Extracellular nanovesicles‐transmitted circular RNA has_circ_0000190 suppresses osteosarcoma progression

Abstract Under the microenvironment, tumour progression is substantially affected by cell‐cell communication. In spite of the mediating effect of extracellular nanovesicles (EVs) on cell‐cell communication by packaging into circRNAs, the effect of EVs circRNA hsa_circ_0000190 (circ‐0000190) in osteosarcoma is still not clear. Circ‐0000190 expressions in tissues and EVs from plasma were compared between osteosarcoma patients and controls. Thereafter, receiver operating characteristic (ROC) curve was drawn and area under the curve was calculated to examine whether the diagnostic results were accurate, and the effect of EVs circ‐0000190 was dug out via the determination of cell phenotypes and animal assays. Results showed circ‐0000190 exhibited an obvious reduction in EVs and tissues of osteosarcoma patients (P < .05). It was also discovered that EVs encapsulated the majority of circ‐0000190, and EVs‐encapsulated circ‐0000190 could be applied to make a distinction between osteosarcoma patients and controls. Besides, EVs circ‐0000190 in osteosarcoma cells transported from normal cells weakened the capacities of osteosarcoma cells to migrate, proliferate and invade, so as to block their biological malignant behaviours (P < .05). In addition, under the action of EVs circ‐0000190, tumour growth was impeded and the expression of TET1 was inhibited via the competitive binding to miR‐767‐5p. In all, EVs circ‐0000190 has a good prospect as it can be regarded as a new biomarker for detecting osteosarcoma. EVs circ‐0000190 transported from normal cells to osteosarcoma cells impeded the in vitro and in vivo development of osteosarcoma, implying that EVs circ‐0000190 exerts an effect on communication between normal cells and osteosarcoma cells in the carcinogenesis process of osteosarcoma.


Abstract
Under the microenvironment, tumour progression is substantially affected by cellcell communication. In spite of the mediating effect of extracellular nanovesicles (EVs) on cell-cell communication by packaging into circRNAs, the effect of EVs circRNA hsa_circ_0000190 (circ-0000190) in osteosarcoma is still not clear. Circ-0000190 expressions in tissues and EVs from plasma were compared between osteosarcoma patients and controls. Thereafter, receiver operating characteristic (ROC) curve was drawn and area under the curve was calculated to examine whether the diagnostic results were accurate, and the effect of EVs circ-0000190 was dug out via the determination of cell phenotypes and animal assays. Results showed circ-0000190 exhibited an obvious reduction in EVs and tissues of osteosarcoma patients (P < .05). It was also discovered that EVs encapsulated the majority of circ-0000190, and EVs-encapsulated circ-0000190 could be applied to make a distinction between osteosarcoma patients and controls. Besides, EVs circ-0000190 in osteosarcoma cells transported from normal cells weakened the capacities of osteosarcoma cells to migrate, proliferate and invade, so as to block their biological malignant behaviours (P < .05). In addition, under the action of EVs circ-0000190, tumour growth was impeded and the expression of TET1 was inhibited via the competitive binding to miR-767-5p. In all, EVs circ-0000190 has a good prospect as it can be regarded as a new biomarker for detecting osteosarcoma. EVs circ-0000190 transported from normal cells to osteosarcoma cells impeded the in vitro and in vivo development of osteosarcoma, implying that EVs circ-0000190 exerts an effect on communication between normal cells and osteosarcoma cells in the carcinogenesis process of osteosarcoma.

| INTRODUC TI ON
Osteosarcoma is the bone tumour that most commonly affects children, adolescents and young adults, 1 whose primary foci are the body parts such as the metaphysis of long bones of limbs where bones grow and repair actively. 2 Previously, surgical amputation is the major treatment method for osteosarcoma, potentially triggering high morbidity rate, trauma and short long-term survival. In the past ten years, the prognosis of patients was enhanced based on the breakthroughs in osteosarcoma, thus substantially increasing the five-year survival rate of osteosarcoma patients to about 60%-70%. 3 In the past decades, however, little effect has been produced in clinical results and prognosis in spite of prominent progresses in neo-adjuvant chemotherapy and surgery. 4 The exact mechanisms where osteosarcoma develops and progresses are unclear none the less, so determining the molecular development mechanism of osteosarcoma becomes the top priority for diagnosing and treating the disease. 5 Circular RNAs (circRNAs), generated from end-to-end ligation of RNA transcripts in the process of transcription, are closed-loop RNAs. 6 Despite over 40 years of investigation on circRNAs, 7 little attention has been paid to them until recently. Enriched in a variety of tissues 8 with the same sequence as corresponding linear isomers, circRNAs are formed via varying splicing mechanisms. 9 Research has manifested that circRNAs are found in cancer cells and show abnormal expressions in such epithelial cancers as lung cancer, 10 gastric cancer 11 and osteosarcoma. 12 CircRNAs consist of a ring structure that is hard to decompose compared to mRNAs, 13 and they significantly modulate cancers, especially the function of miRNAs by combining with them. 14 Chen S et al pointed out that circ-0000190 is clearly expressed in gastric cancer. 15 Meanwhile, Feng Y et al pointed out that circ-0000190 regulates the miR-767-5p/MAPK4 pathway so as to impede multiple myeloma to progress. 16 However, the relationship between circ-0000190 and osteosarcoma is still unknown.
Most types of cells can secret extracellular nanovesicles (EVs), and these EVs can also be seen in the body fluids. EVs consist of a lipid bilayer where genomic DNAs, RNAs (including mRNAs, miRNAs and other small RNAs), soluble and membrane-bound proteins, lipids and metabolites derived from the parent cells exist. 17 Carrying a variety of cargoes, EVs are considered to be the basic transmitters of cellular information and involved in the regulation of pleiotropic and biological functions in multicellular organisms. 18 Hence, EVs exert a good effect as disease biomarkers and have attracted much attention in recent years. 19 CircRNAs, small ncRNA family members, are enriched in EVs. Further, a mass of evidence shows that EVs are capable of participating in the mechanism of tumorigenesis by transmitting circRNA. 20 In this research, the expression of circ-0000190 in osteosarcoma cell lines and normal osteoblast cell line was examined via qRT-PCR, whose results elucidated that circ-0000190 was remarkably attenuated in osteosarcoma cell lines. Subsequently, the further measurement revealed that the expression circ-0000190 showed an obvious reduction in tissues and plasma EVs. The biological effects of EVs circ-0000190 on osteosarcoma have not been illustrated in reports yet. This study aims to find a potential biomarker for diagnosing osteosarcoma and to figure out whether EVs circ-0000190 is involved in extracellular communication so as to stimulate osteosarcoma to progress.

| Study design and subjects
Plasma samples were obtained from 60 osteosarcoma subjects and 60 healthy controls, and the corresponding 60 pairs of paracancerous normal tissues and tumour tissues were collected from osteosarcoma subjects in Liaoning Cancer Hospital & Institute, followed by analysis. This research received the informed consent from all the subjects and gained the approval of the institutional review board of Liaoning Cancer Hospital & Institute.

| Cell proliferation assay
Five randomly selected views in each well were captured using a fluorescent microscope for calculating EdU-positive cells. We performed all experiments in triplicate.

| Migration-related assays
Based on the methods mentioned above, Transwell invasion and migration assays were carried out for analysis. 21

| RNase R digestion
After kept at 37°C for 15 minutes, 3 units of RNase R (Epicentre Biotechnologies) were added to 1 μg RNA, respectively, to degrade the linear RNA. Following RNase R treatment, qRT-PCR was performed to detect the expressions of GAPDH and circ-0000190.

| RNA binding protein immunoprecipitation assay
RNA binding protein immunoprecipitation (RIP) assay was strictly carried out with reference to the instructions of the Millipore kit (Millipore, Bedford, MA, USA). Following cell lysis, each reaction system was added with 8 μg detection antibodies, in which cells were incubated at 4°C overnight, followed by reheating to room temperature for 1 hour.
Then, the complex was captured using Protein G magnetic beads, and the buffer was washed to extract RNA. The extracted RNA was then reversely transcribed, and qRT-PCR was adopted for RNA level detection.

| Separation of EVs
The hFOB1.19, MG63 and U2OS cells were plated onto 10-cm dishes at a concentration of 1.5 × 10 6 cells/ dish. After 48 hours, the culture media were discarded, and the cells were washed three times in phosphate-buffered saline (PBS). Next, the cells were cultured in serum-free media. After 72 hours, cell culture media were collected and 10 ml plasma from each sample was also collected for separation of EVs. The plasma and culture medium were collected and centrifuged at 3000 g for 15 minutes to remove cells and cellular debris. Then, we filtered the supernatant through a 0.22-μm PVDF filter (Millipore). In the case of filtered plasma, an appropriate volume of Thrombin was added to the plasma and centrifuged at 100,000×g, 5 minutes to make them compatible with ExoQuick exosome precipitation. Then, we added the appropriate volume of ExoQuick exosome precipitation solution (System Biosciences) to the Thrombin-treated plasma and filtered culture medium. After refrigeration for 24 hours, the ExoQuick/biofluid mixture was centrifuged at 1500 g for 30 minutes, and the supernatant was removed.
The EVs appear as a beige or white pellet at the bottom of the vessel.

| TEM
Prior to TEM analysis, 100 μl of PBS was added for EVs suspension and 5% glutaraldehyde for incubation at the incubation temperature and kept at 4°C. After that, EVs samples were placed dropwise on a copper grid coated with carbon, followed by 30 seconds of soaking in 2% phosphotungstic acid solution (pH 7.0) in accordance with the preparation procedures of TEM samples. Under a transmission electron microscope (TecnaiG2 Spirit Bio-Twin; FEI, USA), the preparations were observed.

| EVs labelling
Extracellular nanovesicles from 1.5 × 10 6 cells were suspended in 100 μl of PBS with 1 ml of mixed PKH67, a green fluorescent marker (Sigma, in Diluent C), followed by incubation at room temperature for 4 minutes. Subsequently, EVs labelling was terminated through the addition of 2 ml of BSA (0.5%), and Exosome Spin Columns (MW 3000; Thermo Fisher Scientific, Shanghai, China) was used to remove unincorporated dye from EVs labelling reactions. The ExoQuick Exosome Precipitation Solution was added for separation of stained EVs. As the negative control, EVs were collected without PKH67 dye. Thereafter, 9.6 ml of basal medium was taken for EVs suspension, and 250 μl was placed onto the subconfluent layer of MG63 and U2OS cells. Then, the cells were incubated for 3 hours at 37°C, rinsed and fixed at room temperature. DAPI (Sigma) was added for 10 minutes of nucleus staining, and a fluorescence microscope (Zeiss, LSM700B, Germany) was utilized to observe the stained cells.

| IHC
Representative areas were selected via H&E staining, and anti-TET1 (Abcam, Shanghai, China) or anti-Twist was used for IHC according to manufacturer's programme. (Abcam), Twist (Abcam) and GAPDH (Abcam) at 4°C overnight, followed by incubation with appropriate HRP-conjugated secondary antibodies at room temperature for 1 hour. Signals were detected by Immobilon ECL substrate (Millipore, Germany), and the images were acquired using an Optimax X-ray Film Processor (Protec, Germany).

| Characteristics of patients
Characteristics of the osteosarcoma patients and healthy controls are displayed in Table 1. Differences in age, gender and smoking status were not discovered between patients and controls (P > .05).

| Characteristics of circ-0000190 in osteosarcoma
The expression of circ-0000190 in osteosarcoma cell line and human-derived osteoblasts hFOB1.19 was measured via qRT-PCR, and the results showed that circ-0000190 expression was significantly reduced in osteosarcoma cell line ( Figure 1A). Subsequently, we de- This assay suggested that circ-0000190 is indeed a circRNA which is resistant to RNase R digestion ( Figure 1C). We confirmed that the circ-0000190 sequence amplified by the primer was identical to its sequence in circbase through Sanger sequencing ( Figure 1D).
Then, EVs RNA was extracted from plasma, and EVs from plasma of cases and controls were separated and their characteristics were analysed. According to TEM results, cases had the same size of EVs with controls (50-150 nm, Figure 1E). The existence of exosome markers, TSG101 and CD63 was proved by Western blotting ( Figure 1F). Results mentioned above indicated that EVs circ-0000190 might act as a useful biomarker for discriminating patients with osteosarcoma from healthy controls. Thereafter, the expression of EVs circ-0000190 in plasma was detected, and it was discovered that cases had a lower expression of EVs circ-0000190 than controls (P < .05, Figure 1G). Additionally, the corresponding ROC curves were used to investigate the underlying value of EVs circ-0000190 as a non-invasive biomarker. The effect of EVs circ-0000190 on diagnosing restenosis is presented in Figure 1H.
The Youden index was 0.769, and the sensitivity and specificity were 85.0% and 78.3%, respectively.

| Effect of circ-0000190 on osteosarcoma cellular phenotype
The biological effect of circ-0000190 in vitro was explored in this study since circ-0000190 was found to be lowly expressed in tissues and EVs in plasma from osteosarcoma cases. After that, MG63 and U2OS cells were transfected with circ-0000190 overexpression plasmids or negative control vectors (NC), and qRT-PCR was carried out to determine circ-0000190 expression (Figure 2A). In the first

| EVs circ-0000190 mediates intercellular communication
The existing pattern of extracellular circ-0000190 was investigated, and TEM was carried out to determine the size of EVs ( Figure 3A).
According to Western blotting results, TSG101 and CD63 existed ( Figure 3B), and it was demonstrated that circ-0000190 expression was significantly higher in hFOB1.19 cells than in U2OS and MG63 cells ( Figure S1A). Meanwhile, EVs circ-0000190 expression exhibited a notably lower expression in U2OS and MG63 cells than in hFOB1.19 cells ( Figure S1B). Furthermore, circ-0000190 levels displayed approximately fourfold increases in EVs in comparison with producer cells ( Figure 3C). Hence, EVs from U2OS and MG63 cells possessed less circ-0000190 than those from hFOB1. 19  PKH67 was found to be located in the cytoplasm of recipient cells ( Figure 3D).

| Subcellular distribution of circ-0000190
Subcellular distribution of circRNA determines its biological function. To confirm the cellular localization of circ-0000190, we iso-

| circ-0000190 is targeted by miR-767-5p
Given that circ-0000190 was primarily located in the cytoplasmic fraction, we hypothesized that circ-0000190 may act as a ceRNA Note: Total data from 60 tumour tissues of osteosarcoma patients were analysed. For the expression of circ-0000190 was assayed by qRT-PCR, the median expression level was used as the cut-off. Data were analysed by chi-squared test and Fisher's exact test. P-value in bold indicates statistically significant.
in the development of osteosarcoma. QRT-PCR data revealed that miR-767-5p expression was higher in osteosarcoma tumour tissues, which was contrary to the expression trend of circ-0000190 ( Figure 5B). Through bioinformatics prediction (RegRNA, Starbase), we found that sequences in miR-767-5p that were highly matched to circ-0000190 3'UTR. Based on these binding sequences, pGL3circ-0000190-WT and pGL3-circ-0000190-MUT were established ( Figure 5C). Luciferase activity was obviously down-regulated in MG63 and U2OS cells cotransfected with circ-0000190 WT and miR-767-5p mimics, while it did not change after transfection with circ-0000190 MUT ( Figure 5D). RIP analysis was carried out to elucidate whether circ-0000190 was involved in RNA-containing ribonucleoprotein complex. QRT-PCR results showed that circ-0000190 was enriched in anti-Ago2 antibody than controls. Similar results were yielded in miR-767-5p ( Figure 5E). It is suggested that miR-767-5p can bind to circ-0000190 in vitro.

| circ-0000190 regulates TET1, the target gene of miR-767-5p
The potential role of miR-767-5p in the development of osteosarcoma was explored by the screening of miR-767-5p target genes using bioinformatics prediction (TargetScan, Starbase, RegRNA). cells, respectively ( Figure 6A). Luciferase activity of the WT reporter was inhibited, but that of MUT reporter did not change ( Figure 6B).
The above findings imply that TET1 is a potential target gene of miR-767-5p. Subsequently, TET1 expression in osteosarcoma cell line was determined via qRT-PCR. The mRNA levels of TET1 were remarkably inhibited in osteosarcoma tumour tissues ( Figure 6C). Interestingly, there was a significant correlation of the expression levels between circ-0000190 and TET1 ( Figure 6D). Besides, miR-767-5p mimics lowered TET1 expression, while EVs circ-0000190 prominently reversed this effect ( Figure 6E). vectors and circ-0000190-EVs ( Figure 7D). Moreover, Western blotting detection and IHC for TET1 were carried out to detect TET1 expression, the results of which illustrated that TET1 expression was evidently stimulated in the models of circ-0000190 overexpression and circ-0000190-EVs ( Figure 7E, 7). In the meantime, the EMT marker gene Twist was significantly inhibited in circ-0000190 overexpression and circ-0000190-EVs models ( Figure 7E, 7), implying that perhaps TET1 plays a role in the pathogenesis of osteosarcoma through the modulation on the EMT process.

| D ISCUSS I ON
By analysing osteosarcoma microarray data, 22  Therefore, EVs circ-0000190 can be a potential biomaker for osteosarcoma detection.
Existing evidence has proved that the physiological condition of the donor cells can be reflected by EVs circRNAs, and these EVs cir-cRNAs will trigger a sequence of cellular responses after they were captured by recipient cells. 23 Furthermore, it has been reported that EVs circRNAs are potential cancer biomarkers for different cancer patients 24 which indicates that EVs circRNAs can be applied for cancer identification as potential biomarkers.
In this research, the size and shape of plasma EVs were analysed via TEM, and TSG101 and CD63 (exosome markers) were used to verify exosomes. 25  Data in this study verified that circ-0000190 was generated mainly through secretion from cells via EVs. Cancer-related reports involve a large quantity of data on the effects of EVs circRNAs. 27 EVs significantly influence cell-cell communication and jointly change the physiological function of the recipient cells with bioactive factors, including circRNAs. 28 For instance, EVs from normal cells transferred PTENP1 to bladder cells, which suppresses bladder cancer progression. 29 With efforts made in this study, it was QRT-PCR results showed that circ-0000190 was primarily distributed in the cytoplasmic fraction of osteosarcoma cells. We may conclude that circ-0000190 participated in the development of osteosarcoma through post-transcriptional regulation. Hence, it was speculated that circ-0000190 may be a ceRNA, participating in tumorigenesis of osteosarcoma. TET1 belongs to the TET (ten-eleven translocation) family. TET1 plays a crucial role in the DNA methylation process and gene activation. 30