circKIF4A sponges miR-127 to promote ovarian cancer progression

Knockdown of circKIF4A inhibits ovarian cancer proliferation and invasion in vitro

To explore the expression profile of circKIF4A in ovarian cancer, we collected 50 pair of ovarian cancer tissues and adjacent normal tissues and performed qRT-PCR. We found that circKIF4A was significantly upregulated in ovarian cancer tissues (Figure 1A). To explore the function of circKIF4A in ovarian cancer, we used siRNA to knock down the expression of circKIF4A (Figure 1B). CCK-8 assay revealed that circKIF4A knockdown significantly inhibited cell proliferation (Figure 1C). BrDU assay also showed that circKIF4A knockdown reduced proliferation ability of ovarian cancer cells (Figure 1D). Transwell assay showed that cell invasion was significantly suppressed after downregulation of circKIF4A (Figure 1E).

Figure 1. Knockdown of circKIF4A inhibits ovarian cancer proliferation and invasion in vitro. (A) The expression of circKIF4A in ovarian cancer tissues (Tumor) and normal adjacent tissues (Normal). (B) circKIF4A was successfully knocked down by siRNA. (C) CCK-8 assay was performed to detect cell proliferation. (D) BrDU immunostaining of ovarian cancer cells transfected with si-NC of si-circKIF4A. (E) Transwell assay was performed to assess cell invasive ability (left) and the number of invasive cells was quantified by ImageJ (right). **P < 0.01

Knockdown of circKIF4A inhibits ovarian cancer growth and metastasis in vivo

To further explore the function of circKIF4A in ovarian cancer in vivo, mouse xenograft models were established. circKIF4A inhibition significantly decreased xenograft tumor growth (Figure 2A and 2B) and the expression intensity of Ki-67 in cancer cell (Figure 2C). Lung metastasis was also suppressed by circKIF4A knockdown (Figure 2D), indicating that the knockdown of circKIF4A inhibits ovarian cancer growth and metastasis in vivo.

Figure 2. Knockdown of circKIF4A inhibits ovarian cancer growth and metastasis in vivo. (A) The growth curves of tumors were shown. (B) Representative images of xenograft tumors (left) and the tumor weight was summarized (right). (C) IHC staining of Ki-67 in ovarian cancer cells transfected with si-NC of si-circKIF4A. (D) Representative images of lung metastatic nodules on HE-stained sections (left) and the number of metastatic nodules was quantified (right). *P < 0.05, **P < 0.01

circKIF4A acts as a sponge for miR-127

We confirmed the intracellular location of circKIF4A and found it mainly localized in cytoplasm (Figure 3A). Then, we used Circular RNA Interactome (https://circinteractome.nia.nih.gov/index.html) to predict the potential circRNA/miRNA interaction. We found binding sites for miR-127 in circKIF4A sequence (Figure 3B). Luciferase reporter assay revealed that the luciferase intensity decreased after co-transfection of wild type luciferase reporter and miR-127 mimics, while the mutated luciferase reporter had no such effect (Figure 3C). To confirm the direct binding of circKIF4A and miR-127, RIP assay was performed and we found miR-127 predominantly enriched in MS2bs- circKIF4A group (Figure 3D), indicating that circKIF4A directly interacts with miR-127 and could act as a sponge for miR-127.

Figure 3. circKIF4A acts as a sponge for miR-127. (A) The expression levels of nuclear control (18S), cytoplasmic control (GAPDH) and circKIF4A were detected. (B) The predicted binding sites of miR-127 within the circKIF4A sequence. (C) Luciferase assay of cells co-transfected with miR-127 mimics and wild type or mutant luciferase reporter. (D) MS2-based RIP assay transfected with MS2bs-circKIF4A, MS2bs-circKIF4Amt or control. **P < 0.01.

circKIF4A acts as a ceRNA to regulate JAM3

To explore whether circKIF4A sponges miR-127 to regulate the expression of its downstream target, we searched TargetScan for target genes of miR-127 and JAM3 was predicted (Figure 4A). Luciferase reporter assay showed decreased luciferase intensity after co-transfection of miR-127 mimics and wild type luciferase reporter (Figure 4B). Moreover, co-transfection of wild type luciferase reporter and miR-127 inhibitor increased luciferase intensity (Figure 4B). And miR-127 suppressed JAM3 expression, while miR-127 inhibitor increased JAM3 expression (Figure 4C), indicating that JAM3 is a downstream target of miR-127 and is regulated by miR-127.

Figure 4. circKIF4A acts as a ceRNA to regulate JAM3. (A) The predicted binding sites of miR-127 within the JAM3 3’UTR. (B) Cells were transfected and luciferase assay was performed. (C) Cells were transfected and JAM3 expression was detected by qRT-PCR (left) and western blotting (right). (D) RIP assay showing the enrichment of circKIF4A, JAM3 and miR-127 on Ago2. (E) Cells were transfected and a RIP assay on Ago2 was performed. (F) Cells were transfected and JAM3 expression was detected by qRT-PCR (up) and western blotting (below). **P < 0.01.

RIP assay on Ago2 further revealed that circKIF4A, JAM3 and miR-127 were mainly enriched to Ago2 (Figure 4D), indicating that circKIF4A and JAM3 are recruited to an Ago2-related RISC where they interact with miR-127. In addition, knockdown of circKIF4A decreased the enrichment of circKIF4A to Ago2, while increasing the enrichment of JAM3 to Ago2 (Figure 4E), indicating that circKIF4A could function as a ceRNA and compete with JAM3 to bind miRNAs. Furthermore, circKIF4A knockdown led to decreased JAM3 expression, while transfection with a miR-127 inhibitor reversed this decrease (Figure 4F), indicating that circKIF4A sponges miR-127 to increase JAM3 expression.

Ethical standards

This study was approved by the Ethics Committee of The First Affiliated Hospital of University of South China and conducted according to the Helsinki Declaration. Informed consents were obtained from all patients. Animal study was approved and conducted according to the Institutional Animal Care and Use Committee (IACUC) of The First Affiliated Hospital of University of South China.

Clinical samples

50 pair of ovarian cancer tissues (Tumor) who were diagnosed by histopathology and did not receive any chemotherapy and radiation therapy before, and adjacent normal tissues (Normal) were collected from The First Affiliated Hospital of University of South China and immediately stored in RNAlater (Ambion, USA) for qRT-PCR analysis.

Cell culture and transfection

Ovarian cancer cell lines CAOV3 and SKOV3 were obtained from American Type Culture Collection (USA) and cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum. Cell lines were confirmed free of mycoplasma infection and cell authenticity was verified by DNA fingerprinting.

Cells were transfected with Lipofectamine 2000 (Invitrogen, USA). siRNAs targeting circKIF4A were synthesized by GenePharma (China), the sequences are presented in Supplementary Table 1. miR-127 mimic and inhibitor were purchased from GeneCopoeia (USA).

Quantitative real-time PCR (qRT-PCR)

Total RNA was isolated by TRIzol (Invitrogen) and the nuclear and cytoplasmic fractions were isolated by NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific). qRT-PCR was conducted with SYBR Premix Ex Taq™ (Takara, Japan) and an All-in-One™ miRNA qRT-PCR Detection Kit (GeneCopoeia) using Bio-Rad IQTM5 Multicolor Real-Time PCR Detection System (USA). The primers for qRT-PCR were purchased from Invitrogen (Supplementary Table 1).

Cell counting kit-8 (CCK-8) assay

Cells (1 × 10³) were seeded and CCK-8 solution (Dojindo Laboratories, Japan) was added. After 2h of incubation at 37°C, the absorbance at 450nM was measured by microtiter plate reader (Bio-Tek EPOCH2, USA).

Bromodeoxyuridine (BrDU) assay

Cell proliferation was also determined by measuring incorporation of the thymidine analogue BrDU into host DNA. Cells (1×10⁴) were seeded and exposed to an acute pulse of BrDU (10mg/ml) for 2h. Cells were then fixed with cold 4% paraformaldehyde and visualized for the presence of BrDU incorporation by immunocytochemistry, and photographed using standard light microscopy.

Transwell assay

Transwell assays were performed with migration chambers (BD Biosciences, USA). Briefly, cells (1×10⁴) were seeded and medium with 10% FBS was added to the lower chamber as a chemoattractant. After 24 h, cells were fixed in methanol, stained with 0.1% crystal violet and counted.

Mouse xenograft model

Cells (2×10⁶) were subcutaneously injected into the dorsal flanks of 4-week-old female BALB/c nude mice (three mice per group to provide a power of 90% for a significance level of 0.05 with a two-tailed t test). Then the mice were intratumorally injected with 40 μL si-NC or si-circKIF4A every 4 days. Xenografts were excised under anesthesia after 4 weeks, and the tumor weights were measured.

For lung metastasis studies, cells (1 × 10⁵) were injected into the mice tail veins (three mice per group to provide a power of 90% for a significance level of 0.05 with a two-tailed t test). Then the mice were injected with 40 μL si-NC or si-circKIF4A every 4 days. After 8 weeks, the lungs were excised under anesthesia, and the numbers of macroscopically visible lung metastatic nodules were counted and validated by assessment of hematoxylin and eosin (HE)-stained sections by microscopy.

Immunohistochemistry (IHC) analysis

After deparaffinizing and rehydrating, the slides were treated with 90% methanol/3% H₂O₂ solution for 10 minutes at room temperature to block endogenous peroxidase. Then, the slides were soaked in sodium citrate buffer (10 mM Sodium citrate, 0.05% Tween 20, pH 6.0) under 96°C for 5 min for antigen retrieval. After blocking by BSA, antibody against Ki-67 (dilution 1:400, CST, USA) was used and incubated overnight at 4°C then incubated at room temperature with biotinylated secondary antibody for 10 minutes, and finally HRP-Streptavidin for 10 minutes. After DAB staining, the intensity of Ki-67 staining was imaged.

Luciferase reporter assay

The circKIF4A sequences including the miR-127 binding sites was inserted into the pGL3 luciferase vector (Promega, USA) immediately downstream of luciferase. Mutations in the miR-127 seed-region were conducted with Fast Site-Directed Mutagenesis Kit (TIANGEN, China). The JAM3 3’-UTR including the miR-127 binding sites was inserted into the pGL3 luciferase vector. Mutations in the miR-127 seed-region served as a mutant control.

Cells (5×10³) were seeded and cotransfected with corresponding vectors and miR-127 mimic or miR-127 inhibitor. After 48 h of incubation, luciferase intensity was measured by dual-luciferase reporter assay system (Promega).

RNA immunoprecipitation (RIP) assay

Cells were co-transfected with MS2bs-circKIF4A, MS2bs-circKIF4Amt or MS2bs-Rluc and MS2bp-GFP. After 48 h, RIP was performed with Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, USA). The RNA complexes were then purified, and the level of miR-127 was quantified.

For the RIP assay for Ago2, RIP was performed with an anti-Ago2 antibody (Millipore). RNAs were then purified, and the levels of circKIF4A, JAM3 and miR-127 were measured.

Statistical analysis

Statistical analysis was conducted with SPSS 19.0 software. Comparisons between groups were performed using t tests. Unless otherwise indicated, data are presented as the mean ± S.D. of three independent experiments. P < 0.05 was considered statistically significant.