Introduction

Osteosarcoma is one of the most common malignant tumors of the bone in children and adolescents [1]. The etiology of this neoplasm is complex and there are multiple risk factors, such as genetic and environmental factors, that may contribute to the complicated disease [2]. Though advances of modern treatments such as surgery, chemotherapy, and the combination of surgery and chemotherapy are improved, long-term survival rate of patients diagnosed with osteosarcoma remains very low [3, 4]. Thus, a better understanding of the biology of this malignancy is necessary to develop novel treatment strategies.

MicroRNAs (miRNAs) are highly conserved, small noncoding RNAs that play important roles in many biological processes by negatively regulating their target genes post-transcriptionally [5]. Growing evidence suggests that miRNA expression is altered in many human diseases, including cancer [6–8]. Aberrant miRNA functions as tumor suppressor genes or oncogenes and contributes to tumorigenesis and cancer metastasis [9–11]. Numerous miRNAs have been found having different expression in osteosarcoma tissues compared with corresponding noncancerous tissues by miRNA expression profile [12, 13]. Cheng et al. demonstrated that miR-320 was significantly downregulated in human osteosarcoma tissues and overexpression of miR-320 inhibited cell proliferation and induced cell apoptosis [14]. Shen et al. found that miR-128 was significantly increased in osteosarcoma tissues and introduction of miR-128 promoted cell proliferation by directly targeting PTEN 3′-untranslated region [15]. MiR-451 was one of the most significantly downregulated miRNAs and the mechanism of the decreasing of miR-451 has been rarely investigated. Recently, miR-451 was reported to be frequently downregulated in many types of cancers and acted as a tumor suppressor involved in cellular function, such as cell proliferation and growth, cell cycle distribution, migration, and invasion [16–20]. These findings hypothesizes that the expression of miR-451 in osteosarcoma cancer cell is correlated with malignant behavior.

In the present study, we confirmed the expression of miR-451 in osteosarcoma tissue specimens and osteosarcoma cell lines by using quantitative reverse transcription-polymerase chain reaction (qRT-PCR). Then the relationship between miR-451 expression and clinicopathologic features is analyzed, and the effects and mechanisms of miR-451 on osteosarcoma cells growth and invasion are further investigated.

Materials and methods

Cell culture and clinical specimens

The human osteosarcoma cell lines U2OS and SaOS2 and the normal osteoblastic cell line hFOB 1.19 were purchased from the Type Culture Collection of the Chinese Academy of Sciences. Human embryonic kidney cells HEK293T were obtained from American Type Culture Collection (Manassas, VA, USA) and maintained in Dulbecco’s modified Eagle’s medium (HyClone, Logan, UT, USA). hFOB 1.19 cells were cultured in DMEM/F-12 (1:1; HyClone, Logan, UT, USA), and MG-63 and HOS cells were maintained in Eagle’s minimum essential medium. All the mediums were supplemented with 10 % fetal bovine serum (FBS; Gibco, NY, USA), 1 % penicillin/streptomycin, and 2 mM glutamine. Cells were incubated at 37 °C in 5 % CO2-humidified atmosphere.

This study was approved by the Ethics Committee of Ningbo Development Zone Center Hospital with the permit number of 2013F0164. A total of 68 freshly frozen osteosarcoma tumor tissues and adjacent normal tissues were obtained from Ningbo Development Zone Center Hospital between May 2009 and April 2012. No patients had received any treatments prior to surgery. Written informed consent was obtained from all of the patients. After resection, all tissue samples were immediately frozen in liquid nitrogen and stored at −80 °C until RNA extraction.

RNA extraction and quantitative real-time PCR

Total RNA was extracted from each cell line and fresh-frozen specimens using TRIzol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. In the present study, we confirmed the expression of miR-451 with stem-loop RT-PCR using an All-in-One™ miRNA quantitative RT-PCR (qRT-PCR) Detection Kit (GeneCopoeia, Rockville, MD) according to the manufacturer’s protocols. Real-time PCR was carried out with SYBR green detection with a forward primer for the mature miRNA sequence and a universal adaptor reverse primer. The U6 functioned as the endogenous control for miR-451. The specific forward primer for U6 was 5′-GCTCGCTTCGGCAGCACA-3′. The specific primer for miR-451 was 5′-CCAAACCGTTACCATTACTGAGTT-3′.

To measure messenger RNA (mRNA) level of chemokine receptor 16 (CXCR16), 500 ng of total RNA was reverse-transcribed into cDNA using the ReverTra Ace (Toyobo, Osaka, Japan). qRT-PCR was performed by using SYBR Green PCR Master Mix (Toyobo, Osaka, Japan). The primers for CXCR16 used were 5′-GACATGCTTACTCGGGGATTG-3′ (forward) and 5′-GGACAGTGATCCTACTGGGAG-3′ (reverse). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an endogenous control, and the primers for it were 5′-ACCCAGAAGACTGTGGATGG-3′ (forward), 5′-CAGTGAGCTTCCCGTTCAG-3′ (reverse).

Quantification of miRNAs and mRNA by qRT-PCR was performed using an ABI 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. All reactions were repeated three times and reactions without reverse transcriptase were used as negative controls.

Stable cell generation

Lentiviral construct miRZip-451 (MZIP451-PA-1) and its control vector (MZIP000-PA-1) were obtained from System Biosciences (Mountain View, CA, USA). HEK293T cells were cotransfected with the lentivirus-mediated miR-451 or control vector, pPACKH1 Lentivector Packaging KIT (SBI), using Lipofectamine™ 2000 (Invitrogen) according to the manufacturer’s instructions. U2OS and SaOS2 were transduced with miR-451 or control vector and miR-451 expression was evaluated by qRT-PCR.

Target gene indentify and luciferase reporter assay

Prediction of miR-451 binding to the 3′-untranslated region (3'UTR) of genes downregulated by miR-451 was performed using DIANA-microT (available at http://diana.imis.athena-innovation.gr/DianaTools/index.php?r=microT_CDS/index) [21] and microRNAorg (available at http://www.microrna.org/microrna/home.do) [22].

The chemokine ligand 16 (CXCL16) 3′-UTR region was amplified by PCR from human genomic DNA with the following primers: 5′-AAAACTAGTGCCAAGAATGGAAGCTTGTGA-3′ (forward) and 5′-AAAACGGCTCATTCGGAGAGACAAAACAAGAAC-3′ (reverse). The PCR product was then subcloned into a pMIR-REPORT vector (Ambion, Austin, TX, USA) immediately downstream of the luciferase gene to generate CXCL16-WT (wild type). A construct containing the CXCL16 3′-UTR with point mutations in the seed sequence at positions 203–209 was amplified from the wild vector to generate CXCL16-MUT (mutation type). The primers used were 5′-CCCCTCATTTAAAAGGCCGTATACTATAA-3′ (forward) and 5′-TTATAGTATACGGCCTTTTAAATGAGGGG-3′ (reverse). For the luciferase reporter assay, HEK293T cells were seeded in 96-well plates the day before transfection and cotransfected with the CXCL16-WT or CXCL16-MUT vector, the pMIR-REPORT β-gal Control Plasmid (Ambion), and miR-451 or control vector with a 50 nM final concentration using Lipofectamine™ 2000 (Invitrogen). At 24 h after transfection, luciferase and β-galactosidase activities were determined with the Luciferase Assay Kit (Promega, Madison, WI, USA) and ß-Galactosidase Enzyme Assay System (Promega), respectively. The luciferase activity was normalized to the β-Gal activity from the same cells.

Cell proliferation assay and invasion assay

Changes in cell proliferation were evaluated using the Cell Counting Kit-8 (Dojindo, Japan) as previously described [23]. Logarithmic growth phase U2OS and SaOS2 cells (8 × 103 cells/well) transduced with control lentivirus or miR-451 lentivirus were seeded in 96-well plates, and the cell viability at different time points (24, 48, and 72 h) was measured at 450 nm with an enzyme immunoassay analyzer (Bio-Rad, Hercules, CA, USA). Each experiment was repeated three times and the results were obtained from three independent experiments.

Invasion activities of the cells were determined as previously described [23]. Briefly, 3 × 104 cells were suspended in 100 μl of serum-free medium in the upper wells and the bottom wells were added complete medium. Finally, the cell numbers were counted from five nonoverlapping fields of each membrane. Three independent assays were performed.

Small interfering RNA interference for CXCL16 and cell transfection

To silence CXCL16 expression, small interfering RNA (siRNA) targeted to CXCL16 and negative control siRNA were purchased from GenePharma (Shanghai, China). Cells were transfected with siRNA at a final concentration of 50 nM using Lipofectamine™ 2000 according to the manufacturer’s protocols. The protein levels were detected 48 h later.

Western blotting

Western blot was carried out according to the previous description [23]. Equal amounts of protein were separated using 10 % SDS-PAGE gels and transferred to PVDF membrane (Bio-Rad, CA, USA). Then, the membrane was incubated with rabbit polyclonal anti-CXCL16 antibody (Abcam, Cambridge, MA, USA) and rabbit polyclonal anti-GAPDH antibody (CWBiotech, Beijing, China) overnight at 4 °C. Proteins were visualized with enhanced chemiluminescence reagents (Hyperfilm ECL; Amersham Biosciences, Buckinghamshire, England) and the signal detected using an ImageQuant™ LAS 4000 mini biomolecular imager (Fujifilm, Tokyo, Japan). The change in protein expression levels of CXCL16 were normalized to GAPDH.

Statistical analysis

All quantified data were presented as mean ± SD and differences between two groups were studied using the two-sided Student’s t test. The relationship between the miR-451 expression levels and clinicopathologic characteristics was analyzed using the Pearson chi-square test. Pearson correlation was calculated to estimate the correlation between miR-451 values and CXCL16 levels in osteosarcoma tumor specimens. All statistical analyses were performed using SPSS 17.0 software (SPSS, Chicago, IL, USA). P value <0.05 was considered statistically significant.

Results

MiRNA-451 expression is significantly lower in osteosarcoma tissues and cell lines

Previous microarray results suggested that miR-451 was significantly downregulated in osteosarcoma [12, 13]. To further confirm this observation, we used qRT-PCR to assess miR-451 expression in a total of 68 matched osteosarcoma tissues and adjacent normal tissues. The expression of miR-451 was significantly lower in osteosarcoma tissues than that in normal tissues (Fig. 1a, P < 0.01). We also found that levels of miR-451 expression were significantly lower in U2OS and SaOS2 osteosarcoma cell lines than in normal osteoblastic cell line hFOB 1.19 (Fig. 1b). Together these results provide strong evidence that miR-451 is markedly downregulated in osteosarcoma.

Fig. 1
figure 1

The relative expression levels of miR-451 in osteosarcoma tissues and cell lines examined by qRT-PCR. a Expression level of miR-451 was lower in 68 osteosarcoma tissues than in their pair-matched adjacent non-tumor samples (P < 0.01). b The relative miR-451 expression in osteosarcoma cell lines was much lower than that in the normal osteoblastic cell line hFOB 1.19. The relative expression of miR-451 was normalized to the endogenous control U6. Each sample was analyzed in triplicate. **P < 0.01

Full size image

Association of miR-451 expression with clinicopathological features of osteosarcoma patients

To further elucidate correlations between the expression level of miR-451 and clinicopathologic features in osteosarcoma samples, the 68 patients with osteosarcoma were classified into two groups according to the median expression of miR-451. We compared the clinicopathological factors of the high and low miR-451 expression group (Table 1) and found that low expression of miR-451 was significantly correlated with metastasis (P = 0.023) and recurrence (P = 0.027). However, miR-451 expression levels did not show any relationship with age, gender, histologic cell type, or TNM stage. The results provided further support for the idea that altered miR-451 expression may be an important regulator of aggressive biological behavior in osteosarcoma.

Table 1 Association between miR-451 expression and conventional clinicopathological features in 68 patients with osteosarcoma
Full size table

MiR-451 inhibits cell proliferation and migration in vitro

We next undertook gain-of-function analyses to determine the role played by miR-451 in osteosarcoma cell proliferation and invasion in vitro. U2OS and SaOS2 osteosarcoma cell lines were stably transduced with a recombinant lentivirus encoding a miR-451 precursor microRNA and were validated by qRT-PCR (Fig. 2a). Overexpression of miR-451 significantly inhibited cell proliferation in both cell lines (Fig. 2b). The effect of miR-451 expression on osteosarcoma cell migration was assessed using a standard Matrigel invasion assay. As shown in (Fig. 2c), enforced expression of miR-451 in U2OS and SaOS2 cell lines significantly suppressed their invasion compared to cells expressing empty vector.

Fig. 2
figure 2

MiR-451 inhibited osteosarcoma cell proliferation and invasion. a qRT-PCR analysis of miR-451 in U2OS and SaOS2 cells transduced with miR-451 or control vector. Total RNA was isolated 96 h after transduction for the detection of miR-451 by qRT-PCR and normalized to U6 expression. b Growth of the U2OS and SaOS2 cells with miR-451 overexpression or with control vector at an indicated time. c Representative images of basement membrane matrix invasion assay in U2OS and SaOS2 cells transduced with miR-451-expressing or control vector. Quantitative analysis showed the average number of invaded cells counted microscopically in five different fields per filter. **P < 0.01

Full size image

MiR-451 directly targets 3′-UTR of the CXCL16 gene

To understand the mechanism by which miR-451 inhibits cancer cell proliferation and invasion, we identified CXCL16 as a target of miR-451 using the bioinformatics algorithms. CXCL16 mRNA has one potential complimentary binding site with miR-451 within its 3′-UTR at 203–209 bp (Fig. 3a). A luciferase reporter assay was performed to validate whether miR-451 could directly target the CXCL16 3′-UTR. Reporter assays demonstrated that transfection of miR-451, but not of control vector, significantly decreased CXCL16 3′-WT-dependent luciferase activity (Fig. 3b). No difference was observed in HEK293T cells carrying CXCL16 3′-MUT (Fig. 3b). Moreover, we evaluated the effects of miR-451 on endogenous mRNA and protein expression of CXCL16 in U2OS and SaOS2 cells. Cells stably expressing miR-451 showed a significant attenuation of CXCL16 mRNA (Fig. 3c) and protein expression (Fig. 3d). This finding supported the hypothesis that CXCL16 was a direct, downstream target for miR-451 in osteosarcoma cells.

Fig. 3
figure 3

MiR-451 binds to the 3′-UTR of CXCL16 mRNAs and downregulates expression. a The 3′-UTR of CXCL16 is predicted to be miR-451 binding site. b Luciferase activity of wild-type or mutant-type CXCL16 3′-UTR reporter gene in HEK293T cells transfected with miR-451-expressing or control vectors. c The mRNA levels of CXCL16 in U2OS and SaOS2 cells transduced with miR-451-expressing or control vectors were determined by qRT-PCR analysis. d Protein levels of CXCL16 in U2OS and SaOS2 cells transduced with miR-451-expressing or control vectors were determined by Western blot analysis. GAPDH was used as an internal control and a representative experiment was shown in triplicate. **P < 0.01

Full size image

Effect of CXCL16 on osteosarcoma cell proliferation and invasion

We next assessed whether CXCL16 was a functional target of miR-451, we knocked down CXCL16 using a siRNA technique. The knockdown effect was confirmed by measuring protein expression levels. As shown in (Fig. 4a), the relative CXCL16 protein levels were significantly reduced by Western blot analysis in siRNA-transfected cells 48 h post-transfection. A significant reduction in the proliferation rate was observed with the CCK8 assay 6 days after transfection with siRNA-mediated knockdown of CXCL16 when compared to negative control siRNA both in U2OS and SaOS2 cells (Fig. 4b). Invasion assays for U2OS and SaOS2 cells following siRNA-mediated knockdown demonstrated a significant reduction in invasion potential following knockdown of CXCL16 expression (Fig. 4c). These results indicated that knockdown of CXCL16 could recapitulate the diminished proliferation and invasion observed upon upregulation of miR-451.

Fig. 4
figure 4

siRNA-mediated functional effects of CXCL16. a Western blot analysis demonstrated significant reductions in the protein levels of CXCL16 in U2OS and SaOS2 cells 48 h post-transfection. b Cell growth of U2OS and SaOS2 cells transfected with CXCL16 siRNA or a negative control (NC) siRNA was monitored with the CCK8 assay at 48 h after transfection. c U2OS and SaOS2 cells demonstrated significant reductions in invasion following CXCL16 silencing. Quantitative analysis showed the average number of invaded cells counted microscopically in five different fields per filter. **P < 0.01

Full size image

Upregulation of CXCL16 is inversely correlated with miR-451 expression in osteosarcoma

Previous report implicated that CXCL16 were highly expressed in cancer tissues. But its expression levels were not clear in osteosarcoma tissues. Importantly, recent study showed that CXCL16 were highly expressed in cancer tissues [24, 25]. Then, we used qRT-PCR to detect the expression level of CXCL16 in osteosarcoma tissues and adjacent non-tumor tissues. The mRNA expression level of CXCL16 showed a significantly higher level in osteosarcoma tissues, compared with adjacent noncancerous tissues (P < 0.01, Fig. 5a). In further analysis, our data indicated that miR-451 expression level was inversely correlated with CXCL16 mRNA expression level, which was verified by Pearson’s correlation coefficient test (r = −0.77, P < 0.001; Fig. 5b). Taken together, our data demonstrated that miR-451 medicated CXCL16 and suppressed its expression at translation level in osteosarcoma.

Fig. 5
figure 5

Inverse correlation between the expression of miR-451 and CXCL16 in patients with osteosarcoma. a CXCL16 mRNA expression in 68 paired osteosarcoma tissues and their corresponding adjacent non-tumorous tissues by qRT-PCR analysis. The expression of CXCL16 was normalized to GAPDH. **P < 0.01. b Significant inverse correlations between miR-451 and CXCL16 mRNA expression (Pearson’s correlation r = −0.77, P < 0.001) in osteosarcoma tissues

Full size image

Discussion

Emerging evidence suggests that gene dysregulation, including protein-coding genes, microRNA, and long noncoding RNAs, may affect cellular growth advantage, resulting in progressive and uncontrolled tumor growth and metastasis [26–28]. Effective control of both cell proliferation and cell invasion is critical to the prevention of tumorigenesis and to successful cancer therapy. MiRNAs have been found to be localized to genomic fragile sites and involved in cancer development and progression [8, 11]. Therefore, identification of cancer-associated miRNAs and investigation of their clinical significance and functions may provide a new perspective on the cancer therapy.

MiR-451 has also been reported to be a tumor suppressor gene in several cancers [16–18, 29]. Previous report implicated that miR-451 was downregulated in osteosarcoma tissues by miRNA array analysis and was confirmed by qRT-PCR, but the mechanisms of miR-451 on osteosarcoma is not clear [12, 13]. Consist with prior results, we found that miR-451 expression was significantly lower in osteosarcoma tissues compared to pair-matched adjacent non-tumor samples. Thus, we investigated the relationship between miR-451 expression and clinical parameters and found that miR-451 expression was correlated to metastasis and recurrence after operation in osteosarcoma cancer patients. A recent study in osteosarcoma demonstrated that miR-451 suppresses cell proliferation by directly regulating PGE2 and CCND1 [30]. Our findings showed that ectopic expression of miR-451 in osteosarcoma cells could suppress cell proliferation and invasion. These results indicated that miR-451 has important roles in regulating cell proliferation and invasion.

Identification of putative targets is important for exploring the mechanisms underlying the inhibition of osteosarcoma cell growth and invasion mediated by miR-451. In our study, by using bioinformatics analysis and a dual-luciferase report assay, we identified that miR-451 directly downregulated CXCL16 by binding their 3′-UTR sites. Moreover, the ectopic expression of miR-451 decreased expression of CXCL16 mRNA and protein level. These results indicate that CXCL16 is a direct target for miR-451.

Recent reports implicated that CXCL16 were highly expressed in cancer tissues in the regulation of cellular proliferation [24, 31]. Furthermore, recent data also indicated its overexpression has been associated with stem cell character [32]. However, its roles of apoptosis, migration, invasion, and differentiation are not clear. Next we assessed the mRNA expression levels of CXCL16 in osteosarcoma tissues and pair-matched adjacent non-tumor samples and found that CXCL16 was significantly increased in osteosarcoma tissues. CXCL16 silencing by transfection with siRNA in U2OS and SaOS2 cell lines reduced cell proliferation and inhibited cell invasion. These effects are in agreement with previously reported role of miR-451, and thus downregulation of CXCL16 may underlie such functional consequences of miR-451 overexpression. In fact, it will be of interest to identify novel targets of miR-451 to elaborate key regulatory mechanisms to control cell proliferation and tumor metastasis.

Conclusions

In summary, in this paper, we presented that miR-451 was downregulated and highlighted the clinical and functional implications of miR-451 expression. We also found that miR-451 can inhibit cell growth and invasion partly by targeting CXCL16. Our study provided that restoring miR-451 levels may have therapeutic effects in osteosarcoma.