Introduction

Breast cancer is the most frequently detected feminine neoplasia worldwide. One of the factors that have contributed to the development and progression of cancer in the last decades is type 2 diabetes mellitus (DM2) [1, 2]. Hyperglycemia and hyperinsulinemia have been considered factors that promote carcinogenesis and cancer progression in diabetics [3]. A high concentration of glucose or insulin favors the proliferation, migration, and invasion of cancer cells [4,5,6,7,8,9,10].

The plasminogen (Plg) activation system is a key regulator of extracellular matrix (ECM) remodeling, directly due to the ability of plasmin to degrade some of its components, such as laminin and fibronectin, and indirectly through the activation of matrix metalloproteinases (MMP), enzymes that jointly degrade all of the components of the ECM [11,12,13]. Cell-associated plasmin proteolysis contributes to different cell process such as tissue remodeling embryogenesis, invasion, metastasis, angiogenesis, and inflammation [13]. Plasminogen has been localized at the cell surface of mammary carcinoma tissue [14] and a number of cell types including different breast cancer cell lines [15, 16]. The binding of components of the plasminogen activation system at the cell surface facilitates the conversion of plasminogen to plasmin and protects plasmin for inactivation by circulating inhibitors [17]. Plasminogen is bound to the cell surface in a lysine-dependent manner via its kringle domains [15, 17].

Although experiments in plasminogen-deficient mice indicate that the plasminogen is not critical for cancer development and metastasis [18], in wild-type cells and organisms several evidences indicate the participation of plasminogen and its activation by uPA in cancer and metastasis. Plasminogen activation is positively correlated with an aggressive tumor phenotype [19], and the increase of serum levels of uPA and uPAR have been associated with poor prognosis in patients with prostate or breast cancer [20]. It has been suggested that the inhibition of plasminogen activation on the cell surface, including the binding of plasminogen to its surface receptors can interfere with the generation of plasmin, which results in a diminution of pericellular proteolytic activity preventing tumor invasion [21, 22].

Synthetic derivatives of lysine, such as epsilon-aminocaproic acid (EACA), have been used as antifibrinolytic drugs to control bleeding during surgical procedures [23, 24]. Epsilon-aminocaproic acid inhibits plasminogen binding to its receptors, inhibiting plasmin formation; therefore, it prevents ECM remodeling, migratory, and invasive activity of cancer cells, and reduces tumor growth and metastasis in animal models.

Hyperglycemia induces the expression of uPA in pancreatic cancer cells [25] and enhances the migratory and invasive ability of several cancer cells [25, 26]. We found that high glucose concentration (HG) and insulin induce epithelial-mesenchymal transition (EMT) and invasive behavior in metastatic MDA-MB-231 breast cancer cells, associated with increased expression of uPA, uPAR, PAI-1, and uPA activation [27]. The aim of the present work was to analyze whether EACA prevents the pro-invasive effects of HG and insulin by modulating the plasminogen activation system and EMT. The understanding of these mechanisms could allow the development of therapeutic strategies for the prevention of cancer development and progression in DM2 patients, utilizing EACA or similar compounds.

Methods

Cell cultures

Human breast cancer MDA-MB-231 (American Type Culture Collection, ATCC, Manassas, VA, USA) were maintained in Dulbecco’s Modified Eagle’s Medium (Sigma-Aldrich Co., St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS) (Biowest, Nuaillè, France) containing penicillin (100 U/ml) and streptomycin (100 µg/ml) (Life Technologies, Inc. BLR, Grand Island, NY, USA). The cells were grown in 75-cm2 tissue culture flasks in a humidified 5% CO2, 95% air atmosphere in an incubator at 37 °C. Before each experiment, cells were seeded in 6.0-cm-diameter tissue culture plates. The cells were cultured with normal glucose 5.6 mmol/L and further by the following 48 h treatments: (a) control, normal glucose 5.6 mmol/L (LG); (b) osmolality control, using glucose 5.6 mmol/L + mannitol 19.4 mmol/L (Mannitol); (c) high glucose 30 mmol/L (HG); (d) high glucose + 80 nmol/L recombinant human insulin (HG-I) (Humulin NPH, Elli Lilly, Indianapolis, IN, USA); (e) high glucose + EACA 12.5 mmol/L (HG-EACA); (f) high glucose/insulin + EACA 12.5 mmol/L (HG-I-EACA). The concentration of EACA (Sigma-Aldrich Co., St. Louis, MO, USA) was selected, probing the effect of 7.5 and 12.5 mmol/L EACA on the growth of MDA-MB-231 cell population.

Proliferation and viability

Growth curves were constructed to test for the potential effect of EACA on cell proliferation. Cells were plated in a flat bottom 96 well plates, using an initial cell density of 5000 cells per well and allowed to growth under the different conditions of treatment, using EACA at 7.5, 10 and 12.5 mmol/L. Cell number was estimated using the 3–4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) colorimetric assay [28], after 24 and 48 h of treatment, 20 µL of MTT (5 mg/ml) solution (ATCC, Manassas, VA, USA) was added to each well followed by 4–6 h incubation at 37 °C. After the media were removed, 200 μl of dimethyl sulfoxide was added to each well to dissolve the formazan formed. After 30 min incubation at room temperature, the absorbance was measured at 560 nm with a microplate reader iMark (Bio-Rad, Hercules, CA, USA). The MTT cell viability assay is based on the conversion of MTT to violet color formazan crystals by mitochondrial dehydrogenase enzymes. All experiments were performed in triplicate.

The cell percentage of viability was determined using the trypan blue exclusion assay. Briefly, 1 × 106 MDA-MB-231 cells were seeded in 24 well culture plates and treated for 48 h. Cells was collected and removed of the plate by trypsinization. The cells were briefly re-suspended in the normal culture medium. Cell viability was assessed by adding 50 μL of 0.4% trypan blue solution in 150 µL phosphate-buffered saline (PBS) to 50 μL of the cell suspension. After 2 min, the number of living cells, which did not retain the dye, was counted using a hemocytometer to calculate the viability percentage.

Cell migration assay

The migratory abilities of MDA-MB-231 cells were measured using the scratch-wound migration assay in 24 well plates. When tumor cells were grown to 80–90% of confluence, the cell monolayer was scratched with a sterile 1000 µL micropipette tip, washed once with fresh medium, cultured under the different conditions of treatment, and photographed. After a 24 h incubation, the cells were photographed again with a CDD camera (Evolution, Media Cybernetics, Rockville, MD, USA) under an inverted microscope (Olympus CKX41, Tokyo, Japan), and the distance of migration (µm) was measured using Q-Capture Pro7 software (QImaging, Inc., Surrey BC, Canada).

Transwell in vitro invasion assay

The invasiveness of MDA-MB-231 cells was determined using transwell chambers, 8-μm pore size (polyvinylpyrrolidone-free polycarbonate filter with 6.5 mm diameter) (Corning Inc., Corning, NY, USA) pre-coated with 30 μg Matrigel (BD Biosciences, San Diego, CA, USA). 2 × 105 cells were diluted in 200 µL of FBS-free media and were added to the upper chamber of the transwell, while 750 μL medium containing 10% FBS was added to the lower chamber. After incubation for 24 h at 37  °C in a CO2 incubator, non-invading cells were removed from the upper chamber, and invading cells were fixed and stained for 30 min with 0.5% crystal violet in 25% methanol. The number of invading cells in five random fields was counted under a light microscope (Olympus CKX41, Tokyo, Japan) using a 10× objective. Three independent experiments were performed, each done in duplicate. All treatments were applied 24 h before the transwell assay and 24 h during this assay.

Colony formation assay

This method was used to evaluate the tumorigenic potential of the cells or clonogenicity, measuring the proliferative ability of a single cell to form a clone and produce a viable colony [29]. MDA-MB-231 cells were seeded per well in 6-well tissue culture plates using the described treatments. The cells were grown for 15 days, further, the media were removed and cells were washed twice with PBS. The colonies were fixed with 6% v/v glutaraldehyde, stained with 0.5% w/v crystal violet, observed, and photographed with a CDD camera (Evolution, Media Cybernetics, Rockville, MD, USA) under an inverted microscope (Olympus CKX41, Tokyo, Japan). Each treatment was performed in triplicates.

RNA extraction and quantitative real -time PCR

Total RNA was extracted from cells with Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions. Total RNA (1 µg) was reverse transcribed for 60 min at 37 °C, then reverse transcriptase was inactivated at 70 °C for 15 min. The process of real-time PCR was performed using the SYBR Green technology and the PCR Rotor Gene Real-Time Apparatus (Cobbett Research, Sydney, Australia), employing Maxima SYBR Green/ROX Master Mix kit (Fermentas, Thermo-Fisher Scientific, Vilnius, Lithuania) and the following primer pairs: for uPA, forward 5′-CGCTGCTCCCACATTGGCTAA G-3′ and reverse 5′-TGTGCATGGGTGAAGGGAGAGC-3′ [30]; uPAR, forward 5′-CAACGAGGGCCCAATCCT-3′ and reverse 5′-GTAACACTGGCGGCCATTCT-3′ [31]; PAI-1, forward 5′-TGCTGGTGAATGCCCTCTACT-3′ and reverse 5′-CGGTCATTCCCAGGTTCTCTA-3′ [32]; MMP9, forward 5′-CTGCCCCAGCGAGAGACTCTAC-3, reverse 5′-GCTGTCAAAGTTCGAGGTGGTA-3′; MMP2, forward 5′-CAAAAACAAGAAGACATACAT-3′, reverse 5′-GCTTCCAAACTTCACGCTC-3′ [33] and ribosomal protein S18 (RPS 18) as a housekeeping gene, forward 5′-GATATGCTCATGTGGTGTTG-3′ and reverse 5′-AATCTTCTTCAGTCGCTCCA-3′ [32].

The amplification reaction included 40 cycles of heat denaturation at 95 °C for 30 s, followed by annealing of the primers for 30 s at 51 °C (RPS 18), 54 °C (PAI-1) 58 °C (uPAR), or 60 °C (uPA), and extension for 30 s at 72 °C. After the final cycle, the temperature was maintained at 72 °C for 10 min. The relative expression of the genes was determined with the comparative C T method (2−ΔΔCT) using their respective baselines as reference points. Results were expressed as folds of change in relation to the glucose 5.6 mmol/L group.

Western blot analysis

MDA-MB-231 cells were treated for 48 h at 37 °C in 10 cm-diameter tissue culture plates at a density of 1 × 106 cells/plate. Cells were transferred into 100 μL of lysis buffer (RIPA-Tris buffer mmol/L: EGTA 2; NaCl 316, Na2MoO4 20, NaF 50, Tris–HCl 20, Na3VO4 100, PMSF 100, EDTA 100, leupeptin 0.1%, aprotinin 0.1%, SDS 0.2%, and Triton-X100 2%) and maintained under constant shaking for 2 h at 4 °C. Subsequently, the sample was centrifuged for 5 min at 20,800 rpm and the supernatant (80 μg of protein) was denatured in Laemmli sample buffer [34], resolved through 12% SDS-PAGE and electroblotted onto polyvinylidene difluoride (PVDF) membranes. Membranes were incubated for 90 min in Tris-buffered saline (TBS) containing 5% dried skimmed milk and 0.1% Tween 20 to block non-specific protein binding sites. Subsequently, the membranes were incubated for 24 h at 4 °C with the primary antibody 1:500, against to: PAI-1, β actin, uPA, uPAR, HCAM, vimentin, E-cadherin, α-enolase from Santa Cruz Biotechnology (Santa Cruz, CA, USA); AKT, AKT-phospho-Ser 473 from GeneTex Inc (Irvine, CA, USA). Antibodies were diluted in TBS-Tween 20 0.1% including 5% dried skimmed milk. Subsequently membranes were washed with TBS-Tween and incubated with peroxidase-conjugated secondary antibodies 1:10,000. Proteins were detected using the Clarity Western ECL substrate (Bio-Rad, Hercules, CA, USA). The blots were subjected to densitometry analysis and data were analyzed using GraphPad Prism 5 software (GraphPad Software, San Diego, CA, USA). Western blots were repeated at least three times.

Gel zymography of uPA

Cells were treated for 48 h in a serum-free medium, further the culture medium (conditioned medium or medium containing secretions of the cells, including proteases) was collected and mixture with electrophoresis sample buffer (0.5 M Tris HCl, pH 6.8, SDS, glycerol, and bromophenol blue), then cells were lysed with the same buffer and proteases were studied by gel zymography.

The uPA in cellular lysates and conditioned medium was examined by casein-plasminogen zymography [35] after 48 h of treatment. Proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), using gels copolymerized with 1 mg/mL of β-casein and 10 μg/mL bovine plasminogen (both of Sigma-Aldrich Co., St. Louis, MO, USA). After electrophoresis, gels were washed twice with 2.5% Triton X-100 for 30 min and incubated in activating buffer (100 mmol/L glycine, 10 mmol/L EDTA, pH 8.3) for 24 h at 37 °C. Gels were stained with Coomassie Brillant Blue R (0.1% w/v) and destained in 30% methanol with 10% acetic acid. Caseinolytic activity appears as a clear band against a blue background. Bands were quantitated by densitometry using the imageJ software.

Plasminogen activator activity

uPA activity was measured using a two steps amidolytic assay. Plasminogen is converted to plasmin by uPA and plasmin activity was measured using a specific chromogenic substrate, D-Val-Leu-Lys-4 nitroanilide acetate (Sigma-Aldrich Co., St. Louis, MO, USA), which is cleaved by plasmin into a residual peptide and 4-nitroaniline (pNA) [25, 26]. The conditioned medium and cell lysates were obtained after 48 h of treatment in serum-free conditions and frizzed at −70 °C. The assay was carried out at 37 °C, first samples were incubated 18 h with 10 μg/mL bovine plasminogen (Sigma-Aldrich Co., St. Louis, MO, USA) in a 0.2 ml reaction mixture containing 50 mmol/L Tris (pH 7.4), 110 mmol/L NaCl; furthermore 0.3 mmol/L of the chromogenic substrate was added and the absorbance of 4-nitroaniline was monitored by spectrophotometry at 405 nm. The amount of pNA was calculated using the molar extension coefficient (ε 405 nm) of 1 × 104 M−1cm−1.

Data analyses

Data were presented as mean ± standard deviation (SD) (n = 3). Statistical comparisons were performed by one-way analysis of variance (ANOVA) and Tukey test. Significant differences were indicated as p < 0.05 or 0.01.

Results

EACA decreases MDA-MB-231 cell proliferation induced by HG and HG-I

With the purpose of determining the total number of cells after the treatments as a measure of proliferation, we carried out MTT and viability assays. HG and HG-I induce a progressive increase in the cell population of MDA-MB-231 cells, at 24 and 48 h (Fig. 1a). At 48 h, HG and HG-I groups exhibited a significant increase with respect to the control group (p < 0.001). We observed that EACA, at the concentrations employed, prevented the cell population growth induced by HG and HG-I, in similar magnitude (p ≤ 0.05) (Fig. 1a). Indicating that the increase in cell proliferation induced by HG and HG-I could be mediated by the interaction of plasminogen with cell surface and plasmin formation. Due to that 12.5-mmol/L EACA showed a marked inhibition of the effect of HG and HG-I, we employed this concentration in the following assays. The percent of cell viability remains high in all experimental conditions, without significant changes after the different treatments (Fig. 1b). These results indicate that EACA prevented the proliferative action of HG and HG-I.

Fig. 1
figure 1

EACA prevents increased proliferation induced by HG and HG-I in MDA-MB-231 cells. Cells were treated for 24 or 48 h with 5.6 mmol/L glucose (Control) or 30 mmol/L glucose (HG) without or with 80 nmol/L human insulin (HG-I), additionally HG and HG-I treatments were combined with 7.5 or 12.5 mmol/L EACA (EACA 7.5 and EACA 12.5, respectively) (a), in b only EACA 12.5 mM was used. As an osmotic control 5.6 mmol/L glucose plus mannitol 24.4 mmol/L was used (Mannitol). The growth of cell population (a) and the percent of cell viability (b) were evaluated by the MTT and Trypan blue exclusion assays, respectively. *p < 0.05, **p < 0.01 or ***p < 0.001 versus control; ## p < 0.01, ### p < 0.001 versus HG and && p < 0.01, &&& p < 0.001 versus HG-I. The mean and standard deviation (X ± SD) of three independent experiments performed by triplicate are shown

Full size image

EACA inhibits the migratory, invasive, and colony-forming activities induced by HG and HG-I in MDA-MB-231 cells

Later, we conducted migration assays to determine the migratory phenotype of the cells. The HG group exhibited a significant increase in migration distance in relation to the control of 877.56 ± 46.28 μm and 533.95 ± 50.50 μm, respectively (p ≤ 0.001). Insulin did not induce an additional increase in migration, with a value of 867.25 ± 38.73 μm, in comparison with HG. At 12.5-mmol/L EACA inhibited the increase in migration induced by HG or HG-I with values of 412.7 ± 38.73 μm and 280.4 ± 17.22, respectively. In the presence of EACA, migratory activities even were significantly lower than those of the control group (Fig. 2a, b) and HG-I group was more sensitive to the action of EACA. The invasive capacity of MDA-MB-231 cells was also induced by HG and HG-I, finding that cells treated with HG and HG-I possessed greater invasive activity. The HG group showed an increase in the number of invading cells (10918 ± 462.5) in relation to the control group (5637 ± 844.50 cells), whereas HG-I group had a higher significant increase (13718 ± 1200 invading cells) compared to all groups. EACA prevented the effects of HG and HG-I reaching similar values to the control: 5200 ± 196.26 and 4609 ± 366.52 invading cells, respectively (p ≤ 0.01) (Fig. 3a). The ability to form colonies indicates the tumor formation capacity of cancer cells [29]. We found an increase in colony formation in the HG and HG-I treatments compared to the control. 12.5 mmol/L EACA prevented the increase induced by HG or HG-I (Fig. 3b, c). These studies indicated that the migration, invasion, and colony formation induced by a HG or HG-I microenvironment are prevented by EACA and therefore they are dependents, at least in part, on the interaction of plasminogen with the cell surface and plasmin formation.

Fig. 2
figure 2

EACA prevents migration induced by high glucose and insulin in MDA-MB-231 cells. Cells were treated for 48 h with 5.6 mmol/L glucose (Control) and 30 mmol/L glucose (HG) without or with 80 nmol/L insulin (HG-I), additionally 12.5 mmol/L EACA (EACA) was combined with HG and HG-I treatments. As an osmotic control 5.6 mmol/L glucose plus mannitol 24.4 mmol/L was used (Mannitol). Migration (a, b) was evaluated by wound healing assay. Images of cell migration (a) and the distance of migration (b) are presented. The bold lines represent the migration border before (time zero) and after all treatments (a). X ± SD of three independent experiments performed in triplicate is shown. ***p < 0.001 versus control; ### p < 0.001 versus HG and &&& p < 0.001 versus HG-I

Full size image
Fig. 3
figure 3

EACA inhibits invasion and formation of colonies induced by high glucose and insulin in MDA-MB-231 cells. Cells were treated with 12.5 mmol/L EACA in medium with HG and HG-I using as control physiological concentration of glucose (5.6 mmol/L) and the osmotic control (24.4 mmol/L mannitol + 5.6 mmol/L glucose) for 48 h. The invasion (a) and the number of colonies (b) were evaluated; and representative images are shown (c). Data are representative of three independent experiments performed by triplicate. *p < 0.05, **p < 0.01, or ***p < 0.001 versus control; # p < 0.05 or ## p < 0.01 versus HG and && p < 0.01 or &&& p < 0.001 versus HG-I

Full size image

EACA prevents EMT in a microenvironment of HG and HG-I

The activation of AKT in different types of human cancer has been associated with increased proliferation, cell survival, and EMT [36]. In a cell microenvironment of HG and HG-I, we found an increase in AKT Ser473 phosphorylation with respect to control group, the increase in HG group was 1.94 ± 0.33 folds in relation to the control; whereas the HG-I group was 1.78 ± 0.21 folds in relation to control. EACA prevented the activation of AKT induced by HG and HG-I. We did not find significant differences between control and mannitol groups (Fig. 4a, b).

Fig. 4
figure 4

EACA prevents EMT induced by HG and HG-I in MDA-MB-231 cells. Cells were treated with 12.5 mmol/L EACA in medium with high glucose (30 mmol/L) (HG) and high glucose + insulin (HG-I) using as control physiological concentration of glucose (5.6 mmol/L) and an osmotic control with mannitol for 48 h. P-AKT Ser 473, AKT, vimentin, E-cadherin, and HCAM were analyzed by Western blot in cell lysates (a, c). The band intensity was analyzed by densitometry, P-AKT was normalized to AKT (b), and HCAM and vimentin were normalized to β-actin (d, e), *p < 0.05, **p < 0.01, ***p < 0.001 versus control; ## p < 0.01 versus HG and && p < 0.01 versus HG-I

Full size image

Different markers of EMT were evaluated by Western blot, including vimentin, E-cadherin, and HCAM. Vimentin showed a significant increase in HG and HG-I groups with respect to control group, indicating the EMT, whereas EACA prevented the increased level of vimentin (Fig. 4c, d). E-cadherin an epithelial marker also was evaluated, loss of its expression has been associated with dedifferentiation and invasiveness in breast cancer during the development of an aggressive phenotype [37]. E-cadherin was only slightly expressed in MDA-MB-231 in all treatments, given that these cells have lost the epithelial phenotype presenting a mesenchymal and invasive phenotype [38] (Fig. 4c). E-cadherin-expressing MCF-7 cells were shown as positive control.

HCAM showed a significant increase in HG and HG-I groups compared to controls. The HG-I group had an increase with respect to all groups, EACA prevented the increase of HCAM induced by a microenvironment of HG and HG-I. Our results suggest that EACA prevented the increased metastatic potential induced by conditions of high glucose and insulin (Fig. 4c, e).

EACA prevents the action of HG and HG-I on the expression of components of the plasminogen activation system and gelatinases

uPA activity is more efficient when it binds to its receptor (uPAR) on the cellular surface. Active uPA, converts plasminogen to plasmin, which can degrade ECM proteins or activate some extracellular MMP, during the invasive process [39]. MMP activity is linked with an advanced stage of breast cancer because of their participation in the invasion of tumor cells.

Therefore, in this work, we evaluated the levels of messenger RNA (mRNA) of plasminogen activation system components and MMP-2 and -9.

mRNA levels for uPA, uPAR, PAI-1 in the HG and HG-I groups increased with respect to the control. Insulin in the presence of the high concentration of glucose induced a significant additional increase in uPA, uPAR and PAI-1 mRNA levels, while these effects of HG and HG-I were prevented by EACA (Fig. 5 a, b, c).

Fig. 5
figure 5

Effect of EACA on mRNA levels of uPA, uPAR, PAI-1, MMP-9, and MPP-2 in the presence of high glucose and insulin. MDA-MB-231 cells were treated with 12.5 mmol/L EACA in medium with HG and HG-I using as control physiological concentration of glucose (5.6 mmol/L) and the osmotic control (Mannitol) for 48 h. Total RNA was subjected to analysis of uPA (a) uPAR (b), PAI-1 (c), MMP-9 (d), and MMP-2 (p) mRNA by real-time RT-PCR and were normalized using the ribosomal protein S18 (RPS18). X ± SD from three independent experiments performed in triplicate. *p < 0.05, **p < 0.01, ***p < 0.001 versus control; ## p < 0.01, ### p < 0.001 versus HG and && p < 0.05 or &&& p < 0.001 versus HG-I

Full size image

The levels of mRNA for MMP9 and -2 in HG and HG-I groups increased compared to control (p < 0.05), which was prevented by EACA (Fig. 5 d, e).

Once the expression of the plasminogen activator system components was determined by means of RT-PCR, we studied the uPA, uPAR, PAI-1 protein levels, finding that these were increased by HG and HG-I treatments in relation to the control. Insulin in the presence of HG induced an additional increase only for uPAR protein but not for Pro-uPA, active uPA, and PAI-1. EACA prevented the increases of Pro-uPA, active uPA, uPAR, and PAI-1 induced by HG and HG-I (Fig. 6a–d, f). Additionally, ENO A was evaluated, because it acts as a plasminogen receptor and thus mediates activation of plasminogen to plasmin by the proteolytic action of uPA. We found increased levels of ENO A in HG and HG-I groups in relation to the controls. Insulin induced a significant increase in ENO A compared to HG. EACA prevented the effects of HG and HG-I, lower levels of ENO A coincide with lower migratory and invasive abilities (Fig. 6e).

Fig. 6
figure 6

EACA reduced uPAR, uPA, PAI-1, and α-enolase protein levels in an environment of HG and HG-I. MDA-MB-231 cells were treated with 12.5 mmol/L EACA in medium with HG and HG-I, using as control physiological concentration of glucose (5.6 mmol/L) and the osmotic control with mannitol for 48 h. Proteins were analyzed by Western blot (f). The intensity of bands of Pro-uPA (a), active uPA (b), uPAR (c), PAI-1 (d), and α-enolase (e) was analyzed by densitometry and normalized to β-actin. X ± SD of three independent experiments performed in triplicate. *p < 0.05, **p < 0.01, ***p < 0.001 versus control;# p < 0.05, ## p < 0.01, ### p < 0.001 versus HG and && p < 0.01 or &&& p < 0.001 versus HG-I

Full size image

In some cases, EACA not only prevented the action of HG or HG-I, also it had and additional inhibitory activity, which was observed for uPA, MMP-9 mRNA, and Pro-uPA, uPAR, and ENO A proteins in HG condition. In HG-I condition EACA had an additional inhibitory activity on uPA, uPAR, and MMP-2 messengers, and on Pro- and active uPA and uPAR proteins.

EACA prevents the increased uPA activity induced by HG and HG-I

In part, the migratory and invasive processes depend on the plasminogen activation system; therefore, we studied the expression and activity of plasminogen activators. By means of casein-plasminogen zymography, we detected the presence of uPA, as a singlet or doublet bands of molecular-weight (MW) close to 55 kDa in cell lysate and conditioned medium. In the conditioned media, there appear two bands with MW around of 80 kDa that do not correspond to plasminogen activators, because they were presents in gels copolymerized with casein, without plasminogen. In the cells lysate, we found an increase in the intensity of the uPA band, in an HG or HG-I microenvironment, of 2.81 ± 0.15 and 2.9 ± 0.06 folds with respect to the control group (p ≤ 0.001), while EACA prevented the increase of uPA activity induced by HG or HG-I reaching similar values (0.86 ± 0.22 and 0.87 ± 0.15 folds) in relation to the control group (p > 0.05) (Fig. 7a, b). In the conditioned medium, the HG and HG-I groups presented increased intensity of the band of uPA activity with a value of 1.90 ± 0.25 and 1.687 ± 0.38 times respect to the control; however, EACA even significantly diminished the intensity of the uPA band in relation to the control group (p < 0.001) (Fig. 7c, d). Subsequently, we determined uPA activity by a quantitative amidolytic assay. uPA converts the zymogen plasminogen to active plasmin, capable of degrading components of the extracellular matrix. We found a higher enzymatic activity of uPA in the HG and HG-I groups, whereas EACA prevented the increased uPA activity induced by HG and HG-I (Fig. 7e).

Fig. 7
figure 7

EACA prevents uPA increased activity induced by HG and HG-I. MDA-MB-231 cells were treated by 48 h with 12.5 mmol/L EACA in serum-free medium with HG and HG-I, using as control physiological concentration of glucose (5.6 mmol/L) and the osmotic control with mannitol for 48 h. uPA activity was assessed by casein-plasminogen gel zymography in cell lysates (a) and conditioned medium (c) followed by densitometric analysis (b, d). Additionally, uPA activity was measured using an amidolytic assay and the activity of plasmin formed is expressed as the formation rate of p-nitroaniline (e). X ± SD from three independent experiments performed in triplicate. ***p < 0.001 versus control; # p < 0.05, ## p < 0.01; ### p < 0.001 versus HG, and &&& p < 0.001 versus HG-I

Full size image

Discussion

Preclinical investigations indicate that the plasminogen activation system plays a vital role in tissue remodeling during metastasis, though the breakdown of basal membranes and ECM, together with the activation of latent growth factors [12, 40]. Some members of this proteolytic system (plasmin, uPA-uPAR complex, PAI-1) activate signaling pathways implicated in proliferation, migration, invasion, metastasis, and angiogenesis [41]. Plasminogen receptors at the cell surface promote the focalized and efficient formation of plasmin in the majority of tumors, and its enhanced expression is correlated with bad prognosis and reduced possibilities of survival of cancer patients [42] some of these receptors have been considered as being involved in tumorigenesis and suggested as therapeutic targets, including annexin 2 [43, 44], ENO A [19, 45,46,47], cytokeratin 8 [42, 48], histone H2B [13, 49], and recently, a new receptor, identified as Plg-RKT, which is an integral membrane protein [13, 50].

Currently, some modulators of plasminogen activation, synthetic lysine derivatives, such as EACA and tranexamic acid inhibit fibrinolysis by competing with lysine binding sites from plasminogen. In this manner, they inhibit the binding of plasminogen to the cell surface and its conversion to plasmin [51]. The inhibition of fibrinolytic enzymes inhibited tumor growth and metastasis in different types of cancer, such as prostate carcinoma and melanoma. EACA as an inhibitor of plasmin formation at the cell surface, prevented in 37% the growth of tumors derived of a glioma xenograft [52].

In a previous study, we found that in a microenvironment of HG and HG-I, a greater migratory and invasive capacity are promoted in breast cancer cells that are associated with EMT and upregulation of uPA, uPAR, and PAI-1, in addition to a greater concentration of active uPA [27].

High levels of uPA, uPAR, and PAI-1, as well as other components of the system have been correlated with a poor prognosis in breast cancer [53]. uPA is overexpressed in breast cancer and in other cancer types, where it is regulated by IGF-1R through PI3K [54, 55] and MAPK pathways [56]. uPA in turn induces PI3K activation that, in conjunction with GTPase proteins such as Rho, Cdc42, and Rac1, are key effectors that regulate cellular migration by dynamic changes in the actin cytoskeleton [57]. It was expected that PAI-1, the inhibitor of uPA can to prevent invasion and metastasis. However, controversial studies have demonstrated that PAI-1 promotes, and does not inhibit, invasion and metastasis. For example, co-expression of uPA and PAI-1 has been associated with poor prognosis of breast cancer [58]. In the other hand, the deficiency of PAI-1 in mice reduced angiogenesis and invasion in cancer cells [59, 60]. In this study, we confirmed that in the HG and HG-I groups, there was an increase in uPA, uPAR, and PAI-1 expression, associated with EMT and an enhanced invasive phenotype in MDA-MB-231 breast cancer cells, while EACA prevented the upregulation of uPA, uPAR, and PAI-1, the EMT and the increased invasiveness induced by HG and HG-I, inhibiting the aggressive phenotype of these cells.

This study described, for the first time, that EACA, under hyperglycemic and hyperinsulinemic microenvironments, leads to the inhibition of proliferative, migratory, and invasive activity of triple-negative and invasive MDA-MB-231 breast cancer cells, which were associated with prevention of EMT and the inhibition of the plasminogen activation proteolytic system, downregulating the expression of uPA, uPAR, and PAI-1. EACA could act as anti-tumorigenic and anti-metastatic agents under conditions of hyperglycemia and hyperinsulinemia, and they could be used to prevent the progression of breast cancer in diabetic patients.

Increased plasminogen binding is associated with metastatic breast cancer cells; metastatic MDA-MB-231 breast cancer cells bound more plasminogen in a lysine-dependent manner than non-metastatic MCF-7 cells [15]. EACA prevents the binding of plasminogen to the cell surface competing with carboxy-terminal lysine of several of its receptors [17, 42], some of them have been described in breast cancer cells, including ENO A [20], annexin 2 [61], cytokeratin 8 [48, 62], and Plg-RKT [63]. Our data indicate that the aggressive phenotype induced by HG and HG-I is associated at least in part, with increased surface binding of plasminogen. The identity of the plasminogen receptor or receptors involved in the action of HG and HG-I remain to be elucidated. We detected increased expression of ENO A after HG and HG-I treatments, which was prevented by EACA. The role of ENO A as a plasminogen receptor in cancer cells has been widely documented, ENO A acts as a key protein in the promotion of cellular metabolism under anaerobic conditions, and favored tumor invasion through the promotion of activation of plasminogen and ECM degradation [19, 47, 64]. These data suggest that ENO A can to promote the binding of plasminogen at the cell surface and its transformation to plasmin in hyperglycemic and hyperinsulinemic conditions.

In gliomas, it was demonstrated that suppression of ENO A significantly inhibits PI3K and AKT phosphorylation, inhibiting cellular growth and the EMT [65]. These results suggest that EACA could regulate this signaling pathway, preventing the development of the migratory and invasive phenotype. In our study, we found that EACA prevented the activation of AKT (lower level of AKT Ser473 phosphorylation) induced by HG and HG-I, which was associated with the inhibition of proliferation, migration, EMT, and invasion in a high glucose and insulin rich environment.

EMT is a hallmark of tumor progression to a metastatic phenotype resulting in the loss of adhesion between cells and the acquisition of cell mobility and invasive ability [66]. The epithelial marker, E-cadherin, and the mesenchymal marker, vimentin, are important markers of EMT in breast cancer [67,68,69]. Migratory and invasive cancer cells exhibit decreased expression of E-cadherin and increased expression of vimentin that have been correlated with the malignancy and lymph nodes metastasis [37, 68]. E-cadherin, a calcium-regulated adhesion molecule, have a role in epithelial differentiation and have been considerate as tumor suppressor [70, 71], MDA-MB-231 cells have lost the expression of E-cadherin [38], and we did not find changes in its expression in all experimental groups; however, vimentin was induced by HG and HG-I treatments. Vimentin is an important component of the cytoskeleton and is widely distributed in mesenchyme [72] [69]. The overexpression of vimentin in breast cancer cells is correlated with, high tumor invasiveness, chemoresistance, and poor prognosis [73]. HG and HG-I groups showed upregulation of vimentin suggesting the induction of EMT and the rise of a migratory, invasive, and aggressive phenotype. Our finding indicate that EACA prevents the increase in vimentin induced by HG and HG-I, and in this manner it reduces the metastatic potential of MDA-MB-231 cells.

Another molecule that was measured was the hyaluronan receptor HCAM, also known as CD44, a member of a family of cell adhesion molecules that participates in changes in cell–cell and cell–matrix interactions, cell signaling, and in tumor development, particularly during invasion and metastasis [74]. HCAM regulates metastasis, including transformation, growth, cellular invasion, mobility, and chemoresistance, in addition to being a cancer stem cell marker [51, 75]. Thus, the complexity of this molecule permits the activation of multiple signaling pathways including Rho GTPases [76], RAS-MAPK, and the PI3K/AKT pathway [77]. In our work, HCAM was found to be highly expressed in the HG and HG-I groups, favoring an aggressive phenotype in these cells. Contrariwise, EACA importantly inhibited the expression of HCAM in groups treated with HG and HG-I.

These results indicate that in cultured invasive MDA-MB-231 breast cancer cells in a high glucose microenvironment, the interaction of plasminogen to the cell surface and its transformation to plasmin affected multiple aspects associated with tumor progression, including proliferation, migration, clonogenicity, invasion, EMT, and the expression of several genes of the plasminogen activation system. This conclusion was supported by the prevention of these multiple actions of HG and HG-I by EACA. The mechanism why EACA prevented the actions of HG and HG-I requires more studies. EACA can deplete cell surface-bound plasminogen and plasmin, interfering with ECM degradation and activation of several growth factors and cytokines that regulate the activity and gene expression of cancer cell [78]. Additionally, they can to prevent the cell signaling induced by the plasmin binding at cell surface [79]. Our data support, indirectly, this possibility, given that EACA prevents the increase of ENO A protein level induce by HG and HG-I and it has been demonstrated that ENO A gene expression is stimulated by plasminogen/plasmin in fibroblasts and peripheral blood cells [79].

Recently, the participation of plasminogen binding and activation at cell surface in the metastatic behavior of MDA-MB-231 breast cancer cells have been supported, hence the cleavage of carboxyl terminal lysines by active thrombin activatable fibrinolysis inhibitor (TAFIa) attenuates migration and invasion of this cells through the inhibition of plasminogen activation [80].

In conclusion, HG and HG-I induced a series of events leading to an aggressive phenotype in MDA-MB-231-breast cancer cells, which were mediated at least in part, by the binding of plasminogen to the cell surface, which is suggested because EACA prevented the action of HG and HG-I in these cells. The binding of plasminogen to the cell surface, promoted the activation of plasminogen, EMT, a migratory phenotype and thereby invasiveness and metastasis. These events could be mediated by the crosstalk between different oncogenic signaling pathways, including PI3K/AKT, tyrosine kinase receptors (TRK) and MAPK that promoted tumor progression and metastatic-invasive phenotype of breast cancer cells. The inhibition of the binding of plasminogen to the cell surface constitutes a potential therapeutic strategy to prevent the promotion of formation and progression of tumors by hyperglycemia and hyperinsulinemia.