miR-766-3p suppressed tumorigenesis, epithelial-mesenchymal transition, and metastasis by targeting BCL9L via β -catenin signaling pathway in osteosarcoma cells

Emerging evidence has indicated that abnormal microRNAs (miRNAs) play critical roles in carcinogenesis and progression of osteosarcoma (OS). The aim of this study was to clarify the relationship between miR-766-3p expression and osteosarcoma development and to explore its potential mechanism. miR-766-3p was the most downregulated miRNA by analyzing GSE65071 from the GEO database. RT-PCR and western blot was performed to determine miR-766-3p expression and its specic target gene in human OS samples and cell lines. CCK-8 proliferation, colony formation, EdU, wound-healing, and transwell assays were used respectively to evaluate the inuences of miR-766-3p depletion or ectopic expression on OS proliferation, migration and invasion in vitro. And a mouse tumorigenicity model was conducted to investigate effects of miR-766-3p in vivo. Moreover, we identied directly interactions between miR-766-3p and its specic target gene using luciferase reporter assays. The β-Catenin signaling pathway was as being implicated in the miR-766-3p/BCL9L


Introduction
Osteosarcoma (OS), a primary high-grade malignant bone neoplasm arising from mesenchymal cells, has high mortality worldwide [1]. In recent years, regardless of the continuous development of treatments including chemotherapy techniques and various surgical methods, the 5-year survival rate of osteosarcoma is still < 70% [2,3]. And it has a high potential for distant metastasis, particularly lung metastasis, and the 5-year survival rate of OS patients with lung metastasis is less than 30% [4,5]. Therefore, in order to develop new potent therapeutic strategies, the underlying mechanism of OS metastasis must be elucidated.
During the initiation of metastasis process, epithelial-mesenchymal transition (EMT) plays a crucial role.
EMT is a reversible phenotypic change in which polar epithelial cells lose epithelial characteristics and acquire properties of mesenchymal cells, which may help explain tumorigenesis and metastasis of OS [6][7][8]. Therefore, inhibition of EMT progression may be a potentially effective method for the treatment of OS.
MicroRNAs (miRNAs), a kind of endogenous small noncoding RNAs with a length of 22-28 nucleotides, negatively regulate target genes by binding to the 3′-untranslated regions (3′-UTRs) of the target mRNAs and degrading or translational inhibiting them [9,10]. And accumulating evidence demonstrates that miRNAs act as the vital role of miRNA in the occurrence and development of various tumors by multiple signal pathways [11,12]. miRNAs which may regulate cell proliferation, EMT, and metastasis are also identi ed important in the development of OS [13][14][15]. As a tumor-suppressing miRNA, miR-766-3p, which was reported to frequently downregulated in several types of cancer, was remarkably correlated with poor clinical outcomes [16,17]. Nevertheless, to date, there have been no detailed investigations into the functions of miR-766-3p in OS, and the role of miR-766-3p is far from fully understood.
Expression levels of miR-766-3p assessed in hepatocellular carcinoma tissues and cell lines is lower compared to control samples, and was signi cantly correlated with tumor size, TNM stage, metastasis, and a poor prognosis of hepatocellular carcinoma by targeting Wnt family member 3A, metastasisassociated protein 3 and fos-related antigen 2 [18][19][20]. Furthermore, miR-766-3p functioned as a tumorsuppressor gene, which suppressed renal cancer cell-cycle progression by regulating the ampli cation of SF2 and additional downstream signaling pathways, and markedly correlated with a prognosis of renal cell carcinoma [21]. However, as we know, the exact value of miR-766-3p in osteosarcoma growth and metastasis is still unknown.
B-Cell Lymphoma 9-Like Protein (BCL9L), a second component of the vertebrate BCL9 family, is identi ed as cofactor of canonical Wnt signaling in mammalian cells and induces epithelial-mesenchymal transitions [22][23][24]. It has reported that BCL9L promotes early phases of intestinal tumor progression in humans by regulating the switch between the adhesive and transcriptional functions of β-catenin.
Moreover, BCL9L is demonstrated to enhance β-catenin-mediated transcription and increase the proliferation as well as the metastatic potential of breast and colon cancer cells. However, studies revealing the role of BCL9L in osteosarcoma are rarely reported, and the relationship between the miR-766-3p-BCL9L axis and the Wnt signaling pathway involved in OS still remains to be investigated in depth.
In our study, it was found that miR-766-3p was signi cantly decreased in in OS cell lines and tumors and promoted, EMT and. We demonstrated that the overexpression of miR-766-3p suppressed EMT, proliferation and metastasis in OS by downregulating the expression of BCL9L via the β-catenin/TCF-4 signaling pathway.

Tissue samples
This study was approved by the Ethics Committee of The First A liated Hospital of Nanjing Medical University, and all experiments were conducted in accordance with the approved guidelines and regulations, and all the subjects signed the written informed consent. 60 pairs of osteosarcomas and adjacent normal tissue were collected from patients undergoing biopsy before the chemotherapy at the orthopedics department. Biopsy samples were obtained and subsequently frozen in liquid nitrogen. The clinical and demographic information of all patients is shown in Table 1.
Quantitative Real-time quantitative polymerase chain reaction (qRT-PCR) Total RNA from tissues and cells with Trizol (Invitrogen, USA) was extracted from the pulverized samples stored at liquid nitrogen, and was resuspended in DEPC-treated H2O. The concentration and purity of total RNA were con rmed at 260 nm. Reverse transcription (RT) was undertaken using the Goldenstar TM RT6 cDNA Synthesis Kit (Beijing TsingKe Biotech Co. Ltd., China) according to the manufacturer's protocol. SYBR Green Master (TsingKe, Beijing, China) was for the quantitative PCR measure. The expression level of U6 or GAPDH served as the endogenous control. The primers for BCL9L, GAPDH, miR-624-5p, and U6 were obtained from TsingKe (Beijing, China). The sequences of the primers are shown in Supplementary table S1.

Migration assay
Cell migration assay was carried out using Transwell migration and Wound-healing assay. Transwell chambers (8-uM pore size; Costar, NY, USA) were used in the migration assay. In Brief, the Transwell coculture assay was performed using the 12-well Transwell plates (Corning, MA, USA). 24 h later, cells that had passed through the Transwell membrane were xed with paraformaldehyde, stained with crystal violet and subsequently photographed in three random microscopic views. And the cells were counted and photographed under an optical microscope (Nikon, Tokyo, Japan). For wound-healing assay, OS cells were seeded in six-well plates and grown until 80-90% con uence overnight, and were scratched by a sterile 200μL pipette tip. The wound closure was observed at 0 and 24 h, and imaging performed under a microscope.

Invasion assay
To assess the cell invasive ability, Transwell invasion assays were performed. For Matrigel invasion assays, cells were seeded on the upper surface of membranes coated with Matrigel matrix (Millipore, USA). Following incubation for 24 h, cells invading across the Transwell membrane were xed with 4% paraformaldehyde and stained with 0.4% crystal violet for 20 min. The stained cells were counted and photographed under a light microscope.
Cell counting Kit-8 assay (CCK-8) and colony formation assay The transfected cells were plated in 96-well plates (5×10 3 cells with 100 μL/well) and cultured for 24, 48, 72, 96, and 120 h. Cell Proliferative rate was examined by the Cell Counting Kit-8 (CCK-8, Dojindo, Japan) following the manufacturer's instructions. The colony formation experiment was conducted by staining cell lines with crystal violet and counting the number of effective clones after culturing for 14 days.

Immuno uorescence analysis
Transfected cells were xed with 4% paraformaldehyde for 10 min, and permeabilized with 0.3% Triton X-100 for 15 min. Then, the cells were immuno uorescence stained with the primary antibodies to β-catenin (Abcam, MA, USA), uorescein-conjugated secondary antibody, and then with DAPI. Images were collected by uorescence microscope (Carl Zeiss, Germany).

Western blotting
After protein extraction, and the protein concentration was determined using BCA Protein Quanti cation Kit (Thermo, USA). Proteins were then subjected to 10% SDS-PAGE and transferred to polyvinylidene di uoride (PVDF) membranes (Bio-Rad, Hercules, CA, USA). Next, the membranes were blocked in 5% skim milk and incubated with primary antibodies (1:1000) at 4°C overnight. We used a panel of antibodies to detect these proteins, including Rabbit anti-BCL9L, GAPDH, N-cadherin, E-cadherin, vimentin, β-catenin, TCF-1, Cyclin D1 and Axin2 antibodies (Abcam, Cambridge, UK). And the membrane was washed with PBST followed by incubation with the secondary antibody (1:10000) for 1 h at room temperature. Reacting bands were achieved by ECL reagent (Shanghai Tianneng Technology Co., Ltd., Shanghai, China), and protein bands was semi-quanti ed using ImageJ.

Immunohistochemical staining
All tissue specimens were xed in 4% paraformaldehyde overnight, embedded in para n, cut into 4-μmthick sections and incubated with the primary antibody for BCL9L (Abcam, Cambridge, UK) overnight at 4°C . Next, the sections were incubated with the secondary antibody for 1 h and stained using the developed diaminobenzidine (DAB) for 3 min. The staining results were measured by combining the percentage of positive staining and intensity of positively stained tumor cells.

Xenograft transplantation experiments
Animal studies were approved by the Animal Ethics Committee of Nanjing Medical University. Female BALB/c nude mice used for tumor growth assays were purchased from the Animal Model Institute of Nanjing University (Nanjing, China). The nude mice were randomly assigned into four groups, ve in each group. OS cells (1 × 10 6 cells), labeled with re y luciferase, were subcutaneously injected into the nude mice. Tumor growth was observed every 3 days. Tumor size was expressed as tumor volume and calculated by the formula: V = A×B 2 /2(mm 3 ), where "A" was the larger diameter and "B" was the smaller diameter. On day 35 post injection, the progression of xenograft growth was imaged with the IVIS200 imaging system (Caliper Life Sciences, Waltham, MA, USA).

Statistical analysis
All experiments were repeated at least three times, with average values expressed as means ± standard deviation (SD). The association of miR-766-3p expression with clinicopathological features was analyzed by the χ2 test. To compare signi cant differences between the two groups, comparisons were made by independent Student's t-test. Statistical analysis of the differences in mRNA expression levels of miR-766-3p and BCL9L in paired tissues were analyzed by paired t-test. One-way or two-way ANOVA with Bonferroni post hoc test was tested for multivariate analysis. Pearson's correlation analysis used for bivariate correlation. Statistical analyses were performed by SPSS, v. 20.0 (SPSS Inc., Chicago, IL, USA). P< 0.05 was considered as statistically signi cant.

miR-766-3p expression was downregulated in OS cell lines and clinical tissues.
In order to investigate the expression level of miRNAs in OS tissues, we analyzed the data from the GEO database (GSE65071) using the R package limma. The volcano plot demonstrates the miRNA expression differences in normal and OS tissues (Fig. 1A). Totally, 70 miRNAs were downregulated (fold change > 2.0, FDR < 0.05) in OS tissues and the top 8 most signi cantly differentially expressed miRNAs are listed in Fig. 1B. According to the expression levels of miRNA, the up-and downregulated miRNAs was represented in a cluster heap map (Fig. 1C). We found the most down-regulated miRNA was miR-766-3p in these miRNAs. Thus, we tried to understand the molecular mechanism of the inhibition effect of miR-766-3p. As shown in Fig. 1D, miR-766-3p was signi cantly downregulated in OS tissues, compared with that of normal tissue. Then, real-time quantitative PCR was performed to demonstrate the miR-766-3p expression level in 60 paired OS tissues and adjacent normal tissues. Compared with peritumor tissues, miR-766-3p was signi cantly decreased in OS tissues (P< 0.001, Fig. 1E). In addition, it was proven that miR-766-3p was lowly expressed in patients with metastasis (P< 0.001, Fig. 1F).And expressing levels of miR-766-3p prominently decreased in OS cell lines, including Saos-2, HOS, 143B, U2OS, and MG-63, compared with in hFOB 1.19 cells (P< 0.001, Fig. 1G). And as shown in representative images of OS patients with or without lung metastasis ( Supplementary Fig. S1), the red arrows indicated tumor foci or pulmonary metastatic nodules. In order to identify the clinical signi cance of miR-766-3p in OS, median expression level of miR-766-3p was de ned as a cutoff value and the patients were divided into subgroups. In Table 1, the expression level of miR-766-3p was signi cantly negatively correlated with TNM stage, tumor size, and lung metastasis.
Among these OS cell lines, 143B and U2OS cells were used to study further in vitro experiments. The transfection e ciency of miR-766-3p was veri ed lentiviruses by qRT-PCR. The results revealed that miR-766-3p was signi cantly overexpressed in the mimics group and constrained in the inhibitor group (U2OS, P< 0.001; 143B, P< 0.001; Fig. 2A). Western blot analysis demonstrated that miR-766-3p sh#1 and miR-766-3p sh#2 diminished E-cadherin levels and raised the metastasis related protein levels of N-cadherin and vimentin in U2OS and 143B cells (Fig. 2B). To explicate the effects of miR-766-3p on OS cell migration and invasion ability in vitro, we conducted migration and invasion assays. The wound-healing assay and Transwell migration assay showed that down-regulated expression of miR-766-3p could promote the migration of U2OS and 143B cells (U2OS, P< 0.001; 143B, P< 0.001; Fig. 2C-F). The Transwell invasion assay indicated that the knockdown of miR-766-3p signi cantly increased the number of invasive OS cells per eld (U2OS, P< 0.001; 143B, P< 0.001; Fig. 2G, H). Besides, CCK-8, colony formation and EdU assays were detected the effect of miR-766-3p promotion on OS cell proliferation. As can be seen from Additional le 1: Figs. S2A, B, downregulating miR-766-3p revealed no signi cant difference in the rst three days, but signi cantly promoted cell proliferation after ve days, showing that miR-766-3p has the potential as inhibitory factor in OS cell proliferation. The results of colony formation and Edu assays were consistent with the above result (Additional le 1: Figs. S2C-F).

Overexpression of miR-766-3p inhibited OS cell migration and invasion in vitro.
Western blot analysis revealed that E-cadherin levels was raised in the miR-766-3p mimic group, and the metastasis related protein levels of N-cadherin and vimentin was downregulated in U2OS and 143B cells (Fig. 3A).Transwell migration assays were performed and the results showed that miR-766-3p overexpression remarkably promoted the migration abilities of U2OS and 143B cells (U2OS, P< 0.01; 143B, P< 0.01; Fig. 3B, C), and the wound-healing assay were similar to those of the Transwell migration assay results (U2OS, P< 0.001; 143B, P< 0.001; Fig. 2D, E). Then, Transwell invasion assays were conducted to assess the effects of miR-766-3p on invasion. The results were consistent with those of the Transwell migration and wound-healing assays (U2OS, P< 0.001; 143B, P< 0.001; Fig. 2F, G). Moreover, the CCK-8, EdU and colony formation assays revealed that overexpression of miR-766-3p markedly decreased the proliferation of U2OS and 143B cells (Additional le 1: Figs. S2). Taken together, these data suggest that miR-766-3p mediates OS cells invasion, migration and proliferation processes.
BCL9L expression was upregulated in OS cell lines and tissues and was a target of miR-766-3p.
To further detect the underlying mechanism of miR-766-3p in the invasion, migration and proliferation of OS cells, Potential targets of miR-766-3p were predicted using TargetScan, miRDB and miRTarBase Tools (Additional le 1: Figs. S3). Among the candidate genes, we were speci cally interested in BCL9L because it played essential tumor-promoting role in modulation of carcinogenesis and cancer development. The qRT-PCR and Western blotting were used to explore BCL9L expression in 60 paired OS tissues and adjacent tissues. PTPRB expression was signi cantly more in tumor tissues than the adjacent tissues (P< 0.05, Fig. 4A, B). These results were con rmed by immunohistochemistry assays (Fig. 4C). Further, we found that the expression level of BCL9L was negatively correlated with miR-766-3p in OS tissues with an R2 of 0.4751 (Fig. 4D). As demonstrated in Fig. 4E, Kaplan-Meier analysis revealed that patients with high expression of BCL9L had signi cantly poor survival than those with low expression (P= 0.0031). Furthermore, the mRNA of BCL9L was observed to be upregulated in OS cell lines, particularly in U2OS and 143B cells (P < 0.001, Fig. 4F). And western blotting results showed that U2OS and 143B cells contained the most amount of BCL9L protein compared to the other cell lines (P < 0.001, Fig. 4G). Besides, the median values of miR-766-3p and BCL9L mRNA levels were used as cutoff points to divide patients into "high" and "low" subgroups. Table 1 indicated that the expression level of BCL9L was positively related to tumor size, TNM stage, and lung metastasis. Next, we carried out-luciferase reporter assay analysis to con rm that miR-766-3p could directly targeted BCL9L. It revealed that miR-766-3p overexpression could signi cantly reduce the activity of luciferase in OS cells (Fig. 4H). As shown in Fig.  4I, quantitative RT-PCR analysis showed low mRNA expression of BCL9L in cells transfected with miR-766-3p mimics; in contrast, a markedly high BCL9L expression was investigated in cells with miR-766-3p inhibitor. What is more, Western blotting assays determined that miR-766-3p negatively regulated the expression level of BCL9L (Fig. 4J). In summary, these results showed that BCL9Lwas a direct target of miR-766-3p.

Upregulating BCL9L abolished the effects of miR-766-3p mimics on OS migration and invasion.
In order to further con rm that miR-766-3p mediates OS cell migration and invasion by targeting BCL9L, we did a series of rescue experiments. First, OS cells were transfected with miR-766-3p mimics to realize miR-766-3p upregulation. Then, Western blotting revealed that E-cadherin expressions in 143B and U2OS cells were remarkably decreased by the overexpression of miR-766-3p, while BCL9L and the metastasisrelated proteins (N-cadherin and vimentin) did a signi cant increase. Surprisingly, these in uence of miR-766-3p mimics were evidently reversed by overexpressed BCL9L (Fig. 5A). Transwell migration assays were conducted and the results showed that BCL9L inhibited the protective in uences of OS cell invasion resulting from miR-766-3p mimics (U2OS, P< 0.001; 143B, P< 0.001; Fig. 5B); the inhibitory effects of cell migration resulting from miR-766-3p mimics were reduced by BCL9L upregulation in the wound-healing assays (U2OS, P< 0.001; 143B, P< 0.001; Fig. 5C). As shown in Fig. 5D, the above results were also veri ed in Transwell invasion assays. Taken together, BCL9L was con rmed to be helpful in promoting OS cell migration and invasion caused by miR-766-3p miR-766-3p regulated the β-catenin/TCF-1 signal pathway through BCL9L.
To demonstrate the underlying mechanism of how the miR-766-3p /BCL9L axis modulates OS cells migration and invasion, we performed Western blotting and immuno uorescence analysis. It has been reported that the Wnt/β-catenin signaling pathway is involved in cancer development. And previous studies revealed that it was a critical step in the tumorigenesis of various types of tumors, including the muscular and skeletal systems. So, whether miR-766-3p and BCL9L in uenced OS invasion and migration through the Wnt/β-catenin pathway was the focus of our study. Western blotting assay indicated that miR-766-3p mimics in 143B and U2OS cells reduced BCL9L expression, as well as downregulated the levels of β-catenin, TCF-4, Cyclin D1 and Axin2, but the inhibiting effects were all remedied by upregulating BCL9L (Fig. 6A). What is more, the expression of β-catenin protein contained in the nuclei were negatively related with the miR-766-3p level and had a positive correlation with the expression level of BCL9L (Fig. 6B). Immuno uorescence analysis provided vigorous evidence that miR-766-3p mimics enhanced β-catenin import into the nuclei of 143B and U2OS cells and overexpressing BCL9L could abolish this effect (U2OS, P< 0.001; 143B, P< 0.001; Fig. 6C). Overall, these results indicate that the miR-766-3p/BCL9L axis regulates OS cell migration, invasion, and proliferation via the βcatenin/TCF-4 signal pathway.
In order to detect the effects of miR-766-3p on tumor growth in vivo, 143B cells stably upregulated or U2OS cells with suppressed miR-766-3p were subcutaneously injected into nude mice. And we used cells transfected with miR-NC or sh-miR-NC to treat nude mice as negative controls. The xenograft tumor sizes were monitored every 2 days from 14 days after injection, and 4 weeks later the mice were euthanized. Fig. 7A, B showed that knockdown of miR-766-3p accelerated the growth of OS cells (P < 0.01), and the tumor volume and weight were larger and heavier respectively (P < 0.01, Fig. 7C; P < 0.01, Fig. 7D). On the contrary, high expression of miR-766-3p remarkably suppressed tumor growth compared to controls (P < 0.01, Fig. 7E, F), and the tumor was smaller and lighter (P < 0. 01, Fig. 7G; P < 0.01, Fig. 7H). Then, to examine the expression level of BCL9L in xenografts, immunohistochemistry revealed that the expression level of BCL9L was enhanced by miR-766-3p inhibitor; in contrast, miR-766-3p mimics decreased BCL9L expression (Fig. 7I, J). Furthermore, the variation expression of BCL9L in xenografts was con rmed by Western blotting (P < 0.01, Fig. 7K; P < 0.01, Fig. 7L). Taken together, miR-766-3p plays a signi cant role in promoting proliferation of OS cells in vivo. As shown in the mechanism diagram, miR-766-3p and BCL9L played a crucial role in the occurrence and development of OS; additionally, miR-766-3p suppressed OS cell proliferation, EMT and metastasis via β-catenin/TCF-4 signal pathway, and directly targeted BCL9L (Fig. 8).

Discussion
As mentioned in the previous studies, osteosarcoma occurring predominantly in children and adolescents is the most frequent primary malignant bone sarcoma [1,3,4,25]. Although previous studies have revealed that miRNAs have a signi cant in uence on modulating cellular functions and biological processes of various tumors [9,12,26], Unfortunately, little is known about the mechanism underlying the pathogenesis of OS. Therefore, there is an urgent requirement to understand the mechanism of development and metastasis of OS and to identify novel therapies [5,27,28]. In this study, a series of experiments were conducted and the results indicated that miR-766-3p was downregulated in OS tissues and cell lines. Moreover, miR-766-3p was negatively correlated with malignancy of OS, and could be used as a novel target for OS treatment.
Emerging research has demonstrated that the abnormal miRNA expression plays key regulatory role in OS progression [29,30], the underlying mechanisms of miR-766-3p in OS have not yet been detected. MiR-766-3p has recently been recognized as a tumor suppressor by inhibiting the Wnt pathway in several tumors. MiR-766-3p was lower in hepatocellular carcinoma and renal cell carcinoma specimens or cell lines than in normal group. Similarly, our ndings via real-time quantitative PCR are consistent with those from previous studies [16,19,21]. In the current study, we rst found that miR-766-3p was downregulated in clinical OS samples and cell lines, indicating that this RNAs may function as a tumor suppressor. And we performed a series of experiments in vitro, showing that miR-766-3p inhibited aggressive and invasion of OS cells.
EMT has long been linked to cancer malignant progression, tumor migration and metastasis [31,32]. During this process, epithelial cancer cells acquire a mesenchymal phenotype, which was characterized by the loss of cell-cell adherent junctions, loss of cell polarity and actin cytoskeleton remodeling. And these central changes enhance migration, invasion of cancer cells and confer resistance to therapy [33][34][35]. In addition, other characteristics of the EMT phenotype is downregulation of epithelial markers (Ecadherin) and upregulation of mesenchymal markers (N-cadherin and Vimentin) [6,7]. In our study, overexpression of E-cadherin in cells was observed by an upregulated miR-766-3p level; however, when miR-766-3p was overexpressed, the levels of Vimentin and N-cadherin decreased. Consistent with our in vitro results, miR-766-3p suppressed the ability of OS cell metastasis, indicating that miR-766-3p acted as a risk factor in OS progression. Moreover, we demonstrated that BCL9L is one target genes of miR-766-3p. We rst con rmed that miR-766-3p abolished BCL9L expression in OS cells both at the mRNA and protein levels, and miR-766-3p was negatively correlated with BCL9L expression in OS tissues. Then, the luciferase reporter assay showed that the luciferase activity of WT-BCL9L-3′-UTR but not MUT-BCL9L-3′-UTR was inhibited by miR-766-3p. Furthermore, BCL9L was downregulated in OS cells and recovery of its expression could reverse the effects of miR-624-5p.
BCL9L, a type of protein like B-Cell CLL/Lymphoma 9 (BCL9), is a second component of the vertebrate BCL9 family, and is usually regarded as cofactor of canonical Wnt signaling in mammalian cells and induces epithelial-mesenchymal transitions [36,37]. BCL9L was shown to promote tumor progression of breast cancer [38], choriocarcinoma [39], pancreatic cancer [40], and colon cancer [22,41,42]. It was reported that higher expression of BCL9L predict lower survival rates in intestinal tumor patients and could be employed as an independent prognostic biomarker. Additionally, BCL9L was aberrantly elevated in 43% of colorectal tumors, and was required for enhanced β-catenin-TCF-mediated transcription in colorectal tumor cells. Here, our study ndings validate that BCL9L is signi cantly upregulated among OS tissues, relative to the corresponding noncancerous tissues, indicating that BCL9L plays a promotable role in tumorigenesis and progression of OS.
Wnt signaling pathways regulate various processes that are essential for cancer progression, including tumor initiation, tumor growth, cell senescence, cell death, differentiation and metastasis [43,44]. And βcatenin-dependent Wnt signaling pathways have crucial roles in the regulation of cancer cell above behaviors [45,46]. Moreover, β-catenin functions as a transcriptional switch, which reduces the association of TLE with TCF/LEF , while recruiting various transcriptional cofactors including BCL9/BCL9L, Pygopus and histone acetyltransferases [36,37,47]. Considering that we further examined whether miR-766-3p inhibits the Wnt pathway via BCL9L. Overexpression of miR-766-3p decreased the levels of β-catenin, TCF, Cyclin D1 and Axin2, together with the downregulated BCL9L by Western blotting. On the contrary, downregulating miR-766-3p escalated levels of the above proteins. In addition, the application of siBCL9L or BCL9L markedly remedied the effects of miR-766-3p inhibitors or mimics on the Wnt signaling pathway. Immuno uorescence further showed that more β-catenin proteins were transported to the nuclei of OS cells while upregulating miR-766-3p or siBCL9L could abolish this effect.

Conclusions
In summary, we demonstrated the tumor-inhibiting role of miR-766-3p in osteosarcoma development and its underlying mechanism via BCL9L and the Wnt pathway. Our data reported that miR-766-3p had a suppressing role in OS progression. We demonstrated that miR-766-3p inhibited tumor progression both in vitro and in vivo. We further validated that miR-766-3p inhibited cell proliferation, migration, and EMT in OS cells through Wnt signaling activity by directly regulating BCL9L. The present study may provide further insight into the development of osteosarcoma. Therefore, it is a reasonable inference that miR-766-3p/BCL9L signaling may provide valuable information for OS targeted therapy, which may be helpful to control the OS proliferation and metastasis, and even improve survival rate for the patients in the future. Committee and signed informed consent was obtained from each patient.

Consent for publication
All the patients that involved in the study have given their consent to publish their individual data.

Competing interests
The authors declare no con ict of interest. Authors' contributions SZ performed the experiments, analyzed the data, and wrote the manuscript. HC, WL and LF conceived the study and revised the manuscript. ZQ, RK and QZ performed the experiments and analyzed the data. JL and XC conceived, designed, and supervised the research. All authors read and approved the nal manuscript. All authors declare that they have no con ict of interest. All authors approved the nal version of the manuscript.     PCR indicated that the BCL9L mRNA level was negatively regulated by miR-766-3p (n = 4). J. Western blotting supported that miR-766-3p negatively controls the expression level of BCL9L (n = 3). Data are presented as the means ± SD. * P < 0.01.

Figure 5
Upregulating BCL9L restored the effects of miR-766-3p mimics on OS cells. A. Western blotting showed that OS cells invasion and migration was remedied by overexpressed BCL9L (n = 3). B-C. BCL9L could reverse the augmented OS cell invasion and migration caused by miR-766-3p mimics (n = 4). D. These above results were con rmed by cell Transwell invasion assays (n = 4). Data are presented as the means ± SD. * P < 0.01.  immunohistochemistry assays showed that BCL9L expression was reduced in the miR-766-3p mimics group and enhanced BCL9L expression was observed in the miR-766-3p inhibitor group. K-L. Western blotting results con rmed that BCL9L expression was reduced in the miR-766-3p mimics group and enhanced in the miR-766-3p inhibitor group (n = 3). Data are presented as the means ± SD. * P < 0.01.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.