The osteosarcoma cell lines used in the present study, including 143B, U-2OS, KHOS-312H, KHOS-240S, UMR-106, Saos-2, HOS, T1-73, MG-63 and Hs 890.T, and the normal cell line (hFOB) were purchased from The Shanghai Institute of Cell Biology (Shanghai, China). The 293T cell line was also obtained from The Shanghai Institute of Cell Biology. The cells were maintained in RPMI-1640 medium (Sigma; Merck Millipore, Darmstadt, Germany) supplemented by 2% fetal bovine serum (FBS; Sigma; Merck Millipore) streptomycin (100 µg/ml; Sigma; Merck Millipore) and penicillin (200 U/ml; Sigma; Merck Millipore) in 5% CO₂ at 20°C. The surgically resected osteosarcoma specimens were acquired from patients at the Puai Hospital of Tongji Medical College (Wuhan, China) between June 2013 and August 2015. All patients signed formal consent forms. The experiments involving human specimens were reviewed and approved by the Ethics Committee of The Puai Hospital of Tongji Medical College (no. 2015B0013).

Total RNAs were isolated from the osteosarcoma cell lines (MG-63 and Saos-2) and human samples using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA). In total, 2 µg total RNA in a final volume of 10 µl, containing 10 mM dNTP mix (Sigma; Merck Millipore) was used to generate complementary DNA. A total of 1 mg total RNA template was annealed with 1 ml (500 ng) random primer in a sterile RNase-free micro-centrifuge tube and heated at 70°C for 5 min. Then, a mix containing 5X RT buffer (4 ml), 100 U/µl reverse transcriptase (2 ml) and 50 U/µl RNase inhibitor (1 ml) was added (Sigma; Merck Millipore). The standard TaqMan protocol was used, as previously described (14). GAPDH was used as the control. The reactions were performed using the ABI PRISM^® 7000 sequence detection system (Applied Biosystems; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocols. The thermocycling conditions were as follows: 55°C for 2 min, 95°C for 10 min followed by 35 cycles of 95°C for 20 sec and 60°C for 1 min. The expression of miR-643 was calculated using the 2^−ΔΔCq method (15). The procedure was repeated at least three times. The primer sequences were as follows: miR-643, sense 5′-GTTAGCGTGATAGCG-3′ and antisense 5′-CTGAGTAGCTGACGCTT-3′; GAPDH, sense 5′-GATGCAATTGCGCTGCATTGT-3′ and antisense, 5′-ATGAAACGTTACGTTGAT-3′.

The miR-643 mimics, miR-643 inhibitor, scramble and negative controls were all synthesized and purchased from Sigma (Shanghai, China); these were cloned into the pcDNA3.1 vector. The negative control, scramble and pcDNA-miR-643 vectors were then transfected into MG-63 and Saos-2 cells in 12-well plates using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Following incubation for 48 h at 20°C, the culture medium was replaced with fresh medium. The expression levels of miR-643 were determined using RT-qPCR analysis.

Proteins were extracted using 1X radioimmunoprecipitation assay lysis buffer (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) and concentrations were determined by Bradford protein assay (Bio-Rad Laboratories, Inc., Hercules, CA, USA) according to the manufacturer's protocols. Proteins were then subjected to 10% SDS-PAGE (100 µg/lane) and transferred onto polyvinylidene difluoride membranes (Sigma; Merck Millipore). Membranes were blocked with 3% fat-free milk for 1.5 h at 20°C and then probed with antibodies against ZEB1 (1:500; cat. no. 3396) and GAPDH (1:500; cat. no. 8884; both Cell Signaling Technology, Inc., Boston, MA, USA) at 4°C overnight. A peroxidase-conjugated secondary antibody was also used (cat. no. 7076; 1:1,000; Cell Signaling Technology, Inc.) for incubation at 20°C for 1 h. Chemiluminescence was used to visualize the results (GE Healthcare, Fairfield, CT, USA). The blots were quantified with ImageJ software v1.48 (National Institutes of Health, Bethesda, MD, USA).

A Cell Counting kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Kumamoto, Japan) was used to measure the proliferation rates of cells. Following treatment for 24 h, the MG-63 and Saos-2 cells were re-suspended and seeded into a 96-well plate (10⁶ cells per well) for 5 days. Subsequently, 10 µl MTT solution was added into each culture well at a final concentration of 10 mg/ml. The crystalline formazan was resolved in 100 µl sodium dodecyl sulfate (10%) solution for 1 day and the optical density at 490 nm was monitored using the Spectramax M5 microplate monitor according to the manufacturer's protocol (Molecular Devices LLC, Sunnyvale, CA, USA).

Transwell chambers were used to evaluate the invasion of cells (8 µm size; Sigma; Merck Millipore). The upper chamber was coated with Matrigel (Invitrogen; Thermo Fisher Scientific, Inc.) overnight. Subsequently, 1×10⁵ cells transfected with pcDNA-miR-643, miR-643 inhibitor, scramble or negative controls were seeded into the upper chambers. The lower chambers were covered in RPMI-1640 medium with 5% FBS as chemoattractants. After 24 h at 20°C, the cells remaining on the top of the chamber were removed using cotton swabs. The cells, which had migrated into the lower chamber were fixed with 5% PFA and stained with 0.05% crystal violet. A Leica fluorescent microscope (DM-IRB; Leica Microsystems GmbH, Wetzlar, Germany) was used to quantify the results.

The sections of tissue specimens (5 µm) were incubated with ZEB1 antibodies (1:500) overnight at 4°C. The sections of the specimens were then treated with secondary antibody (1:1,000) and further incubated with streptavidin-horseradish peroxidase complex for 1 h at 20°C. The nuclei were stained with hematoxylin and the number of tumor cells was evaluated in each section. Diaminobenzidine was used as the dye (Sigma; Merck Millipore) and immunohistochemical staining was monitored using an optical light microscope (Olympus Corporation, Tokyo, Japan).

Algorithms were applied to predict target genes using the TargetScan v6.2 (www.genes.mit.edu/targetscan), DIANA-microT v5.0 (www.diana.imis.athena-innovation.gr/DianaTools/index.php?r=microT_CDS/index) and miRDB v5.0 (www.mirdb.org) databases as described previously (15–17). Briefly, putative targets of miR-643 were ranked by z-scores and those ranked at the top with overlapping targets were selected for further verification.

The 3′untranslated region (UTR) of ZEB1 with predicted miR-643 binding sites were amplified using PCR and inserted into pRL-TK luciferase reporter vectors (Sigma; Merck Millipore) to obtain the ZEB1 3′UTR wild-type (WT). The ZEB1 3′UTR mutant (MUT) was obtained using primers containing the mutant sequences. The recombinant plasmids were transfected into 293T cells using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). The pRL-TK plasmid containing Renilla luciferase was used as the control. Luciferase activities were quantified using a dual luciferase assay (Promega; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol.

The results are presented as the mean ± standard deviation. Statistical significance was determined using Student's t-test with SPSS, version 16.0 (SPSS, Inc., Chicago, IL, USA). P<0.05 was considered to indicate a statistically significant difference. Spearman's correlation was used to measure the association between the expression of miR-643 and ZEB1. A Kaplan-Meier survival curve was analyzed using the log-rank test. Fisher's exact test was used to evaluate the correlation between miR-643 and different clinicopathological characteristics. All experiments were repeated at least three times.

To quantify the intrinsic levels of miR-643, RT-qPCR analysis was performed. It was found that the expression of miR-643 was significantly decreased in 104 osteosarcoma tissue samples, compared with the normal adjacent tissues (Fig. 1A). In the well-established osteosarcoma cell lines, levels of miR-643 were also downregulated, compared with that in normal bone cells (Fig. 1B). In addition, low levels of miR-643 were correlated with poor survival rates (Fig. 1C). No significant correlations were found between the expression pattern of miR-643 and age or gender (Table I). However, the level of miR-643 was significantly associated with differentiation, tumor-necrosis-metastasis stage and metastasis (Table I). These results suggested that the lower level of miR-643 in the osteosarcoma cell lines and tissue specimens contributed to the malignancy of osteosarcoma. As the Saos-2 and MG-63 cells showed the most marked downregulated expression of miR-643 among the cells, these two cell lines for used for further analysis.

The present study examined whether miR-643 can affect the malignant phenotypes of osteosarcoma cells. The Saos-2 and MG-63 cells were either left untreated or were transfected with miR-643 mimics or scramble controls. Transfection with the miR-643 mimics significantly upregulated the intrinsic levels of miR-643 in the MG-63 and Saos-2 cells (Fig. 2A and B). It was also found that miR-643 transfection led to a marked decrease in the proliferation of MG-63 cells, compared with either the empty or scramble control cells (Fig. 2C). Qualitatively similar results were observed in the Saos-2 cells (Fig. 2D). It was also noted that transfection with miR-643 inhibitor reversed this effect and promoted proliferation (Fig. 2C and D). Accordingly, miR-643 transfection inhibited the invasion of MG-63 cells (Fig. 2E) and Saos-2 cells (Fig. 2F) and treatment with miR-643 inhibitor enhanced cell invasion (Fig. 2E and F). These results suggested that miR-643 inhibited the malignancy of osteosarcoma cells through inhibiting proliferation and invasion.

To identify possible targets of miR-643 in osteosarcoma, the present study used bioinformatics strategies. The online databases DIANA-microT, miRDB and TargetScan were used for target prediction. Overlapping results suggested ZEB1 may be a direct target. The base pairing between hsa-miR-643 and ZEB1 was determined (Fig. 3A). A luciferase reporter assay was then performed, which confirmed that the luciferase intensities were significantly downregulated in WT ZEB1 (Fig. 3B). Introducing mutations in the base pairing regions of ZEB1 showed minimal difference, compared with the control group (Fig. 3B). In addition, the miR-643 mimics decreased the mRNA levels of ZEB1 (Fig. 3C and D). Consistently, the protein expression of ZEB1 was downregulated with miR-643 transfection (Fig. 3E and F). These results suggested that ZEB1 may be a direct target of miR-643.

To further validate the direct target of ZEB1 by miR-643, rescue experiments were performed. pcDNA3.1 was used to overexpress ZEB1 in the MG-63 and Saos-2 cells. The results showed that pcDNA3.1-ZEB1 transfection significantly upregulated intrinsic levels of ZEB1, compared with those in the empty control (Fig. 4A and B). Consistently, ZEB1 transfection increased the proliferation of MG-63 cells compared with the cells in the empty control group (Fig. 4C). Similar results were obtained for the Saos-2 cells (Fig. 4D). It was also noted that the overexpression of ZEB1 markedly enhanced the invasive capacities of the MG-63 and Saos-2 cells (Fig. 4E). These results further confirmed that miR-643 mediated its tumor suppressive effect by targeting ZEB1.

To further confirm the role of miR-643, pairwise analysis was performed on the expression of miR-643 and ZEB1. It was found that the expression of miR-643 was inversely correlated with that of ZEB1 in the tissue specimens (Fig. 5A). In addition, the overall survival rate for the patients expressing a low level of ZEB1 was 52.1%, but was only 25.2% in the group with a high expression of ZEB1 (Fig. 5B). In tissue specimens with relatively higher expression levels of miR-643 (miR-643+), it was found that the ZEB1 staining was significantly lower, compared with that in specimens with low expression levels of miR-643 (miR-643-; Fig. 5C). The samples exhibiting relatively lower expression levels of miR-643 showed characteristics of metastatic osteosarcoma, compared with the normal adjacent tissues (Fig. 5C). These characteristics were less marked in osteosarcoma samples with higher expression levels of miR-643 (Fig. 5C). These results further suggested that the downregulation of miR-643 was correlated with higher expression levels of ZEB1 and poor prognosis.