Total Smad2, phosphorylated Smad2, α-SMA, vimentin, cytokeratin, TGF-β1 and E-cadherin antibodies, as well as secondary antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). The transwell chamber was obtained from Corning Life Sciences (NY, USA). DMEM and fetal calf serum were purchased from Gibco-BRL (Carlsbad, CA, USA). Human TGF-β1 was obtained from Sigma (St. Louis, MO, USA). Other laboratory reagents were obtained from Sigma.

Two human breast cancer cell lines, MCF-7 and MDA-MB-435S, were obtained from the Cancer Research Institute of Beijing, China. These cells were cultivated in T75 tissue culture flasks in DMEM supplemented with 10% fetal calf serum, 100 IU/ml penicillin, 100 μg/ml streptomycin, 2 mM L-glutamine, and 20 mM hydroxyethyl piperazine ethanesulfonic acid, and incubated in a humidified incubator containing 5% CO₂ at 37°C.

Breast tumor samples from 128 patients who underwent surgery in the Affiliated Hospital of Qingdao University Medical College between March 2011 and October 2011, were studied. The local institutional review board approved our protocol for use of patient samples; all patients provided written informed consent prior to participation in the study. Three 5-μm frozen sections were taken from these specimens and stained with hematoxylin and eosin (H&E). The original histologic diagnosis used for clinical management was confirmed by three independent, expert histopathologists. Cases were selected based on a confirmed diagnosis of breast cancer and the presence of both an invasive margin and a central tumor area on the same section. Several criteria to verify the presence of an invasion front were assessed. H&E was used to identify single cells, which appeared to be at the tumor edge. Samples with invasion fronts identified on H&E criteria were then subjected to immunohistochemistry using i) MNF116, a pan cytokeratin marker (1:50) to identify epithelial cells, and ii) CD45, a hematopoietic cell marker (1:100) to confirm that cells identified as possibly invasive were not in fact inflammatory.

Eighteen carcinoma samples from patients that fulfilled these stringent criteria were selected, and a further thirty 5-μm sections of these samples were then cut. An additional H&E section was stained, and the remaining sections were used for immunohistochemistry. Tissue was fixed with acetone for 10 min and rehydrated with ethanol. Slides were blocked with 10% horse serum for 1 h at room temperature and incubated with cytokeratin (1:50), E-cadherin (1:100), vimentin (1:200), and TGF-β1 (1:50) antibodies overnight at 4°C. Slides were incubated with anti-mouse biotinylated secondary antibody (1:200). Finally, detection was carried out with the DAB kit according to the manufacturer’s instructions. Each slide had at least three replicate sections for each antibody. Quantitative analysis of cytokeratin and vimentin staining in the central tumor area compared with the invasive margin was done by two independent observers. The invasive front was identified and the area of central tumor most distant from this invasive front was then selected, and three independent fields were scored for staining intensity and for cellular localization of cytokeratin staining. The intensity of staining was scored as 0 (none), 1 (weak), 2 (mild), 3 (moderate), and 4 (strong) compared with a negative (no primary antibody) and positive control. The cellular localization of cytokeratin was classified as membranous, cytoplasmic or mixed.

The phenotypic changes of breast cancer cells were determined by phase contrast microscopy. The cancer cells in cultures treated with recombinant TGF-β1 and left untreated (control) for 72 h and morphological changes were visualized by phase contrast microscopy. The images were collected using Nikon inverted microscope.

Breast cancer cells were cultured on a 6-well tissue culture plate to confluence. The cell were treated with recombinant human TGF-β1 at the time of switching to serum-free medium, at a final concentration of 5 ng/ml. The cancer cells cultured without TGF-β1 were considered as control. The cells were harvested at 72 h. Total cellular protein was extracted using a lysis buffer and quantified using protein quantification reagents from Bio-Rad. Next, 60 μg of the protein was suspended in 5× reducing sample buffer, boiled for 5 min, electrophoresed on 10% SDS-PAGE gels, and then transferred to polyvinylidene difluoride membrane by electroblotting. The membrane was blocked in 1% BSA/0.05% Tween/PBS solution overnight at 4°C, followed by incubation with the primary antibody (mouse monoclonal antibodies to either human α-SMA, vimentin, cytokeratin, E-cadherin, phosphorylated-Smad2, or Smad2) for 24 h. A horseradish peroxidase-labelled goat anti-mouse IgG was used as the secondary antibody. The blots were then developed by incubation in a chemiluminescence substrate and exposed to X-ray film.

The breast cancer cells were grown to a 70% confluence on culture dishes and the transient transfection was performed with specific stealth small interference RNA against Smad2, or control siRNA over- night using Lipfectamine-2000, according the manufacturer’s instructions. The total of two siRNA sequences for Smad2 and control-siRNA were designed and synthesized from Invitrogen using RNAi designer software program. The concentration of 300 nM was determined to be the most effective siRNA concentration for Smad2 silencing. The transfection medium was changed with culture medium containing 5% FBS for 24 h. TGF-β1 at final concentration of 5 ng/ml was added to the cell cultures in serum-free medium or without TGF-β1 (control). The cells were harvested at 4, 24 and 72 h for further experiments.

In vitro invasiveness was measured by the method described in the study by Albini et al(19), with some modifications. We used chemotaxis chambers with a 8 μm-pore membrane filter coated with 50 mg of matrigel in a 24-well culture plate. MDA-MB-435S cells were pretreated with TGF-β1 or Smad2-siRNA, and plated at a concentration of 3×10⁵/ml per upper well in 200 μl of serum-free medium. As a chemoattractant, 10% fetal calf serum medium was used in the lower chamber. After being recultured with 5% CO₂ at 37°C for 24 h, the filters were removed, fixed in 95% alcohol, and stained with trypan blue. The cells remaining on the top surface of the membrane were completely removed with a cotton swab, and the membrane was removed from the chamber and mounted on a glass slide. The number of infiltrating cancer cells were counted in five regions selected at random, and the extent of invading cancer cells was determined by the mean count.

All values in the text and figures are presented as mean ± SD. In univariate analysis, 2-tailed χ² tests for categorical variables and 2-tailed t-test for continuous variables were used for statistical comparisons. Values of P<0.05 were taken to show a significant difference between means.

Immunohistochemical analysis of the invasive component of carcinoma specimens was compared with the central area of the tumor. There was downregulation of the intensity of the epithelial staining with E-cadherin and cytokeratin at the invasive tumor margin. Furthermore, there was a redistribution of E-cadherin staining from a predominantly membranous pattern in the central area to a predominantly cytoplasmic pattern at the margin. In contrast, a low intensity of vimentin staining was observed in the center of the tumor compared with increased intensity of expression at the invasive margin. Quantification of cytokeratin and vimentin staining intensity showed downregulation of cytokeratin at the invasive margin (P<0.005) and a contrasting upregulation of vimentin staining (P<0.05). Staining for TGF-β1 showed a predominantly stromal expression pattern in both the central and invasive tumor components with foci of increased uptake in the invasive front (Fig. 1).

In the absence of TGF-β1, small portions of cell morphology were somewhat mesenchymal, but most breast cancer cells showed pebble-like shape and tight cell-cell adhesion. However, after TGF-β1 treatment for 72 h, MCF-7 and MDA-MB-435S cells showed morphological changes assessed by phase contrast microscopy. Many cells assumed more elongated shape and lost contact with their neighbor, displaying fibroblast-like appearance compared to the untreated cells (Fig. 2).

To better confirm morphological changes in breast cancer cells, MCF-7 and MDA-MB-435S cells, represent EMT, western blotting was used to examine the changes of EMT-related protein markers. The results indicated that the expression of cytokeratin and E-cadherin, the epithelial phenotype marker, was significantly decreased in MDA-MB-435S cells, while TGF-β1 did not affect the E-cadherin expression levels in MCF-7. Those of mesenchymal phenotype markers, α-SMA and vimentin, were greatly increased in MCF-7 and MDA-MB-435S cells (Fig. 3).

Here we showed that Smad2 phosphorylation was increased by TGF-β1 in MCF-7 and MDA-MB- 435S, while TGF-β1 did not affect the total Smad2 expression levels. In order to confirm whether Smad2 is involved in TGF-β1 mediated EMT of breast cancer, siRNAs were used to knockdown the Smad2 gene in MDA-MB-435S. As shown in Fig. 4B, siRNAi-Smad2#1 highly significant knockdown for Smad2 and phosphorylated Smad2 when compared to the other siRNAs or the control.

After silencing Smad2 by using siRNA-Smad2 or control-siRNA in cancer cells treated with TGF-β1, we noted a remarkably reduced expression of vimentin, and most importantly a significant restoration of the junctional protein cytokeratin suggesting a role for Smad2 signaling in EMT of cancer cells (Fig. 5).

We analyzed the invasion capability of the highly metastatic MDA-MB-435S cells using the methods described above. The results showed that TGF-β1 significantly promoted the invasiveness of cells when compared to control (P<0.05), While Smad2-siRNA significant decreased the number of invasive cells under TGF-β1 stimulation (P<0.05). These results suggested EMT of breast cancer induced by TGF-β1 promote cancer cells metastasis, but knockdown of the Smad2 gene, by silencing siRNA reduced the invasion of gastric cancer cells can partially inhibit these effects (Fig. 6).