Introduction

Osteosarcoma is an aggressive form of bone cancer, which often affects the long bones, including the legs and arms1,2. Osteosarcoma is commonly diagnosed in children and adolescents. Osteosarcoma patients have common symptoms, such as pain, swelling, limited joint movement and bone fractures3,4. The treatments of osteosarcoma include surgery, chemotherapy, radiation therapy and immunotherapy5,6,7. Although early detection and conventional treatment have improved the survival rate of osteosarcoma, some patients have a worse prognosis due to drug resistance and distant metastasis8. Therefore, it is pivotal to explore the molecular mechanisms of osteosarcoma occurrence and progression and uncover the novel treatment9,10,11.

It has been documented that noncoding RNAs (ncRNAs) regulate tumor development and progression12,13. Noncoding RNAs are divided into several groups: long ncRNAs (more than 200 nucleotides) and small ncRNAs (less than 200 nucleotides)14,15,16. Evidence has uncovered that ncRNAs regulate cell growth, invasion, metastasis, apoptosis, autophagy, drug resistance, and immunotherapy in human cancer17,18. LncRNA HOXA-AS3 (HOXA cluster antisense RNA 3) has been reported to regulate specific targets and participate in disease development19,20. For instance, lncRNA HOXA-AS3 regulated lineage commitment of mesenchymal stem cells via interaction with EZH2 (Enhancer of Zeste 2)21. LncRNA HOXA-AS3 was found to regulate endothelium inflammation via integrating NF-kappaB signaling pathway and regulating the expression of IκBα and p65 acetylation22. LncRNA HOXA-AS3 was identified in endometriosis development via construction of ceRNA network23. One study showed that lncRNA HOXA-AS3 regulated the miR-675-3p and PDE5A (phosphodiesterase 5A) expression and accelerated the progression of pulmonary arterial hypertension24. Another study showed that downregulation of lncRNA HOXA-AS3 increased the expression of miR-455-5p and reduced the expression of p27, which reduced the atherosclerosis progression25.

Wu et al. reported that overexpression of lncRNA HOXA-AS3 facilitated tumor malignant progression via regulation of cell apoptosis and cell cycle progression as well as cell migration in glioma. The glioma patients with higher expression of lncRNA HOXA-AS3 displayed a poor prognosis26. In lung adenocarcinoma, lncRNA HOXA-AS3 enhanced the stability of HOXA6 mRNA via formation of an RNA duplex. LncRNA HOXA-AS3 accelerated invasion, migration and progression via binding with NF110. Moreover, NF110 governed the subcellular distribution of HOXA-AS3 in A549 cells27. In hepatocellular carcinoma (HCC), lncRNA HOXA-AS3 was reported to promote cell metastasis, EMT and proliferation via enhancing the activation of the mitogen-activated protein kinase (MAPK) and extracellular regulated protein kinase (ERK) through sponging miR-29c28. Similarly, lncRNA HOXA-AS3 was demonstrated to increase cell migration and proliferation via regulating miR-455-5p and PD-L1 axis in HCC cells29. LncRNA HOXA-AS3 was reported to involve in osteogenesis ossification from mesenchymal stem cells30. Xiao et al. reported that lncRNA HOXA-AS3 targeted the miR-1286 and TEAD1 axis, leading to promotion of osteosarcoma progression31. However, the molecular mechanism of lncRNA HOXA-AS3-mediated oncogenesis is poorly explored in human osteosarcoma. In this present study, we utilized several cellular biological strategies to explore the biological function of lncRNA HOXA-AS3 in osteosarcoma cells. We also utilized several molecular biological strategies to determine the underlying mechanism of lncRNA HOXA-AS3-mediated oncogenesis in osteosarcoma.

Results

Inhibition of lncRNA HOXA-AS3 reduces colony formation

To explore the function of lncRNA HOXA-AS3 on proliferation of U2OS and SW1353 cells, we used shRNA approach to downregulate the expression of lncRNA HOXA-AS3. The RT-PCR results showed that lncRNA HOXA-AS3 expression was significantly downregulated in U2OS and SW1353 cells (Fig. 1A). Among three shRNAs for depletion of HOXA-AS3, sh-HOXA-AS3 #2 exhibited the powerful for downregulation of lncRNA HOXA-AS3. Hence, we used sh-HOXA-AS3 #2 in the flowing experiments to knockdown the expression of lncRNA HOXA-AS3. To determine the function of lncRNA HOXA-AS3 on osteosarcoma cells, colony formation experiments were performed in U2OS and SW1353 cells after sh-HOXA-AS3 transfection. We found that the number of colony formation was decreased in both U2OS and SW1353 cell lines after HOXA-AS3 expression was reduced (Fig. 1B). This study suggested that inhibition of lncRNA HOXA-AS3 reduced colony formation of osteosarcoma cells.

Figure 1
figure 1

Downregulation of lncRNA HOXA-AS3 reduces colony formation of osteosarcoma cells. (A) RT-PCR data showed that sh-HOXA-AS3 transfection reduced the expression of HOXA-AS3 in U2OS and SW1353 cells. (B) Left panel: Colony formation assays showed that HOXA-AS3 transfection reduced colony formation in U2OS and SW1353 cells. Right panel: quantitative data are shown for colony formation. ***p < 0.001 versus control group.

Inhibition of lncRNA HOXA-AS3 reduces migration and invasion

It is known that lncRNA HOXA-AS3 downregulation inhibited the cell invasion and migration in cervical cancer cells via targeting miR-29a-3p32. LncRNA HOXA-AS3 downregulation reduced migration and invasion via targeting miR-29a-3p in gastric cancer33. In addition, lncRNA HOXA-AS3 promoted migration and invasion of A549 lung cancer cells27. Therefore, we used wound healing assay and Transwell invasion assay to measure the cell migratory and invasive abilities in osteosarcoma cancer after lncRNA HOXA-AS3 downregulation. As shown in Fig. 2A and B, sh-HOXA-AS3 transfection reduced the wound healing, suggesting that lncRNA HOXA-AS3 downregulation inhibited the cell migratory ability in SW1353 and U2OS cells. Moreover, Transwell assay data demonstrated that sh-HOXA-AS3 transfection reduced the number of invaded cells in SW1353 and U2OS cells (Fig. 2C,D). Taken together, inhibition of lncRNA HOXA-AS3 attenuated migration and invasion of osteosarcoma cells.

Figure 2
figure 2

Downregulation of lncRNA HOXA-AS3 reduces migration and invasion of osteosarcoma cells. (A) Wound healing assays showed that sh-HOXA-AS3 transfection inhibited migratory ability of SW1353 and U2OS cells. (B) Quantitative data are shown for wound healing assay. (C) Transwell invasion assays showed that sh-HOXA-AS3 transfection reduced invasion of SW1353 and U2OS cells. (D) Quantitative data are shown for invasion ability of osteosarcoma cells. **p < 0.01; ***p < 0.001 versus control group.

LncRNA HOXA-AS3 interacts with miR-218-5p

It has been documented that lncRNAs perform their functions in part via sponging miRNAs to reduce the miRNA target’s expression. Hence, we determined the miRNAs that can interact with lncRNA HOXA-AS3 in osteosarcoma. Several studies have shown the critical role of miR-218-5p in osteosarcoma development34,35,36. According to the TargetScan database, we found the interacting sites between lncRNA HOXA-AS3 and miR-218-5p (Fig. 3A). Our result showed that miR-218-5p inhibitors treatment suppressed the expression of miR-218-5p, while miR-218-5p mimic increased the expression of miR-218-5p (Fig. 3B). To validate the binding sites between miR-218-5p and HOXA-AS3, we did the luciferase reporter gene assay. We found that miR-218-5p mimic decreased the luciferase activity in the lncRNA HOXA-AS3 wild-type group, while HOXA-AS3 mutant group did not exhibit the change of luciferase activity (Fig. 3C). Furthermore, miR-218-5p inhibitor treatment elevated the luciferase activity in the HOXA-AS3 wild-type group, whereas miR-218-5p inhibitor did not change the luciferase activity in the HOXA-AS3 mutant group (Fig. 3C). Hence, lncRNA HOXA-AS3 could interact with miR-218-5p in osteosarcoma cells.

Figure 3
figure 3

LncRNA HOXA-AS3 interacts with miR-218-5p in osteosarcoma cells. (A) Potential binding sites between miR-218-5p and lncRNA HOXA-AS3 and FOXP1. (B) RT-PCR data showed that miR-218-5p inhibitors decreased the expression of miR-218-5p. In addition, miR-218-5p mimic transfection increased the miR-218-5p expression levels. (C) Dual luciferase reporter assays showed that miR-218-5p interacted with lncRNA HOXA-AS3. miR-218-5p regulated luciferase activity in HOXA-AS3 wild-type group. (D) Dual luciferase reporter assays showed that miR-218-5p interacted with and regulated FOXP1. **p < 0.01 versus control group. (E) Western blotting analysis data showed that lncRNA HOXA-AS3 and miR-218-5p regulate the expression of FOXP1 in osteosarcoma cells. WT: wild type; MUT: mutant.

FOXP1 interacts with miR-218-5p

FOXP1 has been reported to drive osteosarcoma development via inhibition of p21, retinoblastoma protein (RB) transcription and inactivation of p5337. To explore the potential target of miR-218-5p, we used TargetScan database and found that FOXP1 could be a potential target of miR-218-5p because there are several interacting sites between FOXP1 and miR-218-5p (Fig. 3A). The luciferase assay data revealed that miR-218-5p mimics decreased the luciferase activity in the FOXP1 wild-type group, while FOXP1 mutant group did not exhibit the change of luciferase activity (Fig. 3D). Furthermore, miR-218-5p inhibitor treatment elevated the luciferase activity in the FOXP1 wild-type group, whereas miR-218-5p inhibitor did not change the luciferase activity in the FOXP1 mutant group (Fig. 3D). Our western blotting data showed that miR-218-5p mimic transfection inhibited the expression of FOXP1 in SW1353 cells, while HOXA-AS3 upregulation increased the expression of FOXP1 in U2OS cells (Fig. 3E). Overexpression of HOXA-AS3 abrogated the miR-218-5p mimic-mediated inhibition of FOXP1 expression in SW1353 cells. Similarly, miR-218-5p mimics abolished the pcDNA HOXA-AS3-mediated inhibition of FOXP1 in U2OS cells (Fig. 3E). Hence, lncRNA HOXA-AS3 could regulate FOXP1 in osteosarcoma cells.

Downregulation of lncRNA HOXA-AS3 inhibits colony formation via miR-218-5p

To test whether lncRNA HOXA-AS3 regulates colony formation of osteosarcoma cells via miR-218-5p, sh-HOXA-AS3 and miR-218-5p inhibitor were co-transfected into osteosarcoma cells. We found that sh-HOXA-AS3 transfection inhibited colony formation in SW1353 and U2OS cells (Fig. 4A,B). Inhibition of miR-218-5p by its inhibitor abrogated sh-HOXA-AS3-mediated inhibition of colony formation in osteosarcoma cells (Fig. 4A,B). Likewise, overexpression of FOXP1 by pcDNA FOXP1 transfection abrogated sh-HOXA-AS3-induced inhibition of colony formation in osteosarcoma cells (Fig. 4A,B). Taken together, downregulation of lncRNA HOXA-AS3 reduced colony formation via miR-218-5p and FOXP1 in osteosarcoma cells.

Figure 4
figure 4

Downregulation of lncRNA HOXA-AS3 reduces colony formation via miR-218-5p and FOXP1. (A) Colony formation assays were performed to measure the colony formation in osteosarcoma cells co-transfected with sh-HOXA-AS3 and pcDNA FOXP1, or sh-HOXA-AS3 and miR-218 inhibitors. (B) Quantitative data are shown for colony formation. ***p < 0.001 versus control group. #p < 0.05; ##p < 0.01 versus sh-HOXA-AS3 alone.

Inhibition of lncRNA HOXA-AS3 reduces migratory ability via miR-218-5p and FOXP1

Wound healing assay was performed to measure whether inhibition of lncRNA HOXA-AS3 regulates cell migration via regulation of miR-218-5p and FOXP1 in osteosarcoma cells. To address this question, co-transfection of sh-HOXA-AS3, miR-218-5p inhibitor, or pcDNA FOXP1 was done in osteosarcoma cells. We observed that inhibition of lncRNA HOXA-AS3 reduced would healing in SW1353 and U2OS cells (Fig. 5A,B). Moreover, miR-218-5p inhibitor transfection abrogated sh-HOXA-AS3-induced suppression of wound healing in osteosarcoma cells (Fig. 5A,B). Furthermore, pcDNA FOXP1 transfection abolished sh-HOXA-AS3-involved suppression of wound healing in osteosarcoma cells (Fig. 5A,B). Altogether, downregulation of lncRNA HOXA-AS3 reduced wound healing via miR-218-5p and FOXP1 in osteosarcoma cells.

Figure 5
figure 5

Downregulation of lncRNA HOXA-AS3 reduces cell migration via miR-218-5p and FOXP1. (A) Wound healing assays were performed to measure migratory ability in osteosarcoma cells co-transfected with sh-HOXA-AS3 and pcDNA FOXP1, or sh-HOXA-AS3 and miR-218 inhibitors. (B) Quantitative data are shown for colony formation. **p < 0.01; ***p < 0.001 versus control group. #p < 0.05; ##p < 0.01 versus sh-HOXA-AS3 alone.

Suppression of lncRNA HOXA-AS3 inhibits invasive ability via miR-218-5p and FOXP1

To test whether lncRNA HOXA-AS3 inhibition attenuated invasive ability via regulation of miR-218-5p and FOXP1, osteosarcoma cells were co-transfected with sh-HOXA-AS3 and miR-218-5p inhibitor, or sh-HOXA-AS3 and pcDNA FOXP1. The results from the Transwell assay showed that sh-HOXA-AS3 transfection reduced cell invasive ability in osteosarcoma cells (Fig. 6A,B). This phenotype was rescued by downregulation of miR-218-5p by its inhibitors (Fig. 6A,B). Similarly, overexpression of FOXP1 also rescued sh-HOXA-AS3-mediated suppression of cell invasive ability in osteosarcoma cells. Hence, silencing of lncRNA HOXA-AS3 reduced cell invasive ability in part via miR-218-5p and FOXP1 in osteosarcoma cells.

Figure 6
figure 6

Downregulation of lncRNA HOXA-AS3 reduces cell invasion via miR-218-5p and FOXP1. (A) Transwell assays were performed to measure cell invasive ability in osteosarcoma cells co-transfected with sh-HOXA-AS3 and pcDNA FOXP1, or sh-HOXA-AS3 and miR-218 inhibitors. (B) Quantitative data are shown for colony formation. ***p < 0.001 versus control group. #p < 0.05; ##p < 0.01 versus sh-HOXA-AS3 alone.

Discussion

Evidence has revealed that lncRNAs are critical in regulating oncogenesis and drug resistance in human cancer38,39,40,41. LncRNA SATB2-AS1 targeted the expression of SATB2 and increased the tumor metastasis and proliferation via influencing the tumor microenvironment in osteosarcoma42. LncRNA XLOC_006786 inhibited tumor invasion, metastasis and proliferation via targeting miR-491-5p and Notch3 pathway in osteosarcoma43. LncRNA HOTAIRM1 regulated miR-664b-3p/Rheb/mTOR pathway and caused induction of proliferation via aerobic glycolysis in osteosarcoma44. LncRNA SNHG14 repressed ferroptosis and caused nutlin3a resistance through regulation of the miR-206 and SLC7A11 in osteosarcoma45. LncRNA EBLN3P regulated the expression of O-GlcNAc pathway via sponging miR-200a-3p, leading to methotrexate resistance in osteosarcoma cells46. Our previous study showed that lncRNA SCAMP1 regulated miR-26a-5p and ZEB2, contributing to promotion of cell invasion and viability in osteosarcoma47. These findings suggested that lncRNAs are involved in osteosarcoma development and progression.

LncRNA HOXA-AS3 was identified in pediatric gliomas as a biomarker for predicting prognosis48. LncRNA HOXA-AS3 regulated the EMT pathway and triggered tumor progression in epithelial ovarian cancer49. Moreover, lncRNA HOXA-AS3 regulated the miR-29a-3p and modulated tumor progression in cervical cancer. LncRNA HOXA-AS3 expression displayed a prognostic value in cervical cancer patients32. Furthermore, lncRNA HOXA-AS3 increased cell proliferation via modulation of miR-218-5p expression in oral squamous cell carcinoma50. Additionally, lncRNA HOXA-AS3 targeted the miR-4319 and SPNS2 expression, which caused the malignant progression in colorectal cancer51. Downregulation of lncRNA HOXA-AS3 blocked the tumor development via sponging miR-29c and upregulating CDK6 in pancreatic cancer52. One group reported that lncRNA HOXA-AS3 facilitated tumor progression via activation of NF-κB pathway and targeting the miR-29a-3p and LTβR in gastric cancer33. Another group reported that lncRNA HOXA-AS3 regulated the miR-455-5p and USP3 expression and enhanced malignant progression in glioblastoma multiforme53. In line with these findings, our study reported that lncRNA HOXA-AS3 facilitated cell proliferation and invasion via targeting miR-218-5p in osteosarcoma.

Several studies revealed that miR-218-5p inhibited cell proliferation, migration and invasion in MG63, U2OS and 143B osteosarcoma cells35,36. One study showed that miR-218-5p was involved in doxorubicin resistance in HOS and U2OS osteosarcoma cells34. In the current study, we identified the tumor suppressive function of miR-218-5p in U2OS and SW1353 osteosarcoma cells. LncRNA HOXA-AS3 has been reported to regulate the chemotherapeutic drug resistance in human cancer. For example, RNA sequencing analysis identified the numerous lncRNAs and miRNAs in small cell lung cancer, which are related with chemotherapy insensitivity, including lncRNA HOXA-AS354. In non-small-cell lung carcinoma (NSCLC), lncRNA HOXA-AS3 interacted with HOXA3 and regulated EMT (epithelial-mesenchymal transition) and led to cisplatin resistance55. In bladder cancer, lncRNA HOXA-AS3 reduced chemosensitivity via working as a ceRNA to sponge miR-455-5p and increase the expression of Notch156. One study revealed that depletion of lncRNA HOXA-AS3 reduced cell proliferation, migration, invasion in osteosarcoma cells via targeting miR-1286/TEAD1 axis. Overexpression of HOXA-AS3 was found in osteosarcoma tissues and was linked to poor outcomes31. Consistently, our study supported the oncogenic role of HOXA-AS3 in osteosarcoma cells. Furthermore, our study demonstrated that lncRNA HOXA-AS3 enhanced tumor progression via targeting miR-218-5p/FOXP1 in osteosarcoma.

In conclusion, lncRNA HOXA-AS3 regulated the expression of FOXP1 via sponging miR-218-5p, which led to promotion of colony formation, migration and invasion in osteosarcoma. Several limitations should be mentioned in this study. The mouse model is required to determine whether lncRNA HOXA-AS3 facilitates tumor growth in osteosarcoma. It is necessary to measure the expression of lncRNA HOXA-AS3 in osteosarcoma tissue samples and determine the association between HOXA-AS3 expression and prognosis in osteosarcoma patients. It is unclear whether lncRNA HOXA-AS3 could promote drug resistance in osteosarcoma, which is required to fully explore in the future. Taken together, targeting lncRNA HOXA-AS3 could be an alternative strategy to treat osteosarcoma. It is important to note that using suppression of lncRNA HOXA-AS3 expression is still premature for treating osteosarcoma patients. This could be due to targeting issues, delivery issues, safety and side effects. LncRNA HOXA-AS3 could regulate the expression of numerous downstream targets. The effective delivery of lncRNAs or their inhibitors to target cells remains a technical challenge. One lncRNA affects multiple signaling pathways and governs several cellular functions. Regulation of lncRNAs could affect normal cell function and have potential genotoxicity. Hence, there are relatively few clinical trials and studies on the role of lncRNAs in human cancer treatment.

Materials and methods

Cell culture

The U2OS human osteosarcoma cell line and SW1353 chondrosarcoma cell line were obtained from the cell bank of the Shanghai Cell Institute. The U2OS and SW1353 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) with 1% penicillin–streptomycin and 10% fetal bovine serum. All cells were maintained at 37 °C incubator with 5% CO2.

Transfection

The osteosarcoma U2OS and SW1353 cells were cultured in 6-well plates. Osteosarcoma cells were transfected with sh-HOXA-AS, shNC, 100 nM miR-218 inhibitor, 50 nM miR-218 mimics (Gene-Pharma CO. Ltd, Shanghai, China) by Lipofectamine 2000 as described previously57. Then, the transfected cells were further analyzed for their viability, invasion and migration.

Quantitative real-time reverse transcription-PCR

TRLzol reagent was used to extract total RNA from the transfected osteosarcoma cells. 1 µg of RNA was used to obtain cDNA via reverse transcription. PCR was performed by SYBR Green Kit as described before47. The expression of GAPDH was for normalization. The primer sequences were as follows: HOXA-AS3: Forward CAC CTC TCT CAT CGA AAA AA CG, Reverse GCA CCA GGA AAG AGG ACA ATTC; miR-218-5p: Forward AGC GAG ATT TTC TGT TGT GCT T, Reverse GAC GTT CCA TGG TGC TTG AC; GAPDH, Forward TGT GGG CAT CAA TGG ATT TGG; Reverse ACA CCA TGT ATT CCG GGT CAA T. The thermocycling conditions of qPCR were as follows: for HOXA-AS3, 5 min at 95 °C, followed by 40 cycles of 95 °C for 30 s and 65 °C for 45 s; for miR-218-5p, 5 min at 95 °C, followed by 40 cycles at 95 °C for 15 s, 60 °C for 30 s, and 72 °C for 30 s.

Colony formation assay

The transfected U2OS and SW1353 osteosarcoma cells were cultured in 6-well plates and maintained for 14 days at 37 °C incubator with 5% CO2. After the cells were washed by PBS four times, 4% paraformaldehyde was added to fix for 45 min. 0.1% crystal violet was used to stain the colonies for 20 min. Microscope was used to take images and the colony numbers were counted.

Transwell invasion assay

The transfected U2OS and SW1353 osteosarcoma cells (5 × 104 cells) were cultured in top level of Transwell plates. The top level of plates was filled with serum-free DMEM, while the bottom level of Transwell plates was added DMEM with 10% FBS. After 20 h, we removed the culture medium and cells on the top level of plates. Then, the bottom plates were stained using 0.1% crystal violet after the invaded cells were fixed by 4% paraformaldehyde for 30 min. An inverted microscope was used to take images for the invaded cells.

Wound healing assay

The transfected U2OS and SW1353 osteosarcoma cells were cultured on 6-well plates. The 200 µl pipette tip was used to produce a wound after cells reached about 95% fluence. PBS was used to remove the floatage cells. Microscope was used to take images near the wound area after 20 h as described before58.

Western blotting analysis

The transfected U2OS and SW1353 osteosarcoma cells were lysed using RIPA lysis buffer to extract total protein. We denatured the protein by boiling for 5 min. Then, 30 µg proteins were subjected to 10% SDS-PAGE gels at 100 voltage for 90 min, and then transferred to PVDF membrane at 45 voltage overnight. 5% non-fat dry milk was used to block the membrane. The primary antibody FOXP1 (1:2000, #2005, Cell Signaling Technology) were used to incubate the membrane at cold room overnight. GAPDH (1:5000, #2118, Cell Signaling Technology) was the loading control. Western blotting analysis were performed as described previously (Supplementary information)59.

Luciferase report assay

Luciferase report assay was performed to determine the interaction between lncRNAs and miRNAs in this study. The mutant binding sequences of miR-218 in lncRNA HOXA-AS3 were cloned into pmirGLO dual-luciferase vector. The plasmids were transfected into U2OS and SW1353 cells. The luciferase activity was determined as described before47.

Statistical analysis

All results were analyzed by GraphPad Prism 5.0. Student t-test was used to determine the significant difference between two groups. Analysis of variance (ANOVA) was used to determine the significant difference between three or more groups. Means ± SD was used to represent the results. P < 0.05 was considered as statistically significant.