The human breast cancer cell lines, MCF-7 (ATCC^® HTB-22™), Hs578T (ATCC^® HTB-126™) and MDA-MB-231 (ATCC^® HTB-26™), were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). The MCF-7, Hs578T and MDA-MB-231 cells were respectively cultured in Eagle's minimum Essential medium (EMEM), Dulbecco's modified Eagle's medium (DMEM) and L-15 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Thermo Fisher Scientific, Inc., Waltham, MA, USA). The MCF-7 and Hs578T cells were cultured at 37°C with 5% CO₂ and the MDA-MB-231 cells were cultured at 37°C in a CO₂-free incubator.

Total RNA was isolated from the MCF-7, MDA-MB-231 and Hs578T cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The PrimeScript RT reagent kit (Clontech Laboratories, Inc., Kusatsu, Japan) was used for the reverse transcription of 500 ng of total RNA to complementary DNA. The relative expression of S100B was determined using a Real-Time PCR system (StepOnePlus Real-Time PCT system; Applied Biosystems, Foster City, CA, USA) using Fast SYBR-Green Master Mix (Applied Biosystems) with the specific primer: S100B forward, 5′-AGGGAGACAAGCACAAGCTG-3′ and reverse, 5′-CGTGGCAGGCAGTAGTAACC-3′. The expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was detected using the primers: forward, 5′-GAGTCAACGGATTTGGTCGT-3′ and reverse, 5′-TTGATTTTGGAGGGATCTCG-3′. Relative mRNA expression was normalized to the expression level of GAPDH and calculated using the 2^−ΔΔCt method (19).

The migration ability was analyzed by wound healing assay and Transwell migration assay. For wound healing assay, 2.4×10⁵ MDA-MB-231 cells were seeded into 24-well plates. At 24 h after seeding, a scratch was made using a 200-ml pipette tip. After scratching, cell debris was removed by washing twice with phosphate-buffered saline (PBS) and the cells were then incubated in serum-free L-15 medium containing 0, 0.1 and 1 nM recombinant human S100B protein (R&D Systems, Minneapolis, MN, USA). Images were acquired at 0 and 12 h following treatment with S100B. Quantification was performed using AxioVs40 V4.8.2.0 microimaging software (MicroImaging GmbH, Gottingen, Germany). QCM™ 24-well Cell Migration assay and Invasion System uncoated 8-µm pore size polycarbonate membranes (Millipore, Billerica, MA, USA) were used for Transwell migration assay according to the manufacturer's instructions. A total of 4×10⁴ MDA-MB-231 or Hs578T cells were seeded in 300 ml of serum-free medium in 24-well upper inserts and 500 ml medium with 10% FBS was placed in the lower chamber. After 12 h, the bottom surface of the membrane was fixed in 5% formaldehyde solution, followed by 0.4 g/l crystal violet (Sigma-Aldrich, St. Louis, MO, USA) staining for 2 h. The cells on the upper surface were removed using a cotton swab after washing the crystal violet-stained membrane. The bottom of the membrane was then visualized using an inverted light microscope (Carl Zeiss Primo Vert Microscope; Carl Zeiss, Munich, Germany) at x100 magnification. In total, 4 random fields of view were counted and the relative fold of migration in each group was compared to the 0 nM S100B treatment group.

The cells were lysed on ice for 30 min in radioimmunoprecipitation lysis buffer (RIPA) buffer (Millipore) which was supplemented with protease inhibitor cocktail (Millipore) at a 100:1 ratio. The total cell lysate was collected following centrifugation at 4°C, 12,000 × g for 15 min. The protein concentration was determined using a bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, Rockford, IL, USA). Equal amounts of protein were loaded and separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were then transferred onto polyvinylidene difluoride membranes (Millipore). After 1 h of blocking [5% skim milk in Tris-buffered saline with Tween-20 (TBST) buffer], the membranes were incubated with the following primary antibodies: N-cadherin (1:1,000, cat. no. 2167184; Millipore), E-cadherin (1:1,000, cat. no. 610182; BD Biosciences, San Jose, CA, USA), vimentin (1:1,000, cat. no. MAB3400; Millipore) and GAPDH (1:5,000, cat. no. MAB374; Millipore) at 4°C overnight. After washing with TBST, the membranes were incubated with secondary antibodies, including peroxidase conjugated goat anti-rabbit IgG (1:5,000; cat. no. AP132P; Millipore) and peroxidase conjugated goat anti-mouse IgG (1:5,000; cat. no. AP124P; Millipore) for 1 h at room temperature. The results were acquired on Alpha Innotech FluorChem FC2 Imaging system (ProteinSimple; Bio-Techne, Minneapolis, MN, USA).

The mRNA expression of S100B in breast cancer samples was obtained from the Oncomine Research Edition (Thermo Fisher Scientific; http://www.oncomine.org, v4.5) which includes the TCGA breast dataset of Ma et al (20). The clinical data of TP53 mutation and the expression levels of S100B in the breast cancer samples were downloaded as z-scores from the cBioPortal (http://www.cbioportal.org, Breast cancer, TCGA, Cell 2015) (21). The expression of S100B in different subtypes (PAM50) of breast cancer was evaluated using the GOBO database (http://co.bmc.lu.se/gobo) (22).

The distant metastasis-free survival (DMFS) analysis of the patients with breast cancer with different expression levels of S100B was performed using Kaplan-Meier plotter (KM Plotter) database (http://kmplot.com/analysis/) (23). Briefly, the prognostic value of each gene was analyzed by splitting the patient samples into 2 groups according to the median value. To investigate the association between S100B expression and DMFS in different subtypes of breast cancer patients, the DMFS was determined in the subtype of estrogen receptor (ER)-positive (n=664) patients, ER-negative (n=218) patients, or patients subjected to endocrine therapy (n=645). The hazard ratio with 95% confidence intervals and log-rank P-value are shown according to the KM plotter database. To further examine whether the expression of S100B is associated with the TP53 status, the subtype of breast cancer was restricted to TP53 status 'mutated' (n=188) or 'wild-type' (n=273) and the DMFS of the patients was analyzed. The overall survival was accessed in the dataset 'Breast Invasive Carcinoma TCGA' from the SurvExpress database (http://bioinformatica.mty.itesm.mx:8080/Biomatec/Survivax.jsp, Interface v2.0, Database Update: November 5, 2017) (24,25). The censored was set to 'Survival_months', groups were divided by 'maximize risk groups' and stratification was set to 'class: ER'. To determine the distant recurrence rate, the dataset 'Vincent Darbon Breast' (Accession no. GSE9893 in the Gene Expression Omnibus database) in the SurvExpress database. The censored was set 'Distant recurrence months', and the raw data of the distant recurrence-free survival curve and mRNA expression levels were sequentially obtained. The low- and high-risk groups were divided according to the median risk index on the SurvExpress website. The survival curve and mRNA expression were re-drawn using GraphPad Prism 7 software (GraphPad Software, Inc., San Diego, CA, USA).

All graphs were generated using GraphPad Prism 7 software (GraphPad Software). The Student's t-test was used for the analysis of differences between 2 groups and one-way analysis of variance (ANOVA) with a Tukey's post hoc test was used for the analysis of differences among 3 or more groups. A P-value <0.05 was considered to indicate a statistically significant difference.

To investigate the expression levels of S100B in the breast cancer samples, the Oncomine database (http://www.oncomine.org) was used. Compared to the normal breast tissue, the expression of S100B in various types of breast cancer tissue was significantly low in 2 independent data-sets (Fig. 1A and B). To further examine S100B expression in different subtypes of breast cancer according to PAM50 subtypes, S100B expression was evaluated using the GOBO database (http://co.bmc.lu.se/gobo). As shown in Fig. 1C, a relatively high S100B expression level was observed in the basal-like type when compared to the luminal A (LumA), LumB and HER2-enriched groups. Moreover, S100B expression in the ER-negative group was higher than that in the ER-positive group (Fig. 1D). Based on these results, we hypothesized that S100B may directly regulate the behaviors of the breast cancer cells, particularly in basal-like or ER-negative breast cancer. mRNA expression levels in 3 cancer cell lines, including MCF-7 (luminal A, ER-positive), MDA-MB-231 (basal-like, ER-negative) and Hs578T (basal like, ER-negative) were then determined. As shown in Fig. 1E, the mean S100B expression in the MDA-MB-231 and Hs578T cells was higher than that in the MCF-7 cells; however, there was no statistical difference among these 3 cell lines.

The effects of extracellular S100B were further examined in the MCF-7, MDA-MB-231 and Hs578T cells. The cells were treated with various concentrations of human recombinant S100B protein. In the Transwell migration assay, S100B treatment significantly inhibited the migration of the two basal-like cell lines (Fig. 2A and B); however, it did not significantly affect the migration of the MCF-7 cells (Fig. 2C). In addition, the inhibitory effects of S100B treatment were observed in the MDA-MB-231 cells (Fig. 2D), but not in the MCF-7 (Fig. 2E) cells, in wound healing assay. The process of epithelial-mesenchymal transition (EMT) in cancer cells is associated with metastatic potential in breast cancer and EMT is related to the basal-like type of breast cancer (26). The MDA-MB-231 and Hs578T cells are classified as the basal B type (27). In order to determine whether S100B affects the mesenchymal or epithelial phenotype transition, the markers of the mesenchymal phenotype, N-cadherin and vimentin, and the marker of the epithelial phenotype, E-cadherin (28), were determined in the MDA-MB-231 cells. The results of western blot analysis revealed that S100B treatment induced the expression of E-cadherin. Thus, S100B treatment led to the acquirement of the epithelial phenotype (Fig. 3).

As S100B treatment inhibited cell metastasis and the acquisition of the epithelial phenotype, we further investigated whether S100B expression could be a prognostic marker for metastasis in patients with breast cancer, particularly in patients with metastatic breast cancer via the online databases, SurvExpress (http://bioinformatica.mty.itesm.mx:8080/Biomatec/Survivax.jsp) and KM plotter (http://kmplot.com/analysis/). In a dataset from the SurvExpress database, patients with a low S100B expression (high-risk group) had a good survival rate in a distant recurrence-free survival analysis (Fig. 4A and B). Furthermore, we examined whether the ER status was an important factor in the S100B-induced inhibitory effect on cell migration. In breast invasive carcinoma, the TCGA dataset from the SurvExpress database, a high S100B expression was associated with a good overall survival rate in only ER-negative breast cancer patients, but not in ER-positive breast cancer patients and in breast cancer patients as a whole (Fig. 4C–E). As S100B treatment resulted in the inhibition of the migration of ER-negative breast cancer cell lines, we further investigated the DMFS using the KM plotter database. S100B expression was not associated with DMFS in all breast cancer patients as a whole (Fig. 4F). By contrast, S100B expression was associated with a good DMFS in ER-positive (P=0.018) and ER-negative (P=0.049) breast cancer patients (Fig. 4G and H). Of note, S100B expression was a marker of prediction of DMFS in ER-positive breast cancer patients who received endocrine therapy (Fig. 4I).

In melanoma, S100B decreases p53 activity and expression (14,15). In this study, the association between S100B and p53 expression was examined using the TCGA breast dataset from cBioPortal (http://www.cbioportal.org). Compared to the p53 wild-type group, a significantly higher p53 expression was observed in the p53 mutant group (Fig. 5A). Both the MDA-MB-231 and Hs578T cells exhibited p53 mutation (27). However, in this study, S100B treatment resulted in the inhibition of cell migration. Therefore, the association between the p53 status, S100B expression and the prognosis of breast cancer patients was evaluated. The DMFS in breast cancer patients with mutant p53 and wild-type p53 was respectively accessed in the KM plotter database. The results revealed that a high S100B expression was associated with a good outcome (Fig. 5B and C). These findings suggest that a low S100B expression is associated with a high metastatic rate irrespective of the p53 status in breast cancer.