A total of 60 pairs of cancer tissue and corresponding normal tissue samples were collected from patients with primary osteosarcoma who received surgery between 2009 and 2011 at Jinshan Hospital, Fudan University School of Medicine (Shanghai, China). Surgically removed tissues were snap frozen in liquid nitrogen overnight and kept at −80°C before use. Each specimen was verified histopathologically and scored according to the TNM staging system. The study protocol was approved by the Ethics Committee of Fudan University, Shanghai, China. All samples were collected after obtaining informed consent from all patients.

Total RNA was isolated from the osteosarcoma cell lines and tumor tissue samples using the miRNeasy kit (Qiagen, Valencia, CA, USA). To detect miRNA expression, RT-qPCR was performed using the SYBR RT-PCR kit (Takara, Shiga, Japan). The primers for TAGLN were as follows: 5′-AATGGCG TGATTCTGAGCAA-3′ (forward) and 5′-CGATGCCTGCTG GAGCCGTCTA-3′ (reverse). The relative expression level of TAGLN was normalized to the internal control, β-actin. Stem-loop RT primers for miR-144 were as follows: 5′-GCTGGGA TATCATCATATACTG-3′ (forward) and 5′-CGGACTAGTA CATCATCTATACTG-3′ (reverse), as previously described (16). The primers used for quantitative PCR were as follows: 5′-ACACTCCAGCTGGGTCATGTA GTAGATA-3′ (forward) and 5′-CTCAACTGGTGTCGTGGAGTCGGCAATTCAG TTGAGATGTCATA-3′ (reverse). The relative expression level of miR-144 was normalized to that of the internal control U6 by using the 2^−ΔΔCt cycle threshold method, as previously described (17).

The human normal osteoblastic cell line, hFOB1.19, and the human osteosarcoma cell lines, MG63, U2OS and HOS, were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen, Carlsbad, CA, USA) supplemented with heat-inactivated 10% fetal bovine serum (FBS) at 37°C in a humidified incubator with 5% CO₂.

The procedure used for miR-144 eukaryotic expression vector construction was as previously described (7). The following primers were used for PCR amplification of the pri-miR-144 and native flanking sequence: 5′-GGATCCCAC AGTGCTTTTCAAGCCATG-3′ and 5′-AAGCTTAGTGCC CTGGCAGTCAGTAGG-3′. The amplified fragments were then cloned into the pcDNA3.1 vector (Invitrogen) and verified by sequencing.

Transfection was performed using Lipofectamine 2000 transfection reagent (Invitrogen) according to the manufacturer’s instructions. For establishing stable transfectants, the miR-144 mimics or negative control (NC) RNA-tansfected HOS and U2OS cells were selected for 6 weeks in the presence of G418 (400 μg/ml). Control cells were transfected with empty vector. The expression of miR-144 in the stably transfected cells was evaluated by RT-qPCR. TAGLN shRNA Lentivirus (TAGLN-shRNA) and Negative Control shRNA Lentivirus (shRNA-NC) purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA) were used to tansfect the HOS and U2OS cells following the manufacturer’s instructions and stably tansfected cells were selected by puromycin (Sigma, St. Louis, MO, USA).

The in vitro proliferation of the HOS and U2OS cells transfected with miR-NC, miR-144 mimics, TAGLN-shRNA or shRNA-NC was measured by 3-(4,5-dimethythiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay as previously described (12). Briefly, the cells were seeded onto 96-well plates and at the indicated periods to time, 100 μl of medium was removed and replaced with fresh medium containing 0.5 mg/ml MTT solution. The plates were incubated at 37°C for 4 h and then the medium was replaced by 100 μl DMSO and the plates were shaken at room temperature for 10 min. Absorbance was measured at 570 nm.

The invasive potential of the cells was evaluated using Transwell inserts with 8 mm pores (Corning Inc., Corning, NY, USA). For invasion assay, at 24 h after transfection, 3.0×10⁵ cells in serum-free medium were added to each upper insert pre-coated with Matrigel matrix (BD Biosciences, Bedford, MA, USA). Approximately 500 ml of 10% FBS medium was added to the matched lower chamber. After 48 h of incubation, the non-invading cells were removed from the upper surface of the Transwell membrane with a cotton swab, and the invading cells on the lower membrane surface were fixed in methanol, stained with 0.1% crystal violet, photographed and counted. Six random fields at ×100 magnification for each insert were counted. The inserts were analyzed in triplicate in 3 separate experiments.

The micro-RNA database (www.microRNA.org) was used to find the putative miR-144 binding site in the 3′ UTR of TAGLN (433 bp). The psiCHECK-2-TAGLN-3′UTR-wt vector or the psiCHECK-2-TAGLN-3′UTR-mut vector were co-transfected with control, miR-NC or miR-144 into the HOS cells using Lipofectamine 2000 (Invitrogen). Subsequently, reporter gene assays were performed 48-h post-transfection using the Dual Luciferase Reporter assay system (Promega, Madison, WI, USA) according to the manufacturer’s instructions. The normalized firefly luciferase activity was obtained by measuring firefly luciferase activity/Renilla luciferase activity. All experiments were performed at least 3 times.

A total of 54 BALB/c athymic nude mice were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). All animal protocols and procedures were approved by the Institutional Animal Care Utilization Committee, and were in accordance with the Johns Hopkins Animal Care and Use Committee guidelines. HOS cells stably transfected with miR-144 mimics, negative control (NC) RNA and control (1×10⁶) were suspended in 0.1 ml PBS and then injected subcutaneously into either side of the posterior flank of the same BALB/c athymic nude mice at 4 weeks of age. Tumor growth was measured using calipers weekly, and tumor volume was calculated according to the following formula: volume = length × width² × 0.5.

Lysates obtained from the cells and tissue samples were extracted using the M-PER Protein Extraction Reagent (Pierce, Rockford, IL, USA), and the protein content was quantified using the BCA assay (Pierce). Equal amounts of extracts were subjected to SDS-PAGE, transferred onto polyvyniledene difluoride membranes and probed with primary antibodies against TAGLN (Sigma), matrix metalloproteinase 2 (MMP2; Abcam, Cambridge, MA, USA) and β-actin followed by the appropriate secondary antibodies (Cell Signaling Technology). Densitometric analysis was performed using LabWorks image acquisition and analysis software (UVP, Upland, CA, USA).

Data are presented as the means ± SD. Statistical comparisons between groups were analyzed using a two-tailed Student’s t-test and a P-value <0.05 was considered to indicate a statistically significant difference. The correlation between miR-144 expression and clinical osteosarcoma stages was analyzed using the Spearman’s rank correlation assay with SPSS 17.0 software (SPSS Inc., Chicago, IL, USA). The correlation between miR-144 expression and TAGLN mRNA levels was analyzed using the Spearman’s correlation coefficient assay with SPSS 17.0 software.

To determine the role of miR-144 in the development of human osteosarcoma, we analyzed miR-144 expression in the human osteosarcoma cell lines, MG63, U2OS and HOS, by RT-qPCR and compared it to that in the normal human osteoblastic cell line, hFOB1.19. The results revealed that miR-144 was significantly downregulated in the osteosarcoma cell lines compared to the normal osteoblastic cells (Fig. 1A; P<0.05). We then analyzed miR-144 expression in the paired human osteosarcoma tumor tissue and adjacent normal tissue samples; the results revealed a significant downregulation of miR-144 expression in tumor tissues compared to adjacent normal tissues (Fig. 1B; P<0.05). To determine whether the downregulation of miR-144 is correlated with the progression and development of osteosarcoma, we used Spearman’s rank correlation, which revealed a significant negative correlation between miR-144 expression in tumor tissue and clinical stage (Fig. 1C). These results suggest that the downregulation of miR-144 expression plays a role in the progression of osteosarcoma.

To further investigate the biological role of miR-144 in osteosarcoma tumor progression, the HOS and U2OS cell lines were infected with lentivirus carrying a negative control miR (miR-NC) or miR-144 mimics and miR-144 expression was analyzed by RT-qPCR 72 h after infection. The results revealed that miR-144 expression was significantly upregulated in the miR-144-infected cells compared to the cells infected with miR-NC (Fig. 2A) (P<0.01). To determine the effects of miR-144 on osteosarcoma cell growth, the viability of the HOS and U2OS cells treated as described above was assessed by MTT assay at the indicated time points. The results revealed that transfection with miR-144 mimics significantly decreased the growth of osteosarcoma cells compared to the control- or miR-NC-transfected cells (Fig. 2B; P<0.05 at 72 and 96 h). Furthermore, the assessment of cell invasion by Transwell assay revealed a significant impairment in the invasive ability of the cells expressing miR-144 mimics compared to that of the control- or miR-NC-transfected cells (Fig. 2C; P<0.05).

To examine the correlation between miR-144 and TAGLN in osteosarcoma, the expression of TAGLN was analyzed by RT-qPCR in the osteosarcoma cell lines and primary tumor tissue samples. The results revealed that the TAGLN mRNA level was significantly higher in the osteosarcoma cell lines, U2OS and HOS, than in the control human osteoblast cell line, hFOB1.19 (Fig. 3A; P<0.05), and it was significantly higher in the human primary osteosarcoma tumor tissue samples than in the adjacent normal tissue samples (Fig. 3B; P<0.05). To identify a potential correlation between the expression of miR-144 and TAGLN in osteosarcoma tissue, we used Spearman’s correlation, which revealed a significant inverse correlation between miR-144 and TAGLN mRNA expression (Fig. 3C; R=−0.465, P<0.001), confirming that the downregulation of miR-144 is associated with the upregulation of TAGLN in osteosarcoma.

To further determine whether TAGLN is the primary regulator of cell invasion and migration in osteosarcoma, TAGLN was stably silenced in the HOS and U2OS cells. The cells were transfected with TAGLN-shRNA and stable transfectants were obtained after 48 h. The expression of TAGLN was determined by western blot analysis, which revealed that the protein levels of TAGLN were significantly decreased in the HOS and U2OS cells (both P<0.05). HOS and U2OS cells stably depleted of TAGLN were cultured and cell viability and invasion were assessed by MTT and Transwell assays. As shown in Fig. 4B, the depletion of TAGLN significantly (P<0.05) reduced the viability of the HOS and U2OS cells compared with the respective negative control shRNA-transfected cells. The results of Transwell assay revealed that the silencing of TAGLN significantly inhibited the invasive ability of the HOS and U2OS cells compared to that of the control- or negative control shRNA-transfected cells (Fig. 4C; P<0.05).

Potential miR-144 binding sites in the 3′ UTR of TAGLN were predicted using micro-RNA.org (www.microRNA.org). A sequential replacement of a 5 base pair region was performed to produce a mutant vector (Fig. 5A). To further investigate whether the predicted binding site of miR-144 in the 3′ UTR of TAGLN is responsible for the interaction, the 3′ UTR of TAGLN was cloned downstream to a luciferase reporter gene (wt-TAGLN), and a mutant version (mut-TAGLN) was generated by binding site mutagenesis. The wt-TAGLN vector or mut-TAGLN with control, negative control or miR-144 mimics was cloned into the HOS cells and the luciferase activity was determined. The results revealed that the luciferase activity of the miR-144-transfected cells was significantly reduced compared with that of the negative control-transfected cells (Fig. 5B). To further confirm that TAGLN is a direct target of miR-144, the protein levels of TAGLN were assessed by western blot anlaysis in the cells transfected with miR-144 mimics, which indicated that the exogenous expression of miR-144 significantly downregulated TAGLN expression in the HOS and U2OS cells (Fig. 5C).

The osteosarcoma cell line, HOS, was stably transfected with miR-144 mimics and negative control (NC) RNA, as previously described (18) and implanted subcutaneously into nude mice to examine tumorigenicity in vivo. The effects were most prominent in the first week after inoculation, in which a significant suppression of tumorigenicity by miR-144 was observed as determined by the significantly smaller tumors formed in the mice (Fig. 6A). Fig. 6B shows the reduction in tumor volume in mice carrying miR-144-overexpressing tumors compared to the control mice or mice transplanted with negative control-transfected cells. To confirm the effects of miR-144 and its target, TAGLN, on invasion in vivo, the expression of proteins involved in cell adhesion and migration was determined. The results revealed that the expression of MMP2 was downregulated by TAGLN-siRNA or by miR-144 overexpression (Fig. 6C). Taken together, these results indicate that miR-144 functions as a tumor suppressor in osteosarcoma through the modulation of its target gene, TAGLN.