CircFAT1 sponges miR-375 to promote the expression of Yes-associated protein 1 in osteosarcoma cells

All animal experiments were carried out according to the Guide for the Care and Use of Laboratory Animals published by National Institutes of Health and approved by the Ethics Committee of Sir Run Run Shaw Hospital. All experiments strictly followed the panel’s specific guidelines regarding the care, treatment and euthanasia of animals used in the study.

Slices of formalin-fixed and paraffin-embedded primary osteosarcoma and chondroma tissues were obtained from biopsies in each 12 patients before administration of neo-adjuvant chemotherapy. Osteosarcoma, chondroma biopsies were histologically characterized by pathologists according to the criteria defined by the World Health Organization. Written informed consent was obtained from each patient before entering this study, and all study protocols were approved by the Ethics Committee of Sir Run Run Shaw Hospital.

Antibodies, mice, and other materials. Information regarding the materials used in this study is included in the Additional file 1: Materials and Methods.

The human cell line hFOB1.19, human osteosarcoma cell lines MG-63, u2os, SJSA-1, HOS and 143B were purchased from FuHeng Cell Center (Shanghai, China). The OS cell lines were authenticated at the by the ShangHai Biowing Applied Biotechnology Co. Ltd., performing a STR profiling analysis, as described by Capes-Davis and according to the ANSI Standard (ASN-0002) set forth by the ATCC Standards Development Organization. Mycoplasma testing was performed using the Venor GeM Mycoplasma Detection Kit (Minerva Biolabs, Berlin, Germany). Cells were cultured as described in detail in the Additional file 1: Materials section.

All the cell lines Cell culture procedures are described in the Additional file 1: Materials and Methods.

Nude mice (nu/nu, male 3- to 4-week-old) were injected subcutaneously with 5 × 10⁶ 143B stable cells. Tumor volumes were calculated from the length (a) and the width (b) by using the following formula: volume (ml³) = ab²/2. Five weeks after injection, the animals were sacrificed, and tumors were harvested (measured and weighed) and fixed in 4% paraformaldehyde. Wet tumor weight was calculated as mean weight ± standard deviation (SD) in each group.

The evidence for Ago2-binding sites was obtained from published online cross-linking immunoprecipitation (CLIP) data sets, which are available from doRiNA (a database of RNA interactions in post-transcriptional regulation, http://​dorina.​mdc-berlin.​de). These data sets include Ago2 HITS-CLIP and PAR-CLIP data from HEK-293 cells and several lymphoma cells. We downloaded the available data sets (http://​dorina.​mdc-berlin.​de) and acquired the Ago2-binding sites of circFAT1 genomic region.

RIP experiments were performed by using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, Bedford, MA). The Ago-RIP assay was conducted in HOS cells stably expressing vector or shcircFAT1. We first constructed the lentivirus vectors harbouring vector or shcircFAT1, and established two stable cell lines of HOS. Approximately 1×10⁷ cells were pelleted and re-suspended with an equal pellet volume of RIP Lysis Buffer (about 100 ml) plus protease inhibitors cocktail and RNase inhibitors. The cell lysates (200 μl) were incubated with 5 μg of control rabbit IgG or antibody against Ago2 (Millipore) coated beads with rotation at 4 °C overnight, respectively. After treating with proteinase K buffer, the immunoprecipitated RNAs were extracted by RNeasy MinElute Cleanup Kit (Qiagen) and reversely transcripted using Prime- Script RT Master Mix (TaKaRa). The abundance of circFAT1 level was detected by qRT–PCR assay.

Cells were seeded in 96-well plates at a density of 5×10³ cells per well 24 h before transfection. The cells were co-transfected with a mixture of 50 ng firefly luciferase (FL) reporter vectors, 5 ng Renilla luciferase (RL) reporter vectors (pRL-TK), and miRNA mimics at the indicated concentration. The miRNA mimics were obtained from Life Technologies. After 48 h, the luciferase activity was measured with a dual luciferase reporter assay system (Promega, Madison, WI). Each miRNA or NC RNA was co-transfected with RL reporter and FL reporter with or without the circFAT1 3’-UTR. For comparison, the FL activity was first normalized with RL activity. The effect of each miRNA on luciferase reporter with circFAT1 3’-UTR was then normalized with that on luciferase reporter without circFAT1 3’-UTR. Finally, the fold-change was calculated by each miRNA compared with NC.

Cy3-labeled locked nucleic acid miR-375 probes and Alexa Fluor 488-labeled circFAT1 probes were designed and synthesized by RiboBio (Guangzhou, China), and the probes sequences were available upon request. The signals of the probes were detected by Fluorescent In Situ Hybridization Kit (RiboBio, Guangzhou, China) according to the manufacturer’s instructions. The images were acquired on Nikon A1Si Laser Scanning Confocal Microscope (Nikon Instruments Inc, Japan).

Pull-down assay was performed as indicated: In brief, 1 × 10⁷ osteosarcoma cells were harvested, lysed, and sonicated. The circFAT1 probe was incubated with C-1 magnetic beads (Life Technologies) at 25°C for 2 h to generate probe-coated beads. The cell lysates were incubated with circFAT1 probe or oligo probe at 4°C overnight. After washing with the wash buffer, the RNA complexes bound to the beads were eluted and extracted with RNeasy Mini Kit (QIAGEN) for RT–PCR or real-time PCR. Biotinylated-circFAT1 probe was designed and synthesized by RiboBio (Guangzhou, China).

HOS and 143B cells were cultured in six-well plates and scraped with the tip of 200 μl pipette tips (time 0 h). Cell migration was photographed using 10 high-power fields at 0 and 24 h after injury. Remodeling was measured as diminishing distance across the induced injury, normalized to the 0 h control, and expressed as relative migration.

The migration and matrigel invasion assays were conducted by using transwell chamber (for migration assay) or transwell pre-coated matrigel chamber (for invasion assay) according to the manufacturer’s protocol (BD Science, Bedford, MA, USA). The homogeneous single cell suspensions (5 × 10⁴ cells/well for migration, 1 × 10 [5]/well for invasion) were added to the upper chambers and incubated for 24 h. The migration and invasion rates were quantified by counting the migratory and invasive cells at least three random fields.

For anoikis analysis, cells were cultured in suspension for 48 h. Cells were then trypsinized and stained with Annexin V- FITC –PI and analyzed by flow cytometry using the Annexin V-FITC/PI Apoptosis Detection kit (BD Biosciences) following the manufacturer’s instructions. Data were collected on a BD FACSCanto and analyzed using FlowJo software.

We performed in vitro experiments as described, including CCK-8 assay, cell transfection and colony formation [18], western blotting [19], flow cytometry, IHC and immunofluorescence microscopy [18, 20], real-time PCR using U6 as an internal control [20], and northern blotting [10, 14]. Given the different sizes of circFAT1 and linear FAT1 mRNA, the samples were separated on a 1.5% formaldehyde agarose gel for circFAT1 and a 0.6% gel for FAT1 mRNA by using voltages of 50 and 90 V, respectively. Tumor-formation assay was described previously [18]. Further details are provided in the Additional file 1: Materials and Methods.

Statistical analyses were performed with 22.0 SPSS. Data are represented as means with SDs, and statistical significance was determined with unpaired Student’s t tests, unless indicated otherwise. P values less than 0.05 were considered statistically significant. Tumor sizes were measured daily to observe dynamic changes in tumor growth and calculated by a standard formula:

To explore the function of circFAT1 in OS cells, we transfected circFAT1 small hairpin RNA (shRNA) constructs into 143B and HOS cells. This transfection targeted the junction sites of circFAT1 and established stable knockdown cells. The expression of circFAT1 was significantly reduced in these cells (Fig. 2a). In contrast, the expression of FAT1 mRNA did not change (Fig. 2a). Accordingly, the proliferation capabilities of OS cells decreased upon transfection with circFAT1 shRNA (shcircFAT1#1) (Fig. 2b). Flow cytometric analysis was conducted to determine the effect of circFAT1 knockdown on the apoptosis rate of OS cells at 48 h post-transfection. As a result, circFAT1 knockdown markedly enhanced OS cell apoptosis, and protein levels of cleaved caspase3 and cleaved PARP (Fig. 2c). Moreover, as shown in Fig. 2d, compared with control cells, circFAT1 knockdown cells exhibited compromised tumor formation. Knockdown of circFAT1 also suppressed the migration and invasion of OS cell lines in Transwell migration and Matrigel invasion assays (Fig. 2e). Consistently, the wound healing assay demonstrated that circFAT1 silencing significantly inhibited cell migration in HOS and 143B cells (Fig. 2f). Together, these results indicate that circFAT1 is involved in OS cell growth and motility in vitro.

It has been reported that circRNA functions as miRNA sponge in cancer cells [10, 14]. Because circFAT1 is abundant and stable in the cytoplasm, we performed an analysis of online AGO2 immunoprecipitation data including high-throughput sequencing from doRiNA. The results revealed a high degree of AGO2 occupancy in the circFAT1 region that is highly conserved across several vertebrate species (Fig. 3a). To validate this result, we conducted RNA immunoprecipitation for AGO2 in HOS cells and found that endogenous circFAT1 pulled-down from AGO2 antibodies was significantly enriched by qRT–PCR analysis compared with circFAT1 stable knockdown cells (Fig. 3b). To confirm whether circFAT1 could sponge miRNAs in OS cells, we selected 16 candidate miRNAs by overlapping the prediction results of miRNA recognition elements in the circFAT1 sequence using miRanda, Targetscan, and RNAhybrid (Fig. 3c). Next, we investigated whether circFAT1 could directly bind these candidate miRNAs. We designed a biotin-labeled circFAT1 probe and verified the pull-down of circFAT1 in OS cell lines. We also found that pull-down efficiency was significantly enhanced in circFAT1-overexpressing stable cells (Fig. 3d and e). The levels of the 16 candidate miRNAs were determined using a pull-down assay and detected by real-time PCR. As shown in Fig. 3f, five microRNAs were pulled-down by circFAT1 in both HOS and 143B cells. We then mutated each of the five miRNA predicted target sites using a luciferase reporter including circFAT1 in 3′- untranslated region (3′-UTR). Compared with the control, three miRNAs (miR-30a, miR-330, and miR-375) reduced the luciferase reporter activities by at least 25% (Fig. 3g). Transfection of the three miRNAs had no significant effect on luciferase activity when the corresponding target sites were mutated from the luciferase reporter. Because miR-375 decreased luciferase activity to the greatest extent and circFAT1 contained an miR-375 8mer-1a binding site (Fig. 3h), we transfected miR-375 mimics into HEK-293T cells and observed that the luciferase activities of WT reporter was significantly reduced when compared with the Mut reporter (Fig. 3i). Supporting this result, the RNA FISH assay revealed that circFAT1 interacts with miR-375 in the cytoplasm of OS cells (Fig. 3j) and tissues (Additional file 2: Figure S3). These results suggest that circFAT1 functions as a sponge of miR-375.

Our previous studies of TAZ confirmed its high expression in human OS [18, 22]. We also found that TAZ increases the migration, invasion, and proliferation of OS cells. Using real-time PCR and FISH, we found that miR-375 was downregulated in human OS tissues compared with chondroma tissues (Fig. 4a and b). miR-375 was also downregulated in OS cells compared with hFOB1.19 cells (Fig. 4c). To investigate the function of miR-375 in OS cell lines, we constructed over-expression of miR-375 in stable cells. We used CCK-8 assays to determine cell viability. The results showed a significant decrease in cell proliferation after upregulation of miR-375 expression (P < 0.05; Fig. 4d). Furthermore, we demonstrated that downregulation of miR-375 clearly induced lower levels of anoikis 48 h after transfection of the miR-375 sponge (Fig. 4e, P < 0.05). In addition, miR-375 overexpression inhibited colony formation compared with control cells (P < 0.05 for both; Fig. 4f). The Transwell assay (Fig. 4g) and wound healing assay (Fig. 4h) showed that overexpression of miR-375 significantly inhibited the migration and invasion of OS cells. These results indicate that miR-375 inhibits the migration, invasion, and proliferation of OS in vitro.

To elucidate the function of miR-375 is OS, we used PicTar and TargetScan algorithms to predict potential targets of miR-375. The YAP1 3′-UTR mRNA contained sequences complementary to the miR-375 seed sequence (Fig. 5a). Interestingly, YAP1 mRNA levels were higher in OS than in chondroma tissue (* P < 0.01; Fig. 5b). Furthermore, our immunohistochemistry results demonstrated that YAP1 was overexpressed in OS tissues compared with chondroma tissues; the YAP1-targeted genes c-Myc and Birc5 were also overexpressed (Fig. 5c).

To verify whether YAP1 is a direct target of miR-375, we constructed 3′-UTR sensors and co-transfected 293T cells with pre-miR-375. We observed reduced luciferase activity for of the YAP1 3′-UTR in the presence of miR-375 (Fig. 5d). To verify target specificity, we generated mutated forms of the YAP1 3′-UTR, in which the miR-375-binding site was abolished. Co-transfection of pre-miR-375 with the mutant construct abrogated the decrease in wild-type 3′-UTR luciferase activity, indicating that miR-375 specifically regulated YAP1 expression (Fig. 5d). Additionally, transfection of miR-375 mimics reduced endogenous YAP1 protein and mRNA levels, as well as the YAP1-targeted genes c-Myc and Birc5 (Fig. 5e, f and g). However, reducing endogenous miR-375 levels (using miR-375 inhibitor) had the opposite effect in both cell lines (Fig. 5e, f and g). These results indicate that miR-375 targets YAP1 in OS. Consistent with our previous studies [20], YAP1 is important for OS cell apoptosis during anoikis and cell detachment. Overexpression of YAP1 decreased anoikis in both HOS and 143B osteosarcoma cells (Fig. 5h).

We next investigated whether the tumor-suppressive function of miR-375 is mediated by YAP1 inhibition. Our immunofluorescence results (Fig. 6a) demonstrated that miR-375 overexpression decreased the expression of YAP1, c-Myc, and Birc5, which was rescued by YAP1 overexpression. Remarkably, we found that both YAP1 and miR-375 knockdown augmented OS cell anoikis compared with miR-375 knockdown alone (Fig. 6b). Furthermore, increasing YAP1 levels abrogated the growth inhibition (Fig. 6c) and colony formation (Fig. 6d) induced by miR-375 overexpression, as well as decreased cell migration and invasion of OS cells (Fig. 6e and f). These results support that miR-375 targets YAP1 and mediates the tumor-suppressive function of miR-375.

To examine whether circFAT1 promoted OS cell proliferation, migration, and invasion via interacting with miR-375, we co-transfected pre-miR-375 or miR-375 sponge and the circFAT1 overexpression or knockdown construct into OS cells. The results showed that both the protein and mRNA expression of YAP1 and target genes c-Myc and Birc5 significantly increased in OS cells co-transfected with shcircFAT1 plasmids and miR-375 sponge compared with cells transfected with shcircFAT1 alone (Fig. 7a and b). Immunofluorescence analysis confirmed that the expression of YAP1, c-Myc, and Birc5 increased in 143B and HOS cells transfected with shcircFAT1 and miR-375 sponge together when compared with shcircFAT1 alone (Fig. 7c). Knockdown of both miR-375 and circFAT1 resulted in a higher growth rate than in the circFAT1 inhibition group (Fig. 7d). Furthermore, anoikis cells were also upregulated after treatment with pre-miR-375 and circFAT1 co-transfected cells (Fig. 7e). In addition, downregulation of both miR-375 and circFAT1 promoted colony formation when compared with cells transfected with shcircFAT1 alone (P < 0.05 for both; Fig. 7f and Additional file 3: Figure S7a-b). In addition, Transwell Matrigel invasion and wound healing assays indicated that OS cells co-transfected with the miR-375 sponge and shcircFAT1 expression construct demonstrated enhanced invasion and migration capabilities when compared with shcircFAT1-transfected cells (Fig. 7g and h). Together, these data suggest that circFAT1 promotes cell migration, invasion, and proliferation via sponging miR-375 and subsequently induces YAP1 expression in vitro.

To investigate the effect of circFAT1 and miR-375 in vivo, circFAT1-deficient, miR-375 knockdown, or control 143B cells were subcutaneously injected into nude mice, after which tumor growth was evaluated. Cells lacking both miR-375 and circFAT1 had a higher growth rate than those lacking circFAT1 expression alone (Fig. 8a and b). In contrast, circFAT1 knockdown reduced the volume of 143B-derived tumors in vivo (Fig. 8c), an effect that was abrogated by co-expression with miR-375 sponge constructs. We observed similar results for the average tumor wet weight in the three groups (Fig. 8d).

We next evaluated the relationship between circFAT1 and YAP1 in vivo. circFAT1 knockdown reduced the mean area immunopositive for YAP1, c-Myc, and Birc5 in tumor tissues, as determined by immunohistochemistry (Fig. 8e), this corresponded to a decrease in the levels of YAP1, c-Myc, and Birc5 protein and mRNA (Fig. 8f and g). Inhibition of miR-375 reversed this effect. Interestingly, immunofluorescence (Additional file 4: Figure S5a) and western blot (Additional file 4: Figure S5b) results indicated that the upregulation of c-Myc expression induced by circFAT1 overexpression is independent on Wnt signal pathway. Besides, Kaplan-Meier survival curves of the TCGA sarcoma dataset showed that the patients with high expression of YAP and target genes had a low 10-year overall survival rate, while the patients with high expression of miR-375 had a high 10-year overall survival rate (Additional file 5: Figure S8). These results suggest that circFAT1 sponges miR-375 in OS and miR-375 mediates the tumor-suppressive function of circFAT1 in vivo (Fig. 8h).