CircRNA hsa_circ_0002577 accelerates endometrial cancer progression through activating IGF1R/PI3K/Akt pathway

EC tissues and matched adjacent normal tissues were obtained from 84 patients undergoing hysterectomy for EC at the Shengjing Hospital Affiliated to China Medical University. No patient received radiotherapy or chemotherapy before operation. The histological diagnosis and FIGO staging of EC were performed by two experienced pathologists. The fluorescence in situ hybridization (FISH) was performed to compare the level of hsa_circ_0002577 in EC and normal tissues as previously described [17]. Briefly, paraffin-embedded tissue samples were cut into 5 μm sections. The hybridization mix was prepared with 2 μL of hsa_circ_0002577 probe (0.5 ng/μL) in the 18 μL hybridization buffer containing 25% (v/v) formamide, 2× SSC, 5× Denhardt’s solution, 1 mM ethylenediaminetetraacetic acid, and 50 nM phosphate buffer at 37 °C overnight. After three washes with 2× SSC and one wash with RNase-free water, slides were stained with DAPI. After treating with ProLong® Gold Antifade (Invitrogen, Carlsbad, USA) overnight in the dark, images were captured at 40× and 100× magnification using a fluorescence microscope. All experiment protocols were approved by the Medical Ethics Committee of the Shengjing Hospital Affiliated to China Medical University and performed following the World Medical Association Declaration of Helsinki [18]. Each participant provided written informed consent.

EC cell lines (HEC-1-B, AN3-CA, KLE, HEC1-A, and Ishikawa) and human endometrial endothelial cells (hEEC) were obtained from ATCC (Shanghai, China) and maintained in DMEM medium supplemented with 10% fetal bovine serum. HEK-293 T cells (ATCC) were cultured in DMEM medium containing 4.5 mg/mL glucose, 2 nM glutamine, 10% fetal calf serum, and 1% streptomycin (Sigma-Aldrich, St. Louis, USA). Human umbilical vein endothelial cells (HUVECs, ATCC) were cultured in Endothelial Cell Growth Media (ECGM, #211–500, Sigma-Aldrich). All cells were maintained in a humidified atmosphere of 5% CO₂ at 37 °C.

Recombinant lentiviral vectors expressing hsa_circ_0002577 (Lv-circRNA), short hairpin RNAs (shRNAs) against hsa_circ_0002577 (sh-circRNA), miR-625-5p mimics, miR-625-5p inhibitor, lentiviral vectors expressing insulin-like growth factor 1 receptor (IGF1R) and their corresponding negative controls (Lv-NC, sh-NC, NC inh, NC miR, Control, respectively) were designed and synthesized by GenePharm (Shanghai, China). When reached 60–70% confluency, EC cells were transfected with designated vectors and/or sequences for 48 h using the TransFast transfection reagent Lipofectamine 2000 (Invitrogen).

RNase R is a 3′ to 5’exoribonuclease that degrades linear RNAs without affecting circular RNAs [19]. RNase R treatment was performed as previously described to enrich circular RNAs [20]. In brief, RNAs were isolated from Ishikawa and HEC-1-B cells using RNAeasy system and on-column DNAse digestion (Qiagen, Hilden, Germany). In the first replicate, 60 μg total RNAs were depleted for ribosomal RNAs using RiboMinus kit (Invitrogen) in 6 separate 10-μg reactions. In the second biological replicate, 20 μg total RNAs were depleted for ribosomal RNAs. A total of six 14.3-μL RNase R reactions or mock treatment aliquots were prepared for each replicate. Samples were denatured at 70 °C and then chilled to 40 °C using a thermocycler. Next, 1.7 μL 10× RNase R buffer was added to each sample. A volume of 1 μL water was added in one of the six reactions, and 1 μL RNase R was added to the remaining reactions. The reactions proceeded for 1 h at 40 °C.

Cell proliferation was evaluated using CCK-8 assay and colony formation assay. In CCK-8 assay, cells (2.5 × 10⁴ cells/well) were plated into 96-well plates and transfected with designated vectors and/or sequences. At indicated time points (0, 24, 48, 72, and 96 h) post-transfection, cells were incubated with 10 μL CCK-8 solution (Dojindo, Tokyo, Japan) for 3 h at 37 °C. The absorbance value was measured at 450 nm wavelength using a micro-plate reader (Bio-Rad, Hercules, USA). In colony formation assay, cells (1.5 × 10³ cells/well) were seeded into P100 petri dishes and transfected with designated vectors and/or sequences for 10 days. Then cells were fixed and stained with 1% crystal violet (Sigma-Aldrich). The quantity of colonies in each petri was counted from 6 randomly selected and normalized to the number of colonies in the control group.

Following transfection, cells were fixed and treated with DAPI (Thermo Fisher Scientific, Waltham, USA) and TUNEL (Roche, Basel, Switzerland) reagents. DAPI stained cell nucleus. The apoptotic cells were detected by dual DAPI and TUNEL staining. The percentage of TUNEL-positive cells on each slide was calculated in 6 randomly selected fields by using a fluorescence microscope at × 100 magnification.

To assess the migration/invasion capability of cells, wound healing assay and Transwell assay were performed. In would healing assay, cells were plated in 6-well plates and transfected with designated vectors and/or sequences for 48 h. Then an artificial straight scratch was made using a sterile pipette tip. At 0 h and 24 h after scratching, the relative migration rate (%) was determined by measuring the percentage of wound recovery area under a light microscope (magnification, × 100). In Transwell assay, transfected cells were transferred to the upper chamber of 24-well Transwell plates with 8 μm-pore membrane (BD Biosciences, Bedford, USA) and cultured in serum-free medium. The bottom chamber was added with DMEM containing 10% fetal bovine serum. Twenty-four hours later, the invading cells at the bottom chamber were fixed and stained with crystal violet dye. The number of migrated cells was counted in 6 randomly selected fields under a light microscope (magnification, × 100).

The cell supernatants were collected from EC cells transfected with designated vectors and/or sequences. HUVECs were cultured with a mixture of EC cell supernatants and ECGM for 4 days. Then HUVECs were trypsinized and reseeded (3 × 10⁴ cells/well) onto Matrigel-coated wells. Twenty-four hours later, the branching points of HUVECs, defined as the three overhanging branches of the intersection point, were observed under a Leica microscope and the branches were quantified using Image-Pro software.

EC cells were grown on glass multi-chamber slides (BD Biosciences, San Jose, USA), fixed with 4% paraformaldehyde, and permeabilized with 0.4% Triton X-100 (Sigma-Aldrich). Then slides were treated with a hybridization mixture containing 18 μL hybridization buffer, 1 μL hsa_circ_0002577 probe (0.5 ng/μL) and 1 μL miR-625-5p probe (0.5 ng/μL) at 37 °C overnight. After treatment with RNase-free water and ProLong® Gold Antifade, slides were stained with DAPI and the images were captured using a fluorescence microscope at 200× magnification.

Experiment 1: The 3′-UTR fragment of hsa_circ_0002577 containing the putative binding site of miR-625-5p or a mutant sequence was cloned into pmirGlo vectors (GenePharm). HEC-1-B cells at 70–80% confluence were co-transfected with miR-34b-5p inhibitor (or NC inh) and hsa_circ_0002577-WT (or hsa_circ_0002577-MUT) for 48 h. Ishikawa cells were co-transfected with hsa_circ_0002577-WT (or hsa_circ_0002577-MUT) and miR-34b-5p mimics (or NC miR) when reached 70–80% confluence. Experiment 2: The 3′-UTR fragment of IGF1R harboring the putative binding site of miR-625-5p or a mutant site was cloned into pmirGlo vectors. The IGF1R-WT (or IGF1R-MUT) and miR-34b-5p inhibitor (or NC inh) were co-transfected in HEC-1-B cells. The IGF1R-WT (or IGF1R-MUT) and miR-34b-5p mimics (or NC miR) were co-transfected in Ishikawa cells. Experiment 3: The 3′-UTR fragments of the following genes (HOXB5, NFIX, IGF1R, MDM4, CCND1, SPSB1, SF3B3, SESN3, KIAA2013, and PRELP) were cloned into pmirGlo vectors. HEK-293 T cells were co-transfected with each type of pmirGlo vectors and miR-34b-5p mimics at 70–80% confluence. In all experiments, the luciferase activities were measured at 48-h post-transfection using the Luciferase Reporter Assay System (Promega Biotech Co., Madison, USA).

Total RNAs were isolated from tissue samples and whole-cell lysates using Trizol (Invitrogen) and Trizol LS (Invitrogen), respectively. In EC cells, the nuclear and cytoplasmic cell fractionations were obtained using the PARIS kit (Thermo Fisher Scientific) prior to the RNA isolation. The ReverTra Ace qPCR RT Kit (Toyobo, Osaka, Japan) was used to reverse transcribe RNAs to cDNA. The reverse transcription of miRNA was performed using All-in-One™ miRNA RT-qPCR Detection Kit (GeneCopoeia Inc., Rockville, USA). The cDNAs were analyzed using 7300 Real-Time PCR System (Applied Biosystem, Foster City, USA). The primers were as follows: HOXB5 forward: 5′-AGCGCCAATTTCACCGAA-3′, HOXB5 reverse: 5′-GGCTGCTTAGCTGGCTTGC-3′; SESN3 (Hs00376220_m1, Applied Biosystems, Foster City, USA), IGF1R forward: 5′-GGCACAATTACTGCTCCAAAGAC-3′, IGF1R reverse: 5′-CAAGGCCCTTTCTCCCCAC-3′; U6 forward: 5′-CTCGCTTCGGCAGCACATATACT-3′. U6 reverse: 5′-ACGCTTCACGAATTTGCGTGTC-3′; GAPDH forward: 5′-ATCACTGCCACCCAGAAGAC-3′, GAPDH reverse: 5′-TTTCTAGACGGCAGGTCAGG-3′.

The RIP assay was performed in HEC-1-B and Ishikawa cells using the Imprint® RNA Immunoprecipitation Kit (Sigma-Aldrich) as previously described [21]. An anti-IgG antibody (Sigma-Aldrich) was used as a negative control. A mouse monoclonal anti-Argonaute2 (anti-Ago2) antibody (Sigma-Aldrich) was used as a positive control. Briefly, cells were collected, lysed in RIP lysis buffer, and incubated with anti-IgG or anti-Ago2 overnight at 4 °C. A volume of 40 μL Protein A magnetic beads were added. Total RNAs were isolated using GenElute™ Total RNA Purification Kit (Sigma-Aldrich). The enrichment of hsa_circ_0002577-WT and miR-625-5p were quantified using quantitative real-time PCR (qRT-PCR).

Total proteins were extracted from homogenized tissues and cell lysates using RIPA buffer containing protease inhibitors and phosphatase inhibitors (Pierce, Rockford, USA). An equal amount of protein lysates was separated on 12% SDS-PAGE and then transferred to polyvinylidene fluoride membranes. After an overnight incubation with the following primary antibodies at 4 °C: HOXB5 (1:1000, #109375, Abcam, Cambridge, UK), SESN3 (1:800, #97792, Abcam), IGF1R (0.1 μg/mL, #AF-305-SP, R&D Systems, Inc., Minneapolis, USA), protein kinase B (Akt, 1:1000, #9272, Cell Signaling), phosphorylated Akt (p-Akt, 1:1000, #9271, Cell Signaling, Danvers, USA), PCNA (PCNA, 1:1000, #18197, Abcam), p27 (1:1000, #32034, Abcam), B-cell lymphoma 2 (Bcl-2, 1:800, #59348, Abcam), E-cadherin (1:2000, #40772, Abcam), N-cadherin (1:1000, #76057, Abcam), cleaved caspase-3 (1:1000, #49822, Abcam), and GAPDH (1:2000, #37168, Abcam), the membranes were washed with Tris-Buffered Saline and stained with secondary antibody (1:2000, #6721, Abcam) for 60 min at room temperature. The images were captured using Alphalmager™ 2000 Imaging System (Alpha Innotech, USA).

Female BALB/c nude mice (4-week-old, Charles River Laboratories, Guangdong, China) were housed in a pathogen-free facility with a temperature of 24 ± 1 °C, a humidity of 50%, 12-h light-dark cycle, and ad libitum access to water and food. Ishikawa cells were transfected with shRNA against hsa_circ_0002577 or the control sequence. After one week of acclimation, animals were randomly assigned into two groups (n = 6/group): sh-circRNA and sh-NC. The sh-circRNA group was subcutaneously injected with 1 × 10⁷ sh-circRNA-transfected cells (suspended in 250 μL culture medium) into the dorsal flank. The sh-NC group was inoculated with shNC-transfected cells following the same procedure. The tumor volume was calculated every week for 6 weeks using the formula: volume = (π × length × width²)/6. Six weeks after inoculation, the metastasis of tumors in mice implanted with Ishikawa cells was evaluated using bioluminescence imaging as previously described [22]. Briefly, animals were anesthetized by isoflurane and then administered with D-luciferin potassium salt (Biosynth, Staad, Switzerland) at the dosage of 150 mg/kg via intraperitoneal injection. Fifteen minutes later, mouse was placed in a supine position. Bioluminescence imaging was performed using an in vivo imaging system (IVIS, PerkinElmer, Waltham, USA). The intensity of signal was reported as radiance (photons/s/cm²/steradian). Then all animals were euthanized and tumor xenografts were harvested and weighted. The tumor samples were sectioned and stained for IGF1R (#110025, Abcam), Ki-67 (#15580, Abcam), E-cadherin, and CD31 (#28364, Abcam) using immunohistochemistry method (magnification 40×). The sectioned slides were stained with TUNEL reagent for apoptosis. Hematoxylin and eosin (H&E) staining was also performed to detect metastatic pulmonary nodules and the density of microvessels. All procedures were approved by the Medical Ethics Committee of the Shengjing Hospital Affiliated to China Medical University, and performed according to the Guide for the Care and Use of Laboratory Animals [23].

Data are shown as mean ± standard deviation and analyzed using software SPSS (version 24.0). All cell culture experiments in this study were performed in triplicate and repeated three times. One-way analysis of variance (ANOVA) was used to evaluate statistical significance. Survival data was analyzed using the Kaplan-Meier method. The linear correlation coefficient was used to estimate the correlation in the expression levels of hsa_circ_0002577 vs. miR-625-5p, hsa_circ_0002577 vs. IGF1R, and miR-625-5p vs. IGF1R. A value of p < 0.05 was considered statistically significant. * p < 0.05, ** p < 0.01.

To identify differentially expressed circRNAs in EC, an expression heat map was generated using three pairs of EC and adjacent normal tissues. The results showed that hsa_circ_0002577, hsa_circ_0005797, hsa_circ_0057780, and hsa_circ_0016595 were the top upregulated circRNAs between EC and noncancerous tissues (Fig. 1 a). Then we measured the expressions of these circRNAs in 84 paired EC and normal tissue samples, and found that hsa_circ_0002577 was the most upregulated one in EC tissues (Fig. 1 b, Supplementary Figure 1A). Also, compared to the samples collected from patients at stage I/II, the expression of hsa_circ_0002577 was significantly higher in advanced EC tissues (III/IV) (Fig. 1 b). Therefore, we focused on hsa_circ_0002577, which was generated from WDR26 gene. (Fig. 1 c). Further analysis showed that the low expression of hsa_circ_0002577 was correlated with significantly better overall survival compared with high expression (Fig. 1 d). By analyzing the clinicopathological significance of hsa_circ_0002577 in patients, we also found that the expression of hsa_circ_0002577 in EC tissues was positively and significantly associated with high histological grade of tumor, lymph node metastasis (LNM) staging, and lymph vascular space invasion (LVS) (Table 1). The FISH analysis showed that hsa_circ_0002577 was prominently overexpressed in EC tissues in comparison to paired normal samples (Fig. 1 e). Next, we measured the expression of the target circRNA in EC cell lines. Compared to hEEC, all tested EC cells exhibited a significantly higher level of hsa_circ_0002577, among which HEC-1-B cells had the lowest expression and Ishikawa cells showed the highest hsa_circ_0002577 level (Fig. 1 f). Then RNase R treatment was performed to enrich circRNAs in Ishikawa and HEC-1-B cells followed by the examination of RNA (linear and circular) expressions. The level of linear RNAs in RNase R-treated group was markedly lower than that in the mock group, whereas the expression of circRNAs was not affected by RNase R treatment (Fig. 1 g). Via the detection of hsa_circ_0002577 level in the nuclear and cytoplasmic fractionations of EC cells, we found that hsa_circ_0002577 was abundantly expressed in cytoplasm (Fig. 1 h). The above findings indicated that the upregulation of hsa_circ_0002577 may exert an oncogenic effect on the progression of EC.

To explore the potential regulatory role of hsa_circ_0002577 in EC cells, we transfected HEC-1-B cells with lentiviral vectors expressing hsa_circ_0002577 (or the control vector), and transfected Ishikawa cells with two shRNAs against hsa_circ_0002577 (or the control sequences). The delivery of Lv-circRNA significantly induced the overexpression of hsa_circ_0002577 in HEC-1-B cells, while sh-circRNA#1 and #2 effectively decreased the level of hsa_circ_0002577 in Ishikawa cells (Fig. 2 a). As sh-circRNA#2 showed more potent knockdown efficiency than sh-circRNA#1, sh-circRNA#2 was used in further analyses, named sh-circRNA. The CCK-8 proliferation assay showed that the upregulation of hsa_circ_0002577 strongly induced the proliferation capacity of HEC-1-B cells at 96 h post-transfection, whereas the knockdown of target circRNA efficiently suppressed the proliferation of Ishikawa cells (Fig. 2 b). Consistently, the number of colonies was significantly increased in Lv-circRNA-transfected cells compared to the controls, but reduced in cells with hsa_circ_0002577 deficiency (Fig. 2 c). Compared to control vector-transfected cells, the number of TUNEL-positive cells was significantly decreased in HEC-1-B cells transfected with Lv-circRNA, indicating that hsa_circ_0002577 overexpression rescued EC cells from apoptosis. Meanwhile, extensive apoptosis was observed in EC cells with insufficient hsa_circ_0002577 expression (Fig. 2 d). These findings suggested that EC cell proliferation and apoptosis were mediated by the level of hsa_circ_0002577.

Furthermore, cells overexpressing hsa_circ_0002577 significantly accelerated wound closure as compared to Lv-NC-transfected group in wound healing assay, whereas the knockdown of hsa_circ_0002577 repressed cell migration, as shown by significantly decreased migration rate compared with the controls (Fig. 3 a). Also, a higher number of migrated and invaded cells were found in Lv-circRNA group as compared to the control cells in the Transwell assay. The hsa_circ_0002577 deficiency, however, impeded the migration/invasion of EC cells (Fig. 3 b). Angiogenesis is a key event in tumor development, in which new capillary networks are formed by vascular endothelial cells, facilitating the delivery of oxygen and nutrients to the tumor [24]. To explore whether hsa_circ_0002577 might affect the formation of tubule-like endothelial structures, we cultured HUVECs with the cell supernatants collected from transfected EC cells. The supernatants of cells overexpressing hsa_circ_0002577 significantly induced HUVECs migration (Fig. 3 c) and increased the number of branches in HUVECs (Fig. 3 d) as compared to the control cells. On the contrary, the supernatants of cells with hsa_circ_0002577 deficiency suppressed the migration and tubule formation of HUVECs. These data implied that hsa_circ_0002577 not only induced the migration/invasion of EC cells, but also promoted the tubule formation of HUVECs.

To investigate the underlying mechanisms involved in the regulation of EC tumorigenesis by hsa_circ_0002577, we searched for the potential bindings sites of hsa_circ_0002577 for miRNAs. The bioinformatic prediction analysis from two databases showed that hsa_circ_0002577 had a putative binding site for miR-625-5p (Fig. 4 a). The fluorescence staining of EC cells with DAPI, miR-625-5p probe, and hsa_circ_0002577 probe demonstrated that both miR-625-5p and hsa_circ_0002577 were predominantly expressed in cytoplasm (Fig. 4 b). We speculated that hsa_circ_0002577 might competitively binds to miR-625-5p as a miRNA sponge to regulate its expression and function in cells. The putative binding site of hsa_circ_0002577 for miR-625-5p and a mutant sequence were shown (Fig. 4 c). The transfection with miR-625-5p inhibitor significantly downregulated the expression of miR-625-5p in HEC-1-B cells, while Ishikawa cells introduced with miR-625-5p mimics demonstrated a remarkably elevated level of miR-625-5p compared to the group transfected with control mimics (Fig. 4 d). The luciferase activity of hsa_circ_0002577-WT was markedly upregulated in EC cells delivered with miR-625-5p inhibitor but decreased in miR-34b-5p mimics-transfected group, as compared to their corresponding controls. Cells transfected with the mutant sequence exhibited no difference in the luciferase activity between the mimics/inhibitor-transfected cells and the control groups (Fig. 4 e). The RIP assay using the Ago2-sepecific antibody showed that both miR-625-5p and hsa_circ_0002577 were enriched in Ago2 immunoprecipitation, confirming their interaction in EC cells (Fig. 4 f). Moreover, the expression of miR-625-5p in HEC-1-B cells transfected with Lv-circRNA was significantly higher than that in control cells. Hsa_circ_0002577 knockdown, however, promoted the expression of miR-625-5p in EC cells (Fig. 4 g). A significant variation in the level of miR-625-5p was also observed in patients’ samples. Tumor tissues showed significantly lower expression of miR-625-5p in comparison to paired normal samples (Fig. 4 h). The hsa_circ_0002577 level was inversely and significantly correlated with miR-625-5p expression (Fig. 4 i). Additionally, EC patients with low expression of miR-625-5p showed significantly better overall survival compared with those with high miR-625-5p expression (Supplementary Figure 1B). The above results indicated that hsa_circ_0002577 reversely regulated the expression of miR-625-5p in EC cells and tissues.

To further explore the involvement of miR-625-5p in hsa_circ_0002577-mediated EC pathogenesis, we predicted the downstream targets of miR-625-5p in four bioinformatic databases. Results showed that ten genes might be the potential targets of miR-625-5p (Fig. 5 a). We transfected HEK-293 T cells with luciferase vectors containing the 3′-UTR fragments of these genes and miR-34b-5p mimics. Cells transfected with vectors containing HOXB5, IGF1R, and SESN3 genes showed luciferase activity lower than 40% (Fig. 5 b). In HEC-1-B cells, the transfection with miR-625-5p inhibitor significantly increased the transcription of these genes compared to the controls, with the most robust induction in IGF1R. In Ishikawa cells, the overexpression of miR-625-5p significantly suppressed the mRNA expressions of HOXB5 and IGF1R, but not SESN3, with the greatest inhibition in IGF1R (Fig. 5 c). Consistently, Western blot analysis of both miR-625-5p-overexpressed and -deficient cells showed that the biggest alteration in the protein expression of these genes was seen in IGF1R (Fig. 5 d). The putative binding site of IGF1R for miR-625-5p and a mutant sequence were shown (Fig. 5 e). The luciferase intensity of IGF1R-WT was increased in cells transfected with miR-625-5p inhibitor but decreased in the ones delivered with miR-34b-5p mimics, as compared to their corresponding controls (Fig. 5 f). The mRNA expression of IGF1R in tumor tissues was significantly higher than that in normal tissues (Fig. 5 g). Also, the mRNA expression of IGF1R in was positively correlated with hsa_circ_0002577 level and inversely correlated with miR-185-5p expression with statistical significance (Fig. 5 h). Additionally, hsa_circ_0002577 overexpression in HEC-1-B cells significantly increased the protein level of IGF1R, whereas the co-transfection with both Lv-circRNA and miR-625-5p mimics eliminated the effect of hsa_circ_0002577 on IGF1R. In Ishikawa cells, the co-transfection with sh-circRNA and miR-625-5p mimics reversed the inhibitory effect of hsa_circ_0002577 deficiency on IGF1R expression (Fig. 5 i). These data suggested that hsa_circ_0002577 induced the expression of IGF1R via inhibiting miR-625-5p.

To ascertain that hsa_circ_0002577 regulated EC progression by targeting IGF1R, we co-transfected Ishikawa cells with sh-circRNA (or sh-NC) and lentiviral vectors expressing IGF1R (or control vectors). As shown above, hsa_circ_0002577 knockdown significantly suppressed cell proliferation and induced apoptosis in Ishikawa cells. By introducing cells with IGF1R-overexpressing vectors, the proliferation capacity (Fig. 6 a) and the number of colonies (Fig. 6 b) was significantly increased. Also, the upregulation of IGF1R in hsa_circ_0002577 deficient cells significantly reduced the quantity of apoptotic cells (Fig. 6 c). In the analyses of cell migration/invasion, the group delivered with both IGF1R-overexpression vector and sh-circRNA showed significantly lower migration rate (Fig. 6 d) and decreased number of migrated and invaded cells (Fig. 6 e) as compared to the cells with insufficient hsa_circ_0002577 expression. To examine the effect of IGF1R on the formation of tubule-like endothelial structure, HUVECs were cultured with ECGM containing the supernatants collected from Ishikawa cells co-transfected with sh-circRNA (or sh-NC) and lentiviral vectors expressing IGF1R. The overexpression of IGF1R significantly promoted tube formation in cells with insufficient hsa_circ_0002577 expression (Fig. 6 f). Furthermore, the expressions of proteins related to the PI3K/Akt signaling pathway were analyzed. The expressions of IGF1R, p-Akt/Akt, PCNA, Bcl-2, and N-cadherin were significantly reduced by the knockdown of hsa_circ_0002577, but restored to the normal level by excessive production of IGF1R. The deficiency of hsa_circ_0002577 elevated the levels of p27, cleaved caspase-3, and E-cadherin in Ishikawa cells, which were then inhibited by IGF1R overexpression (Fig. 6 g). The above findings indicated that hsa_circ_0002577 induced EC cell proliferation and migration by regulating IGF1R and PI3K/Akt signaling pathway.

Finally, we evaluated the regulatory effects of hsa_circ_0002577 in vivo using a mouse EC xenograft model. Ishikawa cells were transfected with sh-circRNA or control shRNA (sh-NC) and then inoculated into nude mice subcutaneously. Compared to the control group, mice injected with hsa_circ_0002577-deficient cells displayed significantly decreased tumor volume and weight (Fig. 7 a). The expressions of IGF1R and Ki-67 (a cell proliferation marker) in tumor samples collected from sh-circRNA group were reduced compared to the control mice. The expression of E-cadherin (a differentiation and invasiveness marker), however, was increased by the knockdown of hsa_circ_0002577 (Fig. 7 b). The sh-circRNA group also showed more apoptotic cells (Fig. 7 c). The bioluminescence imaging showed that hsa_circ_0002577 deficiency ameliorated the growth and metastasis of tumor in mice implanted with Ishikawa cells (Fig. 7 d). The insufficient expression of hsa_circ_0002577 also decreased the number of metastatic pulmonary nodules (Fig. 7 e) and the density of microvessels (Fig. 7 f) in mice. In addition, compared to the control mice, the sh-circRNA group showed less stained CD31-positive cells in tumor tissue sections (Fig. 7 g). Taken together, these results demonstrated that hsa_circ_0002577 knockdown delayed the tumor growth and metastasis in the EC mouse model.