LncRNA EBLN3P promotes the progression of osteosarcoma through modifying the miR-224-5p/Rab10 signaling axis

The treatment of patients with advanced-stage osteosarcoma represents a major challenge, with very few treatments currently approved. Although accumulating evidence has demonstrated the importance of lncRNAs in osteosarcoma, the current knowledge on the functional roles and molecular mechanisms of lncRNA endogenous born avirus-like nucleoprotein (EBLN3P) is limited. At present, the expressions of EBLN3P and miR-224-5p in osteosarcoma tissues were quantified by reverse transcription-quantitative PCR assay, and the expression of Ras-related protein 10 (Rab10) in osteosarcoma tissues was quantified by immunohistochemistry and western-blotting. The bioinformatics prediction software ENCORI was used to predict the putative binding sites of EBLN3P, Rab10 and miR-224-5p. The regulatory role of EBLN3P or miR-224-5p on cell proliferation, migration and invasion ability were verified by Cell Counting Kit-8, wound healing and Transwell assays, respectively. The interaction among EBLN3P, miR-224-5p and Rab10 were testified by luciferase. The increased expression of EBLN3P and Rab10 and decreased expression of miR-224-5p were observed in osteosarcoma tissues and cell lines. Besides, the overexpression of EBLN3P or knockdown of miR-224-5p were revealed to promote the proliferation, migration and invasion of osteosarcoma cells. Bioinformatics analysis and luciferase assay revealed that EBLN3P could directly interacted with miR-224-5p to attenuate miR-224-5p binding to the Rab10 3′-untranslated region. Furthermore, the mechanistic investigations revealed activation of the miR-224-5p/Rab10 regulatory loop by knockdown of miR‐372-3p or overexpression of Rab10, thereby confirming the in vitro role of EBLN3P in promoting osteosarcoma cell proliferation, migration and invasion. To the best of our knowledge, the present study is the first to demonstrate that EBLN3P may act as a competitive endogenous RNA to modulate Rab10 expression by competitive sponging to miR-224-5p, leading to the regulation of osteosarcoma progression, which indicates a possible new approach to osteosarcoma diagnosis and treatment.

Where gels/blots are used in figures were compliance with the digital image and integrity policies (www.natur e.com/srep/polic ies/index .html#digit al-image ).
Cell culture. The human fetal osteoblast cell line hFOB.1.19 and the osteosarcoma cell lines Saos2, MG63, 143B and U2OS were obtained from American Type Culture Collection and cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS; Gibco BRL, Grand Island, NY) in a humidified incubator containing 5% CO 2 . The cells were passaged once every 2-3 days, and cells in the logarithmic growth phase were used for the subsequent experiments.
Short hairpin RNA (shRNA) method. The shRNA-mediated knockdown was performed as previously described 29 . The 293 T cells were cultured to a 70-80% confluence in 6-well dishes before transfected with the aforementioned constructs using Lipofectamine 3000 (cat. no. 11668027, Thermo Fisher Scientific, Inc.  EBLN3P  5′-CAG ACT AAA GGA TCA AGC GAGA-3ʹ 5′-ATC AAT TGC CAC AGG TTG AAGA-3ʹ   miR-224-5p  5′-GCC CCG ACA GTC TAG ATA TGA -3ʹ  5′-GGA TGC TGC TGC TAG AGG T -3ʹ   Rab10  5′-GGA TAC CTA CGG AGC ACG AG-3ʹ  5′-AGC CAT CAC ACT TCT CCA GG-3ʹ   U6  5′-CTC GCT TCG GCA GCACA-3ʹ  5′-AAC GCT TCA CGA ATT TGC GT-3ʹ   GAPDH  5′-CCA GGT GGT CTC CTC TGA -3ʹ  5′-GCT GTA GCC AAA TCG TTG T- www.nature.com/scientificreports/ using a 10 μl pipette tip, and the culture medium was changed to remove the detached cells. After 24 h, the cells were visualized by light microscopy. The procedure was carried out as described previously 32 . Transwell assay. To determine cell invasion ability, the cells (8 × 10 2 cells) were resuspended in serum-free medium and seeded onto the upper chamber of a Matrigel-coated Transwell insert (EMD Millipore). Complete medium supplemented with 10% FBS was added to the lower chamber. After 24 h, the upper surface of the membrane was wiped with a cotton swab, and the cells attached to the lower surface were fixed with 4% formaldehyde for 10 min at room temperature and stained with 1% DAPI (4′, 6-diamidino-2-phenylin-dole) solution for 10 min. The invaded cells were observed and counted under a light microscope. The procedure was carried out as described previously 32 .
Dual-luciferase reporter assays. The procedure was carried out as described previously 33 . Wild-type (WT) and mutant (MUT) reporter plasmids of EBLN3P (EBLN3P-WT-luc and EBLN3P-MUT-luc), containing WT or MUT miR-224-5p mimics or mimics NC-binding sites were synthesized by GenePharma. The synthesized reporter plasmids were co-transfected with miR-224-5p mimics or mimics NC, respectively, by Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) when the cells reached 70% confluence. The luciferase activity was analyzed via the dual-luciferase reporter assay system.
Bioinformatics ENCORI software. The bioinformatics ENCORI software 34 (http://starb ase.sysu.edu. cn/) has been used to systematically identify miRNA-mRNA and miRNA-lncRNA interaction networks. In this work, the ENCOR was used to predict the relationship and the binding sites among EBLN3P, miR-224-5p and Rab10.
Statistical analysis. All the experimental results are expressed as the mean ± standard deviation, and each experiment was performed in triplicate. Statistical analyses and graphical depictions were performed using GraphPad Prism 5.0 (https ://www.graph pad.com, GraphPad Software, Inc.). Student's t-test or one-way analyses of variance (ANOVAs) were employed to evaluate the significance of the differences, as appropriate. The association between the survival rate of osteosarcoma patients and Rab10, miR-224-5p or EBLN3P expression levels was investigated using Kaplan-Meier survival curves and the log-rank test. Statistical significance was set at *P < 0.05 or **P < 0.01.

EBLN3P and Rab10 are upregulated, while miR-224-5p is downregulated in osteosarcoma tissues and cell lines.
To investigate the regulatory role of EBLN3P in osteosarcoma, it was first examined whether EBLN3P and miR-224-5p were dysregulated in osteosarcoma. To identify the potential genes targeted by miR-224-5p in osteosarcoma cells, the bioinformatics ENCORI software was used and Rab10 was identified as one of the target genes of miR-224-5p. The results of RT-qPCR analysis demonstrated that the RNA expression level of EBLN3P was expressed at higher levels in four osteosarcoma cell lines (Saos2, 143B, MG63 and U2OS) compared with those in the normal human fetal osteoblast cell line, while the RNA expression level of miR-224-5p exhibited the opposite trend (normalized to GAPDH, Fig. 1A). The results of western blot analysis demonstrated that the protein expression level of Rab10 were expressed at higher levels in four osteosarcoma cell lines compared with those in the normal human fetal osteoblast cell line (normalized to β-actin expression, Fig. 1B and Supplementary file). In addition, the expression levels of EBLN3P, miR-224-5p and Rab10 were accessed in osteosarcoma tissues and non-neoplastic bone tissues. As expected, the RNA expression of EBLN3P was higher and that of miR-224-5p was lower in osteosarcoma tissues compared with non-neoplastic bone tissues (normalized to GAPDH, Fig. 1C). Besides, the protein expression of Rab10 was higher in osteosarcoma tissues compared with non-neoplastic bone tissues (normalized to β-actin expression, Fig. 1D). Moreover, the associations between the expression of EBLN3P and miR-224-5p and the prognosis of osteosarcoma patients were investigated using Kaplan-Meier survival curves and the log-rank test. As shown in Fig. 1E,F,H, osteosarcoma patients with positive miR-224-5p expression exhibited longer overall survival compared with patients with negative miR-224-5p expression. In addition, osteosarcoma patients with positive EBLN3P or Rab10 expression had a shorter overall survival compared with that of patients with negative EBLN3P or Rab10 expression. Furthermore, the expression of Rab10 in osteosarcoma tissues was investigated via IHC analysis. As shown in Fig. 1G and Table 2, the expression of Rab10 was higher in osteosarcoma tissues (21/29) compared with that in non-neoplastic bone tissues (5/27), and the expression of Rab10 was found to be associated with pulmonary metastasis (P < 0.001) and TNM stage (AJCC) (P < 0.001). These data indicate that EBLN3P, miR-224-5p and Rab10 may play important roles in regulating the development of osteosarcoma.

EBLN3P knockdown suppresses the proliferation, migration and invasion of osteosarcoma cells in vitro.
As mentioned above, the expressions of EBLN3P and miR-224-5p were negatively correlated in osteosarcoma cells and tissues (Fig. 1), suggesting the inhibitory effect of EBLN3P on miR-224-5p. Bioinformatics confirmed that EBLN3P directly interacted with miR-224-5p ( Fig. 2A). Dual-luciferase reporter assays were performed to verify our hypothesis. Overexpression of miR-224-5p was achieved by transfecting 143B cells with miR-224-5p mimics. The luciferase assay results confirmed that transfection with miR-224-5p mimics significantly weakened the luciferase signal of reporters containing EBLN3P-WT, but had no effect on the activity of reporters containing EBLN3P-MUT (Fig. 2B) www.nature.com/scientificreports/ the aggressive behavior of osteosarcoma cells. The negative control shRNA (si-NC) and the blank vector plasmid (pcDNA-3.1) were used as knockdown and overexpression controls, respectively. Moreover, the expression level of EBLN3P and miR-224-5p were evaluated via RT-qPCR (Fig. 2C). The RT-qPCR results indicated a negative regulatory association between EBLN3P and miR-224-5p (Fig. 2C). Taken together, these results demonstrated that EBLN3P may serve as a ceRNA for miR-224-5p to inhibit miR-224-5p expression. The effects of Rab10 on osteosarcoma cells and its interaction with miR-224-5p have yet to be elucidated. Bioinformatics analysis and luciferase assay confirmed that miR-224-5p can directly bind to the 3′-untranslated region (UTR) of Rab10 (Fig. 2D). Next, a dual-luciferase reporter gene assay demonstrated that the luciferase signal of cells co-transfected with miR-224-5p mimics and WT Rab10 vector was notably decreased compared with that of cells co-transfected with miR-224-5p mimics and MUT Rab10 vector (Fig. 2E). This indicated that miR-224-5p likely binds to the 3′-UTR of Rab10. Further RT-qPCR and western blot analysis validated that the overexpression of miR-224-5p significantly reduced the expression of Rab10, whereas knockdown of miR-224-5p significantly enhanced the expression of Rab10 at both the mRNA and protein levels in 143B osteosarcoma cells (Fig. 2F,G and Supplementary file). Taken together, these results indicate that miR-224-5p acts as the upstream regulator to adjust Rab10 expression. www.nature.com/scientificreports/ mimics group and a significantly increased proliferation rate in the miR-224-5p inhibitor group, compared with the control group (Fig. 3A). The migration and invasion assays further verified this trend. The migration and invasion abilities of osteosarcoma cells were suppressed following miR-224-5p mimics transfection, whereas they were enhanced following inhibitor transfection (Fig. 3B-D). Collectively, these data indicate that miR-224-5p acts as a tumor suppressor in osteosarcoma cells, and it may suppress the malignant behaviors of 143B cells, including cell proliferation, migration and invasion.

miR-224-5p inhibits the proliferation, migration and invasion of osteosarcoma cells via down
The knockdown of EBLN3P markedly suppressed the proliferation, invasion and migration of osteosarcoma cells. The 143B and U2OS cells were transfected with EBLN3P targeted shRNA (si-EBLN3P), negative control shRNA (si-NC), pcDNA3.1-EBLN3P overexpression vector (pcDNA-EBLN3P) and blank vector plasmid (pcDNA-3.1). Moreover, the influence of the EBLN3P overexpression or knockdown on osteosarcoma 143B and U2OS cells were also explored. Our data revealed that the knockdown of EBLN3P markedly suppressed the proliferation, invasion and migration of osteosarcoma 143B and U2OS cells, which were enhanced by EBLN3P overexpression, as evidenced by CCK-8, Transwell and wound healing assays, respectively ( Fig. 4A-D). Therefore, these results suggested that EBLN3P has an oncogenic potential and induces osteosarcoma cell proliferation, migration and invasion.

EBLN3P regulates osteosarcoma cells via the miR-224-5p/Rab10 pathway. Rescue experiments
were performed to assess the effects of the EBLN3P-miR-224-5p-Rab10 pathway on 143B cell activity. Overexpression of Rab10 was achieved via transfecting the cells with pcDNA3.1-Rab10 or miR-224-5p inhibitor (Fig. 5A,B and Supplementary file). The results demonstrated that knockdown of EBLN3P markedly reduced cell proliferation, migration and invasion. However, co-transfection with si-EBLN3P and miR-224-5p inhibitor or Rab10 significantly increased cell proliferation, migration and invasion compared with cells transfected with si-EBLN3P alone (Fig. 5C-F). Accordingly, these data suggested that the regulatory effects of EBLN3P were mediated through the miR-224-5p/Rab10 axis to promote the progression of osteosarcoma cells.

Discussion
LncRNAs were initially identified in carcinogenesis due to their differential expression compared with normal tissues 35 . To date, a growing body of evidence indicates that the abilities of lncRNAs regulate complex cellular behaviors, such as cell growth and metastasis, are commonly deregulated in cancer, including osteosarcoma 36,37 . Although numerous of potential biomarkers have been reported, a specific diagnostic biomarker for osteosarcoma has not yet been confirmed 38 . EBLN3P is a novel lncRNA located on chromosome 9: 37,079,935-37,086,874 forward strand 39 . The present study demonstrated that EBLN3P was markedly upregulated in osteosarcoma tissues and cell lines. In vitro, functional assays indicated that the knockdown of EBLN3P suppressed osteosarcoma www.nature.com/scientificreports/ cell proliferation, migration and invasion, demonstrating the potential of EBLN3P as a therapeutic target for osteosarcoma; therefore, we next investigated the underlying mechanism in osteosarcoma cell lines. The ceRNA theory was first proposed in 2011 and has since been widely accepted in the field of non-coding RNA research 40 . LncRNAs may serve as ceRNAs by sponging to miRNAs and inhibiting the downstream target gene 41 . Bioinformatics analyses revealed that there was a conserved binding site of miR-224-5p on EBLN3P. Therefore, it was inferred that EBLN3P could affect miR-224-5p via a ceRNA mechanism. It was validated that miR-224-5p exerted a reciprocal suppressive effect with EBLN3P expression, and knockdown of miR-224-5p induced the proliferation, migration and invasion of osteosarcoma cells in vitro. Importantly, the dual-luciferase www.nature.com/scientificreports/ assay further confirmed that EBLN3P directly interacted with miR-224-5p to reduce its expression, suggesting that EBLN3P serves as the miRNA sponge that binds to and regulates miR-224-5p expression. The miRNAs control gene expression by binding to the 3′-UTR of the target gene, which causes mRNA cleavage or translational repression 42 . According to the results of previous and ongoing research, it was hypothesized that EBLN3P can affect the malignant phenotype of osteosarcoma cells by upregulating the expression of the Rab10 protein. To the best of our knowledge, this study is the first to explore the effect of EBLN3P on the malignant phenotype of osteosarcoma cells and its molecular mechanism of action, so as to provide a theoretical basis and data supporting EBLN3P as a molecular target for early diagnosis and metastasis control of osteosarcoma.
The present study provided evidence that Rab10 was overexpressed in osteosarcoma, and that its positive expression is associated with distant metastasis and a higher TNM stage. Osteosarcoma patients with positive Rab10 expression exhibited a shorter overall survival compared with patients with negative Rab10 expression. The present study demonstrated that Rab10 could promote the proliferation of osteosarcoma cells in vitro. It was also observed that the expression of Rab10 in distant metastases was higher compared with that in primary osteosarcoma tissues, and that Rab10 could increase the migration and invasion abilities of osteosarcoma cells in vitro and in vivo. In the present study, it was also observed that the expression of Rab10 was regulated via the EBLN3P/miR-224-5p axis through bioinformatics analyses; however, whether this compensatory mechanism  www.nature.com/scientificreports/ exists in osteosarcoma requires further investigation. The dual-luciferase assays confirmed the direct interaction between miR-224-5p and Rab10. Further rescue experiments demonstrated that EBLN3P promoted the proliferation and metastasis of osteosarcoma cells via regulation of the miR-224-5p/Rab10 signaling axis. Although we initially uncovered a novel downstream regulatory mechanism in osteosarcoma cells mediated by EBLN3P in vitro, further research is required to fully elucidate this complex mechanism. Future studies using mouse models will be needed to further verify this novel molecular mechanism and prove its therapeutic potential.

Conclusion
In conclusion, to the best of our knowledge, the present study is the first to demonstrate that EBLN3P act as a novel oncogene in osteosarcoma. Furthermore, EBLN3P act as a ceRNA to regulate Rab10 expression via competitively sponging to miR-224-5p, thereby regulating the progression of osteosarcoma. These findings may provide useful information to identify new biomarkers for early diagnosis and therapeutic applications in osteosarcoma. www.nature.com/scientificreports/

Data availability
The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request. inhibitor. The knockdown of EBLN3P markedly reduced cell proliferation, migration and invasion. However, co-transfection with si-EBLN3P and miR-224-5p inhibitor or Rab10 significantly increased cell proliferation, migration and invasion, in contrast to cells transfected with si-EBLN3P alone (Fig. 4C-F). EBLN3P endogenous bornavirus-like nucleoprotein, Rab10 Ras-related protein 10.