Plasma from 16 normal controls, 20 patients with fibroadenoma and 20 breast cancer patients (19 patients with infiltrating ductal carcinoma and 1 patient with invasive lobular carcinoma) was obtained from the Department of Breast Surgery, The First Affiliated Hospital of Zhengzhou University (Henan, China), from September, 2013 to May, 2014. The subjects included in the three groups voluntarily joined the present study after providing informed consent. The experimental procedure was carried out in accordance with the Declaration of Helsinki and was approved by the Medical Research Ethics Committee of the First Affiliated Hospital of Zhengzhou University (approval no.: 2013-MR-0045). The miR-21 reverse transcription (RT) primer and RT-qPCR primers used were described in a previous study (31). miRNAs were isolated from 50 ml of plasma using a mirVana miRNA isolation kit (Applied Biosystems, Foster City, CA, USA). RT-qPCR was performed to measure the expression of mature miR-21 in cells, as previously described (31). miR-21 levels were normalized to U6 levels.

For mRNAs, total RNA from cells was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RT-qPCR reactions were carried out as previously described (31). β-actin was used as an endogenous control. MCF-7 cells were cultured as previously described (31).

The LightSwitch smad7 3′-UTR reporter (Switchgear Genomics, Menlo Park, CA, USA) was utilized to assess whether there is a direct interaction between miR-21 and smad7 3′-UTR. Site-directed mutagenesis kits (Stratagene, La Jolla, CA, USA) were used to introduce a three-base pair mutation in the seed-binding site of smad7 3′-UTR. The pCMV-SPORT6-smad7 cDNA construct used in the smad7 rescue experiment was purchased from Open Biosystems (Huntsville, AL, USA).

Human hsa-miR-21 (accession no.: MIMAT0000076) mimics and negative control, and antagomir and negative control were purchased from Ribobio (Guangzhou, China). The cells were seeded in 6-well plates (50–60% confluency). miR-21 mimics or negative control, and miR-21 antagomir or its negative control were transfected into MCF-7 cells, as previously described (4,31). Sismad7 or the scramble control siRNA sequence (control) was designed, produced and annealed by Ribobio, and designed to target smad7 (gene ID: 4092) was used to silence smad7 in MCF-7 cells. For smad7 rescue experiments, MCF-7 cells were transfected with 40 nM miR-21 mimics and 250 ng pCMV-SPORT6-smad7 cDNA (Invitrogen) and Lipofectamine™ 2000, which was used in all the transfection studies. After 3 days of transfection, the transfected cells were measured.

The cells in 24-well plates (60–70% confluency) were co-transfected with miR-21 mimics (40 nM) and smad7 3′-UTR reporter (200 ng). Luciferase activity was measured after 24 h, according to the manufacturer's instructions, using a Dual-Glo luciferase assay system (Promega, Madison, WI, USA).

Cells in the 6-well plates (60–70% confluency) were lysed using RIPA lysis buffer (Sijiqing, Hangzhou, China). According to our previous studies (4,31), 50 µg of each sample were separated by SDS-PAGE (10%) and transferred to PVDF membranes. The membranes were incubated with rabbit anti-human smad7 antibody (no: BA1399, 1:500) and rabbit anti-human β-actin antibody (no: BA2305, 1:1,000) (both from Boster, Wuhan, China), followed by an HRP-conjugated anti-rabbit IgG secondary antibody (no.: BA1055, 1:2,500; Boster). The samples were visualized by ECL chemiluminescence. β-actin was used as an endogenous control.

Cell invasion and migration assays were performed as previously described (31). For the invasion assay, 2.5×10⁴ cells were seeded on an 8-µm pore size Transwell filter insert (Corning Inc, Corning, NY, USA) coated with ECM (1:7.5; Sigma, St. Louis, MO, USA), while the cell migration assay was not coated with ECM. After 48 h of incubation at 37°C and 5% CO₂, cells adherent to the upper surface of the filter were removed and counted.

Paraffin-embedded human breast cancer tissues were obtained from the Department of Breast Surgery, The First Affiliated Hospital of Zhengzhou University, and 5-µm sections were stained for smad7 using the rabbit antihuman smad7 antibody (no.: BA1399, 1:100; Boster), as previously described (32).

Four-week-old BALB/c nude mice were obtained from a Shanghai Animal Laboratory (Shanghai, China). Animal welfare and experimental procedures were carried out in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the Animal Ethics Committee of The First Affiliated Hospital of Zhengzhou University (approval no.: 2013-MR-0045). Xenografted mice were developed by injecting anti-miR-21 MCF-7 cells or its control cells (the total cell number in each injection was 4×10⁶) to the left side fat-pad of each nude mouse. The tumor volume was determined weekly using digital caliper measurements and was calculated using the formula: V (volume; mm³) =1/2 × LD (longest diameter) 2 × SD (shortest diameter). After five weeks, the mice were sacrificed and the tumors were measured and analyzed.

Results are presented as mean ± SD. Statistical differences between the groups were assessed using Student's t-test, one- or two-way ANOVA, as indicated, by using statistical analysis software SPSS 17.0. P≤0.05 was considered statistically significant.

To confirm that miR-21 was overexpressed in the plasma of patients with breast cancer, we assayed the relative levels of miR-21 in the plasma of 16 normal controls, 20 patients with fibroadenoma and 20 patients with breast cancer, using RT-qPCR. The relative expression of miR-21 was 4.44153±1.46336 in breast cancer patients, significantly overexpressed as compared to 0.22116±0.05622 in the normal control group (>20-fold; P=0.0188; Fig. 1) or 0.31409±0.13237 in the fibroadenoma group (>14-fold; P=0.0192; Fig. 1).

According to web-based predictive software (http://www.targetscan.org/ and http://www.mirbase.org/) and data from other groups (33,34), smad7 is one of the downstream targets of miR-21 (Fig. 2A). To determine whether smad7 is a direct target of miR-21 during breast cancer cell invasion and migration, luciferase assays were performed using smad7 3′-UTR. The luciferase assays showed that miR-21 only reduced luciferase activity in cells containing wild-type smad7 3′-UTR, but not in cells containing mutant 3′-UTR (the mutated nucleotide is marked as red; Fig. 2B and C).

The effects of miR-21 on smad7 expression were further analyzed to determine whether miR-21 regulates smad7 during cancer cell invasion and migration. Firstly, miR-21 mimics was transfected into MCF-7 cells. The relative expression of miR-21 was 1.48398±0.31843 in MCF-7/miR-21 cells significantly upregulated as compared to 0.14668±0.03017 in the negative control group (>10-fold; P=0.0171; Fig. 2D), which suggests that miR-21 mimics may enhance the activity of miR-21 in MCF-7 cells, through various processes such as gain-of-function. The results showed that miR-21 overexpression significantly decreased the protein level of smad7 (Fig. 2E). Notably, the mRNA level of smad7 did not change in the miR-21-overexpressed cells (P=0.18697, Fig. 2F).

To determine whether miR-21 enhanced breast cancer invasion and migration, and whether it modulated the actions of EGF or TGF-β, MCF-7/miR-21 and control cells were treated with EGF (1 nM) or TGF-β (0.5 nM) and measured using Transwell migration and invasion assay. In MCF-7 cells, EGF enhanced invasion (P=0.0006, Fig. 3A) and migration (P=0.01851, Fig. 3B), TGF-β enhanced invasion (P=0.00586, Fig. 3A) and migration (P=0.0079, Fig. 3B), and their combined action was greater than that of either growth factor alone (invasion: P=0.00357, Fig. 3A; migration: P=0.0038, Fig. 3B).

Overexpression of miR-21 increased cell invasion (P=0.00126, Fig. 3A) and migration (P=0.00274, Fig. 3B) in MCF-7 cells, enhanced EGF-mediated invasion (P=0.00771, Fig. 3A) and migration (P=0.01518, Fig. 3B) and TGF-β-mediated invasion (P=0.00189, Fig. 3A) and migration (P=0.00152, Fig. 3B). The combination of the two growth factors exerted a significantly greater stimulatory effect on invasion and migration in MCF-7 cells with high miR-21 levels (invasion: P=0.00929, Fig. 3A; migration: P=0.00294, Fig. 3B).

To determine whether depletion of PTEN affected breast cancer invasion and migration, as well as the actions of EGF and (or) TGF-β, smad7 siRNA (sismad7), which targeted smad7 or the control Ssismad7 was transfected into MCF-7 cells. After 24 h, the cells were treated with EGF and (or) TGF-β, and measured uisng Transwell migration and invasion assays, as described earlier. We found that sismad7 significantly downregulated the expression of smad7 (Fig. 4A), as compared to the control groups. Notably, smad7 silencing by sismad7 enhanced breast cancer cell invasion (P=0.03102, Fig. 4B) and migration (P=0.00281, Fig. 4C) in MCF-7 cells, and increased EGF-mediated invasion (P=0.0301, Fig. 4B) and migration (P=0.01058, Fig. 4C) and TGF-β-mediated invasion (P=0.01746, Fig. 4B) and migration (P=0.00706, Fig. 4C). Moreover, the combination of the two growth factors exerted a significantly greater stimulatory effect on invasion and migration in MCF-7 cells with smad7 silencing (invasion: P=0.00577, Fig. 4B; migration: P=0.003306, Fig. 4C).

To confirm that smad7 downregulation was essential for miR-21-induced stimulation of the actions of EGF and (or) TGF-β, we used the pCMV-SPORT6-smad7 vector, which encodes a smad7 cDNA that is not regulated by miR-21 due to a mutated 3′-UTR binding site. Experiments were carried out in the presence of EGF or TGF-β, and in the absence or presence of transfected miR-21mimics using cells expressing an empty vector or the pCMV-SPORT6-smad7 vector. This experimental design allowed for the specific evaluation of the consequence of expressing a smad7 that was resistant to miR-21 upregulation. When pCMV-SPORT6-smad7 was transfected into MCF-7 cells, the stimulatory effect of miR-21 overexpression on EGF- (or TGF-β-)mediated invasion and migration was partly abrogated (Fig. 5). Thus, smad7 is a functional target of miR-21 whose upregulation promotes EGF (or TGF-β)-induced breast cancer cell invasion and migration.

To investigate the roles of EGF and TGF-β pathways in miR-21-regulating breast cancer cell invasion and migration, the MCF-7 cells were transfected with miR-21 mimics at a concentration of 40 nM for 72 h and then the cells were treated with erlotinib (the inhibitor of EGFR) at 2 mM or SB505124 (the inhibitor of TGF-β receptor) at 10 mM for 24 h. The EGFR inhibitor erlotinib (2 mM) partly blocked EGF-induced invasion and migration in the presence of miR-21 in MCF-7 cells (Fig. 6A). Similarly, the TGF-β receptor inhibitor SB505124 (2 µM) partly blocked TGF-β-induced invasion and migration in the presence of miR-21 in MCF-7 cells (Fig. 6B). In addition, the effects of miR-21 on EGF (or TGF-β) actions on invasion and migration were markedly attenuated by erlotinib and (or) SB505124 (Fig. 6C), indicating that EGF and TGF-β act through their respective receptors to cross talk and enhance breast cancer cell invasion and migration.

We determined whether smad7 was expressed in human breast cancer samples using the same highly specific antibody that was used in the immunohistochemical experiments. Smad7 was readily visible in the cytoplasm of many cancer cells but not in the adjoining stroma. A high smad7 expression was principally identified in well- to moderately differentiated breast cancer (Fig. 7), whereas a low smad7 expression was observed in poorly differentiated breast cancer (Fig. 7).

To validate the contribution of miR-21 to breast cancer biological aggressiveness, we carried out an in vivo study in an orthotopic model using nude mice (Fig. 8A). Three-fifths of the mice with miR-21 antagomir MCF-7 cells and four-fifths of the mice with MCF-7 control cells formed breast tumors and the presence of miR-21 antagomir resulted in retarded growth and smaller tumors (P=0.00033, Fig. 8A and B).