CircYAP1 inhibits osteosarcoma progression through the miR-135b-5p/FRK axis

Summary

Circular RNAs (circRNAs) are implicated in the progression of various malignancies, including osteosarcoma. However, little is known about the role of circYAP1 in osteosarcoma. In the present study, the expression of circYAP1 was found to be downregulated in osteosarcoma tissues relative to adjacent normal tissues. Decreased circYAP1 expression was also found in metastatic osteosarcoma specimens relative to nonmetastatic specimens, and low circYAP1 group had a higher metastasis rate and more advanced clinical stage than high circYAP1 group. In vitro functional assays demonstrated the inhibitory effects of circYAP1 on osteosarcoma cell proliferation, invasion, and migration. Regarding the mechanism, circYAP1 acts as a competitive endogenous RNA (ceRNA) to sponge miR-135b-5p, whereas FRK is a direct target of miR-135b-5p. Finally, a tumor-suppressing role of circYAP1 was demonstrated in the orthotopic lung metastasis models. In conclusion, our data reveal a tumor-suppressing role of circYAP1 and circYAP1/miR-135b-5p/FRK axis in osteosarcoma, demonstrating potential therapeutic targets for osteosarcoma.

Graphical abstract

Highlights

• •CircYAP1 is low expression in osteosarcoma (OS) vs normal cells and tissues
• •Low-expression circYAP1 indicates higher metastasis and advanced clinical stage
• •CircYAP1 could sponge miR-135b-5p
• •CircYAP1 suppresses OS via circYAP1/miR-135b-5p/FRK axis

Molecular biology; Cell biology; Cancer

Introduction

Osteosarcoma (OS) is the most common primary malignant bone tumor in adolescents and children.¹ In recent decades, little progress had been made in OS treatment. The 5-year survival rate is only 55%–65%, and the 2-year survival rate of patients with lung metastasis is less than 25%.^2,3 Treatment for osteosarcoma has entered a plateau period, and it is difficult to achieve breakthroughs with traditional treatment strategies. Therefore, further elucidation of the mechanisms of the occurrence, development, and metastasis of osteosarcoma and exploration of new therapeutic targets have become hot spots in this field.

Circular RNAs (circRNAs) are a class of endogenous noncoding RNAs (ncRNAs) and have been widely found in various human cell types.^4,5 Unlike linear RNA, circRNA is a kind of covalent closed cyclic molecule formed by reverse splicing and thus is more stable than linear RNA.⁶ CircRNAs are closely implicated in the occurrence and progression of various human diseases.^7,8,9 With more attention attracted by circRNA, dysregulated circRNAs have been extensively found in various types of cancers, functioning as tumor suppressors or oncogenes.¹⁰ For example, circZNF609 has_circ_0005075 were found to be upregulated and could promote the malignant phenotype of cancers,^11,12,13 whereas circSLC8A1, has_circ_0007534, and circITCH are downregulated in cancers and could suppress cancer progression by targeting downstream molecules.^14,15,16 It has now been realized that dysregulation of non-coding RNAs (ncRNAs), including microRNAs (miRNAs), long non-coding RNAs (lncRNAs), and the recently discovered circular RNAs, is crucial to the initiation and progression of osteosarcoma.¹⁷ As an oncogene, YAP1 is extensively involved in various malignancies, including gastric cancer and osteosarcoma.^18,19 However, circYAP1 (hsa_circ_0002320) was demonstrated to inhibit cell proliferation and invasion in gastric cancer, thus acting as a tumor suppressor, which is opposite to the oncogenic role of YAP1 in gastric cancer.²⁰ Moreover, it is still unclear whether circYAP1 functions as a tumor suppressor in osteosarcoma as it does in gastric cancer. In the present study, the expression and biological function, as well as the potential regulatory mechanism of circYAP1, were comprehensively explored in osteosarcoma to find a reliable therapeutic target for osteosarcoma.

Results

CircYAP1 downregulation indicates a high metastasis status and a poor prognosis in patients with osteosarcoma

As shown in Figure 1A, circYAP1 is located on chromosome 11q22.1 and is cyclized from exons 4 and 5 of YAP1. To evaluate the expression profile of circYAP1 in osteosarcoma, we performed quantitative analysis on 10 matched pairs of tumor tissues and adjacent noncancerous tissues by RT-qPCR. The circYAP1 expression level in osteosarcoma tissues was significantly reduced to 47% of that in normal (normal 3.01 ± 0.62 vs. osteosarcoma 1.42 ± 0.51) (Figure 1B). Furthermore, the expression of circYAP1 was subsequently detected in 86 formalin-fixed, paraffin-embedded human osteosarcoma tissue specimens by osteosarcoma in situ hybridization (ISH). circYAP1 expression was significantly lower in the metastatic specimens than in the nonmetastatic specimens (Figures 1C and 1D). Based on the ISH scores, all specimens were divided into high and low circYAP1 expression groups. The results showed that the low circYAP1 expression group had a higher metastasis rate and more advanced clinical stage than the high circYAP1 expression group (Table 1). Moreover, patients with low circYAP1 expression had a shorter overall survival time than those with high circYAP1 expression (Figure 1E).

Figure 1.

Low circYAP1 expression is positively correlated with metastasis status and a poor prognosis in osteosarcoma

(A) Schematic illustration shown that circYAP1 is generated from exon 4 and exon 5 of YAP1 by back-splicing.

(B) The expression of circYAP1 was measured in 10 paired osteosarcoma and adjacent normal tissue specimens by RT-qPCR. Paired Student’s t test.

(C) Representative images of circYAP1 FISH in osteosarcoma metastatic specimens and nonmetastatic specimens (200×).

(D) Statistical analysis of circYAP1 expression in both osteosarcoma metastatic specimens and nonmetastatic specimens. Student’s t test.

(E) Demonstration of the correlation between circYAP1 expression and the overall survival time of patients with osteosarcoma. Log rank test. ^∗∗p < 0.01.

Data are presented as the mean ± standard deviation (SD).

Table 1.

The relationship between circYAP1 expression and clinicopathological features of patients with osteosarcoma

Characteristics	n	CircYAP1 expression		p
		Low, n	High, n
Age (years)
≤18	58	25	33	0.8194a
>18	28	13	15
Gender
Male	54	23	31	0.8228a
Female	32	15	17
Tumor size (cm2)
≤8	47	16	31	0.0502a
>8	39	22	17
Clinical stage
IB-IIA	28	6	22	0.0050b
IIB/III	58	32	26
Distant metastasis
Yes	40	24	16	0.0088a
No	46	14	32

^ap values were calculated using Fisher’s exact test.

^bp value was calculated using the Mann-Whitney test.

CircYAP1 is expressed at lower levels in osteosarcoma and is mainly localized in the cytoplasm

Consistent with the results of clinical samples, the expression of circYAP1 was lower in osteosarcoma cells, especially in SOSP-9607 and Saos-2 cells, than in hFOB1.19 cells (Figure 2A). We detected and quantified circYAP1 by RT-PCR after RNase R digestion (Figure 2B). The subcellular localization of circYAP1 was also explored in SOSP-9607 and Saos-2 cells, and the results demonstrated that it was primarily localized in the cytoplasm (Figure 2C). The FISH results further verified the subcellular localization of circYAP1 in osteosarcoma cells (Figure 2D). Collectively, high circYAP1 expression mainly in the cytoplasm may be implicated in tumor progression.

Figure 2.

CircYAP1 expression is significantly reduced in osteosarcoma cells versus control cells and is primarily located in the cytoplasm

(A) Detection of the expression of circYAP1 in various osteosarcoma cell lines by RT-qPCR. One-way analysis of variance (ANOVA) and Dunnett’s multiple comparisons test.

(B) Both circYAP1 and YAP1 mRNA expression was evaluated after treatment with or without RNase R by RT-qPCR. Student’s t test.

(C and D) Evaluation of the subcellular localization of circYAP1 in osteosarcoma cells by RT-qPCR and FISH analysis. Student’s t test. ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001.

Data are presented as the mean ± standard deviation (SD).

CircYAP1 inhibited osteosarcoma cell proliferation, invasion, and migration in vitro

To further elucidate the biological function of circYAP1 in osteosarcoma cells, lentiviral vector-mediated transduction was performed to upregulate the expression of circYAP1. The transduction efficiency was validated by RT-qPCR (Figure 3A). Then, the cell proliferation, invasion, and migration abilities of osteosarcoma were detected by CCK-8, colony formation, and Transwell assays. Subsequent CCK-8 assays demonstrated that increased circYAP1 significantly repressed cell proliferation in SOSP-9607 and Saos-2 cells (Figures 3B and 3C). The same results were also observed in the colony formation assay (Figures 3D and 3E). As shown in Figures 3F–3I, reductions in cell invasion and migration were observed after circYAP1 upregulation in the Transwell assay. Taken together, our data demonstrate that circYAP1 has a suppressive effect on osteosarcoma cells.

Figure 3.

CircYAP1 upregulation suppresses the malignant behavior of osteosarcoma in vitro

(A) The transduction efficiency of circYAP1 was validated by RT-qPCR. Student’s t test.

(B–E) The proliferation rate of osteosarcoma cells was evaluated by CCK-8 and colony formation assays. two-way ANOVA and Dunnett’s multiple comparisons test were used for the CCK-8 assay, and Student’s t test was used for the colony formation assay.

(F and G) Cell invasion was evaluated by Transwell invasion assay. Student’s t test.

(H and I) Cell migration was evaluated by Transwell migration assay. Student’s t test. ^∗∗p < 0.01 and ^∗∗∗p < 0.001.

Data are presented as the mean ± standard deviation (SD).

CircYAP1 serves as a sponge for miR-135b-5p

As an important participant of the competing endogenous RNA (ceRNA) network, most circRNAs have been proven to function as miRNA sponges. According to the subcellular localization of circYAP1 in osteosarcoma cells, it is reasonable to speculate that circYAP1 may exert its functions by sponging downstream miRNAs. As expected, circYAP1 was significantly enriched in the AGO2 group in osteosarcoma cells, suggesting that circYAP1 may function by sponging miRNAs (Figure 4A). According to bioinformatic analysis through circinteractome (circinteractome.nia.nih.gov) and circBank (www.circbank.cn), seven oncogenic miRNAs previously reported in the literature were chosen as potential targets for further verification. RNA pull-down assays showed that only miR-135b-5p was pulled down by circYAP1 in both SOSP-9607 and Saos-2 cells (Figures 4B and 4C), and the potential binding site between circYAP1 and miR-135b-5p is also shown in Figure 4D. In the RNA immunoprecipitation (RIP) assay, both circYAP1 and miR-135b-5p were enriched in the AGO2 group relative to the control group (Figure 4E). Subsequent dual-luciferase reporter gene assays also showed that miR-135b-5p suppressed the activity of the wild-type vector of circYAP1 but not the mutant-type vector (Figure 4F). Collectively, our findings show that circYAP1 acts as an efficient miR-135b-5p sponge.

Figure 4.

CircYAP1 sponges miR-135b-5p in osteosarcoma cells

(A) The possible mechanism of circYAP1 was investigated using an RIP assay. Student’s t test.

(B and C) The regulatory mechanism of circYAP1 in SOSP-9607 and Saos-2 cells was investigated using a circRNA pull-down assay. Student’s t test.

(D) Prediction of binding sites between circYAP1 and miR-135b-5p.

(E) Validation of the potential interaction between circYAP1 and miR-135b-5p by RIP assay. Student’s t test.

(F) The relationship between circYAP1 and miR-135b-5p was demonstrated by a dual-luciferase reporter gene assay. Student’s t test.

Data are presented as the mean ± standard deviation (SD).

miR-135b-5p can reverse circYAP1-induced antitumor effects in osteosarcoma

In contrast to circYAP1, miR-135b-5p was demonstrated to be highly expressed in osteosarcoma (Figures 5A and 5B). In addition, the decreased cell proliferation, invasion, and migration caused by circYAP1 upregulation was markedly rescued by increased miR-135b-5p in both SOSP-9607 and Saos-2 cells (Figures 5C–5J). These data show that circYAP1 might regulate osteosarcoma progression by targeting miR-135b-5p.

Figure 5.

Upregulation of miR-135b-5p attenuates the inhibitory effect exerted by circYAP1 in osteosarcoma cells

(A) Assessment of the expression of miR-135b-5p in paired osteosarcoma and adjacent normal tissue specimens. Student’s t test.

(B) Assessment of the expression of miR-135b-5p in osteosarcoma cells by RT-qPCR. One-way ANOVA and Dunnett’s multiple comparisons test.

(C–F) The effects of both circYAP1 and miR-135b-5p on cell proliferation were determined using CCK-8 and colony formation assays. Two-way ANOVA and Dunnett’s multiple comparisons test were used for the CCK-8 assay, and one-way ANOVA and Dunnett’s multiple comparisons tests were used for the colony formation assay.

(G and H) The effects of both circYAP1 and miR-135b-5p on cell invasion were determined using a Transwell invasion assay. One-way ANOVA and Dunnett’s multiple comparisons test.

(I and J) Effects of both circYAP1 and miR-135b-5p on cell migration were assessed using a Transwell migration assay. One-way ANOVA and Dunnett’s multiple comparisons test. ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001.

Data are presented as the mean ± standard deviation (SD).

CircYAP1 inhibits osteosarcoma progression through the miR-135b-5p/FRK axis

Based on the predictions of TargetScan, PicTar, and miRDB, a series of 17 potential targets of miR-135b-5p were selected for further confirmation (Figure 6A). As a candidate target gene, the sites of interaction between FRK and miR-135b-5p are presented in Figure 6B. As expected, the expression levels of FRK mRNA and protein were also markedly downregulated after enhancing miR-135b-5p expression in osteosarcoma cells (Figures 6C and 6D). In addition, miR-135b-5p significantly reduced the luciferase activity of the vector containing the wild-type 3′UTR of FRK, instead of the mutant-type 3′ UTR in the dual-luciferase reporter gene assay (Figure 6E). Thus, we confirmed that FRK is a direct target of miR-135b-5p. Moreover, the expression of FRK was markedly rescued by circYAP1 overexpression after upregulation of miR-135b-5p in SOSP-9607 and Saos-2 cells (Figure 6F). The expression of FRK was downregulated in osteosarcoma tissues and cell lines (Figures S1A and S1B), and the antitumor effects of circYAP1 were abrogated by knockdown of FRK in subsequent CCK-8, colony formation, and Transwell assays (Figures S1C–S1J). Taken together, our data show that circYAP1 inhibits the malignant phenotype of osteosarcoma cells through the miR-135b-5p/FRK axis.

Figure 6.

CircYAP1 functions as a tumor suppressor through the miR-135b-5p/FRK axis

(A) Schematic illustration shows the overlapping potential targets of miR-135b-5p predicted by TargetScan, PicTar, and miRDB.

(B) Presentation of the direct interaction between miR-135b-5p and the 3′UTR of FRK mRNA.

(C and D) Changes in FRK mRNA and protein expression were measured by RT-qPCR and western blotting. Student’s t test.

(E) Effects of miR-135b-5p on the luciferase activity of osteosarcoma cells after transfection with the wild-type or mutant FRK 3′UTR reporter gene. Student’s t test.

(F) Detection of the effects of both circYAP1 and miR-135b-5p on the expression of FRK protein. ^∗∗∗p < 0.001.

Data are presented as the mean ± standard deviation (SD).

CircYAP1 inhibits osteosarcoma tumor growth and lung metastasis in vivo

We validated whether circYAP1 suppresses osteosarcoma tumor growth and lung metastasis in vivo. SOSP-9607 cells stably overexpressing circYAP1 and negative control (NC) were injected into the proximal tibia to establish an orthotopic lung metastasis model of osteosarcoma. Six weeks after injection, the mice were sacrificed, and statistical analysis was carried out. Both the tumor volume and weight in the circYAP1-overexpressing group were lower than those in the NC group (Figures 7A–7C). As shown in Figures 7D and 7E, compared with mice in the NC group, the mice in the circYAP1-overexpressing group also showed decreased luciferase activity in lung metastases. A reduction in lung weight was also observed in the circYAP1-overexpressing group compared to the NC group (Figure 7F). In addition, the number of microscopic lung metastases in the circYAP1-overexpressing group was markedly lower than that in the NC group (Figures 7G and 7H).

Figure 7.

CircYAP1 inhibits osteosarcoma growth and metastasis in vivo

(A) Image of the primary tumors of the two groups.

(B) Statistical analysis of the volume of primary tumors in the two groups. Student’s t test.

(C) Statistical analysis of the weight of primary tumors in the two groups. Student’s t test.

(D) Representative images of lung metastasis detected using an in vivo imaging system (IVIS).

(E) Quantification of bioluminescence intensity. Student’s t test.

(F) Statistical analysis of wet lung weight in the two groups. Student’s t test.

(G) Representative H&E staining of lung metastatic nodules (100×).

(H) Statistical analysis of the number of metastatic lung nodules in the two groups. Mann-Whitney test. ^∗p < 0.05 and ^∗∗p < 0.01. Data are presented as the mean ± standard deviation (SD).

Discussion

Accumulating evidence indicates that dysregulated circRNAs function as oncogenic drivers or tumor suppressors across diverse malignancies—a paradigm exemplified by circ-CTNNB1 in osteosarcoma¹³ and circITCH in bladder cancer.¹⁶ The findings of this study not only underscore the tumor-suppressive role of circYAP1 in osteosarcoma but also highlight the potential of the circYAP1/miR-135b-5p/FRK axis as a therapeutic target. The downregulation of circYAP1 in osteosarcoma tissues and its correlation with poor prognosis and advanced clinical stages suggest that circYAP1 could serve as a prognostic biomarker. In vitro and in vivo experiments demonstrate that circYAP1 exerts its tumor-suppressive effects by inhibiting cell proliferation, invasion, and migration and by reducing tumor growth and metastasis. These findings align with emerging roles of tumor-suppressive circRNAs in modulating cancer progression,^10,17 yet uniquely implicate circYAP1 as a central regulator in osteosarcoma.

The intricate interactions occur among different RNA species, including protein-coding messenger RNAs and non-coding RNAs such as lncRNAs, pseudogenes, and circRNAs. CircRNA can regulate the downstream genes through sponge adsorption of miRNA, which is a key regulatory node in tumor progression.^13,21,22 In this study, we predicted and validated a direct interaction between circYAP1 and miR-135b-5p by RIP assay and dual-luciferase reporter gene assay. In previous studies, miR-135b-5p has been associated with various cancers.^23,24,25 MiR-135b-5p was found to be overexpressed in highly aggressive osteosarcoma cell lines, suggesting that it was closely related to the metastatic and proliferative capacity of osteosarcoma.²⁶ Our results also confirmed an oncogenic role of miR-135b-5p in osteosarcoma. In addition, we demonstrated that miR-135b-5p could serve as a target of circYAP1 and attenuate the inhibitory effects of circYAP1 on the proliferation, invasion, and migration of osteosarcoma cells. Taken together, our data demonstrated that circYAP1 enhances the malignant behavior of osteosarcoma cells by sponging miR-135b-5p. The identification of miR-135b-5p as a direct target of circYAP1 and its role in reversing the antitumor effects of circYAP1 provide a mechanistic insight into the regulatory network involving circRNAs and miRNAs in osteosarcoma.

FRK (originally called RAK) has been confirmed as a tumor suppressor in various cancers, such as glioma, pancreatic cancer, and breast cancer.^27,28,29 FRK could inhibit the proliferation and colony formation of U2OS cells by regulating the stability and function of PTEN protein.³⁰ We predicted and confirmed that miR-135b-5p could directly target FRK. We found that FRK was expressed at low levels in osteosarcoma cell lines and tissues. The proliferation, invasion, and migration abilities of osteosarcoma cells were significantly enhanced after inhibiting FRK. As a target of miR-135b-5p, the expression of FRK could be inhibited by miR-135b-5p. More importantly, the inhibitory effects exerted by miR-135b-5p on the expression of FRK could be attenuated by circYAP1. In addition, the antitumor effects of circYAP1 could be abrogated by knockdown of FRK. These results indicated that circYAP1 inhibited osteosarcoma progression through the miR-135b-5p/FRK axis.

Our study has demonstrated that circYAP1 acted as a tumor suppressor in osteosarcoma by inhibiting cell proliferation, invasion, and migration through the miR-135b-5p/FRK axis. The downregulation of circYAP1 in osteosarcoma tissues is associated with poor prognosis and advanced clinical stages, highlighting its potential as a prognostic biomarker.^31,32 Liquid biopsy is a minimally invasive approach that does not require surgery; thus, it is considered safe and still efficient. Circulating tumor cells (CTCs), ctDNAs, circulating tumor microRNAs (ctmiRNAs), lncRNAs, and exosomes released into the patient’s blood can all be detected using liquid biopsy.³³ In particular, the stable circular structure of circRNAs lengthens their half-life, especially in cell-free samples (such as blood and urine), creating potential for the use of circRNAs as biomarkers in patient samples from noninvasive sources.^31,32,33,34 We found that CircYAP1 was significantly downregulated in osteosarcoma tissues and had even lower expression in metastatic samples. If its expression level can be detected in patients’ plasma or exosomes and is related to the expression level in tumor tissues, it may serve as a non-invasive diagnostic or prognostic marker. However, there is no direct evidence to support that circYAP1 is suitable for liquid biopsy at present. In the future, we will consider conducting more in-depth basic and clinical research to verify its detectability in body fluids and the feasibility of it as a liquid biopsy marker.

The identification of the circYAP1/miR-135b-5p/FRK axis provided a reliable regulatory mechanism and revealed potential therapeutic targets in osteosarcoma. Restoring circYAP1 expression through delivery of circular RNA vectors or lipid nanoparticles might represent a feasible therapeutic approach.³⁵ Similar approaches have shown preliminary success in cancer treatment with other circRNAs, such as circRNA mSCAR and interleukin-12 (IL-12) circRNA.^36,37 Existing evidences indicate that dysregulation of ceRNA networks can influence tumor cell responses to therapy.^38,39,40 Given that circYAP1 upregulates FRK by antagonizing miR-135b-5p, FRK has been proven to stabilize PTEN and inhibit the PI3K/AKT pathway.³⁰ Therefore, the restoration of circYAP1 expression may enhance the efficacy of PI3K inhibitors (such as LY294002) or mTOR inhibitors (such as rapamycin) or weaken PTEN-mediated cell chemotherapy resistance.^41,42 Although we did not directly analyze the relationship between circYAP1 expression and treatment response, the role of circYAP1 in chemotherapy resistance in osteosarcoma deserves further investigation and exploration in preclinical models.

Conclusion

In brief, we found that circYAP1 is significantly downregulated in osteosarcoma cells and tissues. Functionally and mechanistically, circYAP1 inhibits osteosarcoma cell proliferation, invasion, and migration by sponging miR-135b-5p to upregulate FRK. We validated the mechanism of circYAP1/miR-135b-5p/FRK axis in the malignant progression and metastasis of osteosarcoma in vivo and in vitro, providing new ideas for osteosarcoma diagnosis and treatment development.

Limitations of the study

The relatively small cohort of paired clinical samples may limit the statistical power to generalize our findings. While we mechanistically confirmed circYAP1 to the miR-135b-5p/FRK axis, the absence of multi-omics integration precludes a comprehensive understanding of circYAP1’s broader regulatory network.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Kang Han (gan_7758525@163.com).

Materials availability

Plasmids and/or cell lines generated in this study are available upon reasonable request. Please contact the lead contact.

Data and code availability

• •All data reported in this article will be shared by the lead contact upon request.
• •This article does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

This work was supported by the National Science Foundation for Young Scientists of China (81702935), the Natural Science Foundation of Shandong Provincial (ZR2023MH206), and the Project of Medical and Health Science and Technology Development Program of Shandong province (202204071065).

Author contributions

Conceptualization, K.H.; data curation, S.T., M.C., and C.C.; resources, M.D., X.Z., and M.S.; software, M.D., L.S., and Z.Z.; methodology, Q.Z., L.S., and Z.Z.; validation, M.S., Q.Z., and H.X.; visualization, M.D., L.Y., and X.Z.; writing–original draft, S.T., M.C., Q.G., and K.H.; writing–review & editing, M.D., X.Z., Q.Z., and K.H..

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-beta Actin antibody	Abcam	Cat# ab8227; RRID: AB_2305186
Anti-FRK antibody	Abcam	Cat# ab64914; RRID: AB_1140720
HRP-conjugated Goat Anti-Rabbit IgG(H + L)	Proteintech	Cat# SA00001-2; RRID: AB_2722564
Anti-Ago 2 antibody	Proteintech	Cat# 67934-1-Ig; RRID: AB_2918686
Digoxigenin Recombinant Rabbit Monoclonal Antibody	Invitrogen	Cat# 700772; RRID: AB_2532342
Biological samples
Osteosarcoma tissues	Tangdu Hospital	N/A
Chemicals, peptides, and recombinant proteins
Cell Counting Kit-8	Dojindo Laboratories	Cat# CK04
Lipofectamine 3000	Invitrogen	Cat# L3000075
Anti-fluorescence quenching sealing solution (with DAPI)	Beyotime	Cat# P0131-5mL
TRIzol Reagent	Invitrogen	Cat# 1559606
RIPA buffer	Solarbio	Cat# R0010
Protease inhibitor	Solarbio	Cat# R6730
Immobilon® Crescendo Western HRP substrate	Millipore	Cat# WBLUR0100
Matrigel	Servicebio	Cat# G4130-5ML
DAB	Servicebio	Cat# G1212-200T
Dulbecco’s modified Eagle’s medium (DMEM)	Hyclone	Cat# SH30243.01
RPMI-1640 medium	Gibco	Cat# 11875093
fetal bovine serum	Gibco	Cat# A5670701
Critical commercial assays
RNA pulldown kit	BersinBio	Cat# Bes5102
Hieff qPCR SYBR Green Master Mix kit	Yeasen Biotechnology	Cat# 1201ES08
Immunoprecipitation kit	Abcam	Cat# ab206996
Magna RIP RNA-Binding Protein Immunoprecipitation Kit	Millipore	Cat# 17-701
Nuclear/Cytosol Fractionation Kit	Abcam	Cat# ab289882
Fluorescent In Situ Hybridization Kit	Ribobio	Cat# C10910
miDETECT A TRACK™ miRNA qRT‒PCR Starter kit	Ribobio	Cat# C10712-3
Luc-Pair™ Duo-Luciferase Assay Kit	GeneCopoeia	Cat# LF004
Experimental models: Cell lines
Saos-2	Shanghai Cell Bank of Chinese Academy of Sciences	Cat# SCSP-5057
hFOB 1.19	Shanghai Cell Bank of Chinese Academy of Sciences	Cat# GNHu14
143B	Shanghai Cell Bank of Chinese Academy of Sciences	Cat# TCHu264
U2OS	Shanghai Cell Bank of Chinese Academy of Sciences	Cat# TCHu88
SOSP-9607	Bone Tumor Research Institute of Tangdu Hospital	N/A
Experimental models: Organisms/strains
BALB/c nude mice	Gempharmatech	Cat# NO. D000521
Oligonucleotides
YAP1 F: CACAGCTCAGCATCTTCGAC R: TATTCTGCTGCACTGGTGGA	This paper	N/A
circYAP1 F:ACAGATGCGACTGCAGCAAC R:TGGGTCTAGCCAAGAGGTGG	This paper	N/A
GAPDH F:AATCCCATCACCATCTTCCA R:GGGACTCCACGACGTACTCA	This paper	N/A
FRK F: CTCTGGGAGTACCTAGAACCC R:AGCCTGGTAATCAAACAAAGCC	This paper	N/A
CircYAP FISH probe: mTmTCmAmTGGCmTGm TmAmTCCmAmTCmTCmAmTCCmA	This paper	N/A
CircYAP ISH probe: GATTCTCTGGTTCATGGCTGTA TCCATCTCATCC(tttCATCATCATACATCATCAT)	Servicebio	N/A
Negative Control F: UUCUCCGAACGUGUCACGUTT R: ACGUGACACGUUCGGAGAATT	This paper	N/A
U6 F:CTCGCTTCGGCAGCACATATACT R:ACGCTTCACGAATTTGCGTGTC	This paper	N/A
miR-135b-5p F: CGCGTATGGCTTTTCATTCCT R: AGTGCAGG GTCCGAGGTATT	GenePharma	N/A
miR-135b-5p mimic	MedChemExpress	Cat# HY-R00261
MicroRNA Mimic Negative Control	MedChemExpress	Cat# HY-R04602
Si-FRK GGAGUACCUAGAACCCUAUTT	This paper	N/A
Recombinant DNA
psiCHECK™-2 Vector	Promega	Cat# C8021
pcDNA3.1-circYAP1	Hanbio	N/A
pcDNA3.1	Hanbio	N/A
Software and algorithms
SPSS 22.0 software	IBM	N/A
GraphPad Prism 7.0 software	GraphPad	N/A
Image-Pro Plus (version 6.0)	Media Cybernetic	N/A
Quantity One system	Bio-Rad	N/A
circinteractome	circinteractome	circinteractome.nia.nih.gov
circBank	circBank	www.circbank.cn

Experimental model and study participant details

Patients and tissue samples

The experimental protocol was approved by the Institutional Review Board of Tangdu Hospital, and informed consent was obtained from all participants. Eighty-six formalin-fixed, paraffin-embedded human osteosarcoma tissue specimens and 10 paired osteosarcoma and adjacent normal tissue specimens were obtained from patients diagnosed with osteosarcoma based on histopathological evaluation at Tangdu Hospital. None of the patients received any therapy before biopsy.

Cell lines and cell culture

The human osteosarcoma cell line SOSP-9607 was established and preserved in our laboratory.⁴³ Human osteosarcoma cell lines, including 143B, U2OS and Saos-2, and the human osteoblast cell line hFOB1.19 were purchased from the Shanghai Cell Bank of the Chinese Academy of Sciences. U2OS and Saos-2 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Hyclone, USA) supplemented with 10% fetal bovine serum (Gibco, USA)). SOSP-9607 cells were maintained in RPMI-1640 medium supplemented with 10% FBS (Gibco, USA). 143B cells were cultured in MEM medium (Gibco, USA) supplemented with 10% FBS, and hFOB1.19 cells were cultured with DMEM/F12 medium supplemented with 10% FBS. The OS cells were cultured with 5% carbon dioxide at 37°C, hFOB1.19 cells were cultured with 5% carbon dioxide at 33.5°C These cell lines have undergone authentication and have been confirmed to be free of mycoplasma contamination.

Tumor xenograft model

All animal experiments complied with ethical regulations and were approved by the Medical Ethics Committee of the Air Force Medical University. 6–8 weeks old male BALB/c nude mice were purchased from Gempharmatech (Cat# NO. D000521) and were housed in the specific pathogen-free (SPF) animal facilities in a climate controlled clean room with humidity range of 40–70% and temperature range of 20°C–26°C, with a 12h light/dark cycle and fed with regular chow and water by the facility staff. A total of 3×10⁶ SOSP-9607 cells overexpressing circYAP1 or NC were injected into the intramedullary region of the tibia to establish an orthotopic lung metastasis model. Tumor growth was allowed for 6 weeks, and lung metastasis was detected using an in vivo imaging system. Then, all the mice were sacrificed, and tumor volume, tumor weight, the number of lung metastatic nodules and wet lung weight were calculated in the two groups.

Method details

Lentiviral vector-mediated stable transduction and transient transfection

The lentivirus particles containing the circYAP1 overexpression vector or empty vector were purchased from Hanbio (Shanghai, China). After transduction for 48 h, the efficiency was initially validated according to GFP expression under a fluorescence microscope, and then SOSP-9607 and Saos-2 cells were transduced at an MOI of 50. The cells were selected using puromycin at a concentration of 2 μg/mL. miR-135b-5p mimic and negative control (NC) were purchased from MedChemExpress. Transient transfection was carried out with Lipofectamine 3000 (Invitrogen, USA) according to the instructions.

Subcellular location and fluorescence in situ hybridization (FISH)

The Nuclear/Cytosol Fractionation Kit (Abcam, USA) was used to isolate nuclear and cytoplasmic RNA. Experiments were conducted according to the instructions with GAPDH adopted as an internal reference. A FAM-labeled probe for circYAP1 was synthesized by GenePharma (China), and the FISH assay was performed using an RNA FISH Kit (Ribobio) according to the instructions. Fluorescence images were captured using an FV1000 laser scanning confocal microscope (Olympus, Japan).

RNA extraction, RNase R treatment and qRT‒PCR

For the RNase R treatment assay, 2 mg RNA was treated with or without 3 U/mg RNase R for 15 min at 37°C (Epicenter Biotechnologies, USA). For circRNA or mRNA analysis, the RNA extracts above were reverse transcribed using Hieff qPCR SYBR Green Master Mix kit (Yeasen Biotechnology, China). For miRNA detection, the miDETECT A TRACK miRNA qRT‒PCR Starter kit (Ribobio, China) was used for reverse transcription and PCR quantification. The relative expression level was determined using the 2^−ΔΔCt formula. GAPDH and U6 were used as internal references for circRNA/mRNA and miRNA analysis, respectively. Primers for U6, miR-135b-5p and other miRNAs were purchased from RiboBio (China). Sequences of the other primers are shown in key resources table.

CCK-8 and colony formation assays

For the CCK-8 assay, 3000 cells were plated into a well of a 96-well plate. On the indicated day, 10 μL of the CCK-8 solution (Dojindo, Japan) was added to the culture medium of each well. The cells were then incubated at 37°C for 1 h, and the optical density (OD) value at 450 nm was detected. For the colony formation assay, 1000 cells were seeded into a well of a 6-well plate and then cultured for 14 days. After fixation with 4% paraformaldehyde for 20 min, the cells were stained with 0.5% crystal violet for 15 min. The number of colonies was determined by Image-Pro Plus (version 6.0).

Western blotting

Total protein from SOSP-9607 and Saos-2 cells was extracted using RIPA buffer plus protease inhibitor (Solarbio). Equal amounts of protein were loaded and subjected to 10% SDS‒PAGE and transferred onto a polyvinylidene fluoride (PVDF) membrane. The protein bands were incubated with anti-FRK (1:1000 dilution, Abcam, USA) and anti-β-actin (1:1000 dilution, Proteintech, USA) primary antibodies (incubation overnight at 4°C) and HRP-labeled goat anti-mouse (1:5000 dilution, Proteintech) secondary antibodies. The bands were visualized using Immobilon Western Chemilum HRP Substrate (Millipore, Germany) and analyzed by Quantity One system (Bio-Rad, USA).

Biotin-labeled circRNA pull-down assay

Biotin-labeled circRNA pull-down assays were performed using an RNA pulldown Kit (BersinBio, Guangzhou, China) according to the instruction manual. The biotin-labeled circRNA probe or control probe was incubated with blocked streptavidin magnetic beads at 25°C for 30 min. A total of 1×10⁷ SOSP-9607 or Saos-2 cells were collected and lysed using RIP buffer with protease inhibitor. Then, the lysates were incubated with the above magnetic bead-probe mixture at 25°C for 2 h. The RNA complex was eluted and isolated with TRIzol (Invitrogen), followed by qRT‒PCR analysis. The biotin-labeled probes were synthesized and purchased from RiboBio (China).

RIP assay

RIP assays were conducted using a Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore). A total of 1×10⁷ cells were collected and lysed using complete RIP lysis buffer. Five micrograms of anti-AGO2 antibody (Proteintech) and lgG were added to the magnetic bead suspension for 30 min at room temperature. The lysates above were incubated with bead-antibody complexes overnight at 4°C. After treatment with Protease K Buffer at 55°C for 30 min, the immunoprecipitated RNA was isolated with TRIzol reagent (Invitrogen) and analyzed using qRT‒PCR.

Transwell assay

Transwell chambers (Corning, USA) were used for the migration assay, and chambers precoated with Matrigel (Servicebio, China) were used for the invasion assay. After transfection for 48 h, SOSP-9607 and Saos-2 cells were collected. A total of 8×10⁴ cells were added to each upper chamber with 200 μL serum-free medium, and 600 μL complete medium was added to each lower chamber. After incubation for 20 h (migration assay) or 30 h (invasion assay), the cells were fixed, stained and analyzed under a microscope (Olympus, Japan).

Dual-luciferase reporter gene assay

The psiCHECK2 plasmid (Promega, USA) and Luc-Pair Duo-Luciferase Assay Kit (GeneCopoeia, USA) were used for this assay. After amplification, the sequences of circYAP1-WT, circYAP1-Mut, FRK-3′UTR-WT and FRK-3′UTR-Mut were inserted into plasmids between the Notl and Xhol sites. After cells were seeded into each well of a 96-well plate, the miR-135b-5p mimic and plasmids were transfected into SOSP-9607 and Saos-2 cells. After incubation for 48 h, the cells were lysed, and firefly luciferase activity was detected and normalized to Renilla luciferase activity.

Immunohistochemistry

The probe for circYAP1 ISH was designed and synthesized by GenePharma (Guangdong, China). The detection was performed with anti-digoxigenin-HRP (anti-DIG-HRP) antibody (Invitrogen) and DAB (Servicebio) reagent. Immunoreactivity scores were acquired according to the intensity and proportion of positive cells.⁴⁴ A score of 0–4 was considered to indicate low circYAP1 expression, while a score of 5–12 was considered to indicate high circYAP1 expression.

Quantification and statistical analysis

All statistical analyses were carried out using SPSS 22.0 software (IBM, SPSS, USA) and GraphPad Prism 7.0 software (GraphPad, USA). Statistical methods are shown in the figure legends, and data are presented as the mean ± standard deviation (SD), unless otherwise specified. p < 0.05 was considered statistically significant.

Published: June 9, 2025