MCF-7, MDA-MB-453 and MDA-MB-231 human breast cancer cell lines were chosen because MCF-7 cell line expresses high levels of ER, PR and AR whereas MDA-MB-453 and MDA-MB-231 cell lines express high levels of AR in the absence of ER and PR (26–28). 5α-dihydrotestosterone (DHT, Sigma-Aldrich, MO, USA) was used as it is a non-aromatisable androgen and possesses the highest affinity for AR among natural androgens. All cells were obtained from the American Type Culture Collection (ATCC) and were maintained at 37°C in a humidified atmosphere containing 5% CO₂ in 75 cm² flasks containing minimal essential medium (MEM) supplemented with 10% fetal bovine serum, 100 IU/ml penicillin and 100 mg/ml streptomycin. Cells were passaged every 3–4 days when they reached 80% confluence and were harvested with 0.25% trypsin/EDTA. Before each experiment, cells were grown in phenol red-free (PRF) DMEM, containing 5% charcoal-treated fetal calf serum (PRF-CT), for 3 days and then serum starved in PRF DMEM for 24 h to synchronize the cells. All experiments were performed in 2.5% PRF-CT. Cells were treated with either DHT or vehicle alone at 10⁻⁸ M. DHT was dissolved in 100% ethanol and added to media immediately prior to use.

The MCF-7 and MDA-MB-453 cell samples were analyzed by KangChen (KangChen Bio-tech Inc.) in miRNA microarray experiments. Total RNA was extracted from MCF-7 and MDA-MB-453 cells that were treated with DHT or vehicle alone using TRIzol reagent (Invitrogen, CA, USA) according to the manufacturer’s instructions. RNA quantification and quality assurance were assessed by NanoDrop ND-1000 and RNA integrity and DNA contamination were assessed by denaturing agarose gel electrophoresis. The total RNA was labeled with Cy3 or Cy5 fluorescent dyes by miRCURY™ Array Power Labeling kit (Cat #208032-A, Exiqon) according to the manufacturer’s protocols. Each miRCURY LNA microRNA array (v.11.0, Exiqon, Vedbaek, Denmark) was hybridized with a single sample labeled with either Cy3 or Cy5. Gene Pix 4000B scanner and GenePix Pro 6.0 software (Axon Instruments, Union City, CA, USA) were used to scan images. Each group was hybridized with three miRCURY LNA Arrays in triplicate with independent samples for DHT-treated cells or vehicle-treated cells. Background subtraction and normalization were performed. We selected miRNAs whose expression intensities (ForeGround:BackGround) reached at least 1000 and expression levels differed by at least 2-fold between DHT-treated cells and vehicle-treated cells.

ER⁻, PR⁻, AR⁺ MDA-MB-453 and MDA-MB-231 cells were analyzed. To detect the mature miRNA let-7a level, a stem-loop RT-PCR assay was performed (29). Briefly, 2 μg of small RNA was reverse-transcribed to cDNA using M-MLV reverse transcriptase (Promega) with the following primers: let-7a-RT, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACAACTA-3′; and U6-RT, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACAAAATAT GGAAC-3′, which can fold to a stem-loop structure. The specific let-7a cDNA fragment was amplified along with an endogenous control U6 snRNA using the following primers: let-7a-Fwd, 5′-GCCGCTGAGGTAGTAGGTTGTA-3′; U6-Fwd, 5′-TGCGGGTGCTCGCTTCGGCAGC-3′, and a universal downstream primer reverse, 5′-CCAGTGCAGGGTCCGAGGT-3′. PCR cycles were as follows: 94°C for 4 min, followed by 40 cycles of 94°C for 30 sec, 50°C for 30 sec, 72°C for 40 sec. SYBR Premix Ex Taq™ Kit (Takara) was used following the manufacturer’s instructions, and the RT-PCR was performed and analyzed by 7300 RT-PCR system (ABI). All primers were synthesized by BGI Inc.

ChIP assays were performed using MDA-MB-453 cells to determine if activated AR directly upregulates the expression of let-7a (30). Briefly, chromatin DNA was extracted from harvested cells by sonication and precleaned by incubating with normal rabbit serum (IgG) and protein A/G-Sepharose beads. It was then immunoprecipitated with rabbit anti-AR antibody (PG-21, Upstate). The following primers were used: 5′-CAAAGTTTCTAAACGGCTTC (forward) and 5′-AGGATATTTGGTACACTCTG (reverse) for amplifying the -3922/-3707 fragment and 5′-TTTTACATTGGGCATAGCCG (forward) and 5′-TAGGCATTTGGAAGTTGGAC (reverse) for the -3036/-2786 fragment.

To construct the pcDNA3.1/pri-let-7a expression vector, we first amplified a 203-bp DNA fragment carrying pri-let-7a from genomic DNA using the following PCR primers: let-7a-sense, 5′-CGCGGATCCACTGTGGGATGAGGTAGTAGGT-3′ and let-7a-antisense, 5′-CGCGAATTCTCCAGGCCATAAACAAATGC-3′. The amplified fragment was inserted into the pcDNA3.1 (+) vector at the BamHI and EcoRI sites.

The vehicle-treated MDA-MB-453 and MDAMB-231 cells were transfected with vectors and the DHT-treated MDA-MB-453 and MDA-MB-231 cells were transfected with antisense oligonucleotide (ASO) using Lipofectamine 2000 (Invitrogen) at 24 h after plating. Transfection complexes were prepared according to the manufacturer’s protocols. The final oligonucleotide concentration was 10⁻⁹ M, and the final vector concentration was 0.5 mg/l. The transfection medium was replaced at 4 h post-transfection. The oligonucleotides complementary to let-7a were synthesized by IDT (Coralville, IA, USA) and their sequences were as follows: let-7a ASO, 5′-AACTATACAACCTACTACCTCA-3′; and control ASO, 5′-GTGGATATTGTTGCCATCA-3′.

Charcoal-stripped MDA-MB-453 and MDA-MB-231 cells were plated in replicates of 6 at a density of 7000 cells/well in 96-well microtiter plates. At 24 h after seeding, cells were treated with DHT or vehicle alone at 10⁻⁸ M and reagents were replenished every 3 days. Cell viability and proliferation were measured using the 3-(4,5 dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) colorimetric assay. At days 1–7, the cells exposed to DHT or vehicle were added to MTT. Four hours later, dimethyl sulfoxide (DMSO) was added to each well to dissolve the resulting formazan crystals. After shaking for 20 min, the absorbance at 570 nm was detected using a μQuant Universal Microplate Spectrophotometer (Bio-Tek Instruments, Winooski, VT, USA). MTT assay was also used to measure the viable proliferating cells at days 1–7 after the MDA-MB-453 and MDA-MB-231 cells were transfected.

After 48 h, the charcoal-stripped MDA-MB-453 and MDA-MB-231 cells exposed to DHT or vehicle were detached from the plates using trypsin, rinsed with PBS and fixed in 70% (v/v) ethanol. The cells were then rehydrated in PBS and incubated with RNase (100 μg/ml) and propidium iodide (60 μg/ml) (Sigma-Aldrich, MO, USA). Cells were analyzed using the FACS Calibur System (BD Biosciences, San Jose, CA, USA), and the cell cycle phase was determined by Cell Quest analysis. Proliferation index (PI) was calculated as follows: PI = (S+G2/M)/G1 (S, G2/M and G1 refer to the percentage of cells in S phase, G2/M phase and G1 phase, respectively) (31). Flow cytometry analysis was also used to measure cell cycle phase at 48 h after the MDA-MB-453 and MDA-MB-231 cells were transfected.

After 48 h the charcoal-stripped MDAMB-453 and MDA-MB-231 cells exposed to DHT or vehicle were lysed with RIPA lysis buffer and proteins were harvested. All proteins were resolved on 10% SDS denaturing polyacrylamide gel and then transferred onto a nitrocellulose membrane. Membranes were incubated with anti-KRAS, anti-CMYC or anti-GAPDH antibody (Saier Biotech, Tianjin, P.R. China) with blotto overnight at 4°C. The membranes were washed and incubated with horseradish peroxidase (HRP) conjugated secondary antibody (Saier Biotech). Protein expression was assessed by enhanced chemiluminescence and exposure to chemiluminescent film. Lab Works™ Image Acquisition and Analysis Software (UVP) were used to quantify band intensities.

Formalin-fixed and paraffin-embedded (FFPE) tissue specimens were obtained from 24 female patients (mean age: 54.5 years, range from 41 to 70 years) who underwent surgical resection for breast cancer from 2003 to 2004 at Tianjin Medical University Cancer Institute and Hospital. All cases were ER⁻ and PR⁻ invasive ductal carcinoma (IDC) and all the patients had been treated according to modern guidelines, including the use of adjuvant chemotherapy for IDC and irradiation for lymph node metastasis. The study protocol was approved by the Hospital Human Ethics Committee. Informed consent was obtained from all patients before their surgery and the examination of the specimens.

Locked nucleic acid (LNA) probes complementary to mature let-7a (5′-AACTATACAACCTACTACCTCA-3′) and scrambled negative control (5′-TTCACAATGCGTTATCGGATGT-3′) digoxigenin-labeled at the 5′-position were purchased from Exiqon. Detection of RNAs by ISH utilizing oligonucleotide probes was performed. Briefly, human tissues were deparaffinized, treated with protease (30 min in 2 mg/ml of pepsin), washed in sterile water, and then washed with 100% ethanol and air-dried. Hybridization was performed at 37°C overnight followed by a low stringency wash in 0.2X SSC and 2% bovine serum albumin at 4°C for 10 min. The probe-target complex was visualized utilizing a digoxigenin antibody conjugated to alkaline phosphatase acting on the chromogen nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate. Nuclear Fast Red served as the counterstain. The slides were scored independently by two pathologists and cases were considered positive if 10% cytoplasmic and/or membranous staining was observed.

Deparaffinization; endogenous peroxidase inactivation; antigen retrieval of FFPE clinical tissues; and immunostaining with anti-ER (SP1, 1:200 dilution; ZETA), anti-PR (SP2, 1:200 dilution; ZETA), anti-AR (AR441, 1:100 dilution; LabVision), anti-CMYC (SRP00871, 1:200 dilution; Saierbio) and anti-KRAS (SRP01436, 1:200 dilution; Saierbio) antibody were performed as described previously (13). The immunostained sections were evaluated independently by two pathologists in conjunction with the H&E-stained sections from the same lesions. For each antibody, the location of immunoreactivity, percentage of stained cells, and intensity were determined. The evaluation of expression of each protein was determined using the mean of the individual cases. AR, ER, and PR stains were assessed using Allred scores (32). CMYC and KRAS stains were considered positive if 10% of cells showed cytoplasmic and/or membranous staining.

The data are reported as the mean ± SD of the values from three independent determinations. Statistical analysis was performed using Student’s t-test in comparison with corresponding controls. Associations between ordinal variables were quantified by Spearman rank correlation with Pearson’s χ² test. Probability values of <0.05 were considered statistically significant. Analyses were run using SPSS17.0.

To identify critical miRNAs related to androgen-induced AR activating signal in breast cancer, we examined global miRNA expression in DHT-treated and vehicle-treated MCF-7 and MDA-MB-453 cells using the microRNA array (v.11.0, Exiqon) that consists of 847 capture probes for mature human miRNAs. In MDA-MB-453 cells a total number of 10 miRNAs were identified (Fig. 1A) and in MCF-7 cells none was identified using strict criteria that only miRNA undergoing alterations at least 2-fold with expression intensities of at least 1000 were considered as differentially-expressed candidates. Among these differentially-expressed miRNAs in MDA-MB-453 cells, 4 upregulated miRNAs - let-7a, b, c, d were found. Six downregulated miRNAs were identified in the DHT-treated cells when compared with the vehicle-treated cells (Table I). Among them, let-7, as a tumor suppressor miRNA, is reported to be downregulated in many types of solid tumors, including breast cancer (33–35). In our experiments, only let-7a, b, c, d were upregulated in MDA-MB-453 cells. This finding raises the possibility that these upregulated let-7a, b, c, d miRNAs might contribute to the pathogenesis of ER⁻, PR⁻, AR⁺ breast cancer. We focused on let-7a and investigated its involvement in ER⁻, PR⁻, AR⁺ breast cancer.

In ER⁻, PR⁻, AR⁺ MDA-MB-453 and MDA-MB-231 cells the validity of let-7a ectopic expression was confirmed by real-time RT-PCR, which revealed a 13-fold increase in let-7a expression in DHT-treated MDA-MB-453 and a 5-fold increase in MDA-MB-231 cells over the vehicle-treated groups (Fig. 1B). To determine if the androgen-induced AR activating signal binds to the 5′-DNA region of the let-7a locus to serve as a transcription factor, we analyzed 4.0 kb in the 5′-region of the let-7a gene and identified a typical TATA box. Using the PROMO 3.0 program 7 potential AREs were identified in the 5′-region of let-7a. To determine AR binding, ChIP was performed, and two primer pairs were used to amplify the -3922/-3707 fragment (corresponding to ARE1) and the -3036/-2786 fragment (corresponding to ARE3). Treatment of MDA-MB-453 cells with 10⁻⁸ M DHT induced a 5.2-fold increase in AR binding at ARE1 and a 3.8-fold increase at ARE3 (Fig. 1C). Taken together, the results suggest that androgen-AR signaling directly mediates regulation of let-7a. The effects of DHT on MDA-MB-453 and MDA-MB-231 cell proliferation were measured by MTT assay using cells exposed to 10⁻⁸ M DHT or vehicle for 1–7 days. DHT treatment decreased cell proliferation when compared with vehicle treatment (Fig. 1D and E). Flow cytometric analysis was used to determine the cell cycle distribution of MDA-MB-453 and MDA-MB-231 cells exposed to DHT or vehicle for 48 h. The number of cells in the G1 phase was significantly increased in both types of cells; G1-S arrest was obvious and PI was decreased in the DHT-treated group compared with the vehicle-treated group (Fig. 1F).

In order to investigate the biological significance of let-7a in MDA-MB-453 and MDA-MB-231 cells, we used a precursor expression vector (pcDNA3.1/pri-let-7a) to enhance or let-7a ASO to inhibit mature let-7a activity in vehicle-treated or DHT-treated cells. Real-time RT-PCR was used to validate the alteration in let-7a expression level. The let-7a level in the vector-treated group was significantly increased compared with the control group (Fig. 2A), whereas the let-7a level in the let-7a ASO-treated group was significantly decreased compared with the control group (Fig. 2B). These results suggest that the expression of let-7a was successfully altered as designed in our experiments. MTT assay was used to measure the viable proliferating cells at days 1–7 after the MDA-MB-453 and MDA-MB-231 cells were transfected. The MTT data suggests that when let-7a was overexpressed with vector, cell growth was inhibited, an effect similar to that observed with proliferative inhibition by DHT exposure (Fig. 2C and D). When let-7a was blocked with let-7a ASO cell growth activity was elevated, an effect directly opposite to that observed with let-7a overexpression (Fig. 2E and F). Flow cytometry analysis was used to assess the distribution of cells in the cell cycle at 48 h after the MDA-MB-453 and MDA-MB-231 cells were transfected. The data showed that when let-7a was overexpressed the number of cells in the G1 phase was significantly increased, G1-S arrest was obvious and PI was decreased (Fig. 2G), a pattern similar to that seen in cells exposed to DHT. When let-7a was blocked the number of cells in the S phase significantly increased and PI was increased, an effect directly opposite to that observed in cells with let-7a overexpressed (Fig. 2H).

Translational repression is a major mechanism of miRNA regulation of target gene expression (36). CMYC and KRAS, two oncogenes which are involved in cell proliferation and cell cycle, are known target genes of let-7a (37–40). To confirm the suppressing action of let-7a on CMYC and KRAS the correlation between miRNA and target protein expression was determined. The expression of CMYC and KRAS protein was examined by western blot analysis using MDA-MB-453 and MDA-MB-231 cells that were exposed to 10⁻⁸ M DHT or vehicle for 48 h when let-7a was upregulated. CMYC and KRAS protein were underexpressed when cells were exposed to 10⁻⁸ M DHT (Fig. 3A). Next vehicle-treated MDA-MB-453 and MDA-MB-231 cells were transfected with vector and the expression of CMYC and KRAS protein was examined by western blot analysis. CMYC and KRAS proteins were underexpressed after let-7a was upregulated by vector (Fig. 3B). DHT-treated MDA-MB-453 and MDA-MB-231 cells were then transfected with let-7a ASO and the expression of CMYC and KRAS protein was examined by western blot analysis. The amount of CMYC and KRAS protein was increased after let-7a expression was blocked (Fig. 3C). In conclusion, the above data suggest that let-7a negatively regulates endogenous CMYC and KRAS protein expression through a translational repression mechanism, similar to previous reports (37–40).

In FFPE breast cancer tissue specimens the staining outcome of ISH for let-7a and IHC staining for AR, CMYC and KRAS (Fig. 4; Table II) showed that let-7a expression was negatively correlated with CMYC and KRAS expression. CMYC expression was positively correlated with KRAS expression. No clear correlation was observed between the staining pattern of AR and let-7a, CMYC or KRAS (Table III).