Between April 2018 and April 2020, 60 patients with EC (all female; 26 cases at stage I or II and 34 cases at stage III or IV) were admitted to Beilun District People's Hospital (Zhejiang, China). Patients with recurrent EC were not enrolled into the present study. Patients receiving current treatment therapies were also excluded from the study, as were patients with other severe clinical complications, such as chronic renal or hepatic disease, chronic obstructive pulmonary disease, asthma and congenital diseases. The age range of the patients was 52–68 years, with a mean age of 60.7 ± 5.5 years. The present study was approved by the Ethics Committee of Beilun District People's Hospital (Zhejiang, China) and written informed consent was obtained from all patients. Before treatment, specimens of both EC and paired adjacent normal (within 5 cm around tumors) tissues were collected from each patient using fine-needle aspiration. Histopathological analysis was performed on all tissue specimens to confirm that the correct specimens were collected, according to the guidelines of the German Working Group on Gynecological Oncology (16). The tissue specimens were stored at −80°C or in liquid nitrogen before the subsequent experiments. Association between miR-31 expression level and the clinicopathological parameters (age, sex, histopathological grade, depth of myometrial invasion, lymphatic metastasis and distant metastasis) in patients with EC are shown in Table I.

A total of two human EC cell lines, HEC-1-A and RL95-2 (both from American Type Culture Collection) were used as the EC cell models. Cell culture was cultured in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.), supplemented with 10% FBS (Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C in a humidified incubator with 5% CO₂ according to the manufacturer's instructions. Subsequent assays were performed using cells at ~85% confluence. The cells were transfected with miR-31 mimics and/or circ_POLA2 overexpression plasmid using Lipofectamine^® 3000 (Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. The cells were cultured as aforementioned for 48 h after transfection before subsequent experiments.

A backbone vector expressing circ_POLA2 was constructed using the pcDNA3.1(+) circRNA mini vector (Addgene, Inc.). The following primer sequences were used to construct circ_POLA2: forward, 5′-GGAATTCATGTCCGCATCCGCC-3′ and reverse 5′-ATAAGAATTCAGATCCTGACGACC-3′. miR-31 mimics were synthesized by Shanghai GenePharma Co., Ltd., and negative control (NC; cat. no. miR1N0000001-1-5) miRNAs were purchased from Guangzhou RiboBio Co., Ltd. The sequence of the miR-31 mimics and NC miRNA are as follows: 5′-UAGCAGCACAGAAAUAUUGGC-3′ and 5′-UUGUACUACACAAAAGUACUG-3′, respectively. To overexpress circ_POLA2 and miR-31, the HEC-1-A and RL95-2 cells (1×10⁸) were transfected with circ_POLA2 expression vector (1 µg) or miR-31 mimics (40 nM) using Lipofectamine^® 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The same number of cells were transfected with empty vector or NC miR mimics using the same method as the overexpression vector and miR-31 mimics. Empty vector or untreated cells were used as the control (C) group. Subsequent assays were performed 48 h later.

DNA was extracted from the HEC-1-A and RL95-2 cells transfected with circ_POLA2 overexpression plasmid using a Monarch^® Genomic DNA Purification kit (New England Biolabs, Inc.). An EZ DNA Methylation kit (cat. no. D5001; Zymo Research Corp.) was used to convert the genomic DNA into bisulfite modified DNA. Firstly, MSP was performed to analyze miR-31 methylation, in which the DNA was modified by sodium bisulfite treatment to convert unmethylated cytosine to uracil, while methylated cytosine remained intact. Secondly, following the removal of bisulfite and completion of the chemical conversion, the modified DNA was used as the template for PCR using Taq DNA polymerase (Takara Bio, Inc.) The following primer sequences were used: Methylation forward, 5′-TTGTGTATAATTTGGGGCGTC-3′ and reverse, 5′-CCAACTTACCTACGAATCCGA-3′, and unmethylation forward, 5′-TTGTGTATAATTTGGGGTGTTGT-3′ and reverse, 5′-CTCCCAACTTACCTACAAATCCA-3′. The following thermocycling conditions were used: Initial denaturation at 95°C for 30 sec, then 55°C for 30 sec, 72°C 30 sec. A total of 2 PCRs were performed for each DNA sample, one specific for originally methylated DNA without methylation for the gene of interest and one specific for originally unmethylated DNA with methylation. The PCR products were separated using a 6–8% non-denaturing polyacrylamide gels and the bands were visualized by staining with ethidium bromide. The presence of a band of the appropriate molecular weight indicated the presence of unmethylated and/or methylated alleles in the original sample.

Total RNA was isolated from the HEC-1-A and RL95-2 cells transfected with circ_POLA2 overexpression plasmid or miR-31 mimics, as well as paired tissue specimens from 60 patients with EC using RNAzol (Sigma-Aldrich; Merck KGaA). DNase I (Invitrogen; Thermo Fisher Scientific, Inc.) was used to remove the genomic DNA. RNA integrity was analyzed using agarose gel electrophoresis.

The cDNA samples were prepared using a Bio-Rad cDNA Supermix kit (Bio-Rad Laboratories, Inc.). The following temperature protocol was used for RT: Incubation at 50°C for 15 min then 75°C for 5 min. The following housekeeping genes were used as the internal controls: GAPDH and U6. The expression level of mature miR-31 was determined using the All-in-One™ miRNA RT-qPCR Detection kit using SYBRGreen (GeneCopoeia, Inc.) following the manufacturer's instructions. For circ_POLA2, qPCR was performed using a LightCycler^® 480 SYBR Green I Master (Roche Diagnostics). The following thermocycling conditions were used for both miRNA and cir_POLA2: Initial denaturation at 95°C for 30 sec, then 95°C for 5 sec, 60°C 34 sec for 40 cycles; then 95°C for 15 sec, 60°C for l min and 95°C for 15 sec. The threshold cycle (Cq) values were analyzed using the 2^−ΔΔCq method. Analysis of relative gene expression was performed using the 2^−ΔΔCq method (17).

The following primer sequences were used: Circ_POLA2 forward, 5′-ATGTCCGCATCCGCC-3′ and reverse 5′-TCAGATCCTGACGACC-3′; GAPDH forward, 5′-GCACCGTCAAGCTGAGAAC-3′ and reverse 5′-GGTGAAGACGCCAGTGGA-3′; miR-31 forward, 5′-AGGCAAGAUGCUGGCAUAGCU-3′ and reverse 5′-AAAGGCAAGAUGCUGGCAUAG-3′; and U6 forward, 5′-CTCGCTTCGGCAGCACA-3′ and reverse 5′-ACGCTTCACGAATTTGCGTGTC-3′.

CCK-8 assays were performed using a kit from Abcam (cat. no. ab228554). In a 96-well cell culture plate, 4,000 cells in 0.1 ml fresh RPMI-1640 cell culture medium, supplemented with 10% FBS were added. A total of 3 wells were used for each experiment. The cells were cultured at 37°C. Cell proliferation was assessed by measuring the optical density at 450 nm at different time points (0.5, 1 and 2 h). The CCK-8 (Abcam) solution was added to each well to a final concentration of 10%, 2 h before the samples were measured.

The data were analyzed using GraphPad Prism v6 (GraphPad Software, Inc.). A total of 3 replicates were included in each experiment and the data in all figures are presented as the mean ± SD, while the data in Table I is presented as the median + IQR. Unpaired t-tests were used to compare 2 groups, while comparisons among multiple groups were performed using ANOVA followed by Tukey's post hoc test. Correlations were analyzed by linear regression. The mRNA expression levels of circ_POLA2 and miR-31 in EC and paired non-tumor tissues were compared using paired t-tests. The association between the relative mRNA expression level of miR-31 and the clinicopathological data was analyzed using either a Mann Whitney U test or Kruskall-Wallis test for 2 or >3 groups, respectively. P<0.05 was considered to indicate a statistically significant difference.

EC and paired adjacent normal tissue specimens were collected from 60 patients with EC. RT-qPCR was then performed to determine the mRNA expression levels of circ_POLA2 and miR-31. Compared with that in the adjacent normal tissues, circ_POLA2 was significantly upregulated in EC tissues (Fig. 1A). In addition, miR-31 was significantly downregulated in EC tissues compared with that in adjacent normal tissues (Fig. 1B). Therefore, circ_POLA2 and miR-31 may serve roles in EC. Furthermore, the association between the mRNA expression levels of miR-31 and certain clinicopathological parameters were also analyzed. The median tumor/normal tissue miR-31 expression ratio in EC was 0.034 in patients with a depth of myometrial invasion (18) of ≤1/2 and 0.004 in patients with tumor invasion >1/2. The results also showed that the miR-31 expression level was associated with the clinical stage. The median tumor/normal tissue miR-31 expression ratio was 0.123 in 16 patients with stage I cancer and 0.004 in 34 patients with stage III/IV cancer. These results indicated an association between lower miR-31 mRNA expression levels and increasing stage and depth of tumor invasion. In addition, the median tumor/normal tissue miR-31 expression ratio was 0.120 in 30 patients with G1 and G2 histopathological grade and 0.116 in 20 patients with G3 histopathological grade. The median tumor/normal tissue miR-31 expression ratio was 0.178 in 2 patients with distant metastasis and 0.025 in 58 patients without distant metastasis. These results indicated an association between lower miR-31 mRNA expression levels and increasing histopathological grade and no distant metastasis. miR-31 expression was not associated with other clinicopathological factors, such as patient age and lymphatic metastasis (Table I).

Linear regression was performed to determine the correlation between circ_POLA2 and miR-31 expression levels in both EC (Fig. 2A) and adjacent normal (Fig. 2B) tissues. In EC tissue, circ_POLA2 mRNA expression level was inversely correlated with miR-31 mRNA expression level. However, their expression levels were not correlated in adjacent normal tissues. Therefore, circ_POLA2 and miR-31 may interact in EC.

circ_POLA2 and miR-31 were overexpressed in the HEC-1-A and RL95-2 cell lines (Fig. 3A). Overexpression of circ-POLA2 significantly decreased miR-31 mRNA expression level, compared with that in the empty vector group (Fig. 3B). However, miR-31 overexpression did not affect circ_POLA2 mRNA expression level, compared with that in the NC miRNA group (Fig. 3C). Therefore, circ_POLA2 may downregulate miR-31 expression level in the EC cell lines analyzed.

To investigate the mechanism underlying the interaction between circ_POLA2 and miR-31, MSP was performed. The results showed increased methylation of the miR-31 gene in cells transfected with the circ_POLA2 expression vector (Fig. 4). Therefore, circ_POLA2 may downregulate miR-31 via methylation.

The effects of circ_POLA2 and miR-31 overexpression on the proliferation of the EC cell lines were evaluated using CCK-8 assays. circ_POLA2 overexpression increased EC cell proliferation, whereas miR-31 overexpression decreased cell proliferation compared with that in the empty vector group and NC miRNA, respectively. In addition, circ_POLA2 overexpression suppressed the effect of miR-31, compared with that in the miR-31 overexpression group (Fig. 5).