Overexpression of matrix metalloproteinase‐9 in breast cancer cell lines remarkably increases the cell malignancy largely via activation of transforming growth factor beta/SMAD signalling

Abstract Objectives Matrix metalloproteinase 9 (MMP‐9) has been frequently noticed in the breast cancers. In this study, we aim to investigate the associations of MMP‐9 with the activation of transforming growth factor beta (TGF‐β)/SMAD signalling and the malignancy of breast malignant tumour cells. Materials and methods The distributions of MMP‐9 and TGF‐β in the tissues of canine breast cancers were screened by immunohistochemical assays. A recombinant plasmid expressing mouse MMP‐9 was generated and transiently transfected into three different breast cancer cell lines. Cell Counting Kit‐8 and colony formation assay were used to study cell viability. Migration and invasion ability were analysed by wound assay and transwell filters. Western blot and quantitative real‐time PCR were used to determine the protein and mRNA expression. Result Remarkable strong MMP‐9 and TGF‐β signals were observed in the malignant tissues of canine breast cancers. In the cultured three cell lines receiving recombinant plasmid expressing mouse MMP‐9, the cell malignancy was markedly increased, including the cell colony formation, migration and epithelial‐mesenchymal transition. The levels of activated TGF‐β, as well as SMAD4, SMAD2/3 and phosphorylation of SMAD2, were increased, reflecting an activation of TGF‐β/SMAD signalling. We also demonstrated that the inhibitors specific for MMP‐9 and TGF‐β sufficiently blocked the overexpressing MMP‐9 induced the activation of SMAD signalling and enhancement on invasion in the tested breast cancer cell lines. Conclusion Overexpression of MMP‐9 increases the malignancy of breast cancer cell lines, largely via activation of the TGF‐β/SMAD signalling.

Despite unknown aetiology, several risk factors have been considered to be significantly associated, including hormonal, nutritional and genetic events. 2,[4][5][6][7] The prognosis of breast cancer depends on many clinical and pathological issues, such as clinical stage, histopathological pattern, tumour size, lymphatic or vascular invasion and distant metastases. 2,4,8,9 In the past decades, a number of cellular factors have been proposed to be potential for the prognosis of breast cancers, particularly human breast carcinomas, covering miRNAs, oncoproteins, mutations in several special genes, cancer stem cells and circulating tumour cells. 10 Kaszak and the colleagues have reviewed the frequently addressed biomarkers of canine breast cancers, such as Ki-67, PCNA and p53 for cancer cell proliferation and apoptosis, E-cadherin, CEA and CA 15-3 for metastatic potential and VEGF, EGFR and HER-2 for angiogenesis. Because of the similarity between CMCs and human breast cancers, those biomarkers for humans may be also useful for the prognosis of CMCs.
Matrix metalloproteinases (MMPs) are a family of zinc-dependent proteases; among them, MMP-9 is one of the major members of MMP family. 11,12 MMP-9 protein is mainly secreted by tumour cells and stromal cells, which usually present in the form of zymogen. Via the process of hydrolysis, the activated MMP-9 is able to degrade basement membrane (BM) type IV collagen, which is believed to affect the ability of BMs to impede tumour cell movement. 13 As type IV collagens are the main components of the extracellular matrix and BMs, tumour-derived MMP-9 may destroy these tissue barriers and enhance the invasion and metastasis of tumour cells. Actually, increased expression of MMP-9 has been documented in many different histological types of human malignant tumours and their metastasis, such as oral, larynx, gastric, lung, liver, breast, bone, skin and cervical. 11,[13][14][15][16][17][18][19] Therefore, MMP-9 and its family members, such as MMP-2, are considered as the prognostic biomarkers for various carcinomas.
Transforming growth factor beta (TGF-β) is a multifunctional cytokine belonging to the transforming growth factor superfamily. Activated TGF-β complexes, together with other factors, form a serine/threonine kinase complex that binds to TGF-β receptors and further activate different downstream substrates and regulatory proteins, mainly the SMAD and DAXX pathways, inducing transcriptions of various target genes involving in cell differentiation, proliferation, chemotaxis and activation of many immune cells. 20,21 Increased expression of TGF-β often correlates with the malignancy of many cancers. Activation of TGF-β is induced by many elements, including pH, reactive oxygen species, thrombospondin-1, protease and metalloprotease; among them, MMP-9 and MMP-2 are known to able to cleavage the latent TGF-β. 22 However, the exact association between overexpression of MMP-9 and activation of SMAD pathway is still not clear.
In this study, we screened the expressions of MMP-9 and TGF-β in the samples of canine breast cancers and confirmed overexpression statues of those two proteins in the malignant tissues. Transient expression of mouse MMP-9 in three different breast carcinoma cell lines increases the cell colony formation and migration and promotes epithelial-mesenchymal transition (EMT). Apparent activation of SMAD signalling, represented by increased expressions of SMAD4, SMAD2/3 and phosphorylation of SMAD2, was noticed in the cell lines expressing MMP-9. We also demonstrated that inhibition of the activity of either MMP-9 or TGF-β sufficiently blocked the MMP-9 overexpression induced the activation of SMAD signalling and enhancement on invasion in the breast cancer cell lines.

| Immunohistochemical (IHC) analysis
All the tumour tissues from canine were obtained from China Agricultural University Animal Hospital. Tumours were surgically dissected and fixed in 10% (v/v) neutral buffer formalin for 5 days.
The fixed tissues were routinely dehydrated in ascending grades of ethanol and xylene, and then embedded in paraffin wax. Sections (5 μm) were prepared with microtome (Leica) and mounted on CITOGLAS ® adhesion microscope slides (CITOTEST).

| Transfection
Extraction of the recombinant plasmid pMMP-9-HA DNA was per-

| Migration assay
Wound assay was performed to evaluate the migration ability of cells. Cells were seeded in six-well plate and grew to confluence fol-

| Invasion assay
Eight-micrometer pore-size transwell filters (Costar, Corning Incorporated) were put in 24-well plate, and cells were seeded onto the filters at a concentration of 1 × 10 4 cells/well in 100 μL of FBS-free medium and then transfected with plasmid MMP-9-PCMV-HA. The lower chambers were filled with 600 μL of medium with 10% FBS. TGF-beta (50 pmol/L) or SB431542 (10 μmol/L) was added into each well 24 hours post-transfection.
Thirty-six hours after treatment, cells on the topside of the filter were removed by scrubbing with a tipped swab. The migration of cells to the lower side of the filter was determined by crystal violet staining. Each testing group contained at least three independent wells.

| Western Blot analysis
For the preparation of cell lysates, cells were washed twice with ice-cold PBS and homogenized with RIPA buffer containing a cocktail of protease and phosphatase inhibitor (Sigma-Aldrich) for 20 minutes on ice. Afterwards, samples were sonicated for 20 seconds and centrifuged at 12 000 g at 4°C for 20 minutes.
The supernatants were collected and boiled for 10 minutes in loading buffer (250 nmol/L Tris-HCl 6.8 pH, 10% sodium dodecyl sulphate, 0.5% bromophenol blue, 50% glycerol and 0.5 mol/L dithiothreitol). Equal amounts of protein were separated by 12% or 10% sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene difluoride membranes (Millipore). After blocking with 5% skim milk in Trisbuffered Saline Tween-20, membranes were incubated with the individual primary antibodies at 4°C overnight. Membranes were rinsed three times in TBST and then incubated with different HRP-labelled secondary antibodies at 37°C for 60 minutes.
Signals were developed using an enhanced chemiluminescence detection kit (Bio-Rad).  Table 1

| Statistical analysis
Statistical comparisons between groups were performed using ANOVA (Prism GraphPad 5 Software). All data are presented as mean ± SD of three independent experiments. Significant differences were assigned at P < 0.05, <0.01 and <0.001, denoted by *, ** and ***, respectively.  ( Figure 2C). Besides, no significant differences in colony formation and migrating ability were figured out among the preparations of those three tumour cell lines.

MMP-9
Malignant Benign TA B L E 2 Immunohistochemical assays for matrix metalloproteinase (MMP)-9 and transforming growth factor beta (TGF-β) in 34 malignant breast cancers and 18 benign breast tumours

| Overexpression of MMP-9 promotes epithelial-mesenchymal transition (EMT) of the cultured breast cancer cells
To address the possible alterations of the essential agents related to

| Overexpression of MMP-9 induces upregulation for the expression and phosphorylation of SMAD proteins
To test the possible influence on the expression and phosphorylation of SMAD proteins, three breast cancer cell lines were transiently transfected with plasmid pMMP-9-HA for 48 hours. The harvested cells were subjected to total RNA extraction and cellular lysate preparation.
qRT-PCR assays for the cellular genes SMAD2, SMAD3 and SMAD4

| Inhibition of TGF-β or MMP-9 activity abolishes the enhancing effect of overexpression of MMP-9 on SMAD signal pathway in the cultured breast cancer cells
SB431542 is a selective inhibitor of the TGF-β1 receptor ALK5, whose molecule weight is 384.39. 24 To address the association of were relatively lower. SMAD-specific Western blots of the cellular lysates revealed the similar profiles ( Figure 6B). The cells overexpressing MMP-9 and exposed to TGF-β simultaneously contained remarkably strong SMAD2/3, SMAD4 and p-SMAD2. Most of the preparations treated with pMMP-9-HA and SB431542 showed the relatively lower levels of the tested SMAD proteins than those of the cells receiving pMMP-9-HA alone. Similarly, the levels of SMAD proteins in the cells treated with SB431542 alone were low, which were comparable with that of the mock or even lower in some reactions.

F I G U R E 4
Alterations of the cellular transforming growth factor beta (TGF-β) and its isoform (complexes) in the cells transfected with pMMP-9-HA. Western blots for TGF-β. Cells were harvested 48 h post-transfection. Quantitative assays of the relative grey values of various TGF-β isoforms after normalized with the data of GAPDH are shown. Each test was repeated for three times. Graphical data denote mean ± SD. Statistical analysis was performed using two-way ANOVA. ***P < 0.001, **P < 0.01 and *P < 0.05 were considered significantly different

| Inhibition of TGF-β activity in breast cancer cells blocks the MMP-9-induced enhancement on invasion
To test the influence of overexpressing MMP-9 in breast cancer cells on their invasion abilities and figure out the relevant molecular mechanism, assays of transwell filters were conducted. Compared F I G U R E 5 Analyses of the changes in cellular SMADs in the cells transfected with pMMP-9-HA. Cells were harvested 48 h posttransfection. A, qRT-PCR assays. The total RNA was prepared, and the transcriptional levels of various SMAD genes were evaluated with the individual qRT-PCRs. Y-axis represents the values of 2 −ΔΔCT . Each test was repeated for three times. Graphical data denote mean ± SD. B, Western blots. The cellular levels of various SMADs were determined with the individual Western blots. Quantitative assays of the relative grey value of each SMAD after normalized with the data of the individual GAPDH are shown. Each test was repeated for three times. Graphical data denote mean ± SD. Statistical analysis was performed using two-way ANOVA. ***P < 0.001, **P < 0.01 and *P < 0.05 were considered significantly different with the observations in the mock controls, more numbers of cells were seen in the preparations transfected with pMMP-9-HA, indicating that overexpression of MMP-9 increases the invasion abilities of various breast cancer cells (Figure 7). Transfection with pMMP-9-HA together with exposure to TGF-β induced even more numbers of cells migrating to the lower side of the filter (Figure 7). In contrary, the numbers of the cells migrating to the lower side of the filter were obviously less in the preparations receiving pMMP-9-HA but exposing to SB431542, which looked similar as the mock (Figure 7).
It suggests the enhanced invasion ability of the breast cancer cells by overexpressing MMP-9 relies largely on the activation of TGF-β/ SMAD signal pathway.

| D ISCUSS I ON
In this study, we have confirmed again markedly increased expres- A, qRT-PCR assays. Y-axis represents the values of 2 −ΔΔCT . Each test was repeated for three times. Graphical data denote mean ± SD. B, Western blots. The cellular levels of various SMADs were determined with the individual Western blots. Quantitative assays of the relative grey value of each SMAD after normalized with the data of the individual GAPDH are shown. Each test was repeated for three times. Graphical data denote mean ± SD. C, Assay of the influence of MMP-9 inhibitor. The pMMP-9-HA and/or MMP inhibitor GM6001 were added to MCF7 cells and harvested 48 h post-transfection. The cellular levels of various SMADs and TGF-β were determined with the individual Western blots. Quantitative assays of the relative grey value of each SMAD and TGF-β after normalized with the data of the individual GAPDH are shown right. Statistical analysis was performed using two-way ANOVA. ***P < 0.001, **P < 0.01 and *P < 0.05 were considered significantly different F I G U R E 7 Influences of the transforming growth factor beta (TGF-β) inhibitor on the invasion abilities of the cells transfected with pMMP-9-HA. Representative images of cells in the lower side of the filter. Cells were exposed to recombinant TGF-β and/or the inhibitor SB 431542 12 h post-transfection. The migration of cells to the lower side of the filter was determined by crystal violet staining. Each testing group contained at least three independent wells. Statistical analysis was performed using one-way ANOVA. ***P < 0.001, **P < 0.01 and *P < 0.05 were considered significantly different enrolled canine breast cancers in this study are infiltrative ones.
Higher levels of MMP-9 and MMP-2 have been repeatedly described in the malignant tissues of the patients with infiltrative breast cancer and lymph node metastasis. 11,[26][27][28] Our data here illustrate that transient overexpression of MMP-9 in the breast cancer cell lines strongly enhances the cellular malignant characteristics in vitro, such as the cell colony formation, migration and EMT. Along with these changes, an alteration of the cellular TGF-β profiles in the cells overexpressing MMP-9 is noticed. It is known that TGF-β is synthesized as precursor molecules containing a propeptide region in addition to the TGF-β homodimer. TGF-β homodimer interacts with a latency-associated peptide (LAP) and forms SLC. SLC remains in the cell until it binds with another protein called latent TGF-β-binding protein (LTBP) and forms a large complex LLC that gets secreted to the extracellular matrix (ECM). 23,29 Our data here have revealed that not only the activated form of TGF-β but also cellular TGF-β and SLC increase in the cells expressing MMP-9. It indicates that overexpression of MMP-9 in the transfected cells may not only promote the proteolysis of TGF-β, but also influence the expression of TGF-β.
The observations of stronger TGF-β deposits in the MMP-9-positive areas in the malignant tissues also propose the probable correlation between those two elements during carcinogenesis. We also find that the levels of LLC maintain almost unchanged, even slightly declined, after challenge of MMP-9. As LLC is usually secreted into ECM, the lysates for LLC are carefully prepared without PBS washing in order to avoid the potential loss of LLC during the harvesting process. Those diversity profiles of TGF-β complexes in the breast cancer cells during overexpression of MMP-9 deserve further study.
The activated TGF-β binds to TGF-β type II receptor (TGF-βRII) and further phosphorylates TGF-β type I receptor (TGF-βRI). The binding of TGF-βRII and TGF-βRI promotes the downstream signal- The functions of the TGF-β/SMAD signalling pathway in cancer seem to be paradoxical. Such phenomenon has been described for F I G U R E 8 A hypothetical processing schema of the matrix metalloproteinase (MMP)-9-mediated activation of transforming growth factor beta (TGF-β)/SMAD pathway in malignant and benign tumours. In malignant tumour cells, more MMP-9 molecules are expressed, which induce the more active releases of SLC and TGF-β from the latent form and phosphorylation of TGF-β receptors and SMAD2/3 long time. 32 The suppressive role of this signalling in cancers reflects in inhibition on tumour formation and inducement of growth arrest and apoptosis. 33,34 On the other hand, TGF-β/SMAD signalling shows the activity to promote tumour progression and metastasis, by inducing angiogenesis, inflammation and EMT. 35 In summary, more aggressive expressions of MMP-9 and TGF-β were detected in the malignant canine breast cancers. Overexpression of MMP-9 in the breast cancer cell lines increased the malignancy in vitro, which is likely associated with the activation of TGF-β/SMAD pathway. As illustrated in Figure 8, we hypothesize that in malignant breast cancers, more MMP-9 are expressed, which leads to more releases of TGF-β from its latent form. More activated TGF-β further induces phosphorylation of TGF-β receptor, activates the TGF-β/ SMAD signalling and eventually increases the malignancy of cancer.