Human mammary carcinoma cell line MDA-MB-231 was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) and the cells cultured in the Leibovitz L-15 medium supplemented with 10% fetal bovine serum (FBS), streptomycin (100 mg/ml) and penicillin (100 U/ml). The same culture medium, adding 200 μg/ml G418 (Life Technologies, Carlsbad, CA, USA) was also applied to parental MDA-MB-231 (231-P) cells and the metastatic subline of MDA-MB-231 (231-B) cells derived from the bones of our animal models, both of which had stable luciferase expression. All these cells were maintained in the incubator without gas exchange with outside air.

Five-week-old female Balb/c nude mice were purchased from Vital River Laboratories (Beijing, China) and bred in the SPF Animal Center of Cancer Institute and Hospital, Chinese Academy of Medical Sciences. All the animal experiments were approved by the Institutional Review Board of the Chinese Academy of Medical Sciences Cancer Institute.

To facilitate in vivo observation and the following the primary culture of 231-B cells, we constructed a pLVX-IRES-Neo vector containing luciferase gene and then produced the lentivirus by transfecting 6.5 μg pLVX-luciferase-Neo vector and corresponding amounts of package vectors into 293T cells, which were seeded in 10-cm flasks before the day of the transfection, following the standard method of Lipofectamine 2000 (Invitrogen). Forty-eight hours after the transfection, supernatant of these 293T cells were harvested, centrifuged (3,000 rpm, 4°C), filtered using 0.45-μm filter flasks, aliquoted and stored at −80°C (14).

For transduction, 10 μl supernatant containing lentivirus was added into the MDA-MB-231 cells, which were maintained in 2 ml of L-15 complete culture medium. In addition, 8 μg/ml of polybrene (Sigma-Aldrich) was also present in order to aid transduction. Twenty-four hours later, the medium were replaced by fresh L15 complemented with 10% FBS. Simultaneously, 400 μg/ml G418 was used to screen out the uninfected cells.

The bone metastasis model of breast cancer was described by Yin and colleagues (15). In general, amounts of 231-P cells were harvested using 0.25% trypsin and 0.53 mM EDTA, resuspended in sterilized PBS, and adjusted to a concentration of 5×10⁶/ml. Then, mice were anesthetized by intraperitoneal injection of 1% (w/v) pentobarbital sodium (90 mg/kg). After confirming that mice were under proper anesthesia, we injected 100 μl of the suspended cells into the left ventricle via the third intercostal space with a 29-gauge needle. A successful injection was confirmed by the pumping of arterial blood into the syringe (16). Mice were sterilized with 70% alcohol, and then were bred in the following weeks.

Mice anesthetized by 1% (w/v) pentobarbital sodium (90 mg/kg) were intraperitoneally injected with 150 mg/kg of D-Luciferin (Perkin-Elmer) in DPBS. Bioluminescence images were acquired between 10 and 15 min after injection using Xenogen Corporation Optical in vivo imaging (IVIS Lumina). Image acquisition time at the beginning was 60 sec and then it was reduced in accordance with signal strength to avoid saturation. All images were analyzed with Living Image software. Intensity of bioluminescence was calculated as photons/sec/cm²/steradian of a region of interest (ROI). Average background reads were obtained from sites of the same mice without bioluminescence.

Moreover, radiographic analysis was adopted to further confirm the bone metastases of 231-P cells. Mice were anesthetized using 1% (w/v) pentobarbital sodium (90 mg/kg) and arranged in proper position and exposed to μCT at 60 kV for 800 msec using an Inveon MM Gantry-LG CT.

To isolate tumor cells from the osteolytic lesions, mice with overt bone metastases were sacrificed, and then the affected hindlimbs were separated from the body at the joints. Both ends of the long bones were cut open after skin and muscle was removed using a scalpel. Mouse bone marrow cells as well as tumor cells were washed out by PBS using a 1-ml syringe with a 29G needle. All those cells were collected by centrifuging and washed with PBS before being cultured in 5-cm plates using L-15 medium supplemented with 10% FBS.

Mouse bone marrow cells that failed to attach to plates and thus could be washed off with PBS after the other cells (almost all of them are tumor cells) adhered to culture plates. After continuous culture using 400 μg/ml G418 for more than 1 week, we obtained the subline of bone metastasis named as 231-B cells, which were verified by measuring the luciferase activity (Promega).

Cells with about 80% confluence on 6-well plates were washed twice using ice-cold PBS, and 1 ml TRIzol (Invitrogen) was added to a well to obtain RNA. Cell lysate in TRIzol was stored at −80°C and then sent for screening differentially expressed miRNA and mRNA using Affymetrix GeneChip miRNA 3.0 Arrays and Human mRNA 4×180K (Agilent Technologies). The quantity and quality of extracted RNA were analyzed by CapitalBio. All data were normalized and further analyses were carried out by the College of Bioinformatics Science and Technology, Harbin Medical University.

Total RNA was obtained following the standard method of TRIzol extraction. cDNA was synthesized using Quant cDNA with random primers (Tiangen). Moreover, miRNAs were reverse-transcribed following the instructions with minor alteration (17). In general, one specific reverse primer targeting individual miRNA was designed to complete the reverse transcription. Then, one miRNA-specific forward primer and one universal reverse primer were used in the subsequent real-time PCR detection. Real-time PCR was completed with SYBR Premix Ex Taq™ II (Takara). The PCR procedure followed the instructions of Takara in StepOne Plus Real-time PCR system (Applied Biosystems). The results were analyzed using the 2^−ΔΔCt method (18).

Primers used in this study were the following: miR-429 forward, 5′-ACACTCCAGCTGGGTAATACTGTCTGGTAA-3′ and miR-429 reverse, 5′-CTCAACTGGTGTCGTGGAGTCGGCAATTCAGTTGAGACGGTTTT-3′, universal reverse, 5′-TGGTGTCGTGGAGTCG-3′; U6 forward, CTCGCTTCGGCAGCACA and U6 reverse, AACGCTTCACGAATTTGCGT.

HiPerFect agent (Qiagen) was used to deliver miRNA mimics into cells maintained in 6-well plates. According to the manufacturer’s instruction, cells were grown to ~70% confluence and 20 nM miRNA mimics and 12 μl of transfection agent was mixed thoroughly in 100 μl of Opti-MEM. After 10 min at room temperature, the mixture was added into cells. The subsequent experiments were performed 48 h after the transfection.

The upper chambers of Transwells with 8-μm membrane pores (Corning) were pre-coated with 60 μl Matrigel matrix gel (BD Biosciences) at least 1 h before seeding of the tested cells. A total of 3×10⁴ 231-B cells in 100 μl of L15 medium without FBS were added into the upper chambers and 600 μl of L15 medium with 10% FBS was placed to lower chambers as chemoattractant. Twelve hours later, the upper chambers were removed from lower chambers and then wiped using cotton swabs. The invaded cells were fixed using methanol at room temperature for 15 min, visualized and quantified using crystal violet. Three fields of each chamber were photographed (x10 magnification) and the results were from duplicate chambers and are presented as mean ± SD. This experiment was independently repeated at least twice.

Cells reaching 80% confluence on 6-well plates were washed twice using ice-cold PBS and 80–100 μl cell lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% Triton-100) with protease inhibitor cocktail (Complete Mini EDTA-free; Roche) added into plates. On ice, cells were carefully collected with scrapers and were subjected to lysis for 30 min. Proper amount of Laemmli buffer was added into 20–30 μg cell lysate and boiled at 95°C for 5 min. Then the denaturalized protein was resolved on a 10% gel. After being transferred to PVDF membrane, the protein blot was detected using corresponding antibodies. The immunoblot images were acquired and analyzed using ImageQuant LAS 4000 System. The primary antibodies used in this experiment were ZEB1 and β-actin.

GraphPad 6 was used to evaluate the statistical significance of data. Unless otherwise indicated, t-test was adopted to calculate the statistical significance. A P-value <0.05 was considered significant.

In the present study, we have established a mouse model of bone metastasis. The human mammary carcinoma MDA-MB-231 cells, which constitutively expressed luciferase, were inoculated into the left cardiac ventricle of immunodeficient mice (Fig. 1A). Eight to twelve weeks later, 5 mice eventually developed bone metastasis among the 15 mice under experiment. The bone lesions mainly occur on the long bones of the hind limbs and the ribs, which were observed by Xenogen Corporation Optical in vivo imaging and μCT (Fig. 1B and C).

After verifying the presence of bone metastases, we collected the metastatic tumor cells from the affected bones. G418 were added into the culture medium to wipe out non-malignant cells. After about 10 days, we obtained 231-B cells (Fig. 2A) of which morphology was different from 231-P cells. Then, we tested the luciferase activity of 231-B cells, obtaining much higher luciferase activity compared with the negative control cells (Fig. 2B). Moreover, in vitro invasion assay showed that 231-B cells had stronger invading capability than 231-P cells (Fig. 2C). Thus, we obtained bone metastatic breast tumor cells from the established model.

RNA samples from 231-P and 231-B cells were prepared and analyzed using Affymetrix GeneChip miRNA 3.0 Arrays. Among the detected miRNAs that amounted to 3,949, 118 miRNAs manifested differential expression between 231-P and 231-B cells, of which 16 upregulated whereas 102 downregulated >2-fold (P<0.05). Some of the differentially expressed miRNA are shown in Fig. 3A. Subsequent GO enrichment analysis showed that the functions of these miRNAs focused on ‘regulation of cell death’ (P=1.49xe⁻⁶), ‘signaling’ (P=3.25xe⁻⁶), ‘regulation of localization’ (P=6.31 xe⁻⁶) and ‘regulation of cell migration’ (P=0.00028).

In addition, miR-429, miR-663, miR-486-5p and miR-486-3p were verified in RNA samples derived from 231-P and 231-B cells by real-time PCR assay. Results showed that these miRNAs indeed had distinct expression patterns, indicating that they probably play important roles in bone metastasis of breast cancer cells (Fig. 3B).

Previous reports showed that miR-429 could inhibit migration and metastasis of several types of cancers. Overexpression of miR-429 inhibits invasion and promotes apoptosis in esophageal carcinoma cells by targeting Bcl-2 and SP1 (19). miR-429 inhibits cells invasion by targeting Onecut2 in colorectal carcinoma (20). Our results were in line with the above investigations.

We noted that miR-429 expression was significantly lower in 231-B cells than that of 231-P cells (Fig. 3B). We transfected miR-429 mimics into 231-B cells, which was verified by real-time PCR assay (Fig. 4A), and found that overexpression of miR-429 could dramatically reduce the migration and invasiveness of 231-B cells (Fig. 4B). Taken together, miR-429 probably regulated bone metastasis of breast cancer cells in a negative manner.

To gain insight into the roles of miR-429 inhibiting bone metastasis of MDA-MB-231 cells, several computational algorithms (including TargetScan, DIANA-microT, microRNA.org, miRDB, RNA22-HSA and PITA) were used to predict the candidate targets of miR-429, ZEB1, ZEB2, WIPF1, MARCKS, TRIM33, WASF3, CRKL, WAPAL and NTRK2 were predicted as candidate targets of miR-429 (Table I). Fig. 5A shows the possible regulation network between miRNAs and mRNAs. Then, we combined the in silico analysis of miR-429 targets and global transcriptional profile (Fig. 5B). Among all these results, CRKL was the target gene on both sides, indicating this gene was probably a fundamental node in controlling bone metastasis of breast cancer. Moreover, immunoblot assay showed that miR-429 reduced ZEB1 and CRKL significantly, which is a classical EMT inducer in breast cancer (Fig. 5C). Collectively, miR-429 negatively modulated several key invasion and metastasis inducers to suppress motility of breast cancer cells.