Ellagic acid (EA) was purchased from Sigma Chemical Co. (St. Louis, MO, USA). Stock solution of EA (2 mg/ml) was prepared in dimethyl sulfoxide (DMSO), and filter sterilized before use. The human breast cancer cell line MCF-7 was purchased from the Cell Bank of Shanghai Institute of Biological Sciences, Chinese Academy of Sciences (Shanghai, China). MCF-7 cells were cultured in RPMI-1640 supplemented with 10% fetal bovine serum (FBS) and 1% penicillin streptomycin solution in an atmosphere of 95% air and 5% CO₂ in a 37°C humidified incubator.

Cells were seeded in 96-well plates at a density designed to reach 70–80% confluency. Cells were allowed to adhere and 24 h later were treated with EA at 0, 10, 20, 30 and 40 μg/ml. After 24, 48 or 72 h of treatment, 200 μl of MTT was added to each well, and the cells were incubated for 4 h at 37°C. The medium was discarded, and the dark blue formazan crystals were adequately dissolved with DMSO for 10 min on a rocker platform. The absorbance was measured at 570 nm using an ELISA reader (Tecan Sunrise, Männedorf, Switzerland). The cell viability was calculated according to: OD sample/OD control × 100%. The assay was performed in triplicate.

MCF-7 cells (5×10⁵) were seeded in T25 culture flasks and grew for 6 h to reach 50–60% confluency. Cells were starved in serum-free medium for 24 h to achieve synchronization. After returning to regular growth medium for 6 h, cells were treated with EA at 0, 10, 20 and 30 μg/ml. DMSO (final concentration <0.1%) was used as a negative control. After treated for 24 h at 37°C, floating and adherent cells were collected, washed with ice-cold PBS and fixed with 70% ethanol for at least 12 h at 4°C. The cells were then treated with 80 mg/ml RNase A and 50 μg/ml PI at a density of 1×10⁶ cells/ml for 30 min, and the stained cells were analyzed using a FACScan cytometer (Becton-Dickinson, Franklin Lakes, NJ, USA).

The apoptotic rate was measured by FCM according to the instructions provided by the Annexin V-FITC kit (Becton-Dickinson). Briefly, following treatment with 0, 10, 20 and 30 μg/ml of EA for 24 h, cells were harvested after digestion with 0.25% trypsin. The collected cells were washed three times with ice-cold PBS containing calcium and resuspended in binding buffer (500 μl) at a density of 1×10⁶ cells/ml, in which 500 μl of cell suspension was added to a 5 ml FCM tube and Annexin V-FITC (50 μg/ml, 5 μl) and PI (50 μg/ml, 5 μl) were added. Then the cells were incubated for 30 min at room temperature in the dark. The apoptotic percentage of 10,000 cells was determined using FCM.

Morphology and ultra structure of apoptotic cells was observed by transmission electron microscopy (TEM). MCF-7 cells were seeded at a density of 1×10⁶ cells/flask and treated with 15 μg/ml EA-supplemented medium for 24 h. DMSO was used as a control. After incubation, cells were fixed with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer for 24 h at 4°C. Then, the cells were washed in the same buffer three times, fixed with 1% osmium tetroxide and dehydrated in graded ethanol. The 100% ethanol solution was then replaced by propylene oxide and embedded in epoxy resin, which was polymerized at 70°C for 8 h. Sections were stained with uranyl acetate and lead citrate, and then observed with a Hitachi H-7650 TEM (Hitachi High Technologies, Tokyo, Japan).

MCF-7 cells were plated at a density of 4×10³ cells/cm² in T75 culture flasks. Synchronization was achieved as previously described. The cells were harvested at 80–90% confluency. After exposure to EA (15 μg/ml) for 6, 12 and 24 h, total RNAs of the cells were harvested using TRIzol reagent (Life Technologies, Carlsbad, CA, USA) and the RNeasy kit (Qiagen, Dusseldorf, Germany) according to manufacturer’s instructions. For each time-point, two preparations of RNA samples were independently subjected to array hybridization. After having passed RNA measurement using the NanoDrop ND-1000 (NanoDrop Technologies, Wilmington, DE, USA) and denaturing gel electrophoresis, the samples were amplified and labeled using the Agilent Quick Amp labeling kit and hybridized with Agilent whole genome oligo-microarray in Agilent’s SureHyb hybridization chambers. After hybridization and washing, the processed slides were scanned with the Agilent DNA microarray scanner using settings recommended by Agilent Technologies (Palo Alto, CA, USA).

The resulting text files extracted from Agilent Feature extraction software (version 10.5.1.1) were imported into the Agilent GeneSpring GX software (version 11.0) for further analysis. The microarray data sets were normalized in Agilent Feature extraction software, and then the genes recorded present in all the samples were chosen for further analysis. Differentially expressed genes were identified through fold-change screening.

The gene expression profiling experiment was successfully completed on the samples. The profiling identified a subset of the total number of probes analyzed by Agilent Whole Genome Oligo microarray that are differentially expressed. GO Analysis and Pathway Analysis were performed on this subset of genes. More detailed information was found in Gene Ontology (GO) report and Pathway Analysis report.

Total RNA from MCF-7 cells exposed to 15 μg/ml EA for 6, 12 or 24 h were used for transcriptomics analysis by real-time reverse transcription-PCR (real-time RT-PCR) with selected target genes. Each 2 μg of RNA was reversely transcribed to cDNA using oligo(dT) primers and SuperScript II reverse transcriptase kit (Life Technologies). Primers were designed using Primer 5 software and synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). The primers are shown in Table I. Real-time RT-PCR reactions were then performed in a total of 25 μl of reaction mixture using the ABI Prism 7900HT sequence detection system (Applied Biosystems, Foster City, CA, USA). Data were analyzed using the comparative Ct method and was normalized by GAPDH expression in each sample.

Western blot analysis was performed using the lysis buffer isolated protein. Briefly, 20 μg of protein was resolved in 10–15% SDS/PAGE and transferred to polyvinylidene fluoride membrane. The blots were blocked in blocking buffer overnight at 4°C, and incubated with the primary antibody at 37°C for 1 h. The antibody-antigen complexes were then detected with alkaline phosphatase-conjugated anti-rabbit or anti-mouse IgG secondary antibodies using a BCIP/NBT Alkaline Phosphatase Color Development kit. The band was recorded by a digital camera and analyzed by ImageJ software (NIH, Bethesda, MD, USA). The results were normalized with β-actin. The following monoclonal antibodies were used in the present study (source): anti-β-actin (Sigma Chemical), anti-Smad3, anti-p21^Cip1 (Boshide Co., Wuhan, China), anti-phospho-Smad3 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Rb, anti-phospho-Rb, anti-TGFβ1, anti-TβR I and anti-TβR II (Cell Signaling Technology, Beverly, MA, USA).

The Student’s t-test was used to determine the significance between treatments and untreated controls, and P<0.05 was considered significant.

The inhibitory effect of EA on cell proliferation in MCF-7 cells was assessed by MTT assay. As shown in Fig. 1, EA suppressed MCF-7 cell growth in a time- and dose-dependent manner. The inhibitory response was apparent as early as 24 h with a concentration range of EA from 10 to 40 μg/ml EA was found to reduce the cell number to 13.5 and 80% of the untreated control at concentrations of 10 and 40 μg/ml, respectively.

To determine whether decreased cell number accumulation was related to cell cycle arrest treated by EA, we assessed the effect of EA on cell cycle perturbation by flow cytometry. The data in Fig. 2 clearly showed a significant block in G0–G1 phase of the cell cycle. The increase in the G0–G1 population was accompanied by a delay of passage of cells to S phase. Synchronized control cells that were not treated by EA moved into S phase.

In order to determine whether EA could induce cell apoptosis, FCM and TEM were conducted in EA-treated MCF-7 cells and the control. FCM showed that the cell apoptotic rates were 4.36±0.35 and 10.24±1.23 after treatment with 15 or 20 μg/ml of EA for 24 h, respectively. MCF-7 cells treated with increasing concentrations of EA exhibited a significant increase in the apoptotic cell fraction, indicating apoptosis induction (Fig. 3).

Morphological and ultra-structural characteristics of apoptosis in MCF-7 cells exposed to EA for 24 h were examined by transmission electron microscope (TEM) (Fig. 4).

We found a total of 4,738 genes that showed a >2.0-fold change after 24 h of EA treatment. Among these genes, 2,547 genes were downregulated and 2,191 genes were upregulated in EA-treated MCF-7 cells. The altered expressions of many genes did not occurr after 6 h of EA treatment and changes were evident with 24 h of treatment (Table II).

After clustering based analysis and pathway analysis according to their biological functions, several genes, which are related to cell cycle, apoptosis and DNA replication were found increased or decreased (Fig. 5). Especially, TGF-β/Smads pathway was found as the potential pathway, by which EA induced cell cycle arrest in G0/G1 phase and educed the inhibitory effect on MCF-7 cells (Fig. 6).

The next step was to confirm the changes of a subset of 16 genes that showed pronounced regulation in TGF-β/Smads pathway by real-time RT-PCR and/or western blot analysis. We examined the genes (TGF-β1, TβR-I, TβR-II, Smad3, p15^Ink4b, p19^Ink4d, p21C^ip1, p57^Kip2, cyclin D1, cyclin E, cyclin A2, E2F1, E2F2, Rb1, p107 and p130) by real-time RT-PCR and some proteins (TGF-β1, TβR-I, TβR-II, Smad3, p-Smad3, Rb1, p-Rb1 and p21^Cip1) by western blots. We were able to verify eight proteins by western blot analysis and ten RNAs by real-time RT-PCR. This represents a success rate of 75% (12 out of 16). The results are shown in Fig. 7, respectively. The lack of complete concordance could be attributable to either false positive signals of the array data or the discrepancy between transcript and protein expression. There was a noteworthy observation. A decreased level of cyclin D1 was found at 12 h, but an upregulation of cyclin D1 expression was detected by cDNA microarray, real-time RT-PCR and western blot analyses at 24 h. Cells lacking cyclin D1 are known to stay in cell G0–G1 arrest (12,13). It is possible that at the early time-point, the repression of cyclin D1 was necessary for initiating EA-mediated cell cycle arrest, whereas at the later time-point when G0–G1 arrest was firmly established, cells produced more cyclin D1 for apoptosis to advance (14).