Seventy-two breast cancer samples were examined and 3 adjacent normal mammary tissues were used as the control. The study was approved by the Hunan Normal University Human Ethics Committee, and informed consent was obtained from all patients. The immunohistochemical analysis was performed on polyformalin-fixed and paraffin-embedded tissues. Sections (5 μm) were deparaffinized by two 10-min washes in xylene, then rehydrated through successive graded ethanol solutions. Endogenous peroxidase was quenched with 3% H₂O₂ in methanol for 10 min and washed for 5 min in PBS. Antigen retrieval was achieved by microwaving sections in 0.01 M citrate buffer (pH 6.0) for 10 min at 800 W. The tissues were blocked in 5% bovine serum albumin (BSA) in PBS for 1 h before the addition of the mouse monoclonal anti-Eps8 antibody or mouse IgG control (diluted 1:500, BD Biosciences, CA, USA) at 4°C for overnight. The sections were incubated with HRP-conjugated goat anti-mouse secondary antibody (diluted 1:100, Sigma-Aldrich, St. Louis, MO, USA) for 45 min and then with 3,3-diaminobenzidine (DAB)/H₂O₂ for 10 min. Sections were counterstained with hematoxylin, mounted and photographed using an optical microscope (Olympus CX41, Tokyo, Japan). The percentage of tumor cells stained was scored as: 0 (no cell staining), + or 1 (≤30%), ++ or 2 (31–60%) and +++ or 3 (61–100%). The staining between two score values was given 0.5.

Human breast cancer cells MDA-MB-231, MCF7 and T47D were cultured in RPMI-1640 medium (Gibco BRL, Grand Island, NY, USA) containing 100 U/ml penicillin, 100 μg/ml streptomycin and 10% fetal bovine serum (FBS, Hyclone, Australia). MDA-MB-453, MDA-MB-468 cells and normal mammary cells HBL100 were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco) supplemented as described above. All cells were kept at 37°C in a humidified atmosphere with 5% CO₂.

MCF7 cells were transiently transfected with constructed pEGFP-C3-Eps8 or pEGFP-C3 control plasmids using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) (13). Stably transfected cells were screened by 400 μg/ml G418 (Sigma) (22,23). The transfection efficiency was detected using an invert fluorescence microscope (Olympus IX71, Tokyo, Japan).

Efficient siRNA sequence targeting Eps8 (NM_004447.5) was from the position 93–111 relative to the start codon, stem-loop DNA oligonucleotides were synthesized by Shanghai GeneChem Co. Ltd., China and inserted into the lentivirus-based RNAi vector pGCSIL-GFP (GeneChem) (13). A non-targeting stem-loop DNA was also cloned into pGCSIL-GFP vector as a negative control (NC). Lentiviral particles were prepared as described previously (24). Briefly, the lentivirus expression plasmid and packaging plasmids (pHelper 1.0 and pHelper 2.0) were cotransfected into 293T cells, supernatants were harvested 48 h after transfection and filtered through a 0.45-μm pore size filter (Millipore, Billerica, MA, USA) and concentrated by ultracentrifugation. The infectious titer was determined using hole-by-dilution titer assay. MDA-MB-231 cells at a density of 100,000 cells per well in 6-well plates were infected with Eps8-RNAi-lentivirus or NC-RNAi-lentivirus and 5 μg/ml polybrene (Sigma) at the multiplicity of infection (MOI) 10 and detected on the 4th day by the invert fluorescence microscope.

For cell viability assay, 3,000 cells were seeded in octuplicate in 96-well plates untreated or treated with 30 μM of cisplatin (DDP, Sigma) for 24 h. On days 1, 3 and 4 or days 1, 3 and 6, cells were analyzed with 1 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenytetra zolium bromide (MTT, Sigma) at 37°C for 4°h. Then 100 μl dimethylsulfoxide (DMSO)/well was added to dissolve the formazan crystals. The absorbance at 490 nm was obtained using a spectrophotometer (UV-2102C, Changsha, China).

For cell survival assay, 100,000 MCF7 cells stably expressing GFP-Eps8 or GFP and parental cells were plated in triplicate in 6-well plates in complete medium containing 400 μg/ml G418. MDA-MB-231 cells (50,000) infecting Eps8-RNAi-LV or NC-RNAi-LV were plated in triplicate in 6-well plates. After 3–6 days, viable cell numbers were counted with a hemocytometer after trypan blue staining of dead cells.

Liquid colony formation was performed, 2,000 cells were seeded in triplicate in 6-well plates and grown in complete culture medium for >10 days. Colonies were fixed with methanol, stained with Giemsa (BBI International, Cardiff, UK) and photographed with a digital camera (Canon IXUS 125 HS, Tokyo, Japan). Only colonies containing >30 cells were counted under an inverted microscope (Zeiss Axiovert 25, LLC, USA). All experiments were carried out at least three times.

Lentivirus-infected and parental MDA-MB-231 cells were plated onto 6-well plates for 24 h. Serum was withdrawn when cells were 70% confluent. After 36 h, 10% FBS was added in the medium for an additional 18 h. Cells were collected gently, fixed in 70% ethanol and stained with propidium iodide (PI, BD Pharmingen, San Diego, CA, USA), the DNA content was analyzed on a BD FACSCalibur cytometer (Becton-Dickinson, San Jose, CA, USA) using CellQuest Pro and ModFit software (BD).

Lentivirus-infected and parental MDA-MB-231 cells were harvested and washed with cold PBS. The pellet was resuspended in 1X binding buffer and the sample solution was incubated with PI and FITC-conjugated Annexin V (BD Pharmingen) for 15 min at room temperature in the dark. The samples were analyzed by the BD FACSCalibur cytometer using the CellQuest software.

The mouse experiments were carried out according to the ethical guidelines for laboratory animal use and approved by the Ethics Committee of Hunan Normal University. Approximately 10⁷ of lentivirus-infected MDA-MB-231 cells in 0.2 ml of sterile PBS were injected subcutaneously into the left and right dorsal regions of 4- to 5-week-old female nude mice (BALb/c). Mice were checked every 2 days and the formed tumors were measured with a micrometer (25). After 26 days, mice were sacrificed, and tumors were excised, weighed and photographed.

MDA-MB-231 cells were cultured in 24-well plates until >90% confluence. A 100-μl pipette tip was used to generate wounds. After wound creation, the medium was changed to remove cellular debris. Three wounded areas in each well were marked on the bottom of plates and photographed at 1 and 4 days with an invert microscope.

Stimulation of membrane ruffling with 60ng/ml of human EGF (Sigma, Deisenhofen, Germany) in RPMI-1640 for 30 min was performed after serum deprivation for 48 h. MDA-MB-231 cells were fixed with 3.7% paraformaldehyde for 10 min, permeabilized with 0.1% Triton X-100 for 10 min, and blocked with 5% BSA in PBS. Alexa fluor 595 conjugated to phalloidin (diluted 1:500, Molecular Probes, Eugene, OR, USA) and 1 μg/ml Hoechst 33258 (Sigma) were incubated for 20 min to stain F-actin and the nucleus in the dark. All specimens were viewed with an upright fluorescence microscope (Zeiss Axioskop 2, LLC, USA). At least 20 cells were measured in three independent experiments.

Cells were lysed in RIPA buffer [50 mM Tris-HCl (pH 7.2), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS and cocktail protease inhibitors]. The lysates (50 μg) were denatured in sample buffer and heated to 105°C for 5 min. Samples were then separated on 10–15% SDS-PAGE gels and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore). The membranes were blocked with 1% BSA in TBS-T [10 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Tween-20] and incubated overnight with rabbit polyclonal antibodies against GFP and MMP9, mouse monoclonal antibodies against ERK and phosphorylated ERK, cyclin D1 (CCND1), c-Myc, p53, N-cadherin, E-cadherin, vimentin, β-actin and GAPDH (diluted 1:1,000, Santa Cruz Biotechnology, Santa Cruz, CA, USA) in TBS-T containing 1% BSA with gentle shaking at 4°C. HRP-conjugated goat anti-rabbit and goat anti-mouse secondary antibodies (diluted 1:5,000, Sigma) were used. The signals were detected with SuperSignal West Pico chemiluminescent Substrate (Thermo Scientific Pierce, Rockford, IL, USA) and visualized with tanon-5200 system (Bio-tanon, Shanghai, China).

All statistical analyses were performed using the SPSS 11.0 software (SPSS Inc., Chicago, IL, USA). Data are shown as mean ± SD from at least 3 independent experiments. Statistical significance was determined using Student’s t-test at P-values <0.05.

The expression level of Eps8 was examined in 4 stage IV, 22 stage III, 39 stage II, 7 stage I human breast cancers and 3 adjacent normal mammary tissues by the immunohistochemistry analysis using mouse monoclonal anti-Eps8 antibody. We found that Eps8 was completely localized in the cytoplasm (Fig. 1A). Eps8 expression was detected in 46 (63.9%) of the 72 breast cancers with strong staining (3+), 21 (29%) of the 72 breast cancers were moderately stained (2+), and 5 (6.9%) were weakly stained or negative for Eps8 expression (+/0), which indicated Eps8 was highly expressed in breast cancers (P<0.001) according to Student’s t-test. A complete loss of Eps8 expression was observed in normal mammary tissues (Fig. 1B). Therefore, Eps8 expression was significantly increased in human breast cancer samples. Patient characteristics are summarized in Table I.

We next analyzed the expression of Eps8 proteins in five human breast cancer cell lines. Higher expression of Eps8 proteins was evident in MDA-MB-231, MDA-MB-453 and T47D cells than in MDA-MB-468, MCF7 cells and normal breast epithelial cells HBL100 (Fig. 1C). Thus, the level of Eps8 expression was high in the highly invasive breast cancer cell line MDA-MB-231 and low in the weakly invasive breast cancer cell line MCF7. Both cell lines were further selected to knockdown and overexpress Eps8 proteins, respectively.

To further investigate the role of Eps8 in human breast cancer cells, we overexpressed Eps8 and asked whether Eps8 upregulation enhances the proliferation of breast cancer cells. The pEGFP-C3/Eps8 and pEGFP-C3 plasmids were transfected to established MCF7 cell lines stably overexpressing GFP/Eps8 and GFP by G418 screening (Fig. 2A). As shown in Fig. 2B, GFP and GFP/Eps8 target proteins were expressed at high levels in MCF7 cells.

We then investigated whether Eps8 overexpression in the breast cancer MCF7 cells leads to cell proliferation. The same amount of cells was plated in 96-well plates, and we examined cell viability using MTT assays, Eps8 overexpression resulted in a remarkable increase in viable cells (Fig. 2C). The same amount of cells was plated in triplicate in 6-well plates and cell number was counted on days 1, 3 and 4. We found that the overexpression of Eps8 in MCF7 cells shows an increased cellular growth compared with controls (Fig. 2D). Further, the liquid colony formation assays indicated a great increase in colony number and size (Fig. 2E). Therefore, these results suggested that Eps8 could contribute to breast cancer cell survival and proliferation in vitro.

The above results indicated that Eps8 overexpression significantly promotes MDA-MB-231 cellular proliferation. To gain further supporting data, we attempted to knock down Eps8 expression and asked whether shRNA-mediated Eps8 knockdown inhibits the proliferation of breast cancer cells. Then, lentivirus-based RNAi vector pGCSIL-Eps8 containing the efficient Eps8 shRNA1 and packaging plasmids were cotransfected to 293T cells (13). Lentiviral particles were prepared to infect MDA-MB-231 cells. The fluorescence intensity was markedly increased 4 days after infection and the infection efficiency was close to 90% in MDA-MB-231 cells (Fig. 3A). Additionally, western blot analysis showed that Eps8 shRNA1 could significantly suppress the expression of endogenous Eps8 proteins when compared with negative control shRNA (Fig. 6A).

Next, we examined whether Eps8 is a critical regulator of breast cancer cell proliferation and investigated the effect of Eps8 knockdown on MDA-MB-231 breast cancer cell growth. From Fig. 3B and C, we observed that Eps8 deletion significantly inhibits the proliferation of MDA-MB-231 cells, whereas control shRNA has no effect. The liquid colony formation assays showed that Eps8 knockdown displays much fewer and smaller colonies in MDA-MB-231 cells, while control shRNA has no effect compared with uninfected parental cells (Fig. 3D). We further explored the molecular mechanism involved in Eps8-mediated cancer cell survival, the effects of Eps8 knockdown on cell cycle and apoptosis were investigated. The flow cytometry analysis showed that Eps8 knockdown results in a significant decrease from mean 91.9 to 74.66% in living MDA-MB-231 cells and exhibits an increase in early and late apoptotic cells compared to control shRNA-infected and parental cells (Fig. 3E). Moreover, we compared the DNA content between Eps8 repressed and control cells and found that the proportion of S-phase cells was significantly decreased in Eps8 knockdown cells (31.33%) compared with MDA-MB-231 cells (39.05%), and the proportion of S-phase cells were even lower in Eps8 repressed cells compared with NC control cells (44.83%) (Fig. 3F). These observations suggested that Eps8 knockdown arrested cells at the G1/S checkpoint. Therefore, the above data showed that Eps8 knockdown suppresses proliferation, and cell cycle progression enhancing cell apoptosis of MDA-MB-231 cells.

The inhibitory effect of Eps8 knockdown on breast cancer cell proliferation in vitro suggested that Eps8 deletion might suppress tumor growth in vivo. Tumorigenicity assays were performed by subcutaneous injection of Eps8 shRNA cells into nude mice, and NC shRNA cells were used as controls. Within 4 weeks, solid tumors were readily visible in left and right dorsal regions of all mice (Fig. 4A), but the average tumor volume and weight of the Eps8 shRNA group were markedly reduced by >60% compared with negative controls (Fig. 4B and C). Thus, the above data showed Eps8 knockdown could decrease breast cancer cell proliferation in vivo.

We next assessed the effect of Eps8 on breast cancer cell migration by wound-healing assays. As shown in Fig. 5A, Eps8 depletion produced 62% inhibition of cell migration in MDA-MB-231 cells. In contrast, control groups dramatically promoted cell migration. Thus, Eps8 knockdown was able to significantly inhibit cell motility, which confirmed the role of Eps8 in breast cancer cell migration.

As Eps8 was previously described to be essential for actin dynamics and cytoskeletal organization in pancreatic cancer (12), we examined the effects of Eps8 knockdown on filopodia in MDA-MB-231 cells. Cells were cultivated for 48 h in serum-free medium and then EGF was added for 30 min. Eps8-deletion cells exhibited reduction of filopodial density and length and inhibited the filopodial growth (Fig. 5B), indicating that Eps8 is involved in a rearrangement of F-actin in breast cancer cells.

We further investigated the implication of Eps8 in chemosensitivity of breast cancer cells. MDA-MB-231 cells were treated with cisplatin for 24 h. Sixty-two percent reduction of MTT absorbance was detected in Eps8 shRNA cells compared with 48% inhibition in their control cells (Fig. 5C). Thus, Eps8 knockdown was able to sensitize MDA-MB-231 cells to the cytotoxicity of cisplatin as reported in cervical cancer cells (8).

Because Eps8 influences many genes involved in the development and progression of human cancers, we decided to examine whether Eps8 regulates these target genes in breast cancer cells. As shown in Fig. 6A, Eps8 knockdown had no effect on ERK protein, but it decreased the phosphorylated level of ERK and downregulated the expression of c-Myc and cyclin D1 (CCND1) as downstream mitogenic targets of ERK signaling (26). Moreover, Eps8 knockdown exhibited inhibited MMP9 and enhanced the tumor suppressor p53. The level of EMT epithelial marker E-cadherin was concomitantly increased in Eps8-depleted cells. In contrast, Eps8 knockdown decreased the levels of mesenchymal markers N-cadherin and vimentin. Taken together, these data showed that Eps8 regulates breast cancer cell proliferation and migration, at least in part, by affecting ERK signaling, MMP9, p53 and EMT markers.