular structures that were formed were observed under a microscope light and photographed for subsequent counting. The control sample (untreated) was considered to be 100%, and the formation of tubular structures of each treated sample was calculated in relation to control.
Statistical Analysis -The data were analyzed as the mean ± SD, and statistical significance was determined using ANOVA and t-tests (SigmaPlot 10.0). *p<0.05 was considered to be significant.

Action of rBbKIm and SbTI on Cell
Viability and Cell Death -The effect of rBbKIm on the viability of DU145 and PC3 cells was analyzed by the MTT assay at different time points (24, 48 and 72 h) (Fig. 1). The results of our studies have shown that rBbKIm in DU145 cells were more subtle because the inhibitor was not effective at low concentrations (less than 50 µM). However, after 48 h, rBbKIm effectively inhibited cell viability at concentrations in the range of 50-100 M (reduction by 33 and 56%, respectively), and a slight increase in this effect was observed after 72 h of incubation (Fig. 1A). The action of rBbKIm in the early phase of proliferation (24 h) was effective in PC3 cells, and 25 M and 50 M of rBbKIm inhibited the cell viability by 29% and 49%, respectively. The effectiveness of inhibition was better observed in PC3 cells after 48 h of incubation with 25 M and 50 M rBbKIm in which a decrease in cell viability of 58% and 73% was observed, respectively. Compared to 48 h, no significant modification in this activity was observed after 72 h of incubation (Fig. 1B). We should note that the reduction of fibroblast viability was less at concentrations above 25 µM (Fig. 1C). Comparing the effects of rBbKIm to that of SbTI, at incubation times and concentrations in which the rBbKm action was more significant, one can notice that SbTI did not interfere on DU145 cell viability and poorly on PC3 cell viability (100 µM, 48 h) (Fig. 1D-E). Indeed, to directly address if components from the bacterial preparations could be responsible for the rBbKIm effects, the bacterial endotoxin LPS was also investigated on DU145 and PC3 (Fig. 1F). The results showed that, if present, this bacterial component did not interfere with cells viability. Therefore, these results suggest that rBbKIm has a more specific effect on prostate cancer cells.
Moreover, the experimental data on the cell cycle phases showed a distinct effect of the inhibitor, because DU145 rBbKIm treatment (50 and 100 µM, 24 and 48 h) did not significantly alter the cell cycle compared to the control (untreated cells) (Fig. 2A). Whereas rBbKIm did not affect the PC3 cell cycle in an early phase, except at 100 µM after 48 h, led to cell cycle arrest in the G0/G1 and G2/M phases (Fig. 2B). Moreover, SbTI did not led to DU145 and PC3 cell cycle arrest (Fig. 2C-D).
In addition, DU145 and PC3 cells were treated with 50 and 100 µM rBbKIm for 24 and 48 h to evaluate cell death by apoptosis using Annexin V and PI. When incubating DU145 cells with 50 µM rBbKIm for 24 h, a 45% increase of DU145 cells annexin V + PIwas observed, which suggests that apoptotic cell death was induced by rBbKIm. After 48 h of treatment with 50 or 100 µM rBbKIm, 64 and 68% of cells Annexin V + PIwere observed, respectively (Fig. 2E). PC3 cells after a 24 h treatment with 50 and 100 µM rBbKIm showed a slight increase in apoptotic cell death (Fig. 2F). After 48 h of treatment with 50 or 100 µM rBbKIm, 48 and 44% of cells were Annexin V + PI -, respectively (Fig.  2F). SbTI, 24 h and 48 h, did not induce apoptotic death in both cell lines (Fig. 2G-H).
To corroborate whether cell death induced by rBbKIm occurs by apoptosis, cytochrome c release and caspase activation analyses were performed in DU145 and PC3 cells. Previous to rBbKIm treatment, we observed a colocalization of mitotracker red and cytochrome c in both cell lines, indicating that cytochrome c is present in the mitochondria. Cytochrome c did not colocalize with the mitotracker after rBbKIm (50 or 100 µM) treatment, indicating the release of cytochrome c from the mitochondria to the cytoplasm of DU145 and PC3 cells (Fig. 3).
The apoptotic process involves the activation of caspases, which are essential to unravel the mechanism of the action of cytotoxic drugs. Flow cytometric analysis showed an activation of caspase-3 in the DU145 cells treated with rBbKIm (50 and 100 µM) for 48 h (Fig. 4A), but not with SbTI (Fig.  4C). Similarly, caspase-3 was not activated in PC3 cells treated with rBbKIm (Fig. 4B) and SbTI (Fig. 4D). The activation of caspase-9 was evaluated with the fluorogenic substrate 8 LQHD-AFC, and the cleavage of the AFC fluorophore was monitored. Only PC3 showed a significant activation of caspase-9; DU145 did not present any caspase-9 activity (Fig.  4E).
Hydrolytic Activity of Proteases in the Conditioned Medium of DU145 and PC3 Cells -To evaluate the effect of rBbKIm on DU145 and PC3 extracellular proteases, the culture medium supernatant of DU145 and PC3 cells (conditioned medium -CM) previously treated with rBbKIm for 48 h (RPMI medium without FBS) was submitted to an enzymatic reaction in the presence of HD-Pro-Phe-Arg-ρNa, a specific substrate for plasma and tissue kallikreins. As depicted in Fig. 5 in which the DU145-CM and PC3-CM were incubated at 37ºC for 24 h, there was a 95% inhibition of substrate hydrolysis, demonstrating the effectiveness of the inhibitor for blocking serine protease release by both prostate cancer cells.
Action of rBbKIm on DU145 and PC3 Cell Adhesion -The effect of DU145 and PC3 cell adhesion on extracellular matrix proteins such as fibronectin, laminin and collagen I and IV in the presence of rBbKIm was evaluated. The results show that rBbKIm (3.125 to 100 µM) did not inhibit DU145 cell adhesion to fibronectin, laminin, or collagen I and IV (Fig.  6A-B). The effect of rBbKIm on PC3 cell adhesion was very similar to DU145 cell adhesion, and there was no inhibition of PC3 cell adhesion to matrix proteins ( Fig. 6C-D). Conversely, rBbKIm increased the adherence of DU145 cells to collagen IV at concentrations up to 12.5 µM (Fig. 6B) and PC3 cells to laminin and collagen I and V (Fig.  6C-D).
Action of rBbKIm and SbTI on DU145 and PC3 Cell Migration -The action of the inhibitors on the cell motility of DU145 cells (Fig. 7A-B) indicates that 100 µM rBbKIm did not inhibit DU145 cell migration when incubated for 15 and 23 h. treatment, 50 and 100 µM rBbKIm inhibited 23 and 38% of PC3 migration, respectively (Fig. 7C-D), while SbTI inhibitor did not interfere on DU145 and PC3 cells migration (Fig. 7E-H).

Action of rBbKIm on Ca 2+
Mobilization -The ability of rBbKIm to stimulate calcium mobilization was evaluated and compared to thrombin, a canonical endogenous agonist for PARs. Thrombin greatly increases Ca 2+ released from internal storages. rBbKIm itself was not capable of inducing calcium mobilization, but in the presence of thrombin, an enzyme in which the proteolysis activity is not affected by the inhibitor, an important increase of calcium mobilization was observed (~71%) in both lineages DU145 (Fig. 8A) and PC3 (Fig.  8B). These results suggest that the effects of rBbKIm appear to be dependent on integrin activation, and the interaction of rBbKIm with RGD-binding integrins facilitates the increase of intracellular calcium (Fig. 8A-B).
Action of rBbKIm on Angiogenesis -Numerous characteristics are being explored for the development of antitumorogenic agents. Among these agents are angiogenesis inhibitor drugs that are capable of modulating endothelial cell viability and migration. Thus, our study evaluated the effect of rBbKIm on cell viability, cell migration and the formation of angiogenic structures in human endothelial cells (HUVECs). After 24 h of treatment, rBbKIm reduced HUVEC viability by 28 and 27% when concentrations of 50 and 100 µM were used, respectively (Fig. 9A). The effect of rBbKIm on the migration of HUVECs was measured by evaluating the percentage of cells that migrated through the slit after scraping the cells and transwell migration chamber. In the wound healing assay, rBbKIm (50 µM) inhibited migration by 22% after 18 h of treatment (Fig. 9B), similar results were obtained by transwell assay (20% and 30 % of inhibition with 50 and 100 µM rBbKIm, respectively after 24 h of treatment) (Fig. 9C). Angiogenesis was tested in a matrigel using HUVECs, and an inhibition of angiogenic structures was observed when cells were treated with 50 and 100 µM of rBbKIm. In the 9 presence of rBbKIm 50 µM, there was a 49% decrease in the formation of angiogenic structures. A 70% inhibition was observed when the cells were treated with 100 µM rBbKIm (Fig. 9D-E). These results show that the inhibitor rBbKIm, besides acting on DU145 and PC3 prostate cancer cells, can also be a potent inhibitor of angiogenesis.

DISCUSSION
The action of protease inhibitors has been studied in different models. Specially in cancer, protease inhibitors from Bowman-Birk and Kunitz families are the most studied (14,(42)(43)(44)(45) not only due to the inhibition of the net activation of proteases but also due to the presence of atypical structures that invoke mechanisms that suppress pathways fundamental to the establishment of tumors. In this study, we used the chimeric inhibitor, rBbKIm. In addition to being a potent inhibitor of kallikreins, rBbKIm has a very atypical peptide sequence from another Bauhinia inhibitor that is known to be responsible for the inhibition of cellular adhesion (20). These features attracted our attention, and we selected prostate cancer as a model for our studies due to the importance of tissue kallikreins activation in the development of this disease (9,46). The conserved activity of the purified cloned inhibitor was confirmed by its ability to block trypsin activity in vitro.
The importance of the inhibitor selectivity for trypsin-like activity secreted by these cells was demonstrated to be important and may reflect the cellular signaling events triggered by proteolysis, such as the activation of protease activated receptors (PARs). Studies have shown that expression of PAR1 is clearly increased in advanced prostate cancer (47)(48)(49). The activation of PARs promotes proliferation, invasion and release of angiogenic factors by cancer cells (50,51), and kallikreins are enzymes that are known to cleave this receptor with a specificity similar to thrombin (52, 53). Thus, it is possible to infer that rBbKIm led to the impairment of these prostate cellular events by its direct inhibitory action.
The contribution of the motif sequences RGD/RGE in the suppression of cell adhesion differed from previous reports, which highlights the importance of the recognition by integrins. This recognition regulates the invasive and metastatic potential of tumor cells (16), and peptides bearing this motif have been utilized in cell attachment inhibition (54)(55)(56)(57). Particularly, PC3 and DU145 cells express high levels of α5β1 and αvβ3 integrins that recognize the RGD sequence in cell adhesion proteins (54). The activation of protease activated receptor 1 (PAR1) decreases PC3 cell adhesion to collagen I and IV and laminin (58), which leads to an enhancement of metastasis through improved DU145 and PC3 cell motility, and apoptosis is prevented (59). This information helped us understand the effect of rBbKIm: the RGD domain promoted the adhesion of DU145 and PC3 cells to collagen IV and PC3 cells to laminin and collagen I. An increase of 37% and 71% in adhesion was observed when DU145 cells were treated with 100 µM rBbKIm and seeded on either laminin or collagen IV, respectively. In PC3 cell adhesion, an increase of 43% on laminin and 51% on collagen I was observed. A slight increase of 23% in cell adhesion was also observed when PC3 cells were seeded on collagen IV. Our results suggest that these effects may be a consequence of the inhibition of extracellular proteases capable of activating the PAR1 receptor localized in the cell membrane. The ineffectiveness of rBbKIm on DU145 and PC3 adhesion to fibronectin may be related to the fact that the adhesion to fibronectin is poorly affected by the activation of PAR1, which is mainly due to the α5 subunit expression being lower compared to that of collagen and laminin receptors (58). Cell adherence on the stromal prostate is important to prevent the entry of cancer cells into the bloodstream, thus preventing metastasis in distant tissues (56).
Although the binding of rBbKIm to the cell surface did not impair cell adhesion, the deleterious effect on DU145 and PC3 viability suggests that this macromolecule provides signals that interfere with cell survival.
Integrins are unable to bind to their receptors in an inactive state (60). ECM-integrin binding evokes molecular signaling cascades. By using fluorescence assays, we were able to demonstrate that rBbKIm potentiate increase of cytoplasmic calcium concentration by thrombin. This effect may be the consequence of the amplification of RGD/RGE binding recognition by the receptors that are more exposed after PAR activation (61). Thus, the rBbKIm binding signaling pathway is dependent of integrin activation. Our experimental procedure demonstrated that thrombin induces PAR activation and is a protease that is not affected by rBbKIm. Therefore, in cell culture, integrin activation is stretched by proteolysis produced by enzymes inert to rBbKIm action. Thus, this inhibitory action impaired the action of the PAR and delayed integrin activation. However, this inhibitory action may also contribute to calcium increase and subsequently trigger apoptosis signaling pathways (62).
The effect of rBbKIm was compared to the classical Kunitz plant inhibitor, SbTI, and in contrast to the inert effect of SbTI, rBbKIm was selective causing prostate cancer cell death. The induction of apoptosis in DU145 is not related to cell cycle arrest. Studies have shown that the DU145 cell line has a mutation in the RB tumor suppressor gene (63) due to a mutation in exon 21 (64), which encodes a truncated form of the Rb protein, promoting resistance to cell cycle arrest (65). Apoptosis in these cells suggests that an activation of caspase-3 occurs. However, we could not detect an activation of caspase-9, although the route has been activated via the mitochondrial release of cytochrome c.
Treatment of PC3 with rBbKIm suggests an induction of apoptosis via mitochondrial cytochrome c release and caspase-3 activation. This result was similar to previous results obtained (66), which show that the treatment of neuroblastoma cells with TNF-α induces cytochrome c release and caspase-9 activation but not caspase-3 activation. The PC3 cell death that occurred when with rBbKIm treatment (100 µM/ 48 h) may be related to G0/G1 and G2/M cell cycle arrest as similarly demonstrated by Agarwal et al. (67) in which CR9-7 human fibroblasts were treated with tetracycline.
Another event implicated in the activation of PAR1 in PC3 is the increase in migratory ability, which is important in the process of invasion and metastasis (68,69). rBbKIm inhibits cell migration of PC3 cells; however, this effect is not observed in DU145 cells. The distinct effect on PC3 cells can be understood if we consider the different proteases involved in the mechanism of cell migration in the cells that are affected by the inhibitor.
Cytotoxic drugs generally do not differentiate normal from metastatic tissues. Thus, it is important to develop new drugs that target proteins that are differentially expressed in cancer cells compared to normal adult tissues (70). This reasoning is why the action of rBbKIm on the viability of the amniotic fluid human fibroblast was analyzed. This drug did not inhibit the viability of these cells, which indicates its selectivity for tumor cells.
Angiogenesis occurs when the tumor is large in size and requires the diffusion of nutrients and oxygen (71,72). The impairment of HUVEC cell migration, promoted by rBbKIm, may be a consequence of its effect on cell viability, leading to angiogenesis inhibition. These characteristics may be relevant to the establishment of antiangiogenic properties because the inhibitor reduced the formation of angiogenic structures in the matrigel, suggesting an ability to interfere with tumor metastasis.
In summary, our findings provide a basis for using a distinct inhibitor to investigate the participation of proteases and activation of signaling pathways in different cell lines.